For the Boonton you can get the ranging info from the detector board, U2 latches in all of the settings. The gain settings may be a little clumsy to work with. You can manually override the gain settings easily. If you select the mode (e.g. distortion) and the range (.1%) the gains are preset. It also works for frequency. This is to enable faster testing in automated systems but can be used for this I suspect. Its not documented well and I'm not sure how it works but in some quick testing it seems to work to set a gain after selecting distortion. switching to a different mode clears the selection it seems.
U7 is amp on the second gain stage. its a single opamp so I'm not sure what your looking at. (Detector board).
That could be between six and sixtyfour relays. How may relays are there.
What I would do is monitor the six lines and run the Boonton through it's ranges recoding the state of the control lines for each range.
Last edited:
No on the input/autoranging board, it's a TL072 with 20meg resistors on it's input.
On the autoranging decode, I would measure the attenuation for each of the 64 values. Will take some manual labour time but it's only 64 measurements.
Then draw up a table with the dB value for each 6 bit value and put the table in the controller. The software just reads the 6 bit value, picks the corresponding dB value from the table and puts that out on a standard A/N display.
Jan
First- I figured this out for a calibration for the analog monitor out-
If you enter .01% -80 dB from the fundamental will be 1V at the monitor out. If you enter .1% 1V= approx -60 dB. The actual value on mine is -80 dB= +1.3 dB over 1V but that small difference can be ignored usually. This beats struggling with a complicated outboard circuit.
Second- the TL072 on the input board is part of the DC measurement system. It switches in extra gain and alow pass filter to measure DC voltages at the input. Its reasonably good down to a few millivolts.
If you enter .01% -80 dB from the fundamental will be 1V at the monitor out. If you enter .1% 1V= approx -60 dB. The actual value on mine is -80 dB= +1.3 dB over 1V but that small difference can be ignored usually. This beats struggling with a complicated outboard circuit.
Second- the TL072 on the input board is part of the DC measurement system. It switches in extra gain and alow pass filter to measure DC voltages at the input. Its reasonably good down to a few millivolts.
Its the same thing I do. I just look at 1volt of the monitor output as full-scale on what ever thd range I am on. This was shown/demonstrated when I did some THD work using the ShibaSoku 725D meter and HP spectrum analyzer was used on the monitor port. Just use the ADC/FFT the same way on the monitor port.
But it is the nonlinearity which needs a look-up table to correct the data. That isn't so easy.
The 2H is always too high and stimulated and present at any level. The 3H is level variable. The 4-5H are minimal at lowest levels and continuously increases with input level.... those are not present in signal source. So it is a complex variable which might not lend itself to a simple error correction look-up table. Might be a combination of ills.
Here is an example at approximately .5 volt (no notch filter) of the complexity of the issue. The right is the source (ShibaSoku AG16B) as measured by the A-P and the left, is the QA400 harmonic data. --- not even a close approximation.
Later, I have played with the B&K 1607 as the notch filter with mixed results. Will try again.
Thx-RNMarsh
But it is the nonlinearity which needs a look-up table to correct the data. That isn't so easy.
The 2H is always too high and stimulated and present at any level. The 3H is level variable. The 4-5H are minimal at lowest levels and continuously increases with input level.... those are not present in signal source. So it is a complex variable which might not lend itself to a simple error correction look-up table. Might be a combination of ills.
Here is an example at approximately .5 volt (no notch filter) of the complexity of the issue. The right is the source (ShibaSoku AG16B) as measured by the A-P and the left, is the QA400 harmonic data. --- not even a close approximation.
Later, I have played with the B&K 1607 as the notch filter with mixed results. Will try again.
Thx-RNMarsh
Last edited:
Demian
That doesn't surprise me... Do you think buyers of the QA400 know that? By my measurement, I said that I wouldn't use it below -90 and not with an input greater than -10dBv.
Next up--- who make the best sound card using the best codec? Is the e-Mu 0404 one of the best? The noise floor might be very low --- is it accurate has been my question.
With a notch filter, as A-P and many here and else where have done, it can be better. AND, that brings us full circle back to the features of a notch filter...a few fixed freqs or auto-tune or variable? These are options for the DIY'er. Having used all these types of notch filters, I would not be happy with a few fixed freq BUT it would be the cheapest. It would still have to be checked for accuracy at low THD levels.
I bought two QA400 a long time ago and found it not to fit my quest. It has not been my interest to use a QA400 since it does not go below-100dB accurately. But just to help others and show the true limits of its' capability....in terms of accuracy.... I did some tests regarding accuracy for them.
For many it is fine without a notch filter....as are many 'sound card' type products. I like the software integration which makes it P-N-P to use. Though I had problems with it initially. Those issues in software seem to have been resolved.
THx-RNMarsh
That doesn't surprise me... Do you think buyers of the QA400 know that? By my measurement, I said that I wouldn't use it below -90 and not with an input greater than -10dBv.
Next up--- who make the best sound card using the best codec? Is the e-Mu 0404 one of the best? The noise floor might be very low --- is it accurate has been my question.
With a notch filter, as A-P and many here and else where have done, it can be better. AND, that brings us full circle back to the features of a notch filter...a few fixed freqs or auto-tune or variable? These are options for the DIY'er. Having used all these types of notch filters, I would not be happy with a few fixed freq BUT it would be the cheapest. It would still have to be checked for accuracy at low THD levels.
I bought two QA400 a long time ago and found it not to fit my quest. It has not been my interest to use a QA400 since it does not go below-100dB accurately. But just to help others and show the true limits of its' capability....in terms of accuracy.... I did some tests regarding accuracy for them.
For many it is fine without a notch filter....as are many 'sound card' type products. I like the software integration which makes it P-N-P to use. Though I had problems with it initially. Those issues in software seem to have been resolved.
THx-RNMarsh
Last edited:
How about using both QA400 channels for one test and get some noise and distortion reduction? isn't this something QA people has info on?
-- or what about using a stereo ADC and doing a sum or difference with it? Like in Cirrus Logic AN331? Can this be done by mod'ing the QA400?
Thx-RNMarsh
-- or what about using a stereo ADC and doing a sum or difference with it? Like in Cirrus Logic AN331? Can this be done by mod'ing the QA400?
Thx-RNMarsh