9
\$\begingroup\$

So I recently got myself an oscilloscope and am trying to learn how to use it by starting with the basics, a circuit with a voltage source, and a resistor. When I set my power supply to 5V, the multimeter reads 4.995V and the oscilloscope reads 5.02V rms. I have found a thread already explaining how to tell which one is more accurate, but I am looking for more of an explanation of the other things at play here causing these to be different.

I figure the power supply is just giving an estimated reading, it was a fairly cheap one so I understand if it dent exactly output the voltage it reads.

After this picture I realized the different locations of where I am measuring might be a factor as breadboards tend to have undesirable effects to circuits. It didn't change anything.

What are some other possible sources to the discrepancy and what can I do to try and remediate it (already calibrated the oscilloscope), if anything?

\$\endgroup\$
13
  • 5
    \$\begingroup\$ @jsotola: I agree on measuring at the power supply connectors, but not at different times. You should be able to hang 100 different meters and scopes and whatnot, and they should all see the same voltage (barring the difficulty of actually getting 100 of each of them on the terminals). \$\endgroup\$
    – TimWescott
    Apr 7 at 1:13
  • 16
    \$\begingroup\$ Check your scope's rated accuracy -- scopes are generally good at making measurements fast, but for less expensive scopes, 256 points from minimum to maximum voltage is not out of line -- that would make your 5.02V reading perfectly consistent with the meter's 4.995V. \$\endgroup\$
    – TimWescott
    Apr 7 at 1:15
  • 7
    \$\begingroup\$ Nothing to add but just that I think that's a nice starting set of kit for a serious hobbyist or professional's home setup. \$\endgroup\$ Apr 7 at 1:49
  • 4
    \$\begingroup\$ Your meter has three digits after the decimal, but the power supply only has two. If you round the meter reading to two digits, you will get 5.00 V - same as the power supply. \$\endgroup\$ Apr 7 at 16:03
  • 13
    \$\begingroup\$ There's an old saying that applies here: "A man with a watch knows what time it is. A man with two watches is never sure." \$\endgroup\$ Apr 7 at 16:33
45
\$\begingroup\$

When I set my power supply to 5V, the multimeter reads 4.995V and the oscilloscope reads 5.02V rms.

5.02 V / 4.995 V = 1.005. That's a 0.5% difference. 5.00 V / 4.995 V = 1.001. That's a 0.1% difference. The specified accuracy of your oscilloscope is +-3%, the multimeter is +-0.5% +3 counts, and the power supply meter's accuracy is unknown.

There is no discrepancy. All the measurements are within their error bars for an actual value of between 4.97 V and 5.02 V, so they are all effectively the same.

\$\endgroup\$
3
  • 13
    \$\begingroup\$ And the "specified accuracy" of the multimeter is just what they copied out of the Fluke manual. \$\endgroup\$
    – hobbs
    Apr 7 at 1:55
  • 2
    \$\begingroup\$ Wow, good eye @hobbs -- I didn't even notice the multimeter was a lookalike, not a Fluke. \$\endgroup\$ Apr 7 at 4:23
  • 7
    \$\begingroup\$ Well, Fluke or not, as long as it isn't a fluke it doesn't really matter. \$\endgroup\$
    – Kroltan
    Apr 7 at 18:01
13
\$\begingroup\$

This is fine. None of those devices are precision instruments (and the power supply is probably not the least accurate among them!), and you're seeing a difference of < 0.5% between the highest and the lowest. There's nothing you should be doing, unless you're involved in some high-precision physics work, in which case you should get your lab director to buy you some new gear and a better place to put it than the kitchen counter. For hobby work, you're already doing more than good enough.

\$\endgroup\$
1
  • \$\begingroup\$ Normally I do have a proper bench to work off of, but current situations dictate the need to use this empty space on the counter ^_^ \$\endgroup\$
    – dka13
    Apr 7 at 12:07
9
\$\begingroup\$

Of the three instruments you have on your table, the multimeter will be the most accurate and precise for measuring a DC voltage.

Scopes have larger input offset errors as well as lower ADC resolution than a multimeter, but have a much higher sampling rate. Scopes can also have overall gain error and certain non-linear sampling errors. So there’s a definite trade off when using a scope to measure DC or DC-biased signals. Truth be told, most scopes kind of suck at it really, but if you're measuring DC-biased dynamic signals like the voltage across a resistive current sensor, using a scope is essential.

Your scope can be calibrated better, but it’s still within its margin of error. If this matters for your measurements and you want a quick fix, you can characterize the scope DC offset vs. the meter and subtract the offset in subsequent measurements to improve accuracy. This is good enough for most work. (n.b., that scope input offset can differ from range to range, so you need to characterize the ranges separately.)

If you (or your employer) really, really care about this (for example, for compliance with a QA standard like ISO9001) you will need to spend more money on the scope with more ADC bits, fancier input amplifiers, and NIST/NBS-traceable calibration for it from time to time. You might also need to invest in some precision calibration references for your lab use.

Relevant: Accuracy vs. Precision, https://blog.minitab.com/en/real-world-quality-improvement/accuracy-vs-precision-whats-the-difference

\$\endgroup\$
3
\$\begingroup\$

Your Scope will tell you Vpeak and Vrms(root mean square). Your multimeter gives, if it's a true rms, well that is should give the same rms value as your scope, if the signal is a pure sinus. Your multimeter can also give wrong values depending on the frequency it can handle. Good multimeters are true rms up to 1MHz. Be sure to put the Power supply behind a safety transformer when you need to measure the primary part of it, or use a differential probe so you won't create a short and blow up either your scope or the Power Supply.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.