SBC Board: Types, Examples and How to Choose

September 23rd, 2026

An SBC board is a single-board computer that combines a processor, memory, power circuitry and input/output interfaces on one printed circuit board. It can run an operating system and handle tasks such as an industrial display, network gateway or data logger without a separate desktop motherboard. Raspberry Pi 5, BeagleBone Black and UP 7000 are examples. Choose one by checking software compatibility, interfaces, storage, cooling and product life, not CPU speed alone. For a custom design, PCB layout and assembly testing also determine whether the board will work reliably in the finished product.

At EBest Circuit, we support customer-designed computing boards through PCB fabrication, component sourcing and PCB assembly services. Our manufacturing review focuses on the released design, component packages, assembly access and test requirements. The commercial boards below illustrate different architectures; they are not an EBest Circuit retail product list.

Illustration of an SBC board showing processor, memory, power circuitry and I/O connectors

Key Takeaways

An SBC integrates the main computer functions on one PCB; it is a populated computing board, not a bare PCB.
SoC, SoM and SBC describe different integration levels: a chip, a compute module and a computer board.
Raspberry Pi 5 and BeagleBone Black use Arm processors; UP 7000 provides an x86 example with Intel Processor N-series options.
Software support, usable interfaces, storage endurance and cooling matter as much as processor performance.
An industrial label does not establish a universal temperature rating, power-input range or service life. Check the exact board configuration.
Custom SBC layouts need coordinated memory routing, power sequencing, BGA fanout and thermal design.
A successful boot is not a complete production test. Assembly inspection, memory checks, interface tests and project-specific load testing cover different faults.

What Is an SBC Board?

An SBC is a computer built around a single main circuit board, with the processing and interfaces needed to perform its intended computing tasks. SBC stands for single-board computer, so the phrase “SBC board” is a common but technically redundant name.

A typical board contains these functional blocks:

A processor or system-on-chip for application execution.
RAM for working data and a storage interface or onboard boot storage.
Power regulators, clocks and reset circuitry.
Application interfaces such as Ethernet, USB, display outputs or GPIO.

Not every SBC has soldered storage, wireless networking or a display connector. A headless gateway may boot from eMMC and communicate only over Ethernet. A desktop-oriented board may need an external microSD card, power supply and display before it becomes usable.

SBC vs SoC

An SBC is an assembled computer board; a SoC is an integrated circuit that can serve as its main processor. A system-on-module, or SoM, sits between these levels and normally connects to a carrier board.

Term	Physical form	Typical integration
SoC	Semiconductor package	CPU cores, controllers and other integrated functions; external memory and power circuits depend on the device
SoM	Compact populated module	Processor and supporting circuits, often RAM and storage; carrier provides application connectors and interfaces
SBC	Main computer PCB assembly	Computing functions with usable board-level interfaces and power circuitry

For example, BCM2712 identifies the processor used on Raspberry Pi 5, while Raspberry Pi 5 identifies the board. When sourcing manufacturing, clarify whether the deliverable is a bare PCB, a carrier assembly or a complete computing-board assembly. These require different files, parts and tests.

Illustrated comparison of a SoC chip, a SoM compute module and an SBC computer board

SBC Types

SBC types can be classified by processor architecture or intended use. Those categories overlap: an industrial SBC can use either Arm or x86.

Arm SBCs: widely used for embedded Linux, gateways and compact application-specific computers.
x86 SBCs: useful when the application depends on an x86 software stack, PC-oriented drivers or a supported Windows configuration.
RISC-V SBCs: options for development and deployment where the required operating system, drivers and application packages support that platform.
Industrial SBCs: selected for specified environmental conditions, mechanical integration, lifecycle and support arrangements rather than one particular CPU architecture.

A development board is not automatically unsuitable for a commercial product, but its connector retention, software maintenance and operating conditions still need validation. Likewise, a higher-priced industrial model is not automatically the best match for a low-power sensor gateway.

Single Board Computer Examples

Raspberry Pi 5, BeagleBone Black and UP 7000 show three different approaches to board-level computing: general-purpose Arm performance, embedded I/O integration and compact x86 computing.

Board	Processor example	Selected hardware features
Raspberry Pi 5	Broadcom BCM2712, quad-core Arm Cortex-A76 at 2.4 GHz	LPDDR4X RAM, Gigabit Ethernet, microSD and PCIe 2.0 x1 expansion through an adapter
BeagleBone Black	TI AM3358, Arm Cortex-A8 at 1 GHz	512 MB DDR3L, 4 GB eMMC, two 46-pin expansion headers and programmable real-time units
UP 7000	Intel Processor N50, N97 or N100, depending on SKU	4 GB or 8 GB memory options, 32 GB or 64 GB eMMC options, Gigabit Ethernet and 12 V DC input

These are examples, not a performance ranking. Verify the exact SKU and board revision before ordering. The BeagleBone Black programmable real-time units are separate from its main application processor; they do not make every Linux application deterministically real-time.

How Do You Choose Between ARM and x86?

Choose the architecture that supports your required operating system, application binaries and peripheral drivers before comparing processor benchmarks.

Existing x86 application: confirm the required Windows or Linux version, driver availability and software licensing on the selected x86 board.
New embedded Linux application: an Arm board may fit well when its board support package, kernel and peripheral drivers cover the design.
Timing-sensitive control: check interrupt latency and scheduling requirements. A fast application CPU alone does not guarantee deterministic I/O.
Power-limited enclosure: measure the complete board under the actual workload, including RAM, storage, networking and cooling.

Do not assume all Arm boards use less power than all x86 boards. Idle behavior, CPU load, attached devices and the thermal solution can change the comparison. Run a small application prototype on candidate hardware before committing to its mechanical design.

Which Interfaces and Storage Does Your SBC Need?

Your SBC needs the interfaces, electrical levels and sustained data rates required by the attached devices, plus storage suited to its write workload and power-loss conditions.

USB and PCIe: check the number of independent controllers or lanes, shared bandwidth and whether an adapter is required.
GPIO, UART, I²C and SPI: verify voltage levels and pin functions. A UART header is not directly an RS-232 or RS-485 port.
Ethernet: confirm port count, link speed and any required isolation or power-over-Ethernet hardware.
Storage: compare capacity, write endurance, boot support and replacement access for microSD, eMMC or SSD options.
Display and camera: match connector pinout, interface version, resolution and supported drivers.

A camera gateway writing continuously to storage has different requirements from a kiosk that loads an application once and mostly reads data. Estimate daily writes and test recovery after an interrupted write; capacity alone does not establish storage suitability.

What Makes an SBC Suitable for Industrial Use?

An industrial SBC must meet the finished equipment’s temperature, power, mechanical and maintenance requirements in its actual installation.

Check the board specification rather than assuming “industrial” means −40°C to +85°C. For example, the UP 7000 technical table lists 0°C to +60°C with an airflow condition for its standard configuration, and a separate wider-temperature configuration with active cooling. Those conditions cannot be transferred to every board in the series.

Validate operation inside the enclosure at the specified ambient temperature and workload.
Check input-voltage tolerance, startup demand, brownout recovery and required transient protection.
Use connector retention and mounting suitable for vibration or repeated servicing.
Confirm watchdog behavior, recovery procedures, software updates and component-change notifications.
Agree on product availability and replacement strategy for the expected equipment service life.

A board-level compliance statement does not automatically cover the completed machine. Cables, enclosure, power supply and attached electronics affect the final system assessment.

When Does a Custom SBC Design Make Sense?

A custom SBC makes sense when an available board cannot meet the required connectors, shape, power input, component lifecycle or unit-cost target after engineering costs are included.

Approach	What you develop	Main trade-off
Off-the-shelf SBC	Application software, enclosure and integration	Fast start, but fixed board layout and vendor-controlled changes
SoM with custom carrier	Carrier PCB, application I/O and integration	Reuses the compute subsystem but adds module cost and connectors
Fully custom SBC	Processor, memory, power, PCB and software integration	More design control with greater bring-up and validation work

Compare total project cost: hardware engineering, board-support software, prototype spins, test fixtures, compliance work and expected production volume. Removing unused ports can simplify a product, but it does not automatically pay back the cost of developing a new computing platform.

What PCB Design Details Matter for an SBC Board?

Memory routing, reference planes, power sequencing, BGA escape routing and heat removal are the main PCB-level checks for a custom SBC.

Memory routing: use the processor vendor’s rules for topology, impedance and timing. Do not apply one generic length-matching tolerance to DDR3L, DDR4 and LPDDR4X.
Stackup: agree on the laminate, copper thickness, dielectric spacing and impedance targets before routing. Layer count follows routing and reference-plane needs; “SBC” does not specify a layer count.
BGA fanout: check pitch, pad geometry, solder-mask registration and via construction together. Fine pitch may require HDI features, but not every SBC needs microvias.
Power: verify regulator capacity, decoupling placement and the required relationship between processor and peripheral rails.
Thermal design: provide a practical path from hot components to a heat sink, enclosure or airflow without obstructing connectors or assembly access.

For a custom multilayer FR-4 PCB, we review the specified stackup and fabrication features against the manufacturing files. A change to dielectric thickness or copper weight should go back through the design review if it affects the impedance calculation.

SBC PCB design illustration highlighting memory routing, a reference plane and the thermal path

The layout release should also include mounting-hole positions, connector edge clearances, heat-sink keepouts and accessible test points. These mechanical details can prevent an electrically correct board from fitting its enclosure or production fixture.

How Should an SBC Board Be Assembled and Tested?

An SBC needs assembly inspection followed by electrical and functional tests that exercise its memory, storage and required interfaces.

At EBest Circuit, we can plan inspection and testing around the customer’s released BOM, assembly drawings and acceptance criteria. The production test scope must be agreed before the build; it is not implied by a generic PCB assembly order.

Before reflow: check component identity and orientation, moisture handling and solder-paste deposition around fine-pitch packages.
After reflow: use optical inspection for accessible joints and X-ray inspection where hidden-joint assessment is required.
First power-on: apply a controlled supply, check current draw and rails, then confirm reset, clocks and access to the debug console.
Functional test: boot the approved image, exercise memory and storage, and test the interfaces required by the product.
Load and recovery test: verify agreed thermal behavior, restart conditions and power-loss recovery using defined limits.

Illustration of optical assembly inspection and a connected SBC functional-test fixture

A memory test failure may involve soldering, layout, power integrity or initialization software. Record the failing test and board revision rather than assuming every boot failure is an assembly defect. X-ray inspection cannot establish software correctness, and a boot screen cannot prove every Ethernet or USB port works.

FAQ About SBC Boards

Is Raspberry Pi an SBC?

Raspberry Pi 5 is an SBC. The Raspberry Pi brand also includes microcontroller boards, so check the specific product rather than treating every Raspberry Pi device as the same category.

Can an SBC run Windows?

Some SBCs support Windows, but support depends on processor architecture, firmware, drivers and the Windows edition. Select a board whose manufacturer documents the required configuration.

What is included in a single board computer kit?

A single board computer kit may include a board, power supply, storage, enclosure and cooling accessories. Contents vary; verify the parts list rather than assuming the kit is ready for your application.

Does every SBC use a 5 V power supply?

No. Raspberry Pi 5 specifies a 5 V USB-C supply, while UP 7000 uses 12 V DC input. Use the exact board’s power specification, connector polarity and current requirement.

Is an SBC the same as a PLC?

No. An SBC is a computing-board format. A PLC is a control product with its own I/O, runtime and industrial operating requirements. Building a controller around an SBC does not by itself make it equivalent to a qualified PLC.

How Can EBest Circuit Support Your SBC Project?

We support custom SBC and carrier-board projects with PCB fabrication, component sourcing and assembly, with DFM review and testing defined around your design. We can review the board’s stackup, fine-pitch packages, assembly requirements and test access before a prototype or production build.

Send your Gerber files, BOM, placement files, quantities and test requirements to sales@bestpcbs.com. Include the approved software image and programming procedure if board programming or functional testing is required, so we can quote the manufacturing scope accurately.

Tags: sbc board, sbc types, sbc vs soc, single board computer examples, single board computer kit
Posted in best pcb, PCB Technology, PCBA | No Comments »

What Is FET? Types, Working Principles and PCB Uses

September 23rd, 2026

FET stands for field-effect transistor, a semiconductor device that uses an electric field to control current between its source and drain terminals. The gate voltage controls this current, allowing a FET to act as a switch or amplifier. JFETs and MOSFETs are two common families; MOSFETs are widely used in power supplies, motor drivers and digital circuits. For a PCB design, selecting a FET means matching its voltage rating, gate-drive requirements, losses and package to the circuit—not simply choosing a part with a high current rating.

Field-effect transistor packages on a PCB with conceptual gate-control and current-path inset

Key Takeaways

A FET controls source-to-drain conduction through its gate electric field; its three main terminals are gate, source and drain.
A MOSFET is a type of FET, not a separate category alongside FETs.
Conventional JFETs are normally on at zero gate-to-source voltage. Enhancement-mode MOSFETs are normally off; depletion-mode MOSFETs are normally on.
N-channel and P-channel devices require different gate-voltage polarities, measured relative to the source.
Threshold voltage does not guarantee low on-resistance. Check the specified RDS(on) at the available gate-drive voltage.
FETs serve switching, amplification and power-management functions. Gate charge, safe operating area and cooling matter alongside voltage and current ratings.
PCB performance depends on the correct footprint, gate-loop layout, thermal path, solder joints and electrical verification—not the transistor datasheet alone.

What Is FET in Electronics?

A FET is a voltage-controlled transistor whose gate field changes the conductivity of a semiconductor channel. It can regulate an analog signal or switch a load, depending on the device and circuit bias.

Gate (G): the control terminal.
Source (S): the reference terminal for gate-control voltage.
Drain (D): the other terminal of the controlled current path.

The important control voltage is VGS, or gate voltage minus source voltage. A gate at 5 V does not necessarily mean VGS is 5 V: if the source is at 4 V, VGS is only 1 V. Many discrete MOSFETs have their body connected internally to the source, while integrated devices may expose or control the body differently.

At EBest Circuit, we support the board surrounding these devices through PCB fabrication, component sourcing and PCB assembly services. The component part number, package drawing and circuit test requirements give our team the information needed to review the assembly.

FET Types: How Do JFETs and MOSFETs Differ?

JFETs control a conducting channel with a reverse-biased junction gate; MOSFETs use an insulated gate. This difference changes their input behavior, bias requirements and common applications.

Type	Gate structure	State at VGS = 0	Typical use
JFET	PN junction	Normally on	High-input-impedance buffers, analog front ends, current sources
Enhancement-mode MOSFET	Insulated gate	Normally off	Load switches, converters, motor drivers, logic
Depletion-mode MOSFET	Insulated gate	Normally on	Bias circuits, startup circuits, current regulation

These are common FET types, not an exhaustive list of every field-effect device. N-channel and P-channel describe channel polarity; enhancement and depletion describe behavior at zero gate-to-source voltage. They are different classification axes, not interchangeable names.

Common FET types showing JFET and enhancement-mode and depletion-mode MOSFET branches

How Does a FET Work?

A FET works by changing its channel conductivity in response to gate-to-source voltage. The gate controls the channel; it does not supply the load current.

For an enhancement-mode N-channel MOSFET, the FET working principle can be followed in three stages:

Gate held at the source potential: the intended channel is off, although leakage and body-diode conduction still need consideration.
Positive VGS applied: an electric field establishes a conducting channel. Near threshold, only the specified small test current is assured.
Adequate gate drive applied: the channel can reach low resistance under the datasheet conditions, allowing efficient switching within the device limits.

A conventional N-channel JFET behaves differently: its channel conducts at VGS = 0, and a negative gate-to-source bias narrows the channel until it approaches cutoff. Amplifier circuits bias either device in a suitable operating region rather than simply switching between on and off.

Conceptual enhancement-mode N-channel MOSFET channel absent with gate low and formed when the gate is driven

What Is MOSFET, and Why Is Gate Drive Important?

MOSFET means metal-oxide-semiconductor field-effect transistor. Its insulated gate draws little steady-state current, but a driver must charge and discharge the gate capacitances when switching.

This is why a microcontroller pin may switch a small signal MOSFET slowly yet be unsuitable for a larger device at high frequency. Total gate charge, switching frequency and the required transition time determine the driver demand. Gate leakage is not the same as switching-drive current.

VGS(th) is not a recommended full-load drive voltage. For example, onsemi’s 2N7002E datasheet specifies a 1.0–2.5 V threshold range at just 250 µA. Its maximum on-resistance is specified separately: 3.0 Ω at VGS = 4.5 V and ID = 50 mA, or 2.5 Ω at VGS = 10 V and ID = 240 mA, at 25°C under the stated pulse-test conditions.

Those values do not establish a guaranteed maximum RDS(on) at 3.3 V. For a 3.3 V control output, choose a device with an appropriate guaranteed resistance specification at that drive level or lower, or provide a suitable gate driver. Also keep VGS within its absolute maximum rating during ringing and transients.

Threshold voltage supports only a small test current while low on-resistance requires the specified gate drive

How Do N-Channel and P-Channel FETs Differ?

N-channel devices use electrons as their majority carriers, while P-channel devices use holes. For enhancement-mode MOSFET switching, an N-channel device needs positive VGS; a P-channel device needs negative VGS.

N-channel low-side switch: commonly placed between the load and ground. A ground-referenced source makes gate drive straightforward.
N-channel high-side switch: usually needs a driver that raises the gate above the moving source potential.
P-channel high-side switch: can turn on by pulling its gate below the source, provided the circuit limits VGS correctly.

The polarity rule alone does not apply unchanged to every JFET or depletion-mode circuit. Check the exact device type before interpreting gate voltages. Package shape also does not identify channel polarity or pin order.

What Is the Difference Between a FET and a BJT?

A FET controls conduction through a gate electric field; a bipolar junction transistor uses base-emitter bias and carrier injection, normally requiring continuous base-drive current. Both can switch or amplify signals.

Feature	FET	BJT
Main terminals	Gate, source, drain	Base, emitter, collector
Control interface	Gate-to-source voltage	Base-emitter bias with base current
Input loading	Typically high input impedance	Depends on base current and bias network
Switching-drive concern	Gate charge and voltage	Base-drive current and stored charge

A MOSFET is often attractive for efficient power switching, but a BJT may suit a low-cost signal stage or particular analog requirement. Neither device is universally quieter, faster or cooler; compare the actual parts and their operating conditions.

Where Are FETs Used on PCBs?

FETs are used on PCBs to switch power, amplify signals and control current in converters, drivers and input circuits.

DC-DC converters: MOSFETs switch energy into inductors and can provide synchronous rectification.
Motor and LED drivers: power MOSFETs control load current, often with pulse-width modulation.
Load switches and battery circuits: devices connect or disconnect power paths; reverse blocking may require paired MOSFETs.
Analog inputs: JFETs and FET-input amplifiers reduce loading on high-impedance signal sources.
Level shifting: selected small-signal MOSFET circuits translate logic levels when their topology and speed suit the interface.

The onsemi 2N7002E is a 60 V N-channel MOSFET in SOT-23, with low-side switching and level shifting among its listed applications. It is a small-signal example, not a substitute for a power MOSFET sized for a multiamp motor load.

Which FET Parameters Matter for Selection?

For a switching MOSFET, start with voltage stress, available gate drive, operating current, switching losses and thermal conditions. Then verify that the package and safe operating area support the intended duty.

Parameter	Selection check
VDS rating	Account for supply voltage and measured or predicted overshoot.
RDS(on)	Use the guaranteed value at the available VGS and allow for temperature rise.
Gate charge, Qg	Match the driver to switching frequency and transition-time requirements.
Safe operating area	Check startup, inrush and linear-mode stress, not only steady current.
Thermal data	Use conditions relevant to the actual PCB, copper, airflow and heatsink.
Package and pinout	Match the exact manufacturer land pattern and terminal assignment.

For a simple conduction-loss estimate, P = I² × RDS(on). An illustrative 2 A current through 50 mΩ produces 0.20 W while conducting continuously. This excludes switching and other losses; the resistance must reflect operating temperature. A large current number on a datasheet front page is not a guarantee for a small copper pad on your board.

What PCB Layout and Assembly Checks Matter for FETs?

The most important board-level checks are the correct pinout, a compact gate-drive loop, a viable heat path and verified solder connections. Each addresses a different failure mechanism.

Footprint and orientation: match the approved ordering code and package drawing. Similar SOT-23 or power packages can have different assignments.
Gate loop: keep the driver, gate resistor and source-return path compact. Avoid sharing a sensitive gate return with a high-current path when the device provides a separate source-sense connection.
Thermal connection: follow the package guidance for exposed pads, copper spreading and thermal vias. Do not connect a drain pad to ground simply because it is a thermal pad.
Solder deposition: review stencil openings and paste volume for the actual package. Excess paste, insufficient wetting or hidden-pad voiding may impair assembly quality.
Verification: combine appropriate visual/AOI or X-ray inspection with electrical testing. Measure gate drive and temperature under the specified load; a good-looking joint does not prove correct switching.

Before production, a PCB prototype build can help validate footprint fit, accessible test points and assembly behavior. The illustrations here explain concepts; they are not fabrication drawings or universal package pinouts.

Conceptual FET PCB assembly showing gate-loop routing, thermal copper, solder joints and footprint review

FAQ About FETs

Can a FET conduct current in both directions?

An enhanced MOSFET channel can conduct in either direction, but a conventional discrete power MOSFET also has a body diode. Consequently, turning the gate off does not provide blocking in both directions. Bidirectional isolation often uses back-to-back MOSFETs with a suitable control circuit. Verify the topology rather than treating one transistor as an ideal relay.

Does “FET” always mean MOSFET on a schematic?

No. FET is the broader device family. Informal power-electronics discussions often shorten MOSFET to FET, but a schematic may use a JFET or another field-effect device. The part number, symbol and datasheet—not the abbreviation alone—identify its technology, polarity, operating mode and terminal connections.

Why can a FET turn on when its gate is disconnected?

An insulated MOSFET gate can retain charge or pick up coupled noise when left floating. A gate-to-source pull resistor can establish a defined state, with its value chosen for leakage, noise immunity and driver loading. Keep unpowered devices under proper ESD handling; a floating gate is not a reliable off command.

Can a multimeter confirm that a MOSFET is good?

A multimeter can reveal some shorts or a body-diode path, but it cannot prove rated-voltage blocking, correct gate charge, full-load resistance or switching performance. In-circuit parallel paths can also mislead readings. Disconnect power and discharge stored energy before basic checks; use an appropriate controlled test setup for functional verification.

Can one FET replace another with the same voltage rating?

Not automatically. The replacement must also match channel type, operating mode, footprint, pinout and gate-drive requirements. Compare on-resistance, gate charge, safe operating area and thermal behavior. Changes in switching speed or capacitance may affect ringing and EMI even when the new part has a higher nominal current rating.

Need Support Building a FET-Based PCB?

At EBest Circuit, we have supported PCB and PCBA projects since 2006. We can coordinate bare-board fabrication, component sourcing and assembly for your approved FET-based design, with inspection and test requirements agreed for the project.

Send your Gerber files, BOM with exact FET part numbers, assembly drawings, quantity and test requirements to sales@bestpcbs.com. Include the load current, gate-drive voltage and any thermal requirements so we can review the board and assembly needs before quoting.

Tags: fet types, fet working, what is fet, what is fet in electronics, what is mosfet
Posted in best pcb, PCB Technology, PCBA | No Comments »

Multilayer Ceramic Capacitor (MLCC): Types, Applications, Selection & PCB Guide

September 22nd, 2026

A multilayer ceramic capacitor is one of the most widely used passive components in modern electronics. It provides compact, low-inductance capacitance for decoupling, filtering, power stabilization, RF circuits, automotive electronics, industrial control, and many other applications.

Selecting an MLCC requires more than reading the value in the BOM. Dielectric type, DC bias, working voltage, case size, frequency, and PCB mechanical stress can all change how the capacitor behaves after assembly.

Multilayer Ceramic Capacitors MLCC with cutaway internal layer structure

Key Takeaways

A multilayer ceramic capacitor (MLCC) is a non-polarized capacitor built from alternating ceramic dielectric and metal electrode layers.
C0G/NP0 offers high stability and low loss, while X7R and X5R provide much higher capacitance density.
Nominal capacitance is not always the effective capacitance. Class II MLCCs can lose capacitance as DC voltage increases.
Capacitance, working voltage, dielectric, case size, temperature range, ESR, ESL, self-resonant frequency, and termination type should be reviewed together.
Most MLCCs are SMD chip components, but leaded, axial, radial, stacked, and low-inductance versions are also available.
PCB bending, depaneling, screws, connectors, and rework can crack the ceramic body.
Second-source approval should compare the exact electrical and mechanical behavior of the replacement part, not only its capacitance and package size.

What Is a Multilayer Ceramic Capacitor (MLCC)?

A multilayer ceramic capacitor (MLCC) is a non-polarized capacitor made from many alternating ceramic dielectric and metal electrode layers inside one compact monolithic body.

The phrase multilayer ceramic capacitor MLCC describes the same stacked, monolithic capacitor technology.

The internal electrodes are connected alternately to the two end terminations. This arrangement effectively places many small capacitor layers in parallel, allowing relatively high capacitance in a very small package.

A typical MLCC contains ceramic dielectric layers, internal metal electrodes, external end terminations, barrier plating, and a solderable outer finish. SMD chip versions dominate modern electronics, although the same multilayer ceramic technology can appear in other package formats.

How Does a Multilayer Ceramic Capacitor Work?

An MLCC stores charge between overlapping internal electrodes separated by very thin ceramic dielectric layers.

The basic relationship is:

Capacitance ∝ dielectric permittivity × electrode area ÷ dielectric thickness

This is why multilayer construction is effective. Capacitance can be increased by adding active layers, increasing electrode overlap, using higher-permittivity ceramic, or reducing dielectric thickness.

Because the electrode layers are connected in parallel, increasing layer count raises the total effective electrode area without greatly increasing the component’s external size.

MLCC cutaway showing ceramic layers internal electrodes and end termination

What Are the Main MLCC Dielectric Types?

The main MLCC dielectric types are C0G/NP0, X7R, X5R, and lower-stability ceramics such as Y5V or Z5U.

Dielectric	Main Characteristic	Typical Use
C0G / NP0	Very stable, low loss	RF, resonant circuits, precision filtering
X7R	High capacitance density with moderate variation	General decoupling and power filtering
X5R	High capacitance density in compact packages	Low-voltage power rails and portable electronics
Y5V / Z5U	Very high nominal capacitance, poorer stability	Limited non-precision applications

C0G is preferred when capacitance stability and low dielectric loss matter most. X7R and X5R are more common where higher capacitance density is needed for bypassing, decoupling, and power rails.

C0G NP0 X7R X5R Y5V Z5U MLCC dielectric comparison

What Specifications Matter When Selecting an MLCC?

The most important MLCC specifications are effective capacitance, working voltage, dielectric type, package size, temperature range, ESR, ESL, self-resonant frequency, and termination type.

Specification	Check Item
Capacitance	Required nominal value
Effective capacitance	Actual capacitance at operating voltage
Rated voltage	Electrical margin above working voltage
Dielectric	C0G, X7R, X5R or another class
Tolerance	Initial capacitance variation
Case size	PCB space and electrical behavior
Temperature range	Environmental requirement
ESR	Resistive loss and ripple behavior
ESL	High-frequency inductive behavior
Self-resonant frequency	Frequency range where the part remains capacitive
Termination type	Standard or flex-resistant construction
Qualification	Automotive or other reliability requirement

Working voltage deserves particular attention. Two capacitors with the same nominal capacitance and voltage rating can deliver very different effective capacitance under the same DC bias.

For PCBA sourcing, the exact manufacturer part number is more useful than a generic BOM description such as “10 µF, 25 V, 0603.”

What Is the DC Bias Effect in MLCC Capacitors?

DC bias is the reduction in effective capacitance that occurs in many Class II MLCCs when DC voltage is applied.

This effect is especially important for X7R and X5R parts. A capacitor marked 10 µF does not necessarily provide 10 µF at its actual operating voltage.

The reduction depends on the dielectric formulation, rated voltage, applied voltage, nominal capacitance, case size, manufacturer series, and internal construction.

Engineers should therefore check the manufacturer’s capacitance-vs-voltage curve for the exact part number. For power rails and regulator input/output capacitors, effective capacitance at the real operating voltage is more useful than nominal capacitance.

MLCC DC bias effect showing effective capacitance decreasing under DC voltage

What Are Multilayer Ceramic Capacitors Used For?

MLCCs are mainly used for decoupling, bypassing, power filtering, noise suppression, RF circuits, and signal conditioning.

Common multilayer ceramic capacitor applications range from local IC decoupling to power filtering, RF networks, and signal conditioning.

Typical applications include processor and MCU power decoupling, FPGA and memory rails, DC-DC converters, LDO stabilization, RF matching, resonant circuits, automotive ECUs, industrial electronics, medical electronics, and communication equipment.

For RF and precision circuits, C0G/NP0 parts are commonly selected because their capacitance remains more stable. Automotive applications may add qualification and flex-resistant termination requirements.

MLCC applications in decoupling RF automotive and industrial electronics

SMD, Chip, Leaded and Axial MLCCs: What Is the Difference?

MLCCs are available in several package styles, although SMD chip MLCCs are by far the most common in modern PCB assembly.

An SMD multilayer ceramic capacitor is optimized for automated pick-and-place and reflow assembly.

SMD chip MLCCs support automated placement and reflow. Leaded, axial, and radial versions serve through-hole or legacy applications, while stacked and low-ESL constructions address higher capacitance, ripple, or high-frequency requirements.

MLCC therefore describes the internal capacitor technology, not one fixed external package.

How Are Multilayer Ceramic Capacitors Manufactured?

MLCCs are manufactured by forming thin ceramic sheets, printing internal electrodes, stacking and laminating the layers, sintering the ceramic body, and then adding external terminations and electrical testing.

The multilayer ceramic capacitor manufacturing process controls layer thickness, electrode registration, densification, termination quality, and final electrical performance.

The main production sequence is:

Green Sheet → Printing → Stacking → Lamination → Cutting → Sintering → Termination → Testing

Uniform ceramic thickness, electrode alignment, lamination pressure, firing conditions, and termination quality all affect final capacitance and reliability.

MLCC manufacturing process from green sheet printing stacking lamination cutting sintering termination to testing

MLCC vs Tantalum vs Film Capacitor: What Is the Difference?

MLCCs are generally smaller and lower in ESR/ESL, tantalum capacitors provide more stable capacitance under DC bias, and film capacitors are often preferred for high stability, higher voltage, or power applications.

Factor	MLCC	Tantalum	Film Capacitor
Polarity	Non-polarized	Polarized	Usually non-polarized
Size	Very compact	Compact	Usually larger
ESR / ESL	Very low	Moderate	Application dependent
DC bias effect	Important for Class II	Much smaller	Generally low
Capacitance stability	Depends on dielectric	Relatively stable under bias	Generally stable
Mechanical concern	Ceramic flex cracking	Different failure modes	Less ceramic-flex sensitivity
Typical use	Decoupling, filtering, RF	Bulk capacitance, power rails	Power, filtering, precision

The final choice should be based on actual voltage, frequency, ripple current, temperature, PCB space, and reliability requirements.

What PCB Layout Rules Matter for MLCCs?

MLCC layout should minimize electrical loop inductance and mechanical PCB strain.

For decoupling, place the capacitor close to the relevant IC power pin, keep the power and ground connections short, and use low-inductance vias where appropriate.

Mechanically, avoid placing sensitive ceramic capacitors immediately beside V-score lines, routing tabs, screw holes, large connectors, press-fit components, board edges, or other high-flex regions.

Pad geometry should follow the component manufacturer’s recommended land pattern, while component orientation and flex-resistant termination options should be considered for mechanically demanding assemblies.

What Causes MLCC Cracking and Failure?

MLCC cracking is mainly caused by mechanical strain transferred from the PCB into the brittle ceramic body.

Common sources include PCB bending, depaneling, screw tightening, connector insertion, press-fit operations, mechanical shock, thermal shock, rework, and unsuitable pad geometry.

Soft-termination or flex-resistant MLCC series can absorb part of the board strain, while keeping MLCCs away from high-flex board edges and separation lines reduces the mechanical load transferred into the ceramic.

MLCC reliability showing board flex crack risk soft termination and edge placement

Who Are the Major Multilayer Ceramic Capacitor Manufacturers?

Major MLCC manufacturers include Murata, TDK, Samsung Electro-Mechanics, Taiyo Yuden, Yageo/KEMET, KYOCERA AVX, Vishay, and Walsin.

These suppliers serve consumer, automotive, industrial, telecommunications, computing, RF, and high-reliability markets. Their parts should not be treated as automatically interchangeable even when nominal capacitance, voltage, dielectric, and case size appear identical.

Second-source approval should compare DC bias behavior, effective capacitance, ESR, ESL, temperature characteristics, termination design, qualification, and lifecycle status.

What Should Buyers Check Before Approving an MLCC for PCBA?

Buyers should verify the exact part number, effective capacitance, voltage rating, dielectric, case size, temperature class, termination, qualification, and lifecycle status before approving an MLCC.

Check Item	Why It Matters
Manufacturer + exact P/N	Prevents ambiguous substitution
Nominal capacitance	Basic circuit requirement
Effective capacitance	Real capacitance under operating bias
Rated voltage	Electrical margin
Dielectric	Stability and capacitance density
Tolerance	Initial value range
Case size	PCB area and electrical behavior
Temperature range	Environmental compatibility
Termination type	Mechanical reliability
Qualification	Automotive or other reliability requirements
Lifecycle status	Long-term sourcing
Approved alternative	Supply continuity
Reel/package information	SMT production compatibility

For automotive or mechanically demanding assemblies, check whether a flex-resistant or soft-termination series is required. For power applications, review effective capacitance at the actual working voltage before approving an alternate part.

FAQ About Multilayer Ceramic Capacitors

1. What does MLCC stand for?
MLCC stands for Multilayer Ceramic Capacitor.

2. Are multilayer ceramic capacitors polarized?
No. Standard MLCCs are non-polarized and can be installed in either electrical orientation.

3. Is X7R better than C0G?
No. C0G provides better capacitance stability and lower loss, while X7R provides much higher capacitance density.

4. Why does an MLCC lose capacitance under voltage?
Class II ceramic dielectrics such as X7R and X5R can lose effective capacitance when DC voltage is applied. The amount depends on the exact part construction.

5. Can MLCCs crack on a PCB?
Yes. PCB bending, depaneling, connector forces, screw mounting, press-fit operations, rework, and mechanical shock can crack the ceramic body.

6. Are all MLCCs SMD components?
No. SMD chip MLCCs are the most common, but leaded, axial, radial, stacked, and specialized low-inductance versions are also available.

A multilayer ceramic capacitor should be selected based on its actual operating behavior, not only the capacitance value in the BOM. Dielectric type, DC bias, working voltage, frequency behavior, package size, PCB placement, and mechanical loading all influence finished-PCBA reliability.

For projects requiring PCB fabrication, component sourcing, SMT assembly, inspection, and functional testing, send your Gerber files, BOM, assembly drawings, approved component list, and test requirements to sales@bestpcbs.com for DFM and PCBA review.

Tags: multilayer ceramic capacitor, multilayer ceramic capacitor applications, multilayer ceramic capacitor manufacturing process, multilayer ceramic capacitor mlcc, smd multilayer ceramic capacitor
Posted in best pcb, Electronic Components, PCB Assembly / PCBA, SMT Technology | No Comments »

PCB Engineering Change Order for Controlled Production Updates

September 22nd, 2026

A PCB engineering change order provides a controlled way to move an approved board change into production without mixing old and new manufacturing data. A layout update may appear straightforward, but it can also affect the Gerber or ODB++ data, fabrication drawing, BOM, centroid file, assembly drawing, test instructions, and inventory already in the supply chain. If those records do not change together, a technically correct modification can still produce the wrong PCB or PCBA.

For customers, effective change control means fewer revision-related delays, less risk of unusable inventory, and clearer evidence of what was built in each production lot. EBest Circuit supports customer-approved changes through manufacturing-data review, DFM, PCB fabrication, component sourcing, assembly, inspection, and agreed testing. Product-design approval and the final decision to release a change remain with the customer.

What Is a PCB Engineering Change Order?

A PCB engineering change order, commonly called an ECO, is the authorized record used to implement a change to a released PCB or PCBA. It identifies what must change, which product or board revision is affected, when the change becomes effective, and how existing material or work in progress should be handled.

The ECO is more than a marked-up drawing. It connects the approved engineering decision with the files and production actions needed to build the correct revision. Depending on the change, it may control:

PCB layout or circuit changes;
stackup, copper weight, material, surface finish, or controlled-impedance requirements;
component, footprint, package, or approved-alternative changes;
fabrication, assembly, inspection, programming, or test instructions;
board revision markings and product traceability;
the disposition of open purchase orders, bare PCBs, components, and assembled boards.

The term ECO is also used in PCB design software for transferring differences between a schematic and a layout. That CAD function can be one step in making a design change, but it does not replace the broader approval, document control, effectivity, and production-disposition process discussed here.

How Do an Engineering Change Request, Engineering Change Order, and Engineering Change Notice Differ?

An engineering change request proposes and evaluates a possible change. An engineering change order authorizes the detailed implementation after the technical and business impacts have been reviewed. An engineering change notice communicates the approved change to the people and suppliers who must act on it.

Record	Primary purpose	Typical status
ECR	Describe a problem or proposed improvement and evaluate its impact	Under review
ECO	Define and authorize the files, revision, effectivity, and actions required	Approved for implementation
ECN	Notify affected teams or suppliers that the approved change must be applied	Released for communication

Terminology varies among companies and quality systems. Some organizations combine ECO and ECN into one controlled record, while others use ECN as the main approval document. The abbreviation matters less than a clear workflow: the manufacturer must know which change is approved, which data set is valid, and which production units are affected.

Which Files Belong in the PCB Engineering Change Order Process?

The PCB engineering change order process should include every released record whose content or revision is affected by the change. Updating only the layout file is not enough if purchasing, assembly, inspection, or testing still uses older instructions.

The controlled package commonly includes:

native design data when it forms part of the agreed handoff;
Gerber or ODB++ files, NC drill data, and IPC-356 netlist data;
the fabrication drawing and stackup specification;
the BOM with manufacturer part numbers and approved alternatives;
the centroid, pick-and-place, or CPL file;
the assembly drawing, polarity information, and special process notes;
stencil requirements where pad or package changes affect solder-paste printing;
programming files, test procedures, fixtures, and acceptance limits when applicable;
mechanical drawings or enclosure interfaces affected by board dimensions, holes, connectors, or component height;
revision history, approval record, effectivity, and disposition instructions.

Not every ECO changes every file. A silkscreen correction may affect only a few records, while a footprint or component change may affect the BOM, land pattern, centroid data, stencil, assembly drawing, inspection criteria, and test coverage. The affected-file list should follow the actual technical impact instead of a fixed document count.

How Does an ECO Keep Gerber, BOM, CPL, and Drawings on the Same Revision?

An ECO keeps manufacturing records aligned by releasing them as one identified revision package rather than as unrelated replacement files. The package should make the relationship between the board revision, document revisions, and effective production point unambiguous.

Several controls prevent mixed-revision builds:

one approved package or controlled download location for the released files;
consistent revision identifiers across the ECO and affected documents;
a change summary that identifies superseded and replacement files;
confirmation that the BOM references footprints actually present in the released PCB data;
confirmation that CPL coordinates, rotations, and reference designators match the new assembly data;
removal or quarantine of obsolete files from active production folders;
a documented release date, lot, serial number, purchase order, or other effectivity point.

File names alone are weak revision control. A folder containing names such as “final,” “latest,” and “updated-final” does not tell fabrication, purchasing, and assembly teams which combination was approved. A defined revision package reduces clarification cycles and helps ensure that the quoted, fabricated, assembled, and inspected product refers to the same design state.

How Does Engineering Change Effectivity Control Work in Progress and Existing Inventory?

Engineering change effectivity defines when the approved change begins and which units must use it. This prevents a new revision from being applied casually to some orders while older material continues through production without an agreed decision.

The correct treatment depends on the reason for the change and the condition of existing material:

Immediate implementation: Stop affected production and apply the change before more units are built. This is appropriate when the old revision presents an unacceptable functional, safety, compliance, or assembly risk.
Rework: Modify eligible bare boards or assemblies when the rework method is approved, technically reliable, traceable, and economically justified.
Use as is: Complete or ship existing units when the previous revision remains acceptable for their intended application and the customer authorizes that disposition.
Run out existing inventory: Consume approved old-revision material before switching at a defined lot, date, or order.
Scrap or segregate: Prevent obsolete or nonconforming material from entering later production when it cannot be used or reworked safely.

This decision may affect open component orders, bare PCB stock, stencils, fixtures, work instructions, partially assembled panels, finished goods, and replacement-service inventory. Recording the disposition in the ECO lets customers understand the cost and schedule impact before the production switch is made.

When Does a PCB Change Need a New Revision or Part Number?

A PCB change normally needs a new revision when the updated board must remain distinguishable from the previously released version but still represents the same basic product. A new part number is more appropriate when the change creates a different item that must be separately ordered, stocked, serviced, qualified, or used only in specific configurations.

Revision changes often cover controlled updates such as routing corrections, approved footprint changes, documentation corrections that affect manufacturing, or component changes that preserve the product identity. A new part number may be justified when interchangeability is lost, connector pinout or mechanical fit changes, electrical function changes substantially, regulatory or customer qualification must be separated, or both versions must remain active at the same time.

There is no universal rule based only on the size of a layout edit. The customer’s configuration-management system should define the decision. From the manufacturing side, the identifier must let purchasing, production, inspection, inventory, and field support distinguish the permitted versions without relying on memory or visual guesswork.

Board markings also need deliberate treatment. If the fabrication data changes but the visible revision marking does not, the new and old bare boards may become difficult to separate. If a marking changes, the fabrication drawing and released artwork must agree on its content and location.

How Does an Approved ECO Move Into PCB and PCBA Production?

An approved ECO moves into production by converting the customer’s released change package into controlled fabrication, sourcing, assembly, inspection, and test instructions. The manufacturer should not reinterpret an unapproved concept or decide which product behavior is acceptable; its role is to implement the authorized revision consistently and identify manufacturability conflicts before material is committed.

The production transition normally follows a practical sequence:

The new package is compared with the previous released data to identify affected PCB, BOM, placement, drawing, and test records.
DFM and assembly reviews determine whether the change introduces new stackup, impedance, spacing, tooling, stencil, component, or inspection requirements.
Questions and manufacturing exceptions are resolved before release. Any manufacturer-proposed adjustment returns to the customer for approval when it changes the authorized design intent.
Effectivity and material disposition are applied to quotations, purchase orders, inventory, work in progress, and production travelers.
The correct revision is fabricated and assembled under controlled instructions.
Inspection and agreed testing verify the features affected by the change, with lot and revision records retained for traceability.

The amount of validation should match the change risk. A documentation-only correction may need a focused record review, whereas a new BGA footprint, stackup, high-current path, controlled-impedance route, or critical component may justify first-article inspection, X-ray, dimensional verification, impedance testing, electrical testing, or customer-defined functional testing. This risk-based approach avoids repeating unrelated tests while still checking what the ECO actually changed.

EBest Circuit can fabricate and assemble customer-designed PCBs and PCBAs under an approved ECO, from prototypes to repeat production. If you have a revised Gerber package, BOM, drawings, and change record, send them to sales@bestpcbs.com for DFM review and quotation. A clearly released package helps us identify affected processes early, protect the approved revision during production, and reduce preventable delays caused by mixed manufacturing data.

FAQs About PCB Engineering Change Order

Can an email replace a formal PCB ECO?

An email can communicate a request, but it may not provide adequate approval, revision, effectivity, affected-file, and disposition control. For repeatable production, the final decision should be captured in the customer’s authorized change-control record or another formally approved release method.

Does every BOM substitution require a PCB engineering change order?

Not necessarily. A previously approved alternate may be used under the customer’s existing BOM and sourcing controls. An ECO may be required when a substitute changes the footprint, ratings, function, qualification status, assembly process, inspection criteria, test result, or another released requirement. The customer’s change-control rules determine the approval path.

Can an ECO be applied after PCB fabrication has started?

Yes, but the available options become narrower and may add cost or delay. The affected work can be stopped, reworked where technically acceptable, used as is with authorization, or scrapped. The decision should consider the fabrication stage, change risk, traceability, delivery requirement, and customer approval.

Is a PCB design-tool ECO the same as a manufacturing ECO?

No. A design-tool ECO synchronizes differences between design databases, such as a schematic and PCB layout. A manufacturing ECO controls the authorized revision, affected production data, effectivity, material disposition, implementation, and verification. The CAD update may support the change, but it does not provide the complete production-control record.

What should be sent to a PCB and PCBA manufacturer after an ECO is approved?

Send the complete released manufacturing package, the approved change record or clear change summary, revision and effectivity information, material-disposition instructions, and any updated inspection or test requirements. Providing one coherent package is safer than sending individual replacement files across separate email threads.

How can customers confirm that the correct ECO revision was built?

Use board revision markings where appropriate and connect the production lot to the released data, traveler, inspection results, and agreed test records. The required evidence depends on the product and quality system, but it should allow the customer and manufacturer to identify which approved revision was used without reconstructing the history from informal messages.

A well-controlled PCB engineering change order turns an approved technical decision into a traceable production update. By aligning affected files, effectivity, inventory disposition, manufacturing instructions, and verification, customers can introduce necessary changes without losing control of which PCB or PCBA revision reaches the next build. For support implementing an approved change in prototype or repeat production, contact EBest Circuit at sales@bestpcbs.com.

Tags: Engineering Change Notice, PCB Engineering Change Order, PCB Revision Control
Posted in best pcb, bestpcb, pcb manufacturing | No Comments »

How to Export Gerber Files from KiCad for PCB Manufacturing?

September 22nd, 2026

Here is how to export Gerber files from KiCad for manufacturing: create one Gerber file for each required PCB layer and separate drill files for holes and slots. Most fabricators expect both sets in one ZIP. Gerbers without drill data omit the hole pattern, while a native KiCad board file leaves the final plot choices to someone else.

The instructions match KiCad 10 and follow the files from the saved board through GerbView and into a revision-controlled ZIP. Older releases may label the menu differently, although File > Plot opens the plotting dialog in many earlier versions.

How to export Gerber files from KiCad, PCB design layers and fabrication output on an engineering workstation

What Should You Check Before Exporting Gerber Files from KiCad?

Check the KiCad source design before creating manufacturing files. The plotter reproduces the board as it exists; it does not repair an open outline, an unconnected net, an omitted layer, or an outdated copper-zone fill. Open the final production revision in PCB Editor and save it before running the checks.

Run the Design Rules Checker: Use Inspect > Design Rules Checker and review violations, unconnected items, and schematic-to-PCB differences. Resolve real problems and document any deliberate exclusions instead of assuming the fabricator will interpret them.
Validate Edge.Cuts: Confirm that the external outline is one closed, non-self-intersecting shape and that intended internal cutouts are also closed. KiCad shades a valid closed board area and reports malformed outlines through DRC.
Confirm the copper stack: Verify the enabled copper-layer count and order in Board Setup. A 4-layer board must have the correct internal layers and stack sequence before those layers are plotted.
Review mask and silkscreen: Check mask openings around pads and make sure silkscreen text and graphics do not cover exposed solderable areas. Remove information that does not belong on the manufactured board.
Inspect drilled features: Review plated through holes, non-plated holes, slots, and mounting holes. Confirm that each feature has the intended plating property and finished purpose.
Refill copper zones: Press B or use the zone-fill command, then save the board. Enabling KiCad’s plot-time zone check provides another safeguard, but the source board should already be current.

This preflight checks the editable design. It does not replace the later Gerber review, which checks the actual files that will leave your computer.

How to Export Gerber Files from KiCad Step by Step?

Use PCB Editor to export Gerber files from KiCad into a clean revision folder, then inspect what KiCad actually wrote. Each step below keeps the output tied to the same saved board revision and ends with something you can check.

How to export Gerber files from KiCad, process from PCB Editor through separate Gerber and drill output to a verified ZIP

Open the final board in PCB Editor: Confirm the project name and revision, save the .kicad_pcb file, and make sure the canvas shows the board you intend to order.
Open the Gerber dialog: Choose File > Fabrication Outputs > Gerbers (.gbr). File > Plot opens the same dialog when Gerber is selected as the plot format. The Plot window should show the board’s enabled layer list.
Choose a dedicated output folder: Use a new folder such as fabrication/rev-b/gerber. It should not contain files from an earlier revision. The output path displayed in the dialog should point to that folder.
Select the required layers: Include every copper layer used by the design, the top and bottom solder-mask layers, any silkscreen layers that contain board markings, and Edge.Cuts. Add paste layers only when the package also serves a stencil or assembly workflow.
Set the relevant plot options: Keep the drawing sheet out of the Gerbers, keep the plot at its real scale, use a consistent origin, and enable the zone-fill check. Change compatibility options only when your manufacturer provides a requirement.
Click Plot: KiCad writes one Gerber file for each selected layer. Review the Output Messages panel and open the destination folder. You should see a fresh set whose timestamps match this export.

Do not ZIP the folder yet. First confirm the layer set, generate the drill output, and inspect both together.

Which Gerber Layers Should You Export for PCB Manufacturing?

A bare PCB package normally includes every used copper layer, both solder-mask layers, the required silkscreen layers, and Edge.Cuts. The exact count changes with the board stack and whether artwork exists on the back. The table separates bare-board files from stencil or assembly outputs.

Layer	Purpose and Export Rule
F.Cu	Front copper image; always export it when the board uses front copper.
B.Cu	Back copper image; always export it for a 2-layer or multilayer board.
In1.Cu, In2.Cu, and other internal copper	Inner signal or plane layers; export every enabled internal layer in the correct stack order.
F.Mask	Front solder-mask openings; normally required.
B.Mask	Back solder-mask openings; normally required.
F.Silkscreen	Front legend and reference markings; export when used.
B.Silkscreen	Back legend and reference markings; export when used.
Edge.Cuts	External profile and internal cutouts; required for the manufactured board shape.
F.Paste and B.Paste	Solder-paste stencil apertures; use for stencil or assembly preparation, not normally for bare-board fabrication.
F.Fab and B.Fab	Assembly and fabrication drawing detail; send separately when requested and never substitute it for copper or Edge.Cuts.

For a typical 2-layer bare board, the practical minimum is F.Cu, B.Cu, F.Mask, B.Mask, Edge.Cuts, the silkscreen layers that contain artwork, and the drill file or files. For a multilayer design, add every internal copper layer and provide the intended stackup separately so the manufacturer can map the layer order without guessing.

Which KiCad Gerber Plot Settings Matter for PCB Manufacturing?

Keep the defaults that preserve 1:1 geometry, then change only options that affect file interpretation or compatibility. KiCad exposes many plot controls because the same dialog can also create PDF, SVG, DXF, and other outputs. A Gerber handoff uses a smaller, more focused subset.

Plot drawing sheet: off: The page border and title block are document furniture, not the PCB profile. Edge.Cuts should carry the board shape.
Check zone fills before plotting: on: KiCad warns that plot output can be wrong when outdated zones are not checked and refilled.
Plot on all layers: normally empty: This feature overlays selected items onto every base layer. Use it only for a deliberate documentation requirement, not to place Edge.Cuts or silkscreen into every Gerber by habit.
Gerber X2: use when accepted: X2 adds file-function, net, and other attributes that can help CAM interpretation. Turn it off only when a manufacturer reports that its older CAM system cannot process X2.
Protel filename extensions: optional: This changes names such as .gbr to familiar extensions such as .GTL and .GBL; it does not change the board geometry. Follow the recipient’s naming preference.
Gerber job file: useful but not a replacement: A .gbrjob file can carry stackup, material, and finish metadata. Keep the ordinary layer files and fabrication notes because not every CAM flow relies on the job file.
Coordinate format: manufacturer-bound: Use the fabricator’s recommendation when one is published. Do not change precision merely because an older tutorial shows a different KiCad release.
Origin: keep it consistent: Absolute origin is a common default when the manufacturer does not specify otherwise. If you use a drill/place origin, apply the same origin to the Gerber and drill outputs.

The current option definitions and compatibility notes are documented in the KiCad 10 PCB Editor manual. A manufacturer’s written CAM requirement takes precedence over a generic screenshot.

How Do You Generate Drill Files in KiCad?

Generate the drill data separately after plotting the Gerber layers. From the Plot dialog, click Generate Drill Files, or use File > Fabrication Outputs > Drill Files (.drl). Point the drill dialog to the same revision-specific output folder and keep its origin consistent with the Gerbers.

Choose Excellon for broad compatibility: KiCad states that most PCB manufacturers require Excellon. Use Gerber X2 drill output only when the recipient explicitly accepts it.
Leave Mirror Y axis off: KiCad advises against mirroring when a third party manufactures the PCB.
Leave Minimal header off: Enable it only when the manufacturer asks for a reduced header.
Keep PTH and NPTH separate by default: KiCad normally writes plated and non-plated holes to separate Excellon files. Merge them only when the manufacturer requests a single drill file.
Keep alternate oval-hole mode off: The normal route-command representation suits most manufacturers. Use the alternate mode only for a stated CAM requirement.
Generate and inspect the output: Click Generate, confirm that the messages contain no unresolved errors, and verify that the expected .drl file or files appear with the new Gerbers.

A drill map can help a human review the hole pattern, but it does not replace the machine-readable drill file. If blind, buried, or laser-drilled vias are involved, also send the stackup and via-span requirements because a flat file list does not fully describe the fabrication sequence.

How Do You Verify KiCad Gerber and Drill Files Before Manufacturing?

Load the exported Gerber files and Excellon drill files together in GerbView and inspect the manufacturing data, not the original PCB canvas. KiCad’s Gerber Viewer lets you switch layers on and off, compare their alignment, and view drill data over the copper and outline.

How to export Gerber files from KiCad, Gerber outline copper layers and drill hits checked together before manufacturing

Load the complete export: Open all Gerber layers, then load every Excellon drill file. The correct result is one entry for every expected copper, mask, silkscreen, profile, and drill output, with no file from an older revision.
Check the board outline: Edge.Cuts should show one continuous external profile plus only the intended internal cutouts. The displayed dimensions should match the source design, without a page border or unexpected outer rectangle.
Inspect copper in stack order: Toggle each copper layer and check that its routing, pads, planes, and board position belong to that layer. A multilayer set should contain every intended copper layer once, in the expected order.
Compare mask with pads: Solder-mask openings should align with exposed pads, test points, and other intended copper. Unexpected solid mask over a pad is a reason to return to the source board.
Review silkscreen clearance: Text should remain readable, appear on the intended side, and stay off exposed pads. In the finished-board view, bottom-side text should have the expected orientation rather than looking accidentally mirrored.
Overlay drill hits: Plated holes should sit inside their pads, while non-plated mounting holes and slots should align with the outline and mechanical features. A consistent offset across the board points to mismatched origins.
Check high-risk details: Zoom into fine-pitch footprints, connector rows, mounting holes, slots, board-edge features, and dense via fields. The correct export should preserve the same clearances and feature positions you approved in the source board.

A successful Plot message proves that KiCad wrote files; it does not prove that you selected every required layer or that the output represents the intended revision. When anything looks wrong, repair the source board, delete the affected export set, regenerate Gerbers and drills together, and repeat the viewer inspection.

What Should a Complete KiCad PCB Manufacturing Package Include?

Gerber files describe layer images and board geometry, but they do not communicate every manufacturing requirement. A complete handoff combines the exact files you inspected with the construction details that a fabricator cannot safely infer from those images.

Verified Gerber set: Include the same complete layer set that you loaded and checked in GerbView, without files from another revision.
Machine-readable drill data: Include the Excellon PTH and NPTH output generated for that Gerber set. A drill map is useful for review but is not a production drill file.
Stackup and construction details: State the layer order, material, finished thickness, copper weight, and any blind, buried, laser-drilled, filled, or capped via requirements.
Fabrication notes: Specify surface finish, controlled impedance, critical tolerances, panelization, special routing, and other order-specific requirements.
Optional Gerber job file: Include the .gbrjob file when generated and accepted, but do not use it instead of the individual Gerber, drill, and fabrication files.

If the same project also includes assembly, the assembler may request a BOM, component placement file, assembly drawings, paste layers, programming data, and test instructions. Those are PCBA inputs and should not be confused with the bare-board Gerber and drill package.

How Should You Package and Send KiCad Gerber Files to a Manufacturer?

Create the ZIP from the exact folder you inspected, then keep that verified archive unchanged for upload. This simple rule prevents a late file replacement from breaking the relationship between the viewer check and the package that reaches the manufacturer.

Start from an empty revision folder: Remove or archive the previous output before plotting. Never mix a new copper layer with an old outline or drill file.
Generate Gerbers and drills in one session: Matching timestamps are not a substitute for revision control, but they make accidental carryover easier to spot.
Verify that exact set: Load the folder’s Gerbers and drills into GerbView. Do not inspect one folder and ZIP another.
Use a clear archive name: Include the project and revision, such as controller-rev-b-fab.zip, without spaces or ambiguous words such as final-final.
Attach the fabrication notes: State quantity, material, stackup, finished thickness, copper weight, surface finish, impedance requirements, tolerances, delivery location, and any special process requirement that is not unambiguously encoded in the image files.
Review the manufacturer’s preview: If the upload portal renders the ZIP, compare the outline, layer count, dimensions, and drill pattern with your verified GerbView result before placing the order.

A .kicad_pcb file can be useful for engineering discussion, but it is not a universal substitute for fabrication output. Send it only when the manufacturer accepts native KiCad data and you are comfortable sharing the editable design. Keep the approved Gerber and drill archive as the order’s controlled manufacturing package.

Which KiCad Gerber Export Mistakes Cause Manufacturing Problems?

The most consequential errors are missing outline or drill data, wrong layer selection, mixed revisions, and inconsistent origins. The first symptom usually points to a narrow check, so correct the source or export decision instead of changing unrelated settings.

Issue	Impact and First Fix
Edge.Cuts missing or open	The board outline cannot be interpreted correctly. Close the outline and re-export Edge.Cuts.
Drill files omitted	Holes and vias are missing. Generate and include the Excellon drill files.
NPTH file omitted	Mounting or tooling holes may be missing. Include the separate NPTH drill file.
Inner layer missing	Multilayer connectivity may be incorrect. Check the stackup and re-export all copper layers.
Gerber and drill origins differ	Drill hits are shifted from pads. Use the same origin for both outputs.
Mixed revisions	Layers and holes may not align. Clear the output folder and export one revision.
Wrong extra layers included	The manufacturing package becomes ambiguous. Remove unrelated paste or fabrication outputs.
Files not inspected	Missing or shifted output may go unnoticed. Verify the exported files in GerbView.

Silkscreen over exposed pads is another frequent review finding, but it is not solved by selecting more files. Correct the source artwork or use the approved silkscreen-to-mask handling, then regenerate and inspect the affected silkscreen and mask layers.

FAQs About KiCad Gerber Export

Q1: Does the .gbrjob file include the drill data?

A1: No, the .gbrjob file does not include drill data. KiCad creates it as part of the Gerber export, while hole data is generated separately. Include the Excellon .drl files even when a .gbrjob file is present.

Q2: Can a drill map replace the .drl file?

A2: A drill map cannot replace the .drl file. It is a visual aid for reviewing hole locations and tool sizes. The manufacturer still needs the machine-readable drill file, normally in Excellon format.

Q3: How should slots and internal cutouts be included?

A3: Put board cutouts on Edge.Cuts and configure slots through their pad or drill properties. Use closed shapes for cutouts and set the correct plating property for each slot. Open the Gerbers and drill files together to confirm that every routed feature is present and aligned, and follow the manufacturer’s minimum slot and routing rules.

Q4: Why does the manufacturer’s preview show a full sheet or an oversized board?

A4: A page border or stray item is probably expanding the preview. Check that Plot drawing sheet is off, Edge.Cuts contains one valid closed outer profile, and Edge.Cuts was not added to every layer through Plot on all layers. The preview should show the board outline, not a full page or an unexpected rectangle.

Q5: Can KiCad open or edit the Gerber files after export?

A5: Use GerbView to inspect Gerbers, not as a substitute for the KiCad source design. GerbView can export limited geometry to PCB Editor, but that conversion does not recreate the original netlist, footprints, constraints, or full design intent. Make changes in the .kicad_pcb source, then export a new manufacturing set.

Q6: Do Gerber files include the BOM and component placement data?

A6: Gerber files do not include the BOM or component placement data. They describe the board’s fabricated layer images. Automated assembly normally requires a BOM and a component placement or centroid file in addition to the PCB manufacturing package.

Q7: Should I use different KiCad settings for JLCPCB or another manufacturer?

A7: Follow the manufacturer’s current requirements when they differ from KiCad’s general defaults. The essential checks stay the same: export every required layer, generate the drill files separately, keep the origins consistent, and inspect the exact upload package. A board-house profile does not replace GerbView inspection or the portal preview.

A reliable KiCad handoff comes from four controls: a checked source board, a complete layer and drill export, a viewer inspection of the actual files, and separate notes for requirements that images cannot fully express. That sequence is more dependable than copying a settings screenshot from a different KiCad release or manufacturer.

Once your KiCad Gerber and drill files have been verified, send the ZIP together with your stackup, material, copper weight, surface finish, impedance requirements, quantity, and delivery requirements to sales@bestpcbs.com. EBest Circuit can review the manufacturing package before PCB fabrication and provide a quotation based on the actual board requirements.

Tags: Excellon Drill Files, Gerber File Verification, KiCad Gerber Export, PCB manufacturing files
Posted in PCB Knowledge | No Comments »

Flexible Printed Circuit: Materials, Types and Connections

September 22nd, 2026

A flexible printed circuit (FPC) is an electrical circuit built on a thin, bendable insulating film, usually polyimide, with patterned copper conductors. It can connect boards, carry components, or follow a three-dimensional enclosure where a rigid PCB will not fit.

Flexibility does not mean unlimited bending. A circuit folded during installation and one moving inside a hinge need different constructions. The materials, bend zone and connector ends must be designed together.

Amber flexible printed circuit with copper routing, a smooth bend and gold contact fingers

How Does a Flexible Printed Circuit Work?

An FPC carries power and signals through copper tracks just as a rigid PCB does; its thin film substrate lets those tracks follow a controlled bend. The copper is patterned into separate conductors, while insulating layers prevent contact between adjacent circuits.

A flexible printed circuit board can include pads, vias and mounted components, not just parallel wires. In a camera module, for example, one custom-shaped flex can route signals from the sensor board around a mechanical obstruction to the main board. Component locations remain supported while the connecting section bends.

What Are the Main Types of Flex Circuits?

The main constructions are single-sided, double-sided and multilayer flex; rigid-flex combines flexible interconnect sections with integrated rigid circuit sections.

Construction	Copper structure	Typical reason to choose it
Single-sided flex	One conductive layer	Simple routing with a thin bend region
Double-sided flex	Two conductive layers, usually joined by plated holes	More routing paths or a reference layer
Multilayer flex	Three or more conductive layers	Dense routing where a thicker stack is acceptable
Rigid-flex	Rigid circuit sections integrated with flex layers	Rigid component areas connected without separate cable connectors

Adding layers increases routing space but also changes bending stiffness. A multilayer design that fits a stationary enclosure is not automatically suitable for repeated movement.

Which Flexible Printed Circuit Board Material Should You Choose?

Polyimide with copper foil is a common choice for soldered FPC assemblies; the copper type, adhesive system and protective coverlay must then match the temperature and bending requirements.

Base film: polyimide provides electrical insulation and tolerates the thermal processing used for many assembled flex circuits. Polyester is an alternative for suitable lower-temperature constructions, not an automatic replacement in a reflow-soldered design.
Copper: rolled-annealed copper is commonly evaluated for repeated flexing. Electrodeposited copper is also used, but foil grade and fatigue performance matter more than the abbreviation alone.
Laminate: adhesive-based constructions bond copper to film with an adhesive layer. Adhesiveless laminates remove that separate bonding layer from the copper-to-polyimide interface, helping reduce stack thickness.
Coverlay: a protective film and adhesive cover the tracks, with openings at solder pads and contacts.

For a concrete coverlay example, DuPont Pyralux LF7001 combines 13 µm adhesive with 13 µm polyimide, while LF0110 lists 25 µm for each. These are individual coverlay constructions, not finished FPC thicknesses.

Exploded single-sided flex stack showing coverlay, coverlay adhesive, patterned copper and polyimide base

How Is an FPC Different from an FFC or a Rigid PCB?

An FPC has a custom circuit pattern, an FFC typically has parallel flat conductors, and a rigid PCB uses a substrate intended to stay rigid.

Feature	FPC	FFC	Rigid PCB
Routing	Custom tracks, branches and pads	Usually parallel conductors	Custom tracks, planes and vias
Form	Custom outline with designed bend zones	Usually a flat ribbon	Fixed board shape
Component mounting	Possible on supported areas	Usually used as an interconnect cable	Standard component platform

An FFC can be the simpler option for a straight connection between compatible sockets. Choose FPC when the circuit needs branching, an unusual outline, components or controlled routing. A flex board with a bonded stiffener is still not the same as an electrically integrated rigid-flex board.

How Do Static and Dynamic Bending Change the Design?

Static, or flex-to-install, circuits are bent into position and normally remain there; dynamic circuits must survive repeated movement without conductor fatigue.

For installation-only routing, define the formed shape, bend radius and assembly sequence. For dynamic routing, also specify the travel, cycle requirement, speed and operating environment. A hinge that passes one assembly bend has not demonstrated its service life.

Keep components, solder joints and plated holes outside the working bend zone.
Route conductors across the bend line rather than along it, and avoid abrupt changes in width.
Use smooth curves instead of creases; keep the moving section clear of sharp enclosure edges.
Evaluate the thinnest practical stack and suitable copper foil before adding layers or shielding.

Comparison of a fixed installation bend and a rolling flex loop for repeated motion

How Is the Minimum Bend Radius Determined?

The minimum bend radius comes from the complete flex stack and required bend life, not from the polyimide thickness alone. Copper, adhesive and coverlay all contribute to the thickness and strain of the bend region.

For example, our rigid-flex DFM guide gives 6× composite thickness for single- and double-sided flex sections and 12× for sections with three or more copper layers. Under that guide’s construction assumptions, a 0.10 mm section at 6× gives a 0.60 mm radius. These are design guidelines, not a guarantee for every FPC or dynamic application.

Record the approved radius on the drawing and validate the actual construction at the required motion and temperature. Changing the copper thickness, shielding or coverlay after approval can invalidate the earlier bend assessment.

How Do You Match a Flexible Printed Circuit Connector?

Match the connector’s pitch, position count, contact side and specified FPC mating thickness before finalizing the tail drawing. Matching the number of contacts alone is not enough.

For example, Hirose FH12 is an FPC/FFC connector family offering 0.5 mm and 1 mm pitches. The exact part drawing determines the contact orientation and acceptable tail dimensions; the family name is not a complete interface specification.

Contact geometry: confirm finger width, pitch, exposed length and edge-to-contact dimensions.
Mating thickness: include the circuit, bonded stiffener and adhesive at the insertion area.
Contact side: verify whether the socket contacts the upper or lower face of the inserted tail.
Retention: allow access to the latch and keep cable pull or bending loads away from the connection.

A local PCB stiffener can support the tail and establish its mating thickness. Its edge should not force bending directly beside exposed contacts or solder joints.

FPC tail, local stiffener and open ZIF connector with total mating thickness indicated

Flexible Printed Circuit Manufacturing Process

The flexible printed circuit manufacturing process patterns copper on a flexible laminate, forms any required interlayer connections, adds insulation and reinforcement, and tests the finished circuit.

Prepare the laminate: select the specified copper-clad film and prepare the panel for imaging.
Form the circuit: image and etch the copper tracks; drill and plate interconnections where the construction requires them.
Protect the conductors: align and laminate coverlay with openings for pads and contacts.
Complete exposed surfaces: apply the specified pad or contact finish and bond local stiffeners.
Profile and test: cut the outline, inspect dimensions and check electrical continuity and isolation.

The exact sequence changes with layer count, via structure and finish. The etched FPC process explains copper pattern formation in more detail. Coverlay registration and adhesive flow deserve particular attention because a partly covered contact may not mate or solder correctly.

What Changes When Components Are Assembled on Flex?

Flex assembly needs support beneath the circuit during printing, placement and soldering so the thin panel stays flat and the joints remain unloaded.

Carrier or fixture: support the panel without obstructing pads or distorting the intended outline.
Moisture control: follow the laminate and assembly process requirements before thermal exposure; do not apply one universal baking recipe.
Reflow profile: qualify the temperature profile against the solder paste, components and complete flex material system.
Inspection: check solder joints and pad alignment, then use electrical or functional testing appropriate to the assembly.

After soldering, handle the assembly by supported areas. Pulling a loose tail to lift the board can transfer force into small pads or joints even when the soldering itself was acceptable.

Where Are Flexible Printed Circuits Most Useful?

FPCs are useful where a circuit must fit a narrow three-dimensional space, connect moving sections or reduce separate wire connections.

Cameras and displays: connect modules arranged on different planes.
Wearable devices: follow compact enclosure shapes and connect small component islands.
Medical instruments: route signals through constrained assemblies with application-specific material and reliability requirements.
Battery monitoring: distribute sensing connections across a cell arrangement without treating thin sensing traces as the main power bus.
Moving mechanisms: use a qualified dynamic flex section where the motion profile suits the construction.

For a flat assembly with ample space and no movement, a rigid PCB may be easier and less expensive. FPC is valuable when its geometry solves a real packaging or interconnection problem.

FAQ About Flexible Printed Circuits

Can an FPC be stretched?

Not like an elastic band. Conventional polyimide-and-copper flex is designed to bend, not stretch freely. Stretchable electronics require different materials or conductor geometries.

Can a torn flex circuit be repaired?

Some accessible traces can be repaired in controlled rework, but a repair changes local stiffness and may not survive repeated bending. Replacement is usually more appropriate for damaged moving sections or contacts that must fit a socket precisely.

Can flex carry high-speed signals?

Yes, with a designed transmission-line structure. Specify the impedance target, reference conductors, dielectric thickness and connector transition; bending and any change from solid to hatched shielding should be included in the assessment.

Is every amber flex circuit made from the same material?

No. Color does not identify the laminate grade, copper type or adhesive. The material specification and stackup are the reliable references.

What makes a custom FPC expensive?

More layers, complex outlines, poor panel utilization, fine features, multiple stiffeners and special materials can add cost. Compare quotations using the same construction and test requirements rather than board area alone.

How Can EBest Circuit Support Your Flex Circuit Project?

At EBest Circuit, we support flexible PCB and rigid-flex projects from prototype fabrication through production and assembly. Our PCB and PCBA services date back to 2006, and we can review the circuit, component sourcing and assembly requirements together.

For your flexible circuit project, send the Gerber or ODB++ data, stackup, bend drawing, connector part number and quantity to sales@bestpcbs.com. If assembly is required, include the BOM and placement data. We will review the material construction, reinforcement and manufacturing requirements before confirming a quotation.

Tags: flexible printed circuit, flexible printed circuit board, flexible printed circuit board material, flexible printed circuit connector, flexible printed circuit manufacturing process
Posted in best pcb, FPC, PCB Design | No Comments »

IC Packaging: Structure, Materials and Assembly Process

September 22nd, 2026

IC packaging turns a fragile semiconductor die into a component that can be handled, electrically connected and mounted in an electronic product. The package provides external terminals, protects the die and creates paths for heat to leave it. It is the chip’s physical enclosure and interconnect structure—not its shipping tray or moisture-barrier bag.

A molded QFN, a wire-bonded BGA and a flip-chip processor package achieve these tasks differently. Understanding what is inside each structure helps explain its electrical behavior, thermal requirements and limits during PCB assembly.

At EBest Circuit, we support the board-level stage through our PCB assembly services, including component sourcing, incoming inspection, assembly and project-defined testing. Our engineering support connects package selection with PCB fabrication and assembly requirements; semiconductor die packaging is a separate manufacturing process.

IC packaging cutaway showing a silicon die, bond wires and protective mold compound

Key Takeaways

IC packaging protects a semiconductor die, provides external electrical connections and creates heat-transfer paths. It is different from a shipping tray or moisture-barrier bag.
Package names describe different structural features. BGA identifies the external ball array; flip chip identifies the internal die connection. A BGA is not necessarily flip chip.
Wire bonding uses fine wires from a face-up die; flip chip uses bumps beneath a face-down die. Connection density, parasitics and mechanical requirements influence the choice.
Common materials include copper leadframes, epoxy molding compounds and organic routing substrates. Not every package contains a multilayer substrate.
Packaging combines die attachment, electrical interconnection, protection and testing. Multi-die structures add integration density but also complicate thermal management and test access.
An exposed-pad QFN transfers heat into PCB copper through its die paddle and solder joint. The pad’s electrical assignment and board layout must follow the device documentation.
Board assembly needs package-specific footprint, stencil and inspection decisions. SPI checks printed paste, AOI checks visible features, and X-ray examines hidden solder connections; electrical testing checks operation.
Moisture handling and reflow limits are component-specific. Use the exact part’s moisture sensitivity label, package drawing and assembly instructions. EBest Circuit supports PCB fabrication and PCBA, not semiconductor die packaging.

What Is Inside an IC Package?

A typical molded IC package contains a semiconductor die, a die attachment layer, electrical interconnects, a leadframe or package substrate, and protective molding compound.

Die: the piece of semiconductor containing the active circuit.
Die attachment: an adhesive, solder or other qualified bonding material that secures the die to its support. The required electrical and thermal properties depend on the device.
Internal interconnects: wires, bumps or other structures that connect die pads to the package routing.
Support and routing: a metal leadframe or multilayer substrate carries connections toward the external terminals.
Protection: molding compound, a lid or another enclosure protects vulnerable structures.

A leadframe package does not need the same internal routing stack as a substrate-based BGA. A package substrate, in turn, is not the motherboard: it redistributes fine die connections to the package’s external connections. Our explanation of IC substrates and how they differ from PCBs covers that distinction.

IC Packaging Types

IC packaging types can be grouped by external terminal format, internal connection method or integration architecture. These classifications overlap: “BGA” describes the external ball array, while “flip chip” describes how the die connects inside the package.

Classification	Examples	What It Describes
External terminals	DIP, QFP, QFN, BGA	How the finished component connects to the board
Die connection	Wire bonding, flip chip	How die pads connect to package conductors
Integration architecture	Single-die package, SiP, stacked-die package	How multiple functions or dies are arranged

For example, a QFN commonly uses a copper leadframe with bottom-side lands and may include an exposed thermal pad. A flip-chip BGA instead connects the die through bumps to a routing substrate, with a separate ball array underneath for board attachment. For footprint and assembly comparisons, see our IC package types guide.

Which IC Packaging Materials Are Used?

Common IC packaging materials include copper alloys for leadframes, epoxy molding compounds for encapsulation, organic laminates for substrates, and metals such as copper, gold or aluminum for bond wires.

Material	Package Location	Primary Function
Copper alloy	Leadframe and die paddle	Electrical conduction and mechanical support
Filled epoxy molding compound	Molded body	Encapsulation and electrical insulation
Organic laminate and copper routing	Package substrate	Signal and power redistribution
Bond-wire metals	Die-to-terminal connections	Electrical interconnection
Underfill resin	Gap beneath a bumped die	Mechanical reinforcement of interconnects
Ceramic and metal lids	Selected cavity or hermetic packages	Structural support and environmental protection

Epoxy molding compound is a common IC packaging plastic material, but it is an engineered composite rather than ordinary unfilled plastic. Resin chemistry and fillers affect flow, expansion, moisture behavior and reliability. Sumitomo Bakelite’s EME encapsulants, for example, are specifically developed for semiconductor packaging.

Materials must work as a system. A low-expansion encapsulant alone cannot eliminate stress if its adhesion, curing behavior or compatibility with the die and substrate is unsuitable. The package manufacturer qualifies the complete material combination, not just one favorable property.

How Do Wire Bonding and Flip Chip Differ?

Wire bonding connects a face-up die to package terminals with fine wires; flip chip connects a face-down die through bumps directly to matching pads on a substrate or other receiving structure.

Wire bonding and flip chip compared by die orientation and internal connection method

Wire bonding accommodates many established analog, power-management and logic packages. Its wire loops introduce electrical parasitics, and the bond-pad arrangement and loop geometry constrain the layout. Wire material and bonding settings must be compatible with the die metallization.

Flip chip supports connections across the die surface rather than only along an accessible perimeter. Its shorter interconnects can reduce inductance and support dense signal and power connections.

The trade-off is additional control of bump formation, alignment, substrate routing and mechanical stress. Underfill is used in many flip-chip structures to reinforce the connection region; the exact material and application sequence depend on the package. Neither method is universally better for every IC.

IC Packaging Process

The IC packaging process typically prepares and separates wafer dies, attaches each die, forms electrical connections, protects the assembly, and tests the finished devices. A molded wire-bonded leadframe package follows the example below.

Four simplified IC packaging stages: die attach, wire bond, mold, and singulate and test

Prepare the dies: wafer-level inspection and electrical probing identify die performance before packaging; wafer preparation and dicing separate individual dies.
Attach the die: place it on the designated paddle or support using the qualified attachment process.
Form interconnects: bond wires between die pads and separate leadframe terminals.
Encapsulate: mold the body around the die and wires, then complete the required cure and finishing operations.
Separate and finish: singulate individual packages; terminal finishing or lead forming applies where the package design requires it.
Test and pack: electrically test, inspect, mark and prepare accepted devices for shipment.

The illustrations simplify these stages and are not tooling drawings. Flip-chip, cavity and wafer-level packages use different sequences; a ceramic cavity package, for instance, may require lid sealing rather than plastic molding. Process order is defined by the particular package flow.

Why Are Multiple Dies Combined in One Package?

Multiple dies are combined to integrate functions, shorten communication paths or place different semiconductor technologies in one component or module.

A system-in-package (SiP) may combine logic, memory, radio-frequency devices and passive components. It can use several interconnect and assembly technologies rather than a single universal construction.

Side-by-side integration: dies occupy neighboring locations on a common routing structure.
Stacked-die integration: dies sit above one another; their connections may use wires or other vertical interconnect structures.
Interposer-based integration: an additional fine-routing structure connects dies before signals reach the package substrate.

These arrangements increase the importance of thermal interaction, interconnect yield and testing access. A failed die or connection can affect the completed assembly. Our CoWoS-S packaging article examines one interposer-based architecture in more detail.

How Does Heat Leave an IC Package?

Heat leaves an IC through paths into the PCB, the package surface and any attached cooling hardware. In an exposed-pad QFN, an important path runs from the die through its attachment and paddle, through solder, and into PCB copper.

Simplified QFN thermal path from die through exposed pad and solder into PCB copper and thermal vias

The exposed pad needs the land pattern, solder connection and copper arrangement specified for that component. Thermal vias can connect the top land to additional copper layers. Their dimensions, filling or tenting treatment and stencil layout must also account for solder loss into holes and assembly quality.

Not every exposed pad is an interchangeable ground connection. Its electrical assignment comes from the device datasheet. Likewise, a thermal resistance value is meaningful only with its stated board and test conditions; it is not a fixed prediction of temperature on every PCB.

For a lidded processor package, heat transfer through a thermal interface material and heat spreader may be central to the cooling design. That is a different assembly from the QFN example and should not inherit its thermal-pad rules.

How Are IC Packages Tested?

IC packages undergo electrical testing to confirm device operation, while inspection and reliability qualification address assembly defects and resistance to specified stresses.

Electrical testing: checks functions and relevant parameters after assembly, using the device’s test program.
Visual and dimensional inspection: checks body condition, markings, terminal geometry and other specified characteristics.
Internal inspection: X-ray or acoustic methods may investigate hidden connections, voids or delamination, according to the inspection plan.
Reliability qualification: evaluates defined stresses such as temperature cycling or humidity exposure on the applicable qualification samples.

These checks answer different questions. Passing an electrical test does not prove that a package has no internal structural defect, and a clear X-ray image does not establish full device functionality. Burn-in and system-level testing are product-dependent, not mandatory stages for every IC.

What Can Damage an IC Package During PCB Assembly?

Moisture exposure followed by reflow, excessive temperature, electrostatic discharge and mechanical stress can damage an IC package or its internal connections.

Absorbed moisture is especially important for moisture-sensitive surface-mount components: rapid heating can generate internal pressure and contribute to cracking or delamination. The moisture sensitivity level, permitted floor life and peak package-body temperature must be taken from the component’s label and applicable handling instructions.

Before placement: confirm the exact ordering code, package drawing, moisture status and storage history.
During reflow: use a validated profile compatible with the component, solder paste and board; oven settings alone do not show the component’s actual temperature.
During handling and rework: control ESD, avoid excessive board bending, and follow the specified limits on heating and mechanical loading.

Baking is not a universal remedy to apply at an arbitrary temperature. Follow the approved recovery procedure when floor life or storage conditions have been exceeded.

What Should Be Checked Before Mounting an IC Package?

Check the exact package drawing, PCB land pattern, stencil apertures, orientation and inspection plan before releasing an IC for board assembly. A package-family name or pin count alone is not enough to approve the footprint.

Footprint: compare terminal pitch, pad dimensions, exposed-pad geometry and pin-1 location with the manufacturer’s drawing. A mismatch can produce open joints or incorrect connections even when the body fits.
Paste printing: evaluate stencil thickness and aperture geometry together. For a rectangular aperture, area ratio is L × W / [2t(L + W)], where t is stencil thickness. Reducing thickness increases this ratio but reduces theoretical paste volume; neither choice should be made from pitch alone.
Exposed pads: review aperture segmentation and thermal-via treatment to limit excess paste and solder loss into open vias. Do not apply one void-percentage limit to every QFN or power device.
Inspection: use SPI to evaluate paste height, area and volume before placement. After reflow, use AOI for accessible joints and orientation, and X-ray for hidden BGA or QFN connections. Follow with the agreed electrical or functional tests.

Our SMT stencil service supports the paste-printing stage. Send the package drawing with the PCB and assembly files so aperture and thickness choices can be reviewed together, rather than copied from a different component.

At EBest Circuit, our PCBA process includes SPI, AOI and X-ray inspection. The project’s component geometry and acceptance requirements determine the inspection plan; an X-ray image alone does not prove that a populated board functions correctly.

FAQ About IC Packaging

Is IC packaging the same as PCB assembly?
No. IC packaging builds the component around semiconductor dies. PCB assembly mounts packaged components and other parts onto a circuit board.

Does every IC package contain a substrate?
No. Many molded packages use a metal leadframe instead of a multilayer organic package substrate.

Are all BGA packages flip chip?
No. BGA describes the external solder-ball array. The die inside can use wire bonding, flip chip or another qualified interconnect arrangement.

Are plastic IC packages waterproof?
Ordinary molded plastic packages should not be treated as hermetic enclosures. Environmental suitability depends on the complete package qualification and the product’s protection measures.

Does a 3D package always use TSVs?
No. Dies can be stacked and connected with bond wires. Through-silicon vias are one possible vertical interconnect technology, not a requirement for every stacked-die package.

How Can EBest Circuit Support Your PCB Assembly?

A well-chosen IC package still needs a compatible footprint, soldering process and inspection plan. At EBest Circuit, we bring PCB fabrication, component sourcing and assembly support together so these requirements can be reviewed before your build.

Send your BOM with exact manufacturer part numbers, Gerber files, assembly drawings, quantities and test requirements to sales@bestpcbs.com. We can review your project’s PCB and assembly requirements, identify missing package information, and discuss the next steps for a quotation.

Tags: IC packaging, ic packaging materials, ic packaging plastic material, ic packaging process, ic packaging types
Posted in best pcb, PCB Technology, PCBA | No Comments »

Ethernet PCB Manufacturing and Assembly for Network Hardware

September 21st, 2026

An Ethernet PCB that links reliably on a laboratory bench can still become unstable in production if the fabricated stackup, connector assembly, or test method differs from the released design. Link speed, PHY interface, isolation magnetics, connector style, PoE requirements, and the enclosure all affect how the board must be laid out, fabricated, assembled, and verified.

EBest Circuit turns customer-approved schematics and component requirements into manufacturable layouts, controlled-impedance PCBs, and assembled network boards. With in-house PCB and PCBA production, 20 years of PCBA experience, and a supply network of more than 1,000 partners, we can coordinate fabrication, component sourcing, SMT and through-hole assembly, and agreed testing under one production route. To discuss an Ethernet hardware build, contact sales@bestpcbs.com.

How Link Speed Changes Ethernet PCB Requirements

Higher Ethernet speeds leave less margin for insertion loss, crosstalk, discontinuities, and timing variation. The target speed must therefore be fixed before the stackup, magnetics, connector, and test method are released.

Link speed	PCB impact	Build priority
10 / 100M	Two active MDI pairs	Correct pair geometry and matched interface parts
1G	Four bidirectional pairs	Four-pair consistency and 1G link testing
2.5G-10G	Tighter loss, crosstalk, and thermal limits	Material and stackup matched to the PHY budget

The cable-side MDI pairs are not the same as the MAC-to-PHY interface. MII, RMII, RGMII, SGMII, and other system-side interfaces use different electrical and timing rules. For example, RGMII timing depends on the selected MAC, PHY, and internal-delay settings; it should not receive a generic length-compensation rule copied from another design.

PoE adds another layer. The data rate may be correct while the board still has inadequate current capacity, magnetics rating, isolation, or thermal performance. PSE or PD role, PoE type, supply path, copper area, and expected temperature must be defined with the approved circuit.

Where Ethernet PCBs Are Used in Industrial and Embedded Equipment

Ethernet PCBs are used wherever equipment needs a stable wired link, but the board requirements change with the environment and the job performed by the port.

Industrial controllers and gateways: PLC interfaces, HMIs, remote I/O, and edge gateways may combine Ethernet with noisy power stages, long field cables, and metal enclosures. Grounding, isolation, surge protection, and connector retention become as important as the PHY itself.
Machine vision and IP cameras: A compact board may need high data throughput and PoE in the same interface. Connector position, heat around the PHY and power stage, and repeatable link testing are common production concerns.
Embedded computers and test equipment: These boards often combine Ethernet with processors, memory, USB, wireless modules, and sensitive analog sections. Placement and return-current planning must prevent one interface from disturbing another.
Switches, access points, and network appliances: Multiple ports increase pair density, power demand, connector alignment, and test coverage. The production fixture and test plan must reflect the actual port count and target speed.

Not every industrial Ethernet port uses an RJ45. Sealed M12 interfaces and single-pair Ethernet use different pin assignments, coupling arrangements, and mechanical constraints. Automotive 100BASE-T1 or 1000BASE-T1, for example, should not inherit a four-pair RJ45 layout merely because both are called Ethernet.

Ethernet PCB Layout Support Based on Customer-Approved Schematics

EBest Circuit can support PCB layout after the customer has approved the network architecture, schematic, PHY, magnetics, connector, and functional requirements. Our role is to translate those decisions into a board that can be fabricated and assembled consistently, not to replace the customer's circuit or system design authority.

Physical signal path: Place the PHY, discrete magnetics or magjack, protection parts, and connector so the MDI pairs remain short, balanced, and free of unnecessary stubs and vias.
Interface-specific routing: Apply the selected PHY's data sheet and reference design to the MDI path, while treating MII, RMII, RGMII, or serial MAC-side interfaces according to their own timing and impedance requirements. Detailed Ethernet PCB routing rules are reviewed against the actual device rather than copied as universal numbers.
Reference and isolation structure: Coordinate reference planes, isolation gaps, chassis or shield connections, and any keepout beneath magnetics with the approved safety and EMC concept.
Support circuitry: Keep the oscillator or crystal, bias resistor, decoupling network, termination, and ESD parts in the locations required by the selected components.
Manufacturing handoff: Align the net classes and critical geometry with a producible stackup, then return material or geometry changes for approval before CAM release.

The customer remains responsible for PHY and MAC selection, circuit function, firmware, system compliance, and final product validation. This boundary keeps layout support practical without allowing production changes to alter the approved design intent.

How Stackup, Materials, and Controlled Impedance Shape Ethernet PCB Manufacturing

Controlled impedance is a property of the finished stackup, not a trace-width value in isolation. Dielectric thickness and Dk, finished copper, line width and spacing, reference planes, solder mask, and etching compensation work together to produce the result.

For an Ethernet production build, EBest Circuit can support multilayer rigid PCB fabrication from 1 to 32 layers, controlled impedance, and standard or high-Tg FR-4 options; HDI or other special structures are evaluated against the released design. FR-4 is not automatically unsuitable for a faster link, but the selected material and geometry must satisfy the PHY's actual channel-loss budget.

The manufacturing sequence should remain controlled:

the customer defines the target impedance and tolerance for each applicable net class;
EBest Circuit proposes a manufacturable stackup and calculates the corresponding geometry;
the approved material family, dielectric, copper, line width, spacing, via structure, and reference planes are locked for production; and
an impedance coupon and TDR record can be supplied when included in the order requirements.

Our impedance control PCB process can verify representative fabricated geometry, but a coupon does not prove the performance of the complete assembled Ethernet channel. Magnetics, connector transitions, soldering, firmware, cable, and link partner still require the appropriate assembly or functional test.

Ethernet PCB Assembly for PHYs, Magnetics, and Network Connectors

Ethernet PCB assembly must hold both the electronic and mechanical interfaces to the released design. A board can have correct copper and still fail if a PHY has hidden-joint defects, a connector sits unevenly, or an unapproved magnetics substitute changes bandwidth, isolation, pinout, or PoE current capacity.

BOM and revision control: Source the exact approved PHY, oscillator, magnetics or magjack, connector, ESD device, bias and termination parts, and PoE components. Alternatives require engineering approval; matching only the package or nominal function is not enough.
Fine-pitch assembly: Use an appropriate stencil and reflow profile for QFN, BGA, LGA, or other bottom-terminated PHY packages. SPI, AOI, and X-ray are applied where they match the package and inspection plan.
Network connector assembly: RJ45, M12, and other ports may use SMT, through-hole, pin-in-paste, or mixed processes. The Ethernet connector PCB build must control seating height, board-edge position, shield tabs, through-hole fill, and the mechanical load transferred from the cable.
Integrated or discrete magnetics: A magjack already contains the magnetics, while other designs place a separate transformer between the PHY and connector. The assembly route and inspection points must follow the actual architecture rather than assume both parts are present.

A typical build route is approved BOM verification, solder-paste inspection, SMT placement and reflow, AOI or X-ray as required, through-hole connector assembly, cleaning, and customer-defined testing. PoE boards also need process control for higher-current paths, power devices, isolation areas, and local heat.

Ethernet PCB Inspection and Testing for Bare Boards and PCBAs

No single test proves that an Ethernet board is ready for the end product. Each stage answers a different question, so bare-board, assembly, and link evidence must not be treated as interchangeable.

Test process	Main fault found	Production role
Bare-board electrical test	Opens and shorts	Screens PCB connectivity before assembly
Coupon / TDR	Impedance deviation	Monitors stackup and trace geometry
SPI / AOI / X-ray	Paste, placement, and solder defects	Controls assembly workmanship
Defined link test	Power-up, negotiation, and packet faults	Exercises the agreed PCBA functions

These methods address different production layers: electrical testing and TDR cover the fabricated PCB, inspection controls assembly workmanship, and link testing exercises the agreed PCBA functions. EBest Circuit can run customer-defined link tests using approved firmware, fixtures, link partners, and pass/fail limits. Production testing does not replace EMC, safety, PoE, or IEEE compliance validation.

How EBest Circuit Supports Ethernet PCB Prototypes and Repeat Production

EBest Circuit uses prototype builds to establish a manufacturing baseline, then carries the approved data and process into repeat production.

Prototype build: DFM and CAM data, stackup, impedance classes, BOM, first-article inspection, and the test method are aligned before release.
Repeat orders: Gerber or ODB++, material, stackup, approved alternatives, placement data, firmware, and test revisions remain tied to the order.
Traceability: The digital shopfloor can locate material and product-batch records within five seconds when an engineer needs to investigate a component, revision, or lot.
In-house execution: PCB fabrication, sourcing, mixed SMT and through-hole assembly, and final inspection follow one coordinated production route. Prototype and small-batch builds are supported.
Quality systems: Company certifications include ISO 9001, ISO 13485, IATF 16949, and AS9100D for applicable regulated programs.

FAQs About Ethernet PCB

Does every Ethernet PCB use 100-ohm differential impedance?

Most cable-side copper Ethernet MDI pairs use the differential impedance specified by the PHY and magnetics interface, commonly 100 ohms. MAC-side interfaces, single-pair Ethernet, and vendor-specific structures may follow different requirements, so the selected device data and approved stackup remain the authority.

Can standard FR-4 support a Gigabit Ethernet PCB?

Yes, standard FR-4 is suitable for many 1G boards when the route length, stackup, geometry, and connector path stay within the PHY channel budget. Faster links or longer on-board channels may require tighter loss control or a lower-loss material.

Should the Ethernet magnetics be integrated into the RJ45 or placed separately?

Both architectures are valid. A magjack can shorten the interface path and reduce the part count, while discrete magnetics can provide more flexibility for connector choice, placement, isolation, and PoE implementation. The PCB and assembly route should follow the approved architecture and BOM.

Why can a PCB pass bare-board electrical testing but still fail Ethernet link testing?

Bare-board electrical testing finds opens and shorts; it does not exercise the assembled channel. Link failure can still come from impedance deviation, solder defects, an incorrect component, clock or power problems, PHY configuration, firmware, the connector, or the cable.

Can EBest Circuit perform Ethernet link testing before shipment?

Yes, when the firmware, fixture, link partner, target speed, active ports, and pass/fail limits are agreed before production. The test can cover power-up, link negotiation, and defined packet functions within the approved test scope.

When your schematic and interface architecture are approved, EBest Circuit can coordinate controlled-impedance PCB fabrication, sourcing, assembly, and the agreed link test. Send the released files, target quantity, and test requirements to sales@bestpcbs.com, or arrange an on-site factory audit before first articles or repeat production.

Tags: Ethernet PCB, Ethernet PCB Assembly, Ethernet PCB Layout
Posted in best pcb, bestpcb, RF & Telecom PCB, Solutions | No Comments »

Star Grounding in PCB Design: When to Use or Avoid It

September 21st, 2026

Star grounding is a targeted fix for shared-return noise, not a default PCB layout style. It helps when a known high-current or noisy load would otherwise share return copper with a sensitive circuit, and the two branches can meet at one short, controlled junction. Fast digital, RF, and widely distributed currents usually need a continuous ground plane instead.

A sound choice starts with the actual current loops: identify the disturbing current, the reference that must stay quiet, the fastest edge, and every path that can reconnect the grounds. Those four facts lead to a practical decision between a star, a plane, or a hybrid layout and define what must be measured on the prototype.

Star grounding topology with four PCB return branches meeting at one controlled point

What Is Star Grounding, and What Problem Does It Solve?

Star grounding gives selected circuits separate return branches that meet at one controlled junction. The aim is to stop current from one branch flowing through the reference used by another branch before both currents reach their common source.

The problem is common-impedance coupling. Copper, vias, connectors, and solder joints all have impedance. When two circuits share a return segment, the current from the first circuit creates a voltage across that segment:

Reference error = shared return current × shared-path impedance

Consider a 24 V control board with a solenoid and a sensor amplifier. If the solenoid and amplifier share return copper, each solenoid current step can shift the amplifier reference and appear as a false sensor signal. Separate branches can keep the solenoid current out of the measurement path until both returns reach the supply reference.

The topology is useful only when the designer can name both sides of that relationship: the current that causes the error and the reference it disturbs. A radial drawing without that explanation is not a design decision.

How Do Frequency and Return Paths Affect Star Grounding?

Low-frequency load currents may tolerate separate branches, while fast-edge currents need a short return close to the outgoing signal. This is why a star can reduce sensor-reference error yet make a clock, data bus, or switching node worse.

Nominal operating frequency is not enough. A slowly updated bus may still switch in nanoseconds, and a low-sample-rate converter may contain fast clocks. Long branches add inductance and loop area, which can increase ringing, crosstalk, emissions, and susceptibility even when their DC resistance looks acceptable.

Check three things for every important signal or load:

The complete loop: source, outgoing conductor, load, and return conductor.
The fastest transition: driver rise and fall time, switch-node edge, converter clock, or ESD event.
Return continuity: plane splits, narrow necks, missing return vias, and connector boundaries that force a detour.

If a proposed branch sends a fast return across the board before it can close, use a nearby plane for that loop. The same PCB can still use a dedicated branch for a compact low-frequency load return.

When Does Star Grounding Work Well?

Star grounding works well when one identifiable load threatens one identifiable reference and both branches can reach a compact junction. The following conditions should all be true:

The disturbing current is known. Typical sources include a relay, motor, solenoid, heater, lamp, or power-output stage.
The sensitive reference is known. It may belong to a sensor, precision reference, measurement shunt, low-level audio input, or feedback network.
The junction is physically close. A supply return, bulk-capacitor return, regulator return, or connector boundary can accept the branches without long radial routing.
Fast local loops stay local. Decoupling, clocks, interfaces, and switching loops still close through a nearby plane or tightly coupled conductor pair.

On the control-board example, the solenoid can return directly to the power-entry capacitor while the sensor branch reaches the same reference on separate copper. The ADC and microcontroller interface can remain over continuous ground. This is a hybrid solution driven by current paths, not by the labels “analog” and “digital.”

When Should You Avoid Star Grounding and Use a Ground Plane Instead?

Avoid a whole-board star when fast signals, RF energy, dense routing, or distributed loads need many short return paths. Start with a continuous ground plane for most multilayer digital boards, then add a controlled junction only where a specific noisy branch must be kept out of a sensitive region.

Engineering comparison of star, continuous ground plane, and hybrid PCB return topologies

Board condition	Starting topology	What to verify
One compact low-frequency load beside a sensitive reference	Star or Kelvin return	Measure reference movement during the worst load step.
Clocks, fast buses, RF traces, or rapid converter interfaces	Continuous ground plane	Keep an adjacent return path beneath the complete signal route.
Fast local interfaces beside a noisy power section	Hybrid	Keep fast returns on the plane and control where the power return joins it.
Loads spread across a large board	Plane or region-based power return	Calculate voltage differences using the actual routing length and current.
Several boards, cables, shields, or conductive mounts	System-level grounding plan	Trace every connection that can create a parallel return path.

A layout should be judged by return-current behavior rather than visual symmetry. A neat radial pattern can create a large high-frequency loop, while careful placement over a solid plane can keep a noisy load current away from a measurement circuit.

Where Should the Star Ground Point Be Placed?

Place the star point where the selected branch currents return to their shared source without crossing another branch’s reference first. That point is often near a bulk capacitor, regulator return, supply connector, or measurement shunt, not at the geometric center of the PCB.

Engineering comparison of incorrect shared return routing and a controlled star point near the supply return

Mark the sources and loads. Include connectors, regulators, bulk capacitors, switching stages, sensors, converters, and external cables.
Draw outgoing and return current together. Check steady load, the worst load step, and the fastest transition because the paths may differ.
Locate the protected reference. Identify the exact point where the sensor, amplifier, ADC reference, or feedback network measures voltage.
Join the branches after their local reference points. The noisy current should reach the shared source without using the sensitive branch’s copper.

A net tie or zero-ohm resistor can make the chosen junction visible in the schematic and testable on the board. It cannot correct long branches or poor placement. Choose the electrical location first, then use the component or copper feature to enforce it.

How Should Analog, Digital, Power, and Chassis Grounds Meet?

Connect ground regions according to the currents they carry and the reference required by signals that cross between them. The names AGND, DGND, PGND, and chassis ground describe roles; they do not automatically require separate copper planes.

Analog return: keep load and switching current away from low-level inputs, references, and feedback nodes.
Digital return: give clocks and interfaces a short path beside their signal traces, even when average current is small.
Power return: contain converter hot loops, motor current, and gate-drive current near their sources and local capacitors.
Chassis, shield, and protective earth: route ESD, common-mode, shield, and safety current at the connector or enclosure boundary without sending it through a sensitive reference.

For a mixed-signal IC, follow the device data sheet and reference layout before splitting ground. AGND and DGND pins may describe internal circuit functions rather than a requirement for two board planes. The relevant questions are where the pin currents flow and what reference the crossing signals use.

Also trace paths outside the PCB. A cable shield, metal standoff, programming lead, or oscilloscope earth connection can bypass the intended junction. A board-level star is valid only if the assembled system preserves it.

What Layout Mistakes Make Star Grounding Fail?

Star grounding fails when the copper, components, or external connections create a different current path from the one shown on the schematic. These mistakes are the first places to look when the expected noise reduction does not appear:

The junction is beyond the sensitive reference. The noisy current crosses the measurement reference before reaching the star. Move the junction or change placement.
The branches are too long. A branch with acceptable DC resistance can still have excessive inductance. Keep fast returns on a plane.
A signal crosses a split or narrow ground neck. The return detours around the gap. Reroute the signal or restore continuous ground beneath it.
A second connection bypasses the star. Noise changes when a cable, enclosure, debugger, or instrument is attached. Trace current through the complete test setup.
Decoupling current travels to a remote junction. The IC supply loop becomes larger than necessary. Close each high-frequency decoupling loop locally.
A bead or zero-ohm link replaces analysis. The fitted part adds impedance but does not define the current path. Select it only after the target current and allowed voltage difference are known.

Change one path at a time during troubleshooting. Switch the suspected load while monitoring the protected reference, then repeat with the enclosure and external cables connected. The difference separates shared-copper error from a secondary path or radiated-coupling problem.

How Can You Review and Test a Star Ground Design?

A star-ground design is ready when the current-path drawing, physical junction, and measured result agree with one defined acceptance limit. Six checks provide that evidence:

Draw the critical loops. Mark the source, outgoing path, load, return, and local decoupling loop for each noisy, fast, or sensitive circuit.
Name the coupling. Record the aggressor current, shared path, victim reference, operating condition, and maximum acceptable error.
Compare the drawing with the PCB. Check branch length, plane continuity, layer changes, connectors, shields, mounts, and the exact junction feature.
Estimate the low-frequency error. Calculate shared resistance and voltage drop at maximum load and compare the result with the allowed reference shift.
Measure the load event. Compare idle and worst-load conditions with a short-ground-spring probe or an appropriate differential probe. Save the waveform that occurs at the same time as the disturbing event.
Repeat the test on the assembled system. Add the enclosure, shields, cables, debugger, and representative external equipment so hidden return paths are included.

The result should lead to one decision. Keep the star if the protected reference stays within its limit and fast local returns remain continuous. Use a plane or hybrid if fast or distributed currents need a nearby reference. Change the architecture if the current paths or acceptance limit are still unclear.

FAQs About Star Grounding

Q1: Should each star-ground branch use a separate schematic net name?

A1: Use separate net names when they help the CAD tools enforce the intended connection. Join them with a documented net tie or approved junction footprint, then confirm that the PCB netlist and copper contain no unintended second connection.

Q2: How wide should a star-ground branch be?

A2: Size it from current, allowable voltage drop, temperature rise, and transient behavior. There is no universal width. A high-current branch may need wide copper, while a sensing branch may need a separate Kelvin path that carries almost no load current.

Q3: Can a ground pour surround star-ground branches?

A3: Yes, but the pour must not create a parallel connection between branches. Use clearances or a controlled net-tie strategy, then inspect the final Gerber data and connectivity report rather than relying on the schematic alone.

Q4: Can several regulators share one star point?

A4: They can when their return currents meet at a low-impedance source node without modulating another regulator’s reference. If one converter has large switching or load-step current, give it a local hot loop and check whether a higher-level junction is needed.

Q5: How should star-ground noise be measured with an oscilloscope?

A5: Measure between the protected reference and its source reference during the suspected load event. Use a short ground spring or a suitable differential probe; a long probe ground lead can create a loop and show noise that is not present at the circuit node.

Conclusion: Choose the Grounding Topology from the Current Loops

Choose a star only for a specific shared-impedance problem that has short branches, a clear victim reference, and one practical convergence point. Choose a continuous plane when fast or distributed currents need nearby returns. Use a hybrid when those two needs exist on the same board.

The next design review should produce two items: an annotated drawing of the critical current loops and a measurement plan for the worst load event. If both are specific, the topology choice can be checked. If either remains vague, deciding between star grounding and a ground plane is premature.

Tags: ground plane design, mixed-signal PCB, PCB grounding, PCB return current, star grounding
Posted in PCB Knowledge | No Comments »

CoWoS-L Explained: RDL and LSI for Larger AI Packages

September 21st, 2026

CoWoS-L combines a redistribution-layer (RDL) interposer with local silicon interconnects (LSI). The RDL spans the larger routing platform; embedded silicon bridges provide dense links at selected die interfaces. This division helps integrate more logic and high-bandwidth memory without using one continuous silicon interposer across the full area.

Conceptual CoWoS-L package with embedded local silicon interconnects

What Is CoWoS-L Packaging?

CoWoS-L packaging is TSMC’s CoWoS architecture that embeds local silicon interconnects within an RDL-based interposer for large multi-die computing packages.

A CoWoS-L cross section contains logic dies and HBM at the top, an interposer incorporating RDL and local silicon beneath them, and a separate package substrate below. RDL means redistribution layer: patterned conductors redistribute connections across the platform. LSI means local silicon interconnect: dense silicon-based routing placed where neighboring die interfaces need it.

Calling this only an “organic interposer” misses the embedded silicon and molded integration structure. It is not silicon-free, and its package substrate is not the system PCB. For the foundry context, see our introduction to TSMC’s manufacturing technologies.

How Do RDL and LSI Work Together in CoWoS-L?

LSI handles dense local die-to-die connections, while RDL distributes connections over the wider interposer footprint.

Local die-to-die signal path within the silicon bridge above wider RDL routing

Structure	Location	Interconnect role
LSI	Selected neighboring die interfaces	High-density local links using fine silicon-based wiring
RDL	Across the wider interposer	Broader redistribution and integration with the embedded structures
Package substrate	Below the interposer	Connections toward the finished component’s board interface

TSMC describes CoWoS-L LSI with multiple layers of submicron copper wiring and connections including SoC-to-SoC, SoC-to-chiplet, and SoC-to-HBM. The bridge occupies the region serving those interfaces; it is not simply a small silicon support underneath an arbitrary part of the package.

Consider two logic dies whose high-density interfaces face each other. An LSI region can connect those edges while the wider RDL carries other connections. Rotating one die may move its interface away from the intended landing region, forcing changes to bridge position and routing. Die orientation and LSI placement must therefore be co-designed.

The bridge does not generate bandwidth by itself. Link width, signaling rate, electrical characteristics, and the transmitting and receiving circuits still determine usable throughput.

Why Can CoWoS-L Support Larger Packages?

CoWoS-L extends the interposer footprint through a wider RDL platform while concentrating fine silicon routing at local interfaces, instead of enlarging one continuous silicon interposer everywhere.

TSMC’s public overview identifies a 3.5-reticle CoWoS-L generation that entered production in 2024. This is a dated technology generation, not a permanent maximum. Reticle multiples describe interposer scale relative to an exposure field; they do not specify the component’s outer dimensions, BGA pitch, or HBM count.

More placement area: the platform can accommodate additional or larger logic and memory components when supported by the package design.
Localized fine wiring: high-density silicon regions follow the die interfaces rather than covering the entire footprint.
Separate size limits: interposer area, package substrate outline, and cooling assembly envelope remain different dimensions.

Larger still means more manufacturing coordination. Mold, copper, silicon, and the organic substrate respond differently to temperature. Warpage, interconnect stress, routing yield, and cooling all require qualification. Less full-area silicon does not prove that every CoWoS-L product is cheaper than a CoWoS-S alternative.

How Is a CoWoS-L Package Manufactured?

The CoWoS-L process flow integrates local silicon interconnects into a molded RDL platform, attaches the top dies, and completes the assembly on a package substrate.

Plan the floorplan: align SoC, chiplet, and HBM interface locations with the required LSI regions and reserve power-routing space.
Integrate the embedded structures: incorporate the local silicon elements, and applicable embedded passive components, into the reconstituted interposer structure.
Form the RDL connections: create the redistribution wiring that connects the embedded regions and wider package interfaces.
Attach the top dies: TSMC describes a chip-last approach in which the interposer platform is prepared before top-chip assembly.
Complete substrate integration and verification: assemble the package substrate and thermal structure and verify links, power behavior, and reliability.

CoWoS-L functional stages from floorplan through assembly and verification

Critical controls include embedded-element position, RDL registration, surface planarity, and joint formation. A displaced bridge or a local height error can compromise the fine die interface even if the overall package outline is correct.

This sequence describes functional stages, not a proprietary process recipe. Exact mold materials, bonding temperatures, tolerances, and inspection acceptance criteria must come from the qualified package process.

How Does CoWoS-L Support Power Delivery?

CoWoS-L supports power delivery through its interconnect network and the integration of stand-alone embedded deep trench capacitors (eDTCs) beneath the SoC.

Embedded eDTC beneath the SoC within the RDL interposer above a separate package substrate

TSMC identifies this eDTC integration in its official CoWoS technology overview. Placing capacitance near the load can shorten the local current loop and reduce the inductive penalty of supplying rapid current changes. The benefit depends on the actual connection geometry, capacitor characteristics, and complete power network.

Local transient support: nearby capacitance supplies part of a fast current demand before more distant supply paths respond.
Power-path coordination: package conductors, board planes, capacitors, and voltage regulators must meet the device’s supply limits together.
Impedance control: capacitance and interconnect inductance can create resonances, so adding capacitance is not automatically an improvement at every frequency.

For an illustrative target, a permitted 30 mV voltage change during a 100 A current step gives Z = ΔV/ΔI = 0.3 mΩ. These are example inputs, not a CoWoS-L rating. A real design needs its own tolerance and frequency-dependent package, board, and regulator models.

Where Is CoWoS-L Packaging Used?

CoWoS-L is used for large AI and HPC packages that need dense local connections among logic dies, chiplets, and HBM on a larger interposer platform.

Multi-die AI accelerators: local bridges connect the compute interfaces, while the wider floorplan accommodates logic and memory integration.
HBM-based computing packages: SoC-to-HBM connections combine local routing density with space for the required memory arrangement.
Chiplet-based HPC designs: dense connections between selected chiplets support integration without making every part of the routing platform silicon.

TSMC’s disclosed SoC-to-SoC, SoC-to-chiplet, and SoC-to-HBM connection forms support these application categories. They are not a claim that every chiplet processor or AI accelerator uses CoWoS-L. Naming a particular product requires a disclosed packaging variant, not an inference from its HBM count or performance.

What PCB Assembly Checks Matter for CoWoS-L Packages?

PCB assembly checks must follow the specific finished component or module’s land pattern, handling limits, reflow instructions, warpage requirements, and thermal-mechanical design.

Board interface: confirm whether the delivered item is a directly soldered component or a module with its own board/connector interface.
Land pattern and escape routing: use the released ball map and pad geometry; LSI dimensions do not become PCB trace dimensions.
Moisture handling and reflow: use the device’s specified storage, exposure, and thermal-profile limits, not a generic “CoWoS-L temperature.”
Coplanarity and support: review the component’s allowed deformation, board support, heatsink loads, and attachment keep-outs.
Inspection and testing: agree on accessible joint-inspection methods and electrical or functional tests appropriate to the assembly. Visual inspection alone cannot verify hidden BGA joints.

Our advanced HDI PCB guide covers board-level routing options. At EBest Circuit, we review PCB fabrication and PCBA requirements against the actual design and component documentation—not the interposer’s marketing dimensions. Send your Gerber files, stackup, BOM, quantities, and component assembly specifications to sales@bestpcbs.com. This board-level support is separate from TSMC’s CoWoS-L package manufacturing.

FAQs About CoWoS-L

Does CoWoS-L eliminate silicon?

No. It retains silicon in local interconnect regions. The wider RDL platform changes where silicon is used, not whether silicon is present.

Is LSI another processor?

No. In CoWoS-L, LSI is a local silicon interconnect structure that routes signals between die interfaces; it is not an additional computing die.

Is one LSI bridge enough for every package?

Not necessarily. The number and location of bridges follow the interfaces being connected. Multiple logic dies and HBM stacks can require several local regions.

Does an embedded eDTC replace PCB decoupling?

No. Embedded capacitance supports part of the local supply network. Board capacitors and the voltage regulator must still satisfy the component’s wider power-delivery requirements.

Is 3.5 reticles the maximum CoWoS-L size?

No. It identifies a published production generation rather than a permanent ceiling. Obtain the qualified dimensions and availability for the intended design instead of treating a roadmap target as a released package.

Tags: CoWoS-L, cowos-l cross section, cowos-l lsi, cowos-l packaging, cowos-l process flow
Posted in best pcb, PCB Design, PCB Technology | No Comments »

Cookie	Duration	Description
cookielawinfo-checkbox-analytics	11 months	This cookie is set by GDPR Cookie Consent plugin. The cookie is used to store the user consent for the cookies in the category "Analytics".
cookielawinfo-checkbox-functional	11 months	The cookie is set by GDPR cookie consent to record the user consent for the cookies in the category "Functional".
cookielawinfo-checkbox-necessary	11 months	This cookie is set by GDPR Cookie Consent plugin. The cookies is used to store the user consent for the cookies in the category "Necessary".
cookielawinfo-checkbox-others	11 months	This cookie is set by GDPR Cookie Consent plugin. The cookie is used to store the user consent for the cookies in the category "Other.
cookielawinfo-checkbox-performance	11 months	This cookie is set by GDPR Cookie Consent plugin. The cookie is used to store the user consent for the cookies in the category "Performance".
viewed_cookie_policy	11 months	The cookie is set by the GDPR Cookie Consent plugin and is used to store whether or not user has consented to the use of cookies. It does not store any personal data.