in that case, i may just use a 500 ohm 50watt metal clad Dales bypassed with 470ufd/100volt caps in place of the ccs...
it is to bias the output tubes, any other scheme can do...
i am hoping to catch the essence of the design by using CCS at the output cathodes...
i am hoping to catch the essence of the design by using CCS at the output cathodes...
A bit more than just bias the output tubes, it ensures full differential operation. Self-bias isn't going to do the trick AFAIK.
In fact, if the common CCS exists in first/driver stage it does not matter much for following output stage to have one, proven with experiments. If so, the common output stage CCS will add a very little improvement but the drawback is that if one shoulder’s tube is missing/burned by accident - the second one will receive the double current and (un)happily pass to another (better) world quiet quick.
So I would advise having the separate (2 x) CCSes on output stage if there is already a common CCS in e.g. phase-splitter. Two CCS on output will better control the big bottles plus you do not have to worry about ‘em be matched.
The interesting IMO exception is the common choke in output stage, I just like the idea from "purist"s point of view, but the matching is the MUST in this case.
All the above greatly applies for the differential input case, and, in fact, for running the Amp via balanced XLR with using the separate CCSes for driver AND for output stage you do not bother with matching of any. You get king of great convenience: use any tube you want and as soon as it fits into the operating point limits - you are good, great for doing rolling.
Last edited:
I would love to see your experimental results showing there is no benefit having the CCS on the output, since Pimm put it there on purpose.
Differential input/Differential output stages
My knowledge on this is limited, but I would like to put the following to discussion; A tail CCS is supposed to ensure AC balance even with severely unmatched tubes. This at the cost of DC imbalance even with -real world's- matched tubes. At the input stage what matters mostly is AC balance. DC imbalance can be faced with coupling capacitors. At the output stage first concideration is DC balance otherwise the OPT will be damaged. AC imbalance will be "fixed" with the differential to SE conversion at the OPT. A common CCS is very unlikely to fulfill this requirements. Separate CCS for each tube should allow DC cancelation but most probably we are back to AC imbalance. Typical fixed bias should also do that fine. One thing that I find interesting is that a tail CCS -AC balance- theoriticaly converts the stage to a constant load as seen from the psu meaning less interaction. One CCS common for both cathodes should do that. Can separate CCS do it? And how about simple fixed bias compined with an OPT of high enough(?) inductance?
My knowledge on this is limited, but I would like to put the following to discussion; A tail CCS is supposed to ensure AC balance even with severely unmatched tubes. This at the cost of DC imbalance even with -real world's- matched tubes. At the input stage what matters mostly is AC balance. DC imbalance can be faced with coupling capacitors. At the output stage first concideration is DC balance otherwise the OPT will be damaged. AC imbalance will be "fixed" with the differential to SE conversion at the OPT. A common CCS is very unlikely to fulfill this requirements. Separate CCS for each tube should allow DC cancelation but most probably we are back to AC imbalance. Typical fixed bias should also do that fine. One thing that I find interesting is that a tail CCS -AC balance- theoriticaly converts the stage to a constant load as seen from the psu meaning less interaction. One CCS common for both cathodes should do that. Can separate CCS do it? And how about simple fixed bias compined with an OPT of high enough(?) inductance?
Last edited:
They didn't have high performance CCS's back then - you can bet your bottom dollar that if they had they would have used them.
Without it I can guarantee that Gary Pimm would never have designed this circuit. The design philosphy is using the best elements to allow the tube's to do what they do best. I also doubt that the absence of a CCS would be as insignificant as some might suggest since the output stage is the main contributor to the overall distortion profile compared to just about any kind of input stage.
With regard to separate output stage CCS's, if you bypass them with caps between the OT cathodes you get both DC balance and enforced AC balance. It really represents the best of both worlds. A single CCS approach works equally as well but requires constant attention to bias balancing through some other mechanism such as independent screen adjustment.
The penalty, from my experience, with separate CCS's is that once they fall sufficiently out of balance the circuit will tend to enter oscillation and this is likely well before the tubes have reached a full service life.
Shoog
Last edited:
I agree thats a good approach - but is it superior to screen adjustment as in the original design ?
Shoog
Screen adjustment should be cheaper since you don't need a high power wire wound pot like you would in the cathode's.
Shoog
Yes, the darn WW pots are kinda expensive, but it does use less parts than two CCS's with the associated circuitry.
I would go with two simple TL783's or two of the IXY chips, I bet they would work out cheaper than the WW pot.
Shoog
Please help me to understand this. Does it need ideally matched tubes to work?
I think that for any reasonable pot value the differential function is degrated. Perhaps a bypass cap should be used here also?
Edit; How about a combination of a common CCS and fixed bias?
Last edited:
- Status
- Not open for further replies.