My New EL84 PP

[Sherman]

Some of you may remember a while back I was experimenting with transformers as phase splitters. I eventually ordered a pair of PSA-2N phase splitters from Electra-Print. They are actually autoformers rather than transformers.

I decided to try them in a PP EL84 amp. Jack has a schematic for one on his site but I decided to roll my own. See attached schematic.

I finally got to the stage where I can test the amp. I'm "borrowing" a power supply from another amp since my new PTX isn't here yet. The B+ I'm getting from the borrowed PS is perfect for the new amp, I just had to drop in an additional RC filter to obtain the B2+ for the input stage. Attached is the PS schematic I'm using now. The new PS will be similar except it will use a 5AR4 and electrolytics rather than motor runs for the caps. (No room in the chassis for the big motor runs.)

[Sherman]

Here is the PS I'm testing it with. The actual PS will be similar.

[Sherman]

Here is the topside. Keep in mind the PTX and 5AR4 are yet to be mounted. The chokes will mount on the underside. The cage over the OPTs is a wire basket from Container Store (about $4 IIRC) held down with a couple of plastic wire routing tabs.

Front-

[Sherman]

Back-

I put the inputs on the back. I like the front to look clean with only an On/Off switch.

[Sherman]

Inside-

The autoformer phase splitters are mounted between the turret board and the back of the amp. They are immediately below the OPTs, probably not the best place but I can't hear any hum so I may not move them.

[dsavitsk]

Nice design. I think transformers/chokes as phase splitters makes a lot more sense than more tubes. Seems much more elegant to me.

You might try replacing the resistor plate load with a simple CCS -- performance will likely go up and the need for large electrolytics and the extra filter stage to drop voltage will go down. I think there is plenty of space on the side for some motor runs.

-d

[Sherman]

Thanks,
I think the autoformers are a nice solution, no extra tube sockets, lower demands on the PS, easy to implement. I'm tempted to try a pair in a PP KT88 amp.

I have some of the Forum CCS boards, I'll have to look into that. Would a current source (plate load) be better case than a current sink after the cathode?

I've also thought about removing the cathode bias on the EL84s and going with fixed bias. I'd have to do a voltage divider off the HT supply though since the PTX doesn't have a bias tap. Should be doable.

This is bad... thinking about mods and the amp isn't even completely finished yet!

[dsavitsk]

Originally posted by Sherman
I have some of the Forum CCS boards, I'll have to look into that. Would a current source (plate load) be better case than a current sink after the cathode?

You want it as a plate load where it presents a super high impedance to the tube. With the 6N1p, you can run it at about 11mA and bias it with a red led (instead of the resistor/cap pair) which will also offer a nice performance bump.

[MQracing]

Originally posted by dsavitsk
Nice design. I think transformers/chokes as phase splitters makes a lot more sense than more tubes. Seems much more elegant to me.

I agree. the circuit is pretty elegant. My only concern is the particular phase splitting autoformer chosen--- has only 10 henries of pri L on the driven half--- which will not provide much of a load impedance to the anode of the tube.

In defense of Jack at EP--- this unit (according to a spec sheet he used to have on line) was really designed to be driven by a cathode follower--- where the source impedance (i.e., the output impedance of the CF) is say a hundred or so ohms--- in which case 10 henries
(whose ind reactance at 20 hertz is approx 1250 ohms) would be perhaps a sufficiently high impedance---

but if the source impedance (the r sub p of your anode) is derived from the anode and it is several kilohms in magnitude--- that same 10 henries is not going to provide much of a load to the anode of the tube.


:::You might try replacing the resistor plate load with a simple CCS -- performance will likely go up and the need for large electrolytics and the extra filter stage to drop voltage will go down. I think there is plenty of space on the side for some motor runs.::::

it is a neat circuit--- and a builder could have several choices of how to load the anode---- with a resistor, a CCS, or a plate choke. But the achillles heel as drawn is the lack of L in the driven half of he autoformer.

This L will be in parallel with whatever is sitting on top of the anode---- so that 1250 ohms in parallel with a CCS is still going to be less than 1250 ohms effective impedance.

Of course as your frequency goes up--- so won't the inductive reactance of the autoformer's driven half.

but as is--- the limiting factor is going to bee the ten henries.


But--- I too like the circuit very much--- it conceptually very close to the SETH amp as shown on our forum at AA.

MSL

[dsavitsk]

[QUOTE]Originally posted by MQracing

Originally posted by dsavitsk
I agree. the circuit is pretty elegant. My only concern is the particular phase splitting autoformer chosen--- has only 10 henries of pri L on the driven half--- which will not provide much of a load impedance to the anode of the tube.

Probably worth a call to Jack. He is selling it as a plate load, and my experience is that model numbers don't mean a whole lot with him. He may have wound it differently than in the past, or just called it the wrong thing. Who knows. But, I'd give him a ring.