EEVblog Electronics Community Forum
Products => Test Equipment => Topic started by: IDEngineer on September 11, 2020, 09:56:27 pm
-
Just ran into something *I* consider wrong, but I don't have another digital scope here to compare.
I'm using a Rigol DS4024. Generally I'm very happy with the scope AND with Rigol. Today I connected a current meter and wanted to change that channel's probe ratio so the current had a more direct relationship to the volts/division display on the screen.
To my astonishment, changing the probe ratio made no difference to the displayed amplitude whatsoever. x1, x10, x1000, the (freshly captured each time) waveform never changed in its vertical amplitude. As far as I can tell, the only thing that changes is the icon at the bottom of the screen that indicates V/div for that channel.
The current probe connects via a standard BNC, so there's no funny business about interacting with the smart probe pads. To confirm, I connected a function generator and used a pure, continuous sine wave of known amplitude. Sure enough, I can vary the probe ratio anywhere I like but the displayed waveform does not change size at all.
The worst part about this is that the indicated V/div DOES change. Literally this means you can configure the scope to display BAD DATA. If you take a screenshot and look at it sometime later, it will display INVALID DATA and there's absolutely no way to know that.
As a concrete example: I set the sine wave for 4Vpp and the probe ratio at x1. The screen shows a sine wave that is 4 divisions tall. I then set the probe ratio to any other value. The displayed size of the waveform does not change on the screen. I tried setting the probe ratio to x1000 and the icon at the bottom of the screen indicated 1000V/div, so the screen is literally reporting a 4KVpp sine wave.
This is VERY easy to do (IMHO) properly. On a digital scope the vertical sensitivity is set via a rotary encoder, so the firmware is already involved and controlling the front end attenuator. It would be child's play to also consider the probe ratio, and then always make certain that the attenuator and the on-screen sensitivity icon agree with each other. Instead, for a standard BNC connection the probe sensitivity literally does nothing but screw with that channel's icon - nothing more.
I don't have another digital scope here to compare. I called Rigol and they were super helpful, super friendly, but stated this is intended behavior. The guy couldn't argue with me when I said it allows the scope to lie to the user. I can't rationalize the ability to make a scope screen outright lie to you.
Is this considered "normal"?
-
To my astonishment, changing the probe ratio made no difference to the displayed amplitude whatsoever. x1, x10, x1000, the (freshly captured each time) waveform never changed in its vertical amplitude. As far as I can tell, the only thing that changes is the icon at the bottom of the screen that indicates V/div for that channel.
...
Is this considered "normal"?
now you know ;D it is super normal you should understand how it work.. if you put 1X probe on 1X dso, you see 1X signal... change the probe to 1/10X type of probe leaving dso at 1X, displayed signal will go down to 1/10X than actual signal on circuit (before probe attenuation). set the dso to 1/10X, voila! the correct signal as on the circuit... signal still the same in display, just V/div is readjusted. or maybe your idea is...
I said it allows the scope to lie to the user. I can't rationalize the ability to make a scope screen outright lie to you.
if you measure BNC end of your 1/10X probe with a Fluke DMM. your Fluke DMM will be measuring 1/10X than whats really is, so your Fluke DMM is lying ;) its you that need to keep track of that in what game you are playing. you want automation? that the dso will do that for you? fine, higher end dso have those many pins near BNC input, in fact yours also have. one of the intention is so the compatible probe will tell dso what kind of probe it is, dso will automatically change to the appropriate attenuation, you cant even change that you cant make the dso lies. but how about cheap 3rd party probes that dont have those pins? dso wouldnt know so it lets you change accordingly, thats a bless actually, otherwise you'll need the compatible aka original Rigol current probe designed specifically for your dso. you use that, your dso cant lie anymore, thats good for you and good for Rigol too ($$$)
to change that channel's probe ratio so the current had a more direct relationship to the volts/division display on the screen.
maybe the trick, or one of the trick you are looking for is change the V/div setting in fine grain until the signal has a direct relationship with the graticules imagining it with some nice division/integer number.. ymmv.
-
Just ran into something *I* consider wrong, but I don't have another digital scope here to compare.
To my astonishment, changing the probe ratio made no difference to the displayed amplitude whatsoever. x1, x10, x1000, the (freshly captured each time) waveform never changed in its vertical amplitude. As far as I can tell, the only thing that changes is the icon at the bottom of the screen that indicates V/div for that channel.
Nor should it.
If the waveform did change, say for example with applying a 1000:1 attenuation the waveform would disappear which would be of greater confusion to the operator.
Any measurement taken is referenced to the input attenuation factor and the V/div setting.
Change the input attenuation value in a DSO and the V/div value automatically changes to suit.
Entirely normal and expected behavior.
-
Hmm. I understand your points but am still bothered by it.
Given this, no one can trust what a scope screen says. If I take a screenshot and look at it weeks later (or someone else looks at it), there is absolutely no concrete way to confirm that the scope is telling the truth. Unless you're using a smart probe that automatically sets the ratio thanks to the little pads around the BNC, the indicated V/div may be accurate. Or not. And that accuracy may differ from screenshot to screenshot, day to day, week to week.
I can now make a case for both ways of doing it. There's a justified reason for both approaches. So rather than force users to live with one or the other, perhaps the V/div icon could include a little "/10" or something indicating that the display and the icon are not sync'd. Something, anything, to warn the user that the indicated data is NOT LITERAL. Because at the end of the day, it's irrational to have a precision piece of T&M equipment that displays erroneous information.
Thanks for the feedback!
-
Hmm. I understand your points but am still bothered by it.
Given this, no one can trust what a scope screen says. If I take a screenshot and look at it weeks later (or someone else looks at it), there is absolutely no concrete way to confirm that the scope is telling the truth. Unless you're using a smart probe that automatically sets the ratio thanks to the little pads around the BNC, the indicated V/div may be accurate. Or not. And that accuracy may differ from screenshot to screenshot, day to day, week to week.
I can now make a case for both ways of doing it. There's a justified reason for both approaches. So rather than force users to live with one or the other, perhaps the V/div icon could include a little "/10" or something indicating that the display and the icon are not sync'd. Something, anything, to warn the user that the indicated data is NOT LITERAL. Because at the end of the day, it's irrational to have a precision piece of T&M equipment that displays erroneous information.
Thanks for the feedback!
Yep, all that and it's one of the reasons we pushed real hard to get Siglent to display the probe attenuation in the channel tab where at a glance you can see what any probe is set for without diving into the channel menu.
So once channel attenuation is set to match the probe it's all very clear to the operator and in any screenshot.
-
I need to look into that. I'm delighted to hear that someone may already be doing this. I already have a Siglent function generator... sounds like I may need to add one of their scopes to the family. I understand their scopes couple intelligently to my SDG2042X to do nice Bode plots too.
What model of Siglent has this indication on the screen? I'd like to see what the screen looks like.
EDIT: I went to their website but I couldn't find any close-in screenshots on their various 2000 series scopes.
-
I need to look into that. I'm delighted to hear that someone may already be doing this. I already have a Siglent function generator... sounds like I may need to add one of their scopes to the family. I understand their scopes couple intelligently to my SDG2042X to do nice Bode plots too.
Yep, via USB or LAN and the Bode plot feature in the scope takes control of the AWG.
Not all Siglent scopes offer the Bode plot feature so just check carefully for those that do.
What model of Siglent has this indication on the screen? I'd like to see what the screen looks like.
All X, X-E and X Plus models. Most models also offer custom attenuation values.
You can see them in screenshots I've posted in this thread however most of the channel tabs are displaying 1x as we're at max sensitivity trying to measure uV of ripple. :scared:
https://www.eevblog.com/forum/testgear/a-good-scope-for-measuring-low-levels-of-noise-(e-g-linear-ps-noise)/ (https://www.eevblog.com/forum/testgear/a-good-scope-for-measuring-low-levels-of-noise-(e-g-linear-ps-noise)/)
-
The power of 10 R&S scopes also have custom ratios and easy to see display of the ratio for each channel. In case you're interested.
-
Just ran into something *I* consider wrong, but I don't have another digital scope here to compare.
I'm using a Rigol DS4024. Generally I'm very happy with the scope AND with Rigol. Today I connected a current meter and wanted to change that channel's probe ratio so the current had a more direct relationship to the volts/division display on the screen.
To my astonishment, changing the probe ratio made no difference to the displayed amplitude whatsoever. x1, x10, x1000, the (freshly captured each time) waveform never changed in its vertical amplitude. As far as I can tell, the only thing that changes is the icon at the bottom of the screen that indicates V/div for that channel.
The current probe connects via a standard BNC, so there's no funny business about interacting with the smart probe pads. To confirm, I connected a function generator and used a pure, continuous sine wave of known amplitude. Sure enough, I can vary the probe ratio anywhere I like but the displayed waveform does not change size at all.
The worst part about this is that the indicated V/div DOES change. Literally this means you can configure the scope to display BAD DATA. If you take a screenshot and look at it sometime later, it will display INVALID DATA and there's absolutely no way to know that.
As a concrete example: I set the sine wave for 4Vpp and the probe ratio at x1. The screen shows a sine wave that is 4 divisions tall. I then set the probe ratio to any other value. The displayed size of the waveform does not change on the screen. I tried setting the probe ratio to x1000 and the icon at the bottom of the screen indicated 1000V/div, so the screen is literally reporting a 4KVpp sine wave.
This is VERY easy to do (IMHO) properly. On a digital scope the vertical sensitivity is set via a rotary encoder, so the firmware is already involved and controlling the front end attenuator. It would be child's play to also consider the probe ratio, and then always make certain that the attenuator and the on-screen sensitivity icon agree with each other. Instead, for a standard BNC connection the probe sensitivity literally does nothing but screw with that channel's icon - nothing more.
I don't have another digital scope here to compare. I called Rigol and they were super helpful, super friendly, but stated this is intended behavior. The guy couldn't argue with me when I said it allows the scope to lie to the user. I can't rationalize the ability to make a scope screen outright lie to you
Is this considered "normal"?
I feel your pain, & it is a weird artefact of the DSO concerned, but changing the probe ratio isn't the normal way of achieving your original intention.
The usual method would be to adjust the volts/div control to give you a more appropriate relationship between current being looked at with your current probe & display amplitude.
I can only guess that Rigol assumed people would only manually change the probe ratio if they were using a x10 or x100 probe which didn't have the special pin on the probe BNC, needed for automatic setting of probe ratio.
I don't have a Rigol, or indeed, any DSO, but with the 7A18N plugin on my Tek 7613, if I use a plain x10 probe, the "on screen" v/div display agrees with whatever the v/div knob says, in other words, to get the true voltage, I need to multiply by 10.
If I use a probe with the special pin on the BNC, the "on screen" v/div display now reads 10x what it read before.(it doesn't cater for x100)
At the same time, the now correct volts/div mark on the scale around the knob lights up.
In both cases, the actual number of vertical divisions occupied by the probed voltage remains the same, as indeed, it should.
The 7A18N will not allow you to arbitrarily change the probe ratio.
If you only have a standard probe, --------"tough titties", you have to multiply by 10.
It seems Rigol trying to be "Mr Nice Guy", & saving their users from having to do elementary arithmetic has backfired on them.
-
perhaps the V/div icon could include a little "/10" or something indicating that the display and the icon are not sync'd. Something, anything, to warn the user that the indicated data is NOT LITERAL. Because at the end of the day, it's irrational to have a precision piece of T&M equipment that displays erroneous information.
i cant deny it. its a good thing to have more info on screen, but it wont solve everything.. for example whats the point of it if you cant know what type of probe used during probing? you can have a screenshot showing say 10Vpp signal swing 10V/div @ 1/10X and your client is happy about it, where you forgot you were actually using 1X probe and the signal was actually only 1Vpp? let me guess.. a wishlist to add another field in dso to enter probe model used and show that as well in screenshot? (higher end scopes + probe do this automatically) or a list of probes we can fill and select? but then what if we forgot to set it correctly before making a capture?
due to lack of this feature in my cheapo DS1054Z, i included in my VisaDSO the ability to store picture taken from handphone or camera or whatever of probing setup, so i can have the complete picture.. rig or jig used, what probe, which pcb revision and/or components configuration used etc... and then the ability to type simple text to describe whats going on in the testing, date, this channel represents what? etc etc in textual form.. all bundled in one file. we can always think its not enough and ask for more... as wisemen said you get what you pay, you ask more you need to pay more ;) i try to remind and educate myself to take note everything during probing and capture/document them in my VisaDSO, whichever that will confuse me later on, or also documentationing them in ms word excel doc or xls as well etc if necessary. this request/necessity happened from experience. i cant blame on the instrument because i cant afford to pay 5 figures for all the features that i wish. even if i can, i hate those virtual keyboard that we need rotate knob few turns just to run through a letter, press enter, swing back few turns again for another letter, i think my DS1054Z has this to set custom name for channel, man i dont want to use that i need full keyboard to do that kind of task i usually change channel setup very quickly. in the end i should blame myself (not the tool) and do better in the future imho. ymmv cheers.
-
I had an Hitachi analog scope with a digital readout. It seems like there was no way to over ride the probe type. If you used a probe without the feedback to the sense ring, I remember having to manually handle that complex 10X math in my head. My first scope had no readout and I had to know what my probe was and count the volts per division and calculate it in my head.
It's nice to be able to program in the scale factors for any probe you come up with.
-
What model of Siglent has this indication on the screen? I'd like to see what the screen looks like.
Now have time to dig old posts out, 2 channels are also inverted and indicated as such.
Probe compensation waveform at 1x, 10x, 100x and 1000x.
(https://www.eevblog.com/forum/testgear/siglent-sds1204x-e-released-for-domestic-markets-in-china/?action=dlattach;attach=448618)
Another of with the custom attenuation:
(https://www.eevblog.com/forum/testgear/siglent-sds1204x-e-released-for-domestic-markets-in-china/?action=dlattach;attach=448513;image)
From:
https://www.eevblog.com/forum/testgear/siglent-sds1204x-e-released-for-domestic-markets-in-china/msg1588057/#msg1588057 (https://www.eevblog.com/forum/testgear/siglent-sds1204x-e-released-for-domestic-markets-in-china/msg1588057/#msg1588057)
-
I'm confused as to why this is an issue. The displayed amplitude did, in fact, change. And the scope told you that it did by displaying a new V/div. I don't think I've seen a DSO that doesn't do this now that I think about it. As long as it is internally consistent and not buggy, it makes sense.
-
No, that's the point. The indicated V/div changed but the peak to peak displayed waveform did not change. The waveform signal looked identical and occupied the same number of vertical divisions whether the indicated V/div was 100mV/div or 1000V/div (not a typo).
-
It looks like it auto adjusted the gain for you. I think you are expecting it to behave like a CRO, where you have to manually adjust the gain when you switch to a probe with a larger attenuation.
-
It looks like it auto adjusted the gain for you. I think you are expecting it to behave like a CRO, where you have to manually adjust the gain when you switch to a probe with a larger attenuation.
No, that's the whole point.
The Rigol, did in fact, "behave like a CRO"!
The OP was using a current probe, & switched the probe setting, thinking that with the same probe, that would change the display vertical amplitude.
There is no reference in the first posting to changing probes, as the probe in use was, I reiterate, a current probe.
If you don't change the probe attenuation, changing the setting will not change the display amplitude, just as my 7A18N/7613 combination, using a x10 probe with a sensing pin will still give the same display amplitude as another x10 probe without one.
It will however, give the correct volts/div value.
If I could bodge a sensing pin on a x1 probe, I could make my CRO "lie", too.
-
Today I connected a current meter and wanted to change that channel's probe ratio so the current had a more direct relationship to the volts/division display on the screen.
...
To my astonishment, changing the probe ratio made no difference to the displayed amplitude whatsoever. x1, x10, x1000, the (freshly captured each time) waveform never changed in its vertical amplitude. As far as I can tell, the only thing that changes is the icon at the bottom of the screen that indicates V/div for that channel.
...
I called Rigol and they were super helpful, super friendly, but stated this is intended behavior.
...
Is this considered "normal"?
The probe attenuation factor is used to document the signal levels and provide calibrated cursor and automatic measurements. It has nothing to do with and no effect on the displayed waveform. The Rigol is operating like every other oscilloscope ever. Its only purpose is to provide the correct units.
-
Hmm. I understand your points but am still bothered by it.
Given this, no one can trust what a scope screen says. If I take a screenshot and look at it weeks later (or someone else looks at it), there is absolutely no concrete way to confirm that the scope is telling the truth. Unless you're using a smart probe that automatically sets the ratio thanks to the little pads around the BNC, the indicated V/div may be accurate. Or not. And that accuracy may differ from screenshot to screenshot, day to day, week to week.
I can now make a case for both ways of doing it. There's a justified reason for both approaches. So rather than force users to live with one or the other, perhaps the V/div icon could include a little "/10" or something indicating that the display and the icon are not sync'd. Something, anything, to warn the user that the indicated data is NOT LITERAL. Because at the end of the day, it's irrational to have a precision piece of T&M equipment that displays erroneous information.
Thanks for the feedback!
There are many ways an oscilloscope can display untruthful data. It's a comparatively complex tool, not a fire and forget plaything. Avoiding the pitfalls is a relevant part of learning to use them. Even experienced folks like Dave get caught out at times.
-
It looks like it auto adjusted the gain for you. I think you are expecting it to behave like a CRO, where you have to manually adjust the gain when you switch to a probe with a larger attenuation.
No, that's the whole point.
The Rigol, did in fact, "behave like a CRO"!
The OP was using a current probe, & switched the probe setting, thinking that with the same probe, that would change the display vertical amplitude.
There is no reference in the first posting to changing probes, as the probe in use was, I reiterate, a current probe.
If you don't change the probe attenuation, changing the setting will not change the display amplitude, just as my 7A18N/7613 combination, using a x10 probe with a sensing pin will still give the same display amplitude as another x10 probe without one.
It will however, give the correct volts/div value.
If I could bodge a sensing pin on a x1 probe, I could make my CRO "lie", too.
No, if you change between a 1x, 10x, 100x, 1000x voltage probe without touching the vertical gain on a CRO, the displayed amplitude will change (and will probably disappear altogether). As David Hess points out above, the readout is meaningless here and the readout sense pin only serves to make the readout read the right units.
Current probes are a whole other kettle of fish that I must admit I didn't notice in the OP in my sleepy stupor last night. At least the one I use (P6042) wants a 50 ohm input and the scope set to 50 mV/div for a calibrated current waveform; you control the gain from the current probe's own amplifier. I do not know the details about what current probe the OP is using, but I would wager the manual for it tells you how to use it.
-
It looks like it auto adjusted the gain for you. I think you are expecting it to behave like a CRO, where you have to manually adjust the gain when you switch to a probe with a larger attenuation.
No, that's the whole point.
The Rigol, did in fact, "behave like a CRO"!
The OP was using a current probe, & switched the probe setting, thinking that with the same probe, that would change the display vertical amplitude.
There is no reference in the first posting to changing probes, as the probe in use was, I reiterate, a current probe.
If you don't change the probe attenuation, changing the setting will not change the display amplitude, just as my 7A18N/7613 combination, using a x10 probe with a sensing pin will still give the same display amplitude as another x10 probe without one.
It will however, give the correct volts/div value.
If I could bodge a sensing pin on a x1 probe, I could make my CRO "lie", too.
No, if you change between a 1x, 10x, 100x, 1000x voltage probe without touching the vertical gain on a CRO, the displayed amplitude will change (and will probably disappear altogether). As David Hess points out above, the readout is meaningless here and the readout sense pin only serves to make the readout read the right units.
All of which just re-states what I said in my posting, I can't see where the "No" comes in!
Current probes are a whole other kettle of fish that I must admit I didn't notice in the OP in my sleepy stupor last night. At least the one I use (P6042) wants a 50 ohm input and the scope set to 50 mV/div for a calibrated current waveform; you control the gain from the current probe's own amplifier. I do not know the details about what current probe the OP is using, but I would wager the manual for it tells you how to use it.
My point, again, was that the OP was not at any time, changing a voltage probe, & the output voltage of the current probe was the same, no matter what probe setting was selected on the Oscilloscope.
All the "probe setting" does is to change the apparent v/div.
-
Current probes are a whole other kettle of fish that I must admit I didn't notice in the OP in my sleepy stupor last night. At least the one I use (P6042) wants a 50 ohm input and the scope set to 50 mV/div for a calibrated current waveform; you control the gain from the current probe's own amplifier.
I'm measuring dozens of amps so a P6042 isn't a good tool in this application (plus so far I haven't been able to find a good one that wasn't overseas). The current probe in question is a simple Hantek CC-65. It has two ranges and expects to drive into a high impedance load (manual states "typical 10K") so its coax is not being used as a transmission line and the scope input is set to 1M.
Its output ranges are 1mV/100mA (1V = 10A) and 1mV/10mA (1V = 100A). My goal was to set the probe ratio so that 1V = 1A, making it easier to read quickly on-screen. It seems my only option is to 1) set the ratio to 1:1, 2) adjust the channel sensitivity to get a decently sized physical display, and then 3) adjust the ratio again based on the probe's output and whatever sensitivity I ended up at based on step 2. And that's only if I care that the on-screen icon is "correct"... if I just ignore it nothing really changes, since the physically displayed waveform stays the same.
Basically the probe ratio adjusts the text at the bottom of the screen. Nothing more.
-
You can't perform what ever math function you want on a channel and display that? I have spoiled myself. :-DD
I guess you could amplify the signal externally or just put 10 turns through it assuming it would fit into the clamp and still carry the current you want.
Depending on your requirements maybe someone could recommend a better probe if that is what you are after.
-
Basically the probe ratio adjusts the text at the bottom of the screen. Nothing more.
That's pretty normal. It wouldn't make sense to do anything else.
-
Depending on your requirements maybe someone could recommend a better probe if that is what you are after.
It's not the probe that is the question here. It's very clear about its current to voltage relationship. I just hoped to use the probe ratio to normalize that on the screen. Oh well... I grew up on CRO's and have a lifetime of doing sensitivity calcs in my head, so I'll just keep doing that.
Thanks for the discussion!
-
Depending on your requirements maybe someone could recommend a better probe if that is what you are after.
It's not the probe that is the question here. It's very clear about its current to voltage relationship. I just hoped to use the probe ratio to normalize that on the screen. Oh well... I grew up on CRO's and have a lifetime of doing sensitivity calcs in my head, so I'll just keep doing that.
Thanks for the discussion!
I was referring to your comment "...plus so far I haven't been able to find a good one that wasn't overseas".
Looping the 10 turns through will gain it up by 10X. On the 1mV = 10mA scale you will get 10mV = 10mA. You may run out at the high end.
-
It's not the probe that is the question here. It's very clear about its current to voltage relationship. I just hoped to use the probe ratio to normalize that on the screen. Oh well... I grew up on CRO's and have a lifetime of doing sensitivity calcs in my head, so I'll just keep doing that.
Why ? :-//
Can't you select amps as the displayed unit ?
Then with the correct attenuation factor for the probe let the scope display the value in amps.
A DSO takes all the headache of doing calcs on the fly and you need really get you head around how to take advantage of this.
-
It's not the probe that is the question here. It's very clear about its current to voltage relationship. I just hoped to use the probe ratio to normalize that on the screen. Oh well... I grew up on CRO's and have a lifetime of doing sensitivity calcs in my head, so I'll just keep doing that.
let us help you. what is your current probe sensitivity? 10mA/V? 100A/V? what amperage you want for each graticule? my (cheapest grade) rigol channel's unit can be changed to A. your probe is not exactly 1/10X or 1/100X attenuation? fine! we have fine/coarse vertical adjustment. we always do multiplication in our head, if signal is 2.5 graticules high at 10V/div, we know its 2.5x10 = 25V... its not going to be a brain breaking task... ;)