Suggestion for push pull output transformer.

[Tom Bavis]

Class A has 360 degree conduction at all power levels, class B has close to 180 degree conduction at any power level, and class AB varies from 360 degrees to somewhat more than 180 degrees at full power. Class AB has a 360 degree conduction region, but it's incorrect to call it a class A region. (though it doesn't bother me as much as hearing "RMS Watts"...)

Using a 5K transformer at 2.5K is possible, but will have a price: twice the resistive losses, and an octave loss at thehigh frequency end. The Lundahl likely has an octave to spare - many others won't.

 

[Alan0354]

I don't care exactly how you call it as long as there is a region where both devices are on and there is no crossover distortion right where the small signals are.

People in SS world called that class A region. In that region, both devices conduct 360deg.

I am not using 2.5K primary for stage that is for 5K, that would be way off. I am using 2 pairs of 6L6GC. Each pair optimized at 5K, so 2 pairs in parallel will optimize with 2.5K primary. That's perfect match.

 

[kward]

The transformer reflects the impedance of the secondary to the primary (and vice versa) through the square of the turns ratio. Meaning, it will indeed be a 2.5K primary if the secondary is loaded as discussed. Again, not saying you should do it, there are some limitations as discussed. You asked for options. Just pointing out an option you may not have considered.

 

[Alan0354]

The transformer is designed with the number of turns on both primary and secondary for that specific impedance, by using using a 5K primary and connect the 8ohm secondary tap to a 4ohm speaker might theoretically gives you 2.5K primary, but it's not optimized. You have too many turns in both primary and secondary and parasitic dc resistance goes up, more capacitance than if the transformer is design the right way. I don't think I want to do that.

I specifically go to a lower power 6L6GC and use 2 pairs to get OPT with lower impedance so it has less turns, less parasitic than a 5K primary. It would defeat the whole purpose if using a 5K OPT and tap it the wrong way.

Put is in another way. Lets just say the secondary is 10 turns, and primary is 100 turns for 2.5K to 4ohm. For 5K, to use a 5K primary to 4ohm, it would be 20turns for secondary and 200 turns for primary to give the same 2.5K to 4ohm. Yes, the ratio is the same, but you double the number of turns if you play the game. You defeat the whole purpose to use a lower primary impedance.


We are talking about quality of the OPT, we talk about parasitic, inductance and all, the lower the primary impedance, the lower the turn ratio, the less turns you need compare to a much higher primary impedance. The less turns, the less parasitic, less of all the unwanted stuffs. I don't want to play the game of matching to get the right impedance.