Hi Bricolo,
I do not know the Aleph CS either by reputation or schematic. I have seen NP say here that the sliding bias arrangement (to use a PP terminology, rather than my preferred term 'variable current source') does not sound as good as a constant, high current CCS. This seems to be about right given that in all engineering more of this inevitably means less of that. And for those who have used the Rane EQ, balancing 'this' against 'that' is fraught with enormous difficulties.......
The topology of a follower circuit perched atop a variable CCS is fascinating. My essential philosophical point is that the signal voltage is solely controlled from the follower, along with positive current to the load, while negative current flow from the load to the negative rail is controlled by the CCS. This is quite a 'wiggle' factor, certainly, and to me would seem (at least for Class AB) like a very different configuration to a full complementary output stage or even a quasi-complementary stage, where signal voltage is alternately controlled from each side of the output stage while the inactive side falls comatose as the other side springs to life.....
I suspect 'bad sound' of the follower/variable CCS is not because of the control servo for the variable CCS, but rather because of high distortion artefacts introduced by base/emitter [gate/source] variations in the follower. Where no global feedback is used, these distortions are significant. I believe there may be considerable damage done to the music if the servo is not accurate, or overshoots; the design of such a servo would involve great speed and considerable precision.
Sadly I have nothing more to say on this; but I am confident there are others here who could add to this discussion. It is an attempt to debate the creation of SE sonics in a Class AB PP stage - a lofty goal, certainly, but one which just might be the holy audio grail providing it could be adequately marketed. 😎
Cheers,
Hugh
I do not know the Aleph CS either by reputation or schematic. I have seen NP say here that the sliding bias arrangement (to use a PP terminology, rather than my preferred term 'variable current source') does not sound as good as a constant, high current CCS. This seems to be about right given that in all engineering more of this inevitably means less of that. And for those who have used the Rane EQ, balancing 'this' against 'that' is fraught with enormous difficulties.......
The topology of a follower circuit perched atop a variable CCS is fascinating. My essential philosophical point is that the signal voltage is solely controlled from the follower, along with positive current to the load, while negative current flow from the load to the negative rail is controlled by the CCS. This is quite a 'wiggle' factor, certainly, and to me would seem (at least for Class AB) like a very different configuration to a full complementary output stage or even a quasi-complementary stage, where signal voltage is alternately controlled from each side of the output stage while the inactive side falls comatose as the other side springs to life.....
I suspect 'bad sound' of the follower/variable CCS is not because of the control servo for the variable CCS, but rather because of high distortion artefacts introduced by base/emitter [gate/source] variations in the follower. Where no global feedback is used, these distortions are significant. I believe there may be considerable damage done to the music if the servo is not accurate, or overshoots; the design of such a servo would involve great speed and considerable precision.
Sadly I have nothing more to say on this; but I am confident there are others here who could add to this discussion. It is an attempt to debate the creation of SE sonics in a Class AB PP stage - a lofty goal, certainly, but one which just might be the holy audio grail providing it could be adequately marketed. 😎
Cheers,
Hugh
Seeking information
Hi,
So much has been spoken about the various amplifiers and its performances and characteristics in this thread. Kindly if any of you experienced guys list out the links of the diy amplifiers which are being repeatedly talked about will of great help to the aspirants like me to just try out few of those things.
Thanks
Regards,
avguy
Hi,
So much has been spoken about the various amplifiers and its performances and characteristics in this thread. Kindly if any of you experienced guys list out the links of the diy amplifiers which are being repeatedly talked about will of great help to the aspirants like me to just try out few of those things.
Thanks
Regards,
avguy
If I'm not mistaken, what you're describing is a variable bias SE class A with a current source
the bias being set by the current sent to the load. In exemple twice the load current
But I don't see big differences between this, and an Aleph (Hugh, you can find the schematics on http://www.passlabs.com/aleph.htm )
Maybe someone can explain me
BTW, is the JLH96's current source constant, or have I looked too fast on the schematic?
the bias being set by the current sent to the load. In exemple twice the load current
But I don't see big differences between this, and an Aleph (Hugh, you can find the schematics on http://www.passlabs.com/aleph.htm )
Maybe someone can explain me
BTW, is the JLH96's current source constant, or have I looked too fast on the schematic?
Re: Seeking information
Hello there,
You can first try to seek the information you want from the "SEARCH" button of this forum.
Regards,
Chris
avguy said:Hi,
So much has been spoken about the various amplifiers and its performances and characteristics in this thread. Kindly if any of you experienced guys list out the links of the diy amplifiers which are being repeatedly talked about will of great help to the aspirants like me to just try out few of those things.
Thanks
Regards,
avguy
Hello there,
You can first try to seek the information you want from the "SEARCH" button of this forum.
Regards,
Chris
AKSA said:
Perhaps I am misunderstood on this. I see variable bias as
referring to push-pull follower stages and similiar complementary
push pull stages and used to reduce distortion and/or extend
the class A operating envelope.
The Aleph is a variable current source which creates a negative
impedance image of the load, increasing the real and imaginary
impedance of the load by an arbitrary inverse factor from 0
(which creates a constant current source) to 1 (which makes
the load seem to go away entirely). You can pick the number
arbitrarily, but .5 seems to give the best all round compromise
between efficiency, specs, and subjective sound. The subjective
part of that, is of course, totally the perception of the listener.
NP,
I shall have to think about this for a bit - it's a tad cryptic.....
What is an 'arbitrary inverse factor'???
Cheers,
Hugh
I shall have to think about this for a bit - it's a tad cryptic.....
What is an 'arbitrary inverse factor'???
Cheers,
Hugh
I'm with Hugh on this, waving my hands in the air to clear away the smoke. I am bemused by the negative CS theory. After all, if you can design a "CS" to act as the negative of the load Z then isn't the "CS" then effectively the amplifier? Isn't is just as hard to make a "CS" negate the load Z as it is to make a voltage source drive the load in the first place? 
Hugh and Traderbam
I think the best explanation is found in Nelsons US-patent. It has the number 5,710,522
Type in the above number at http://patft.uspto.gov/netahtml/srchnum.htm
This will lead you to the document.
Edit: To view the images in the document you will probably have to install a freeware program which can be found at http://www.alternatiff.com/
Torben
I think the best explanation is found in Nelsons US-patent. It has the number 5,710,522
Type in the above number at http://patft.uspto.gov/netahtml/srchnum.htm
This will lead you to the document.
Edit: To view the images in the document you will probably have to install a freeware program which can be found at http://www.alternatiff.com/
Torben
Thanks for the link, Herrmann.
As I see it the invention can be summarised as a push-pull output with either the pull or the push being driven off the output current rather than the output stage to output voltage difference. The benefit is claimed in reducing power dissipation (which is easy to see, just like any push-pull arrangment) and some performance due to the primary amplifier exerting a lesser output current swing (which is highly dubious IMO as it appears to neglect the performance impairments of the active CS).
A conventional push-pull circuit is essential two active CSs anyhow. Why you'd want to drive one ****-about-tit from the output current isn't obvious to me.
As I see it the invention can be summarised as a push-pull output with either the pull or the push being driven off the output current rather than the output stage to output voltage difference. The benefit is claimed in reducing power dissipation (which is easy to see, just like any push-pull arrangment) and some performance due to the primary amplifier exerting a lesser output current swing (which is highly dubious IMO as it appears to neglect the performance impairments of the active CS).
A conventional push-pull circuit is essential two active CSs anyhow. Why you'd want to drive one ****-about-tit from the output current isn't obvious to me.
Herrmann said:
Err... Ahum... The purpose of a patent generally is to make money from your idea and/or to prevent others to do so. The purpose of a patent is NOT to clearly and unambiguously explain your idea in engineering terms.
Jan Didden
The way I understand the Aleph topology (or the way I understood it when I built my Mini-A) is like this:
- there is a constant DC current source
- this current source is AC modulated
- the degree of AC modulation is governed by a network sensing output current
- therefore the current source has an AC current gain, proportional to AC in output, that can go from 0 to 100% (usual value 50%)
- that means for a given bias current you get more load drive than you would, had it been constant DC bias
- or, for a given output power, you save on constant DC bias
- for "50% AC current gain" setting, half the output current "goes away", meaning, you get twice the AC output current to your real load than you would expect from the value of the constant DC current source.
You can call it AC current gain modulated by 50% of load, or you can call it AC load reduction (negative impedance) by 50%.
I can't model or calculate the circuit, this is just my understanding of the principle. I may be wrong, but my Mini-A is happily heating away... 😉
- there is a constant DC current source
- this current source is AC modulated
- the degree of AC modulation is governed by a network sensing output current
- therefore the current source has an AC current gain, proportional to AC in output, that can go from 0 to 100% (usual value 50%)
- that means for a given bias current you get more load drive than you would, had it been constant DC bias
- or, for a given output power, you save on constant DC bias
- for "50% AC current gain" setting, half the output current "goes away", meaning, you get twice the AC output current to your real load than you would expect from the value of the constant DC current source.
You can call it AC current gain modulated by 50% of load, or you can call it AC load reduction (negative impedance) by 50%.
I can't model or calculate the circuit, this is just my understanding of the principle. I may be wrong, but my Mini-A is happily heating away... 😉
degree of difference
Only saw this thread last night, so while a number of rankings are already here:
Rather than just rank them, could they be rated, eg out of 10, to get a sense of the *degree of difference* in quality
Only saw this thread last night, so while a number of rankings are already here:
Rather than just rank them, could they be rated, eg out of 10, to get a sense of the *degree of difference* in quality
How about a ranking that includes commercial amps too? I'm very interested in just how good amateurs creations are when put to the acid test with the big boys.
For example, do you own a commercial amp and have you built something that most people would say sounds better?
Can anyone give some commercial calibration to the AKSA or Randy Sloane designs?
For example, do you own a commercial amp and have you built something that most people would say sounds better?
Can anyone give some commercial calibration to the AKSA or Randy Sloane designs?
i do not own any commercial amp, did never, and as a result my Zens are the best sounding i know.
Updated Ranking
Finally got round to firing up the Aleph:
Power Amps:
1) AKSA 100N (slightly modified)
2) AKSA 55N/P61 (Modified)
3) Aleph 5 (Aleph 30 input stage)
4) P3A with Sanken o/p trannies, dual mono etc
5) JLH 80W Mosfet amps with improved low current regs and unregulated high current supplies, dual mono.
6) About 20 variations (including my own variations) of the Borbely 60W mosfet/DC100 amps (not built on official boards).
7) Mini Crescendo
8) Maplin Mosfet amps
15) TDA1541
Finally got round to firing up the Aleph:
Power Amps:
1) AKSA 100N (slightly modified)
2) AKSA 55N/P61 (Modified)
3) Aleph 5 (Aleph 30 input stage)
4) P3A with Sanken o/p trannies, dual mono etc
5) JLH 80W Mosfet amps with improved low current regs and unregulated high current supplies, dual mono.
6) About 20 variations (including my own variations) of the Borbely 60W mosfet/DC100 amps (not built on official boards).
7) Mini Crescendo
8) Maplin Mosfet amps
15) TDA1541
- Status
- Not open for further replies.