Would all of the commercial True RMS voltmeters give me an accurate voltage reading of pink noise? In Table 6 of the datasheet for the AD737, optimum values for the averaging and filtering capacitors external to the chip are different relative to the waveforms to be measured. I was wondering if it would be worthwhile for me to construct an adapter such as Dick Cappels has done tailoring it for true RMS of (not sinusoidal) audio waveforms (my intended application), or if a commercial meter would be adequate for this.

https://www.analog.com/media/en/technical-documentation/data-sheets/AD737.pdf

Moderator edit: New thread created from this.

 

Would all of the commercial True RMS voltmeters give me an accurate voltage reading of pink noise?

Only over there input AC bandwidth, which should be specified in their spec sheet, but may not be.

 

In Table 6 of the datasheet for the AD737, optimum values for the averaging and filtering capacitors external to the chip are different relative to the waveforms to be measured. I was wondering if it would be worthwhile for me to construct an adapter such as Dick Cappels has done tailoring it for true RMS of (not sinusoidal) audio waveforms (my intended application),

You just select the capacitor for the lowest frequency of interest and it will give the true RMS value of the signal from there out to it's rated frequency response.
Why do you think you need Dick's adapter?

 

You just select the capacitor for the lowest frequency of interest and it will give the true RMS value of the signal from there out to it's rated frequency response.
Why do you think you need Dick's adapter?

Thanks Crutschow!

I'm doubtful that any of the commercial meters allow the user to select the values of averaging and filtering capacitance of the two capacitors based on what you want to measure. Perhaps this is only true for the AD 737, but the fact that Table 6 is included in the data sheets for it, suggests that having the correct values of capacitance is necessary to get an accurate reading, based on what is being measured.

I haven't yet really carefully read through the AD737 data sheets, but I don't think that they specify how much error results from not having the optimum values of Cav and Cf.

I'd want to build an adapter if I thought that the reading on a commercial meter would be off by maybe +/- 1 dB. Also probably the worthwhile True RMS voltage meters are on the more expensive side.

 

The RMS (root mean square) value of any waveform can be calculated as follows:

1) form the square of the voltage, i.e. v(t) x v(t)
2) take the average value over a given duration T
3) take the square root of the average value,

hence the name, square root of the mean of the square.

 

suggests that having the correct values of capacitance is necessary to get an accurate reading, based on what is being measured.

I believe that capacitor is just to smooth the output which has a ripple based upon the lowest frequency measured.
Thus there's a tradeoff between response time to a change in input amplitude, and ripple.
Don't think it affects accuracy otherwise.

 

Would all of the commercial True RMS voltmeters give me an accurate voltage reading of pink noise? In Table 6 of the datasheet for the AD737, optimum values for the averaging and filtering capacitors external to the chip are different relative to the waveforms to be measured. I was wondering if it would be worthwhile for me to construct an adapter such as Dick Cappels has done tailoring it for true RMS of (not sinusoidal) audio waveforms (my intended application), or if a commercial meter would be adequate for this.

https://www.analog.com/media/en/technical-documentation/data-sheets/AD737.pdf

Moderator edit: New thread created from this.

I would not think the RMS value of a signal that is neither periodic nor has an average value of zero would be particularly meaningful. I'm curious as to why you might think so. It might be instructive to ask if such a measurement was repeatable. For example, if you make 10 measurements over a 1 second interval, would you expect the values to cluster or would they be randomly distributed?

 

Would all of the commercial True RMS voltmeters give me an accurate voltage reading of pink noise? In Table 6 of the datasheet for the AD737, optimum values for the averaging and filtering capacitors external to the chip are different relative to the waveforms to be measured. I was wondering if it would be worthwhile for me to construct an adapter such as Dick Cappels has done tailoring it for true RMS of (not sinusoidal) audio waveforms (my intended application), or if a commercial meter would be adequate for this.

https://www.analog.com/media/en/technical-documentation/data-sheets/AD737.pdf

Moderator edit: New thread created from this.

You can buy commercial meter but a good one won't be cheap.
https://www.fluke.com/en-au/product/electrical-testing/digital-multimeters/fluke-287

True-RMS AC bandwidth

100 kHz

 

I would not think the RMS value of a signal that is neither periodic nor has an average value of zero would be particularly meaningful. I'm curious as to why you might think so. It might be instructive to ask if such a measurement was repeatable. For example, if you make 10 measurements over a 1 second interval, would you expect the values to cluster or would they be randomly distributed?

Take two independently running sources of noise and add their outputs. The circuit that I want to test attenuates the level of the addition of the two noises and I need to determine how much attenuation occurs. On my oscilloscope I can see that some attenuation is occurring, but just by looking at the two wave forms I can't say to what extent. I want to claim as much attenuation as possible because that is what my circuit is supposed to do and it is a fairly complex circuit To make this understandable I have to include that the attenuation can only occur where the two signals at input are uncorrelated. Where two signals are added that are the same, then there can be no attenuation of the summation. In other words if the output of a sine wave generator is applied to the two inputs of my circuit, then it must be the case that the processing of the doubled voltage of the sine wave is not attenuated at output of processing.