Positive bias for tube CF?

[pftrvlr]

I took apart my headphone amp and measured the voltages of the 6SN7 cathode follower. I found that the grid was 2 volt higher than the cathode. Voltage between anode and cathod was about 60v. The tube drew 11ma current. And it looked like working fine. I did a PSPICE sim, the result was very close.

Anybody sees any issues with this setup? Anywhere I can find a 6SN7 curve in positive bias zone?

[Geek]

Hi,

Sounds rather interesting. Have a schematic?

Cheers!

[Brian Beck]

Although you don't usually see positive grid bias on a 6SN7 for audio, the combination of 60 volts on the plate, 11mA and +2 volts on the grid is a plausible combination.

6SN7 data sheet - look at page 4

The upper figure on page 4 shows curves for positive grid bias, but it will be hard to extrapolate the +2 volt line.

Better yet, use the standard plate characteristics on the bottom of the same page. Imagine a +2V curve to the left of the 0V line, spaced the same distant as the 0V curve is from the -2V curve.

Still, I wonder if something might be wrong in your circuit to put the 6SN7 in this unusual operating point. Can you post a schematic?

[pftrvlr]

Here is the PSPICE schematic. The measured voltages are within 1%.

[Brian Beck]

Hmmm, to me this circuit could be improved if you had another 100 volts of B+ or so. Right now, the 10K load on the first stage is about equal to rp at those operating conditions. This is not optimum for 6SN7 linearity, although the unbypassed Rk will provide feedback. The gain of the first stage is roughly 5x (14 dB) - maybe less since it's driving some grid current (which is very non-linear). If you raised B+ to about 300 to 320 volts, then you could put a 30K plate load on the first stage. You might have to play with Rk a bit to keep the first stage's plate voltage roughly where it is. This RL is about 3 times rp, a better choice. The higher B+ would also pull the cathode follower out of the grid current region, with a new grid voltage of somewhere around -3V. Gain in the first stage will go up to roughly 10X, or 20dB. The additional 6dB may or may not be to your liking.

BTW, the output resistance of the 6SN7 CF will be roughly 375 or 400 ohms. Unless you plan to drive only very high-Z headphones, the 6SN7 may not be the best choice for a CF here.

[pftrvlr]

Yeah, there is easy way to mod the CF here.

I might just take the output from the amplication stage and parafeed to an output tansformer. I happen to have two 4k-8 lying around.