diytube.com

What do you think about this change to the Poseidon board to accommodate the unused triode (an CCS grounded cathode amplifier)?

I am using 6SN7's to make the amp less sensitive.

Are there any optimizations you would make, like changing the voltages, plate/cathode resistors, etc, assuming 6SN7s are used? I will be driving a PPP stage with this configuration. I have a plan on how I will find the optimum feedback.

You can use the front of an Aikido there. There are alot of SRPP setups you can, but the Aikido is simple and sounds good. But the Aikido seems to run the tube the hardest. R2 should be 100 ohms, C2 can be much lower, like 0.1uf or .047uf, and you might be able to get rid of R3 and C1.

I ended using a 6C4 inplace of the 12AX7. Just didn't like the 6AT6 or 6AV6. Of course I can put in a 6AB4 for more gain, but 20 is good once I adjusted the feedback.

Hi,

- Make C2 47n instead of 680.
- Try R5 bypassed with a cap.
- R2 I'd definately use 470 to 1K. 100 is too low, IMO. Not enough Gm and mu to be multiplied in a 6SN7 by the top tube.

Why are you not taking the output from the top tube's cathode?

Cheers!

Here is an update with a spice waveform. I made a "true" SRPP directly coupled to the phase splitter, along with a POT to adjust the balance. In simulation I got it about as close as it will get. The open loop gain of the driver is about 40dB. If I run feedback into it I expect to lose a good 10 to 20db, depending on how much feedback I end up using. Any other suggestions?

That'll work

The 10uF cap... try a 100uF across that resistor... 16V Silmic-II or MUSE are great.

Just an idea.... try Spice again with a 33K, 2W in the PI tail. Might get a weeeeeee bit better balance.

Cheers!

Does the 100uF cap allow for better frequency response than the 10?

The 33k resistor made the balance worse. Is it supposed to make it better, if not there is some other factor that is probably throwing it out of whack.

Hi,

Yes, 100uF will have the F3 at 20Hz on that first stage. Right now it's up about 200 Hz (F3 is where the cathode R == Xc).

OK... I just thought 33K night because it's what worked on the KTA PI.

Cheers!