Caps in signal path, influence on sound

[Boba7]

Hi guys!

So today I upgraded an old Korg DT10 tuner I got for cheap a few weeks ago. I changed the footswitch for a better one (hate those spring actuator footswitches) and made a couple mods to improve the buffer sound. I was very surprised by the difference in sound before and after.

Schematic is below

So I changed all the caps in the signal path to mlcc or film 1uf instead of cheap 10uf electros, and before that I had changed R19 to 3k3, which seems to be a more conventional value (the one I use in my buffers)
I did notice a little change after the resistor swap, but I swear that changing the caps really affected the sound. To me with the electros in the signal path, there was almost a sort of high pass filter going on.

What do you guys think? Are my ears playing tricks on me?

[chromesphere]

Did you test their value after you removed them?  Also could be an ESR thing, I believe the resistance of electros goes up over time (might explain pass filter effect that you were describing). Prehaps the Electros have dried out?

[Boba7]

Thanks, that's interesting. I dont know much about ESR, I should read about it. What's an acceptable ESR for electro caps?

I measured them and they are still 10uf or so. 

[ahiddentableau]

I'm not generally a huge believer in this kind of thing apart from the kind of thing just mentioned (ESR, different levels of parasitics changing impedance, etc.)  But there's an interesting chart in one of Merlin Blencowe's books that gave me pause (I looked it up, it's on page 79 in Designing High-Fidelity Valve Preamps.).  Merlin measured the distortion that comes from different makes/models of film caps w/ differing construction methods and plotted distortion vs. frequency for all of them on a single graph.  It was interesting.  According to his tests, there are small but real differences in the way different types of caps distort different frequencies.  They are small, but he concludes that it is possible that those small differences could add up over the course of a circuit to be audiable.  WIMA red box films had the most distortion out of the kinds tested (kind of surprising given their more hifi reputation), and has a fair bit of 2nd harmonic distortion. 

So there you go, it's at least possible.  Although in your case you probably didn't change enough caps to make change in type add up to much, and what you're hearing is related to changes other than type (difference in value, ESR, impedance change (or placebo effect)).

[Boba7]

Thanks! Yeah I'm not a believer in those kinds of things either, I've tested in the past to hear the difference between different types of capacitors and just couldn't, it was impossible.

I don't know enough about electronics to really understand all this, and I was very surprised by the difference in sound. There were not many capacitors in the signal path but maybe they were getting old, dry, or maybe replacing 10u with 1u made all the difference. I thought there was no reason to have large caps like that and 1uf should be enough to let all the signal pass (and I think I'm right, as far as I can tell)

I should probably buy Merlin's books!!

[lars]

To me with the electros in the signal path, there was almost a sort of high pass filter going on.

The difference in sound is because of changing the values of the caps, not because of some magical properties of their construction. Pretty much every capacitor/resistor combination in the signal path forms some kind of low or high pass filter. Since you changed from 10uf (a value that generally allows more bass to pass), to 1uf (a value that generally allows less bass to pass), you are going to hear a "brighter/clearer" sound. It's not because a film cap can transmit beautiful crisp highs while an electro is mud machine.
I used to work in the home/car audio scene, and there are some crazy theories about hifi components that "experts" continually promote as doctrine. Capacitors have always a been a huge debate.

[Boba7]

The difference in sound is because of changing the values of the caps, not because of some magical properties of their construction. Pretty much every capacitor/resistor combination in the signal path forms some kind of low or high pass filter. Since you changed from 10uf (a value that generally allows more bass to pass), to 1uf (a value that generally allows less bass to pass), you are going to hear a "brighter/clearer" sound. It's not because a film cap can transmit beautiful crisp highs while an electro is mud machine..

Yeah but that's what seems very odd to me : the sound was more trebly and a bit thinner with the old electro 10uf! I compared it with my other buffered pedals, either Boss (bjt buffer) or DIY (jfet buffer) and it sounded a bit "off", thin. Then after changing a resistor and all the caps, the sound is much more pleasant to my ears and very close to my other buffers (can't hear a difference now)

[culturejam]

To me with the electros in the signal path, there was almost a sort of high pass filter going on.

That's because they are each forming a high-pass filter (the coupling caps, that is. Decoupling caps are forming low-pass filters).

It's possible you are hearing a change in the cumulative/total frequency response of the buffer because several caps were changed. The values probably shifted a bit (based on tolerance of each cap) and that moved the cutoff frequency. 

[Boba7]

That's because they are each forming a high-pass filter (the coupling caps, that is. Decoupling caps are forming low-pass filters).

It's possible you are hearing a change in the cumulative/total frequency response of the buffer because several caps were changed. The values probably shifted a bit (based on tolerance of each cap) and that moved the cutoff frequency.

Of course. But I was surprised that with the 10uf the sound was actually thinner, a bit brighter and with a bit less bass than with the 1uf.
I measured the ESR of two 10uf caps, they're between 4 and 5 ohms. Is that normal? Just trying to understand...

[lars]

Another thing to consider is that you probably changed the bias points of those FETs by adjusting several of the cap values in the circuit. Changing the bias point of a FET can change the sound drastically.

[Boba7]

Another thing to consider is that you probably changed the bias points of those FETs by adjusting several of the cap values in the circuit. Changing the bias point of a FET can change the sound drastically.

Interesting, though again I really don't know enough to understand what you're saying
There's just a FET at the input. 

All this really makes me wanna read more about electronics, there's so much I don't know and understand.