Does cap voltage affect tone?

[Helmholtz]

Just to demonstrate that you don't need a 10,000$ HP analyzer for comfortable Z measurements, here a few curves taken with my cheap Velleman PCSU200 USB scope+synchronized signal generator:


The generator source impedance was 100 Ohms. The horizontal blue line is the "Z-response" of a 1 Ohm resistor for reference/calibration.
The red curve shows one section of a 2x32µ/350V cap I pulled from my 1967 Marshall. This type has the highest inductance.

[Chuck H]

Excellent! But... Could I trouble you about measured UF values for the caps tested? I think it could be beneficial to the discussion.

[Helmholtz]

Excellent! But... Could I trouble you about measured UF values for the caps tested? I think it could be beneficial to the discussion.

Oh boy, they are all back in their boxes, but I made sure that all were within +/- 10% of rated value. All are at least 20 years old. The C value can be estimated from the lay of the -6dB/octave slopes.

But what specific discussion?

[J M Fahey]

Just to demonstrate that you don't need a 10,000$ HP analyzer for comfortable Z measurements, here a few curves taken with my cheap Velleman PCSU200 USB scope+synchronized signal generator:
[ATTACH=CONFIG]49647[/ATTACH]

Thanks for this very useful post, which clear more points than it caims to
Such as:

* confirms previous statements and tests that much maligned Electrolytic capacitors are much better than their "internet fame" would say.

* no excessive inductance *inside the audio range*

* no weird (or even mild) resonant peaks *inside or outside the audio range*

* ESR low enough that it does not interfere with expected/calculated hum/noise/ripple filtering action.
Curves stop sloping down at 6dB/Oct (expected capacitor behaviour) and become horizontal (resistive behaviour) at 30kHz to over 100kHz ... and then we are talking between 1 ohm and fractions of an ohm magnitude, degradation is minimal.
Consider in RC smoothing stages in a preamp, R values go from 10k to at most 1 K ... compare that to 1 ohm or less in series with the cap.
"Butterfly in China" effect indeed

* some people worry and overthink about parts leads or even parts body self inductance ... but a **1 ohm** resistor is perfectly flat to 1 MHz , so a higher value one will be even less affected than a 1 ohm one.

Just found the Velleman thingie in Amazon ... loved it and will probably order one, thanks for showing it here

Excellent! But... Could I trouble you about measured UF values for the caps tested? I think it could be beneficial to the discussion.

I think the curves clearly show that the green cap has somewhat less capacitance than the black one.
How much less? ... dunno but probably 10% to 20% less, within what Helmholtz selected previous to experiment.
Notice curves are practically parallel in the "mostly capacitive" area and separate in the resistive area, I guess the green cap has about 3 times the black cap ESR. (because it´s about 10dB higher in this Log Voltage graph)

[Chuck H]

Rethinking it I don't suppose it matters beyond what the graph shows. My first thought was indeed the green vs black difference. For some reason I momentarily thought that an actual measured value would help disprove a significant capacitance change at different voltages. Only after posting did it occur to me that ALL the test equipment is operating at voltages significantly lower than some of those cap ratings. So...

Managed to prove out pretty much everything in question (including the need for a bench tool we'd need to make payments on) except whether or not electrolytic capacitor plates are squished together at rated voltage enough to alter the value significantly.

[nosaj]

Rethinking it I don't suppose it matters beyond what the graph shows. My first thought was indeed the green vs black difference. For some reason I momentarily thought that an actual measured value would help disprove a significant capacitance change at different voltages. Only after posting did it occur to me that ALL the test equipment is operating at voltages significantly lower than some of those cap ratings. So...

Managed to prove out pretty much everything in question (including the need for a bench tool we'd need to make payments on) except whether or not electrolytic capacitor plates are squished together at rated voltage enough to alter the value significantly.

You could always take a pair if hemostats and clamp one on a cap to prove or disprove that idea.

nosaj

[Chuck H]

Ha ha! I was thinking more along the lines of measuring a known parameter under controlled conditions that is relative to capacitance AND voltage. Something like reverse analyzing the capacitance by the charge time with other factors, such as ESR already known. This goes beyond my skill set.

But I like your idea too. Wear gloves

[nosaj]

Ha ha! I was thinking more along the lines of measuring a known parameter under controlled conditions that is relative to capacitance AND voltage. Something like reverse analyzing the capacitance by the charge time with other factors, such as ESR already known. This goes beyond my skill set.

But I like your idea too. Wear gloves

Well going by that one guys suggestion that the plates pull together in a cap, why wouldn't squeezing one simulate his idea.

Heck just a cap out of the junk box put it on a capacitance meter and squeeze it . I doubt it will change but when I get back from the store. I'll put one one the der5000 and see if it changes. I''ll try a film and a electrolytic, The electrolytic will be damaged so it's going to take one for the team.

Report back this evening.

nosaj

[Enzo]

Prolly not on a 22uf e-cap, but on some film cap, squeezing with your fingers might upset the readings some with your body capacitance.

[Chuck H]

Well going by that one guys suggestion that the plates pull together in a cap, why wouldn't squeezing one simulate his idea.

Heck just a cap out of the junk box put it on a capacitance meter and squeeze it . I doubt it will change but when I get back from the store. I'll put one one the der5000 and see if it changes. I''ll try a film and a electrolytic, The electrolytic will be damaged so it's going to take one for the team.

Report back this evening.

nosaj

Where can I get a der5000!?! I don't want to be under equipped for building guitar amps. I can put it right next to the HP3577 I just ordered based on some readings in this thread. Wow! I'm gonna have the most badass bench ever.

[Leo_Gnardo]

Where can I get a der5000!?! I don't want to be under equipped for building guitar amps. I can put it right next to the HP3577 I just ordered based on some readings in this thread. Wow! I'm gonna have the most badass bench ever.

Will you be satisfied squinting at meters? Or do we build/fix these things to satisfy our ears. How many gazillions of amps have been built without the "benefit" of (over)analysis. I gots to wonder... just WHO are the clients that have benefited by having their amps put thru the super duper NASA/NSA-approved network analysis process. By name, not just "a long list of very impertinent people."

[Enzo]

I have a couple HP distortion analyzers, I think a 331 and 332. Maybe you could use those?

[nosaj]

Where can I get a der5000!?! I don't want to be under equipped for building guitar amps. I can put it right next to the HP3577 I just ordered based on some readings in this thread. Wow! I'm gonna have the most badass bench ever.

I say it's bunk. I put a 220uf cap in the meter(Chuck it's just a $70 LCR meter nothing super badass but beats spinning dials on a Genrad 1650a) Squeezed it slightly no change, harder denting shell moved up 1 uf, crunch it really badly goes up to 330uf but the cap is now damaged. Put a .1uf film cap in squeezed it no change .
Seem UF only changes if you really apply some force to it and when you see change the cap is damaged.

nosaj

[cerrem]

The ubiquitous 25uF value (and notice that almost never it is , say, 10uF or 47uF ... why?) , if you think about it, is weird because it acts way below audible frequencies
http://www.sengpielaudio.com/calculator-RCpad.htm
tells us that typical 1k5 + 25uF has a crossover frequency (somebody would love to call it a Pole instead, but won´t speculate on the preference) of .... drumroll!!! .... 4.24Hz !!!!!
Hardly an Audio Frequency, huh?

Dr. Juan ... I believe you may have confused a Crossover frequency and a POLE... not the same thing... The POLE is the -3dB point you may be referring to...
Were you referring to Phase Crossover or Gain Crossover ??
The 4.24 Hz is off by significant factor...
The Web Calculator does not apply to this topology....
It's not that simple...since you need to include the impedance "looking into" the Cathode...
With a 12AX7 and a 100K plate resistor in this Fender circuit... Your looking at roughly 1.54K in parallel with you 1.5K cathode resistor...when you make this calculation...
I still get your point.. that this was all below the guitar range...

[Helmholtz]

Excellent! But... Could I trouble you about measured UF values for the caps tested? I think it could be beneficial to the discussion.

O.K., here you are (I think I picked the same caps as before).

Measured values (using a PeakTech 2170 LCR/ESR meter with 4-wire Kelvin clamps):
(the numbers mean C@1kHz/ESR@1kHz/ESR@10kHz/ESR@100kHz)

black: 23.6µ/0.46/0.37/0.33 (22µ/450V, Siemens Long Life, 1992)
red: 31.6µ/0.93/0.0.77/0.0.71 (32µ/350V, 1967)
mag: 33.0µ/0.62/0.54/0.41 (32µ tantalum)
green:20.7µ/1.46/1.24/1.06 (22µ/35V)