Value of coupling caps

[john dozier]

I have a schematic for an amp that uses 6sn7 drivers for 4-El34 outputs. The coupling caps between these stages are 1uf Can I reduce this to .22 or perhaps .47 without affecting performance. Thanks for the help. John Dozier

 

[Loren42]

I have a schematic for an amp that uses 6sn7 drivers for 4-El34 outputs. The coupling caps between these stages are 1uf Can I reduce this to .22 or perhaps .47 without affecting performance. Thanks for the help. John Dozier

Or parallel two .47 µF caps...

 

[SY]

I have a schematic for an amp that uses 6sn7 drivers for 4-El34 outputs. The coupling caps between these stages are 1uf Can I reduce this to .22 or perhaps .47 without affecting performance. Thanks for the help. John Dozier

That can't really be answered without looking at the schematic. Can you link to it?

 

[DF96]

A lot could depend on whether there is global NFB, and where the OPT LF rolloff starts. It is easy to create or solve LF stability problems by changing coupling cap values.

Why do you want to reduce them?

 

[john dozier]

High quality caps in 1uf are hard to find and expensive. I cannot link to the schematic as I do not have it on computer and do not have a scanner. I can calculate the cut off of the RC network though if that would be helpful. Thanks

 

[john dozier]

There is a 6.8k resistor in series with each coupling cap. This seems to give a cutoff around 20hz. Check my math please! thanks

 

[DF96]

The 6.8k resistors are grid stoppers. They don't affect LF. The relevant resistance for LF is the grid leak (to ground, or the bias supply). Probably in the region 200-500K.

If you reduce the capacitors you might simply get a bit less LF. Or, because of the increased phase shift, you might get an LF peak around 10-20Hz. With digital sources this does little harm. If you listen to LP then you may get cone flapping. Mains voltage variations can also do this. We can't tell you the answer, because it depends on all the other LF poles in the circuit too.

 

[john dozier]

I appreciate all the help. My knowledge of tube theory is limited. Since high quality 1uf are usually limited to 200v, I think I will use 2-,47uf at 600v, bypassed with .1 and .01 600 v. I hope I can fit them in the chassis I have in mind. thanks again

 

[mastodon]

0.22uF will generally be ok for a decent LF response, assuming something from 100k upwards has been used for grid stoppers of the power tubes. Bypassing the coupling caps with yet smaller values makes very little sense IMHO, but whatever floats your boat; it's a free world. Expect very little improvement with values higher than .47uF.

 

[john dozier]

The grid stopper is 100k. I think perhaps a .47 600v might be a good value. This way I can use a high quality cap, have it fit, and do away with the bypasses. What say You? Thanks again

 

[mastodon]

Sounds like a plan. I'm a cheapskate personally, so I always use cheap Wima coupling caps, but if you're into the whole audiophile thing, then I think you're on the right track.

 

[john dozier]

The Audiocap Theta (relcap) is almost as good as teflon and only about $8 a pop. I think that is very cost effective as teflons of the same value would be ten times that. And yes, you can hear the difference if your system has enough resolving power. Again, many thanks

 

[jeffreybehr]

John, did I miss that you know the value of the load resistor? That's what determines the hi-pass filterpoint.

Lots of us feel that in a sand amp that's otherwise direct-coupled, a FP somewhere in the 2- or 3Hz range range is plenty low enough, and for a tubed amp that has other HP filters, c. 5Hz is low enough.

Connect a DMM switched to DCV to the cap that's there and turn on the cold amp. It's the highest DCV you see there that you have to exceed with the rating of the cap. 600 sounds mighty high to me.

And consider the MultiCap RTX 'styrenes; they sound quite good and aren't priced up with the Teflons, etc.

http://www.soniccraft.com/multicap_rtx.htm

 

[john dozier]

Its 100k. I know about styrenes and love them except when you have the heat of 8 el34s to contend with. I think polyprop is a better choice in this and the thetas sound just as good as styrenes. Will use teflons in the front end and the thetas for output coupling. Should be a nice matchup. BTW, I have been using teflons in audio for twenty years. I know my way around capacitor quality.

 

[tomchr]

AES (Antique Electronic Supply) has reasonably priced caps. I use the Solen polypropylene ones... 1 uF is a little over $2/EA. You probably won't need that many...

Antique Electronic Supply

The LF cutoff is set by the coupling cap and any resistance to AC ground. Usually, you'll find a 100 kOhm to 1 Mohm grid leak resistor. Some tubes have more grid leakage than others, hence requires a lower resistance grid leak resistor. You can calculate the LF cutoff as

f-3dB = 1/(2*pi*R*C)

R in ohm, C in Farad, f-3dB in Hz.

Changing R or C will affect circuit performance. The question is if this change is acceptable to you or not.

~Tom