Cathode Bypass capacitor voltage rating

[rafafredd]

I am building a tube amp that has only a few volts bias (like 1.5v DC or so) on the cathodes, bypassed with a big lytic cap.

Can I use a capacitor with very low voltage rating (like 4v or 6.3v) for bypass, or should I consider also the AC voltage across it?

I ask because I see all the time people using 25v and 35v for cathode Bypass, but I though that going lower wouldn't really be a Problem. Is it right?

On the other hand, on the output Stage, if the cathode cap wasn't there, it would develope up to 10v - 15v AC Peak to Peak on the cathodes before clipping. So, should this AC voltage be considered for the voltage rating when the cathode cap is in place?

I though I could just ignore it, as the voltage isn't really there (or is it???) when the cathode is bypassed, and consider only the very low bias voltage, and then using 6.3v Bypass capacitor, that are smaller and cheaper than 25v or 35v in the same capacitance value.

Any thoughts???

 

[abbey road d enfer]

I am building a tube amp that has only a few volts bias (like 1.5v DC or so) on the cathodes, bypassed with a big lytic cap.

Can I use a capacitor with very low voltage rating (like 4v or 6.3v) for bypass, or should I consider also the AC voltage across it?

I ask because I see all the time people using 25v and 35v for cathode Bypass, but I though that going lower wouldn't really be a Problem. Is it right?

On the other hand, on the output Stage, if the cathode cap wasn't there, it would develope up to 10v - 15v AC Peak to Peak on the cathodes before clipping. So, should this AC voltage be considered for the voltage rating when the cathode cap is in place?

I though I could just ignore it, as the voltage isn't really there (or is it???)

Indeed, within the intended frequency range, the voltage is NOT THERE, but in case of infrasonics, there may be some voltage. For preamp stages, the voltage does not vary much. When the great tube amps were designed, there was no low-voltage lytics, they were all about 25V, so now, It's considered as gospel, but really, on a 12AX7 cathode, you could use a 3V cap with no problem. For the output stage, there is a voltage variation, particularly in class-AB; that's why you need a 25 or 35V type.

 

[rafafredd]

Thanks for your repply, man.

What if the output Stage is single ended class A? Then, could I just use the low voltage caps???

I really need to know, because I designed the BOARDS, and they are here, PRINTED already!

Now, after I got the boards, I had gotten unsure about the caps voltage...

So, if I really need higher voltage, I'll have to make the uF value lower, and loose some of the low end. Otherwise, I could mantain the same value, and mount the caps outboard. Would work fine, but not look as good.

If I could use the low voltage ones without worrying about premature failures, and with no mods to the circuit or boards, then that would be NICE INDEED!

 

[emrr]

There seem to be thoughts around potential differing amounts of DC leakage based on voltage rating, and what effects may occur.  I can't speak to that to either confirm or deny.....

 

[abbey road d enfer]

Thanks for your repply, man.

What if the output Stage is single ended class A? 

Depends if it's run clean or overdriven. If clean, the voltage almost doesn't change. If overdriven, the voltage can almost double.

 

[PRR]

> should I consider also the AC voltage across it?

Self-bias cathode resistors, the "AC" can not be much larger than the DC.

In general, the cathode cap voltage does not need to be much over (B+) divided by Mu.

This means 12AX7 up to 400V supply, 4V is all you will find. More likely 0.6 times that, or 2.4V. A 10V cap seems ample.

Figure power tubes as Mu(g2) about 10. So 40V cap will do for 400V supply. (We should not be cathode-biasing for very high B+ because the optimum load impedance becomes too high to wind.)

The basic electrolytic layer is "naturally 450V". Into the 1950s, low-volt e-caps were less common, and rarely under 25V. With the explosion in pocket transistor radios and such, low-low-Volt e-caps did become readily available. However Tube Amp Supply Shops tend to stock higher-V values, because there is always (should be!) room for a larger cap, and a high-V cap replaces a low-V cap, saving inventory costs.

 

[CJ]

that 100K plate resistor will limit current ( and thus voltage) on the cathode resistor,

so 400 v / 100k = 4 ma,    4 ma x 2K cathode resistor = 8 volts, 

so even in the highly unusual event of a 12AX7a shorting plate the cathode, the most V on the cap will be 8 volts,

some folks believe there actually may be sonic benefits to running caps close to the voltage they see,

but cathode caps?  probably not an issue,

 

[joaquins]

  Do you have a schematic that could show us, of that last stage at least... If not is very hard to know, for your specific stage, if your cap would work or not.

  Not knowing I'd said use a higher voltage cap, I'd probably go for the outboard option, it's not very important the looks inside of an amp, I wouldn't sacrifice low end for looks. Now, how much the low end is affected by reducing that vaule, is other story, for a guitar amp makes no sense account for 20Hz, if you look the size of the OT in the fender champ you'll know what I'm saying.

JS

 

[rafafredd]

Thank you a lot for talking the time to help me.

Here is the output stage and the curves, for reference. This is plain simple, no feedback, with a switch that connects G2 to plate for TRIODE operation.

I know 10k transformer load seems hi, but that's what I have on hand, and I really don't need Hi Power.

In fact, low Power is a plus on this one, and I like to load the output softly. Sounds good!

But I use big caps, for all low end I can get, and a little more

The 4700uF in the cathode is the object of this thread:

Will 6.3v do?

 

[rafafredd]

Circuit:

 

[CJ]

40 ma thru 68ohms generates 2.72 volts so i think you are fine with 6.3

 

[rafafredd]

Thanks, CJ.

That was my though.

Sorry for the large pic. I did it all in the cellphone, and didn't remembered to realize it.

Going small cap route. Just hope it will stand the test of time!

 

[PRR]

Equivalent cathode impedance is like 48 Ohms (68r||166r) (Gm~~6mS).

For 20Hz I would bypass with 200uFd.

6,800uFd looks like under 1Hz.

If you are going that low, the screen cap needs to be considered also.

Cathode cap voltage: if you could have 6.3V across 68 Ohms, that would be 90mA current. The poor tube is rated 40mA. The cap will not blow first. The normal working voltage will be around 1.5V. Unless you expect bad things to happen, or your empire will fall if this cap fails, 6.3V seems commercially safe and 3V may be fine.

You do not lose a whole lot of gain (3dB) if you don't even bypass that cap.

 

[rankot]

Equivalent cathode impedance is like 48 Ohms (68r||166r) (Gm~~6mS).

For 20Hz I would bypass with 200uFd.

6,800uFd looks like under 1Hz.

If you are going that low, the screen cap needs to be considered also.

Cathode cap voltage: if you could have 6.3V across 68 Ohms, that would be 90mA current. The poor tube is rated 40mA. The cap will not blow first. The normal working voltage will be around 1.5V. Unless you expect bad things to happen, or your empire will fall if this cap fails, 6.3V seems commercially safe and 3V may be fine.

You do not lose a whole lot of gain (3dB) if you don't even bypass that cap.

What happens if tube blows? Is there any chance that B+ will hit cathode bypass capacitor?

 

[abbey road d enfer]

What happens if tube blows? Is there any chance that B+ will hit cathode bypass capacitor?

Wow! Return of the living thread
Well, if the tube goes short (which isn't very likely) current is limited by the anode resistor. You may want to take that into account.
In the case of inductive plate load (xfmr or parafeed) the transformer or choke may die before the cathode capacitor. Anyway, it would be a significant event that would require complete verification and overhaul.
In the case of the specific circuit in thread #13, even if both anode and screen were to short to cathode, the voltage would be limited to about 3.5V, so a 6.3V cap would be OK.

 

[rankot]

Wow! Return of the living thread
Well, if the tube goes short (which isn't very likely) current is limited by the anode resistor. You may want to take that into account.
In the case of inductive plate load (xfmr or parafeed) the transformer or choke may die before the cathode capacitor. Anyway, it would be a significant event that would require complete verification and overhaul.
In the case of the specific circuit in thread #13, even if both anode and screen were to short to cathode, the voltage would be limited to about 3.5V, so a 6.3V cap would be OK.

Thanks!