Estimated reading time: 6 minutes

Zen and the Art of Accurate Impedance Measurement* – With Apologies to Prisi

In his 1974 philosophical novel "Zen and the art of Motorcycle maintenance” Robert M. Prisig contrasts his regular and ongoing daily approach to motorcycle maintenance with his friend's alternate view of leaving well alone between annual service center based maintenance. What has this got to do with accurate impedance measurement you may ask?  Please read on to discover more… While I am writing this with the aim of fine tuning your impedance measurements, the underlying principles hold true for any type of measurement.

When measuring impedance with any type of TDR or impedance test system there are a number of sources of potential measurement error. This month’s “Pulse” sets out to explain those error sources and how you can minimise them in order to achieve repeatable and accurate measurements. First let’s consider probes and probing, which can be the Achilles’ heel of any measurement; you should look for a probe which has a good RF match to the test cable and has good signal integrity at the launch of the test trace. Some lab probes are superb from an RF perspective, but in a production environment you should also look for adequate robustness to survive the rigorous nature of repeated testing.

Thinking next of calibration, TDRs used for impedance measurement are precision RF measurement tools and need to be annually calibrated. Reference air line standards are used for TDR calibration and are available in a variety of standard impedances. Precision air lines are a high cost item, so an acceptable alternative for less critical applications is a set of precision semi rigid cables calibrated against air lines. Air lines are made traceable by a precision metrology technique called air gauging which allows the internal bore of the air line to be measured and the impedance calculated using a standard formula traceable to national standards.

Hands OFF! Resting your hand or fingers on a test coupon will cause the impedance of surface structures to drop. Ensure operators are aware of this effect – it’s easy to demonstrate, simply make multiple tests on a surface trace and see how placing your fingers on the board surface alters the measurement. An extension of this source of error occurs if you test a bottom surface trace when the coupon is in contact with the work surface. Yes even the work surface can influence measurement results. The work surface has its own dielectric constant and a microstrip on the bottom of the test coupon or PCB will measure low if it is too close to the table surface. Two ways to resolve this – place the coupon in a test fixture or support jig or turn the coupon over so the bottom layer faces up.

Secrets of TDR Calibration: many TDRs exhibit a characteristic where calibration will only be correct if your TDR is calibrated with the same DC conditions at the end of the test cable as those occurring when you measure. Simply stated, if you calibrate a TDR using a precision broadband resistor as a standard you should also terminate your test sample (coupon) with the same resistance. However if you calibrate or verify your TDR with an open circuit standard such as an air line you should measure your test sample without resistive termination. As it is inconvenient to fit a terminating resistor to a test coupon, always calibrate a TDR used for coupon testing with an air line or precision semi rigid reference. The error between the two methods can be as much as 3 or 4 ohms. The reason behind this effect is that TDR sampling heads are precision fast switching diode bridges and are sensitive to temperature change. A change in DC conditions at the test sample will alter the DC output voltage of the pulser.

Worn cables and probes: RF Cables and probes have a finite life and should be regarded as consumable; with 24/7 operation they will in time become resistive and ultimately fail. Cable life can be extended by careful handling, avoiding tight bends, etc. Operators should be trained to take care of cables and not to wrap them tightly against the probe or flex them more than necessary to make tests; taking care of cables will help ensure accuracy is sustained between annual service and calibration. Avoid falling into the trap of one PCB fabricator whose well meaning operator intentionally used worn probes to help a batch of “out of spec” boards pass testing!

Drift: Some TDRs are susceptible to measurement drift over time and with temperature range. Ask if your TDR makes compensation automatically for this or if compensation is inbuilt. If your TDR does not have inbuilt compensation it may be necessary to verify measurements if the temperature of the test environment varies significantly during the day.

Cell phones: Are cell phones influencing your measurement? Cell phones are RF transmitters and regularly "sign on" to the nearest base station – even when you are not making a call. Test coupons are around the same length as a mobile antenna and will readily receive the signal from your phone. This is especially the case with surface structures like microstrip so do ensure phones are switched off in the test area.

Test coupons: Test coupons provide a reliable way of verifying PCB impedance and will offer you more reliable repeatable measurements than attempting to measure the board itself. The advantages of using test coupons are described in detail on the Polar website at www.polarinstruments.com. Important note: If your striplines are reading high – did you remember to short the planes Vcc and Gnd on the coupon? (NOT on the board!) Vcc and grounds need to be shorted on the coupon especially on narrow coupons to simulate actual RF conditions on the finished board. Tip: If increasing the height of the top plane to 10x its normal height on your field solver gives you a reading of impedance similar to your measured value you have probably forgotten to short circuit the planes on the coupon.

Last but not least is the area of protection against ESD damage, and the faster the risetime of your TDR system, the more care needs to be taken in this area to keep running costs to a minimum. I’ll expand this in more detail in a future edition.

Back to “Zen”, and whilst Prisig points out the wide difference between the two approaches to maintenance (frequent ongoing minor adjustment vs. longer interval based "central servicing") – the best quality results are obtained when both methods are used together and the user is sensitive to day to day changes in performance. From the above sources of measurement error, you can see how keeping impedance test operators well trained and attuned to the variations caused by cable damage or other degradation will ensure maximum confidence in your impedance measurements during the full interval between routine calibrations

If  you would like to seek further enlightenment visit Prisig's page at Amazon.com

The book sold over 4 million copies in twenty-seven languages and was described by the press as "the most widely read philosophy book, ever. According to Prisig – despite its title, "it should in no way be associated with that great body of factual information relating to orthodox Zen Buddhist practice. It's not very factual on motorcycles, either." Enough philosophy for this August!