Voltage measurement using an oscilloscope vs multimeter

[createthis]

So, disclaimer here, I more or less have no idea how to operate an oscilloscope. I've always wanted one, so I finally bought one. I have a little electronics theory under my belt (book knowledge), but I'm not an electrical engineer. I'm a programmer by trade.

I've been playing with the scope, and the one thing that keeps sticking out like a sore thumb is that I can't trust the voltage measurements. They seem to bounce around depending on the scale, etc. So, I thought I'd reach out and learn something rather than continue to bumble around like an idiot.

It's painfully obvious in this video I made, attempting to measure inrush current from my 12v fridge's compressor: https://youtu.be/O_DnUwPpbKQ (skip to 19:40 for the part about the oscilloscope)

Why does my multimeter give me different voltage measurements than my oscilloscope?

How do I properly setup my Rigol oscilloscope so that it's voltage measurements match my multimeter?

Thanks!

--
Jesse
CreateThis.com

[timb]

Scopes are really meant to measure the relative change of a signal over time, not an absolute voltage. For absolute voltage measurements, you're looking at one decimal place of accuracy, I.e., 12.0V, 12.1V, etc. at best.

Modern digital scopes are better at this with their digital readouts, historically analog scopes were even less accurate, since you had to count which grid line the trace was on to get your measurement (if you're set at 10V/div and the trace is on the third division from the bottom your signal is roughly 30V, assuming you zeroed it out at the bottom).

To get the most accurate absolute voltage measurement, make sure you're in DC coupled mode and use the voltage readout function for that channel. (AC coupled mode places a capacitor in series with the input, which charges up to the incoming voltage; so if you had, say, a pure 15V DC signal, the scope would indicate a 0V flat line. But, say that 15V DC signal had a 100mV RMS AC sine wave riding on top of it, then you could set your scope to 50mV/div sensitivity and see the sine wave clearly, whereas if you were in DC coupled mode you'd have to be in 5V/div mode and that 100mV sine wave riding atop the 15V wouldn't be visible. AC coupled mode is great for looking at power supply ripple for this very reason!)


Sent from my Tablet

[wiss]

You should have chanel 2 at 2 V / square and set the ref level to the bottom of screen, that way you'll get way better measurements.

[createthis]

To get the most accurate absolute voltage measurement, make sure you're in DC coupled mode and use the voltage readout function for that channel.

At 25:08 you can see CH2 is in DC coupled mode. I'm not sure what you mean by "use the voltage readout function for that channel". I used max and min voltage from the vertical menu on the left.

[createthis]

You should have chanel 2 at 2 V / square and set the ref level to the bottom of screen, that way you'll get way better measurements.

Sorry for being a noob, but how and why? Thanks.

[timb]

I didn't watch your video (especially at 30+ minutes), those were just general tips. Like I said, scopes aren't designed for accurate absolute voltage measurements. Use a multimeter for that. The right tool for the job, as they say.

The max and min readouts should be alright, but there should also be an option for an "average" readout, which might be your best bet (as it'll smooth out the noise if you just want a basic voltage reading).


Sent from my Tablet

[timb]

You should have chanel 2 at 2 V / square and set the ref level to the bottom of screen, that way you'll get way better measurements.

Sorry for being a noob, but how and why? Thanks.

Why: For a positive going waveform, setting the reference at the bottom and the V/div to the lowest setting for the maximum voltage excursions expected will allow you to use the maximum amount of the display and thus increase resolution.

How: Move the Ch 2 trace to the bottom. Set the scale to 2V/div.


Sent from my Tablet

[Galaxyrise]

I'll try to watch the video this evening, but a general on how the scope works.  The input signal must be scaled, shifted, and then digitized (usually 8-bit.)  Each of those steps adds error, and the error will be specific to that setting.  This is true on a DMM as well: different DMM ranges can give different readings for the same voltage (though it's usually off by less and has more resolution.)

[wiss]

You will have (usually) 8 bits to cover the full display (on Keysight that is 10 squares, of those you can see . That is 256 states, you set about one square to observe the voltage, that is ~ 25 states, ofcourse the quantization will be noticeable.

[Stray Electron]

   OP, a couple of things.  A scope gives you a two dimensional display, voltage and time.  A meter only gives you one dimension (voltage) unless it's a slowly varying signal that you can follow with your eyes. (Look up the update rate of your particular meter and also persistance of the human eye for more details about that.) But outside of that, unless I KNOW what to expect, I prefer to use my relatively cheap meter for most measurements rather than risk a very expensive scope.  Also my regular meter reads up to 1200 volts and my scope only up to 200 volts (350 volts MAX) with the regular x10 probes so that's something else to consider.

   Other than that, as you've already found, motor starting transients are very fast! And you're not going to see them on a meter.  The other thing is that they can also be very high so there's a good chance of burning out either a meter or a scope unless it has good protection built into it.

   For example, I was testing a Power Line Disturbance Analyzer a few months ago and left it running.  Later that day I turned on an Astron  switching power with a small load (<3 amp) on it and I was surprised when the PLDA alarm went off.  I checked the stored record and it showed a power line spike of over 500 volts when I turned on the switching power supply!  I tried it several more times and the same thing repeated every time!

[createthis]

I didn't watch your video (especially at 30+ minutes), those were just general tips.

I understand, and thanks for taking a look. BTW, if you jump to 19:40 as I suggested in the OP, it's only 10 minutes. Still pretty long though.

[createthis]

   Also my regular meter reads up to 1200 volts and my scope only up to 200 volts (350 volts MAX) with the regular x10 probes so that's something else to consider.

Just to be clear, the current probe I'm using with the scope AND the meter (same probe) is a Fluke 80I-110S hall effect 100mv/A sensor. 200 volts / .100mv = 2000 amps. I'm admittedly ignorant of how x10 probes affect the 400V limit on my scope, but I thought that would probably be a fine safety margin. I only measured voltage directly on the battery to see the voltage sag as the current spiked.

   Other than that, as you've already found, motor starting transients are very fast! And you're not going to see them on a meter.  The other thing is that they can also be very high so there's a good chance of burning out either a meter or a scope unless it has good protection built into it.

   For example, I was testing a Power Line Disturbance Analyzer a few months ago and left it running.  Later that day I turned on an Astron  switching power with a small load (<3 amp) on it and I was surprised when the PLDA alarm went off.  I checked the stored record and it showed a power line spike of over 500 volts when I turned on the switching power supply!  I tried it several more times and the same thing repeated every time!

I did hear a number of beeps on my meter during the measurements. I just realized I have no clue what those meant.

[createthis]

I think I figured out what's happening here. I was distrustful of the o-scope because, like you all, I tend to trust my multimeter first for voltage readings. However, I think in this case the o-scope was correct. The Fluke 87 wasn't sampling fast enough to catch the voltage peak.

I found this (rather old) document that talks about the Fluke 87's Min/Max sample rate: http://www.atraonline.com/gears/2003/2003-08/2003_8_54.pdf

Here's the relevant excerpt:

Pressing the button labeled PEAK MIN MAX below it (audible alert button) while in MIN MAX mode will put the meter into Peak MIN MAX mode. This will record minimum and maximum voltage peaks that are much shorter than in regular MIN MAX. Exactly how short depends on the age of your meter. Older Fluke 87 meters increase the sampling rate to 1/1000th of a second; later meters are 4 times faster, recording at a speed of 250 microseconds. That’s considerably faster than the normal 1/100th of a second in normal MIN MAX mode.

I tried using the Fluke 87 in Peak mode and I got 1.87V. At .100v per amp, that's 18.9 amps, or 243.81 watts. That's a lot closer to what I was seeing with the o-scope and explains the discrepancy.

Thanks for looking!

--
Jesse
CreateThis.com