Hi!

Looks like there is a noise signal around. Use your scope to measure the frequency, looks like MHz-range. The frequency may help identify what it is, then you can try and remove it. You could also use a shielded environment (like a 20ft container).

regards

You can use a probe for oscilloscope for DC and AC input. It works on an effect of Hall (Hall probe).

Dear Vladimer Mikelashvili

there are different approaches to do that. you need to assess their associated benefits and drawbacks to select the method.

the selection criteria is "what do you want to measure?". only the pick of DC high voltage? the waveform? the rise-time and fall time? the overall wave form.

the simplest, and the most popular, method is to sample the high-voltage using a resistor network and calculate the actual high voltage. the drawback is the elimination of low-voltage waveform.

the most complicated method is the use of Logarithmic amplifier and considering the logarithmic calculation to unveil the real measured values.

NOTE: application of HIGH-VOLTAGE for measurement or usage requires essential SAFETY Consideration. please refer to safety handbook for your safety.

Best Regards

Ahmad

Thank You for answers. Dear Dr. Ahmad Esmaeilkhah i need to measure voltage drop in time. when i rise the voltage rises only amplitude of sinusoidal curve. is there any mathematical equation to convert sinusoidal signal into single curve?

Dear Vladimer

Honestly, i did not understand what you mean. do you want to the "break down delay" as illustrated in attached figure?

Dear Dr. Ahmad Esmaeilkhah see attached pic1. As you see during nanosecond discharge current getting high around handreds of amphere and voltage drops.after drop voltage value restores until next discharge. I measure current and is no problem but voltage curtve oscilates althou it wortks.drop is appartent. I need one line as it is on pic 1.

I believe the first thing you need to find out is: is this really part of your experiment or is it external noise?

Since it is a pretty steady sine wave which is superimposed on your actual measurement I suspect some external noise.

What does your test setup look like?

Try to short probe tip to probe ground- does the noise disappear?

Can you share a schematic, drawing, or photo of the test set up?

A few items to consider.

The ground of the Oscilloscope must be the same ground reference of what you are measuring. Power cord Earth ground is too long. For example the scope ground should be connected to your high voltage power supply ground with a short wire.

The scope and probe bandwidth are very low for nanosecond measurements. Better to be at least 500MHz, best at 1GHz.

And consider the high voltage electric fields being generated are effecting the probe, cables, and scope and require careful routing and shielding.

I know the high voltage prove had a high value resistor divider network built it. I'm not familiar with the probe to know the impact on measuring.

Deart Michael Stevens, thank for answer.before prtepare pics ort video to upload, i will try express experimental setup. We ave 220v, minus and ground. Power ccord of oscilloscope and high voltage probe is grounding in the same cable. Before i turn on the high power suply i simply touch to the high voltage probe tip and signal oscilates already. If i take ordinary 9 volt battery and cconnect oscxilosxcope probe grounding on minus and tip on the plus, oscilloscope automaticcally detecting dc voltage. Stright line is moving up during connecting the battery and come down when discxonnect. It good.but when i remove one probe electrode from battery, it begins oscillating.

Hello Vladimer, does your high voltage have a ground reference or is it floating?

here is experiment. see movies

The picture 2.jpg shows about 50MHz interference, while videox.43 shows line frequency interference. You must either lower the impedance of your measurement system (possibly causing other problems), filter out the interference (high-pass filter for power line interference), or isolate your system from the environmental interference. If you were in a Faraday cage you would probably not see these problems unless your test setup is the cause of your noise. as Michael Stevens mentioned, you may have a scope ground-reference problem, since your experiment shows connection to the battery gives a clean measurement. What conditions were present during that measurement (before you touched the battery) that did not show the interference?

Unless you are familiar with making high-speed measurements, the most likely source of error will be improper termination of your cables. The 50 MHz ringing you see is characteristic of an unterminated cable of about 10 nsec electrical length -- about 2 meters with standard coaxial cable. The simplest thing to do is to attach the center cable of the coax to the point you want to measure through a resistor, then put a 50 ohm impedence on the other end of the cable (the oscilloscope or digitizer). The value of the resistor should be chosen so that your signal is about 1 volt at the 50 ohm load. If your signal is about 1000 volts, that means the resistor should be about 50 kohm.

This method will give moderately good time response -- microseconds or so. If you need to do better, you should use a capacitive rather than resistive voltage divider.

Or, if you have the money, just go buy a professionally-engineered high-voltage probe.

Dear R.D. Horton see attached my high voltage probe. should i use resistor anyway?

today i have experiment with discharges in water. see my results to better understand my problem

this is directly from osciloscope

i grounded oscilloscope also

With the scope showing 100ms/div and the "oscillation" showing 5 cycles (at about 3kV p-p), there is either a serious 50Hz component to your H-V power supply, or there is still a significant coupling from the local power line environment into your measurement system. Does the noise go away if you touch the scope probe tip to the point where the scope ground lead touches your measurement system? if it does not go away, there is something wrong with your probe.

I'm glad your waveforms are better with altered grounding, but I see you still have some noise.

Looking at the lower-right figure in Untitled-1.jpg, I count about 10 cycles per .08 seconds, or 125 Hz -- pretty close to 120 Hz. That's the frequency where I would expect ripple in a full-wave rectified DC supply. Your figure doesn't show the values of R1 and C1, but the product should be >.1 second (e.g. R1 = 10 ohms, C1 = 100 uF) in order to really knock down the ripple component. I don't fully understand your experimental setup, but if the quantity measured is effectively AC coupled, it will be more sensitive to any 120 Hz ripple than you might expect just by looking at the DC supply output.

One thing that would make your DC voltage cleaner, if this is actually the problem, would be to replace R1 by an inductor L1. You'd like to have the product LC to be > .01 second^2 (e.g. L1= 1H, C1 = 100 uF). Even better would be to use a regulated power supply. You should see no signal at when you're not pulsing.

Dear R.D. Horton R1=51 kohm, 10 w and C1=0.1 microF, 15 kV

Dear Carl Van Wormer see the pic