The Supercapacitor

[Guitarist]

Wondering if this might work as the sole cathode component in a preamp: 1.0 Farad 2.5 vdc 1.0f 1.0 f 2.5v Super Capacitor - eBay (item 110467803449 end time Mar-03-11 10:26:00 PST)

The many 1.2V rechargeables I've used similarly are still working perfectly so I perhaps this could too.

You may or mayn't like this tone:
Take 1) Is just the one 12au7 tube. 1a has the battery feding 1b the cathode follower. This is direct tone, no eq, Strat.

2) Now sent into a Digitech GSP1101 modelling a Sunn amp and speaker,on a medium-gain setting. (I normally feed this into a stereo tube amp and normal guitar speakers.) Sorry, was playing with this vs having a smarter nap.

We talk a lot about tube amps here and some listenables are better than none But back to the supercapacitor, is there a place for it in a tube amp?

[J M Fahey]

Don't understand what you mean, please post a schematic.
Thanks.

[mooreamps]

Sounds pretty sweet to me...

-g

[Guitarist]

Schematic

Sure, here it is. The plate resistor needs to be beefy as 1/2 watters will fry and 2 watts might be better still. I always use a low gain, tube buffer whenever possible. I called it the Sugertone as it makes modellers a fair bit better, thus keeping them tolerable.

EDIT: removed sidetracked details.

[J M Fahey]

I'm very sorry but still don't get it.
The thread title refers to a "Supercapacitor"
You link to an EBay offer of :"1.0 Farad 2.5 vdc 1.0f 1.0 f 2.5v Super Capacitor"
Yet I can't find it anywhere on the schematic nor do you refer to it at all but to a 1.2V rechargeable battery.
What am I missing?
You also refer to a "Tube Buffer", with unspecified B+ on the schematic, but you mention "rechargeable 9V pack".
You also mention powering the 6V filaments with 3 NiMh batteries in series (you show the 3 cell holder).
Available voltage would be 3 x 1.2V = 3.6V.
So your tube buffer runs out of a 9V power supply, with 3.6V feeding the 6V filaments.
Can't criticize you at all; VOX runs "Cool Tubes" out of 6V and 3.3V respectively and makes a lot of money out of them.
But they put a Led under the tube so that *something* glows.
Congratulations on your part overspecifying too.
You also suggest 1W 47K resistors, preferably 2W, because "1/2 watters will fry ".
That's playing it *real* safe.
Even if the tube shorted, they would dissipate 81/47000=0.0017W (1.7 milli Watts)

[Steve Conner]

I think the idea is to use a supercapacitor in place of a cathode resistor and its bypass cap.

Not a very good idea IMO. The rechargeable batteries have an internal chemical process that sets a well-defined upper limit to their voltage: generation of oxygen or whatever it is, will clamp the voltage to about 1.3-1.4. If the current isn't too high, they can run like this forever, the oxygen just recombines internally. So they work like a LED with a colossal capacitor across it, 360 Farads or whatever.

But a supercapacitor doesn't have any mechanism like that that I know of. The capacitor is there, but I think the "LED" is missing. If it does have one, I can't see how the voltage would be well-defined.

[Guitarist]

Alright, I'll stick to the original question and start another thread regarding a fully battery powered preamp.
Would this type of cap work somewhere in our amps such as what is illustrated by this schematic (subbing for the battery on the cathode)? Or perhaps elsewhere.
I had 1/2 plate resistors in this design before, and can testify that they got HOT ( I felt 'em and they fizzled out. When I lost sound I checked and fixed the buffer with higher wattage resistors and now it works fine. Why this is I don't know, perhaps the battery amps it so. Similarily, I just fixed an amp with a Marshall style CF with a failed 1/2 watt 100k cathode resistor. So now I play it safe with overkill, at times.

I own such a Vox pedal and it's now on eBay as it sucks. I don't think this is soley the tubes fault though, but those granted (good point) those values are rather low.

[teemuk]

What first comes to mind is this:
Yes, the super capacitor can hold its charge for a very long time but how will it actually charge?

Unless you use precharged super capacitors you will need a DC source for actually charging the caps, at which point you could just as well use this DC source alone as the biasing means (e.g. Magnatone amps). If you decide to use pre-charged devices then you might as well use a battery or a bising cell.

[mooreamps]

We talk a lot about tube amps here and some listenables are better than none But back to the supercapacitor, is there a place for it in a tube amp?

Hi.. Mad Rocket Scientist here.... Ya, well of course I'm going to have to look into this... To answer the question "is there a place for it in a tube amp ?" Well, there might.. But maybe a better question would be "which kind of tube amp". The bias voltage looks correct, and it seems the "device" has no problem passing the A/C component of the signal... I don't know, seems to be working to me....

-g

[R.G.]

I googled "supercapacitor" and "supercapacitor failure modes".

From Wikipedia:

Has the highest dielectric absorption of any type of capacitor.

Hmmm... from hifi tweeko political thought, "dielectric absorption" is by definition "Bad".


From cap-xx.com:

What is the frequency response of your supercapacitors?
Typically supercapacitors are used for low-frequency, less than 100Hz, applications but this is not true of all supercapacitors. CAP-XX supercapacitors can be used in pulsed applications such as GPRS and GSM with pulses
= 500 µS. In fact, our supercapacitors are still very effective at pulse widths down to 100 µS. CAP-XX has found that a term we call effective capacitance or Ceff is a much better indication than a traditional frequency response plot of supercapacitor behavior in pulse power applications. Ceff reflects the percentage of DC capacitance available at a given pulse width and can be thought of the time domain representation of the frequency response curve. For a 1 millisecond pulse a typical CAP-XX supercapacitor will have approximately 50% of the DC capacitance. Ceff varies with device type and temperature so contact CAP-XX for more information.

Hmmm... this is from a maker and seller of supercaps. I translate that above as "we don't want to talk about 'frequency response', we want to talk about battery applications." It appears that the apparent capacitance varies with the frequency of applying/removing charge, and that at a 1mS pulse the apparent capacitance will be down by 50%. Uh... 1Ms is the cycle time of a 1kHz signal, right?

Well, maybe that's OK, since the initial capacitance is 1F or so, and it could be down to half a farad and still bypass a 2.2K cathode resistor fine.

Oh, wait. There's no other part there, just the supercap. The resistance it's bypassing is its own esr, which is microcscopic. So - what is the frequency response??

What is the failure mode of a supercapacitor?
If a supercapacitor is exposed to excessive voltage or temperature for extended periods it gradually degrades to essentially an open circuit. The time taken for this to occur depends on how much over voltage or over temperature is applied. Other than physically puncturing the supercapacitor there are no short circuit failure modes. There are no catastrophic failure modes.

Good news. It won't explode. However, it does charge up like a capcitor, with I = C*dv/dt. or dv = Idt/C. If you're using it in a tube amp the plate current flows into it for as long as the plate supply is on.

Which leads to an interesting question: How long does it take to get to 1.2V?

Well, dt = CdV/I, and taking C =1F, dV=1.2V, and I =1ma, then dt = 1*1.2/0.001, and the answer is 1200 seconds - it takes 20 minutes to charge up to the desired operating voltage if you only use the plate supply. But the plate supply won't stop there. It will keep cramming in the 1ma, and in an hour, the supercap will be up at 3.75V. Oh. That's an overvoltage for the cap.

But wait - that much voltage will shut down the tube and it won't let current through! It won't overvolt the cap. Uh-oh. The tube turning off won't let audio through. Oops. So the tube will sound good as it passes through the good bias point on its way to cutoff. And the tube will trickle more current into the supercap as long as the plate supply is on because the supercap will never be able to fully cut off the tube. So if you leave it on a really long time,

it gradually degrades to essentially an open circuit

It's got a built in failure mode, which is sometimes known in the rural USA South as a self-eating watermelon.

Well, no problem. We just keep the (expensive!) supercap from overvolting by putting a resistor to eat up that amount of current to keep it at 1.2V. Other than the objective being to NOT use anything other than the supercap itself being the original objective, and that resistor happening to be the exact value you'd have used for self biasing in the first place, no problem.

OK, we can use a 1.2V battery across it to keep it at 1.2V forever! Right - except that both batteries and supercaps have self-discharge, internal leakage that runs them down. In this case, the super cap, which has a much higher self discharge rate than batteries, runs down the internal battery during the time the amplifier is off. Put it into a closet for months and when it comes up, you now not only have a flat supercap, you have a flat battery that was "holding" the supercap at the right voltage.

OK, we can use a supercap AND a rechargeable battery that we let go flat! Oh. That means that when we power up the amp, not only does the 1ma plate current have to charge the super cap, it has to charge the battery too, so maybe it's an hour or two of power on time til we can get it to sound good.

No problem! We'll just put a trickle charger on the amp so that when you turn off the amp, you leave the trickle charger plugged in to keep the rechargeable battery charged to keep the supercap charged for the next time the amp's turned on!

And we'll do one of these for each preamp cathode that needs biasing and a cathode cap! Supercaps on ebay are about $3-$5, the rechargeable battery is about $2, and we'll do this five or six times in a preamp. But we do get to use the same trickle charger for all of the tubes/supercaps!!!

I'm thinking I'll stick with other solutions. Just me. Your mileage may vary.

[mooreamps]

I googled "supercapacitor" and "supercapacitor failure modes".

From Wikipedia:

Hmmm... from hifi tweeko political thought, "dielectric absorption" is by definition "Bad".




I'm thinking I'll stick with other solutions. Just me. Your mileage may vary.

Ya, well,, we can sit here at our keyboards and write paper until the Second Coming... Or : we could cobble up a prototype circuit and measure the test performance... I vote for the later...

-g

[R.G.]

Ya, well,, we can sit here at our keyboards and write paper until the Second Coming...

We could. I've seen people do that.

Or : we could cobble up a prototype circuit and measure the test performance... I vote for the later...

I've always been a fan of taking a quick look at what I'm getting into before investing time and money into prototypes, but again, that's just my personal preferences.

I learned by experience as a kid that poking a stick into the water to see how deep it was before jumping in was a good idea, both for finding out if it was too deep or too shallow.

But then I prototype slowly. It would take a few days to locate a super cap and either get out and buy it or get it shipped in, and then build the circuit, put it through its paces and wait for how overloading, etc affected it, ordering several more from other manufacturers to see how alike they were, life testing under temp variations, etc.

I *like* to read about the experiences of people who cobble up a prototype circuit and measure the test performance. So you go, guy. I'm eager to see what you turn up with. After all, proving that easily available data is wrong or fraudulent is always fun, right? Lemme know what you find. I'll be waiting.

[J M Fahey]

Well, let's not forget that for biasing purposes (which by definition means a definite and stable voltage), a capacitor, as in *ANY* capacitor, is the absolute *WORST* voltage biasing source, as connected in that schematic, because *by definition* (I do not need to experiment it myself, at all) it is constantly changing voltage all the time.
That's called "charging a capacitor".
A humble resistor is predictable and precise, in that a given current through it will cause a predictable and precise voltage drop across its terminals.
So much so, that Congress passed a Law suggested and lobbyed by Senator Ohm (R., Ark.), popularly known as Ohm's Law, so much admired by Democratic and antiDemocratic Governments alike, that it's followed the World over.
In fact, it rules in Outer Space too.
There's been a growing Ecologic movement to make all manufacturers in the World bias by means of Resistors, with very few exceptions allowed, and even so, most have to include some kind of resistor in their guts anyway.
Some of the most extreme non-resistor biasing systems use a (charged) battery, forward biased diodes, either regular or LEDs or reverse biased ones, AKA Zeners or Avalanche Diodes.
Even then, there's a still unvoted but already much followed Law, popularly called the Economics Law that in general terms frowns on the use of these last exotic biasing means.
Mind you, they are not forbidden !!
Oh well.
.

[mooreamps]

I *like* to read about the experiences of people who cobble up a prototype circuit and measure the test performance. So you go, guy. I'm eager to see what you turn up with. After all, proving that easily available data is wrong or fraudulent is always fun, right? Lemme know what you find. I'll be waiting.

Oh, this could be real simple.. I still have a working chassis on my bench... All I need to do is lift a leg on one bias lamp, and tack on a couple of pig-tails from a 1.5 volt cell, just see what it does... What "would" be nice, if I had some way of generating a sound clip of the before and after.... I still need a permission to obtain a replacement video cam, and I'm not seeing that coming not to soon....


Also, the charge rating for a size D cell ; 10,000 milliamp hours. That would translate into about 10,000 hours service for On time at a constant current draw of 1 mill though a preamp tube gain stage....


-g

[R.G.]

Oh, this could be real simple.. I still have a working chassis on my bench... All I need to do is lift a leg on one bias lamp, and tack on a couple of pig-tails from a 1.5 volt cell, just see what it does...

Yeah, it would be real simple if that was the objective. However, the objective is to try out a supercap in a cathode, all by itself, not to try out a battery.

Did I mention that to test a supercap, first you have to obtain a supercap? Try re-reading my post.
Of course, if the let's-just-prototype-it-and-see theory lets you just anything at all instead of what you wanted to test, then I guess you can claim anything and make it all up in the prototype.

While we're at it, I think it was me that mentioned that one might try putting a battery in parallel with a supercap to keep it charged. But that does interfere with the issue of seeing how a supercap sounds with its I-don't-wanna-talk-about-the-frequency-response from the manufacturer, doesn't it?

What "would" be nice, if I had some way of generating a sound clip of the before and after.... I still need a permission to obtain a replacement video cam, and I'm not seeing that coming not to soon....

Most people buy their own or borrow one, instead of asking permission. Or do you admit publicly to working on your private amplifier development using the equipment and supplies of your employer?

Also, the charge rating for a size D cell ; 10,000 milliamp hours. That would translate into about 10,000 hours service for On time at a constant current draw of 1 mill though a preamp tube gain stage....

Oh, yeah. That.
I'm afraid I obscured things a little to see if you'd jump through the hoop.
See, in the circuit with a battery in the circuit, the plate **charges** the battery instead of discharging it. The current goes into the + end of the battery, not out of it. With a relatively constant charging current of 1ma, a non-rechargeable battery will not be too happy in the long run. I'm even not too sure a rechargeable battery would be happy with that for a long time. The battery is there to keep the supercap from needing to be recharged every time the amp is powered up if it's left off for a long time. I've noticed that guitar players want their amps to come on and work when the power switch is turned on, not to wait for 20 minutes while it comes up to speed.
Of course, if I were faced with the problem, I'd go read about it and find out how batteries are likely to deal with long term high voltage charging. But that's just my old see-how-deep-the-water-is-first mentality holding me back again.

You know, I just had a thought. You're a military kind of guy. How does the try-it-first-then-read-about-it approach work with explosives? : - D

See, my mania for research first led me to go look for "self discharge of batteries", where I found this tidbit"

Life of primary batteries
Even if never taken out of the original package, disposable (or "primary") batteries can lose 8 to 20 percent of their original charge every year at a temperature of about 20°–30°C.[54] This is known as the "self discharge" rate and is due to non-current-producing "side" chemical reactions, which occur within the cell even if no load is applied to it.

and

The self-discharge of the supercapacitor is substantially higher than that of the electro-chemical battery. Supercapacitors with an organic electrolyte are affected the most. In 30 to 40 days, the capacity decreases from full charge to 50 percent. In comparison, a nickel-based battery discharges about 10 percent during that time.

So again, the issue is not how long a battery lasts. It's how a supercap sounds, and how long *a supercap* stays charged up without a battery. I suggested the battery to keep the supercap charged, and it looks like it loses half its charge (which has to be made up by the battery) every month or so.

So it may sound fine when you try an amp with a battery in it. But a supercap carries the issue of recharging it to get it to the "sounds fine" voltage, and that carries with it the complexity and expense of putting in a battery, charger, or some other means to keep the supercap up so you don't have to go precharge your amp for a long time to get it to play. And batteries are not what the proto is supposed to test.

Go sit down and think about it for a while.