Cathode Bypass caps - how to choose?

[JesseG]

Hi all:
Christmas just a week away and I get a little time to blow up some parts

I expect that I am like many in this crazy hobby - I know just enough about the whole thing to get myself in trouble however not enough to say I know what I am doing...

One thing that has me perplexed - how to choose a "good" cathode bypass cap. I know that these caps appear in the signal path and they are usually 100uf or better, although fairly low voltage (most of the time). I have bins full of the little blue and black cheapies rescued from many a dead pc board, but I just know that these are not good-sounding caps.

Trying to find a 220uf Solen or Auricap is crazy-making AND expensive and the size of these things could require a seperate chassis. So, could one (or more) of you gurus out there help a bit and fill me in on how you select appropriate cathode bypass caps?


PS: Merry Christmas everyone!

[kmaier]

Take a look here... Sidebar A, would also recommend reading the entire article:

http://www.wavelengthaudio.com/bugle.pdf

I use all Axon caps in my SET amps... 630V HighVolt in the power supply, 250V TrueCaps for all bypass and 630V foil/film for signal coupling.

Regards, KM

[SY]

how to choose a "good" cathode bypass cap

No such thing. Besides the defects of big electrolytics, cap bypasses also degrade overload recovery. Once I started using diode bias, I never looked back.

[Miles Prower]

Blackgates are supposed to be pretty good. Though, up till now, I've been able to avoid using cathode bypass capacitors. If you use these, they still need to be bypassed with a 0.1uF capacitor, and make it a good one like an AuriCap or a Solen.

If you have a convenient source of negative DC, you can use fixed bias on your small signal VTs. If you have gain to burn, you can simply forego bypassing and degenerate the cathode at the expense of some lost gain.

[Stixx]

As long as you are within the voltage limits I've used Os-Con with very good results...even bypassed with mica

[inertial]

Jesseg,

I think 10 in parallel of these .........
http://www.wima.de/DE/mks2xl.htm

Cheers
Paolo

[Eli Duttman]

With all respect to the other posters, decent 'lytics are available at modest expense. Panasonic and Nichicon offer low impedance, 105^o C, parts, without "sticker shock". Check the spec's out. These caps. exhibit satisfactory behavior up into the low RF range.

[sharpi31]

I had better results with conductive polymer caps than black gate (standard) in my Spud amp.

[JesseG]

Thanks everyone!

So this is not one of those "right way"things, but trial and error plus some personal preference - cool! I like stuff like that.

kamier - thanks for the link - the Bugle 45 is almost exactly the same schematic as my EL34 SE.

Sy - can you elaborate further on diode bias?

Eli - I agree that most of the Panasonic and Nichicon are good, as 'lytics go, but after trying my EL34 with them in, I removed them, accepted the drop in gain and like the sound much better.

sharpi31 - what are conductive polymer caps, please?

Happy Ho-Ho

[SY]

Sure. A forward biased diode has a low dynamic impedance and a very constant (and predictable) voltage drop. So you can use that in a cathode circuit in place of the cathode resistor to bias the tube. Unlike the bias resistor/bypass cap combination, overloading the stage will not cause blocking distortion.

You can use different sorts of diodes to get different amounts of bias. A silicon diode will give 0.6-0.7V, an IR LED will give about 1.2V, a red LED about 1.7V, a green LED about 2.0V. The LEDs, besides being excellent voltage sources (and very quiet) also give a nice visual indication that current is running through the stage.