Time to revisit Aleph J. I've been pondering this for a while, one of my new aims is to keep single ended throughout the chain (aim: unadulterated phase, so no push-pull M2, F4 or F6 designs). Options available to hod rod the Aleph J to provide more power into 4 ohms as I see it are:
- Lower supply voltage, say 20v, increase bias to standard Aleph J. More into 4ohms, but power into 8 ohms is down to 20w.
- Use Hughs/XRKs Alpha Nirvana instead - front-end is based on the Aleph J, but the output stage is Single-Ended Push-Pull. Good for 34W into 2.1ohms loads according to this post. Disadvantage/advantage, will sound different to the original Aleph J (Nirvana has a higher damping factor), little bit of phase shift of only 5.3deg at 20kHz.
- Build standard Aleph J, but use one less output device per side (?) and/or reduce bias and use the Aleph J as a driver for a BA1 single ended output stage. Trim voltage of BA1 to 20-22v and bias as high as chassis/mosfets will take. Limits ultimate volume level into 8ohm loads, but should breeze through lower impedance loads.
- Use double die output transistors with standard Aleph J (like IRFP150?) - is this an option, anyone done it?
No can do for parallel. My aim is single ended only- so no bridged amps. Not sure series would work - you’d blast the socks off the input sensitivity of the final Aleph J and limited final output still… has someone actually done this?
Build bigger Aleph - but targeting latest Aleph J incarnation and SE only. If I can get to 25w into 4ohms then I’d be happy.
for more powah on tougher loads, you need more Iq
means, more heatsinks, more PSU, more output mosfets
so, either build bigger one with same rails as Aleph J, or build second Aleph J and connect two channels in parallel, practically doubling W in 4R load
already done, and not just with Aleph J
means, more heatsinks, more PSU, more output mosfets
so, either build bigger one with same rails as Aleph J, or build second Aleph J and connect two channels in parallel, practically doubling W in 4R load
already done, and not just with Aleph J
I always miss the obvious... partly due to getting lost gleaning ideas from the forum.
So, to ask, based on this Wikipedia article:
1. I'd need to make sure both Aleph J's DC offset was as close to zero possible
2. I might need a offset nulling circuit to each amplifier?
3. I can check and match the components on the gain stage, so there is no gain difference.
4. Do I need output resistors or consider current drive.
All new to me!
So, to ask, based on this Wikipedia article:
1. I'd need to make sure both Aleph J's DC offset was as close to zero possible
2. I might need a offset nulling circuit to each amplifier?
3. I can check and match the components on the gain stage, so there is no gain difference.
4. Do I need output resistors or consider current drive.
All new to me!
Found useful info from Planet10 over in the ACA forum:
"In parallel mode, you halve the output impedance making it more suitable for speakers that expect a low output impedance from the amp. Better into low impedance loads. The boards are in phase so the amp retains its SE character, but may suffer from some low level “blurring” of details which inevitably happens unless the 2 board shave exactly the same transfer curve."
Will be interested in your experiences and what I need to consider in post #10.
Would a lower biased Aleph J, potentially reducing number of output devices as well (if possible) + BA1 as a follower not be more elegant?
"In parallel mode, you halve the output impedance making it more suitable for speakers that expect a low output impedance from the amp. Better into low impedance loads. The boards are in phase so the amp retains its SE character, but may suffer from some low level “blurring” of details which inevitably happens unless the 2 board shave exactly the same transfer curve."
Will be interested in your experiences and what I need to consider in post #10.
Would a lower biased Aleph J, potentially reducing number of output devices as well (if possible) + BA1 as a follower not be more elegant?
Last edited:
Thanks, familiar with this - but it doesn't really address the technical points in post #10? Although it does indicate that #4 output resistors or consider current drive might not be needed. F4/BA1 has no gain, but Aleph J does. I do like to overthink these things
Aleph 2 is old/original version, lacks some of the circuit improvements that Pa made to the J I believe?
A2 is 100W output into 8R. It's using mosfet front end vs. AJ JFET front end. And of course the A2 has more outputs and higher rail voltage.
It has been a LONG time since I've compared AJ to A3 / A30 in an A/B test. I know both are fantastic. I enjoy my A30's for hours daily. The original factory A3 is what converted me from 300B amps to Pass/Solid state.
The Aleph circuit is, in general, scalable. I don't know how high the AJ can be scaled up. It need higher voltage PSU rails, more outputs, etc. I don't know when the JFET "runs out of gas". The mosfet based front end is good for the 200W Aleph 1.2 and maybe beyond.
Aleph 60 could be a good middle ground.
It has been a LONG time since I've compared AJ to A3 / A30 in an A/B test. I know both are fantastic. I enjoy my A30's for hours daily. The original factory A3 is what converted me from 300B amps to Pass/Solid state.
The Aleph circuit is, in general, scalable. I don't know how high the AJ can be scaled up. It need higher voltage PSU rails, more outputs, etc. I don't know when the JFET "runs out of gas". The mosfet based front end is good for the 200W Aleph 1.2 and maybe beyond.
Aleph 60 could be a good middle ground.
