quick question about el84 cathode caps

[mort]

why the huge range? how do I select a value? drawing says 50-500uF ...

http://www.dreamtone.org/pdf/18watt_Schematic_Rev5.pdf

[tedmich]

it seems to have a subtle filtering effect, with some saying 47uF is a lower limit before bass suffers in the famous Marshall 18watter, see here
The Amp Garage :: View topic - Cathode Bypass Capacitor Options

[pdf64]

I've seen assertions that values larger than necessary (eg which fully bypass the cathode below say 10Hz) can create conditions similar to fixed bias, in that the bias shift during heavy overdrive is reduced.
I've not seen evidence for the above and my experience differs.
My guess is that values in the 1000s uF would be needed to reduce dynamic changes in cathode Vdc level in response to signal level, and that such large values would be likely to cause overdissipation during start up.
When calculating the lower frequency breakpoint, bear in mind that the total impedance to be bypassed (ie the 'R' component of 1/2PiRC) is found by the parallel combination of the cathode resistor and the cathode impedance (=1/gm) of the tube/s.
However, the use of a zener in the cathode circuit can be effective at limiting bias excursion.

BTW that is a very inadvisable method of implementing a standby switch, as the reservoir cap is being 'hot switched' on to a tube rectifier.
Pete

[mac dillard]

From the "Tone Tweaking" charts from the 18 watt forum:"Output tubes cathode bypass filter: The original Marshall 18W's used 500uF, and the Watkins Dominators they derived from used 50uF, there is no right value. Larger values will give more bass response, What is used depends on the type of speakers, cab, OT and even guitar. The -3db low frequency cutoff point is approximately1/(2 x pi x R x C), so for example 125 ohms and 50uF has a lower cutoff of 25 Hz. Try either a 50 or 100uF. Some people find that lower values keep the low end from getting muddy or boomy. This is not a conventional high pass filter in that there is still some gain/ band pass at low frequencies. Instead of a continuous cutoff there is a step in the gain response, lower gain below the cross over point, and higher above."
I have built two versions of this amp and used 150 uF in both..That is what was on the schematics I was using. They sound pretty good to me.

[tedmich]

Be careful as I believe several uF values have been patented by previous 18watt members.... a little forum humor!

[mort]

Thanks ladies and gentlemen

[pdf64]

'The -3db low frequency cutoff point is approximately1/(2 x pi x R x C), so for example 125 ohms and 50uF has a lower cutoff of 25 Hz'
This omits the tubes' cathode impedance from the equation, 1/gm =~88ohms, 2 in parallel giving 44 ohms, resulting in a total cathode impedance to by bypassed of 88//88//124 = 32.5 ohms.
Pete