Does cap voltage affect tone?

[tedmich]

OK Chuck! I wish you were there when I got spanked over Tantalum caps in another thread... Enzo and others greatly dislike them. They have nice low voltages which are perfect for cathode bypass and they are much better than EL in many applications...they do explode sometimes when greatly abused (MnO2 ->O2 + BOOM)

[Chuck H]

There is alot to be learned from the (ehem...) more experienced members here. In fact guys like Enzo and Bruce are the biggest part of what makes this forum the best of it's kind. But the idea that tantalums are crappy caps that will always be a problem waiting to happen IS a bit antiquated. Modern tants have very reasonable specs and IME are reliable in the given application. Further, their audio properties are far superior to electros. I don't think I would even consider tants in most applications, but if they ever had a niche, this is it. Consider that Mesa Engineering has been using them for over a decade. The biggest failure issue with those amps is burned PC traces. I've never even heard of a failed bypass cap. But consider that they (and we) are using 15 and 20 or 25 volt caps in 1.5 volts of operation. Still, this seems to make sense if it's giving results. And it is for me.

Chuck

[daz]

Well, i hate to be the naysayer, and i'm not saying i am doing a 180 here. But i really need to test any changes i do in a live situation. And being that i seldom play out anymore, the way i do it is with a cranked up sound system pounding out tunes of the type i prefer. I did that tonite, and i found that as good as the changes i made seemed, in a mix there were issues for me. Big surprise here....why i don't expect this kind of thing by now i don't know. But sure enough it was not happening. I loved the tone i was getting. But as soon as i played it in a mix it was unusable. I don't know exactly what caused that, but the changes i made were to put tantalums and epoxy caps in 4 places including the first 2 stages, as a gain boost which was a switchable cap on the 2nd stage cathode, and a .68uf cap in the NFB loop. Instead of removing them one by one to see where the issue may be, i reinstalled all the original caps at once. Mainly because i don't have the luxury of cranking the amp for long periods due to the neighbor issue. In any case, when i reinstalled the original caps i was back to what works. the difference was that the tone sort of fit in the mix nicely verses being TOO fat and big to fit in the mix. It was also less rich with harmonics. The only way i can explain it that might make sense is to say the tone was like a sine wave in that there was no harmonic content i could hear when in the mix, and the highs just disappeared. Mind you that it sounded great by itself. But as always when try something new it always is nite and day between how it sounds with the amp by itself and when in a mix.

So what i think i have to do is take the time and try each cap change one at a time and try to find which caps that will make a positive difference in a mix and which were the ones that were responsible for ruining the tone within a mix. One thing is for sure, and i've dealt with this for years, no change can be evaluated till it is heard within a mix. Playing the amp by itself is so very tricky because what sounds fantastic can easily be and often is just the opposite within a mix. what it comes down to Chuck, is that these caps may well be great, but you have to experiment and find where they work right for you and whether they need a change in value too. All, i can say at this point is that putting everything back to stock i got my usable tone back. But i feel in certain positions they will be an improvement. I just have to take the time to experiment with them and find which positions they do the magic in.

[Chuck H]

I couldn't agree more. "In the mix" is the only way to know anything with amp design. I've had some horendous experiences playing out with some of my early mods I originally thought were great.

But... Consider that your ears get trained to your gear. I recently have been evaluating speakers for an amp I designed and not surprisingly my favorite speaker with this amp is the same speaker I've been playing all my amps through for the last twenty two years. I knew at that point that my brain was playing shinanagins on me. So I asked my wife to do a listening test... She chose the "other" speaker Who knows at this point. I have a vision. But apparently it is narrowed a bit by conditioning. I'm really trying to like the other speakers. But it's just not working out.

There's a tip in here somewhere.

Chuck

[daz]

I know what you mean, and it is very very tricky at times. there are times you just cannot understand how you can be hearing what you are hearing when things were so different before, even when you haven't done anything. But even being "trained" to hearing the amps' sound, i just could not adjust the amp to get remotely close to fitting in the mix without being a completely different and uninspiring sound. I will try each one one at a time and see if i can get back some of the good w/o the bad.

[cerrem]

If your using a way bigger voltage rating electrolytic cap....it will behave like a much lower uF cap at lower voltage...
The plates will pull towards each other at full voltage and "press" the dielectric material to a smaller thickness" thus larger capacitance...
Depending on the material ....Capacitance vs Voltage Gradient across the dielectric....some cases are more dramatic than others...this is the effect that occurs when a cap re-forms after discharge....
The affects of caps are barely noticeable at low levels and when viewing normal sin waves in the audio range.... things become more apparent when the amp is cranked and you are making square waves ....the harmonics extend past the audio range at 1/f 1st order roll-off...then roll off at 1/f^2 at some given dominant POLE in the circuit...
Depending on the Tantalum cap.. many have ESR as high as 10 to 20 OHMS.... this will dampen any ringing in the square waves and make for a smoother guitar sound... In MIL design the Tantalum caps must always have some serious derating...
I have used this MIL-STD 1547B for a lot of work I did in the past... It is a good guidline to use for Hi-REL design...
http://everyspec.com/MIL-HDBK/MIL-HD...1547A.2102.pdf

[J M Fahey]

If your using a way bigger voltage rating electrolytic cap....it will behave like a much lower uF cap at lower voltage...

It doesnīt happen to any significant degree.
If capacityance were voltage dependent, 2 things would happen:
1) it would create A LOT of distortion, and that effect would be widely known and studied ...it is not.
2) being a significant parameter, Manufacturers would show this in datasheets, including some curves or graphs ... they do not.

The plates will pull towards each other at full voltage and "press" the dielectric material to a smaller thickness" thus larger capacitance...

"Sounds" logical as a piece of conversation but would love to see some data about it ... lots of things "sound logical" but then donīt happen in reality.

Depending on the material ....Capacitance vs Voltage Gradient across the dielectric....some cases are more dramatic than others...this is the effect that occurs when a cap re-forms after discharge....

Thatīs different, and is a sllloooowwww process which takes hours or days, absolutely unrelated to Audio frequencies.
A cap does not "reform-unreform" many times a second like when Audio is applied.

The affects of caps are barely noticeable at low levels and when viewing normal sin waves in the audio range....

You bet.

things become more apparent when the amp is cranked and you are making square waves ....the harmonics extend past the audio range at 1/f 1st order roll-off...then roll off at 1/f^2 at some given dominant POLE in the circuit...

Yes on power tubes, and OTs can (and should) have wider bandwidth than raw Guitar signals ... but are played through Guitar speakers which fall like a brick, 24dB/Oct or worse, above 3 or 4 kHz.

Depending on the Tantalum cap.. many have ESR as high as 10 to 20 OHMS.... this will dampen any ringing in the square waves and make for a smoother guitar sound...

How can they do that if they are on *cathodes* ?
And to boot usually on Preamps, where gain stages are followed by strong tone shaping or plain HF killing, notice those 100pF to 1000pF caps in parallel with plate resistors in many High Gain amps ... those which are designed to make lots of funny squarewaves.

In MIL design the Tantalum caps must always have some serious derating...
I have used this MIL-STD 1547B for a lot of work I did in the past... It is a good guidline to use for Hi-REL design...
http ://everyspec. com/MIL-HDBK/MIL-HDBK-1500-1799/download.php?spec=MIL_HDBK_1547A.2102.pdf]

Now I see.
I congratulate you on your wide RF and MIL Electronics Experience and understand you see things through that lens, but in Guitar land things are way slower and easier.

[Helmholtz]

Quote Originally Posted by cerrem View Post
If your using a way bigger voltage rating electrolytic cap....it will behave like a much lower uF cap at lower voltage...

It doesnīt happen to any significant degree.
If capacityance were voltage dependent, 2 things would happen:
1) it would create A LOT of distortion, and that effect would be widely known and studied ...it is not.
2) being a significant parameter, Manufacturers would show this in datasheets, including some curves or graphs ... they do not.

And it would show when measuring uncharged HV electrolytics with an LCR meter. While such measurement may actually show a reduced capacitance value by 20% to 30%, a bias voltage of 5V (meaning that the cap is charged to 5 VDC during C measurement - some LCR meters provide this useful option) typically suffices to bring C (and ESR) up to spec for a good cap.
So definitely not "like a much lower ĩF cap".

[cerrem]

It doesnīt happen to any significant degree.
If capacityance were voltage dependent, 2 things would happen:
1) it would create A LOT of distortion, and that effect would be widely known and studied ...it is not.
2) being a significant parameter, Manufacturers would show this in datasheets, including some curves or graphs ... they do not.

"Sounds" logical as a piece of conversation but would love to see some data about it ... lots of things "sound logical" but then donīt happen in reality.

Thatīs different, and is a sllloooowwww process which takes hours or days, absolutely unrelated to Audio frequencies.
A cap does not "reform-unreform" many times a second like when Audio is applied.
You bet.

Yes on power tubes, and OTs can (and should) have wider bandwidth than raw Guitar signals ... but are played through Guitar speakers which fall like a brick, 24dB/Oct or worse, above 3 or 4 kHz.

How can they do that if they are on *cathodes* ?
And to boot usually on Preamps, where gain stages are followed by strong tone shaping or plain HF killing, notice those 100pF to 1000pF caps in parallel with plate resistors in many High Gain amps ... those which are designed to make lots of funny squarewaves.

Now I see.
I congratulate you on your wide RF and MIL Electronics Experience and understand you see things through that lens, but in Guitar land things are way slower and easier.

Sounds like you are clueless..... I could explain for 100 pages and you still wont get it... You don't even know what a Network Analyzer is... Look up a HP 3577 and read... It's an ANALOG signal sweep and it sweeps as low as 5 Hz ... not in the RF range... its not DIGITAL it measures impedance ..Been designing and building audio amps guitar amps for 45 years.. Also design capacitors for a major manufacturer....Stay out of my post and stay out of yours...

[Enzo]

OK, we often hear of various phemonena. So answer this:

a typical 20uf 500v cap as used in a zillion B+ supplies. I will believe you that the capacitance might effectively change with voltage applied, but I have trouble with "much lower" and "much higher". So HOW MUCH does the capacitance change when using a 20uf 500v cap on say 25v?

I did read through the 300+ page document you linked, but unless I missed something it seemed to focus on tantalums and film caps. I didn't spot plain old vanilla electrolytics.


I would consider a 10% shift to be within tolerance.

[pdf64]

Sounds like you are clueless...Stay out of my post and stay out of yours...

That's just silly, please don't 'take on'. Please don't act like a 'god on high' and rage if someone should dare to raise a query.
This is a public forum, for better or worse there's no 'your thread' and 'my thread'.

Other forums may happily lap stuff up and not be troubled by the paragraphs of missing material represented by each ellipsis.

The nature of this forum is more akin to 'peer review'; if a technical statement is made, even one by a competent, venerable member, the onus is on them to justify it, provide at least something to verify the point.
We're not expecting a full report with exhaustive citations etc but there's really no point just writing conclusions without 'showing us your workings', ie give a guideline by which someone similarly resourced and competent could reproduce your finding.

I respect you and I'm very pleased that you've decided to grace us here with your insights.
However, the points made in response to your posts contain queries that crossed my, and I'm sure many other reader's, mind.
A prime example being your remark about tantalum caps and their ESR acting to dampen ringing with square waves; such a statement is frustrating and basically useless without any context, examples etc.

I acknowledge this may make creating material hard work; an example of how this can be done well may be found http://music-electronics-forum.com/s...ad.php?t=46108
Ultimately, it leads to a better result, with everyone involved in the thread, including even the OP, perhaps gaining a greater understanding of the topic.

As such, this forum occasionally exemplifies, in some small way, the real benefit of the www, as by documenting and publishing such new, previously unknown / unpublished or forgotten insights, the sum total of human understanding of our universe has been incremented ever so slightly.

[Chuck H]

Hey cerrem. To put it another way... Glad you're here. Please take it down a notch

And Happy Birthday.

[J M Fahey]

Not answering to anybody in particular, just taking a *realistic* look at what can be found on any Guitar Amp related Tech bench.
And I am not talking only a regular "Service" Tech, who usually is involved in repairing stuff back to original working state but with no need for reengineering, but also including Factory Engineers (think Fender/Peavey/Marshall/Ampeg/Laney/etc.) who do design Production Amplifiers.

Somewhere I heard an HP3577 being mentioned:
http://www.testequipmentdepot.com/us...zers/3577a.htm

Yes, signal output can go as down as 5 Hz ... somebody might want to test a Network performance that low ... considering it touts a 200MHz range I somewhat doubt low frequency measurements are its forte.

Wonīt copypaste what it can do because I would fill the page, just notice that none of its about 100 impressive features includes or even mentions measuring capacitors, itīs focused on Networks, only component they mention is transistors, probably because their parameters influence the Networks they drive, for example RF transistors often include Smith Charts.
Very justified there, useless for us.

Doubly doubt its presence on any Service (or Amp design one) is because, besides complexity , this used-refurbished unit carries a price tag of

Price 3577A Agilent / HP 3577A 200 MHz Network Analyzer, Refurbished $3,695.00

FWIW the current price tag for an equivalent modern one is
https://www.keysight.com/en/pcx-x205...0&cc=US&lc=eng

US$ 11,433 Starting Price*

But thatīs not the point ... IF it were that expensive AND its use were worth it, THEN it would find its way into Tech benches , one way or the other ... somehow I donīt see that happening.

A small side note, since the HP3577 specs

.001 dB, .005 degree, 1ps, .001 Hz resolution

I can understand somebody used to such precision fumes and boils when watching sloppy 0.1dB change on anything and calls that a MAJOR change .

We mere Mortals who can not get even near Olympus usually accept 3dB as a clearly detectable change and would be hard pressed to detect 1dB under very relaxed and stable conditions (as in: silent room and not a drummer within a mile).
So we are 1000 to 3000 times "deafer" than such a marvellous Instrument .

Why would anybody use a Network Analyzer to test a Guitar Amp capacitor?

All I can think is (and this really belongs in the other thread):

if all you have is a hammer, everything looks like a nail
https://en.wiktionary.org/wiki/if_al...ks_like_a_nail

Me? ... I am happy with a humble but faithful analog LCR Bridge

which I have for over 35 years now and can give me all relevant answers, it only takes a couple minutes fumbling with controls, and if I were setting up a new Bench today I would find use for a digital Capacitance meter and an ESR meter which give instant results.

Of course, Olympus Gods would sneer at such humble setups.

[loudthud]

Depending on the Tantalum cap.. many have ESR as high as 10 to 20 OHMS.... this will dampen any ringing in the square waves ... and make for a smoother guitar sound... In MIL design the Tantalum caps must always have some serious derating...

In an oscilloscope, almost every power supply voltage is regulated, possible exception might be a fan. Those old Tektronix tube scopes were about half power supply. At some point in test or calibration, every power supply is tested for transient response and there is a spec that the supply has to pass. In the early 80's (when I worked there) Tektronix had to substitute Aluminum electrolytics for Tantalum caps on the output of regulated power supplies in most of their scopes because of cost. The capacitance was increased about 2X to get the impedance in the same range at high frequencies. The ESR of the Aluminum caps was higher and that damped the high frequency ringing.

Tantalum caps have lower ESR at high frequencies, can operate at higher temperatures and they last longer. That's why they are used in high priced equipment that is meant to last longer that consumer gear. The only time I've ever seen a Tantalum cap derated was when they had to operate at extended temperature like in down-hole (oil well) equipment where they have to work at 200 degrees C.

[Chuck H]

He's right about the high esr in some models though. especially some of the bullet shaped dry tants that I have used. They are indeed rated in tens of ohms (at least on the Mouser description). I suppose this anomaly doesn't apply to all tant caps, but I thought I'd point this out. I remember being surprised because I thought that by default I would get better esr with the tantalums. Turns out I have to pick through them. Not that it matters a whole lot with a cathode bypass cap. 20 ohms compared to 2 ohms across a 1500 ohm resistor!?! I'd need to hear it to believe there's an audible difference.