F4 power amplifier

[cal3713]

Thanks Pico, appreciate the information. Maybe just learning towards stock for safety though...

And I've got my testing rig built, still waiting on the reusable zipties for clamping/unclamping, and my regulated lab power supply for providing constant voltage & amps, but am otherwise pretty much ready to go...

 

[cal3713]

Also, does anyone know if I'll need to heat or cool the mosfets when testing at 22v and .4A to get them to 50C?

 

[Zen Mod]

8.8W demands some cooling

easiest way is to have moderate sized heatsink, breezed with 12V fan all the time , so temp change would not mean much

clamp mosfet on heatsink, wait while reading became steady and that's it

that's - at least in my book , enough of temp. repeatability

 

[cal3713]

Great, thanks zm. I can do both, so I should be able to just set it exactly at whatever temp I want (50C I read). I guess warm up to 50 and then keep cool once current starts flossing...

 

[Zen Mod]

I'm finding good enough having heatsink blasted with fan

that way temp. change isn't big - mosfet is getting fast to steady readout

 

[cal3713]

Well, it ain't pretty, but I believe I'm getting the job done. Currently on an initial practice N channel I had left over from a previous build, but hovering around 4.40Vgs at a constant 22V, .42A, & 50C.

 

[2 picoDumbs]

Good job

 

[cal3713]

Thanks 2PD.

I saw in Nelson's resistor calculation that he's subtracting 4v for the post-resistor voltage. Should I be running up at 26v so that the MOSFET is getting a full 22v?

 

[2 picoDumbs]

Not super critical, if you're close to operating conditions and they are all done in the same manner this will suffice.
If you have already tested a stack and don't want to have to start all over again, just stick with what you have, just record the testing conditions, so 12 months later you can be sure how they were tested.

It's up to you to decide.

 

[cal3713]

Thanks. Fortunately I had another 10W 10R laying around so I was able to do 26v at the power supply and sit right at .41A. Still working out the official test, so I didn't have to retest anything and am slightly more "proper" this way. Appreciate all the help & input.

 

[cal3713]

All done. Pretty low variation, so I should be able to get matched parts out of the full complement. I'll work on shipping next week.

If anyone else wants a matched set, please pm me. 50 claimed so far with 34 left.

 

[myleftear]

Are those Harris so much better than the „original“ ones? (Worth a swap)?

 

[cal3713]

Unfortunately I haven't heard them yet, and will also have premium resistors in the circuit when I do, so I won't really be able to do a full a/b vs a traditional build...

 

[cal3713]

Hi folks, everything's taken... I'll be asking for mailing addresses via pm and will ship out by the end of next week. Hope they sound great...

 

[2 picoDumbs]

You should have charged more.

Hahahahaha

 

[cal3713]

Due to a miscommunication, I have 6 matched Harris 9240s left (two matched trios). Anyone want them?

Hi all. As those of you following the "F4 Power Amplifier" thread will know, Nelson recommends Harris IRFP9240 parts for the P-channel mosfets as they show reduced distortion compared to the 9240s from other manufacturers (see:
https://www.diyaudio.com/forums/pass-labs/97540-f4-power-amplifier-463.html#post6311086). Based on this recommendation, I purchased and matched some Harris parts for a new premium build of the F4.

Due to a large minimum order quantity at Rochester Electronics, I currently have 34 extra if anyone would like a matched set (50 currently under reservation). If so, please PM me. They have been matched at a constant temperature of 50C and at the F4 operating characteristics of 22V and .42A. I'm selling at cost which is around $3 per mosfet + shipping.

Pictures of the testing rig below.

 

[Rodeodave]

Are those Harris so much better than the „original“ ones? (Worth a swap)?

See page 9 of http://www.firstwatt.com/pdf/art_matching.pdf

 

[Stanislav]

How do you get to page 9 in a two page pdf file?

 

[Rodeodave]

That's a good question.

I was linking to the wrong article, sorry. I can't find the link to the one I had in mind, so I attached it below.

Edit: Found the link too: http://www.firstwatt.com/pdf/art_mos_test.pdf

 

[myleftear]

Thanks for the link.
I was thinking about a swap (now the there were a little batch available again, but I interpret page 9 as „both would do fine“ — and by lack of knowledge, I could’ve be more precise than that ...