Learning to build a differential probe

I wanted to ask the community what to read/try next to improve a very basic design. My goal here is to first understand how to build a cheap differential probe. They are a few schematics, but I'd rather understand than copy them.

I've read these pages :

I've ordered cheap parts (not quality ones as we'll see) and try to build my project on a protoboard. Here is the first schematic to test the opamp and power supply, very basic;



Basically the slew rate is reealy slow, I've measured 0.5v/us so it is not a TL072 (even I purchased an item with such a description on Aliexpress and beeing dumb... I've received many fake chips)



So now I'll have to take for consideration that I have an LM741 die in my dual op amp packages. I'll not have any high frequency performance, but for the sake of understanding it should work.

Then I build this schematic, the idea is to have a 1/2 differential probe to measure small signals (Vce on transistors to be tested). The purpose here is not to measure mains or anything, for now I stay small

The measurements I have are somhow understandable :

Yet I observe huge distortion that I do not know how to explain. In all the readings I had about building differential probes I haven't seen that. My guess is that my fake operational amplifiers are the cause of it. However in the fist test (in following mode on the first schematic) the distortion was not so dramatic.



Does anyone knows a good reading about that ? I bet that by getting op amps from a reliable vendors and better quality ones this would dissapear, but this is maybe du to my schematic ?

I understand that I do not compensate yet for the input capacitance of the opamps. But could this cause this distortion or do you have any good recommandations for a video tutorial or reading about it ?

Edit1 : Bias input voltage has been measured at 1.4mV

Edit2 : I've found found that Reasons not to use a 741 op-amp?

Best regards

There may be minor improvements with OA changes, passive value changes for gain and layout changes. Why not learn from the best? The original Teltronix probe manual for example uses discretes and not internally integrated OA's.

So in this FWIW/FYI answer, rather than analyze your failures, I'll show how good Diff Probes were originally designed by Tektronix in the late 60's.

Refs: https://www.davmar.org/TE/TekConcepts/TekProbeCircuits.pdf https://www.ebay.ca/b/Differential-Probe/181954/bn_7023241187





The downside with these probes is they needed a shorting cap when not in use as they were extremely low input capacitance of x pF and more than a 25V E-field would damage them which does not require contact. ESD was a no-no.

So I used one finger on the ground at all times when using them. Our female tech always complained when another rookie eng burnt out the front end and she had to repair them.

Back in the early 80's these probes were $250. Now $750 used.

Now you can find Le Croy Diff Probes "for almost free " but shipping is not.



other info https://cdn.teledynelecroy.com/files/manuals/ap033-active-differential-probe-operators-manual.pdf p. 8 Zin is very important on low level high BW Diff Probes up to 1 GHz.



The ringing is not fake OA but a poor choice of OA for this challenging application, which desires no ringing or overshoot on step inputs for both small and large, as the dynamic gain margin changes with step current and from GBW for different V gains.

Consider the cutoff characteristics near \$f_{-3dB} = 0.5 /t_{rise}\$ (0~63%) (approx)

Compare with datasheet. Differences may include 10 nH of RV2 (est.) adding more rippling.





This is why I asked for specs on your expectations. There are tradeoffs with internally compensated FET input OAs, such as CM range, too, which cause fringe gains to alter significantly with Vcm input.



source: https://www.ti.com/lit/ds/symlink/tl072.pdf