New build- 2 channels, FOUR output transformers!

GordonW

Speakerfixer
Subscriber
I finally have the critical parts, to move ahead a bit more with my idea on using the Leslie/Triad 100-12 output transformers.

These transformers have been a bit of a pickle- they have great specs, but they're made for SIXTEEN ohm speakers only. And they lose a LOT of power when loaded super-hard, into lower speaker impedance.

Well- I took a cue from some really high-end amps (VAC Statement, etc) and decided to do something about it. Instead of one transformer (with one pair of output tubes) per channel- TWO output transformers (each with a pair of output tubes) per channel, with the secondaries wired in parallel- voila, EIGHT ohm output.

The first problem, naturally, is packaging. That's a LOT of transformers and tubes to fit onto a chassis. I think I came up with something that will work well (17"w x 8"d x 2"h chassis):

4-21-17_el84_ppp_4opts_chassis_layout_900.jpg

Two output transformers and four EL84s per side, two 12AU7 per side for pre-amp, driver and phase inverter, and two 12AX7s for phono stage. That's fourteen tubes!

According to the sims (and I've done a LOT of them), it should easily be able to do 30w/ch- maybe as high as 35 or 40 watts per channel. And by the measurements that have been done on these transformers- it should be able to do this from 20Hz to above 50KHz, EASY.

I found a couple of chassis, that were made for the "Trainwreck clone" guitar amp, that are exactly the right size, and built like a tank. Unless I find something better soon, that's what I probably will be going with.

Now, I just gotta get a suitable power transformer. Currently, I'm looking at an Antek toroid, with 230VAC secondaries. That should give me right at 300v on the plates and screens.

The circuit will likely wind up being my modified Pilot AA902 layout that I've used so many times before, with a Dynaco phono stage. Should be plenty of gain- in fact, I may wind up having to run things at minimal gain on the preamp and driver stage, to keep the gain from getting too high. But, that's a better problem than not having ENOUGH gain, naturally...

Regards,
Gordon.
 
I don't recall if I gave you this info... but here's my measurement of those transformers (used a Wayne Kerr Impedance Analyzer). Impedance and phase are flat (15%, 20 degrees) to 100 KHz.

Parallel tubes make wonderful oscillators - it'll want separate screen and grid stopper resistors, and probably plate suppressors as well (47 Ohm carbon with 5-10 turn parallel choke wound on the outside).
 

Attachments

  • Leslie_xfmr.pdf
    47.1 KB · Views: 25
I'm hoping that the separate phase inverters and isolated primaries per transformer (I am not going to connect the primaries from one transformer to the other, at all), with separate screen and grid stoppers (as well as separate cathode bias resistors and caps per transformer) should do it. I hope to not have to use plate suppressors, but I'll definitely keep that in mind.

Regards,
Gordon.
 
This is an interesting project. Hope it goes well. I would be at least a little concerned that gain differences in parallel circuits without local feedback will result in power loss and excessive distortion when the signals are combined. What's your thinking on this issue?
 
I was also wondering what the feedback would end up doing since it would end up feeding back the products of both amplifiers. I wonder if maybe a common voltage amp stage fed out to different inverters would help any, though that might just take the same problem and move it one stage downstream.

To be fair though, "sound reinforcement amps" were often designed to have their outputs strapped together and I don't think they did anything particularly strange with the feedback in that mode.
 
I tried something like that driving a Dynaco 70 with an external phase inverter,8 ohm speakers on the series-ed 4 ohm taps, it worked, but it oscillated once for some reason.If the transformers are parallel at the output, it should be able to have a normal feedback arrangement, it would use an 8 ohm speaker on the paralleled 16 ohm tap
 
Well, thinking a little further on this topic and realizing that pentode finals behave like current sources, I reckon that paralleling the secondaries should work without difficulty. My concern would be valid only if individual output stage impedance was low, as it would be with local feedback and/or triode finals.
 
Well, thinking a little further on this topic and realizing that pentode finals behave like current sources, I reckon that paralleling the secondaries should work without difficulty. My concern would be valid only if individual output stage impedance was low, as it would be with local feedback and/or triode finals.

That's exactly what I was thinking, and what seems to happen in the amps that do this in practice.

Though, I've seen Kevin Hayes (VAC) do this with UL outputs. Works fine, sounds incredibly good. Of course, he's a master with decades of experience at doing this... and he has the ability to tine-tune parts for exceptionally good matching, too.

Regards,
Gordon.
 
Bonus - You'll get plenty of bodybuilding practice moving this rig around. Interesting project!
 
Back
Top Bottom