Homebrew HV hiZ scope probe

On a sunny day (Sat, 20 Oct 2012 12:42:11 -0800) it happened Robert Baer wrote in :

Nobody can get that group.

Reply to
Jan Panteltje
Loading thread data ...

This HV load bank took a while to build.

formatting link

HV Probes are easy. A sealed dry AIR based (or gas) container for a precision HV resistor string. Potted versions lose accuracy and are harder to make calibration adjustments on.

Reply to
TheGlimmerMan

See similar title in a.b.s.electronic for PDF. Almost took longer to document than to build..

Reply to
Robert Baer

alt.binaries.schematics.electronic ? I get that group, but I don't see the post.

Cheers

Reply to
Martin Riddle

A scope probe needs high frequency compensation, too, like a cap across each resistor and a variable cap at the bottom to tweak the step response.

--
John Larkin                  Highland Technology Inc 
www.highlandtechnology.com   jlarkin at highlandtechnology dot com    
 Click to see the full signature
Reply to
John Larkin

  • Bet it took a fair amount of time for that labeling in the image.

Would adding heat shrink on the 1% resistors mess them up at high (up to 20KV) voltage? As an alternate, what about wrapping them with Kapton(TM)?

Reply to
Robert Baer

Yep. For some reason I _never_ see any posts by Panteltje >:-} ...Jim Thompson

--
| James E.Thompson, CTO                            |    mens     | 
| Analog Innovations, Inc.                         |     et      | 
 Click to see the full signature
Reply to
Jim Thompson

Ditto

RL

Reply to
legg

I was being facetious >:-} ...Jim Thompson

--
| James E.Thompson, CTO                            |    mens     | 
| Analog Innovations, Inc.                         |     et      | 
 Click to see the full signature
Reply to
Jim Thompson

I believe its Baer's post that's gone astray. RL

Reply to
legg

freenews.netfront.net carries ABSE but only the smaller posts (not large binaries)

For the binaries Astraweb.com 180GB $25 never expires. Since march ABSE has used aboiut 0.68GB, so at that rate I'll be putting it in my will.

--
?? 100% natural 

--- news://freenews.netfront.net/ - complaints: news@netfront.net ---
Reply to
Jasen Betts

On a sunny day (22 Oct 2012 07:22:02 GMT) it happened Jasen Betts wrote in :

OK I see that group, but not his pdf.

Well, I do not need it that much.

Reply to
Jan Panteltje

Somehow the post vaporized, so i put it up on my corporate site at

formatting link
. That probe was a "quick and dirty"; rise time seems to be better than

20nSec (my "pulse generator" is a HP3312A function generator). However, the output step only goes up 70% then a slow rise to max with a time constant of about 600nSec.
  • Am using SPICE to help design a 40KV probe and i seem to be homing into a design with values. The listing is below (and on that site). The attenuation ratio is 2500 and is "protected" to ground like the Tektronix probes. Why 2500? Because it takes two 1Gohm resistors to withstand the 40KV (target rating is 30KV with some "elbow room"). Note the resistance ratio is exactly 2500:1 and (now) the capacitance ratio is close also.

Physical layout is to have a floating ring around each resistor,to provide a controllable input coupling and capacitive divider (which is the secret of a (theoretical) infinite risetime. Now around this whole assembly will be a grounded shield (to isolate input from external bazzzz fazzzz). I think that five sections is a reasonable division of each resistor for emulation of the actual distributed part.

Risetime seems to be infinite, but there is this slow "hump" that i am fighting. Any ideas as how to solve?

Version 4 SHEET 1 2064 680 WIRE -352 -16 -448 -16 WIRE -208 -16 -272 -16 WIRE -64 -16 -128 -16 WIRE 80 -16 16 -16 WIRE 224 -16 160 -16 WIRE 448 -16 304 -16 WIRE 592 -16 528 -16 WIRE 736 -16 672 -16 WIRE 880 -16 816 -16 WIRE 1024 -16 960 -16 WIRE 1280 -16 1104 -16 WIRE 1328 -16 1280 -16 WIRE 1520 -16 1328 -16 WIRE 1552 -16 1520 -16 WIRE -352 16 -352 -16 WIRE -272 16 -272 -16 WIRE -208 16 -208 -16 WIRE -128 16 -128 -16 WIRE -64 16 -64 -16 WIRE 16 16 16 -16 WIRE 80 16 80 -16 WIRE 160 16 160 -16 WIRE 224 16 224 -16 WIRE 304 16 304 -16 WIRE 448 16 448 -16 WIRE 528 16 528 -16 WIRE 592 16 592 -16 WIRE 672 16 672 -16 WIRE 736 16 736 -16 WIRE 816 16 816 -16 WIRE 880 16 880 -16 WIRE 960 16 960 -16 WIRE 1024 16 1024 -16 WIRE 1104 16 1104 -16 WIRE 1280 32 1280 -16 WIRE 1328 64 1328 -16 WIRE 1520 64 1520 -16 WIRE 1568 64 1520 64 WIRE 1568 80 1568 64 WIRE -352 112 -352 80 WIRE -272 112 -272 80 WIRE -272 112 -352 112 WIRE -208 112 -208 80 WIRE -208 112 -272 112 WIRE -128 112 -128 80 WIRE -128 112 -208 112 WIRE -64 112 -64 80 WIRE -64 112 -128 112 WIRE -32 112 -64 112 WIRE 16 112 16 80 WIRE 16 112 -32 112 WIRE 80 112 80 80 WIRE 80 112 16 112 WIRE 160 112 160 80 WIRE 160 112 80 112 WIRE 224 112 224 80 WIRE 224 112 160 112 WIRE 304 112 304 80 WIRE 304 112 224 112 WIRE 448 112 448 80 WIRE 528 112 528 80 WIRE 528 112 448 112 WIRE 592 112 592 80 WIRE 592 112 528 112 WIRE 672 112 672 80 WIRE 672 112 592 112 WIRE 736 112 736 80 WIRE 736 112 672 112 WIRE 784 112 736 112 WIRE 816 112 816 80 WIRE 816 112 784 112 WIRE 880 112 880 80 WIRE 880 112 816 112 WIRE 960 112 960 80 WIRE 960 112 880 112 WIRE 1024 112 1024 80 WIRE 1024 112 960 112 WIRE 1104 112 1104 80 WIRE 1104 112 1024 112 WIRE 1280 144 1280 96 WIRE 1568 144 1520 144 WIRE -448 160 -448 -16 WIRE -32 160 -32 112 WIRE 784 160 784 112 WIRE 1568 192 1568 144 WIRE 1328 224 1328 144 WIRE -448 272 -448 240 WIRE -32 272 -32 224 WIRE 784 272 784 224 FLAG 1328 224 0 FLAG 1568 192 0 FLAG 1552 -16 scope FLAG 1280 144 0 FLAG -448 272 0 FLAG 784 272 0 FLAG -32 272 0 SYMBOL voltage -448 144 R0 WINDOW 0 8 7 Left 2 WINDOW 3 11 105 Left 2 WINDOW 123 0 0 Left 2 WINDOW 39 0 0 Left 2 SYMATTR InstName V2 SYMATTR Value PULSE(0 250 0 1p 1p 10 20 1) SYMBOL cap 1296 96 R180 WINDOW 0 24 56 Left 2 WINDOW 3 24 8 Left 2 SYMATTR InstName C31 SYMATTR Value 2478p SYMBOL res 1312 48 R0 WINDOW 3 30 126 Left 2 SYMATTR Value 4.008Meg SYMATTR InstName R31 SYMATTR SpiceLine tol=0.1 pwr=1 SYMBOL cap 1552 80 R0 SYMATTR InstName C32 SYMATTR Value 22p SYMBOL res 1536 160 R180 WINDOW 0 36 76 Left 2 WINDOW 3 36 40 Left 2 SYMATTR InstName R32 SYMATTR Value 1Meg SYMBOL cap -368 16 R0 WINDOW 0 20 11 Left 2 SYMATTR InstName C1 SYMATTR Value {Cr} SYMBOL cap -288 16 R0 SYMATTR InstName C2 SYMATTR Value {Cr} SYMBOL res -368 0 R270 WINDOW 0 -24 71 VTop 2 WINDOW 3 68 60 VBottom 2 SYMATTR InstName R1 SYMATTR Value {Rp} SYMBOL cap -224 16 R0 WINDOW 0 20 11 Left 2 SYMATTR InstName C3 SYMATTR Value {Cr} SYMBOL cap -144 16 R0 SYMATTR InstName C4 SYMATTR Value {Cr} SYMBOL res -224 0 R270 WINDOW 0 -24 71 VTop 2 WINDOW 3 68 60 VBottom 2 SYMATTR InstName R2 SYMATTR Value {Rp} SYMBOL cap -80 16 R0 WINDOW 0 20 11 Left 2 SYMATTR InstName C5 SYMATTR Value {Cr} SYMBOL cap 0 16 R0 SYMATTR InstName C6 SYMATTR Value {Cr} SYMBOL res -80 0 R270 WINDOW 0 -24 71 VTop 2 WINDOW 3 68 60 VBottom 2 SYMATTR InstName R3 SYMATTR Value {Rp} SYMBOL cap 64 16 R0 WINDOW 0 20 11 Left 2 SYMATTR InstName C7 SYMATTR Value {Cr} SYMBOL cap 144 16 R0 SYMATTR InstName C8 SYMATTR Value {Cr} SYMBOL res 64 0 R270 WINDOW 0 -24 71 VTop 2 WINDOW 3 68 60 VBottom 2 SYMATTR InstName R4 SYMATTR Value {Rp} SYMBOL cap 208 16 R0 WINDOW 0 20 11 Left 2 SYMATTR InstName C9 SYMATTR Value {Cr} SYMBOL cap 288 16 R0 SYMATTR InstName C10 SYMATTR Value {Cr} SYMBOL res 208 0 R270 WINDOW 0 -24 71 VTop 2 WINDOW 3 68 60 VBottom 2 SYMATTR InstName R5 SYMATTR Value {Rp} SYMBOL cap 432 16 R0 WINDOW 0 20 11 Left 2 SYMATTR InstName C11 SYMATTR Value {Cr} SYMBOL cap 512 16 R0 SYMATTR InstName C12 SYMATTR Value {Cr} SYMBOL res 432 0 R270 WINDOW 0 -24 71 VTop 2 WINDOW 3 68 60 VBottom 2 SYMATTR InstName R6 SYMATTR Value {Rp} SYMBOL cap 576 16 R0 WINDOW 0 20 11 Left 2 SYMATTR InstName C13 SYMATTR Value {Cr} SYMBOL cap 656 16 R0 SYMATTR InstName C14 SYMATTR Value {Cr} SYMBOL res 576 0 R270 WINDOW 0 -24 71 VTop 2 WINDOW 3 68 60 VBottom 2 SYMATTR InstName R7 SYMATTR Value {Rp} SYMBOL cap 720 16 R0 WINDOW 0 20 11 Left 2 SYMATTR InstName C15 SYMATTR Value {Cr} SYMBOL cap 800 16 R0 SYMATTR InstName C16 SYMATTR Value {Cr} SYMBOL res 720 0 R270 WINDOW 0 -24 71 VTop 2 WINDOW 3 68 60 VBottom 2 SYMATTR InstName R8 SYMATTR Value {Rp} SYMBOL cap 864 16 R0 WINDOW 0 20 11 Left 2 SYMATTR InstName C17 SYMATTR Value {Cr} SYMBOL cap 944 16 R0 SYMATTR InstName C18 SYMATTR Value {Cr} SYMBOL res 864 0 R270 WINDOW 0 -24 71 VTop 2 WINDOW 3 68 60 VBottom 2 SYMATTR InstName R9 SYMATTR Value {Rp} SYMBOL cap 1008 16 R0 WINDOW 0 20 11 Left 2 SYMATTR InstName C19 SYMATTR Value {Cr} SYMBOL cap 1088 16 R0 SYMATTR InstName C20 SYMATTR Value {Cr} SYMBOL res 1008 0 R270 WINDOW 0 -24 71 VTop 2 WINDOW 3 68 60 VBottom 2 SYMATTR InstName R10 SYMATTR Value {Rp} SYMBOL cap -48 160 R0 SYMATTR InstName C21 SYMATTR Value {Cs} SYMBOL cap 768 160 R0 SYMATTR InstName C22 SYMATTR Value {Cs} TEXT 264 256 Left 2 !.tran 0 10m 0 10n TEXT 264 -208 Left 4 ;40KV 2E9 ohms HV scope probe TEXT -144 -120 Left 2 ;Ohmite MOX2-131007FE\nin pi net distributed form TEXT 168 160 Left 2 !.PARAM Cr=10p, Cs=7.9p, Rp=200Meg TEXT 672 -120 Left 2 ;Ohmite MOX2-131007FE\nin pi net distributed form

Reply to
Robert Baer

It's conventional to include a 'tip' resistor, that will absorb contact (arc) surges safely.

The hardest thing about HV probe structure, is getting practical values that are physically possible, to do the job. Look at physical embodiments, make a few measurements, then go to the model.

Not much point in speculating over something that you can't build. Sizing the capacitive divider is a real physical issue. Your Cr/Cs ratio may be impractical.

As to risetime, an intentional RCseries, paralleling C22 position can act on leading edges. Look at the current midpoint voltage (jn C10/C11). Also, run an ac sweep.......the LF and HF gains will meet at some point that will look like a can of worms, and is just as easily manipulated, given physical constraints.

RL

Reply to
legg

High-Ohm resistors have been reported to be non-linear, possibly that is what you are getting. You may need some more poles of RC compensation to try to fix this.

Jon

Reply to
Jon Elson

Have you attempted a spice simulation of the part you've actually built and tested? This would be more than just instructive and assist in developing the newer iteration and an understanding of the spice derivative.

RL

Reply to
legg

  • The only guides i have so far,are the Tektronix P6015 probe and the "quick and dirty" 1Gohm probe i built. Unfortunately, i have no way to making reasonable guesses related to capacitances involved in their construction. The few P6015 drawings leave a lot to be desired, and measurements on my Q&D probe would be extremely difficult to do due to some of the very low values.
  • But i have one working one, and the model for it. And as far as sizing the capacitive divider, as long as there is a simple capacitor from tip to output, and its ratio to a total corresponding capacitance from there to ground, there is no problem; that is simple - for a simple model.
  • The Cs was tweaked for "best looks", and could be easily done mechanically. The Cr is done via floating coaxial "shield" around the resistor (see the aluminum foil wrap on the Q&D probe i built).
Reply to
Robert Baer

"non-linear" can mean many things..voltage coefficient, temperature coefficient, and resistance per unit across the length (i assume that is what you are referring to). Where is that "report"? What date? I do not think the resistance per unit length varies much, certainly not so much as to give the waveform seen in the simulation. That can be varied by quite a bit by changing Cs and/or Cr. However, it might be a GOOD THING (as Martha Stewart would say) to use ten pi pads instead of five for the resistor.

Reply to
Robert Baer

Yes, but i started with a very simple C across the 1G resistor and one to ground; it was plain that i needed (so to speak) more cable capacitance,and better adjust-ability of that. Made no refinements. Basically, that probe was an exercise of "can it be done"? And it has the noted deficiency of no ground return inside (hence the big fat warnings).

Reply to
Robert Baer

See

formatting link
for updated PDF. Have added two more scope pictures and modified closing notes.

This shows that the fundamental risetime is faster than 20nSec.

Reply to
Robert Baer

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.