Alright, so I am building the beast to end all beasts, finally. 6x 6L6GB class AB2, 500v plates, 420v screens. Combination bias, with mostly fixed but a bit of unbypassed cathode bias added for tone & safety. I am driving each side with 6AV5GA pentode cathode followers.
I am using a bipolar supply, with a negative rail around 280v. This is established with a fixed voltage divider / cap combo off a maximum 400v negative rail.
I have the 6AV5 plates tied to the 420v screen supply. The 4k7 cathode resistor is also the grid return for the power tubes. The screens are tied with 220k resistors (47ks evaporate, 470ks drive screen too low) to the anode, and cap coupled to the cathode (.1uF for testing, probably .5 or 1uF finally). Screen voltage is around 65v, cathode around -37, plates at 420 or so. I am feeding the expected bias voltage for the driven tubes into the grid of the 6av5 (-12v, cathode bias makes up the rest). Is this correct? I haven't done DC-coupled drivers before, and I am not sure how to feed fixed bias through them. In any case, I end up with approx -37 volts on the grids of the power tubes, not 12, and there is quite a bit of warmth coming from the power supply. It seems like I am forcing the 6AV5 into grid current, so I am not sure if I need a different voltage on its grid to get -20 ish volts on the grids of the driven tubes. I should note that there is no red plating or glowing screen grids that I can see.
Everything other than the negative rail voltage divider is running really cool, and I went overboard on lead dress, winding RF stoppers on screen grid wires (50 turns between sides), etc. 340uF on plate supply, 240uF on screen supply. Scary, really! Even with 8 power tubes, heater voltages are still around 6.7v, need to draw more current
I am using a bipolar supply, with a negative rail around 280v. This is established with a fixed voltage divider / cap combo off a maximum 400v negative rail.
I have the 6AV5 plates tied to the 420v screen supply. The 4k7 cathode resistor is also the grid return for the power tubes. The screens are tied with 220k resistors (47ks evaporate, 470ks drive screen too low) to the anode, and cap coupled to the cathode (.1uF for testing, probably .5 or 1uF finally). Screen voltage is around 65v, cathode around -37, plates at 420 or so. I am feeding the expected bias voltage for the driven tubes into the grid of the 6av5 (-12v, cathode bias makes up the rest). Is this correct? I haven't done DC-coupled drivers before, and I am not sure how to feed fixed bias through them. In any case, I end up with approx -37 volts on the grids of the power tubes, not 12, and there is quite a bit of warmth coming from the power supply. It seems like I am forcing the 6AV5 into grid current, so I am not sure if I need a different voltage on its grid to get -20 ish volts on the grids of the driven tubes. I should note that there is no red plating or glowing screen grids that I can see.
Everything other than the negative rail voltage divider is running really cool, and I went overboard on lead dress, winding RF stoppers on screen grid wires (50 turns between sides), etc. 340uF on plate supply, 240uF on screen supply. Scary, really! Even with 8 power tubes, heater voltages are still around 6.7v, need to draw more current
Last edited:
Update, not sure where I was getting -37 from, but I swapped some resistors around in the negative supply for more power handling. Power tubes grids are at -250, cathodes grounded for now, driver's cathode resistor ties to -290 and is only dropping 30-40v across it. In triode mode the situation reverses dramatically, with bias applied to grids of power tubes dropping to near zero accompanied with rapid red plating before the plug is yanked. It does make sound this way, however . I'm pretty stumped, I feel like I'm missing something quite important here. Thanks for any help!
. I'm pretty stumped, I feel like I'm missing something quite important here. Thanks for any help!
Last edited:
I do. They work just great for Hi-Fi too. Avoids blocking distortion caused by over drive, and makes for better behavior upon clipping.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.