High-quality power supply for a gainclone not needed ?

analog_sa · 2004-01-31 6:44 pm

Whenever I used those AC filters on primary (in amps), the sound degraded substantially.

I know exactly what you mean. Mains filters don't work too well with me either; maybe i've only tried the wring ones.

You are right to mention mains filters in connection to chokes - my first experiments with chokes in tube circuits were also disheartening - good bass but really bad everything else. Apparently some listeners are not bothered

Eventually it turned out the choice of chokes is critical - same as caps they're not made equal.

Blitz · 2004-01-31 6:50 pm

Yup. I think you would not eveb come to the idea to compare the little coils in the AC-filters with chokes of a quality like these:

http://www.ae-europe.nl/smoorspoelen_engels.htm

Basically, they make anything you want in any size. So, I have as well very small tiny chokes for my ONO. Extremly nice.

Blitz · 2004-01-31 6:53 pm

Forgot to write: Currently I build my 100 Watt Aleph-X with those C-Core-Chokes (40mH@6A). Pretty impressive piece of iron (meant for choke-input, that why the inductance is so high).

cocolino · 2004-01-31 8:04 pm

Forgot to write: Currently I build my 100 Watt Aleph-X with those C-Core-Chokes (40mH@6A). Pretty impressive piece of iron (meant for choke-input, that why the inductance is so high).

Hallo Blitz,

this must be monster chokes!
Do You have a picture of them?
How much did You pay for one of those chokes?

analog_sa · 2004-01-31 8:15 pm

Forgot to write: Currently I build my 100 Watt Aleph-X with those C-Core-Chokes (40mH@6A). Pretty impressive piece of iron (meant for choke-input, that why the inductance is so high).

Very cool. How do the chokes sound compared to a wire short?

Pedja · 2004-01-31 8:42 pm

Blitz said:
Besides chokes, here is a regulator suggested for Naim-equipment:

May I humbly suggest you reading this page?

Pedja

tecnofossil · 2004-01-31 11:27 pm

Pedja said:
May I humbly suggest you reading this page?

Pedja [/QUOTE

Thanks for the suggestion. Another possible solution(provided you aren't needing more than 36 volts per rail), suggested in a old National Semiconductor app manual, is to parallel LM317 (or LM337 if a negative supply is required) chips to meet the current requirements of the power amp. I am drawing on old, faint memories here so please bear with me. I'll see if my old National manual from 1982 has the info.

-Tom

Hybrid fourdoor · 2004-02-01 9:33 am

I once asked this question and the general concesus was that you cannot just parrellel regulators, rather there is more circuitry that is involved. You may know this or not, but thought I'd share, in case some got some bright ideas, but I could be totally mistaken.

Looking at National's datasheet for the LM317, they show some possible high current designs.

tbla · 2004-02-01 10:01 am

......ok, i don't get it!!!!!!!

why don't you gentlemen try this.......

a pair of lt1033/lt1085
or
a pair of lt1084
or
a pair of lm1084
or
a pair of lm338
or
a pair of lt1083

its very easy - and it works.........

tecnofossil · 2004-02-01 12:16 pm

tbla said:
......ok, i don't get it!!!!!!!

why don't you gentlemen try this.......

a pair of lt1033/lt1085
or
a pair of lt1084
or
a pair of lm1084
or
a pair of lm338
or
a pair of lt1083

its very easy - and it works.........

Because, here in the US we have places called Radio SHack which stock these chips, so quick test could be hacked together quickly for a test. National also makes a IC reg, a LM138, (I think anyway), that will supply about 5 amps.

-Tom

Blitz · 2004-02-01 12:22 pm

Pedja said:
May I humbly suggest you reading this page?

Pedja

Very nice page ! Great reading. At the end: Was was your expereince soundwise to use the non-feedback regulator compared to a non-regulated supply ?

Best Regards

iCebReakEr409 · 2004-02-01 12:42 pm

Hybrid fourdoor said:
I once asked this question and the general concesus was that you cannot just parrellel regulators, rather there is more circuitry that is involved. You may know this or not, but thought I'd share, in case some got some bright ideas, but I could be totally mistaken.

Looking at National's datasheet for the LM317, they show some possible high current designs.

yeah i was wondering the same thing. lm317 are pretty cheap. or you could use the 5amp lm338.

Pedja · 2004-02-01 8:08 pm

Blitz said:
Very nice page ! Great reading. At the end: Was was your expereince soundwise to use the non-feedback regulator compared to a non-regulated supply ?

Thanks. There was not a direct comparison of these two. After I tried LM338 I gave up of using unregulated supply in GC. So the direct comparison I could report about was between non-f/b and LM338 based one and I preferred non-f/b. A bit more of work should be done but I think it worth it.

Pedja

cocolino · 2004-02-01 8:34 pm

Thanks. There was not a direct comparison of these two. After I tried LM338 I gave up of using unregulated supply in GC. So the direct comparison I could report about was between non-f/b and LM338 based one and I preferred the first. A bit more of work should be done but I think it worth it.

Hi Pedja,

Interesting!
Would You mind to elaborate in a bit more detail what were the differences soundwise between unregulated/LM338 regulated and LM338 regulated/non-feedback regulated?
Did You find that the non-feedback circuit was better in all respects or are there also any drawbacks ?

Pedja · 2004-02-01 9:30 pm

Hi Christoph,

I have posted something about unregulated vs. regulated supply here:
http://www.diyaudio.com/forums/showthread.php?postid=231896#post231896

Non-feedback regulator was more natural and dynamic and probably cleaner than LM338 based. The simplest non-f/b option vs. LM338 could still be a matter of taste since LM338 might appear as more refined. With the improvements on non-f/b regulator I have described, I do not think that there are drawbacks.

There were some reports agreeing with my findings but I know also at least one that does not. So, as always, it is the best if you check it yourself.

Pedja

cocolino · 2004-02-01 10:24 pm

There were some reports agreeing with my findings but I know also at least one that does not. So, as always, it is the best if you check it yourself.

Pedja,

thanks for the link and the explanation!
As I`m currently planning a small gainclone based powered (active) system, actually I`ll give this a try provided that I find sufficient space inside the case (I have already) and that I can get rid of the additional heat wasted in the regulators.

High-quality power supply for a gainclone not needed ?

Member

Member

Member

Member

Member

Member

Member

Member

Attachments

Member

Member

Member

Member

Member

Member

Member

Member