Practical use of a TDR for impedance measurements

[rx8pilot]

I am very curious about the practical details of measuring impedance with a TDR. Never owned or used one, my only knowledge is based on what I have been able to read. I believe the reading has given me a high-level overview of how the instruments work, but I have not found much that help understands how to make the measurements successfully and how to understand the traps and limitations for various instruments.

A few of the questions:
From what I understand, TDR's make relative measurements so it should be able to deal with various common impedances of DUTS? Not just 50ohms......
Is it practical/common to directly measure traces on a PCB or only using test coupons printed adjacent to the PCB design?
With a single channel TDR, is it possible to measure a differential structure on a PCB?
What kind of fixtures, probes, calibration devices are common for TDR's configured for measuring PCB structures, connectors, FPC's, etc.

If anyone has some practical experience they are willing to share, or perhaps links to reading materials or videos that accomplish the same - it would be appreciated. 

[Leo Bodnar]

I use TDR for two purposes

1) checking RF parts for being outright rubbish.  After RF cable or converter is out of the package it's hard to remember where it came from.

2) verifying my board design ideas - mostly if I got impedance right or not.  I would often put what I think should be 50R impedance microstrip around one edge of the panel. As well as +-0.1mm width microstrips.

I would recommend getting CSA803A or 11801/11802 and SD-24 head.

Leo

[rx8pilot]

Hi Leo - thanks for chiming in!

(I have one of your tiny fast pulsers which has been a fun learning tool)

At the moment, for TDR, I only have an HP 54120B sampling scope with the 20GHz TDR head. It is far more basic than the Tek CSA, but in terms of seeing relative differences in impedance on a PCB structure, it should offer some useful data if I can apply it well.

At the moment, I am trying to figure out how to best calibrate and physically probe the PCB. Calibration kits and probes are expensive, so I want to choose wisely. Not really sure where to start and how much money is reasonable to expect.

Do you have any suggestions for short/open/loads for cal? Not sure what I can get away with. The goal is not to be a metrology house, but see if my PCB designs are close enough to the intended design impedance.

Any probe suggestions? It looks like the money is in the rigging mechanics and otherwise the electrical path is fairly straightforward. Is that in the $hundreds or should I expect $thousands for a probe suitable for PCB coupons?

[Leo Bodnar]

I am no expert user myself, and definitely not of metrology ilk.

I suspect that you can create a reasonably simple TDR setup yourself with nothing more than a few pogo pins soldered to the end of SMA launch connector and 25R, 50R and 75R resistors.  If you are chasing after good 50R impedance paths then these values are more important than the short and the open that are targeted at VNA application.

The beauty of TDR is that you can see through a rubbish launch point since it is very broadband.
With VNA you could de-embed your fixtures but this involves sweeping quite a large spectrum chunk and doing complex maths (or having VNA that can do it for you.)

In that sense TDR is good for shooting from the hip.

These are my thought, not the facts!  Please someone correct me if I am grossly mistaken.

Ironically, I bought N1020A TDR probe set for a different reason - as a poor man's microprobing facility for measuring a few Tektronix hybrids.  I think all of its cost is in mechanical construction rather than RF side of it.

Cheers
Leo

P.S.Much enjoyed your factory tour videos!