cathode bypass cap arrangement

[ee]

I have tinkered around a bit now with the cathode bypass caps and would like an opinion. I installed Mundorf EVO Al/oil 100uf cap to replace the cathode bypass cap (220uf). I am considering getting a couple more of these 100uf and wiring parallel. Space be damned is it good idea bad idea to parallel in that position or do it and see what happens? Thanks
Eric

[Daikini]

If I remember correctly, the 220 uF bypass cap value is used because it's a part that's used in lots of kits. You can get a way with lower values. 

I think 100 uF is plenty.

I have a 47 uF Solen in that position in my Stereomour, and it works fine.

See this thread: http://bottlehead.com/smf/index.php?topic=5499

[Paul Birkeland]

Sorry, this doesn't answer your question.  Back in the Paramour days Bottleheads were using a charged 9V battery for the cathode bias.  It replaced the resistor and capacitor.

I think, someone correct me if I'm wrong, this should also work for the Stereomour.

This is an obscure reference to using various 5 pin driver tubes and how to bias them in a different product, and hasn't been relevant for many years now.

In the original Paramour monoblocks, the cathode bypass cap was 100uF.  PJ generally recommends 1/GM as a starting cap value (in micromhos, this yields microfarads, which is convenient).  This recommendation suggests that 200uF is a good starting value.

[howardnair]

I also used a 47uf  in my stereomour -it is a mundorf M-tube cap

[Paul Joppa]

Grainger, I remember (now that you remind me...) the battery bias on drivers - works like LEds in that the bias is fixed. The battery is on trickle charge from the cathode current, so it's important to understand the battery's limitations!

For the 2A3 however, the fixed bias voltage is a problem as the tube ages - it will change the current too much for consistent operation. That's the real reason for using a resistor, it's self-compensating.

The capacitance should have a low impedance relative to the intrinsic cathode impedance which is 1/gm; this should be the case at some low frequency - I use 5Hz unless I need to cut it close. That works out to

C = gm/(2*pi*f)

or, at 5Hz, C = gm/31.4

That's 159 uF if gm is 0.005 mA/v (5000 micromhos). It's 40uF at 20Hz (the resonance of the plate choke/parafeed capacitor in Stereomour) which is at least an arguably good point to choose. I used 47uF in Paramount.