Modify the PSU of a console: Any ideas?

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Indicator

Member
Joined
Jan 7, 2009
Messages
6
Hi there,



Sadly the PSU of my D&R 8000 is defect, so I have to disconnect it from the console and bring it to a tech. I thought, now that I can not use the console and the PSU is will be given to a tech anyway I could let him modify it a bit. Since I have heard that there is much potetial to im prove the soundquality of the console.



Have you any ideas what could be done? I will give all the ideas to the tech.

If you need any infos, I try to post them here.


Thanks a lot. Have a nice day.
 
Not sure what a D & R is, but I improved my Allen and Heath by simply replacing the bridge rectifier
with fast diodes. Of course PS Caps help as well.
 
abbey road d enfer said:
guavatone said:
Not sure what a D & R is, but I improved my Allen and Heath by simply replacing the bridge rectifier
with fast diodes. Of course PS Caps help as well.
That's interesting! In what respect has it improved your desk?

High Frequency Crud(noise). 
 
guavatone said:
abbey road d enfer said:
guavatone said:
Not sure what a D & R is, but I improved my Allen and Heath by simply replacing the bridge rectifier
with fast diodes. Of course PS Caps help as well.
That's interesting! In what respect has it improved your desk?

High Frequency Crud(noise). 
Yes, it's the only thing a change of rectifier can bring. Strange how this is overlooked by designers (or is it bean counters?). I remember the first Ampex MM1200 that came to France was inducing noise in the otherwise extremely quiet Harrison desk it was paired to. We investigated the problem with a pocket radio. After changing the rectifiers, everything was OK.
 
Indicator said:
Hi there,



Sadly the PSU of my D&R 8000 is defect, so I have to disconnect it from the console and bring it to a tech. I thought, now that I can not use the console and the PSU is will be given to a tech anyway I could let him modify it a bit. Since I have heard that there is much potetial to im prove the soundquality of the console.



Have you any ideas what could be done? I will give all the ideas to the tech.

If you need any infos, I try to post them here.


Thanks a lot. Have a nice day.
I think you should not put to much hope in a dramatic improvement of the sound by tweaking the PSU - unless it is unacceptably poorly designed. Changing the rects as guava suggests may or may not bring an improvement; if the originals are good, there's no improvement to be had. Tweaking the voltages is the most often discussed subject and it's tricky. The expected improvement in dynamic range is not spectacular, and the drawbacks are numerous and hazardous; putting the desk on fire is just one!
The most desirable improvement is reducing the noise of the rails. In some cases, there are some possibilities. If the PSU uses 3 pin regulators of the 78/79 family, it is possible to replace them with LM317/337. It is a relatively easy mod that does not involve much work on the PCB, and the improvement is generally noticeable.
 
I've just replaced the power supply of my console with a lab supply. 2 x 0-30 V / 5 A. I'm totally happy so far, the console is slightly quieter than before (which is a lot, because it is anyway very quiet) and the unit did cost 250 €. All plans to diy something would have turned out way more expensive. I just hope that it is built well enough to last for years, but so far it is perfect.

Michael
 
abbey road d enfer said:
The most desirable improvement is reducing the noise of the rails. In some cases, there are some possibilities. If the PSU uses 3 pin regulators of the 78/79 family, it is possible to replace them with LM317/337. It is a relatively easy mod that does not involve much work on the PCB, and the improvement is generally noticeable.

A related question,

If the 78/79 family based PSU was a good design in the first place, would converting that to LM317/337 amount to anything?

Say, this one (but with equal 1.5A limited regulation).

powersupply.gif


Why would LM317/337 help?
 
Yes, that would work. The LM317/337 are many times quieter than the 78/79. They also have a better protection.
If I understand well, you want to run this psu at ca.200mA. I'm afraid the very low input to output differential may impair the regulation. 3pin regs are perfectly happy with 20v at the input. I understand you want to improve pre-regulation filtering with res. R2 and R12, but they don't leave much headroom for low mains voltage or increased current draw. R1 & R11 are almost useless. R4 & R7 ARE useless. This psu is somewhat overkill: 10 000uF for 200mA! Some power amplifier manufacturers use that for 20 amps!
As R5/8 and R6/9, I understand they are intended for discharging the caps (which is of dubious necessity) but it would make more sense to assign the 1k's to the bigger caps, the 10k's to the smaller caps. If I were you, I would put them off altogether.
You mention 1.5A limitation, that's what the plastic TO 78/79 are rated at, same as the lm317/337, but ther's no way you're gonna have that with the resistor values you have put.
 
What about a supply like this? It's for my Quantum Audio Labs QM-168.

quantumaudioqm168psu.jpg


I bought this Acopian TD15-250 in hopes of bypassing the original power supply altogether:

acopian.gif


I've yet to rack the supply, but do you think this would this be a dramatic improvement based on the schematic of the original supply?

Thanks,
Chris
 
abbey road d enfer said:
Yes, that would work. The LM317/337 are many times quieter than the 78/79. They also have a better protection.
If I understand well, you want to run this psu at ca.200mA. I'm afraid the very low input to output differential may impair the regulation. 3pin regs are perfectly happy with 20v at the input. I understand you want to improve pre-regulation filtering with res. R2 and R12, but they don't leave much headroom for low mains voltage or increased current draw. R1 & R11 are almost useless. R4 & R7 ARE useless. This psu is somewhat overkill: 10 000uF for 200mA! Some power amplifier manufacturers use that for 20 amps!
As R5/8 and R6/9, I understand they are intended for discharging the caps (which is of dubious necessity) but it would make more sense to assign the 1k's to the bigger caps, the 10k's to the smaller caps. If I were you, I would put them off altogether.
You mention 1.5A limitation, that's what the plastic TO 78/79 are rated at, same as the lm317/337, but ther's no way you're gonna have that with the resistor values you have put.

Hey thanks for the analysis. Not my design if that was unclear, just an example of a decent one. There were some tidbits you mentioned I wasn't aware of. There's always more to learn. I was mainly interested in why you said LM317/337 would be an improvement over the 78/79 in general. I guess it's always simply the slightly cleaner standard regulator.

If I was to build this for 0.5 - 1.5A target, I would leave R1, R11, R2, R12, R4, R7 out completely. Just the plain 4700uF pre-regulator would be plenty. Then heatsink adequately, if needed. R6 R9 are definitely handy, for the case of sudden changes in current draw they'd help balance out, especially if C3 C6 are any bigger. The same pre-regulator R5 R8 I think would be unnecessary here.

There are of course some by-the-book LM317/337 upgrades not shown here because they are not valid for plain 78/79.
 
abbey road d enfer said:
guavatone said:
abbey road d enfer said:
guavatone said:
Not sure what a D & R is, but I improved my Allen and Heath by simply replacing the bridge rectifier
with fast diodes. Of course PS Caps help as well.
That's interesting! In what respect has it improved your desk?

High Frequency Crud(noise). 
Yes, it's the only thing a change of rectifier can bring. Strange how this is overlooked by designers (or is it bean counters?). I remember the first Ampex MM1200 that came to France was inducing noise in the otherwise extremely quiet Harrison desk it was paired to. We investigated the problem with a pocket radio. After changing the rectifiers, everything was OK.

Complete nonsense.
Fast recovery rectifiers on a 50 or 60Hz input will worsen any HF noise in your power supply due to much faster switching edges when the rectifier conducts. They will also rectify any HF noise coupled from the mains into the PSU and, without adequate HF suppression, will be coupled into your audio circuitry via the power rails.
 
What a lovely way to express your disagreement!
Apparently, you don't understand the mechanisms of rectification. A fast recovery diode takes less time to enter into conduction, thus the switching involves a smaller charge accumulation across the junction. As to the possibility of rectifying HF noise coming from the mains, there is no significant difference below a few MHz, which is the kind of frequency domain we are interested in.
 
Isn't a schottky diode nearly always a better choice anyway, especially in cases where voltages are below 100V? recovery behaviour is in order of magnitudes better, moving the high frequency crud even further in MHZ range.
 
abbey road d enfer said:
A fast recovery diode takes less time to enter into conduction, thus the switching involves a smaller charge accumulation across the junction.
I think that's what I was trying to say. dv/dt is significantly higher than a conventional rectifier (for the MBR1100 Schottky it is 10V/ns). vf is 0.4-1.0V, depending on if. For a 0.5V step change at switch on in 50ps, there can be significant harmonics in the current waveform, and the possibility of ringing due to the inductance present in the filtering capacitors. With a conventional diode the switch on time is orders of magnitude slower, with a reduction of harmonics and less ringing in stray inductances.
 
But then, why are abbey road d enfer and quavatone reporting from practical experience that noise actually goes down?

The reduction of harmonics using slow diodes theory does not seem to hold up.  ???
 
I guess the solution would be small paralleld bypass-caps to shorten the HF crap across the diodes. But at the end of the day the entire circuit counts. The better (faster) the regulators are the lesser is the trouble introduced from the rectifiers. Hence the rectifiers type is of lesser significance.
Older german broadcast modules were designed each with it´s own regulator working way up into MHz region because the desks were working close to FM transmitters. With regulators like these (+ Mu-metal shielded modules) there was no trouble on the rails.
 

Latest posts

Back
Top