Aksa Lender P-MOS Hybrid Aleph (ALPHA) Amplifier

Member
Joined 2014
Paid Member
I’m pretty much where your at. Although, I’m going the smps, cap Mx route for the 20w’er. I’m thinking a chassis similar to a tube amp with the cpu coolers taking the place of the output/power trafo’s on the topside with the Alpha board mounted upside down in the enclosure. I’d have to figure out if fans are needed and how to make it all work aesthetically.
 
Hi X,

The Big Boy (BB) will use the FDA69N25 and IXTH48P20P, not the three paralleled IRFP devices?

Not too happy about that, maybe I will do my layout (of course for own use only) to test and build that amp. :);)
Reason I know the IRFP mosfets, they being used in so many HI-FI amplifier projects and I have them at home ready to use
 
Last edited:
Founder of XSA-Labs
Joined 2012
Paid Member
Yes, as Hugh recommends this for better performance without having to match output devices. The cost may actually be a little less to go with single devices. It also makes fitting CPU coolers easier.

I also have a bunch of IRFP’s in hand already - but the Big Boy Alpha is a special amp. Not too many 50W+ full SE Class A’s out there. So let’s do it right, per recommendation by Hugh.
 
Last edited:
I’m pretty much where your at. Although, I’m going the smps, cap Mx route for the 20w’er. I’m thinking a chassis similar to a tube amp with the cpu coolers taking the place of the output/power trafo’s on the topside with the Alpha board mounted upside down in the enclosure. I’d have to figure out if fans are needed and how to make it all work aesthetically.

That is a good idea, was thinking along those lines, I'm currently trying to print an 'as close to life size' copy of the board layout so I can play with layout with the heatsinks, etc.
 
Founder of XSA-Labs
Joined 2012
Paid Member
Not too happy about that, maybe I will do my layout (of course for own use only) to test and build that amp. :);)
Reason I know the IRFP mosfets, they being used in so many HI-FI amplifier projects and I have them at home ready to use

JPS64 and I considered making PCB with parallel positions for optional triple parallel MOSFETs. But the challenges of getting CPU cooler pads big enough to fit three and need for matching them, we thought to keep it simple as better plan. It might be possible to add two sets pads to the side of central one to accommodate this - let me check with JPS64 to see how far along he is.
 
JPS64 and I considered making PCB with parallel positions for optional triple parallel MOSFETs. But the challenges of getting CPU cooler pads big enough to fit three and need for matching them, we thought to keep it simple as better plan. It might be possible to add two sets pads to the side of central one to accommodate this - let me check with JPS64 to see how far along he is.

I would be not interested to use CPU cooler type of heatsink at all. That is a BIG no to me! I always use a regular heatsink. If you listen to the music at the low volume you can hear the fan noise. I have an HP 4 years old computer, was top of the line when I purchased, it uses low noise fan, I clean the dust every 6 months still I hear the fan noise. At the 52Watt power amp, you have to run the fan at full speed - so you get noise. :eek: I know is cheaper to use air forced cooling but you have to deal with the noise and risk if the fan stops you burn everything. Yes, you can use protection etc and you end up spending the same amount of money plus get the noise. That is my opinion. :)
 
Hmm, if you try to use one large heatsink for even just one channel, you must move a dissipation of 3pairx0.98currentx72total rail=212 watts. If you use a 350x100 heatsink with vertical fins (the best I know is the Conrad), the thermal capacity is 0.34C/watt.

So, 212W will correspond to 72.1C above ambient. If ambient is a highish 30C (warm day in Oz, very hot in cooler countries but less than a tropical climate), this is 102.1C total, which is far too hot for electronics. You have to have the heatsink no more than 50C, which is no less than 52.1C too high for comfort........ You'd need TWO of these very large heatsinks for one channel. That's why we consider Big Boy a Monster and to be frank only liquid cooling can move this sort of heat.

The 20W ALPHA is OK with this very large heatsink. It will be hot, but acceptable with a rise of 25.5C above ambient for 75W dissipation.

These heat issues of big Class A amps are very difficult to overcome.

Cheers,

Hugh
 
So I scrubbed through the schematic and got most of it captured into a bom, just digikey part numbers at the moment.

I subbed a few resistors, 22.1k for 22k, etc.

Also there's a few missing parts, a couple caps and connectors, etc. Anyone feel free to add to it. If I get some time tomorrow I'll add Mouser part numbers.


As for C1021 (C102 substitute), I plan on using:

ClarityCap 10 mfd CSA Range with CopperConnect Technology

bypassed with:
940C30S1K-F Cornell Dubilier Electronics (CDE) | Capacitors | DigiKey


I like that combo, and look forward to testing it out.

and for R131, R132 I plan on using:
0.12 Ohm Mills 5 watt MRA-5, Sonic Craft
0.33 Ohm Mills 5 watt MRA-5, Sonic Craft


A bit of a start at least.
 

Attachments

  • alpha_bom.zip
    8.1 KB · Views: 124
Founder of XSA-Labs
Joined 2012
Paid Member
Thanks for putting this together Pcgab. Those input caps look like a very reasonably priced but very high end performance. I might try them. Never thought of using a kV rated PP bypass cap before. Does that give it better attack on transients?

The Mills resistors at the big 3W locations look good as well. I found that cheap resistors here can make a difference with distortion.

Thanks again.
X