Continuing the off topic discussion from here: http://www.diyaudio.com/forums/chip-amps/79303-chip-amp-photo-gallery-270.html#post3818236
Well some still listens to mono recordings
Its maybe you that lives in the past when people thought that bigger was always best.......
Last edited:
High capacitance is fine, but the average 10-15,000uf cap is garbage. Spend the $ to make it a very good, very fast capacitor or use multiple smaller values. This seems to escape everyone, always trying to find a compromise between enough capacitance for bass, and yet not destroy the highs.
Peter Daniels has used 100/1000uf and 1500/50uf for Chip/PSU. Blackgate N/Blackgate STD, Panasonic FC/Panasonic FC. In either situation the caps are small enough and responsive enough to provide good quality music, bass maybe respective of speaker load however. But a lot of people on here are aware that 10-15,000uf has some serious merit, per channel.
What would I try? Hm, well I'd try a few things perhaps. But I think the approach PD took was legit. However I'd probably use Elna SIlmic II's or Nichicon FG's to play around with these days.
I wouldn't be doing any mythical bypass caps of small value (film caps or whatever) that are just ringers.
Peter Daniels has used 100/1000uf and 1500/50uf for Chip/PSU. Blackgate N/Blackgate STD, Panasonic FC/Panasonic FC. In either situation the caps are small enough and responsive enough to provide good quality music, bass maybe respective of speaker load however. But a lot of people on here are aware that 10-15,000uf has some serious merit, per channel.
What would I try? Hm, well I'd try a few things perhaps. But I think the approach PD took was legit. However I'd probably use Elna SIlmic II's or Nichicon FG's to play around with these days.
I wouldn't be doing any mythical bypass caps of small value (film caps or whatever) that are just ringers.
Ideally there should be infinite capacity for the amp, so it gets the same voltage always. If it is not infinite, as the 50/60Hz goes, 90% of the time the chip gets the juice only from the capacitor and not from the transformer. And I guess this explains why bigger is better. Don't really understand why there is an argument at all.
Yes, ripple current is up to us to decide. Yes, it is 100Hz after a rectifier. The argue is following that the signal is traveling through the power supply capacitance. And high capacitance damps the high side of a freq response. Connect many smaller caps in parallel was the idea. Anyway. It is not difficult to measure the frequency response at the certain points and see the dB difference. I was using 10mF power supply caps with no problems.
Some time ago, I built a test amp with relay switchable capacitors so I could change the value from the listening position after reading a load of rubbish here.
The small values just gave and apparent loss of bass with no other noticeable affects.
I wouldn't personally bother with many small capacitors in parallel as I was measuring about 0.01Ohm ESR @ 200kHz from my 10,000uF caps.
Frank
The small values just gave and apparent loss of bass with no other noticeable affects.
I wouldn't personally bother with many small capacitors in parallel as I was measuring about 0.01Ohm ESR @ 200kHz from my 10,000uF caps.
Frank
Last edited:
Before jumping to conclusions please read the following posts:
http://www.diyaudio.com/forums/chip-amps/43423-high-cap-unregulated-psu-chipamps.html
http://www.diyaudio.com/forums/chip-amps/49813-my-first-try-lm3886-layout-any-comments-suggestions-16.html#post581230
http://www.diyaudio.com/forums/digital-source/48990-blackgate-caps-out-production-so-4.html#post574431
http://www.diyaudio.com/forums/chip-amps/43423-high-cap-unregulated-psu-chipamps.html
http://www.diyaudio.com/forums/chip-amps/49813-my-first-try-lm3886-layout-any-comments-suggestions-16.html#post581230
http://www.diyaudio.com/forums/digital-source/48990-blackgate-caps-out-production-so-4.html#post574431
Peter Daniel's amps are certainly works of amazing craft. When it comes to the power supply's caps, all he's offering is a subjective opinion which cannot be explained.
Interestingly, the discrete amps' crowd never took interest in those undersized PS.
Gootee has a very nice excel sheet to get an idea of required capacitance btw: http://www.diyaudio.com/forums/chip-amps/243940-chip-amp-way.html#post3662817
Interestingly, the discrete amps' crowd never took interest in those undersized PS.
Gootee has a very nice excel sheet to get an idea of required capacitance btw: http://www.diyaudio.com/forums/chip-amps/243940-chip-amp-way.html#post3662817
Well, I think that most of the HIFI community is based on subjective opinions .
My father is specialist in reproduction of pictures and he says that it will never be possible to reproduce the colour of a flower.
The same is happening here.
I build my first amp according to P. Daniel's instructions and according to me it sounds very well.
I'm using quiet easy to drive speakers and that maybe is the reason I'm very pleased with it.
Anyway, the subjective experience of quality keeps the HIFI discussions interesting as this is impossible to reject.
Last edited:
Carlos probably did hear a change when he inserted the snubber on the OUTPUT of the PSU.
But that does not make the snubber right.
If the fast switching of the transformer was causing an audible problem, then a snubber at the transformer was the right place for it.
If the local decoupling in the amplifier was lacking, then a snubber at the amplifier would be the correct place for it.
Carlos' design is proving that unsnubbered was possibly wrong, but in the dozen or so years since then, many have come to realise what the snubber is and what it does and why it should be located in the position that allows it to do it's job.
Separating the snubber from the "problem" with long lengths of connecting cable is not the right way.
Which means that the amp can be worked on more quickly after power-off. Woo!
What do you expect, people around here will argue about anything.
Seriously though, if you look a little closer you'll find that the strongest protagonists of undersized power supply caps also promote ideas like:
a) Bronze heatsinks "sound better" than aluminium ones.
b) Good bass requires lots of current. So if you want your chipamp to have good bass response, you need to make the input impedance as low as possible to force it to draw as much current as possible.
IIRC, the rot started some years ago when a reviewer questioned the undersized caps used in an overpriced, under-engineered commercial amp. Needless to say the manufacturer denied that cost cutting had anything to do with it, but then went on to invent some nonsense about "signal storage" to justify their claim that supply capacitors should be as small as possible. Unfortunately that mantra's been repeated over and over ever since by people who don't actually know anything about electronics.
You have to remember that some people who don't know any better actually believe the utter bilge routinely published in magazines like 6Moons. After absorbing enough of that nonsense some of them consider themselves to be experts, and set about sharing their "wisdom" with anyone who cares to listen. Unfortunately a lot of innocent noobs get infected that way, and then go on to spread the nonsense further.
I remember a test that was published where they asked 10 best trombone players of the world to play on two different trombones, one made of steel and one made of copper.
At first all the trombone players sworn that a copper trombone sounds superior to the steel one.
They let them play with and without blind fold on both of them.
Without blindfold they almost all thought the coper one sounded better, with blindfold it was fifty fifty.
This was an easy test to rule out the subjective opinion because it only had 2 parameters.
To my opinion in the case of using high or low capacitance with CG this test is equally easy to do and I can not imagine that this is not already done before.
At first all the trombone players sworn that a copper trombone sounds superior to the steel one.
They let them play with and without blind fold on both of them.
Without blindfold they almost all thought the coper one sounded better, with blindfold it was fifty fifty.
This was an easy test to rule out the subjective opinion because it only had 2 parameters.
To my opinion in the case of using high or low capacitance with CG this test is equally easy to do and I can not imagine that this is not already done before.
You don't see the value in quicker response to changes in demand for power? K... These aren't constant current supplies.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.