Aleph-X asymetric waveform
I'll pop a few resistors out and measure them (but I did already measure them prior to stuffing the boards). The input FETS are definitely matched, but again I could try swapping them out for another matched pair.
I'll pop a few resistors out and measure them (but I did already measure them prior to stuffing the boards). The input FETS are definitely matched, but again I could try swapping them out for another matched pair.
Aleph X Boards
Still looking for a set of Aleph X boards. My contact disappeared. When I asked about the boards after seeing his pictures of them.....no more correspondence. I would love to build a stereo 100 watt system. I have read most of the threads. Anyone interested in trying to have some more boards built?
Thanks
Still looking for a set of Aleph X boards. My contact disappeared. When I asked about the boards after seeing his pictures of them.....no more correspondence. I would love to build a stereo 100 watt system. I have read most of the threads. Anyone interested in trying to have some more boards built?
Thanks
Pass DIY Addict
Joined 2000
Paid Member
Yep, they are! Here is my trio of 100w monoblocks. For reference, the speaker behind them is nearly 5 feet wide and the silver woofers are 12" in diameter.
Keep an eye out for boards, they show up every few weeks or so. Over 700 boards were made, but it's been over a decade...
I haven't actually followed recent/not so recent developements but, you may also be interested in Peter Daniels AlephJ boards wich come with a bridging board to do an AlephJ-X
Hi,
I am Jean-Claude, living in germany (info to excuse my englisch) and since the beginning of the aleph-x project I have already built up 3 Aleph-X for me an two friends on own PCBs (3 IRFPs in parallel). All running very well (and very hot).
Generally I like building up own concepts, and not building up 1:1. And since the Aleph-design is quite old and not optimized for the X, I hope there is some room for improvement.
Okay, now I want to build up the "real stuff". The existing one was a bit economic since I dont knew how good it will be.
I have some modifications in mind:
a) filter the PS for the positive input stage. Like on the ZEN, with a IRF610. Loss of 4V is not relevant there.
b) modify CCS. Since its a X, more C. perhaps a bipolar. No 9.1V ZD due to its positive TK, to my experiences with drift, a negative TK is needed.
c) Mod a) allows to feed DC-feedback to source of the CCS Transistor instead of drain. This does not affect the common-mode rejection like the original version. With clean DC there should be no hum with this method.
d) replace the IRFs by 2SJ74. 3 or 4 in parallel.
e) the wildest idea. The JFET input stage will loose around 15dB of open loop gain. not a problem with distortion since the Jfets are better. But I expect a significant drop in damping factor, a thing I dislike in my existing am (its ~60). There is some negative feedback of owners of the Aleph-J. So i like to bring back the open loop gain by
- reducing Source resistors from 0r47 to 0r22 in the output stage
- replace the load resistors of the input stage by fine-tuned CCS.
Whats your optinion? Has anyone tried such things? Will it work? In PSpice it will, with very good data.
Regards, J-C
I am Jean-Claude, living in germany (info to excuse my englisch) and since the beginning of the aleph-x project I have already built up 3 Aleph-X for me an two friends on own PCBs (3 IRFPs in parallel). All running very well (and very hot).
Generally I like building up own concepts, and not building up 1:1. And since the Aleph-design is quite old and not optimized for the X, I hope there is some room for improvement.
Okay, now I want to build up the "real stuff". The existing one was a bit economic since I dont knew how good it will be.
I have some modifications in mind:
a) filter the PS for the positive input stage. Like on the ZEN, with a IRF610. Loss of 4V is not relevant there.
b) modify CCS. Since its a X, more C. perhaps a bipolar. No 9.1V ZD due to its positive TK, to my experiences with drift, a negative TK is needed.
c) Mod a) allows to feed DC-feedback to source of the CCS Transistor instead of drain. This does not affect the common-mode rejection like the original version. With clean DC there should be no hum with this method.
d) replace the IRFs by 2SJ74. 3 or 4 in parallel.
e) the wildest idea. The JFET input stage will loose around 15dB of open loop gain. not a problem with distortion since the Jfets are better. But I expect a significant drop in damping factor, a thing I dislike in my existing am (its ~60). There is some negative feedback of owners of the Aleph-J. So i like to bring back the open loop gain by
- reducing Source resistors from 0r47 to 0r22 in the output stage
- replace the load resistors of the input stage by fine-tuned CCS.
Whats your optinion? Has anyone tried such things? Will it work? In PSpice it will, with very good data.
Regards, J-C
Hi Jean Claude,
Most of what you suggest is already part of the Aleph JX (towards the end of the Aleph J thread.)
I don't know what you hope to gain changing the Source resistors to 0R22. The uppper ones set the bias (coarsely), so you'd have roughly double the bias per fet. The lower ones are isolated from the ouput by the output gain devices. If you want to reduce just the net value of the current sense resistors, you can compensate for that by changing the other resistors in the active gain setup - R8, R37 (1K) and R12 R34 (1K2) on Grey's 5/24/02 schematic (pre MacMillan resistors, but the first one I found). They are inside the feedback loop, so that's not going to buy you much. This schematic uses 0R22 source resistors with 15V rails - just be sure you can get rid of the heat.
Hope this helps.
Bob
Most of what you suggest is already part of the Aleph JX (towards the end of the Aleph J thread.)
I don't know what you hope to gain changing the Source resistors to 0R22. The uppper ones set the bias (coarsely), so you'd have roughly double the bias per fet. The lower ones are isolated from the ouput by the output gain devices. If you want to reduce just the net value of the current sense resistors, you can compensate for that by changing the other resistors in the active gain setup - R8, R37 (1K) and R12 R34 (1K2) on Grey's 5/24/02 schematic (pre MacMillan resistors, but the first one I found). They are inside the feedback loop, so that's not going to buy you much. This schematic uses 0R22 source resistors with 15V rails - just be sure you can get rid of the heat.
Hope this helps.
Bob
Last edited:
Hi,
thanks for reply.
For sure has the Source resistor of the gain devices an influence of the gain. Its a local feedback where the source resistor reduces the transconductance of the gain device.
Since I have a DC-servo for common-mode DC, I think I can also live with 0R22.
The common-mode-feedback-resistors are not used in the Aleph-J, This is a only Aleph-X topic. And a not so fine soulution too, since higher values give a better sound quality. So there is the idea to get rid of them...
Annother idea: what about using an output stage from the F3? Is the LU1014 still available? There is a problem with the absolute gate voltage, but when this problem is solved by a kind of level-shifting, there may result a nice amplifier
And the CCS-loads are really discussed in the J-thread? I fear I have to read all...
Regards, J-C
thanks for reply.
For sure has the Source resistor of the gain devices an influence of the gain. Its a local feedback where the source resistor reduces the transconductance of the gain device.
Since I have a DC-servo for common-mode DC, I think I can also live with 0R22.
The common-mode-feedback-resistors are not used in the Aleph-J, This is a only Aleph-X topic. And a not so fine soulution too, since higher values give a better sound quality. So there is the idea to get rid of them...
Annother idea: what about using an output stage from the F3? Is the LU1014 still available? There is a problem with the absolute gate voltage, but when this problem is solved by a kind of level-shifting, there may result a nice amplifier
And the CCS-loads are really discussed in the J-thread? I fear I have to read all...
Regards, J-C