Cathode bias resistor value for 6SN7?

[kegger]

Curious what people have used or better yet what is the smallest value someone
would use in this spot to try and get some decent results.

I'm currently simming in multisim an amp circuit and using a 620 ohm one now.
I know the general concensus seems to be 1K but I'm limited here to a B+ of 315
and at 1K my current is pretty low, I'm not expecting to get perfect #''s with this
setup but I'm trying to get the best I can. I'm also using 27k plate resistors with
currently getting about 5ma current, 180 plate voltage and around -6 on the grids.

I really didn't want to go any lower on the resistor as I didn't want to drop my plate
voltage much below 180. I could lower the plate resistors to get more but did kinda
want to keep them up there as far as I could?
This is the differential section of a PPP 6V6 amp with a cathodyne 6J5 up front.

Anyways just throwing this out there to get others oppinions on this setup then to
decide from there where to go. I'm just not sure how the 6SN7 will run in this conf.

Thanks in advance for any and all types of responses.
Also if anyone would want any multisims run I'd be more then happy to help whatever you got.

 

[kegger]

At 180 plate voltage, 5ma current draw and -6 on the grids I'm in the curves for a 6SN7.
I just don't have a lot of experience building circuits with this tube.

Also Both halves of the dual triodes in the 6SN7 are tied together at the cathode.
Don't know if that will make a difference, would that halve the noramal value so 500?

 

[dsavitsk]

I used a couple of red LEDs in series on the cathode (of a 6j5) to bias it to about 4V. I have about 260V on the plate and this is working quite well as a line stage. I suppose I should add another one or two, but I haven't heard any clipping yet, so I'll probably leave it be for a while.

You might want to look here as well: http://boozhoundlabs.com/howto/

 

[ray_moth]

Your arrangement seems like a reasonable compromise, although you might like to try using a higher value cathode resistor, to reduce the plate current of each tube to 4mA, and increasing the plate load resistors to 33k. This would still give you 180 V on the plate and maybe better linearity. (I agree with you that it's best not to go any lower than 180 V plate voltage.)

You are correct that sharing the cathode resistor between two tubes means that you have to halve the ohmic value that would be used for a single tube at the same operating point.

The 6SN7 is an annoying tube, in that it ideally likes a plate-cathode voltage of 250 V or more, a plate current of 8 mA or more and a load of 47k. This means that to get the most out of it, you would need B+ of around 600 V!

Direct coupling,especially, demands significant compromise in the operating point (not your situation, I realise). Most designers, such as Williamson, have tended to choose a far from ideal operating point because of this.

 

[kegger]

Thanks, both of you.
I'm reading through the site mentioned now.

Yes with 33k on the plates and an 820 cathode resistor I get 4.01ma current draw and 181 on the plates.

No problem running that low on the current aye? Yes I agree here I'm voltage starved to begin with.

Nice to be confirmed on one of your assumptions, meaning the plate voltage, and the halving of the resistor.
But I do wonder about the current as well.

I've read yes that the 6SN7 does like current as well, but alas yu gotta po what yu gotta do.

Thanks, again!

 

[Miles Prower]

The 6SN7 is an annoying tube, in that it ideally likes a plate-cathode voltage of 250 V or more, a plate current of 8 mA or more and a load of 47k. This means that to get the most out of it, you would need B+ of around 600 V!

Direct coupling,especially, demands significant compromise in the operating point (not your situation, I realise). Most designers, such as Williamson, have tended to choose a far from ideal operating point because of this.

If the 6J5 is actually half of a 6SN7 (and it would appear that it is) then this isn't true. I did a design using a 6J5 that set a Q-Point current of 1.0mA, with a Vpk~= 52Vdc. With Vgk= -2.0Vdc, and Rp= 100K, it worked just great. That allowed DC coupling without busting any heater/cathode ratings.

The estimated THD is quite low, and the gain a bit on the high side for my purpose. So I figured that an unbypassed cathode resistor would be better, and that gets the THD down even farther. Actual performance was almost exactly as predicted.

 

[kegger]

Yes a 6J5 is half a 6SN7 and they should react the same in circuit.

One of the things I've heard with these tubes is that if you run them on the lean side they tend to be micrphonic.
(Just what I've read, no first hand experience here)

It's nice to see others having good results with low settings, makes me feel comfortable about how mine will run.

My 6J5 is running a little lean as well but the voltage is up there.

 

[Miles Prower]

One of the things I've heard with these tubes is that if you run them on the lean side they tend to be micrphonic.
(Just what I've read, no first hand experience here)

Mine are Westinghouse JAN NOS 6J5s. If you give them a really good whack with a pencil, there is an ever so slight disturbance while o'scoping. Not bad at all, and hardly worth worrying about. In actual operation, it's of no consequence at all. Now, I've heard that a distressingly large number of US made 6SN7s are terribly microphonic regardless of Q-Point. Then, again, I wouldn't expect something like that from anything built to military specs.

 

[kegger]

Sounds good!

 

[arnoldc]

have you considered choke loading the plate (or CCS) and if you can run at 8mA or so, an LED could work.

 

[kegger]

I have looked at both those options as they would be nice additions, but 2 things.

I was trying to just do a simple arrangement on these mono blocks to get em up and runnin and I am
a little constrained for space as it is.

But those are definatly 2 great ideas.

I may do a CCS to my modded ST-70 which runs an LTP. I'm thinking of building a voltage doubler off
of the 55 volt bias tap which would give me in the neighborhood of 135volt, or so say's PSUDII anyway.

And a 135volt CCS just squeeks out fine on an 18K tail in that circuit, so it just may workout nice.
I run a 12bh7 as a driver (Split to feed both channels) to a 12au7 on each for the LTP on that one..