audio-talk

This is not a question about quality of film vs electrolytic per se, I understand that film caps etc do have a "sound" directly in the signal path.

What I don't understand is the compromises made on capacity only to have an undersized film cap in a location that needs capacity, that may have questionable results with regards to sound quality in certain locations.

My latest amplifiers needed around 2200uF spread throughout each amp to stabilize operating points etc, to make it function (perform) well, in my opinion.
These are <3 watt amps. Granted some of my caps are in series and one 1500uF is a low voltage cap. Nonetheless, a trivial task to accomplish with modern electrolytics.

I have noticed a similar trend in other amplifiers I have built in the past as well, film caps would be nice but where would I fit it all?

My hat would be off to anyone who did it right, don't get me wrong.

For myself though, I notice a bigger improvement in regards to sound quality by having the Joules as opposed to an undersized film cap.

Is this a valve amp or a solid-state amp ?
The electolytics use organic ( paste ) dielectrics that have a very high dielectric constant in comparison to film materials like polypropelene . These tend to have more distortion than the low-dielectric constant materials ... Black Gates being an obvious exception to this rule .

Sorry these were valve amps I was referring to. Although I did build one SS amp and found the same, the power supply needs too much capacitance to make film caps practical.

What about bypassing a big electrolytic with a small film cap ?

I suppose that is the typical thing that is attempted. I myself never really bought into it much.
It's one of those grey areas. Having caps in parallel would lower ESR which is a good thing, but I feel the bypass cap is too small to make a significant difference overall to compensate for a lack of capacitance. "In addition to" might be alright if it doesn't do something funky.

Take for example the cathode bypass capacitor on my 6J5 input tube, it only needs a few tens or hundreds of mF to bring the low frequency up, but 1500mF really stabilizes the bias, helping to keep the load-line flat as opposed to an elliptical orbit.
Any residual induced warble from the power stage only swings the bias more negative. (as opposed to swinging towards the grid.)

I'm a big fan of stout power supplies. I feel it is an elegant, albeit dumbed-down way to regulate voltage.

Gargoyle wrote:
I'm a big fan of stout power supplies.

You'll get no argument from me on that one

SteveTheShadow wrote:

Gargoyle wrote: I'm a big fan of stout power supplies.

You'll get no argument from me on that one

There's "stout" and then there's hernia-inducing contributors to global warming...

FWIW, I've still got some SERIOUS low ESR RIFA/BHC caps, NOS, as mentioned in http://www.audio-talk.co.uk/phpBB2/viewtopic.php?t=5702 as used by various members here and elsewhere...

(*) also http://www.audio-talk.co.uk/phpBB2/viewtopic.php?t=4708 although a lot of these were sold at Audio Jumble last year - still some left.

pre65 wrote:What about bypassing a big electrolytic with a small film cap ?

Aaarg !
Always sounds crap . I tried it last on the cathode by-pass of my shunt-reg power supply and regretted it immediately. Something to do with the mismatch of ESR causing a small phase shift in the HF area , under 'load' ie. music.

I guess the 'very big cap' idea is a way of getting the low-end frequency response down, and as such a pretty good idea that ought to help the bass and midrange tone. For those sort of values electrolytics are the way to go, I'm sure.
Personally I went away from passive supplies about 8 years ago with a series-reg supply ; then about 3 years ago I moved to a shunt-reg supply. This has the advantage of having no significant change in resonse down to <5Hz, plus it still has a relatively clean return signal path - through the shunt valve and its cathode by-pass cap .

By the way, this is for SE amps ....

In my limited experience paralleling up caps is good practice in SS amps and the results seem to be better with say a bunch of 2200uF caps instead of a single 10000uF cap. But in a valve amp size is an issue and the benefits don't always seem to materialize IME and as Mark says some times bypassing with a small value cap can be a backward step IE like bypassing a BlackGate...having said that I have had some good results doing this on some types of lytic caps with basic CLC supplies but a bigger bang for the money IMO is to regulate the supply, this seems to have a much greater affect on sound quality than bypassing or spending on specialist caps...just my 2p...

Nickds1, I will definite keep those caps in mind for another project. I don't have the space this time around though.

Those are some interesting thoughts on regulating the supply.
I should try and do some measurements to see how stable my amplifiers voltages actually are, maybe get an idea if I should attempt some kind of regulation.
The seem pretty stable in spice.(With my limited experience, only checked with sines.)

Realistically I don't have the room for tube regulators in these chassis, but perhaps a SS unit would fit.

Interesting that you noticed a sound quality change by bypassing the cathode resistor on the shunt regulator valve. I have read similar issue about the phase shift.

Neal wrote: as Mark says some times bypassing with a small value cap can be a backward step IE like bypassing a BlackGate...having said that I have had some good results doing this on some types of lytic caps with basic CLC supplies but a bigger bang for the money IMO is to regulate the supply, this seems to have a much greater affect on sound quality than bypassing or spending on specialist caps...just my 2p...

I suppose I could try modelling it, too see if I can see any phase issues by pacing some little films on my CLC.
Recently I was able to sort out a phase issue with the plate choke I was using in spice.
I suppose the trick will be using accurate parameters for the capacitors and knowing where/what to look for.