Differential oscilloscope probes: The common and the better

Relevant videos
Oscilloscope measurements in power electronics: I. Single ended probe
• Oscilloscope measureme...
Oscilloscope measurements in power electronics: II. Differential probe
• Oscilloscope measureme...
Correction
Slide 15: the unit is powered by a USB port not a battery as said.

[@kenhenderson7858]

Thanks Prof Ben Yaakov for another great video!. We have optical isolated channels in our CS448 oscilloscope with over 100dB CMRR and it is always a challenge to explain how that helps to our customers, so the material covered in this video is good to see.

[@sambenyaakov]

Hi Ken thanks for input. Can you write to me off line to [e-mail]? and BTW: the phrase " The gates are slewing 0->500V in about 10ns while the measurement is made. " in your writeup does not sound right😊

[@akosbuzogany2752]

Once I made a diff probe and it turned out 2-3 cm guitar steel strings are great as probe tips. They are flexible, springy and have some surface treatment against rust which makes them solder like charm. Tiny capacitances (0.2-0.3pF) for C1 and C3 I did by PCB traces.

[@sambenyaakov]

Hi Akos, thanks for sharing. Interesting.

[@tseckwr3783]

Thanks Professor Sam. Appreciate the comparisons.

[@sambenyaakov]

Thanks

[@JadeHare1]

Absolutely excellent video! Thank you 👏🏼👏🏼

[@sambenyaakov]

Thanks

[@raviteza8]

Great Video, Prof. Ben Yaakov. I did similar simulations a few months ago. The most important thing is the input capacitance. Even 1pF is too much. In order to differentiate if the oscillations are caused by the probe or the circuit, on needs a flat response not only of the input bandwidth but also of the CMRR of the probe into atleast a few hundreds of MHz range if not GHz.
If we can design a differential probe with very low capacitance, say 0.2pF, then in theory we can have a very good CMRR. In fact this is what is done in active LV probes with 2 or more GHz bandwidth. At the end of the day, the impedance rolls off and the CMRR deteriorates. The Isomue from Tektronix resolves this issue by simply isolating. My wish is to design a probe which can compensate for this CMRR roll off, perhaps using a special circuit in addition to a very low input capacitance of less than 0.2 pF. I am not sure if its physically possible to develop a HV differential probe with these specifications. An LV active probe is possible and is commercially available. Of course, one can always resort to what Tektronix did, simply isolate, also Clever Scope CS448 uses a similar principle and achieves great results. The other points you mentioned are equally important like inductance etc.
I simulated different configurations with single and multiple stage HV Instrumentation Amplifiers in LTSpice. Also the ratio of the differential gain to the common mode gain is very important in the first stage and should be maximized with proper choice of components. But the compensation of the feedback is really hard at high frequencies.
But anyhow, as always, Prof. Ben Yaakov, great video!

[@sambenyaakov]

Hi Ravi, Thanks for sharing your experience . Interesting. Differential probes are indeed a hard nut to crack

[@kenhenderson7858]

Thanks Ravi Teza for mentioning our CS448 isolated oscilloscope. You are quite right, making these differential measurements in the presence of high common mode voltages is much easier with optical isolation. We use optical isolation inside the scope as compared to IsoVu with the optical fiber on the probe.
Thanks also to Prof. Ben for this great video. It is always a challenge to understand how CMRR changes the measurements, and this video is very clear.

[@keithking1985]

really liked this video. you covered it very well Sir. thank you.
i really was blown away by the voltage capability's of the fibre optic probe. wow.

[@sambenyaakov]

Thanks

[@SaibalHmmzz]

Given 1st like, thank you for sharing this knowledge Sir. 🙏

[@sambenyaakov]

😊Thanks

[@Stelios.Posantzis]

Prof. Ben-Yaakov I hope you write a book with all of this material you have on videos. I am sure it will be a thrilling read and of terrific instructional value.

[@sambenyaakov]

Thanks. I doubt if I will ever have the time to get to it. But I will certainly try.

[@Stelios.Posantzis]

@@sambenyaakov It's certainly a worthwhile endeavour! There's plenty of books in the field but I reckon most are rather terse and fairly dry and impenetrable so I have a feeling that if you ever wrote one, it would be a success.

[@saivedavyaschowdarychintaa5668]

Sir you are an ocean full of knowledge

[@sambenyaakov]

Thanks. Let's not exaggerate I will settle for a sea😊

[@saivedavyaschowdarychintaa5668]

@@sambenyaakov no exaggeration but that's true according to me I do follow your videos regularly

[@sambenyaakov]

@@saivedavyaschowdarychintaa5668 Thanks. Happy to share. Have a happy and healthy new year.

[@saivedavyaschowdarychintaa5668]

@@sambenyaakov thank you and wish you the same

[@bencepapp5533]

Great video!
My question is: What can be done when the measurement can't be done with short wires because of physical accessibility? How the parasitic inductance can be compensated? If I connect the diff probe to the measurement points with coax cable can that help?

[@sambenyaakov]

Try to twist the wires.

[@mata7648]

As far as I know, the differential probe usually uses JFET pairs on the input side. Prof, could you explain that a bit?

[@sambenyaakov]

JFET amp have very low bias current, low input offset and high BW. Perfect for a differential probe.

[@huanzhou4768]

Thank your for your sharing.
Both the Differential oscilloscope probes you showed have errors. Durig my test in the laboratory, we use the passive prob and we solder the probs direclty to mesuring points, in order to decease stray indutance; Meantime, for the oscilloscope we isolate ( not connect to the Ground): I think this way is better to measure the Vgs of High side Mosfet for example;

[@raviteza8]

Isolating the oscilloscope is the most dangerous Thing to do. The bnc connector is not grounded any more and it's a serious safety concern

[@NiHaoMike64]

@@raviteza8 There are plastic cased scopes that are safe to use floating. In that case, it would be a good idea to apply heat shrink over the collars on the probes and put plastic caps over unused inputs.

[@sambenyaakov]

Using an isolated (by transformer) scope is not only dangerous but of little use. Have you ever checked the impedance between the "ground" of the oscilloscope and "earth" you will be surprised. I would suspect that the impedance is also low for a battery operated scope.

[@kkh8623]

Hello, Prof Ben Yaakov. I'm Frank KWON. Thanks for your great lectures!.
BTW, I have several questions about GND of the circuit to be measured and diff probe GND to select probe appropriately. Can you let me know your email to ask specifically, please??

[@sambenyaakov]

[e-mail]

[@kkh8623]

@@sambenyaakov Thank you! I will send it!

[@kkh8623]

@@sambenyaakov Prof.Ben Yaakov. I have sent the mail to you. I hope you can read and reply it. Many thanks!

[@kkh8623]

Hello, Professor, I sent it on Jan 8th. I hope you can read it. The title is 'About Diff Probe VCM GND'. Thank you!

[@sambenyaakov]

@@kkh8623 I did not get it please send again
[e-mail]

[@tamaseduard5145]

👍🙏🙏🙏👍

[@sambenyaakov]

😊🙏

[@smeggyhead1]

13:11 "this is 10MHz" - did you mean 100MHz?

[@sambenyaakov]

Indeed. A slip of the tongue. Thanks for pointing this out.