Current loops in valve amps

[acorn]

Hi all

I have just read Lynn Olson,s artical with interest, regarding current loops in valve amps output stages, you can read it at www.nutshellhifi.com.

I would like to know if I decided to implement this suggestion , I would fix a 5uF to 20uF capacitor between the B+ cap and the cathode bias of the output valve, but would you also have to leave the original cathode by pass cap out, and what effect would this mod have on the sound?.

I presume this would apply to both output valves, thus by doing so isolating the input loop from the output loop, therefore letting the B+ cap to do the main job of filtering the ripple from the power supply. It raises the question would this kind of mod be any more benificial than a Poly cap strapped across the B+ cap down to the negative rail.

Regards Acorn

[NealG]

I talked to James about this at Eggfest as I want to try it on my WAD 4wse 6080 based amp.

recommendation is to connect from B+ to top of cathode R as you say but leave in the original bypass cap in. The addition on the ultrapath cap injects some supply noise back in to the valve just where you don't want it, the original cap shunts it to ground, so use a good quality cap. like a BG.

The ultrapath cap is calculated as original bypass cap value divided by Mu+1 of the o/p valve. Use a good quality PP cap. with the correct voltage rating IE somewhat more than B+.

I've gone for Solen for now, low cost so if I don't get on with it I haven't lost out too much.

[acorn]

Hi NealG

I assume you will need two caps one for each valve, 4 x for Push - Pull. I wounder what affect this mod will have on the sound, I will be interested to know how your amp sounds after the mod, as you say I will have to fit some suitable voltage rated PP caps. Has any one else contamplated trying this mod?.

Regards Acorn

[James D]

Ahh the SE and the PP application are different - they act in different ways. Neal has described the application to SE amps where the WE/ultrapath connection does two things:

1) It provides a very short loop for the output current return through a high quality cap that bypasses the psu.

2) It provides psu noise reduction via the mechanisim Neal has outlined above.

In PP the situation is quite different - the output current loop does not involve the psu - it is already short and defined by two valves and OPT. Note it doesn't involve the cathode resistor so the cathode bypass is not required in a PP amp. I'll say that again - the cathode bypass is not required in a PP amp. In PP the bypass that is required is from B+ (CT of the OPT) to the joined cathodes of the output valves - here it supports the return loop for out of balance current and in doing so prevents this running through the psu and making the sound worse.

SO to recap. In SE use the WE/ultrapath connection and use two caps one as cathode bypass the other from cathode to B+. In PP NO CATHODE BYPASS and one cap from cathodes to B+.


James

[NealG]

Cheers James!

I think I'll try that on my PP 300B as well. Is the cap calculated in the same way?

[James D]

Not quite the same way, it is bypassing the parallel combination of the two halves of the OPT loop circuit and each half is the series combination of valve output impedance and one side of the OPT primary impedance (this is Raa/2 for a Class A PP amp and Raa/4 for any other Class).

James

[Greg.]

Hi Neal,

Looking forward to a more detailed explaination for the 300B PP Also I owe you some money etc. I'll email you.

Best wishes,

Greg

[Ianm2]

I think I saw somewhere approx. 1/3 the value of the cathode cap. for S.E.

[James D]

Ian,

Depends on the gain of the output stage as Neal quotes above.

James