Oscilloscope/Probe voltage derating confusion

[Kaylan]

Hello everyone, I've owned my first oscilloscope for about a month now and I'm trying to clear up some confusion on the voltage derating of scope probes. I understand that a probes voltage rating will decrease with higher frequencies. Because the series capacitor in parallel with the resistor will offer less impedance as the frequency increases. So if you were to measure a signal that has a voltage higher than the probe is rated for at that particular frequency, can you damage the probe, or the scope, or both? Do oscilloscopes have protection from over current?

[Kim Christensen]

Usually it's the probe that dies. I guess with a severe overload, the scope could be damaged as well.

[pdenisowski]

Depending on how high the voltage is, you can damage the scope, the probe, and/or yourself.

[radiolistener]

So if you were to measure a signal that has a voltage higher than the probe is rated for at that particular frequency, can you damage the probe, or the scope, or both? Do oscilloscopes have protection from over current?

No there is no protection in oscilloscope. But usually it leads to a probe damage only. However if you're unlucky it may damage and your oscilloscope. Usually high impedance input of oscilloscope has about 40 Vpk limit for RF.

[Kaylan]

So if you were to measure a signal that has a voltage higher than the probe is rated for at that particular frequency, can you damage the probe, or the scope, or both? Do oscilloscopes have protection from over current?

No there is no protection in oscilloscope. But usually it leads to a probe damage only. However if you're unlucky it may damage and your oscilloscope. Usually high impedance input of oscilloscope has about 40 Vpk limit for RF.

Observing the RF output of my radios is the main reason i even bought a scope in the first place. I educated myself as best i could and learned a LOT before the scope was even delivered. However, i think voltage derating is a subject that doesn't get brought up very often when talking about input limits. Mostly, the absolute maximum is all you ever see. I was very familiar with probe circuitry, but it just didn't register until almost to late that frequency plays a major role in input limits. My stryker SR-955HP 10 meter radio is capable of about 15 watts output. And once or twice, i was stupidly measuring the output directly from the antenna connector with a 50 ohm dummy load connected. I was seeing somewhere around 60V and cant remember if this was pp or the amplitude, but i don't think it was rms. Most likely well outside of the probes limit at 27.205MHz which is somewhere around 20V rms at that frequency according to the manual. Luckily it was only a few short measurements and i don't think anything was damaged. Major trap for new players. 

[radiolistener]

You can use 1:100 probe (2000V at DC), it has better limit at RF, about 50V or even 100V at low frequency band.

Another way is to use high power RF attenuator. It allows to see signal with less distortions like it happens with probes.

[Alex Nikitin]

There are some probes more suitable for this kind of application. I use ELDITEST GE3425 (GE3421 and GE3422 are OK too in this respect)



Cheers

Alex

[DavidAlfa]

Sure, very high frequencies would cause a low impedance path through the caps, overloading the probe and/or oscilloscope input.
As you see, there's about 30pF, so at 10Mhz this would be ~500Ohms.


But here would you measure 500V+ and several MHz? High voltage output in a broadcast station?
Otherwise it's very rare to find >1MHz high voltage (SMPS etc), so needing to think on probe derating is a rare situation in everyday use.

A scope is pretty much useless for RF, you'll see "something", but that's all, no measurements for power, armonics, modulation...
You need a RF analyzer for that!

[golden_labels]

Another similar thread exists, and T3sl4co1l provided a longer post on the issue. May be of interest to you people in this thread.

DavidAlfa:
A MHz range pure sine wave at 500 V? Perhaps rare.

But what about a 333 kHz, 1.5 kV square wave? The second component is 500 V at around 1 MHz, then higher frequency components at 300 V, 214 V, 167 V and so on. The power from them don’t simply sum, as they partially cancel each other out, but it’s not negligible either.

I don’t want to say this is an issue. I don’t know. Just noting, that the particular argument may not be something to be taken as obviously true.

[DavidAlfa]

Sure, but again not your average scope stufff, for 1.5KV you'll need specialized probes anyways, which you should RTFM before use to not die or kill your expen$$$ive probe .
For the remaining 99%, it'll be the usual ~600V 10x probe. I have no idea if these have any frequency derating, but again the required conditions are far for common.

[golden_labels]

In the other thread we have an example of derating for a 10:1 probe rated for 300 V.

[BrokenYugo]

Even the cheapo switchable 10:1 probes I don't trust at any dangerous voltage level typically include a derating curve on the datasheet in the bag along with the other accessories. IIRC most start to drop just past audio frequency.

[Kaylan]

You can use 1:100 probe (2000V at DC), it has better limit at RF, about 50V or even 100V at low frequency band.

Another way is to use high power RF attenuator. It allows to see signal with less distortions like it happens with probes.

This is exactly what i was considering. A 1:100 probe will be next on my shopping list. Will also shop RF attenuators if the budget allows. Expansion of ones probe collection can never be a bad thing.

A scope is pretty much useless for RF, you'll see "something", but that's all, no measurements for power, armonics, modulation...
You need a RF analyzer for that!

But isn't the 11 meter band in America between 26.965 and 27.405 well enough below the -3db point for a 100MHz scope that decently accurate measurements can be taken? Then manually calculate average/peak envelope power if one were without the equipment to measure it? Modulation can be observed quite well for reference purposes IMO. You can definitely see if your radio is over/under modulating and make adjustments. As for harmonics, i have the Siglent SDS812X HD so wouldn't the 2Mpts FFT get you a vague idea of where your power is going? At any rate, you're right, one could definitely get better measurements from specialized test equipment. But i'm a tinkerer on a tight budget. Way to tight for this kind of hobby but i've been into electronics since i was a kid and i enjoy it.

There are some probes more suitable for this kind of application. I use ELDITEST GE3425 (GE3421 and GE3422 are OK too in this respect)



Cheers

Alex

Not a bad price, thanks Alex.

Even the cheapo switchable 10:1 probes I don't trust at any dangerous voltage level typically include a derating curve on the datasheet in the bag along with the other accessories. IIRC most start to drop just past audio frequency.

Neither do i at this point. The Siglent PP510's that came with my scope start at 300V rms and drop off rapidly. But the PP510's i just looked up on there website are completely different according to the data sheet, they start at 600V and don't start dropping off until around 20kHz higher than the ones that i have. 

In the other thread we have an example of derating for a 10:1 probe rated for 300 V.

That is a really close match to the siglent PP510's that came with my scope i believe. BUT, not the ones you find in a search for the same model on their website.

Back to the subject at hand. I took the time to dig further through the spec sheet https://siglentna.com/wp-content/uploads/dlm_uploads/2024/06/SDS800X-HD_Datasheet_EN01D.pdf for my scope and found: Max. input voltage 1 MΩ ≤ 400 Vpk (DC + AC), DC ~ 10 kHz. So i'm still wondering about frequencies greater than 10kHz directly on the scopes inputs. However, one would do good to keep within the limits of the probe you're using as the probes job is to attenuate high voltages. And as others have stated, use the proper attenuation when dealing with high frequency + high voltage measurements. Know your probe, know your equipment, when in doubt, attenuate it out.

[CaptDon]

The killer of probes is that you have the main series capacitor (in parallel with a resistor) and a bigger capacitor on the scope end which includes a fixed capacitor, a trimming capacitor for 'compensation' and those are shunted by the scopes input capacitance. As you go up in frequency the capacitive string to ground shows a lower and lower impedance (Xc) and conducts more current which will load your circuit under test and stress the probe components. Generally speaking. the resistors survive since their resistive value barely changes with frequency. The series capacitor toward the probe tip usually dies first as it is often physically smaller and has the most voltage drop across it as compared to the larger shunt capacitors. As an example with a 100VRMS RF signal and assuming a 10:1 probe the series capacitor has 90VRMS across it and the shunt capacitors on the scope side of the probe only see 10VRMS. Scope damage usually occurs when the series capacitor in the probe being over stressed punches through and shorts out putting the full signal under test right into the scope!!! I strongly disagree with the statement "A scope is pretty much useless at R.F.". I have been using scopes 90% of my life for R.F.!!! My 500MHz TDS644B can examine 146.97MHz cycle by cycle!!! I have a 1KW dummy load with a 20dB drop (10:1 voltage ratio / 100:1 power ratio) frequency compensated tap and 50 ohm output. The output side (the tap) is also an excellent 10 watt dummy load if you use the tap as an input. I use a short cable between the scope and the load and set the scope for Hi-Z NOT 50 ohms!!

[radiolistener]

In the other thread we have an example of derating for a 10:1 probe rated for 300 V.

I just checked datasheet for PP510 100 MHz 1:10 600V probe and it shows curve which drops down below 300V at 80 kHz, drops down below 50V at 1 MHz and drops down below 25 V at 100 MHz.

If you don't know highest frequency on the signal, you can expect absolute maximum rating for 1:10 probe at about 25 V. Which is about 1.56 Watt for non-terminated 50 Ω transceiver output. Or 6.25 Watt for properly terminated output.

So you can safely use 1:10 probe to measure transceiver output if it's output don't exceed 1 Watt. If you are sure that output is properly terminated, then you can use it for transceiver output up to 6 Watt.

For 100 Watt you need to use powerful attenuator.

If you use 1 kW or more output, your probe will emit magic smoke in less than 1 second. Its already tested 

[Kaylan]

In the other thread we have an example of derating for a 10:1 probe rated for 300 V.

I just checked datasheet for PP510 100 MHz 1:10 600V probe and it shows curve which drops down below 300V at 80 kHz, drops down below 50V at 1 MHz and drops down below 25 V at 100 MHz.

That PP510 is not the same as my PP510. Or so it would seem  See attached image 1000005825.jpg
However, my probe would appear to be ok at or below 20V rms at around 27MHz.

In the other thread we have an example of derating for a 10:1 probe rated for 300 V.

If you don't know highest frequency on the signal, you can expect absolute maximum rating for 1:10 probe at about 25 V. Which is about 1.56 Watt for non-terminated 50 Ω transceiver output. Or 6.25 Watt for properly terminated output.

So you can safely use 1:10 probe to measure transceiver output if it's output don't exceed 1 Watt. If you are sure that output is properly terminated, then you can use it for transceiver output up to 6 Watt.

For 100 Watt you need to use powerful attenuator.

If you use 1 kW or more output, your probe will emit magic smoke in less than 1 second. Its already tested 

I assume that a setup like mine with a properly terminated output would work for stock 4 watt radios as the probing setup is basically in parallel with the terminated system.
Image 1000005832.jpg shows this (horrible, but functional) "RF sampler" i made from some copper, a chassis mount connector and a piece of RG-316 with an SO-259 connector on the end. The trusty RigExpert AA-600 tells me its good up to about 100MHz. After that the SWR gets a little rough. You also see a white wire coiled around the center conductor. Thats an add-on i did after i realized i might put to many angry pixies into my scope. You can put the ground lead of your probe on the sampler, then probe the wire for a greatly attenuated, but visually equivalent sign wave on a scope. Lastly, image 20240830_160505.jpg (Why didn't i rename these  ) shows a basic model, that is not to scale, or painted, but if stared at for long enough could possibly give a little more insight into whats going on.

[pdenisowski]

I strongly disagree with the statement "A scope is pretty much useless at R.F."

I think most top-tier scope manufacturers would agree with you   I know we (R&S) sell plenty of scopes for RF applications.

[Doctorandus_P]

I just had a look at my PP510 probes. These are the standard probes delivered with my Siglent SDS 1104X-E. They also have a graph with a curve similar to the other probe shown above. 600V at low frequencies, but at 100MHz it's down to only 25V.

https://siglentna.com/wp-content/uploads/2020/02/PP510_Users_Guide_082017.pdf

[Kaylan]

I just had a look at my PP510 probes. These are the standard probes delivered with my Siglent SDS 1104X-E. They also have a graph with a curve similar to the other probe shown above. 600V at low frequencies, but at 100MHz it's down to only 25V.

https://siglentna.com/wp-content/uploads/2020/02/PP510_Users_Guide_082017.pdf

I noticed that those probes have the compensation cap in the probe. On mine, it is in the BNC connector. And as said before, are only good for 300V RMS. Yet, they are labeled PP510 100MHz just like the others. I don't know whether to be happy or upset. It's kind of like a genuine scam i got caught up in really. 

[Kaylan]

I strongly disagree with the statement "A scope is pretty much useless at R.F."

I think most top-tier scope manufacturers would agree with you   I know we (R&S) sell plenty of scopes for RF applications.

I don't suppose R&S would have a freebie available for someone who probably should have shopped from there store in the first place, would they? 

[tautech]

I just had a look at my PP510 probes. These are the standard probes delivered with my Siglent SDS 1104X-E. They also have a graph with a curve similar to the other probe shown above. 600V at low frequencies, but at 100MHz it's down to only 25V.

https://siglentna.com/wp-content/uploads/2020/02/PP510_Users_Guide_082017.pdf

I noticed that those probes have the compensation cap in the probe. On mine, it is in the BNC connector. And as said before, are only good for 300V RMS. Yet, they are labeled PP510 100MHz just like the others. I don't know whether to be happy or upset. It's kind of like a genuine scam i got caught up in really. 

FYI, latest versions of PP510 have the compensation trimmer in the BNC.

[Doctorandus_P]

My stryker SR-955HP 10 meter radio is capable of about 15 watts output. And once or twice, i was stupidly measuring the output directly from the antenna connector with a 50 ohm dummy load connected.

15Watt into 50 Ohm is 27.39V (DC or RMS) For a sinewave, that's 38.3 Volt peak. And that is a bit above the official specifications of my PP510 probes.

I noticed that those probes have the compensation cap in the probe. On mine, it is in the BNC connector. And as said before, are only good for 300V RMS. Yet, they are labeled PP510 100MHz just like the others. I don't know whether to be happy or upset. It's kind of like a genuine scam i got caught up in really. 

FYI, latest versions of PP510 have the compensation trimmer in the BNC.

There is nothing wrong with having the trimmer in the BNC. If anything, it is regarded as a better option. It's not a scam.

I just noticed,  that the labels in the frequency range are completely bonkers for my PP510 probes:


This is only in the PDF, it looks normal on the dead tree thing I got with the probes.

[Kaylan]

There is nothing wrong with having the trimmer in the BNC. If anything, it is regarded as a better option. It's not a scam.

I had heard somewhere that having the trimmer in the BNC was better, it's just that having half the voltage rating was a little bothersome.

I just noticed,  that the labels in the frequency range are completely bonkers for my PP510 probes:
(Attachment Link)

This is only in the PDF, it looks normal on the dead tree thing I got with the probes.

What exactly are you referring to here?

[MathWizard]

Does anyone make probes with added inductance to add impedance back at higher frequencies? Or does that just add more unwanted ringing, or oscillations ?

[Doctorandus_P]

If you want to increase the impedance of the probe, you can also just cut the cable. 
This may also add some attenuation at all frequencies.