Whats wrong with large filter caps for Gainclone?

[kropf]

Originally posted by Pedja
[B]
I ran shortly that 2200uF/8 Ohms load case and the spectral content of the harmonics relates very well to your measurements, though cap’s ESR and parasitic series resistance could change this content somewhat. But what puzzles me is that you get 5V p-p swing around the 22.7V. I see swing between some 23V (or 22.7V you mentioned) and 34V (it can’t reach 35V anymore even at the peak) so it seems to me I am missing something about your circuit.

The PSU was actually charging and discharging between 20.2 and 25.2 V. I think the explanation is in the transformers. I am using 120VA EI transformers. The resistance of the secondaries must be limiting the charging rate and the load regulation may be rather poor for these transformers, since they were running very near the VA rating . The trace on the scope looked more like a triangle than a sawtooth. A toroidal transformer could probably charge at a higher rate. I also have a 1 meter run of 18 gauge wire from the PSU to the amplifier. It shouldn't be too hard for me to calculate the resistance of the wire and determine if there was a significant voltage drop across it.

Jeremy

[tbla]

please use "monstersize" trafo when performing those tests - so you are absolutely certan that there is no limitations.....ei regulates better than torroid btw......but anyway, very interesting test - thanks.

[JoeBob]

Wouldn't a film cap in parallel with the main psu capacitors shunt alot of the high-frequency content to ground?

I know alot of people don't like bypass caps, but here sounds like somewhere they might come in handy.

[theChris]

such is reccomend on the datasheet, but not located near the main caps but rather near the chip itself.

[janneman]

Originally posted by JoeBob
Wouldn't a film cap in parallel with the main psu capacitors shunt alot of the high-frequency content to ground?

I know alot of people don't like bypass caps, but here sounds like somewhere they might come in handy.

Not really. The cap is the CAUSE of the hf content. If you read the earlier posts, you will see that increasing the cap generally increases the hf content, because the current pulse gets narrower/sharper and that means a higher hf spectrum.

It would be different if the hf signal was fed to the cap via an impedance, then you would have the normal low pass filtering action.


Jan Didden

[Joe Rasmussen]

Originally posted by JoeBob
Wouldn't a film cap in parallel with the main psu capacitors shunt alot of the high-frequency content to ground?

I know alot of people don't like bypass caps, but here sounds like somewhere they might come in handy.

I agree with Janneman, it might actually make it worse. A classic Dr Jeckyll vs Mr Hyde. It may increase the peak current at HF and be worse, on the other hand it may present a Lower Z at those freqs to the subsequent load. Which is preferred? It may be a 'horses for courses' situ. Almost near unpredictable. This is why you cannot really design overall 'by the book' and why a combination of gut feeling and thorough experimentation cannot be substituted. What comes to mind is a time (and I suspect that it still does happen) when there were those who said that all amps sounded the same. Yet Matti Otala (the TID guy) said there are thousands of reasons why they sound different. I have never heard two amps sounding the same. That's the honest truth. So power supplies make a difference, that we can all agree? So why be surprised if some use it to shape the kind of sound they like? Provided the end result works, right?

Joe R.

[millwood]

Originally posted by Joe Rasmussen
So power supplies make a difference, that we can all agree? So why be surprised if some use it to shape the kind of sound they like?
Joe R.

there is no question that power supplies make a difference. But how many of us can hear it at -50db over 1Khz, on an amp with over 40-50db psrr?

I cannot.

Materiality is the key, everyone.

[carlosfm]

Guys, I don't want to start a discussion here, we're just debating this, and this is an interesting one.
But it seams to me that what is good for digital is not good for analog.
Would you put a single electrolythic cap on the PSU pins of a DAC?
I would bypass it with a small ceramic/polyester cap near the chip's pins.
Is this bad for a GC?
Wouldn't a PSU be better if it has low impedance at high frequencies?

[Pedja]

Beside the content and level of the HF noise generated by the larger cap value, the problem with larger_main+small_bypass_cap is the way in which these two caps are integrated. The larger cap has inductive rise at some point (which is for us too low). So we put one small cap that can keep the impedance low further above. So we’ll indeed have further above that our wanted low impedance, but the meeting point of two caps might be one new resonant frequency.

As Carlos pointed out, in the digital (and high speed opamp) circuits it is usually beneficial to put small bypass close to the chip, because here you have more HF problems than in classic slow analog circuits. Bypass in the analog circuits can make more problems than it solves.

Pedja

[carlosfm]

Originally posted by Pedja

As Carlos pointed out, in the digital (and high speed opamp) circuits it is usually beneficial to put small bypass close to the chip, because here you have more HF problems than in classic slow analog circuits. Bypass in the analog circuits can make more problems than it solves.

Pedja

OK for power op-amps.
Modern signal op-amps are very fast, and do benefit from bypassing.
Like the LM6171.
Excellent, BTW.
And my beloved OPA627.