Calculating B+ for direct coupled SE amp

[JWK]

Many years ago I drew up a load line for a SE KT88 amp in UL. I have to go back now and relearn how to do that (the load line part) because I want to go back to this project now that my kids are a bit older and I have a bit more time. Only this time I want to direct couple with a mosfet.

I do remember my B+ being about 350v. I also remember my estimated bias voltage for that B+ and my OPT (2500) was about -35v. I can get a PT from Weber that has all different windings all at 150mA. Should be about perfect. I can get all different voltages needed for the bias (fully recitified), B+ and mosfet driver. The only thing I'm worried about is having all those SS bridge rectifiers creating a lot of noise.

The thing I can't wrap my mind around is what this does to the desired B+. My goal for the amp is to get as much clean headroom I can possibly squeeze out of it. If I remember doing the load line correctly, I set the bias voltage so that clipping would be symmetric. Wouldn't the mosfet driver now change this? I can't help but think that with the same B+, the tube would draw too much current when pushed into clipping.

Is this correct? Wouldn't you have to lower the B+? By how much, the bias voltage? So if your B+ was 350v and your best bias point sits at -35v for the most clean headroom, would you lower the B+ to 315v?

Thanks to anyone who can help me understand this.

[Steve Conner]

What are you proposing to direct couple to what? Do you mean the Class-AB2 thing where you drive the power tube grid with a MOSFET?

For maximum unclipped output, the idle current on a SE amp wants to be about half of the maximum current the tube can provide. Driving the grid positive allows it to deliver more current. Therefore, you have to increase the idle current in proportion, by using a smaller cathode resistor or whatever.

If that makes the tube dissipate too much, then you have to lower the B+ voltage to get it back in spec. Or lower the screen voltage. Of course this will lower the power output back to what it was before you started driving the grid positive.

If you really want maximum clean headroom, you could always, shock horror, get another KT88 and operate them in push-pull.

[JWK]

Thank you, Steve. Yes, the AB2 thing where you drive the power tube grid with a mosfet. I realized today that the way I phrased that question must have come off like vague jibberish.

I'm pretty sure what you're saying agrees with what I thought I had figured intuitively. Driving the grid positive with respect to the cathode allows the tube to deliver more current. So if your class A SE power amp is biased just right for maximum unclipped output (staying within dissipation ratings) and you just stick in a mosfet driver without changing bias voltage or B+ values, the tube will still clip in the same place at the grid going negative, but now will not clip as the grid is pulled positive and will continue to draw current, going over dissipation ratings (assuming we had the parameters set to maximum safety). So we lower the B+, but by how much? And how do we plot this out on a load line?

[Steve Conner]

So we lower the B+, but by how much? And how do we plot this out on a load line?

Well, if your KT88 datasheet doesn't include data for positive grid voltages in the charts, then you can't plot it at all.

Nothing you do to any of the other electrodes will prevent the control grid from exceeding its dissipation rating. That is always a worry with positive grid drive. Most audio tubes aren't designed to run grid current at all, which is why you don't see the positive grid regions on the datasheet.

[JWK]

Well, if your KT88 datasheet doesn't include data for positive grid voltages in the charts, then you can't plot it at all.

Nothing you do to any of the other electrodes will prevent the control grid from exceeding its dissipation rating. That is always a worry with positive grid drive. Most audio tubes aren't designed to run grid current at all, which is why you don't see the positive grid regions on the datasheet.

Well, now I don't feel so stupid. I've been staring at the UL sheets for KT88s all afternoon trying to figure out what I'm missing.

It sounds to me like I might not want to use this technique on an SE amp that will be used often to drive the piss out of the power tube. I will concentrate on efforts to get a smooth transition into power tube distortion without blocking distortion. I plan on setting it up in UL with fixed bias. My other goal with this amp is to minimize the "hisss" and "Shhhhhh" as much as possible since this will be used as a practice amp. It's not that I need whisper volumes; I don't. But when I'm practicing in a room alone that noise out of an amp drives me nuts.

I suppose I should start another thread.

Thanks for the info and help!

[Cygnus X1]

I thnk you are referring to the WPTGP transformer?
I run that with the recommended bridge 4-diode setup and it is very quiet.
Go to AX84 and take a look at the SEL schematic for a good KT88 style supply for cathode bias.