Bob Cordell Interview: Power Supplies

Status
Not open for further replies.
Jump to Latest
#541
Hi,
twin caps before and single cap after the 0r22 filter limits the peak current available from the PSU.
I recommend that you move one of the twinned 10mF to the amp side of the 0r22, thus giving rCrC//C instead of rC//CrC
But do check the ripple current vs ripple current rating of the first cap.

It might be worth doing something similar with the low current supply.
2m2F + 10r + 2m2F + regulator is better than 4m7F + regulator.

Ensure you have a clean ground separate from the dirty ground on the PCB. That then allows the option on where you locate the final ground reference for the whole amp.
 
#543
roender said:
So, the 0V point for SR regulators is well located on amp's PCB or on the common point of second battery caps?
Click to expand...
the dirty ground and the common point for the onboard decoupling should be the same point.
The ground for the regulators is technically a dirty ground and should be taken to the dirty ground or power ground. But I suspect that it may be better to take it to the central ground reference where all the grounds meet. If it is bolted you can experiment.
 
#544
Which one from this two PSU for front end (LTP + VAS)?
I need good stability, low noise and and high PSRR
 

Attachments

  • frontendpsu.jpg
    frontendpsu.jpg
    33.1 KB · Views: 832
#546
roender said:
Which one from this two PSU for front end (LTP + VAS)?
I need good stability, low noise and and high PSRR
Click to expand...


Why not run a sim for those attributes? Noise is simple, Zout is simple (just insert a signal current into the output, measure the signal voltage at the output pin, and divide). PSRR: put a signal source in series with the input battery, measure signal output, sweep over frequency. Divide and show in dB.

Personally, I think from a first glance that B is better.

Jan Didden
 
#552
May be worth a look at Doug Self's book for some useful input on supply line rejection of VAS and input stage and some pragmaitc solutions for limiting the impacts of the inevitable signal components on the rails in a simpe but elegant way. Im sure there are more complicated ways of solving / minimising these problems of course!
 
#553
Though I no longer am invovled in it I used to be an SMPS designer. In that world capacitor ESR is everything for high frequncy high current low ripple performance. OK so the applications in audio is very different but I have often wondered if there are many superb capacitors in the SMPS field that dont carry the brand names audi enthusiaists seem to hold in high esteem. Guess Im saying what is the correlation betwen a quality SMPS capacitor and its audi capability or is that even relevant? Why do I think this is possibly relevant - well to get low ESR challenges the engineering of capacitor design to its limits in that field and maybe as a by product also gives good audio performance? Streeter A.
 
#554
streetera said:
Though I no longer am invovled in it I used to be an SMPS designer. In that world capacitor ESR is everything for high frequncy high current low ripple performance. OK so the applications in audio is very different but I have often wondered if there are many superb capacitors in the SMPS field that dont carry the brand names audi enthusiaists seem to hold in high esteem. Guess Im saying what is the correlation betwen a quality SMPS capacitor and its audi capability or is that even relevant? Why do I think this is possibly relevant - well to get low ESR challenges the engineering of capacitor design to its limits in that field and maybe as a by product also gives good audio performance? Streeter A.
Click to expand...


I'd tend to think that these capacitors might be excellent for audio, at least in those audio applications where ESR matters a lot, especially if they employ a good-quality dielectric like polypropylene.

I recall an earlier thread posting where the so-called four-pole capacitor types were discussed, and I believe that they were originally developed for switchmode power supplies.

Cheers,
Bob
 
#557
streetera said:
May be worth a look at Doug Self's book for some useful input on supply line rejection of VAS and input stage and some pragmaitc solutions for limiting the impacts of the inevitable signal components on the rails in a simpe but elegant way. Im sure there are more complicated ways of solving / minimising these problems of course!
Click to expand...

In a folded cascode amp there is no room for other simple way of line regulation.
I need a separate PSU for first stages (with 10V over outputs) because of voltage lost on folded cascode "VAS" and inherently low PSRR topology
 
#558
roender said:
In a folded cascode amp there is no room for other simple way of line regulation.
I need a separate PSU for first stages (with 10V over outputs) because of voltage lost on folded cascode "VAS" and inherently low PSRR topology
Click to expand...

The PSRR depends on the quality of the current source used in
the folded cascode. If use use a high quality CCS and voltage
references, the PSRR can be quite good, and if the CCS is good
at low voltage, the losses aren't so bad.

😎
 
#559
Hi all,

This thread is so good and it aroused my enthusiasm so much that I dare to contribute with my humble, non-educated, experience :angel:

Roender,
At Pinkfishmedia there are many threads that study the concept of a gyrator followed by a super-regulator (and its variations), with the gyrator taking care of very low ripple/noise at high freqs, where the super-reg seem to be not so good, and the super-reg giving low Zout, wideband. Noise is very, very low.

http://www.pinkfishmedia.net/forum/showthread.php?t=29501&highlight=gyrator+super+regulator

Check out also this one:
http://www.pinkfishmedia.net/forum/showthread.php?t=36174&highlight=gyrator+super+regulator

I have used the "standard version" and one with a gyrator followed by LM317/337, with (sonical) success.

Film caps are preferred for the gyrator. The sonic benefits are many but the most constant is impressive bass and less HF noise...very worthwhile.

Streetera,
I have played with low cost DVD players' SMPS with the recommended mods. Sorry for not being able to make a technical description of what I did but I will try:
At the entry of the SMPS (high V) there are usually a couple of 47-100nf caps that are X or Y type. I swapped them for Epcos polypropilene rated for 100nf/1250VDC. Nothing exploded...and sound improved significantly.

Then there is an electrolitic cap, around 120uF. I tried both ways:
1) swap for a Black Gate FK (or similar series, I don't remember)
2) bypass the stock one with Epcos polipropilene 100nF.
Both ways produce significant improvement in sound.

I would love to test those four pole caps but there are too expensive. I also would love to know how to integrate a gyrator on a SMPS.

Sorry if all this is already known. :angel:

One of the good parts of this thread is recognising the fact that some renowned members admit PS related doubts that are the same that the ethernal beginner fan confronts when planning/building his firts projects 🙂
 
#560
maxlorenz said:
Hi all,

This thread is so good and it aroused my enthusiasm so much that I dare to contribute with my humble, non-educated, experience :angel:

Roender,
At Pinkfishmedia there are many threads that study the concept of a gyrator followed by a super-regulator (and its variations), with the gyrator taking care of very low ripple/noise at high freqs, where the super-reg seem to be not so good, and the super-reg giving low Zout, wideband. Noise is very, very low.

http://www.pinkfishmedia.net/forum/showthread.php?t=29501&highlight=gyrator+super+regulator

Check out also this one:
http://www.pinkfishmedia.net/forum/showthread.php?t=36174&highlight=gyrator+super+regulator

I have used the "standard version" and one with a gyrator followed by LM317/337, with (sonical) success.

Film caps are preferred for the gyrator. The sonic benefits are many but the most constant is impressive bass and less HF noise...very worthwhile.

Streetera,
I have played with low cost DVD players' SMPS with the recommended mods. Sorry for not being able to make a technical description of what I did but I will try:
At the entry of the SMPS (high V) there are usually a couple of 47-100nf caps that are X or Y type. I swapped them for Epcos polypropilene rated for 100nf/1250VDC. Nothing exploded...and sound improved significantly.

Then there is an electrolitic cap, around 120uF. I tried both ways:
1) swap for a Black Gate FK (or similar series, I don't remember)
2) bypass the stock one with Epcos polipropilene 100nF.
Both ways produce significant improvement in sound.

I would love to test those four pole caps but there are too expensive. I also would love to know how to integrate a gyrator on a SMPS.

Sorry if all this is already known. :angel:

One of the good parts of this thread is recognising the fact that some renowned members admit PS related doubts that are the same that the ethernal beginner fan confronts when planning/building his firts projects 🙂
Click to expand...

Thanks for contributing. These are all good points. I have always believed that the best designs have a very quiet power supply AND very good PSRR (including up to fairly high frequencies).

Cheers,
Bob
 
Status
Not open for further replies.
Top Bottom