MM Vs. Scope

[eev_carl]

Hi,

I have two resistors in series connected to a sine wave voltage source.  I'm getting two different readings with my multimeter and scope and was wondering if someone could explain the scope reading.  My multimeter gives me an RMS reading that matches the function generator (1.765V), but the scope gives me a lesser value (1.69V).  Is this just a difference in accuracy?

Both the multimeter and scope leads are connected to the voltage source and to the ground.

Thanks,
Carl

[eb4fbz]

An oscilloscope is designed to look at waveforms, it's voltage calibration and resolution is secondary. However, a multimeter is carefully calibrated and use a higher resolution (although slower) ADC.

[Gyro]

You haven't indicated what frequency you are running the function generator at. The frequency response of the meter and scope are different. At some lowish frequency point, the meter reading will start to drop off.

[Kleinstein]

The amplitude calibration of scopes is often not that good - depending on the model some 2-5% are possible. So the difference could be just from the tolerances. In addition a probe (e.g. 10:1) might add some error. So does possible AC coupling if the frequency is low.

In addition the waveform from the generator may not be perfect and the DMMs might not include all the harmonics. Also AC accuracy of DMMs is often not that high, especially if using only a small part of the range. At higher frequencies the scope might be more accurate than the DMM, while around 50 Hz the DMM is usually more accurate.

[eev_carl]

My input Vs is 5Vpp at 1kHz.

[Gyro]

At 1kHz sine wave, the meter is probably the closest to correct (depending on model). 5V is a good level (DMMs tend to develop non-linearity at the bottom of a range).

You can do some experiments to 'characterize' your meter. Wind the frequency up and down to see how the meter reading responds, I would expect that at, say, 50kHz the reading will be well down. The scope reading ought to stay reasonably constant over the 50Hz - 100kHz range, while the meter will probably vary. The reading at 50Hz is probably there the meter will tell the 'truth', so you can get a feel for how close the scope reading is at that point.

[james_s]

Most scopes are only 8 bit resolution, they are not precision voltage measurement devices. This particular discrepancy doesn't sound too far off.

[eev_carl]

At 1kHz sine wave, the meter is probably the closest to correct (depending on model). 5V is a good level (DMMs tend to develop non-linearity at the bottom of a range).

You can do some experiments to 'characterize' your meter. Wind the frequency up and down to see how the meter reading responds, I would expect that at, say, 50kHz the reading will be well down. The scope reading ought to stay reasonably constant over the 50Hz - 100kHz range, while the meter will probably vary. The reading at 50Hz is probably there the meter will tell the 'truth', so you can get a feel for how close the scope reading is at that point.

• It looks like 2k is a "sweet spot" where the DMM matches the scope; both are 1.72V
• At 10kHz, my DMM doesn't show anything.
• At 100Hz, my scope is way off.

[Gyro]

I suspect that "sweet spot" isn't the most accurate reading though, it will be where the DMM reading has rolled off to match the slightly inaccurate (but within tolerance) scope reading. You should get a better scope reading at 100Hz if you change the timebase setting, also check that the input coupling is set to DC for low frequencies.

It's a useful learning exercise though, obviously not 'calibration' standard, but it gives you a much better feel about how much you can trust you can trust your instrument readings in different situations.

P.S. Try the same exercise with squarewave, triangle etc. Only a true-RMS meter will give a reading more or less matching the scope, but you can compare the reading you 'should' get with an average responding meter for different waveforms.  eg. https://www.fluke.com/en-us/learn/best-practices/measurement-basics/electricity/what-is-true-rms

[Doctorandus_P]

What are the values of your series resistors?
Your DMM and oscilloscope may have different input impedances, which could be a mayor factor in the discrepancy.

A lot of multimeters are also lying about "RMS".
This grew historically. The RMS converter chip often has a DC blocking capacitor in front of it, while for a true RMS value also the DC component should be measured.
Only recently the AC+DC RMS DMM's are becoming more common.

What kind of DMM / Oscilloscope are you using?
Toy's, or decent stuff?

It is very common for a DMM to only be able to measure AC accurately to a maximum of a few hundred Hz.
If your scope is "way off" @ 100kHz, I suspect you have a toy like the DSO150 or DSO nano or similar, or your function geneator is not capable of generating proper signals for such a frequency.

[eev_carl]

What are the values of your series resistors?

What kind of DMM / Oscilloscope are you using?
Toy's, or decent stuff?
If your scope is "way off" @ 100kHz, I suspect you have a toy like the DSO150 or DSO nano or similar, or your function geneator is not capable of generating proper signals for such a frequency.

Hi,

Resistors are 1k and 10k.

I'm using a Klein MM700 (True RMS).  It's listed at 50-400Hz, so I switched my Vs to 5Vpp at 400Hz.  My scope is a Rigol MSO1104Z.  The function generator is a DG1022Z.

I watched a W2AEW video on probes and adjusted my compensating capacitor.  That seems to bring the values closer.

[bitseeker]

Also, for amplitude measurements on a digital scope, be sure to use as much of the vertical scale as possible. Setting the V/div too high doesn't fully utilize the ADC and will not give as accurate a result. You want the waveform to span as many divisions as possible without going out of bounds in order to use as many of those bits as possible.

[Kleinstein]

The limited resolution of the ADC is not such a big problem for RMS calculation for the scope. It can use quite some samples and thus adds resolution from oversampling. With the usual 1:2:5 steps at the scope, it should be possible to use at least 30% of the ADC range.

The scope might be off at low frequency due to AC coupling roll off. Otherwise I would not expect too much deviation towards low frequency here. If the DMM is specified till 400 Hz, it might already be off quite a bit at 2 kHz - enough to compensate for some difference in calibration of the scope and DMM. Heaving the same reading on the DMM and scope does not mean the reading is accurate.

The cheap DMM might also have quite some error for the RMS value, especially if the signal is less than 20% of full scale. So with the typical 1:10 steps, some AC voltages might be hard to measure accurately with such a meter.

The input impedance of the scope and AC input of DMMs is often not that different: scopes are usually 1 M + some 20 pF and DMMs  are often 10 M and sometimes (especially the better ones) 1 M in AC mode, with usually a little more capacitance.

[Circlotron]

Also, for amplitude measurements on a digital scope, be sure to use as much of the vertical scale as possible. Setting the V/div too high doesn't fully utilize the ADC and will not give as accurate a result. You want the waveform to span as many divisions as possible without going out of bounds in order to use as many of those bits as possible.

Not only that, make sure you have several cycles visible on the screen. On my scope when you wind the frequency down and the ends of the cycle stretch off each side of the screen the digital volts measurement reads very low.