This thread is split from:Super Regulator
____________________________________________________________________________
I am uploading my preview document for a tester useful towards exercising SR circuits. It offers useful info on cap ESR and regulator Zout, using the AD825 in a SR2020 circuit.
It will also be available from my website via this link.
Walt Jung
Last edited by a moderator:
Tester schematic
Patrick, the short answer is NO, for a couple of reasons. One is the basic reason that a high degree of technical skill is necessary to make the output useful. As so stated in the Tester Docs. Plus the data obtained is highly specialized.
You can build your own of course, along these lines: A CMOS 555 configured for a low duty cycle, followed by an RC filter to define risetime, and a high current linear switch stage that dumps current into the reg-under-test output. A low noise gain of 100x preamp that reads the AC error from the reg output. You will need to calibrate it all of course, and this is very difficult considering the burden of low-level signals.
I have been informed that you got a tester schematic via a back door, so this was without my permission. The tester schematic was shared with only a few people, and in confidence. Sorry that this kind of thing happens here. But, it is also ironic that the sch you did get is obsolete. I have upgraded the circuit with some refinements since the published photos were taken.
I think it best for me to take leave of this thread, with all such nonsense going on. I am asking the DIYaudio people to remove my name from any and all stuff they are marketing. I do not endorse this regulator PCB as being offered here. They never asked me to begin with, they just used my name. But let's hope they can now find religion and do the right thing.
I can be reached for questions via my website, index.html.
Patrick, the short answer is NO, for a couple of reasons. One is the basic reason that a high degree of technical skill is necessary to make the output useful. As so stated in the Tester Docs. Plus the data obtained is highly specialized.
You can build your own of course, along these lines: A CMOS 555 configured for a low duty cycle, followed by an RC filter to define risetime, and a high current linear switch stage that dumps current into the reg-under-test output. A low noise gain of 100x preamp that reads the AC error from the reg output. You will need to calibrate it all of course, and this is very difficult considering the burden of low-level signals.
I have been informed that you got a tester schematic via a back door, so this was without my permission. The tester schematic was shared with only a few people, and in confidence. Sorry that this kind of thing happens here. But, it is also ironic that the sch you did get is obsolete. I have upgraded the circuit with some refinements since the published photos were taken.
I think it best for me to take leave of this thread, with all such nonsense going on. I am asking the DIYaudio people to remove my name from any and all stuff they are marketing. I do not endorse this regulator PCB as being offered here. They never asked me to begin with, they just used my name. But let's hope they can now find religion and do the right thing.
I can be reached for questions via my website, index.html.
Last edited:
This means you do not approve or endorse the pcb sold by DIYaudio for building your regulator?
Any specific details or parts or design you do not approve and go against your original design published in The Audio Amateur and other Old Colony magazines?
An_Improved_Reference_Filter_for_Audio_Regulators. pdf
Here is a very simple technique for improving the reference noise rejection in a SR (or similar) audio regulator.
I was testing some Nichicon audio caps for leakage, and found that they were excellent. Well below 1uA in fact. This allows the filter input resistor to be increased, to 5k from 500 ohms, and voila! Better noise rejection. Read about it in the PDF.
Link to same file on my website: https://refsnregs.waltjung.org/An_Improved_Reference_Filter_for_Audio_Regulators.pdf
Here is a very simple technique for improving the reference noise rejection in a SR (or similar) audio regulator.
I was testing some Nichicon audio caps for leakage, and found that they were excellent. Well below 1uA in fact. This allows the filter input resistor to be increased, to 5k from 500 ohms, and voila! Better noise rejection. Read about it in the PDF.
Link to same file on my website: https://refsnregs.waltjung.org/An_Improved_Reference_Filter_for_Audio_Regulators.pdf
I might be tempted to connect a 100 Kohm resistor in parallel with the low leakage capacitor. In that case, the current in the 100K resistor is (7 / 1E5) = 70 microamps ... so the capacitor's leakage is a small to negligible fraction of the total.
With this new 100K resistor, it's just a plain ordinary resistive voltage divider, and plain ordinary DC analysis applies
Capacitor leakage has been rendered ~ insignificant and you can choose a series resistor "Rser" even larger than 5K if it pleases you. Certainly it will improve the rejection of Vref noise compared to 5K.
Just use "Vref_out" instead of "7V" when calculating the values of the resistors in the regulator's feedback network, to get the desired output voltage. Or, as DIYers often do, include a multi-turn trimmer pot.
With this new 100K resistor, it's just a plain ordinary resistive voltage divider, and plain ordinary DC analysis applies
- Vref_out = (Rpar / (Rser+Rpar)) * Vref_in
Just use "Vref_out" instead of "7V" when calculating the values of the resistors in the regulator's feedback network, to get the desired output voltage. Or, as DIYers often do, include a multi-turn trimmer pot.