Oscilloscope - safe to use any scale as long as not over max ratings?

[py-bb]

Hi there,

Mine for example says "400v DC or pk AC" - I'm assuming that means AC up to -400v to +400v right?

The real question I have though concerns "can I damage it through using a more sensitive voltage setting?"

I was testing a digital signal from 0v to +3.3v which is usually high, so I AC coupled it and now it's "mostly ground" with dips downward - letting me study the rises and falls.

I noticed eventually I was viewing this on 2mv/div signal. that'd put the -3.3v ish that the lows at >1,500 divisions down

If I indeed used a 400v signal that'd be 200,000 divisions on the 2mv scale.

This is an analogue scope.

The manual hasn't really shed any light on this, obviously at some point things will saturate, but my question is will this harm the scope?

I've deliberately not mentioned what mine is to get a better overview of the subject, I am also aware that the amplifiers may saturate and be unable to change their output fast enough to keep up with the signal - this isn't harmful in and of itself though.

[RoGeorge]

Mine for example says "400v DC or pk AC" - I'm assuming that means AC up to -400v to +400v right?

The real question I have though concerns "can I damage it through using a more sensitive voltage setting?"

The oscilloscope settings doesn't matter.  It can be at 2mV/div or at 20V/div and it is still safe to apply those 400V from the max specs.


Side note, beware when switching from DC to AC.  The capacitor inside the oscilloscope might remain charged at high voltage from a previous AC measurement, and when you switch from DC to AC the charged capacitor will put that high voltage into the circuit you try to measure.

For example if you measure first a 400Vpp in AC mode, then change to DC and measure 3.3V it's all OK (but the internal capacitor for AC might have remained charged).  Then, if while you measure the 3.3V circuit you switch the oscilloscope to AC again, that 400V charged capacitor might still be charged and put 400V in your 3.3V circuit.

The good practice is to always discharge the internal capacitor when measuring high voltage.  Many oscilloscope can not switch from DC to AC without going through GND (at which moment the presumably charged capacitor would be discharged), but not all models have GND between AC and DC modes.

[jonpaul]

Scope make/model?

Chinese knockoffs are shaky for protection.

TEK, HP, other non China are fine at any setting V/div for max V

j

[py-bb]

Mine for example says "400v DC or pk AC" - I'm assuming that means AC up to -400v to +400v right?

The real question I have though concerns "can I damage it through using a more sensitive voltage setting?"

The oscilloscope settings doesn't matter.  It can be at 2mV/div or at 20V/div and it is still safe to apply those 400V from the max specs.


Side note, beware when switching from DC to AC.  The capacitor inside the oscilloscope might remain charged at high voltage from a previous AC measurement, and when you switch from DC to AC the charged capacitor will put that high voltage into the circuit you try to measure.

For example if you measure first a 400Vpp in AC mode, then change to DC and measure 3.3V it's all OK (but the internal capacitor for AC might have remained charged).  Then, if while you measure the 3.3V circuit you switch the oscilloscope to AC again, that 400V charged capacitor might still be charged and put 400V in your 3.3V circuit.

The good practice is to always discharge the internal capacitor when measuring high voltage.  Many oscilloscope can not switch from DC to AC without going through GND (at which moment the presumably charged capacitor would be discharged), but not all models have GND between AC and DC modes.

That's a very nice "two birds one stone" putting ground in the middle like that.

The capacitors in the probe basically "bleed"/self-discharge through the resistors right? So they're a none issue (if they did store enough charge to last that long it'd be a very bad thing indeed!)

Thanks for mentioning that, it'd never occurred to me!

Scope make/model?

Chinese knockoffs are shaky for protection.

TEK, HP, other non China are fine at any setting V/div for max V

j

Really old Gould OS300 (manual: https://www.steampoweredradio.com/pdf/gould/Gould%20OS300%20Dual%20Trace%20Oscilloscope%20Manual%20and%20Schematic.pdf )

Say what you will but (other than the size) I have only 2 complaints:

Cannot trigger on the addition of the channels (if you invert channel 2, you have X minus Y)

and

Says 20mhz, really is 2mhz with a x10 expander on the x ramp. I usually deal with serial ports at a highest baud of 9600 x 48 so this isn't a problem.


Thanks for clarifying.

[EPAIII]

If the name on the scope is Tektronix or HP or another old US brand you will be OK as long as you stay within the stated max Voltage. Likewise for British, German, and probably Japanese brands.

If the CoO is China, then all bets are off.

[py-bb]

A few of you lot have mentioned knock-off Chinesium scopes.

Are we talking like the "DIY-ish" USB or USB-powered ones that I've seen on ebay? Because I can totally believe they're trash. They're also suspiciously cheap.

There are a few like SDS/OWON/FNIRSI (see https://siglentna.com/USA_website_2014/Documents/UserManual/SDS1000X&Xplus_UserManual_UM0101X-E02A.pdf here, it's sold under at least 2 brand names, same just a different name on the case) that looked alright.

I was trying to justify them as a solution to a problem I don't have. The manuals are in good English I'd have thought they'd be OK.

[rstofer]

There is no way in the world I would inject anything near 400V into a scope.  I would prefer to cut the voltage with a x10, x100 or even x1000 probe.

About the highest voltage I work with is 48VDC for stepper motors on my CNC Mill.  More often than not, I am probing 3.3VDC.  I like low voltages!

[KaneTW]

I trust a scope input to accept its rated voltage safely, but not a potential surge voltage. An oscope by itself is generally CAT I (exceptions exist like R&S Scope Rider etc which are CAT IV 600V). You'd be fine measuring a 300V regulated DC supply (although I'd still divide it down), but not a wall socket or a building installation without a suitable probe as you can have a transient overvoltage that will endanger your life and device.

[bdunham7]

With DC coupling you are obviously limited by the fact that the scope probably can only display 80V_p-p (10V/div x 8 divisions), plus you can use a certain amount of offset.  So the 400V_pk is a safety or damage limitation and you typically need to stay well below that.  In practice, most experienced users will use at least a 10X probe almost all the time and not get anywhere near the 400V limitation.

 With AC coupling, the scope will not see a DC bias so you can look at the ripple of a 300VDC power supply using the 2mV/div range without any issues.  That's fine, but if the AC coupling capacitor gets overloaded or breaks down, sparks will fly.  Unless you really, really need the sensitivity, using a 10X probe (at least) puts 9M of resistance between the DUT and your scope.  Note that the AC coupling capacitor still charges up to the full 300VDC, so the 400V limitation remains the same and the full 300VDC shows up at the input BNC.  There are specific x100/10M probes with internal dividers that avoid this issue when using scopes with this design.

Many newer design scopes have AC coupling that works differently than your older scope and these don't react the same way--AC coupling is not limited in the same way.  A 10X probe with the same 300VDC power supply on these scopes will only have 30VDC at the input BNC. 

I wouldn't worry about whether the scope is 'Chinese' or not, but I would avoid cheap junk.  I have Tektronix scopes that are made in China and I have Siglent scopes made in China and they are both protected just fine.  However, the Siglents are of the latter design that doesn't have the AC coupling limitation.  In any case, I wouldn't typically set the scope up with anything over 50V or so at the inputs, just because it is almost always unnecessary.

[py-bb]

With DC coupling you are obviously limited by the fact that the scope probably can only display 80V_p-p (10V/div x 8 divisions), plus you can use a certain amount of offset.  So the 400V_pk is a safety or damage limitation and you typically need to stay well below that.  In practice, most experienced users will use at least a 10X probe almost all the time and not get anywhere near the 400V limitation.

 With AC coupling, the scope will not see a DC bias so you can look at the ripple of a 300VDC power supply using the 2mV/div range without any issues.  That's fine, but if the AC coupling capacitor gets overloaded or breaks down, sparks will fly.  Unless you really, really need the sensitivity, using a 10X probe (at least) puts 9M of resistance between the DUT and your scope.  Note that the AC coupling capacitor still charges up to the full 300VDC, so the 400V limitation remains the same and the full 300VDC shows up at the input BNC.  There are specific x100/10M probes with internal dividers that avoid this issue when using scopes with this design.

Many newer design scopes have AC coupling that works differently than your older scope and these don't react the same way--AC coupling is not limited in the same way.  A 10X probe with the same 300VDC power supply on these scopes will only have 30VDC at the input BNC. 

I wouldn't worry about whether the scope is 'Chinese' or not, but I would avoid cheap junk.  I have Tektronix scopes that are made in China and I have Siglent scopes made in China and they are both protected just fine.  However, the Siglents are of the latter design that doesn't have the AC coupling limitation.  In any case, I wouldn't typically set the scope up with anything over 50V or so at the inputs, just because it is almost always unnecessary.

We've drifted a bit, with a large signal say like 300v the main reason not to try and view it (DC coupled) is dynamic range, if you can get it into view your'll maybe be able to tell 299v-301v say - by AC coupling you can go go a lot lower (that's the point)


My question was more about whether it'd hurt the amplifiers or the scope, exceeding the ratings means if it did it's not unexpected. I'm not trying to do that.


To re-iterate the example take a 300v digital signal switching between 0v and 300v for some reason, suppose we were only interested in viewing the bits near 0v, and we put our scope to 2mv/div

We can now very clearly see the lows of the signal and inspect any ringing and whatnot, but the 300v values are MASSIVELY out of range.

The question was can the scope damage itself here.

I'd like to know more about what protects it (is is just simply "the amps saturate and hit a maximum value"?) - I imagine it is, but I saw nothing in the manual about how it handles the amps basically being over-driven.


The main benefit for using a x10 probe in this situation is dynamic range, by dividing like that we can actually fit it on our screens say, if you step down that 300v signal by 100x say you're now going from 0v to 3v, and you have good odds of being able to use the 0.5v/div scale and offsetting it and getting a decent view of that 300v

This "happens" to be safer/easier and all that but it's not why we're doing it.


Does this make sense?

From other's thought it seems like 0-300v switching signal viewed at 2mv/div is totally fine (provided you only want to see the bottom of it!) for a scope rated say 400v (as it is in my case)


Thanks

[bdunham7]

To re-iterate the example take a 300v digital signal switching between 0v and 300v for some reason, suppose we were only interested in viewing the bits near 0v, and we put our scope to 2mv/div

OK, that involves something called overload recovery which has been discussed before in this forum.  I'll try and find a thread.  In short, when you 'overload' (meaning way off scale but not over the specified limits) the input amplifier, it does not recover instantaneously from this overload.  Some designs recover very, very quickly and can be used in the manner you are describing while others recover more slowly and less predictably.  Your old CRO Tek scopes typically recovered very quickly. 

Here's a post from a previous thread on the subject showing a Siglent SDS1104X-E with relatively poor overload recovery characteristics.

https://www.eevblog.com/forum/testgear/siglent-sds1104x-e-anomaly-on-what-should-be-a-simple-task/msg3451402/#msg3451402

[py-bb]

To re-iterate the example take a 300v digital signal switching between 0v and 300v for some reason, suppose we were only interested in viewing the bits near 0v, and we put our scope to 2mv/div

OK, that involves something called overload recovery which has been discussed before in this forum.  I'll try and find a thread.  In short, when you 'overload' (meaning way off scale but not over the specified limits) the input amplifier, it does not recover instantaneously from this overload.  Some designs recover very, very quickly and can be used in the manner you are describing while others recover more slowly and less predictably.  Your old CRO Tek scopes typically recovered very quickly. 

Here's a post from a previous thread on the subject showing a Siglent SDS1104X-E with relatively poor overload recovery characteristics.

https://www.eevblog.com/forum/testgear/siglent-sds1104x-e-anomaly-on-what-should-be-a-simple-task/msg3451402/#msg3451402

This is usual amplifier saturation right?

The key thing is it wont harm it? This is what I really want to know. Practical things like amps not being ideal are to be expected.

Thanks for finding the thread, I will read it but I'm asking about harm (and we've already agreed for the sake of dynamic range, we'd divide down anything big anyway)


Addendum: I was looking at one of those SDS scopes, they're sold under several brands but it's the same scope as a cheap digital scope with 4 channels (one of them does)  -the manuals were in good English and they had decent specs!

[james_s]

I don't actually know off hand if it applies to the scope front end too but something to watch out for with probes is voltage de-rating at higher frequencies. The higher the frequency, the more you have to worry about insulation breakdown.

[bdunham7]

The key thing is it wont harm it? This is what I really want to know. Practical things like amps not being ideal are to be expected.

AFAIK, it should not harm the amplifier as the clamping should just be shunting the input voltage through a fairly high resistance of ~1M or so.  However, I wouldn't push it to the limits just because things happen, and I certainly would be wary of any circuits with significant energy.  So looking at a 0-30V input on your 2mV scale is something I would not expect to be a problem--and as you can see from the thread I linked, that is exactly what I was doing.

[py-bb]

The key thing is it wont harm it? This is what I really want to know. Practical things like amps not being ideal are to be expected.

AFAIK, it should not harm the amplifier as the clamping should just be shunting the input voltage through a fairly high resistance of ~1M or so.  However, I wouldn't push it to the limits just because things happen, and I certainly would be wary of any circuits with significant energy.  So looking at a 0-30V input on your 2mV scale is something I would not expect to be a problem--and as you can see from the thread I linked, that is exactly what I was doing.

I just looked at the pictures in your post, wow there's a big difference between the two!

Obviously I wont hold you liable or anything should you be wrong but that's not a very definitive answer. Is the scope designed to "be OK and not harmed" by such use? Or is it implied that I shouldn't do it?

I'm reasonably sure the answer is "yes it'll be fine" (even if not that useful) simply because the amps can only amplify so much.

I was looking at those SDS (it's the same as OWON or some other company, they rebadge them) - that seems like a pretty big caveat, give it a miss?

[Psi]

Be careful with 1x/10x switchable probes.

It might be safe to use 230V mains on your scope at 10x but not at 1x and you don't want to bump the switch. It can ruin your day.

I always interpreted the 400V peak to peak as meaning 230V AC was outside the 400V peak to peak spec.
Since 230V is around +315  to -315,  so 630V peak to peak?

[bdunham7]

Obviously I wont hold you liable or anything should you be wrong but that's not a very definitive answer. Is the scope designed to "be OK and not harmed" by such use? Or is it implied that I shouldn't do it?

I'm reasonably sure the answer is "yes it'll be fine" (even if not that useful) simply because the amps can only amplify so much.

Well, there's no point in doing it if it behaves like that.  However, typical amplifier designs that are properly protected won't have any issues with this sort of treatment because of the low clamped currents as I mentioned earlier.  Derating for voltage and frequency doesn't seem to be mentioned anywhere, but I'm pretty sure no scope can take 400 volts peak at 200MHz or anything like that.

I was looking at those SDS (it's the same as OWON or some other company, they rebadge them) - that seems like a pretty big caveat, give it a miss?

No, Siglent is its own company and they're overall pretty good IMO. I highly recommend the model I mentioned as an entry-level scope with a lot of function for the money.  Nothing to do with OWON or Hantek, but they OEM lower end scopes for LeCroy and BK Precision and have a pretty good lineup themselves.  Overload recovery just isn't their thing I guess.  To be fair, I've been told that overload recovery is an issue with many modern DSO's although I haven't had the opportunity to try them all myself.  As  you can see, all of my old scopes handle my example perfectly, along with the Tek TPS which is a fairly old design at least in DSO years.  Classic sampling scopes are also virtually immune to this by design, but those aren't common at all.  The only modern manufacturer I know of that addresses this issue precisely is Cleverscope, a very expensive high-end PC based design.  They state that their overload recovery time is one sample period, or IOW it just isn't an issue with their design.



 

[bdunham7]

Be careful with 1x/10x switchable probes.

It might be safe to use 230V mains on your scope at 10x but not at 1x and you don't want to bump the switch. It can ruin your day.

I always interpreted the 400V peak to peak as meaning 230V AC was outside the 400V peak to peak spec.
Since 230V is around +315  to -315,  so 630V peak to peak?

I'm pretty sure the 400V_pk means DC + AC_pk or simply the maximum absolute instantaneous voltage, not peak to peak.  However, I still wouldn't let any mains, let alone 230V, anywhere near cheap 1X/10X probes or the direct input of the scope.  They're rated CAT I at best.  You need CAT II/300V at a minimum.

[py-bb]

Be careful with 1x/10x switchable probes.

It might be safe to use 230V mains on your scope at 10x but not at 1x and you don't want to bump the switch. It can ruin your day.

I always interpreted the 400V peak to peak as meaning 230V AC was outside the 400V peak to peak spec.
Since 230V is around +315  to -315,  so 630V peak to peak?

I'm pretty sure the 400V_pk means DC + AC_pk or simply the maximum absolute instantaneous voltage, not peak to peak.  However, I still wouldn't let any mains, let alone 230V, anywhere near cheap 1X/10X probes or the direct input of the scope.  They're rated CAT I at best.  You need CAT II/300V at a minimum.

In this case it says 400v DC & 400v AC pk to pk - although both only get within 400v of ground, it's pretty clear from "peak to peak" that 200v AC about ground is max.

[bdunham7]

In this case it says 400v DC & 400v AC pk to pk - although both only get within 400v of ground, it's pretty clear from "peak to peak" that 200v AC about ground is max.

What model scope do you have?  Your first post said something different.

[py-bb]

In this case it says 400v DC & 400v AC pk to pk - although both only get within 400v of ground, it's pretty clear from "peak to peak" that 200v AC about ground is max.

What model scope do you have?  Your first post said something different.

Yeah I misread the manual

https://www.steampoweredradio.com/pdf/gould/Gould%20OS300%20Dual%20Trace%20Oscilloscope%20Manual%20and%20Schematic.pdf#page=7&zoom=auto,-274,715

Right there it says it (left column near the top)

Either way I'd step it down

[rstofer]

To add to the confusion, the Rigol DS1054Z inputs are rated:  1 MOhm, 13 pF, 300V RMS

We can have all kinds of fun calculating the RMS of a pulse with far less than 50% on time.  Or triangle waves, trapezoidal waves and other odd shapes.

https://masteringelectronicsdesign.com/how-to-derive-the-rms-value-of-pulse-and-square-waveforms/

[bdunham7]

To add to the confusion, the Rigol DS1054Z inputs are rated:  1 MOhm, 13 pF, 300V RMS

We can have all kinds of fun calculating the RMS of a pulse with far less than 50% on time.  Or triangle waves, trapezoidal waves and other odd shapes.

Yes, but it also says 'CAT I', so the implication is that it is 300V_RMS for a sinusoid at mains frequency, not some once-per-minute 10kV pulse.  Of course CAT I also implies some transient protection and I would wonder how well it actually holds up.  Perhaps joeqsmith can start destroying oscilloscopes now.

[KaneTW]

CAT I implies *no* transient protection. CAT I is unrated.

[radiolistener]

The manual hasn't really shed any light on this, obviously at some point things will saturate, but my question is will this harm the scope?

Be careful, you can damage your oscilloscope very easily.
Note that safe input rating is specified for DC Voltage.

For AC Voltage max rating drops down when frequency rise. While at 50 Hz you can still be safe at 300 V, this is not the case at 1 MHz. For 10 MHz or 100 MHz it drops down even more dramatically. The max Voltage vs frequency curve usually is not specified by manufacturer, but it dramatically drops down above 100 kHz.

I asked Siglent about their oscilloscope max rating for 1MΩ input at RF frequencies (1-200 MHz and above), they said that the max safe rating for RF signals is about 40 V. This is much smaller than 400 V marked on input label which is relevant for DC.


The same thing with oscilloscope probes. For example 1:100 probe max rating at DC is about 2000 V, but above 40 kHz it starts to drop down. At 100 kHz it just 800 V. At 1 MHz it just 200 V. At 30 MHz it just 100 V. You can find that max voltage vs frequency curve in the probe manual.

For classic 1:10 probes things are much worse. They have max voltage 600 V up to 40 kHz, but then their rating drops down. At 100 kHz it just 200 V. At 1 MHz it just 50 V. At 30 MHz it just 25 V. So, you can burn it if you connect it to transmitter output with 50-100 V (25-100 Watt) at RF frequency.


I recommend to not use direct oscilloscope connection for circuit which can have above 40 V. Always use proper probes with divider for high Voltage, or attenuator with proper power rating. If you burn your probe it will cost much less than repair of burned oscilloscope input.

[py-bb]

I forgot my own question.

I just came across why I asked in the first place.


Consider this case:

We apply the same signal to both channels, invert channel 2 and add them (so now we have (ideally) a flat line of ground potential)

How does that scale?

When I do this it doesn't seem to be just noise and I wonder how it's doing it. Remember it's analogue!

[oldjackbob]

CAT I implies *no* transient protection. CAT I is unrated.

Not sure where you got that info, but at https://wiraelectrical.com/measurement-category/ it says that CAT1 has a transient voltage rating of at least 4 times the rating of the device in question:

[David Hess]

The real question I have though concerns "can I damage it through using a more sensitive voltage setting?"

The general answer is that the maximum input voltage is not affected by the sensitivity setting, but this may not apply to cheap poorly designed oscilloscopes.

If I indeed used a 400v signal that'd be 200,000 divisions on the 2mv scale.

Some oscilloscope inputs are designed to work exactly like that.  The Tektronix 7A13 differential comparator operates as a normal oscilloscope input but with 20,000 divisions of position range.  It can literally display 1 millivolt/division riding on top of 10 volts.

[jonpaul]

bonjour cher Monsieur, just seeing this long topic now.

Many questions are answered by the great  classic Tektronix Oscilloscope Circuits Concepts book  Oscilloscope Vertical amplifiers

https://w140.com/tekwiki/images/f/fd/062-1145-00.pdf

Especially see overload recovery, protection, coupling,...

We have worked since 1970s with SMPS, HV to 75 kV,  12 kw arc lamps, radar pulse transformers.

Used mostly Tektronix scopes, Yokogawa digital.

For probes, 1x, 10x,  100x, current probes, Zo probes.

Never blew a scope, or probe.

No experience with ChiCom junk, or USB like Picoscope.

Just the ramblings of an old retired EE

Jon