Not sure why your worried about plate impedance on the LTP. Even with say a couple of 6ej7 pentodes the 30k loads will still drive the kt88's without slew rate limiting.
When DC coupling the LTP you need a reverse diode between k and g of U2 for protection at turning on.
@jcalvarez,
Thanks! I have this up and running now with a 10M45, as I didn't have any 10M90. It seems to work, but I had to reduce the current through the 6922 significantly. Going forward do you have a working LTspice model for the 10M45/10M90 so I can also model this to get an idea what I can expect? I have added the measured voltage to the below schematic, does this seem about right? R3 and R6 are 33K in my prototype.
Thanks! I have this up and running now with a 10M45, as I didn't have any 10M90. It seems to work, but I had to reduce the current through the 6922 significantly. Going forward do you have a working LTspice model for the 10M45/10M90 so I can also model this to get an idea what I can expect? I have added the measured voltage to the below schematic, does this seem about right? R3 and R6 are 33K in my prototype.
If I understood correctly, the numbers you put there are measured in the real circuit, right?
CCS current: 3.4V/390 = 8.7mA
U2 Ia: 293V/33K = 8.87mA ????
U3 Ia: 297V/33K = 9.0 mA ????
Something is not right, either oscillations or something else. The CCS current looks fine, but the LTP plate currents are not, should be half of the CCS current in each plate.
Attached is the 10M90S model I'm using.
I added some more current and voltage labels
CCS current: 3.4V/390 = 8.7mA
U2 Ia: 293V/33K = 8.87mA ????
U3 Ia: 297V/33K = 9.0 mA ????
Something is not right, either oscillations or something else. The CCS current looks fine, but the LTP plate currents are not, should be half of the CCS current in each plate.
Attached is the 10M90S model I'm using.
I added some more current and voltage labels
I arrive at different numbers?
400-307=93V > 93/33000=2.8mA > Current U2
400-303=97V > 97/33000=2.9mA > Current U3
Total=5.7mA
There's no oscillation or anything, it measures fine, in line with the original solution.
400-307=93V > 93/33000=2.8mA > Current U2
400-303=97V > 97/33000=2.9mA > Current U3
Total=5.7mA
There's no oscillation or anything, it measures fine, in line with the original solution.
P.s. R8 is a 1K trimmer in my case, not a 390 ohm resistor, I adjusted for best THD result and have not measured the actual value.
Sorry, I should never post when I'm in a conference call....
Somehow in my math 400-307 = 293 🙄
But your CCS seems to be giving 8.7mA (3.4V/390 ohms), which I would expect to result in ~4.35mA per plate, and much lower voltage. Are you sure you got 400V on top?
Somehow in my math 400-307 = 293 🙄
But your CCS seems to be giving 8.7mA (3.4V/390 ohms), which I would expect to result in ~4.35mA per plate, and much lower voltage. Are you sure you got 400V on top?
Is your LTP power supply wired like this (in your original schematic)?
I had to reduce R16/17 from 12K to get 400V with the CCS at 5.6mA ....
I had to reduce R16/17 from 12K to get 400V with the CCS at 5.6mA ....
Okay, I'm back to the old configuration and added a 500K potentiometer to be able to change the AC balance on the LTP, R35 and R38 in the schematic. Please have a look at the THD vs. power plots, I wasn't expecting results this good to be honest, this really is a significant reduction in THD.
Proto 051
THD vs. power - Old vs. new
Proto 051
THD vs. power - Old vs. new
My take is that the two triodes inside the 6922 are not perfectly balanced, which makes sense, and by changing the anode resistors AC balance is restored so now the loop gain that was available has less distortion to work with which nets lower overall distortion.