Another cap replacement: uF changes..

[cuibono]

I'm building new boards for some monoblocks I built, and am wondering if I can use .1uF caps instead of .2uF that the schematic calls for. These are used for interstage coupling, keeping the B+ off the grids of the output tubes (all tubes being 6BM8's).

The reason is I'd like to use some russian teflon caps, and would be easier to get only .1 uF caps, which are used in other stages.

I'm also wondering if it is okay use even larger values, say up to several uF - increasing the size of coupling caps has worked nicely in other projects (non tube)..

Thanks for helping a newbie.

[Miles Prower]

Originally posted by cuibono
I'm building new boards for some monoblocks I built, and am wondering if I can use .1uF caps instead of .2uF that the schematic calls for. These are used for interstage coupling, keeping the B+ off the grids of the output tubes (all tubes being 6BM8's).

Hard to say anything definitive about this without a schematic. If the monoblocks incorporate gNFB, using the smaller coupling capacitors probably won't make much difference, but be aware that there will be less feedback at the lower frequencies, hence higher Zo, and possibly sloppy bass as a result of the underdamping. If it's an open loop design, you're probably sacrificing something at the low end. Whether or not those Russian teflons make up for it sonically is purely a subjective call.

I'm also wondering if it is okay use even larger values, say up to several uF - increasing the size of coupling caps has worked nicely in other projects (non tube)..

Again, is it a gNFB design or not? Making the capacitors that much larger might upset your phase margins at the low end that could lead to low frequency or even subsonic oscillations. Even if you don't get outright oscillation, you could end up with sloppy bass due to a rising amplitude characteristic at the low end. That, too, is a phase margin problem.

Open loop or closed loop, making the coupling capacitors that large may lead to problems with blocking distortion if the grids are driven more positive than the cathode.

[DigitalJunkie]

several uf? That might be asking for trouble.
I'd try maybe 0.47uf first,even that might bring you close to motorboating/instability,depending on the design.

Remember,tubes are high impedance,even 1uf is a pretty large value for coupling tube circuits.

[oshifis]

I replaced the WIMA MKS-4 1 uF 400 V coupling capacitors with Russian FT-3 0.22 uF 600 V teflon capacitors in my tube amplifier. It all depends on the grid resistor (100 k in my case). It sounds excellent, no low freqency rolloff can be heard or measured. It is an open-loop design, though.

[Sherman]

How about just paralleling a couple of the .1uF caps to get .2uF?

[cuibono]

Ahh, so much good information. The design does use global NFB, and the grid resistors are 330k. I could parallel the .1uf caps to get .2uf, but then I'd need 10 caps, and they are sold in packs of 8. I suppose I could work something out with the seller.

In fact, I'll probably go that route, it sounds easiest.

Thanks much

[Colt45]

330k / 0,1µF cutoff is 4.82 Hz... plenty low for me.

[oshifis]

Originally posted by cuibono
Ahh, so much good information. The design does use global NFB, and the grid resistors are 330k. I could parallel the .1uf caps to get .2uf, but then I'd need 10 caps, and they are sold in packs of 8. I suppose I could work something out with the seller.

In fact, I'll probably go that route, it sounds easiest.

Thanks much

When I received my caps (could be from the same guy in Lithuania?) I was surprised by their size. They are BIG.
BTW if you use fixed bias like I do, 330k is a bit on the high side. I would not go below 150k. On the other hand, if it is auto-bias using a cathode resistor (which permits higher grid resistor), you have an electrolytic cap parallel and that destroys the quality improvement gained from the teflon coupling caps. I tried both, and ended at fixed bias.
I also read somewhere that teflon caps are perceived to go deeper than polystyrene caps of the same value.

[cuibono]

Originally posted by Colt45
330k / 0,1µF cutoff is 4.82 Hz... plenty low for me.

so no oscillations?

oshifis, yes the caps are big, so its good timing I'm redesigning the boards. the circuit is fixed bias, and was designed by a friend I consider an expert, although a minimalist. Unfortunately, we aren't in contact anymore...