LM338 regulated snubberized PSU for audio amplifiers

[AR2]

What about class A amps? How this apply to them? What capacitance to use behind the regulator in lets say Aleph 30 or Aleph 5 situation?

 

[carlosfm]

From my point of view, the regulated supply has a permanent 8 volt voltage drop which is probably greater than that of the unregulated PSU.

I meant the voltage drop on the chip's pins, on the amp.
On an unregulated PSU that is modulated with the music, depending on how much current the amp is asking from the PSU.

 

[carlosfm]

What about class A amps? How this apply to them? What capacitance to use behind the regulator in lets say Aleph 30 or Aleph 5 situation?

You don't need more than 10,000uf before the reg.

 

[Greg Erskine]

I meant the voltage drop on the chip's pins, on the amp.
On an unregulated PSU that is modulated with the music, depending on how much current the amp is asking from the PSU.

I understood what you meant, I was just trying to say that to achieve a sag free PSU, we have to add a regulator that permanently drops, in this case 8 volts. The power used by the regulator is not avaliable to the amp. That's all.

 

[homer09]

I understood what you meant, I was just trying to say that to achieve a sag free PSU, we have to add a regulator that permanently drops, in this case 8 volts. The power used by the regulator is not avaliable to the amp. That's all.

i think logically, less regulated power is better than more but sagging power.

also, chips have a voltage limit, so if you like listening to them at say 32 volts, use a 40V trafo regulated down to 32. so in this case, yes those extra 8 volts are not available, but you dont want them to be anyway.

but i see what you mean if you are using the same trafo for both situations.

 

[carlosfm]

also, chips have a voltage limit, so if you like listening to them at say 32 volts, use a 40V trafo regulated down to 32.

 

[homer09]

haha, i know these are way to high for LM series chips. i just pulled numbers out of the air. no need for fire warnings!

 

[carlosfm]

haha, i know these are way to high for LM series chips. i just pulled numbers out of the air. no need for fire warnings!

Didn't you find out yet?
With a 40V trafo you will get 56V

 

[Upupa Epops]

I wasn't a few days here, but I see - neverending story .

 

[Franz G]

I wish, you still have a happy and sunny sunday, regardless of neverending stories

Kind regards
Franz

 

[Upupa Epops]

Here is not so sunny, Franz . What are you making new ?

 

[carlosfm]

What are you making new ?

I think Franz is recycling now.

 

[Upupa Epops]

Do you mean " ECC 88 like New kid in the Town " ?

 

[Franz G]

... I am planning and building my external DAC, with tube (sorry, valve) output stage, based on a CS8412 and a TDA1541A

This is the idea about the audio stage:

cheap but very good PCL82, voltage gain with the triode part of it, DC-coupled to a cathode follower, the pentode part, to reduce the output Z.

 

[carlosfm]

... I am planning and building my external DAC, with tube (sorry, valve) output stage, based on a CS8412 and a TDA1541A

If that's a NOS dac, you need to compensate for the high-frequency roll-off.

 

[Franz G]

Carlos

I send you a mail, because it is off topic here

Franz

 

[Franz G]

What are you making new ?

... and because all my LM3875 samples are used, I must start to use LM3886

So, I prepared a new stereo pcb, snubberized and high cap, two trannies, separated grounds, OPA627 buffer, T-net on the board...

Franz

 

[jackinnj]

Have you actually measured the PSRR of a chipamp?

I'll ask the question again -- we can all cut and paste something from National Semi's website -- where's the "before" and "after" -- with the various snubbers --

the problem hay not "where" you describe,
nay solution "what" you proscribe

(with apologies to the Robbie Burns of Audio)

 

[carlosfm]

I'll ask the question again -- we can all cut and paste something from National Semi's website -- where's the "before" and "after" -- with the various snubbers --

If you have the time for that, be my guest.
I have no time for this, sorry.

DELETE FROM SNUBBER WHERE PSRR IS NULL;

 

[jackinnj]

If you have the time for that, be my guest.
I have no time for this, sorry.

The snubber in your configuration does little to cure a possibly non-existent PSRR problem.

Better to put the 1 ohm resistor IN SERIES with the power supply pins, and use a 100nF/10uF immediately adjacent to Vdd/Vss.

Better still, put a ferrite bead on each power supply pin -- and save the watts for the music.