It was probably about 30 yrs ago when Mr. DJK suggested to me that I put bypass caps in my Ampzilla. I thought he was absolutely nuts at the time. I tried it per his instructions and the amplifier changed completely as a result. It was bass heavy as it came from the factory and this modification resulted in a smoother fuller sound. It awakened the midrange and high frequency response of the amplifier.
Still being skeptical I tried it on a pair of Tigersarous mono block 250 watt amplifiers. I put bypasses on one channel and left the other channel stock. I ran a mono passage thru them and it was night and day difference.
Still being a skeptic I put bypass caps on a Dunlap Clarke amplifier that suffered by some idiot installing caps that were double the original size which made the amplifier sound very muddy in the low end. Installing bypass caps brought back the midrange and high frequency detail that had been lost.
Has anyone read the numerous posts concering the Marantz SE63 CD player? The "Brains" seem to think that changing capacitor sizes and types makes big differences there also. I have always maintained the attitude that if I cannot measure a difference it doesn't exhist.
The bypass caps are proof that sonic difference can result by changing, adding bypass caps. Try it.
Still being skeptical I tried it on a pair of Tigersarous mono block 250 watt amplifiers. I put bypasses on one channel and left the other channel stock. I ran a mono passage thru them and it was night and day difference.
Still being a skeptic I put bypass caps on a Dunlap Clarke amplifier that suffered by some idiot installing caps that were double the original size which made the amplifier sound very muddy in the low end. Installing bypass caps brought back the midrange and high frequency detail that had been lost.
Has anyone read the numerous posts concering the Marantz SE63 CD player? The "Brains" seem to think that changing capacitor sizes and types makes big differences there also. I have always maintained the attitude that if I cannot measure a difference it doesn't exhist.
The bypass caps are proof that sonic difference can result by changing, adding bypass caps. Try it.
burnedfingers said:
IMHO it's much better if you try different arrangements until you find a combo that works best for your application.
Wow, you take a lot to move from the skeptic seat!...
In the Chip Amplifiers there was a similar problem with Gainclones too, that seemed to sound muddy in mids and highs when you increased PSU caps beyond 1000uF, even if lows improved. Adding a snubber allowed using larger caps and brought back the missing mid-highs.
Capacitor types seem to sound different, or make things sound different, even in PSU.
Why don't you look at it in a different way: perhaps the measuring tools are the wrong ones, particularly because they are mostly static. Perhaps other measuring standards will have a better correlation with our natural tool: our ears. I prefer to think that, just to explain why I hear what I hear.
As you seem to have proved: sometimes changes cannot be measured.
Carlos
carlmart said:
*Particularly* in the PSU. What you hear on the output of your amplifier is (predominantly?) the sound of your PSU capacitors. The output devices demand current, the PSU responds with its non-linear PSU impedance, PSU voltage drops as a result of this impedance, the PSU caps kick in to stabilize PSU voltage adding, in the process, their characteristic audible stamp to those PSU voltage variations, which your speaker then reproduces.
serengetiplains said:
Can't say I do not agree with you.
I said what I said perhaps because I expect certain minimum demands to be fulfilled. Like providing a stiff supply that will not vary as much due to the reasons you describe. E.g.: if the transformer is large enough and/or independent will not really drop its voltage to need stabilizing. So whose is the fault then: the transformers or the caps?
Maybe that's why high current regulated outputs tend to sound better than unregulated. But many will disagree.
So IMHO a bypass will probably be less an improvement in a properly designed supply than on the others.
But once again: I agree with your concepts.
carlmart said:
I suspect that "many" have heard only NFB stabilized power supply (like 78XX, eventually in a "beefed" version for power amplifier). In this case, for a lot of reasons, this kind of regulators can only worsening the sound of system because they are designed for facing with essentially static loads as usually a power amplifier is not...
In my opinion PSU regulators for audio must be designed not for "stabilizing" output but just for *filter it*. And avoid any ringing problems that usually arises with normal stabilizers (which, by their own, suffer the common nuisance of NFB systems when faced with "transient load"; indeed a normal stabilizer in audio is not more than an amplifier which "drive" other amplifier and this must be taken in account to avoid unpleasant surprises, as precisely is the "poor dynamics" complaint usually moved to PSU regulators).
Hi
Piercarlo
I can't say that I agree with you on this at all. I would like to see the calculations that prove that this minimal capacitor value is correct not just some assumptions that it is. Lets get a formula going here to figure the cap size.
QUOTE]Capacitor types seem to sound different, or make things sound different, even in PSU. [/QUOTE] I don't believe in the magic of one brand of cap over another. I have listened to very expensive coupling caps in tube amplifiers and laughed at the people purchasing them after they failed numerous A/B tests. I do believe in proper cap sizing and proper bypassing
Yes, I am skeptical and refuse to blindly jump on something and say it sounds better. The mind is a funny thing because you can convince yourself there is going to be a difference and wow there is. I generally use people to listen to a blind A/B test because it is a good indicator that there has or hasn't been a change. Maybe the ear is a good instrument but unfortunately the mind isn't.
the incorrect values of bypass caps can cause the system to ring -- that is -- the indutive nature of the power supply impedance resonates with the bypass capacitor under certain conditions -- josephK demonstrated this with his analysis -- and as SY points out the why-fores are explained in a paper in Bob Pease's "Troubleshooting Analog Circuits."
You can't leave out that your solid-state amplifier is a crystal radio in disguise with every base emitter junction a diode which just wants to play music -- that's why it's important to bypass at the opamp V+/- pins. reduce the RFI and EMI which gets onto the supply at the chip and you have fewer problems with noise and offset.
one of the reasons that the LM4702 design in the Nat Semi ap-note has such low THD (certainly lower than I could hope to measure) is that they spent hours and hours figuring out the correct chip bypass caps, orientation of the transformer, choice of two transformers, etc.
if you have a good scope with a great horizontal amplifier (like a 7A22N) you can make a "sniffer" to look for energy that your power supply is unwittingly radiating (well, a power supply is incapable of wit, but perhaps the designer is).
You can't leave out that your solid-state amplifier is a crystal radio in disguise with every base emitter junction a diode which just wants to play music -- that's why it's important to bypass at the opamp V+/- pins. reduce the RFI and EMI which gets onto the supply at the chip and you have fewer problems with noise and offset.
one of the reasons that the LM4702 design in the Nat Semi ap-note has such low THD (certainly lower than I could hope to measure) is that they spent hours and hours figuring out the correct chip bypass caps, orientation of the transformer, choice of two transformers, etc.
if you have a good scope with a great horizontal amplifier (like a 7A22N) you can make a "sniffer" to look for energy that your power supply is unwittingly radiating (well, a power supply is incapable of wit, but perhaps the designer is).
burnedfingers said:
No need to become a hero, really.
What we did with a friend of mine might be called empiric, even if not scientific.
First of all some background:
1) I have been a film location sound recordist for a long time, since the Nagra times, which probably gave me some idea on how things are "transformed" when reproduced in the following chains.
2) My friend makes high quality speakers in Argentina.
What I mean is that it's not easy for us to be fooled by imagination on how things sound. We do not "want" or "expect" things to sound in any way. They sound and that's it: you just describe what you hear.
The speakers are good enough to let you listen to very, very small changes you make.
What we did was test some gainclones, same source and speakers, only changing the supply caps. On that test, when using the 1000uF capacitors the highs and mids were very good, and low frequencies were less controlled. When going to 2200uF or higher, the bass got tighter, but the mids and highs were less clear. Do you think you can find a formula to explain this?
We were not measuring that, but we did repeat it several times to check what happened. You apparently had to choose what type of sound you wanted.
We are not talking 50W or 60W here, which would certainly demand tenths of thousand uF to play properly. Probably 30W or so, which would be plenty in most circumstances.
In the Chip Amplifiers forum this matter found a third option: using a snubber. Many people tried it and reported back on their findings, so anyone can go there and read about them.
Whether you believe them or not is your problem.
Hi,
if the ONLY way to get "good" sound out of a chipamp is to build it with just +-1mF of smoothing on the supplies then I have no intention of wasting my time experimenting with chipamp topology.
I know that one can never get a wideband power amplifier to work succesfully with that amount of ripple on the supply rails, no matter what claims the manufacturer makes for PSRR.
if the ONLY way to get "good" sound out of a chipamp is to build it with just +-1mF of smoothing on the supplies then I have no intention of wasting my time experimenting with chipamp topology.
I know that one can never get a wideband power amplifier to work succesfully with that amount of ripple on the supply rails, no matter what claims the manufacturer makes for PSRR.
AndrewT said:
Well, our tests served to see what we could get from this chip amps.
If you followed these chips application and tests since the beginning, trying to clone the Gaincard amplifier, they seemed to bring quite a lot for very little money.
If you increased the capacitance you got a different sound, but that is not what National shows on the datasheet. Many people built it with more capacitance, of course.
But several people went on to test regulators and passive solutions, like the snubber. Which seems to "restablish order" onto things. Now you could use greater capacitances and still get good mid & highs.
AndrewT said:
Probably 30/40W. Like the original Gaincard did.
Don't look at me for answers over this matter. Peter Daniel also went on to release an amplifier line based on the gainclone, using also small capacitors, with around that power. With great reviews too.
My estimate on our test is the peaks were really 40W. The quality did not change. As I said, the bass was only muddier, not as controlled as when increasing the capacitance. Highs and mids were very good.
carlmart said:
I don't want explain nothing, just add an anecdote. Some time ago (six or seven month) I repaired and test for a friend an old Grundig 1700 integrated amplifier. Its supply, for a 20 Watt per channel, was among the shrinkest I ever see: just 3300+3300 uF for the overall amplifier.
Hard to believe but true, this was one of the best sounding amplifier that come in my hands. Limited output power was its only real limit. Circuit were just the essential for working (tone controls were implement in the feedback network of the main amplifier which, with a 4558 deputed to equalize the phono MM input, was the only active stages in the equipment).
Why a "amplifier for flies" get a good sound when other "monster" (surely well thinked and designed) get, on the field, more disappointing performance?
Hi
Piercarlo
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.