If you would buy Krell and later built a Gainclone, chances are you'd be using Krell as a nice stand for chip amp.😉
http://www.sakurasystems.com/show/chrissystem.html
http://www.sakurasystems.com/show/chrissystem.html
PaulHilgeman said:
66,000uf per voltage rail
dont know, some cheap ones i got out of a power supply
+/- 15v at present, i am building a +/- 30v one at the moment
8R
Panasonic FC caps; 120's in the PSU and 1000's on each rail on the chip pins. (1000's are bypassed by .1ufd polyprops.)
Very happy with sound but will experiment further. That's half the fun with these amps.
25-0-25 transformer (120VA) provides +/-36 volts on the rails.
Very happy with sound but will experiment further. That's half the fun with these amps.
25-0-25 transformer (120VA) provides +/-36 volts on the rails.
🙂Peter Daniel said:
BWAAAHAHAHAHAHA! That is TOO funny.
Now I know I have to build one of these and see what the fuss is all about.
Carlos,
10000µF would be the minimum I would expect to see with a p/p class AB amp. I can't argue with listening tests, but it does surprise me that everyone is using so little capacitance! I guess i will have to find out for myself.
Why not simply having the best of both worlds?
When it`s true that a Gainclone amp sounds best (above the bass range) with only 1000µF rail filter cap, IMO the obvious next logical step would be an active multiamp system (especially as those chips do not cost more than good passive filter parts) with seperate rectifiers and 1000µF cap for the mid/high amp(s) and some "normal" cap values for the bottom range amp(s).
Or in a threeway system let`s say: 1000µF/tweeter-amp; 3300µF/mid-amp; 10000µF/woofer amp.
When it`s true that a Gainclone amp sounds best (above the bass range) with only 1000µF rail filter cap, IMO the obvious next logical step would be an active multiamp system (especially as those chips do not cost more than good passive filter parts) with seperate rectifiers and 1000µF cap for the mid/high amp(s) and some "normal" cap values for the bottom range amp(s).
Or in a threeway system let`s say: 1000µF/tweeter-amp; 3300µF/mid-amp; 10000µF/woofer amp.
I went from 4700uf caps on the chip to 2x680uf BGN super -E configuration. So less is more in my case.Bypass on the bridges with 2,2uf polycarb and 0.1uf tin and foil was best combination at least for me.
Hot stand
Peter Daniel said:If you would buy Krell and later built a Gainclone, chances are you'd be using Krell as a nice stand for chip amp.😉
Oh, man, that would be a hot stand!😎
I won't buy a Krell, because I don't have the money.
But I was thinking on trying something like cocolino suggests, I had thought of something similar.
Maby even bi-amp with LM3875s for the tweeters and LM3886 for the woofers.
Deam it, the mail system guys here in Portugal are on strike, I'm waiting for my transformer for a week!!!
One channel is ready and no transformer.
I understand, my Digikey order is currently taking a tour of Canada courtesy of Purolator
based on the tracking info the package was in a depot just a few miles away before this tour started
Bypass, bypass, bypass...
Protos, man, try to put the 4700uf caps back, and bypass it with, say, a 10uf electrolythic and a 0.1uf poly (or even ceramic!).
If you prefer, you can forget the 10uf cap.
You may be surprised.
I think people shouldn't use a big cap alone.
The better results you had with the 680uf caps are because they are faster and have less resistance.
But just try to bypass the big ones.
My first channel is done that way (with 1000 uf caps, for now), but the deam transformer doesn't arrive!...
protos said:
Protos, man, try to put the 4700uf caps back, and bypass it with, say, a 10uf electrolythic and a 0.1uf poly (or even ceramic!).
If you prefer, you can forget the 10uf cap.
You may be surprised.
I think people shouldn't use a big cap alone.
The better results you had with the 680uf caps are because they are faster and have less resistance.
But just try to bypass the big ones.
My first channel is done that way (with 1000 uf caps, for now), but the deam transformer doesn't arrive!...
Well I had bypassed the 4700uf as well and of course they sounded better but I think the BGN with super-e configuration are a special breed by themselves and maybe should not be compared just on total uf. In any case many members here with experience in BG's say that they should be counted as having 3 times more capacitance than stated.
I don't think one should be worried about not having enough capacitance just because there is an audiophile tradition that says more is always better .
There is no question that with the gc you don't really need a lot of uf for ripple smoothing purposes. Now comes the question of how capacitance affects the perceived speed or sound of the amp.
I guess here it is a matter of experimentation and everybody can choose a point that suits their sonic preferences and system particularities.It seems that some prefer a lot and then many are happier with less.
I don't think one should be worried about not having enough capacitance just because there is an audiophile tradition that says more is always better .
There is no question that with the gc you don't really need a lot of uf for ripple smoothing purposes. Now comes the question of how capacitance affects the perceived speed or sound of the amp.
I guess here it is a matter of experimentation and everybody can choose a point that suits their sonic preferences and system particularities.It seems that some prefer a lot and then many are happier with less.
Caps
Ok, Protos,
As I said, i'll test that.
In fact, I'll start with 1000uf, that's what I have now.
If only that deam transformer arrived!
Anyway, capacitance is not audiophile tradition.
That's electronics principles.
Of course someone always have to make a revolution to these established ideas.
Maby this is one of those revolutionary ideas that work.
But it's curious to see people saying that a Gainclone works better with batteries.
I also think that may be true, as always.
A battery is much less noisy than any PSU.
Isn't this also a contradition?
I mean, if you made a good stabilized PSU you could get near of the sound that bateries produce.
But of course that regulated PSU woudn't have only 1000uf caps per rail.
Everytime I think of these things I feel a will to add more caps.
Well, doesn't matter, I'll test as soon as possible.
Ok, Protos,
As I said, i'll test that.
In fact, I'll start with 1000uf, that's what I have now.
If only that deam transformer arrived!
Anyway, capacitance is not audiophile tradition.
That's electronics principles.
Of course someone always have to make a revolution to these established ideas.
Maby this is one of those revolutionary ideas that work.
But it's curious to see people saying that a Gainclone works better with batteries.
I also think that may be true, as always.
A battery is much less noisy than any PSU.
Isn't this also a contradition?
I mean, if you made a good stabilized PSU you could get near of the sound that bateries produce.
But of course that regulated PSU woudn't have only 1000uf caps per rail.
Everytime I think of these things I feel a will to add more caps.
Well, doesn't matter, I'll test as soon as possible.
Re: Caps
If you would be using 1000u BG type N, at $ 80 a pc, you wouldn't feel that will.😉
I ordered some of those and will try them. Currently I'm using 1000u standard grade BG, and I tried to parallel them with 10,000 Jensen 4 pin caps (which by some are regarded as competition to BG). To tell the truth I didn't hear any improvement, and mostly degradation . Even additional 1000u BG (in the PS), degraded the sonics.
carlosfm said:
Everytime I think of these things I feel a will to add more caps.
If you would be using 1000u BG type N, at $ 80 a pc, you wouldn't feel that will.😉
I ordered some of those and will try them. Currently I'm using 1000u standard grade BG, and I tried to parallel them with 10,000 Jensen 4 pin caps (which by some are regarded as competition to BG). To tell the truth I didn't hear any improvement, and mostly degradation . Even additional 1000u BG (in the PS), degraded the sonics.
Re: Re: Caps
Man, that's very expensive (at least for me).
I think I'll continue to search for audio nirvana with more real-world caps.
The secret is in acheaving a good result without spending lots of money.
Of course BGs would be better...
But you can't always have what you want.😉
Where did I hear this?
There's a guy's wife here that works in a car battery store and buys them at half the price.
Nobody has a wife that works on an electrocnics components store?
Ok, ok... Just asking!...
Peter Daniel said:
Man, that's very expensive (at least for me).
I think I'll continue to search for audio nirvana with more real-world caps.
The secret is in acheaving a good result without spending lots of money.
Of course BGs would be better...
But you can't always have what you want.😉
Where did I hear this?
There's a guy's wife here that works in a car battery store and buys them at half the price.
Nobody has a wife that works on an electrocnics components store?
Ok, ok... Just asking!...
Re: Re: Re: Caps
I remember Stones used to sing: "You can't always get what you want"...😉
For me those caps are expensive too, but still cheaper than 8 batteries and a charger. Also it is next step progression and desire for never ending improvements.
Also considering that the amount of BG N type caps for an amp like Aleph would be insanely expensive, I don't feel like wasing money here.😉
carlosfm said:
But you can't always have what you want.😉
Where did I hear this?
I remember Stones used to sing: "You can't always get what you want"...😉
For me those caps are expensive too, but still cheaper than 8 batteries and a charger. Also it is next step progression and desire for never ending improvements.
Also considering that the amount of BG N type caps for an amp like Aleph would be insanely expensive, I don't feel like wasing money here.😉
nobody special
One thing that gets overlooked is that these gainclones prefer to be operated strictly Class B and so, many of the tenets of traditional Class AB power delivery become less important. The chip has excellent ripple rejection at exactly the same frequency where you would derive benefit from large capacitance. This amp thrives on fast power pulses with perfect phase delivery. If one capacitor has sufficiently low ESR it does the job admirably. Oh, and I should also mention that I suspect matching the ESR characteristic of these caps would also yield sonic benefits.
Caps & batteries
Peter,
8 batteries?
To get a voltage similar to a 2*24V transformer you would need 6 batteries (3 in series per rail).
Anyway, you need 4 BGs (two per channel), at $80 each (the cheaper ones)...
I think that batteries and charger, comparing to the price of 4 BGs, would be cheaper.
I received some publicity from a supermarket in my mailbox, and they have car batteries at around 25 Euros each.
Well... BGs are more conveinient than batteries, and even take less space.
Like Bricolo said once, you go to bed without thinking you have to charge your toroid.
Peter Daniel said:
Peter,
8 batteries?
To get a voltage similar to a 2*24V transformer you would need 6 batteries (3 in series per rail).
Anyway, you need 4 BGs (two per channel), at $80 each (the cheaper ones)...
I think that batteries and charger, comparing to the price of 4 BGs, would be cheaper.
I received some publicity from a supermarket in my mailbox, and they have car batteries at around 25 Euros each.
Well... BGs are more conveinient than batteries, and even take less space.
Like Bricolo said once, you go to bed without thinking you have to charge your toroid.
Re: Re: Caps
Peter,
I think that perhaps that would not be the most cost efficient way to use BGN's.
If you use only 4 that would be 320 USD and if you wanted Super E that would set you back 640 USD (ouch!)
However with 8 x 680uf BGNx 35V (assuming your rails are below that) you get 1360 uf per rail per chip in super -e at about 90 USD total or thereabouts. That's a huge saving !
Peter Daniel said:
Peter,
I think that perhaps that would not be the most cost efficient way to use BGN's.
If you use only 4 that would be 320 USD and if you wanted Super E that would set you back 640 USD (ouch!)
However with 8 x 680uf BGNx 35V (assuming your rails are below that) you get 1360 uf per rail per chip in super -e at about 90 USD total or thereabouts. That's a huge saving !
it would be interesting to see what ratio of BG capacitance to
standard electrolytic capacitance would still offer the "BG sound"
maybe there is a $weet $pot...
standard electrolytic capacitance would still offer the "BG sound"
maybe there is a $weet $pot...
- Status
- Not open for further replies.