George Egely's Magic Wand

    Quote


    I am in the market for a high voltage high frequency scope probe (about 1000:1) . Most - even 'used' ones are in excess of $1000.00 . I have one similar already, but it's only good for low frequency use - up to 60Hz..


    Hi!

    I could recommend the HVP70 70 MHz differential probe from EEVBLOG. The quality of the product is excellent. They recently had a ”sales promotion” with a good discount. I gave about 270 US $. However I don’t know if that offer is still available. The general specs. Are as follows:

    Bandwidth DC to 70MHz (-3dB)

    Attenuation Ratio 1:10/100

    Accuracy ±1%

    Rise Time: 5ns


    Input

    Impedance 4 MΩ, 5.5pF each side to ground

    Category CAT III

    Differential Range ±70V (DC+AC Peak) and 70Vrms @1/10

    ±700V (DC+AC Peak) and 700Vrms @1/100

    Common Mode Range ±700V (DC+AC Peak) and 700Vrms @1/10 & 1/100

    Absolute Max. Voltage ±1400V (DC+AC Peak) and 1000Vrms @1/10 & 1/100 (Differential or Common Mode)


    Output

    Swing ±7V (into 50

    Offset (typical) <±5mV

    Noise (typical) 1.5mVrms

    Source Impedance (typical) 50Ω (for using 1M input system oscilloscope)

    CMRR (typical) -80dB @ 50Hz, -60dB @20kHz -50dB @1MHz, -40dB @ 10MHz


    Here is a link: https://www.eevblog.com/product/hvp70/


    Reagards


    Bo, SM6FIE

    Alan Smith is looking for something like that beast of a HV Tektronix probe (3k$ new) at Dr. Egely's lab:


    For example:

    It's a challenge to find such a piece of tech for cheap... used probes can be faulty, and need to stay away from clones.

    I'll keep an open eye, but I can't promise I'll come across the guaranteed hidden magic item on ebay, etc.

    Quote from Alan Smith

    What about this one?


    https://uk.rs-online.com/web/p/oscilloscope-probes/2291019


    DATASHEET_TESTEC_TT-HVP_15_B.pdf

    Testec I would also trust. Bandwidth is not >100MHz as prior requirement.


    If voltage and bandwidth is a matter of negotiation considering cost, and compromise on device operation voltage range, then I was hoping you'd look at the previous link mid-range offerings as a "guide" on price vs. capabilities:



    TT-HVP08 at 8kVpeak could be also tempting considering Dr. Egely's tests being <3kVDC, however -->


    What drives more the formation of those negative charge clusters which suppose to catalyze nuclear reactions, the voltage or current? Voltage dictates the accelerating force, and effects their speed, and the current density dictates their formation number and density? Increasing both their speed and their number, will increase a chance of interacting with a H proton? Thus a 20kV probe would be a good future investment to explore the higher voltage region? How about effect on voltage vs. current on electrode erosion?


    I'm no plasma physicist, someone please help me out here!

    Quote from Alan Smith

    I am in the market for a high voltage high frequency scope probe (about 1000:1) . Most - even 'used' ones are in excess of $1000.00 . I have one similar already, but it's only good for low frequency use - up to 60Hz..


    I think it needs to be good for 20kV .


    Anyone got any ideas (apart from building my own )?


    Quote from Alan Smith

    SM6FIE


    The 1400V max is a problem. The input to George's device is anything up to 3kV, the output is higher. But it does look like a nice piece of kit.


    Quote from Alan Smith

    Tibi.fusion


    I think you are right about this being difficult for the average tinkerer. The high voltage is of course offputting but more so is the fact that you probably need to spend at least € 3000 on equipment to measure, and I could see that rising to twice that very easily. Never mind, I don't spend money on golf or whisky. :)

    If you build your own:


    I am not convinced that flat response up to high frequencies means much at these very high voltages. Just by waving wires near things you will change HF characteristics, and wires look like antenna to space with a quite low impedance. So you just can't get high amplitude high frequency stuff at 5kV (impedance too low), therefore it is not going to alter the input power budget a lot. You can measure HF currents in the load easily enough and they tell you what is important.


    For HF HV measurement the problem is you are limited to a high impedance load so as not to draw much power from the circuit under test, so you cannot simply match a coaxial cable without a ridiculously high division ratio.


    Still, without matching coax you can treat it as capacitive up to 30MHz or so and have a matched capacitance divider (which is how scope probes do it).


    It is easy to have say 4.7pF 500V capacitors across each of your 500V resistors in the top half of the divider. For X1000 and 10 resistors that gives you 470pF across the lower resistor. You actually use a lower value because of coax + scope input capacitance but this is fixed and can be measured or calculated.


    R values could be 10 X 4.7Mohm, and 47kohm for the lower one.


    Given that you need a v high impedance divider you need to match capacitance like this for all measurements above some quite low cutoff:


    20M total divider resistance X 1000 => 20k low side. In parallel with typical a 30pF scope X10 probe -> 500kHz or so. Or if you use a X1 probe it would be 100X worse, because the input capacitance of the scope _+ cable goes up to 200pF or so, and the low end resistor goes down to 2k or so. So you do maybe want a 10X probe. (You need to calibrate the probe from the scope cal waveform, but that is easy).


    So:


    No capacitive matching - homebrew divider ok to 50kHz at 20M input. That would draw 1W at 5kV but it is known power draw. Increase to 20m is quite possible, you still have a 50kHz bandwidth fine for the 1kHz waveform.


    Capacitive matching - you can go up to 20MHz or so, maybe more. It is trickier to set up. And I don't see 50MHz signals at 5kV as being that significant.



    Tom

    Quote from Alan Smith

    What about this one?


    https://uk.rs-online.com/web/p/oscilloscope-probes/2291019


    DATASHEET_TESTEC_TT-HVP_15_B.pdf

    That looks good for rough measurement:

    DC accuracy 6%

    AC accuracy - it does not say. The 3dB bandwidth I guess gives you accuracy to within 3db below that. One would hope.


    As above - I think rough measurement is all you need for this except at DC. You can calibrate the probe for better than 6% at DC. I don't think it can be nonlinear.

    For the DIY enthusiast's viewing pleasure:

    Tektronix P6015A teardown :love:

    P6015A High Voltage Probe

    (It's not science-fiction, price reflects good parts, robust construction, strict quality control, impedance matching and top-tier bandwidth, brand)


    DIY HV probe youtube stuff:

    External Content www.youtube.com
    Content embedded from external sources will not be displayed without your consent.
    Through the activation of external content, you agree that personal data may be transferred to third party platforms. We have provided more information on this in our privacy policy.

    External Content www.youtube.com
    Content embedded from external sources will not be displayed without your consent.
    Through the activation of external content, you agree that personal data may be transferred to third party platforms. We have provided more information on this in our privacy policy.


    If the goal is to minimize cost at (all cost) then how about looking at capacitor voltage with a DIY setup, since relaxation oscillator period is in the us-ms range, and we can't really find out much there, it is perhaps just for the scope of measuring capacitor voltage (min, max) and repetition rate. Then for the interesting part, on the discharge event with the "small explosions", why not use a second non-inductive shunt resistor to measure the current (voltage drop across shunt) using a standard cheap 200MHz 10:1 probe? Iron core isolation transformer / battery on oscilloscope supply is of course recommended to increase the impedance for those pesky HF currents finding their alternative route and messing up readings (of course this means oscilloscope in not PE protected anymore so know what you're doing !!!)

    Quote from Alan Smith

    SM6FIE


    The 1400V max is a problem. The input to George's device is anything up to 3kV, the output is higher. But it does look like a nice piece of kit.

    A quick and dirty method would be to modify the input circuit of the probe. Typically the input circuit looks like the schematic below. You could for example change R19 and R78 from 1 Mohm to say 9.1 Mohm. These resistors need to be handpicked and closely matched. They are probably of the SMD 1208 type so it should be a fairly easy task to replace them.


    This will have the effect of scaling up the max voltage with a factor of about 3 (3.030). So now you are in the position to be able to cover the primary voltage span you are interested of. Of course there is some safety issues involved in this that has to be considered. The frequency response and CMRR will be affected. So it would be wise to do a total recalibration and test. Eventually also tweak the RC part of the input circuit.


    /Bo, SM6FIE


    SM6FIE . Thank you so much for your skill and input.


    I ordered the Testec £300 one in the end, as THHuxleynew (who also knows of what he speaks) thinks it will suffice. Fingers crossed and wearing rubber gloves for luck!


    https://uk.rs-online.com/web/p/oscilloscope-probes/2291019



    The problem with modding s commercial probe is not my ability or willingness to do it under other circumstances, but with the credibility gap that exists around all this stuff. Using a hacked probe would raise arguments we don't need. It's a bit like the magician who sacrificed a virgin to prevent an outbreak of plague in Venice. It didn't work and when the plague arrived anyway he admitted the chosen virgin had been 'slightly modified'.

    Just to note - I think it will suffice because I don't believe better than +/-30% measurement is needed at 5kV except for DC which can be calibrated.

    Just a reminder. Much of this is new.


    External Content www.youtube.com
    Content embedded from external sources will not be displayed without your consent.
    Through the activation of external content, you agree that personal data may be transferred to third party platforms. We have provided more information on this in our privacy policy.

    Just to note - I'd be unhappy not cross connecting the high-side Cs and Rs to prevent overvoltage from differential C leakage etc. It will make the C's less reliable and I can see no strong reason for not cross-connecting. I guess the issue with doing it is that it means errors in those Rs and Cs can have a non-compensatable effect on frequency response but this is a VERY minor effect.

    Quote from THHuxleynew

    Just to note - I'd be unhappy not cross connecting the high-side Cs and Rs to prevent overvoltage from differential C leakage etc. It will make the C's less reliable and I can see no strong reason for not cross-connecting. I guess the issue with doing it is that it means errors in those Rs and Cs can have a non-compensatable effect on frequency response but this is a VERY minor effect.

    Talking about leakage, it reminded me to remind Alan about parasitic inductances, capacitances. Make sure of basic "hygiene" rules are respected, differential/symmetrical lines, no big wire loops in the high speed current path, always check the return path of current, attention to parasitic capacitances, no close parallel wires where they're not suppose to couple, at these high voltages and high dv/dt, high di/dt events many unwanted effects can take place.


    Crocodile clip quick-connect in a random spider-web fashion might not be the best way to create the circuit, regardless of how simple the circuit is.


    Please just make sure not to confuse oscillations through parasitics with the "small explosions" that suppose to give the excess juice. :thumbup:

    Been looking at details for the scope of independent replication. Started to brainstorm about cell construction and candidates for BOM. I think I found an easy cheap way and went ahead to modeling it tho show you:

    It's a draft, please imagine in the missing HV insulation and gas ports.

    Features such as see-through and gap adjustment were considered, and it should hold the gas in, I guess. I wouldn't bet any money on that gap adjustment solution, but I thought it's still better than no adjustment.


    What do you think?

    FreeCad?

Subscribe to our newsletter

It's sent once a month, you can unsubscribe at anytime!

View archive of previous newsletters

* indicates required

Your email address will be used to send you email newsletters only. See our Privacy Policy for more information.

Our Partners

Supporting researchers for over 20 years
Want to Advertise or Sponsor LENR Forum?
CLICK HERE to contact us.