2
\$\begingroup\$

I have seen that trace impedances in PCB's are maintained 50 ohm & so on. How to measure this impedance using a simple multimeter?

\$\endgroup\$
1
  • \$\begingroup\$ You can get fairly close by calculation though. Unless you're laying out a gigabit PCB, you will probably get by with the predicted impedance. And unless you have a long track, the parasitic effects of corners, pads, vias etc will dominate. \$\endgroup\$ – tomnexus Jan 27 '15 at 16:08
9
\$\begingroup\$

There is no* way to measure trace impedance with a simple multimeter.

Measuring trace impedance requires, at minimum, an oscilloscope and an impulse generator of some sort. Here's an article about how you can measure the impedance of a cable (or PCB) with a signal generator.

Doing "Proper" impedance measurements requires a special tool, called a network analyzer.

* No reasonable way, in any event. There may be ways to abuse a specific model of meter in a specific manner to get a approximate result, but it won't be generalizable.

\$\endgroup\$
1
  • \$\begingroup\$ @Connor_Wolf I think you should explain how to use a network analyzer to measure characteristic impedance of a PCB trace in the lab? It's not that easy to do for short structures. I feel a TDR is much easier to use (but be really careful with ESD). \$\endgroup\$ – Rolf Ostergaard Jan 27 '15 at 17:23
5
\$\begingroup\$

You should use a TDR (time domain reflectometer) or a Lecroy Sparq which is a cross between a network analyzer and a TDR. You can also use controlled impedance with your board house and specify to them what the impedance is. Then demand a test coupon and measurement results.

As a poor mans approach you can sort of build your own TDR with your oscilloscope. But you need a fast edge / pulse source, probably in the ps range depending on how long your traces are. Preferably with an output impedance close to your trace. I once used a analog devices high speed opamp with ps rise on the output for this.

Anyway input the fast pulse into your trace ( with no termination) and if you measure with your scope at the source you should see the reflection from the unterminated edge come back.

Then add a termination resistor at the far end of the trace that you think matches the impedance. When you get it right no more reflection :)

A little tricky but doable if you can't afford the test gear.

\$\endgroup\$
1
  • \$\begingroup\$ Oh I should add you'll need a scope with enough bandwidth and sample rate to see the ps rise time :) \$\endgroup\$ – Some Hardware Guy Jan 27 '15 at 14:01

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.