DIY Precision AC-RMS to DC Transfer Standard

[1audio]

The Fluke 510 is an old design. I have one and tweaked it down to around -110 dB THD. However implementing a modern version with IC's would not be too difficult. You need a 2 phase oscillator (state variable is the typical implementation) and a sample and hold triggered by the zero crossing of one phase to capture the peak of the other phase. One clever implementation used an ADC to sample the peak. Sample and hold chips are mostly obsolete today since the ADC's incorporate them now. There are other ways to capture the peak voltage. The reference circuit looks like Flukes standard FLU practice so its complicated as well.

[Andreas]

A modern part with extremely small distortions would be a 24 Bit Audio ADC. One could run the left channel on the sine and the right channel on the chopped DC reference. Those ADCs are pretty cheap and sample up to several hundred KHz.

But look at the gain error. It might be surprisingly high. (typical in the 3-6% range).

with best regards

Andreas

[dietert1]

Yes, the audio ADC can still be useful to compare/transfer a sine or any other AC waveform to that chopped DC reference mentioned above. One could use the same ADC channel with a MUX and/or avoid channel imbalance by calibration.

Regards, Dieter

[Verticon]

I just came across this thread these days and because I fancied with a DIY ac standard already a long time I wanted to know what can be realized quick and dirty by using existing lab equippment. The setup is based on a method which was already described in this thread. I tried to measure the peak values of a pure sine wave by comparing these with a well known DC value. I connected the output of my home made state variable oscillator (low source resistance output) with the input of a 7A22 Differential Amp in an old TEK scope but with a precision DC source in series. By changing  the polarity of the DC source the positive or negative peak value of the sine wave can be compensated to zero by observing it on the scope. Switching the polarity is neccessary to correct for a small DC part of the oscillator (Offest voltage of the Output Amp). The DC values have been mesured with a 3478 (on DC very close to the 34401A) and the calculated AC voltage (add the two unsigned DC values and divide the result by the two-fold suare root of 2) was compared with the measured value by my freshley calibrated 34401A.

I have to admit that I was a bit surprised about the good and reproducable results of this quick test. A typical result: For the 1V AC range I' ve got a deviation to the directly measured values from the 34401A of approx. 0,03% (at 110Hz). This is well below the 90day spec of the 34401A and in the range of the measuring uncertainty noted in my calibration certificate. The pictures show a sketch of the setup, the oscillator (distortion < 0,001%) and DC source and a typical curve on the scope display to get a feeling about noise (1mV/div scale). I am pretty sure with some more effort on the oscillator output configuration and better shielding the accuracy can be significantly improved.

[1audio]

That s the method Krohn Hite described for checking their function generators.