Hello diyaudio
I've taken a keen interest to Class-A amplifiers and particularly Nelson Pass's designs. I'm an electronics amateur but linear amplification is still a very new topic to me. I'm eager to dig in to a build like the Aleph-X but first I'm a bit hung up on the heat and power consumption issue. I want to build a unit with a lower design power than Grey's AX and I thought I should pick your brains since I'm so inexperienced myself.
The 40W output of Grey's design is reasonably high, and the 135W TDP is manageable, but if I want to build a low wattage multi-way stereo, or power some bookshelf speakers in the 5W neighborhood, the design heat is excessive. However it seems the Grey design hit a popular balance so I'm wary to vary from those voltage and bias numbers. I don't want to risk mucking with the design without a little basic consulting first, I fear making a poor amp that my ears couldn't tell the difference
1. I've read Eric's Home Theater post on the Aleph-X and I think I have a decent grasp on the principles. However the wattages I'm interested in weren't really discussed, opting for higher numbers and I don't know if I'm treading into making a lower quality design by mistake. For starters, I think based on William's spreadsheet, I could easily build a setup for a 5W Aleph-X using 18V transformer, 1A bias. Then with a ~1K/W heatsink I'd be sitting pretty at 62C, 40W. I'm sure others have done like this, they haven't got any issues? No negative effect of using much less than the 4.5A bias current of Grey's design?
2. Going any lower than about 8V across each IRFP240 looks like you'll run into issues with gate drive and linearity. If my design is looking for a really low output power, say 1W or 0.1W for desktop speakers or headphones, is the best play to simply lower my bias current while keeping voltage 15V and above? I'd like to lower those heat dissipation numbers and make the heatsinks, and wattage smaller. Is a bias current of 0.5A or even 0.1A feasible, provided I can keep transistor temps above 50C?
Thank you for your time.
I've taken a keen interest to Class-A amplifiers and particularly Nelson Pass's designs. I'm an electronics amateur but linear amplification is still a very new topic to me. I'm eager to dig in to a build like the Aleph-X but first I'm a bit hung up on the heat and power consumption issue. I want to build a unit with a lower design power than Grey's AX and I thought I should pick your brains since I'm so inexperienced myself.
The 40W output of Grey's design is reasonably high, and the 135W TDP is manageable, but if I want to build a low wattage multi-way stereo, or power some bookshelf speakers in the 5W neighborhood, the design heat is excessive. However it seems the Grey design hit a popular balance so I'm wary to vary from those voltage and bias numbers. I don't want to risk mucking with the design without a little basic consulting first, I fear making a poor amp that my ears couldn't tell the difference

1. I've read Eric's Home Theater post on the Aleph-X and I think I have a decent grasp on the principles. However the wattages I'm interested in weren't really discussed, opting for higher numbers and I don't know if I'm treading into making a lower quality design by mistake. For starters, I think based on William's spreadsheet, I could easily build a setup for a 5W Aleph-X using 18V transformer, 1A bias. Then with a ~1K/W heatsink I'd be sitting pretty at 62C, 40W. I'm sure others have done like this, they haven't got any issues? No negative effect of using much less than the 4.5A bias current of Grey's design?
2. Going any lower than about 8V across each IRFP240 looks like you'll run into issues with gate drive and linearity. If my design is looking for a really low output power, say 1W or 0.1W for desktop speakers or headphones, is the best play to simply lower my bias current while keeping voltage 15V and above? I'd like to lower those heat dissipation numbers and make the heatsinks, and wattage smaller. Is a bias current of 0.5A or even 0.1A feasible, provided I can keep transistor temps above 50C?
Thank you for your time.
You might want to have a look at this thread instead. Stuff the board with as many devices as needed.
Beast with 1000 JFETs redux ?
Beast with 1000 JFETs redux ?
I agree that you have to choose your speakers/headphones first. Also, what genre of music will you be listening to? Dub Step will have different requirements than small group light jazz.
Thank you for the replies.
In a 3-way setup, I want to push Peerless NE85W midwoofers (85dB, 4ohm) at about 20W, and DA25TX tweeters (89dB, 8ohm) at ~6W to match output dB. Class D subwoofer for lows. I know the match could be better but it's what I've got and I'm going with it. The tweeters are the reason I am asking about lower bias/lower output setups, I don't want to dissipate 100W+ for a 6W peak. I will be listening to mostly rock in digital format.
I think I want to build an Aleph X, can a bias around 1A produce good results?
In a 3-way setup, I want to push Peerless NE85W midwoofers (85dB, 4ohm) at about 20W, and DA25TX tweeters (89dB, 8ohm) at ~6W to match output dB. Class D subwoofer for lows. I know the match could be better but it's what I've got and I'm going with it. The tweeters are the reason I am asking about lower bias/lower output setups, I don't want to dissipate 100W+ for a 6W peak. I will be listening to mostly rock in digital format.
I think I want to build an Aleph X, can a bias around 1A produce good results?
...I don't want to dissipate 100W+ for a 6W peak...
That's exactly why I created a low power version of a DIY supersymmetric amp project that is based on Pass Labs X series of amplifiers. My amp has a single pair of MOSFETs per output phase biased at 500mA or so with 15V rails, so a total of 30W dissipation or channel. With no source degeneration on the output FETs, that's good enough for a couple watts in Class A. Can output up to 25W or so in Class AB, depending on some part choices when building. When designing the PCB, I included options to try out all kinds of asymmetric feedback, single ended bias, and such, but I've been happy enough with the stock X type circuit for over a year now with a wide variety of speakers, so I've never done much experimenting. See my thread for other details.
Make a flea-watt class A amp for the tweeter, and use a discrete active xo from Nelson and the shop:
DIY Biamp 6-24 Crossover – diyAudio Store
DIY Biamp 6-24 Crossover – diyAudio Store
Thank you all for the helpful suggestions. I'll get back to you when/if I have made some progress.
- Home
- Amplifiers
- Pass Labs
- Micropower Class A?