changing from 6550s/KT88s to EL34s

[jamesmafyew]

I searched on this topic -- a lot of people are interested in putting KT88s in their EL34 amps, but not the other way around. I've got a guy who's interested in putting EL34s in a 70s Sunn Model T (and being able to switch back and forth between KT88s and EL34s later on). I know that in this case, heater draw is a non-issue. I realize that ultralinear operation may not be ideal, so it would likely need a screen node created from the power supply. Then there's the bias supply, which sits at about -55V, so that would need to come down quite a bit to get EL34s idling properly. Otherwise, voltages don't seem to be an issue.

My understanding (could be wrong) is that the Model T OT has a primary Z of about 2k, for 4x 6550s. Most 4x EL34 amps are more like 1.7k, right? Not that far away. So in general, in an amp designed for 6550s or KT88s, how much power loss should you expect going to EL34s? Are there other considerations to this that I'm missing?

[Chuck H]

Your not missing anything. Proceed as planned.

Your right about the parameters being close enough. The power loss will be strictly a matter of the difference in the dissapation ratings.

You'll very likely get more distortion from the EL34's (if the amp is run cranked). Only because the EL34 requires less drive voltage than a KT88, and so will be driven further into clipping by the provided drive.

[jamesmafyew]

Good point about the PI drive -- I did notice that the Model T squeezes a lot out of its LTP, via both high plate voltage and seemingly very low NFB.

A point of technical clarification -- if you were dealing with 4x EL34s and all else being equal, would there be any difference at all in power output between a 1.7k OT primary and a 2k primary? Or is the knee of the characteristic curves broad enough that it doesn't really even matter?

[Chuck H]

Not sure. If I were better with data sheets I may be able to tell you for certain. It will depend on Vp. Any tube is happiest (and therefore able to do it's best work) at a certain load for a given plate voltage dependant on class of operation. The way data sheets are configured still confuses me. If your Sunn has highish plate volts (compared to , say, a Marshall) you may be better off with the slightly higher load. At any rate it's such a small difference as to be ignored. Any difference would be a watt or two most likely. You'd get more variable changing speaker cabinets since different speakers have widely different impedance curves.

Hopefully someone who knows the formula will key up just for the sake of clarity though.

[loudthud]

I did notice that the Model T squeezes a lot out of its LTP, via both high plate voltage and seemingly very low NFB.

The Model T has one of the highest NFB ratios you will find. There is a 22K feedback resistor and it's connected to the 16 ohm tap of the OT. EL34s might want to oscillate with a fixed screen voltage.

[es345]

... regarding the OT impedance 1,7K versus 2K there will be no big difference in the output power.

If you look to the EL34 data sheet:

there is one example for 2 tube PP with 500V Ub and 400V supply for the screens resulting in 70W output at a 4K OT
Following this path (- non ultralinear, screen resistors as described in the datasheet) with 4 tubes you will have an exact match and 140W output.

[GregS]

would there be any difference at all in power output between a 1.7k OT primary and a 2k primary?

If both the power transformer can source the additional current of the 1.7K load and maintain B+ and the tubes don't introduce a bottleneck to the increased current, then the 1.7K should have more power output. That's my understanding anyway.

Greg