Hi,

I am an EE student with a passion for hardware. Recently I got myself into a DIY HV differential probe design project. I have been studying all the different circuits on the internet and also improved my knowledge of diff amps...

My goal is to design a 100 kV 200-150 Mhz probe (I know it sounds absolutely mad, but this project is meant for learning and discovering. 100 kV voltage rating will only cost me the extra mile, forming a 1000:1 divider out of dozens of 50-100 MOhm 0.1% resistors with a respective voltage ratings for each of them) I dont have a strict budget or restrictive form factor for the PCB, basically, I dont have any limitations...

One thing I can not understand is how to design all the opamp feedback networks and how to pick all those small pF order of magnitude capacitors ? I know that we have to match the divider part not only for the resistors but for capacitors also to match the oscilloscope input impedance and form an even resistive-capacitive voltage divider... I also understand the instrumental amplifier topology consisting of 2 independent opamps and the third one functions as diff amp. But how do you pick all those trimmer cap values? How do you pick the diff amp input branch capacitors? Why are there so many resistors in parallel? Does the physical SMD footprint size of the resistor/capacitor influence the target cap value? Usually, a wise approach is to make a perfboard demo version of the circuit before designing the PCB, but if we are fighting for the pFs at 100s of Mhz in this case, perfboard can take a rest until the next project... I have also seen designs incorporating discrete matched JFET`s instead of the precision opamps or even the diff amps. Is that a good practise? Any other advice or useful information conquering this project? Thank you in advance.

 

What kind of signal is 200MHz at 100KV?!
You do know that 100kV can be quite lethal.

 

What kind of signal is 200MHz at 100KV?!
You do know that 100kV can be quite lethal.

Just like I mentioned, this project is not about measuring actual 100 kV @ 200 Mhz, it is about gaining the knowledge and experience... If we take out the 1000:1 divider part (lets say I am designing a 1000 V probe), my question is still the same

 

What must the Probe's Impedance be ?, ( at 200Mhz ),
so as to not have a negative effect the operation of the super-High-Voltage-Circuit ?

With Voltages and Frequencies that high,
You first stay as far away as possible,
and measure the Power being radiated with a specially designed Antenna .
.
.
.

 

Just like I mentioned, this project is not about measuring actual 100 kV @ 200 Mhz, it is about gaining the knowledge and experience... If we take out the 1000:1 divider part (lets say I am designing a 1000 V probe), my question is still the same

Those types of High Voltage/Wide Band instruments are complex and usually very expensive because they require custom parts. Good luck finding practical design information. We normally buy the divider and use a calibrated DC/AC meter as the readout for low frequency measurements.
https://www.rossengineeringcorp.com/products/measurement/hv-voltage-dividers.html

 

I have this. 1000:1 (not diff) probe. 50mhz. N2771 Keysight.

DP0001A 400 MHz High Voltage Differential Probe
Just got some of these. Not certain if I like them.

 

I see no real purpose in trying to design a probe with off-the-wall characteristics that has no practical purpose, and would likely be very difficult to actually build.
And I find it difficult to generate enough interest in such an unobtainium project to generate advice on how to do it.
Better to pick one with characteristics that could be built and used.

 

Ok, lets forget about 100 kV... Lets say I am designing a 1kV 200 Mhz diff probe for oscilloscope (50 Ohm input). I will have USB connector on my PCB to power all the amplifiers from 5V (I guess it will be single rail supply) My question is exactly the same, how do I calculate all the pF capacitors? Is there a way to somehow simulate the circuit (as I understand, this project is very PCB layout sensitive, so something like LTSpice would not do anything useful?)

 

Are You messing around with an old Transmitter ?
( You could get yourself DEAD )

Don't keep us in the dark ........
What specific problem are You trying to solve ?
Sometimes odd situations require specially designed test gear.
.
.
.

I am not messing with anything. I just want to do a DIY HV diif probe for myself just for the learning purposes. If 100 kV is unobtainable, then lets make it 1 kV but still 200-150 Mhz. However, my question remains the same