M&M 6

I have a top-end spec Oscilloscope, and I believe what I see

This is a very common issue I have seen in the field. When I spoke to the engineers, asking “have you measured it?”. The answer is always, “yes, of course”. Then the next question “Can I have a look at your test set-up?”, 99% of the time I got answer “Sorry, I haven’t got a picture of the test set-up.”

This is the first problem, i.e. lack of detailed test report. Engineers are often getting themselves very busy testing, writing technical report, reporting to their boss. The very beginning of the whole process, i.e. test set-up was always treated as minor, not important and should not be included in the report. You would get lucky if they put the test set-up picture in the Appendix.

Then the second question is, “how good is your measurement?”

“Well, I have got this top-end oscilloscope, the company paid fortune for it, so my result must be good.”

Not necessary!

Measurement error contributes at least 20% of overall engineering productivity. Cases of engineers spending a week on something, only to find that it was due to a test set-up error were abundant. Here’a typical example from a client of ours

“When using oscilloscope to measure the switching waveform, I noticed that the spectrum analyser result was different when I probed the testing point.”

Oscilloscopes work very similarly, they have two ends that measure the electric field of the device under test (DUT). But by touching the two measurement points to the device, you inevitably guide field into the probe itself, therefore affecting the very signal you are trying to measure. This is the fundamental problem with probe measurement. From this point, it is safe to say that no oscilloscope measurements is accurate.

But we can always try to minimise the error. The first step is to reduce the coupling loop. This is particularly true if we are measuring high-speed circuit. The long tail crocodile lead should never be seen in a high-speed circuit measurement.

Don’t forget another subject, i.e. you yourself as an engineer. Have you seen a case that when you just stand closer to a circuit, or your arm is closer to the circuit, the measurement was different? This is because we as human beings are also a capacitor, when we are close to a circuit, we inevitably increase the parasitic capacitance of the circuit.

Figure 1 Using short ground connection helps reducing the magnetic loop in the measurement set-up

Here’s an interesting story to share. A few years back, I worked on a project with a very big engineer called Bill. When Bill was using an oscilloscope to measure some small signals on the circuit, he noticed the difference when he moved. Then he passed the probe to me, I could see the difference, but I couldn’t re-produce the same level as he did. Bear in mind, we were measuring some very small signals on a precision circuit. It turned out, big person also has a larger capacitance (of course). The sheer size difference between Bill and me caused the small measurement difference we saw on the signals.