noob PSUD question...

[roundel325]

I am tring to design a power supply for a el84/12at7 SET/SEP amp. I downloaded the PS modeling program from duncanamps, but I don't know what value to use for the "load resistor" in the modeling program. I'm shooting for 250v and 185v B+, using a 300VAC Xformer. Any suggestions? Thanks.

-Jared

[electronaut]

Hi Jared,

The resistor represents the load of the circuit you're planning on powering. Do you know the current requirements of your amp? If you do, then use Ohm's law to find an equivalent resistor value that would result in the same current at that voltage, and use that value.

If you don't know the current requirements, post your schematic and I'm sure someone will help you figure that out.

-E. Lectreau

[EC8010]

Even better, having found the current drawn by the load, double-click on the load resistor, change it to "constant current", and type in your load current.

[roundel325]

Thanks the tips. I have not scanned the schematic, so I have no way to post it. I have all the tube data sheets, is there a value of the current draw for each tube I could look for on there and then add them up? Or is it more complicated than that? If so, I can scan it and post tomorrow.

Also, does anyone have a schematic of an el84 wired single ended PENTODE? I'm shooting to make my design switchable from SET to SEP, but all the schematics I have found are wired in SET. Thanks.

Jared

[roundel325]

Also, wouldn't the current draw be changing with the music? Or does a class A amp draw constant current no matter what?

Jared

[dhaen]

You can determine the current drawn by output each tube by looking at the cathode resistor value, along with the tube spec. Since it's the output stage that takes the "lion's share" of current, the other stages can be guessed (perhaps allow 10-20% extra on the total current for these in your case).
Looking at the Mullard data sheet on Franks tubes 40mA or so for each EL84 at 250v seems a reasonable starting point.

The current consumption of a class A stage does not change with music, but averages to a constant current. There is a change in current when overload is approached. For design purposes we consider it constant.

Changing to pentode mode normally only involves connectig the screen grid (G2) to +B. Sometimes a series resistor is used. It's presence and value will depend on the value of +B.
Pentodes have high distortion levels unless negative feedback is used.

[SHiFTY]

Try here for starters.

EL84s are pretty nice valves, I use them in my guitar amp. And don't be scared of a little NFB, its not a crime!

[roundel325]

Thanks again guys.

When you say pentode wiring increases the distortion, is that because of the length of the load line? I ask because I am using 10k primary OPTs, so my load line is longer then the average el84 SET (I think). So, knowing this, would it be a good idea to impliment some NFB, and if so, how do I do that?

Once I get a schematic drawn up with SET->SEP switching, I'll post it in a new thread and let everyone fire away on the whole design. Right now, I'm just looking for the mechanics.

Thanks.

-Jared

[dhaen]

It's not trivial like triodes. A good explanation is on Steve bench's site here .
You'll see that increasing load impedance does not necessarily gain you anything in this mode (it may be worse).
10k sounds awfully high. It'll be nice in triode mode, but with limited output power. I don't think there is very much to be gained with NFB in this case.
The normal reason for going for pentode mode in an output stage is the increased power that can be got, but I don't think you'll achieve that with 10k.

If you do want to try negative feedback, you need surplus loop gain (which you will get when you switch to pentode mode), and for global feedback, a circuit without too much phase shift (most circuits with only one transformer in the signal path qualify).
You can choose to apply local feedback to each stage, or a global - overall feedback.