R6 is too low, the current through the 6V6 is too high, and you have positive grid Vs the cathode. Increase R6 to 15K and try again
PS I think your 12AX7 is biased too far, 1.6V is a lot for a 12AX7. If you want a diode for self-bias, use a small signal diode like, for example, the 1N4148, and select the anopde resistor to have around 115~150V. Then you can play with R6 to select the biggest current you want for the follower, while keeping the 6V6 negative bias lower than -3V.
PS I think your 12AX7 is biased too far, 1.6V is a lot for a 12AX7. If you want a diode for self-bias, use a small signal diode like, for example, the 1N4148, and select the anopde resistor to have around 115~150V. Then you can play with R6 to select the biggest current you want for the follower, while keeping the 6V6 negative bias lower than -3V.
Last edited:
I think the problem is that you used a 12AX7 (ECC83) for U1 and not a 12AT7 (ECC81). (I used a 12AT7.)
Since U1's plate is DC-coupled to U2's grid, you can't swap tube types.
Try it with a 12AT7 for the input stage.
Since U1's plate is DC-coupled to U2's grid, you can't swap tube types.
Try it with a 12AT7 for the input stage.
Okay, I was thinking more about the Feliks Euforia which inspired this idea in the first place. The Euforia is a dual 6SN7/6AS7(6N13S) amp. However, you look on Head-Fi and other places, you can see people using adaptors to put KT66/77/88 not only in place of the driver tubes, but the input tubes as well. Now, this wouldn't be possible with all dual triode circuits--for example, I wouldn't try it with the Bottlehead Crack even though it uses similar tubes, because the Crack puts each stereo channel in one half of a single tube. My suspicion is that the only way to really work would be if you're treating the dual triode as essentially a single big triode. Like this sort of thing:
Then if you put a tetrode or a pentode in the socket, it would just behave like a triode.
Searching around the internet you can find circuits like this for OTL amps. The above screenshot for example comes from this design, which uses different tubes:
I also found this design:
Which seems like the Euforia-type circuit but with an extra tube at the input. I quickly mocked this simplified version of it (using 6080 tubes because I don't have a 6AS7 model handy):
The distortion profile isn't great, but it's not the worst either. Might be worth experimenting which a real 6AS7 model, with different values, with feedback, and also of course with putting a KT66 into the driver slot in triode mode.
ASC and models used attached.
Then if you put a tetrode or a pentode in the socket, it would just behave like a triode.
Searching around the internet you can find circuits like this for OTL amps. The above screenshot for example comes from this design, which uses different tubes:
I also found this design:
Which seems like the Euforia-type circuit but with an extra tube at the input. I quickly mocked this simplified version of it (using 6080 tubes because I don't have a 6AS7 model handy):
The distortion profile isn't great, but it's not the worst either. Might be worth experimenting which a real 6AS7 model, with different values, with feedback, and also of course with putting a KT66 into the driver slot in triode mode.
ASC and models used attached.
A big difference in that Japanese 6AS7G design is global negative feedback. Take that away and of course linearity suffers.
So the gain was very high on my first draft, but just tweaking some of the resistors and lowering the gain and I get the distortion numbers pretty darn low without having to resort to global negative feedback, though the FFT chart looks a bit 'grassy'.
Does seem like this could be a viable amplifier...
Does seem like this could be a viable amplifier...
Of course, simply swapping in the KT66 for the 6080 and not changing anything else produces gloriously terrible results:
How does the Euforia do it? Wish I knew.
How does the Euforia do it? Wish I knew.
Okay, just adding a line from the screen grid back to the anode with a small resistor suddenly makes the KT66 work with not that much distortion (relatively speaking) AND the anode current is only 9mA on the KT66 anode and suddenly I'm back to thinking this might be a good idea again.
(Edit attached the asc--you can find the KT66 model at the beginning of this thread)
(Edit attached the asc--you can find the KT66 model at the beginning of this thread)