Spice Simulation of NFB Amp with embedded PF

[Koonw]

Here is LTSpice simulation for the OTL amp with embedded PF. I use LTSpice as it costs me nothing. You can google and download a copy for free, good for drawing schematic diagrams too.

First need to download inc files for tubes of interest, open using notepad, replace all chars "^" with "**". Need to change the triode symbol asy file so the PIN designator reads G, A, K not Grid, Anode or Cathode. Then use Spice directive to locate all the inc files, just press "s" will pop up the small window.

I have done simulation for two amp 25W 6c19p and 100W 6c33cb, for 6c19p simulation is quite close except the power is about 20W, as I use bias auto-adjustment in the output tube I gain 5 more watts. Hope the information is of some helps to our diyers and friends

 

[Koonw]

It has been sometimes since I revise this schematic, I thought it worthwhile to see what is maximum output from this circuit. 2 tubes 25W,,,8 tubes 200W, 10 tubes 300W and 12 tubes 400W, every tube added after this gives at least 100W or more. Noted after 8 tubes the output stage has a gain of 1.1 which means it does not has to increase driver level and capable of driving full output within the grid voltage swing -100V and hence no current gird drawn. The driver voltage has reached 580V still due higher output level required by the top driver as output level increased, and output tube to 230V (230-110), still has about 120V working voltage at full voltage. This is about the limit for such totem pole topology I think.

 

[jazbo8]

Are you going to build the prototype? It sims well, would love to see how the real circuit compare with the sim.

 

[Koonw]

Are you going to build the prototype? It sims well, would love to see how the real circuit compare with the sim.

Maybe I will, not with 12 X 6c33c, but 12x4=48 6C19P, I have plenty of them (1000?). As for now, the top driver is driven much higher so maybe sonic wise it is also influenced by it? In the mean time I am trying to sim an IT in the upper drive only so there is balance drive everywhere. I already have built a 25W model with 10 X 6c19p, and a pairs of ECC99, sound very good. I built it before I sim so I already the sim is correct. It is more fun this way I agreed, you have to built one way or another to see actual results.
Recently I think one +V rail supply rely has lost contact maybe due to too much power drawn from PSU.

 

[jazbo8]

Cool! So the 6C33c is just the big brother to 6C19P and should scale accordingly, kinds pricey these days though, wish I had bought some when they were cheaper.

 

[Koonw]

Here is another update of the simulation. I have asked the question earlier in another thread as to whether EC alone can correct the unequal drive requirement for top and bottom. The analysis shows that is can be done and it agreed with the practical amp I built. The reason it does NOT agreed earlier is due to sagging of -170V rail, 3 out of 4 200V/1000uF are faulty, so that explain it, after repaired it agreed with sim. So I feel it is necessary to get the correct information through. Also I have tied the 2 driver cathode resistors to get a current sink instead of capacitor, that makes the bass really comes out now.

 

[Koonw]

OTL AMP WITH FULL ERROR CORRECTION

Here is final OTL amp with full error correction, distortion level as low as 0.05% and zero phase shift in audio region. I have verified with real amp I built, it does work for me. If you have built this amp please let me know what you think.