As a contract manufacturer, we review many circuit board designs. Often, too often, orientation markings are missing from PCB layouts, and sometimes they are incorrect.
IPC standard is explicit that the cathode (negative) is pin 1 and the side which is marked on the silkscreen. Even if you have a multicolor LED with common anode, the orientation mark must point away from it.
Since diodes and LED’s not having polarity markings is such a frequent problem, we have a simple set of rules for determining the correct orientation. We’re going to share these rules here:
I’m going to say “almost” a lot because there are occasionally, very rarely, exceptions to these rules.
Circuits with negative voltage power supply reverses the rules. Determining if a supply is negative is part of this guide.
If you have schematics, you can easily determine orientation by matching a trace connection to any nearby part. Unfortunately we almost never receive schematics, just the BOM from which we can determine the type of diode, so…
Locate Ground & Identify Negative Supply
Before determining orientation you need to know which inner layer or trace is ground and what the polarities of the power supply planes are.
In most cases, digital designs will be all positive supply.
If there are any silkscreen markings explicitly indicating the power input polarity, which connector pin is ground, that helps.
Many connectors such as USB have standard pinout which can be used to determine ground and +5V. The housing of the connector will be grounded.
Mounting holes, housings, shields should all be grounded.
The outer sleeve of DC barrel jacks should be ground for safety reasons. The PCB designer may not follow this convention, but any off the shelf wall adaptor will be this way.
In 4 and 6 layer boards, layer 2 is almost always ground.
The ground plane will have few or no cuts in it relative to the power plane which is usually divided into different voltages.
If plane layers are inverted, which they should be, unconnected vias have a circle around them. Connected vias will either have nothing around them or an X-pattern on them.
There are normally more via connections to ground than any power supply.
An IC with a central power pad will be connected to ground with multiple vias, but voltage regulator tabs and transistor tabs are often not grounded.
The negative of aluminum or tantalum caps are often connected to ground unless there are negative voltages, but if there are negative voltages you will usually see two polarized caps with plus and minus connected together, that mid-point would be ground. A plane connected to the negative terminal of a polarized cap will be either ground or negative.
For LED’s the stripe connects either directly to ground or through a resistor to ground, except LED’s connected to negative power supply or that come on when something is connected backwards.
LED’s may be switched on and off by either the anode or cathode so there is no preference for parallel LED’s to share either end.
Other types of diodes almost never have the cathode connected to ground as this would just be a short circuit, except in circuits with negative voltage power supply which reverses the rules.
Zener Diodes & TVS
A Zener diode has a specific reverse voltage. They are normally installed backwards, with the cathode (negative) connected to the power supply positive and/or the anode (positive) connected to ground.
Not all TVS are Zener, but all Zener are TVS. TVS is transient voltage suppressor which limits voltage to a specific level. A bipolar Zener TVS is actually two diodes in series in opposite directions to suppress voltage in both directions. Non-zener TVS are always bipolar and unreliable for anything but minor noise and spike suppression. Non-zener TVS (ie MOV, varistor) should never be used for reverse hookup protection. TVS are designed to withstand large short duration pulses of power but are normally off, normal Zener are meant to be on all the time at a constant level though they may dissipate a lot of power.
Two diodes in series may go in the same or opposite directions, but both of these cases are only going to be Zener diodes or Zener TVS. If both ends are connected to ground and power with nothing in between, they go in opposite directions, they are suppressing over-voltage (transient, TVS). There must be a fuse or thermistor next to them, but it could be off board and a connector should be next to them. If they are in series with something else (LED, IC) they are in the same direction, they are dropping voltage to protect those devices they are in series with. This may be done either to get a specific unusual value of voltage or to spread out power dissipation. Connecting two Schottky in series defeats the point of using Schottky, but two regular diodes could be used like a 1.4V zener.
Schottky diodes are low voltage and therefore low power, efficient, typically used in power supply sections.
The anode (positive) of Schottky diodes may be connected to the positive power input terminal. In battery powered devices, this prevent loss of battery power to anything else that might be connected to that power supply, or as a very simple AC-DC rectifier.
Schottky diodes almost always have anode (positive) connected to either positive power input or ground and cathode (negative) to positive power output. There is usually a large inductor in the area and neither end of the large inductor should be connected to ground. Normally one end of the inductor will connect to the Schottky and the other to the power output. If the output of a power supply is negative there may be an aluminum or tantalum cap to indicate the polarity.
Diodes which are not obviously routed to ground are intended to rectify current: block current from flowing the wrong way, only allow current to flow in the correct direction. Schottky’s are rectifier’s they are just more efficient at it for high current power circuits. Regular diodes are used in regular analog and digital signal pathways.
Whenever multiple diodes (not LED’s) are connected together the common pin is almost always cathode (negative). This is to protect multiple signal outputs from driving into each other when sharing/driving a bus line.
In general, if you can trace one end of a diode to a dedicated output of any IC, that end will be the anode (positive), or if you can trace one end of a diode to a dedicated input of any IC, that end will be the cathode (negative), unless of course the signal uses negative voltage.
Two diodes in parallel always go in opposite directions without exception. If two diodes are installed in the same direction in parallel, only one will ever function properly. This is for hysteresis which prevents the circuit from rapidly switching back and forth (going crazy) around it’s trip point. Schottky diodes will provide smaller hysteresis. There is never any reason for Zener diodes to be connected in parallel as the forward voltage will always come on before the spec voltage.
Determining the orientation of IC’s is usually much easier than diodes because there will normally be explicit ground and power pins. The explicit input and output pins can be used to determine orientations of signal diodes. Many common IC’s such as opamps and logic gates have standard pinouts which makes it easier (ie, top right pin VCC, bottom left pin GND/VSS).
Trying to determine orientations at the last minute while the SMT line is waiting is very frustrating.
Please try to follow IPC guidelines when designing your footprints.
If you need help, please ask. We prefer your board to be designed right the first time as much as you do.
Do not rely on footprints you downloaded off the Internet, you’ll regret it.
Check your Gerber files in Viewmate or another Gerber viewer before sending them out. We like Viewmate because it can open zip files directly without unpacking them first.
About the Author
Tige Gibson is an electronics engineer with over 20 years experience in design and manufacturing.