low power... PI issue?

[black_labb]

in the bass amp i've just finished building (currently troubleshooting) there doesnt seem to be the gain that there should be. i think it is in the PI that it is lacking. the PI is a LTP with 100k/91k plate resistors, 1M grid leaks, 470ohm bias resistor and a 12k cathode resistor. they are directly coupled to cathode followers which are set up with the output connected to the cathode and cathode resistors at 100k. the set up is essentially like this http://www.ozvalveamps.elands.com/ho...rcit200wka.gif but the LTP has slightly different values. i started with an 820ohm bias resistor but ended up changing it to 470 hoping it would help the issue.

it passes the signal nicely, but doesnt seem to have as much gain as i would have expected. i am a bit reluctant to do it again but i wasnt able to get much more than 2-4v peak-peak on the control grids of the power tubes.
my experementation is a bit unorganised as i have been sceptical about the dummy load so i didnt measure all the details in a logical order, so sorry for the lack of information.

the supply voltage is 350vdc, the voltage on the plates of the LTP (and the grids of the cathode followers) is 130v with a 12ax7, or 120v with a 12at7. the cathode followers may be pulling some current through the grids. i tried swapping the cathode follower tube between a 12ax7 and a 12at7 which didnt change the voltage on the plates/grids at all.
the plate voltage on the ltp seem too low to me.
i have an amp with the same output schematic as the one in the circuit posted above. the amp i have has 2 gain stages running off 12ax7's with a TMB tone stack and a volume control which seems to have alot more gain that this amp. so i know that the circuit works, but there seems to be something different happening here.

thanks everyone.

[black_labb]

just pulled the driver tube and the voltage on the PI was 220v, so the grids of the driver stage is pulling a fair bit of dc. guess this means that the grids were drawing ~1ma (100k resistor, 100v difference when installed).

edit: been looking for a solution for the issue without adding coupling caps as i dont want to add more ac coupled stages within the feedback loop. i cant seem to find any other circuits with a DC coupled buffer stage. hiwatts seem to have a resistor with a cap in parrallel. i may try a resistor or a resistor with a cap.

[Steve Conner]

Hey labb

What plate voltage are you running your CFs from? And what negative voltage have you returned their cathode resistors to? (afaik in both cases, the more the better)

Tried bypassing the CF stage and driving the power tubes directly from the PI? This might help find some wiring mistake.

[black_labb]

the cathodes are running to ground, the plates are at 340 (same supply voltage as the amp.

i havent tried disconnecting the CF from the circuit, but i am thinking about it. i will have a look at the circuitry and if it doesnt work out then thats my next step.


i just pulled out the amp that is similar to the one in the schematic posted above (only difference is the one i have has a preamp and there is slightly more negative feedback). i tested it out and the circuitry a bit. the plates on the LTP run at 173 volts, and when i removed the CF tube the voltage didnt change, so there is an issue with my version of the design, not a part of the design. i didnt check the voltage supplying the CF's or the preamp, but the voltage before any of the dropping resistors is 375v, so the voltages are probably pretty similar to what i am running. the LTP supply voltage comes from a voltage tap one resistor downstream from the CF supply. would the higher supply voltage to the CF be biasing the CF colder than my design which runs them off the same supply? having the CF hotter on mine could be causing the grid to conduct the 1mA.



i think there are some issues with my scope (second hand ebay thing), so i'm not sure what to think about the output.


thanks steve

[black_labb]

should have had a closer look before, seems i never connected the plates of the CF stage. explains why the grids were drawing current.