100x oscilloscope probe voltage ratings confusion

[Agent24]

I want to buy a basic 100x probe, currently for the purpose of looking at SMPS primary side waveforms, including those in the APFC circuitry, and maybe other stuff in the future.
I don't want to put my faith in the undocumented properties of an AliExpress special. But at the same time, I'm not getting the best information from RS or Farnell, either.

As I understand it, the probe must be rated not only for the voltage being probed (taking into account frequency derating etc), but also should be at least CAT II rated due to being used in the mains side of an appliance, and thus potentially being exposed to transient voltage spikes.

My confusion and annoyance comes from the conflicting ratings of a bunch of probes I'm trying to compare, and also the measurement category ratings themselves. Perhaps I'm just misunderstanding them.

I've noted: http://www.ni.com/white-paper/5019/en/ which lists a small voltage table and an example, and which seems to make sense in describing the basic concept.


However.... 


Take for example this probe: http://nz.rs-online.com/web/p/oscilloscope-probes/8251215/

The RS page claims "Rated voltage 1000V, CAT III (600V CAT IV) (Max.3540 V, CAT I)" - and yet the datasheet makes no claims of it being CAT IV rated whatsoever. In fact, the only mention of a measurement category rating in the datasheet is beside a derating curve titled "Max. 300 V, CAT III" showing a maximum voltage of 300v at DC - right next to another graph titled "Max. 1.000 Vr.m.s." showing a maximum of .. 1kV.

Can the thing actually work at 1kV? Or only 300V? What about transients? Where did that CAT IV rating come from? Thin air?
And what about the 3540V CAT I rating? Are they trying to tell me I can measure 3.5kV with it provided I only use it in a CAT I environment? I thought it was a 1kV probe? Or is it only really 300V? 

Am I just genuinely ignorant of something here, or am I being fed a blob of marketing bamboozlement which is working as intended?

[Dubbie]

Just buy one of Daves high voltage differential probes. I use mine all the time for repairs and messing about with mains stuff. The differential is great for avoiding shooting current down your ground lead by mistake.

[Agent24]

I would, but it's a bit out of my budget right now. I do already have an isolation transformer, at least.

[Someone]

Can the thing actually work at 1kV? Or only 300V? What about transients? Where did that CAT IV rating come from? Thin air?

The manufacturer specifies it for 1000V peak working (rms) but it looks like RS have copied over the spec for the Isoprobe II 100:1 probe instead which does go into CAT IV etc.

[tautech]

It's not just about the probe, you are constrained by the CAT rating of the whole measurement system.
If the probe ratings are not high enough to give you confidence in the product go to 1000:1 probes.

What is often sadly lacking online is derating graphs and for HV probes, and in fact all scope probes, it's a major oversight.

I've got a few Pintek 300 MHz 100:1 probes and can whip the RS price if they'd interest you....drop me a line.
http://www.pintek.com.tw/files/pintek/CP3308R.pdf

[tggzzz]

I want to buy a basic 100x probe, currently for the purpose of looking at SMPS primary side waveforms, including those in the APFC circuitry, and maybe other stuff in the future.

Make very sure you look at and understand the voltage-vs-frequency specifications.

The probe/scope manufacturers provide good solid reliable information on how to select probes and use them safely. Since there are subtle traps for the beginner and intermediate user, you would be well advised to read learn and understand what's in those documents. I have absolutely no intention of poorly regurgitating bits of safety recommendations and practices. Some starting points can be found at https://entertaininghacks.wordpress.com/library-2/scope-probe-reference-material/

Note that isolating transformers don't "make HV measurements safe"; there are other threads on this forum about that!

[Agent24]

It's not just about the probe, you are constrained by the CAT rating of the whole measurement system.
If the probe ratings are not high enough to give you confidence in the product go to 1000:1 probes.

What is often sadly lacking online is derating graphs and for HV probes, and in fact all scope probes, it's a major oversight.

I've got a few Pintek 300 MHz 100:1 probes and can whip the RS price if they'd interest you....drop me a line.
http://www.pintek.com.tw/files/pintek/CP3308R.pdf

Damn, I must have missed that point. My best oscilloscope is rated CAT II 400Vp. There is no derating curve for that, either. Does this mean I really need a new oscilloscope, or the differential probe? The PFC voltage in an SMPS is typically under 400V, I know that, but I guess it doesn't leave much room for higher voltages under possible fault conditions.

Those Pintek look OK, but does it matter? Maybe my equipment does not measure (haha) up. Damn.

I want to buy a basic 100x probe, currently for the purpose of looking at SMPS primary side waveforms, including those in the APFC circuitry, and maybe other stuff in the future.

Make very sure you look at and understand the voltage-vs-frequency specifications.

The probe/scope manufacturers provide good solid reliable information on how to select probes and use them safely. Since there are subtle traps for the beginner and intermediate user, you would be well advised to read learn and understand what's in those documents. I have absolutely no intention of poorly regurgitating bits of safety recommendations and practices. Some starting points can be found at https://entertaininghacks.wordpress.com/library-2/scope-probe-reference-material/

Note that isolating transformers don't "make HV measurements safe"; there are other threads on this forum about that!

I think I understand them... the probes derate in voltage capability as the frequency of the measured signal goes up - most of the ones I've been concerned with seem to top out their maximum voltage at about 2MHz. As far as I can see, though, that shouldn't be a problem, considering that SMPS switching frequency is only 200kHz or less, usually.

The isolation transformer remark was only about what Dubbie said, in regards to accidental current through the ground lead. I don't believe for a second that using one will magically make HV safe.

[tautech]

It's not just about the probe, you are constrained by the CAT rating of the whole measurement system.
If the probe ratings are not high enough to give you confidence in the product go to 1000:1 probes.

What is often sadly lacking online is derating graphs and for HV probes, and in fact all scope probes, it's a major oversight.

I've got a few Pintek 300 MHz 100:1 probes and can whip the RS price if they'd interest you....drop me a line.
http://www.pintek.com.tw/files/pintek/CP3308R.pdf

Damn, I must have missed that point. My best oscilloscope is rated CAT II 400Vp. There is no derating curve for that, either. Does this mean I really need a new oscilloscope, or the differential probe? The PFC voltage in an SMPS is typically under 400V, I know that, but I guess it doesn't leave much room for higher voltages under possible fault conditions.

Those Pintek look OK, but does it matter? Maybe my equipment does not measure (haha) up. Damn.

IMO if you have worries of CAT ratings you need go to a higher V rated probe and then it seems CAT ratings are no longer stated.
The higher the division rating of the probe the safer you and your scope will be.
I've always used 100:1 probes for anything near NZ mains voltages and higher but many after studying the probe derating curves will just use 10:1 probes. 
Knowledge of the risks is key and also when working with switchers, consideration of back EMF spikes.
Good SMPS design usually incorporates snubbers to control back EMF however care must always be taken with anything near mains potential and low impedance source.

HV Differential probes do make things easier and remove a lot of the chance of error.
I stock the Pintek DP-25 too if you want to go in that direction.

[Hydron]

One key thing to remember is to not use AC coupling when both of the following are true:
- The probe input resistance (in Mohms) is the same as the probe division factor (i.e. 100x probe with 100M input impedance)
- The DC voltage on the circuit is greater than the scope's rated DC input voltage

This is because in most scopes, AC coupling removes the input divider DC path to ground, so the full DC voltage will be present over the AC coupling cap (AC division still works as normal). This is a recipe for blown front-ends! Probes with a lower resistance than stated above already have an internal divider, and thus a DC path to ground regardless of scope coupling.

On another note, I've got a couple of these and use them on mains occasionally: http://uk.rs-online.com/web/p/oscilloscope-probes/4376581/
They're cheap and seem to do the job, though the sprung hook isn't help onto the probe tip that securely (just with friction, rather than with a clip action). NZ price for them is bollocks though, with a good 50% fuck-you factor added on (UK-based Kiwi here, with far too much experience of getting ripped off for little reason).

[Agent24]

I'm not significantly worried about the CAT ratings, I'm more worried that I am misunderstanding them, or other aspects.

The Pintek CP-3308R for example does state a rating of 1.5kV CAT II. Going by the table at http://www.ni.com/white-paper/5019/en/ I would expect it can handle transients of at least 8kV (maybe more?), and so it seems as if it would provide adequate transient protection (although I suspect any transients would be attenuated\absorbed by the SMPS input protection circuitry anyway). Without the datasheet saying so, though, I do not know for sure. Do you know?

I don't see a big reason to worry about using a 100:1 probe on any of my oscilloscopes and probing 400V or so in an SMPS. The divided voltage should be quite low, as long as I use DC coupling as Hydron states. (I already read about that sneaky fact) Therefore, I don't really see a problem in terms of input voltage with what I want to do, provided the probe is good quality and does what it says it does. The Pintek CP-3308R being good for 2kV up to 2MHz seems like it would be sufficient.

I suppose I just want to check that I'm not making a mistake in my reasoning\understanding.

[tautech]

I'm not significantly worried about the CAT ratings, I'm more worried that I am misunderstanding them, or other aspects.

The Pintek CP-3308R for example does state a rating of 1.5kV CAT II. Going by the table at http://www.ni.com/white-paper/5019/en/ I would expect it can handle transients of at least 8kV (maybe more?), and so it seems as if it would provide adequate transient protection (although I suspect any transients would be attenuated\absorbed by the SMPS input protection circuitry anyway). Without the datasheet saying so, though, I do not know for sure. Do you know?

No I don't for sure but to meet a CAT II rating it will have been subjected to at least the transients outlined in the CAT II tests. Somewhere online there should be the specs for such CAT tests.

[Someone]

I'm not significantly worried about the CAT ratings, I'm more worried that I am misunderstanding them, or other aspects.

The Pintek CP-3308R for example does state a rating of 1.5kV CAT II. Going by the table at http://www.ni.com/white-paper/5019/en/ I would expect it can handle transients of at least 8kV (maybe more?), and so it seems as if it would provide adequate transient protection (although I suspect any transients would be attenuated\absorbed by the SMPS input protection circuitry anyway). Without the datasheet saying so, though, I do not know for sure. Do you know?

No I don't for sure but to meet a CAT II rating it will have been subjected to at least the transients outlined in the CAT II tests. Somewhere online there should be the specs for such CAT tests.

Also the high voltage withstand voltages for the tests only require safety is maintained, the probe can still be destroyed beyond repair as long as the user is safe. When the probe says it has a working voltage of 2kV, exceeding that may damage it, or it might just fail to meet some specifications on voltages greater than 2kV.