Upgrade Fluke 510A AC voltage reference
Posted 16th July 2012 at 04:45 PM by 1audio
While doing some research on making an ultra low distortion analyzer I stopped to look at the Fluke 510A that has been siting on my shelf for years. Lots more on the 510a here: https://assets.fluke.com/manuals/510A_AD_imeng0000.pdf
In short a low distortion (.005% or better) 10V +/- .002% reference oscillator. There are very good numbers and by 1970 standards exceptional.
I was looking at it because I was researching AGC circuit for oscillators. This is an interesting implementation with its standards lab precision accuracy and stability. However its not simple and the critical parts are expensive (an ultra stable voltage reference, ultra stable resistors etc.)
In any case I was wondering if the distortion can be reduced still further. At the time it was built .002% was about as good as possible in practice and better would be hard to verify.
In short, after a number of experiments I got the distortion down to .0003% THD (30 KHz LPF) and .00015% looking at the first 5 harmonics.
There are several tricks in volved. The first efforts (increasing the internal gain and making it more linear) didn't have miuch effect. I tried some tricks around the agc fet and those also did not do a lot, but they did help. Finally I borrowed from Scott Wurcer's distortion neutralizer trick and that really seemed to help.
The AGC is outside of the amp/oscillator so none of this affects it. although calibration is probably necessary at its age. Most of the cal is pretty simple (if you have the necessary references. . . ).
Short6 list of changes-
1) 1 mA current diodes for R51 and R84 (increase forward gain and input differential linearity)
2) 2 mA current diode replaces R97 (more improved forward gain)
3) replace output resistor R103 with a 1 mH inductor to improve stability with the distortion neutralization cap.
4) add a Zobel network across the output of 1000 pF in series with 249 Ohms.
5) add a 10 pF cap from the node labeled tp12 to the base of Q22 (distortion neutralization cap)
6) I replaced R91 with a pot and upgraded C32 to a film cap. The pot has a small effect and I am not sure its making a difference any more.
After this a normal calibration will yield -120 dB harmonics, pretty much a uniform spectrum with no single harmonic dominating.
This was with one example of a 2400 Hz unit. I can't say that others would respond the same.
The 10V output is constant with load until the current limit is hit (and the uncal light lights).
I'll post more on my efforts towards some useful reference oscillators.
In short a low distortion (.005% or better) 10V +/- .002% reference oscillator. There are very good numbers and by 1970 standards exceptional.
I was looking at it because I was researching AGC circuit for oscillators. This is an interesting implementation with its standards lab precision accuracy and stability. However its not simple and the critical parts are expensive (an ultra stable voltage reference, ultra stable resistors etc.)
In any case I was wondering if the distortion can be reduced still further. At the time it was built .002% was about as good as possible in practice and better would be hard to verify.
In short, after a number of experiments I got the distortion down to .0003% THD (30 KHz LPF) and .00015% looking at the first 5 harmonics.
There are several tricks in volved. The first efforts (increasing the internal gain and making it more linear) didn't have miuch effect. I tried some tricks around the agc fet and those also did not do a lot, but they did help. Finally I borrowed from Scott Wurcer's distortion neutralizer trick and that really seemed to help.
The AGC is outside of the amp/oscillator so none of this affects it. although calibration is probably necessary at its age. Most of the cal is pretty simple (if you have the necessary references. . . ).
Short6 list of changes-
1) 1 mA current diodes for R51 and R84 (increase forward gain and input differential linearity)
2) 2 mA current diode replaces R97 (more improved forward gain)
3) replace output resistor R103 with a 1 mH inductor to improve stability with the distortion neutralization cap.
4) add a Zobel network across the output of 1000 pF in series with 249 Ohms.
5) add a 10 pF cap from the node labeled tp12 to the base of Q22 (distortion neutralization cap)
6) I replaced R91 with a pot and upgraded C32 to a film cap. The pot has a small effect and I am not sure its making a difference any more.
After this a normal calibration will yield -120 dB harmonics, pretty much a uniform spectrum with no single harmonic dominating.
This was with one example of a 2400 Hz unit. I can't say that others would respond the same.
The 10V output is constant with load until the current limit is hit (and the uncal light lights).
I'll post more on my efforts towards some useful reference oscillators.
Total Comments 3
Comments
-
Here is a marked up schematic showing the changes. I may do more tests and tweaks if time permits.
Posted 17th July 2012 at 04:34 AM by 1audio
-
Interesting stuff Demian, which of the mods you're showing correspond to Scott's 'distortion neutralizer' changes?
If you've not seen it yet, GK's done some sterling work on THD here: An Audio T.H.D. AnalyserPosted 18th July 2012 at 03:43 AM by abraxalito
-
Update- I needed to check the voltage accuracy of some of my meters. They were all reading different things so it was time to do some calibration. Staring from a Fluke 732A 10V (+/- 50 mV max under my circumstances, probably closer to 5 mV but I have no way to tell) I used the AC-DC transfer capability of a Fluke 8506A to transfer the DC of the 732A to the 510A. a small tweak (100 mV high or 10 PPM so not really significant) and then back to the bench to check meters. Yes, they all say different things. But on the Shibasoku I determined that the harmonics are at -120 dB or better and the THD+N is -110 dB or better. There are upper harmonics in the output, probably artifacts of the S/H leveling circuit. Still very useful as a transfer standard. Frequency is still within a few PPM after sitting for 5 years.Posted 9th March 2017 at 01:45 AM by 1audio
