Peter Daniel said:
No more so than one could conclude the same from your post-
http://www.diyaudio.com/forums/showthread.php?postid=146024#post146024
I simply attempted to offset some of the cost of those caps which I sourced for a friend by buying a larger quantity. No one was interested so I just bought enough for those needs. Unlike you, I've never suggested or reccomended Muses to anyone on this board. (though I did suggest the Rubycon ZL as a possible alternative to the Panasoninc FC in the past).
I will attempt to be clearer- 2 months ago, you stated your subjective dislike for the Muse caps in power supply and decoupling applications. Nevertheless, you chose to use those very same caps in your implementation of Carlos's design to compare to your preferred implementation of your own Gainclone. What is germane here is not my view of the Muses (which I've never given) but your subjective views on the Muse caps, prior to your using them to evaluate Carlos's topology. I am saying that, in this situation, you are not acting with intellectual honesty, as an 'honest broker'. All you have done is verify your own prior stated, subjective reaction to the Muse caps, not Carlos's topology.
I'm certainly not accusing you of not doing your best.
Don't get annoyed Carlos, at least 99% of us are happy, I'm actually surprised Peter is giving up on this amp.
Are you still using your TDA1543 dac Peter?
If my amp had boomy bass then this would certainly show up with my IPL ribbons
Are you still using your TDA1543 dac Peter?
If my amp had boomy bass then this would certainly show up with my IPL ribbons
I do not know the result of using buffer but I have done abit of experimenting with power supply;
I started with standard NIGC with 1500 pana caps on 3875 PCB
Xfrmr plitron 400VA 2x22 using Brianm's boards.
Sound was clean, amp very quiet no noise no hum BUT after having it runninf for a couple of weeks I was dissatisfied with the
sound particularly with heavy bass but just with demanding passages. Pipe organ and full classical orchestra, 1812 on Telarc
Philips test record by Technics really showed it up.
I checked the voltages dynamically and found the sag of up to 10%. If you do the maths this represents a power drop of about 20%. I do not know how this nonlinearity can not affect the quality of sound. This creates a high degree of compression. There was absolutely no voltage sag from the Xfrmr, it was all past the rectifiers.
So I plugged in a double set of PS boards so I have a full set of diodes in parallel, I then added 10,200 uf at the PS PCB. the voltage sag was much better only 3% now.
The dynamics were better but I was stiill not happy, I added T03 form of LM338 regs on heatsinkfollowing the Application Note.
It helped the regulation but not the overall sound.
So I removed the regs.
Carlos suggested much smaller caps on the 3875 board, on second amp I used 47uf bypassed with 0.1 ufpoly and used a 100VA 2x25 Xfrmr,
On PS PCB i mounted 2x2000uf abd another 5600uf immediately after. I drop about 4 volts to 32 volts.
]I have been playing this combo for severa days or more and it is sweet, the best so far. I just played telarc CD with 1812 and the bass is the best I have ever heard on my speakers, There is also nothing wrong in the mids or highs
I checked the voltages under load and the Xfrmr now sags maybe 3%or more but the voltage at the 3875 is very steady withsomething liker 0.3% regulation as seen by DMM.
I will try with more caps before reg.
I am now starting amp #3 and it will be 4 channel 400VA with regs.
I have a large room, relatively inefficient speakers and yet the little 100VA Xfrmr is belting out the music.
On a previous post on this I made an arithmetical error I said 100 VA is good for 2 Channels if regulatedI feel this is still true but I then suggested 600VA for a 6 channel amp. this is wrong one would only need 300VA regulated.
I wish more people using unbuffered GC would check the voltages on the 3875 boards under heavey load. And report it.
I also believe that a good battery supply would outperform a standard GC, and might surpass a rgulated GC. After all it should.
Sheldon
I started with standard NIGC with 1500 pana caps on 3875 PCB
Xfrmr plitron 400VA 2x22 using Brianm's boards.
Sound was clean, amp very quiet no noise no hum BUT after having it runninf for a couple of weeks I was dissatisfied with the
sound particularly with heavy bass but just with demanding passages. Pipe organ and full classical orchestra, 1812 on Telarc
Philips test record by Technics really showed it up.
I checked the voltages dynamically and found the sag of up to 10%. If you do the maths this represents a power drop of about 20%. I do not know how this nonlinearity can not affect the quality of sound. This creates a high degree of compression. There was absolutely no voltage sag from the Xfrmr, it was all past the rectifiers.
So I plugged in a double set of PS boards so I have a full set of diodes in parallel, I then added 10,200 uf at the PS PCB. the voltage sag was much better only 3% now.
The dynamics were better but I was stiill not happy, I added T03 form of LM338 regs on heatsinkfollowing the Application Note.
It helped the regulation but not the overall sound.
So I removed the regs.
Carlos suggested much smaller caps on the 3875 board, on second amp I used 47uf bypassed with 0.1 ufpoly and used a 100VA 2x25 Xfrmr,
On PS PCB i mounted 2x2000uf abd another 5600uf immediately after. I drop about 4 volts to 32 volts.
]I have been playing this combo for severa days or more and it is sweet, the best so far. I just played telarc CD with 1812 and the bass is the best I have ever heard on my speakers, There is also nothing wrong in the mids or highs
I checked the voltages under load and the Xfrmr now sags maybe 3%or more but the voltage at the 3875 is very steady withsomething liker 0.3% regulation as seen by DMM.
I will try with more caps before reg.
I am now starting amp #3 and it will be 4 channel 400VA with regs.
I have a large room, relatively inefficient speakers and yet the little 100VA Xfrmr is belting out the music.
On a previous post on this I made an arithmetical error I said 100 VA is good for 2 Channels if regulatedI feel this is still true but I then suggested 600VA for a 6 channel amp. this is wrong one would only need 300VA regulated.
I wish more people using unbuffered GC would check the voltages on the 3875 boards under heavey load. And report it.
I also believe that a good battery supply would outperform a standard GC, and might surpass a rgulated GC. After all it should.
Sheldon
Yes, Carlos- can't you see the difference between a lot of people being happy with an amp and someone doing an AB test and saying he prefers another amp in his system?
Of course your amp may sound very good. In that case a lot of people might like it. It doesn't mean that it is the best amp around. Everyone who says they like your amp AREN"T saying they think it is the best amp around.
You goaded Peter into making one of your amps to try. It is clear that you feel no need to try a non-inverted gainclone exactly as Peter builds his, with his components (and with battery power?) Why not? I just saw the quote you made that AB is the way to compare. Wy don't YOU compare the way you insisted Peter compare?
I wouldn't ordinarily care- I would just think to myself "Oh, two good amp designs" but you are very aggressive attacking others when they don't agree with you. It is very unpleasant to read while I try to learn more about amp designs. Why don't you try to act like a grown up person with some dignity rather than a
Of course your amp may sound very good. In that case a lot of people might like it. It doesn't mean that it is the best amp around. Everyone who says they like your amp AREN"T saying they think it is the best amp around.
You goaded Peter into making one of your amps to try. It is clear that you feel no need to try a non-inverted gainclone exactly as Peter builds his, with his components (and with battery power?) Why not? I just saw the quote you made that AB is the way to compare. Wy don't YOU compare the way you insisted Peter compare?
I wouldn't ordinarily care- I would just think to myself "Oh, two good amp designs" but you are very aggressive attacking others when they don't agree with you. It is very unpleasant to read while I try to learn more about amp designs. Why don't you try to act like a grown up person with some dignity rather than a
Variac said:
And what have I done too?
Look, Mr., what I see in the pics that Peter posted is a nice box and a complete mess.
THAT'S NOT MY AMP!
As Peter is your god, your inspiration, ask him where's the trafo.
No more hints.
Variac said:
Did I say that?
Please show me where.
I just said, read my lips:
1. regulated is better than unregulated
2. Buffered is better than unbuffered, be it NI or I
3. Inverted + buffer is what sounds best
Who tells you this (me) has made more tests with these chips than you might think of, tried it all, instead of loosing years with inverting (copy of Thorsten's schematic) and testing resistors (and it's polarities
), caps, etc, and then copied again (from the datasheet, or the Gaincard - the same).Horizonts are very limited, and the crowd follows the leader.
Have in mind that while I say regulated is better, I don't say the LM338 is the best.
Have in mind that the OPA627 gives excellent results everywhere I use it (and here too), but sure there is better.
But you have to know how to use it.
Variac said:
AB is the way to compare when there are doubts, or small (not immediately perceptible) differences.
As I said, I TRIED IT ALL, and the buffer thing for me it's old story, you're very delayed in time.
Now, Variac, prove me that Peter made an A-B test. .
Do you know what's an A-B test?
Variac said:Why don't you try to act like a grown up person with some dignity rather than a
Sure.
I know whose people I take serious.
You're not one of them.
And a guy like Peter, who believes in the polarity of resistors, is not one either.
The way he comes to my thread giving his oppinion as an oppinion-maker, as god, as someone who knows it all, THAT puts me off.
It's an ignorant who opens threads to learn (he beats all the records), people help him, and then he sells it.
Pedja must be crazy, in your judgement, but I believe him.
He tried the LM338, he tried discrete reg. PSU, he tried discrete buffers.
Surely he's trying to muck up the sound of his amps.
Deam it, some discussions here remind me a question that I sometimes ask myself: how could Virgin Mary have the child?
pmkap said:
I suggested KZ Muse as they supposedly have very good reputation and in some applications they might work fine. I compared them against Panasonic FC and BG STD in GC PS and myself and two other listeners didn't like them. All the caps tested were not broken in, so comparison was sort of fair. As soon as I got impression about their sound I reported it here.
Now, the caps I used to build Carlos circuit are not the same. Here I used Nichicon 'Fine Gold' which is a different cap and it has different sonic signature than KZ Muse (which has black sleeve and FG has golden color sleeve). Nichicone FG were the best caps I had on hand in those values (33u and 4,700u). I used them before in a DAC circuit and they sounded fine, I only found Cerafines and BG to sound better (and I didn't compare them to Panasonics yet). I bought 'Fine Gold" caps to modify my ML preamp, as I noticed that Empirical Audio is using them in their $2,400 mod on this unit. After comparing them to Panasonics Audio Grade and BGs, I didn't find them superior, but otherwise they shouldn't be worse than majority of other brands (and probably better). Here's what some customers of Michael Percy report regarding Nichicon KZ series:
In order of how good they sound in power supply decoupling application on output stage of power amplifier, with best at the top :
100/100 KZ Nichicon MUSE -- fantastic, very pure
Elna Silmic II RFS-63V101MI6 - almost as good, missing a teeny bit of bass wallop
Elna Silmic ROB63V101MH5 -- not as pure -- lags behind the first two
Elna Tonerex ROS-63V101M Thick and murky-sounding. "
I compared them against the best I know of, so if I found them inferior, it doesn't mean they are bad sounding. It's just means that the signature they present is not what I'm after.
The reason I didn't use the best parts in Carlos circuit is because I wanted to build it as close to what he has, although I suspect that the parts I used and the layout I made is much superior to what he has.
pmkap said:
That's backwards. Intellectual rigor demands an amp built for comparison purposes use the same parts in the initial assesment. If I understand Peter's design philosophy correctly, and it's been the central theme of so many posts, parts selection and mechanical construction are the primary factors determining a Gainclone's sound. To sub parts before listening to the original would have been dishonest and imposed preconceptions. Part cooking is what makes the whole scene so fascinating because, to be blunt, the simple topology of a Gainclone leaves very little room for 'engineering'.
SheldonD said:
If you didn't notice anything wrong with mids and highs after adding 10,200 uf at the PS PCB on non regulated amp, you will not notice anything wrong with highs when using regulators either.
But this is not the case here. With battery supply I don't have any voltage sag, I don't have any power drop. I'm mostly discussing here comparison between my NI, NB, battery operated amp and Carlos circuit.
I can't understand how Carlos can claim that his amp should sound better, if he doesn't have clue how mine sounds. I tried to reason in very respectful way, and I already commented few times that his amp is very good in certain ways, but it's still not there when it comes to highest fidelity. I'm not saying I'm any near there, but at least I understand that. Carlos just can't accept that possibility.
He's been attacking my implementation on the grounds of improper layout, or whatever else he could think of, but to me those are just cheap excuses. This is a prfectly well made circuit. In many ways the implementation is much superior to what Carlos has in his box.
t. said:
I'm still using that DAC. It sounds very coherent.
The bass is boomy when you compare it to the battery power. Otherwise it may pass as good.
I will still do more tests between the AC powered amp and Carlos high end circuit, as what I did so far wasn't very accurate, didn't last long enough and this discussion definitely requires better evidence (on my side).
Peter Daniel said:
Why didn't you post the rest of my post, I thought it would have been obvious that the other times I tried these chips in cdps that the iplementation was crap being the reason why the results was poor😉
Peter you are being intellectually dishonest. You quote my post and then comment out of context. I said I was unhappy with the sound before the regs were installed as in my present setup.
Secondly when comparing to Carlos amp the very least you should do is to use a power supply as in the Brian GC package.
The whole point of regulation is lost if you compare it to a battery.
Secondly when comparing to Carlos amp the very least you should do is to use a power supply as in the Brian GC package.
The whole point of regulation is lost if you compare it to a battery.
Peter, how about measuring the voltages on your NIGC with the 4.7uf on the PS PCB under heavy musical passages with lots of bass.
Do not discuss the battery supply,in this context
Do not discuss the battery supply,in this context
SheldonD said:
No, I'm not dishonest. You said:
>>So I plugged in a double set of PS boards so I have a full set of diodes in parallel, I then added 10,200 uf at the PS PCB. the voltage sag was much better only 3% now.
The dynamics were better but I was stiill not happy, <<
You said dynamics were better, and you didn't mention about any deterioration in highs and mids. Whoever tried adding bigger caps on rectifer's board, knows that the sound quality drops substantially. If you didn't comment on that, it means that your main concern are dynamics without further thought of how adding bigger caps affects the rest of the sound spectrum. So you either didn't noticed it, your system didn't show this, or it simply doesn't matter to you.
I also compared Carlos circuit to Brians premium kit and I posted the impression of a friend of mine. I didn't post my impressions, as the session was not properly conducted to present any reliable assesment. I still plan to do that in more detail.
Post#666: http://www.diyaudio.com/forums/showthread.php?s=&threadid=35482&perpage=25&pagenumber=27
As to the battery comparison, I suspect Carlos didn't even noticed that his point is lost, as he still claims that his amp SHOULD be better than my battery powered Patek.
t. said:
So what's the secret to proper implementation? I thought going by Carlos schematic was that.
SheldonD said:
I was never interested in that. In my main system I'm using GC on top only and both 13" and 15" subwoofers are powered by the monoblocks with 24 mosfets in the output stage.
In my other full range system, when GC are used full range, I'm using batteries.
Third system is in a rather small room, and with 6" AT drivers, I never observed any bass problems. I tried both amps today in that system as well (both AC powered), and the bass was pretty much the same with both of them. For some reason, in this system, inverted amp always sounds better, be it buffered or not buffered. But those speakers are also rather bright (Triangle tweeter) and I'm using oversampling DAC, so maybe that's the reason. But again, NI, not buffered amp sounded more detailed and clean, Carlos amp was more pleasing, but I don't know if I could listen to it permanently.
Peter: this thread is about buffered and regulated GC.
It is not about your other nonGC systems. And if we consider regulation then Battery power is also not part of it. As regulation
is pointless on a battery powered GC.
I can only discuss with regulation of the GC. I have as yet no experience with buffering.
I have arrived at the conclusion that the standard GC needs
more than the 4.7uf on the power supply and the 1000 or 1500 on the chip.
With this configuration the voltage sag is considerable and has considerable impact on the quality of sound.
I challenge people to measure the voltage on heavy passages of music to the chips if GC unregulated.
If you only listen to quiet music and have very efficient speakers
you won't notice it. The stock GC sounds very good under these conditions.
This is not a shooting match it is to present experiences to share with others.
It is not about your other nonGC systems. And if we consider regulation then Battery power is also not part of it. As regulation
is pointless on a battery powered GC.
I can only discuss with regulation of the GC. I have as yet no experience with buffering.
I have arrived at the conclusion that the standard GC needs
more than the 4.7uf on the power supply and the 1000 or 1500 on the chip.
With this configuration the voltage sag is considerable and has considerable impact on the quality of sound.
I challenge people to measure the voltage on heavy passages of music to the chips if GC unregulated.
If you only listen to quiet music and have very efficient speakers
you won't notice it. The stock GC sounds very good under these conditions.
This is not a shooting match it is to present experiences to share with others.
- Status
- Not open for further replies.