Technology Sharing

Perhaps when you are making PCB boards from various manufacturers, you will encounter orders asking whether you need impedance?

In today's applications, designs are often made faster and faster, and controlling layout parameters is more important than ever.

There are several ways to do impedance control during PCB design and production. The most common method is to annotate the information within the schematic with symbols and special comments, incorporate these comments and design rules when laying out the PCB, and pass it through the drawing so that your PCB supplier understands it and can manufacture it appropriately.

A brief background on impedance control

Why is it necessary? Simply put, a signal of given energy and speed will leave the output pin of a component, travel through the PCB trace, and end up at the input pin of another component.

If the signal's energy is not completely absorbed by the receiver, some of the remaining energy may be reflected back into the trace and back to the output pin. These reflections may interfere with the real signal and reduce the overall signal integrity.

For very short traces you probably don't need to worry about controlled impedance, but as a general rule of thumb, any trace longer than 1/10 the wavelength of your signal should contain some kind of controlled impedance.

Another general rule of thumb is to add controlled impedance to traces that are longer than 1/3 of the signal rise time. High-speed USB, Ethernet, DDR memory, and other low-voltage differential communication systems are examples of data rates that require impedance control.

A trace has several characteristics to consider: height, width, length, distance between itself and other copper features (including copper layers below or on top of signal layers containing controlled impedance), dielectric constant, PCB manufacturing tolerances/limitations, etc. All of these need to be taken into account when calculating and designing controlled impedance. Most industry-standard CAD software has impedance control and differential pair feature sets that can calculate these for you (and even show simulations of expected signal integrity and any potential issues), but understanding how they work together can help to better understand when they are needed or not.

Here is a screenshot from Altium Designer (taken from Altium TechDocs) of a 4 layer board containing some simple 70 ohm impedance driven trace widths.

Today, impedance control is typically done at the board house using modeling software and testing the impedance on the bare board during production, but it is also important for PCB designers to know how to mark these specifications on the drawing.

This is usually accomplished by having a note and arrow on the PCB drawing pointing to a trace (or group of traces) on the board stating “90 ohm differential impedance required for layers 1 and 8 with 7 mil width and 8 mil spacing” along with a legend for the layer stackup and its requirements for copper thickness and dielectrics.If you have any impedance requirements, please log in to the JPE official website to inquire!