Film Caps for Tube Amp Power Supplies

[Mike Sulzer]

I’m sorry, I call bullshit that anyone can hear the difference between film and electrolic caps in the power supply, or any significant difference between 100/120Hz ripple modulation either. I’m not saying the difference isn’t there, but until I see some convincing data, from double blind testing, i will remain highly skeptical that either contribute anything significant to the overal sound

Sorry, cannot help you out with double blind results with guitar amps. But here are some short files, sine wave (1000Hz) with sine wave amplitude modulation, 120Hz mod frequency except 100Hz in one case. The last number is the % modulation. 6% is a large effect; 3% is easily audible; 1% is harder to hear except under good conditions. The basic effect is dissonance. You can hear some difference in the dissonance between 120Hz and 100Hz.


tmod1000_120_6.mp3
tmod1000_120_3.mp3
tmod1000_120_1.mp3
tmod1000_120_0.mp3
tmod1000_100_6.mp3

[SoulFetish]

All I'm saying is that it's real and audible. A lack of documentation for the phenomenon in an arcane field like "overdriven guitar amps" hardly seems like proof of absence for the affect.

Don't get me wrong, Chuck. I WANT to believe. I used all film caps in my amp, and I love it. 18 Watts and just under the size of a Volkswagen Bus.
Now, off the record, I'll tell you a quick story. The first amp I built used sprague atoms in the power supply. I don't remember what inspired me to do this, but I decided to change out the electrolytics for all film caps. At the same time I swapped out my 1N4007 diodes for UF recovery diodes. The only change to component values was that the reservoir cap changed from 80µF to 60µF. I did the math and I figured I could save real estate, and still be right around where I wanted to be for filtering. The rest of the changes were component swaps.
Between you and me, after that, the difference in perceived bass response was so drastic, that I had to modify the design to compensate. But that's purely anecdotal and untested, and I don't implicitly trust my own experiences as canon. I suppose I can't completely dismiss them either.
Mike, I haven't gotten around to listening to your files yet. But, I look forward to being convinced.

[Chuck H]

I had a similar experience testing coupling cap dielectric types. Though I used modern caps that typically have very close specs I didn't test values so I can't rely on perceived results. Same with the "outer foil" arrangement on coupling caps. I didn't hear a difference at all in that case. But it was a matter of swapping them in the same amp and listening. No hard bench gear data before and after. But I do know I've heard a plain difference in decoupling caps in the power supply. Mostly WRT value. But I've read enough about impedance (ESR?) also having an affect to believe it. Though I won't qualify believing as fact.

Once upon a time SGM mentioned that 'Bias supply cap value can make or break bass response.' (paraphrasing) He took a beating for it too. But when you consider the time constant of the circuit instigates significant affects on grid loading, the decoupling through those caps could well have an audible affect (effect?) in some designs. He probably actually experienced what he said he heard, but in his usual way was proclaiming it as a blanket phenomenon on ALL amps without qualifying any aspect of the circuits in question.

[Helmholtz]

Sorry, cannot help you out with double blind results with guitar amps. But here are some short files, sine wave (1000Hz) with sine wave amplitude modulation, 120Hz mod frequency except 100Hz in one case. The last number is the % modulation. 6% is a large effect; 3% is easily audible; 1% is harder to hear except under good conditions. The basic effect is dissonance. You can hear some difference in the dissonance between 120Hz and 100Hz.


[ATTACH]50915[/ATTACH]
[ATTACH]50916[/ATTACH]
[ATTACH]50917[/ATTACH]
[ATTACH]50918[/ATTACH]
[ATTACH]50919[/ATTACH]

Thanks, very interesting.
With headphones I can hear an increasing roughness with higher modulation depth. There is definitely a difference between 100Hz and 120Hz.
Maybe that's why Marshalls sound better over here in Europe?

[eschertron]

Thanks, very interesting.
With headphones I can hear an increasing roughness with higher modulation depth. There is definitely a difference between 100Hz and 120Hz.
Maybe that's why Marshalls sound better over here in Europe?

Interesting. The AM sidebands for 100Hz are between G and G#, falls in nicely with the key of E (rock or blues). The 120Hz... not so much.

As good a reason as I've heard today to move to Ireland

[Chuck H]

Maybe that's why Marshalls sound better over here in Europe?

Cute...

I can hear it too. I even had my wife click the starts while turned around and I could still hear it. I won't say it was unpleasant though. A little dissonance can add spice. A lot and all bets are off. Bury it under typical guitar amp distortion levels (even clean) and I don't think I would hear it. Let alone be noticed by anyone in a mix.

[SoulFetish]

Dissonance makes music interesting.

[Leo_Gnardo]

Dissonance makes music interesting.

You bet! Handy for jazz. Hm... where did I read that?

[Mike Sulzer]

When might it matter in a live performance? Lead on plain strings up the neck. Not so many harmonics fit in the system bandwidth, and so intermod becomes more audible and important to making an interesting sound. For example, Jimmy Page leads, LZ live recording, blues material.

Power chords: too much else going on.

Wound strings, low notes: A lot happening here already because many harmonic fit in the system bandwidth. Also, there are two types of harmonics. First, those from the string which are not exact multiples of the fundamental, an increasing effect as the harmonic number increases. Second, those created by the distortion of the fundamental and lower harmonics, which are exact. The interaction of the two helps create a dissonance that can be interesting.