5 wire output transformer

[navin]

Originally posted by ray_moth

Triode: minimum 5k; not critical, higer impedance will reduce both power and distortion

Ultralinear: 6k to 7k, the higer impedance will reduce both power and distortion

Pentode: 3.5k, critical.

the primary of the OPT I am looking at is 5k. The builder of this OPT claims it is used in UL/PP mode by Diva amps.

I intend to use it in PP/Pentode AB1 and PP/Triode AB.

from the mullard datasheet:
Ra=2k in Class A Pentode for 1 valve,
for 2 valves in PP 6.6K with screen grid tapping of 43%,
for 2 valves in PP 3.4K with fixed bias and cathode bias

Ra= 3K Class A Triode
for 2 valves in PP 5K at 400V

so from this is seems that i would need OPTs with a 3.4k primary or 6.6k primary. now i am confused. if I use the OPT in pentode AB1 what should be the ideal primary impedance? The supplier says he can wind a custom primary for me if i wish.

[ray_moth]

IMHO, given your stated intentions and the fact that you can get the OPTs wound to order, you should go for 6k plate-to-plate and use that same ratio for both UL mode (bass amp) and triode mode (treble amp).

UL will give close to pentode power and wil not reqire large amounts of NFB (20dB should be enough) to get a decent damping factor. Pentode, on the other hand, is a more difficult mode to design and build, because of its high OP impedance and the consequent need for a lot of NFB (26dB or more) to get an acceptable damping factor. The more NFB you need, the more challenging the design.

[navin]

Originally posted by ray_moth
IMHO, given your stated intentions and the fact that you can get the OPTs wound to order, you should go for 6k plate-to-plate and use that same ratio for both UL mode (bass amp) and triode mode (treble amp). .

I've told him to make 2 OPTs with 5K and 2 with 6.6K. makes sense. 5K for triode and 6.6k for UL. correct?

[ray_moth]

Hi Navin,

Yes, I was thinking of flexibility but your solution is more optimal.

[navin]

Originally posted by ray_moth
Yes, I was thinking of flexibility but your solution is more optimal.

thanks. this is 1st piece of good news i got on this project :-).