Setup & Implementation of G Pimm's Self-Bias CCS

[danFrank]

A while back somebody posted on this forum how they setup and implemented G. Pimm's self bias CCS in their circuit.
How to choose resistance values, etc. I think they posted that they started with a 22K resistor and went from there.
Could somebody post me a quick link to that, if possible. I've been looking with no luck.
Thanks, Daniel

[kopite]

If you build it exactly to Gary's design the reference voltage is 14v. (approx)
Use this voltage divided by the current to select R1.
If the current you want is 20mA then use a 680 ohm.
Adjust the pot to give you a 13.6v drop across the 680 Ohm.

[jeapel]

hi

for my 6sn7 preamp test at 5ma bias :

1- c4s ccs like bottlehead

2-self bias normal ccs and mu-follower output

G.P. normal ccs is good one and mu-follower give
very good bandwidth because low zout to attack
next stage

but to my ears and only to me :-) i prefer the sound
of c4s ccs it s more soft sound

bye

[jeapel]

@ 5ma no heatsink on 2n2907 but small heatsink on mje350

[Brian Beck]

Sorry if I sound like a broken record here:

Pimm’s CCSs were an improvement over other kinds of CCS designs since he was able to greatly reduce shunt capacitance. His tests showed only 0.08pF of shunt C in his self-bias design, and even lower in some of his more elaborate designs. The C4S (as I understand it) and other simple cascoded variants expose the load to the full Ccb of the bottom PNP, here an MJE350. This transistor’s base is shunted to ground at high frequencies by the capacitors (and sometimes LEDs) in the divider path to B+, which is itself shunted to ground via filter caps. So Ccb is shunted to ground by this path, and it directly loads the plate. My MJE350 data sheets don’t specify Ccb, but it will be a lot higher than 0.08pF and it will be variable with collector voltage for sure, the real problem. I think that capacitance modulation is one the main reasons why CCSs can sound so different from one another, even when the capacitances create RC time constants that would suggest no effect within the audio band. Just my opinion…

[jeapel]

hi Brian

my frequency test give : 6sn7 with ccs + 6as7 cathode follower

fc -3db = 90Khz with c4s

fc -3db = 110Khz with G.P. ccs

lower shunt capacitance for G.P. but not so bad for c4s

like you say the sound signature is intriguing from
ccs to another

[Brian Beck]

jeapel,

That's good data, and decent performance. I think you're seeing the effect of additional capacitance that exceeds the CCS capacitance, such as inter-element capacitances, stray capacitances and the input capacitance of the 6AS7 cathode follower. Your source impedance matters greatly too. That's OK because these other capacitances are usually not as non-linear as the smaller CCS shunt capacitances (although they might be microphonic). Even when these other sources of capacitance are present, a smaller, but modulated, capacitance can have an imprint on phase shift behavior, IMO.

[jeapel]

ok 6as7 input cap explain the test result

"CCS shunt capacitances (although they might be microphonic)"

maybe it s my situation because my tube preamp sound is
very good but microphonic just a small finger tap is on
speaker

how a "transistor" shunt cap can be microphonic ?

[Brian Beck]

I wasn't including the transistor capacitance in my comment about microphonics. Transistors can be slightly microphonic, but usually much less so than a tube.

[jeapel]

sorry Brian my english is so so

if i understand a 6sn7 ccs load (very hight AC resistor) is

no more microphonic than 6sn7 with 47kohm load or

i m wrong