I'm just making my way though this fantastic thread, last night I found (and bookmarked) this...
http://www.diyaudio.com/forums/showthread.php?postid=749245#post749245
Nelson Pass - 10-21-2005 09:57 PM said:JFETs on the front end was not the variation I had in mind.
As far as I'm concerned, that's such an easy one that you can
do it on your own - Just plug a 2SJ109 in where the mosfets
were. I don't think that will get us another 90,000 F5's.
http://www.diyaudio.com/forums/showthread.php?postid=749245#post749245
steenoe said:
Thats right, Holger. What would be more interesting is how that dual fet was implemented😀 I dont know if the P-channel frontend can be implemented in the Aleph-X circuit through plain surgery or not, but some of the more experienced A-X DIY'ers around here has tried. If Wuffwaff, Netlist, and not the least: Patrick (EUVL, who is one clever dude, indeed!!) and some others cant figure it out, who can??
I guess most of us, has thrown in the towel long ago
As far as I am concerned, a few breadcrumbs is needed😀 Grey mentioned that things has to looked at in the correct way. Cant agree more!! If you have a J-fet A-X, by all means lets have a look😉 I might build the proto...he-he
Steen🙂
call me ignorant......I lost thread.......what's the question?
I guess, to sum it up: Will the Babbel-X work?? Do mind, that I have a scope now.... He-he.. And that was no succes, according to the J-fet-X-BosoZ!!
Steen🙂
Steen🙂
Yes, the 2SJ109 can be plugged in, but only up to a point. You run into current limitations depending on how many output devices you want to drive, and voltage limitations as you increase the rails. Then there's the power dissipation problem...
The rather obvious first step is to cascode. That will help with voltage and power dissipation problems and reduce the effect of the capacitance. But to go beyond that takes a reimagining of the circuit--where you've come from and where you want to go with it.
In other words, it's not scalable, at least not to the extent that the original Aleph-X is. With a few simple resistor changes, the original circuit can cover pretty much everyone's needs, from the stock 40W or so, on up to hundreds of watts, assuming that people are willing to throw enough heatsinks and power supply into the equation.
Will the 2SJ109 work? Yes. But be aware of the limitations.
Grey
The rather obvious first step is to cascode. That will help with voltage and power dissipation problems and reduce the effect of the capacitance. But to go beyond that takes a reimagining of the circuit--where you've come from and where you want to go with it.
In other words, it's not scalable, at least not to the extent that the original Aleph-X is. With a few simple resistor changes, the original circuit can cover pretty much everyone's needs, from the stock 40W or so, on up to hundreds of watts, assuming that people are willing to throw enough heatsinks and power supply into the equation.
Will the 2SJ109 work? Yes. But be aware of the limitations.
Grey
steenoe said:
last first- you need to place more potent jfets in jfetxbosoz ( ya remember our PMs) ............or to decrease current through present ones to -say- 6mA per each jfet ( to achieve symetrical clipping); and there is another problem, 'couse I have impression that IRFs can't be entirely happy with just 6mA........
besides that- scope and siggene will tell ya what's most happy amount of Uds for needed swing.........I mean on shifting Uds of cascode mosfets to achieve most graceful clipping
anyway- you are on beginning of funny game...... you have scope and siggene now.......
now first:
Babel x must work ; even ifI didn't tried it yet
up to +/- 25V without cascodes
above that cascodes on 2SJ109........ (there are few other ways to avoid cascode,but why.....)
in both cases up to two gates driven with half 2sj109
little mumbo jumbo regarding McMilans, compensating caps etc must be tested.......
I have plain Aleph X pcbs- gift from AR2 ,but not 'nuff time 'till this days to try
in fact- that can be just a little step for me- comparing to solving Babelfish J puzzle .......that was a challenge for me in that moment;
frankly , Babelfish JX isn't on par with his predecessor as a riddle...
Thanks guys🙂 I have a spare set of A-X boards, so I might try the J-fet-version with lower railvoltages than the ones I have now and take it from there.
Steen🙂
BTW Blues;
Steen🙂
BTW Blues;
Also an Albert King line from the song "Born under the bad sign" 😉
steenoe said:
Song written by Booker T. Jones/William Bell...so I give the credit to them🙂 Albert King gets the credit for lending his voice and turning it into a hit...so with the next generation of Blues artists.
Excuseme! I'm from Thailand. Now i'have 40W Aleph-X. and i want to modify it. i want to know if i'm build Aleph-X amp can i use +-34 volt supply for 100w Aleph ? Thanks for your help
Pass DIY Addict
Joined 2000
Paid Member
My first reaction is that it might be easier to start over rather than "convert" an existing 40w amp. You haven't provided much detail, so I would assume your current transformer and heat sinks are too small to make the jump to 100w. There are also a number of differences in the resistor values of the circuit between 40w and 100w output, and you will need more output transistors that are carefully matched.
While there are certain things you can reuse, I would think its easier just to build a new amp.
Have a look at my Aleph-X web page as it provides a run down of the differences between the 40w version and the 100w version.
Eric
While there are certain things you can reuse, I would think its easier just to build a new amp.
Have a look at my Aleph-X web page as it provides a run down of the differences between the 40w version and the 100w version.
Eric
Not sure whether this is the most appropriate thread, but here goes anyway.
In light of recent hints that the new XA.5 amplifiers may use an output stage similar to the Aleph0, I thought it would be interesting to study the latter in order to understand how it might be adapted for XA.5 purposes. It certainly is interesting but also has me baffled – can anyone explain how it works? In particular (referring to the simplified schematic attached), what is the purpose of R1, R14, R15, R16 and R17? More to the point, how are the values derived?
As I don’t have an Aleph0 prototype to play with at the moment, I input the circuit to a simulator (yes, I know…) to see if it would provide any answers. In short it raised more questions although I appreciate the simulator may be misleading me. Most intriguing (and contrary to my expectations), Q4 apparently functions as a current source (below the bias level) whilst Q5 provides most of the signal. The result is single ended class A, but not in the manner I was expecting. I thought Q5 was supposed to act as the current source with Q4 providing the drive.
I’m sure the circuit works given its heritage but I would appreciate any explanation of how.
In light of recent hints that the new XA.5 amplifiers may use an output stage similar to the Aleph0, I thought it would be interesting to study the latter in order to understand how it might be adapted for XA.5 purposes. It certainly is interesting but also has me baffled – can anyone explain how it works? In particular (referring to the simplified schematic attached), what is the purpose of R1, R14, R15, R16 and R17? More to the point, how are the values derived?
As I don’t have an Aleph0 prototype to play with at the moment, I input the circuit to a simulator (yes, I know…) to see if it would provide any answers. In short it raised more questions although I appreciate the simulator may be misleading me. Most intriguing (and contrary to my expectations), Q4 apparently functions as a current source (below the bias level) whilst Q5 provides most of the signal. The result is single ended class A, but not in the manner I was expecting. I thought Q5 was supposed to act as the current source with Q4 providing the drive.
I’m sure the circuit works given its heritage but I would appreciate any explanation of how.
Pass DIY Addict
Joined 2000
Paid Member
Thanks for the compliment, John! I've been working on it off-and-on again for a while. Needless to say, my amps are not yet completed. The next big expense is a big stack of extruded aluminum and then I might be able to actually get one of them up and running.
I created the web site because I found the Wiki to be lacking the level of detail that a noob like me needs. A short while ago, I edited the Wiki and added a link to my web page.
Eric
I created the web site because I found the Wiki to be lacking the level of detail that a noob like me needs. A short while ago, I edited the Wiki and added a link to my web page.
Eric
Xman3141 said:
Two things:
--If you're looking at 34V rails, you're going to have something over 200W output. Quiescent dissipation will be on the order of 600W per channel, which leads me to...
--I agree with Eric--you might be better off starting from scratch.
Grey
Ian,
The thing is a bit counterintuitive in that the N-ch devices on top run harder than the P-ch devices on the bottom. The comparatively light bias of the P-ch devices is made up for by ('made up for by'...ugh, what an ungainly concatenation of words, mea culpa) a current source running in parallel with the P-ch outputs. Given that lower wattage signals will be handled primarily by the heavier bias from up top working against the current source, that part is presumed to be kinda sorta class A. Since the current source is defined (unlike the "Aleph current source"), once the output demand goes beyond its contribution, the P-ch devices kick in and you're in something that's arguably class B, but with the current source still contributing. Want more class A? Increase the current source contribution. Want less? Haul back on the leash. Wanna get really funky? Use an Aleph current source instead of a fixed one. Okay, that suggestion might take a little work (ask yourself what happens when the current source runs out of steam), but it's not that bizarre.
Nelson suggested earlier in the thread that I make a run at a Ver. 2.0 of the Aleph-X. I sat down and sketched out a couple of ideas that I thought might be interesting to pursue. I managed to get a few hours of time in over the next week or two and a couple of things were looking promising, but then Nelson said that the XA .5s were using the older Aleph 0 output stage...which kinda unraveled the stuff I'd been working on, as it was based on the "Aleph current source."
Bummer.
Okay, so I went back and looked at the Aleph 0...and looked...and looked...and looked...and (yes, I know this is heresy) I just couldn't get excited. It would be the work of an afternoon to throw together two of the output stages and hang an X front end between them, but no matter how I looked at it, I couldn't really see anything interesting that I could bring to the table. I mean, really, it's a no-brainer to toss the MOSFET front end and replace it with some sort of JFET-based "UGS" front end, but there's already a thread chock-full of UGSs--any of which will work just peachy--and the output stage is fairly straight forward once you sit down and stare at it for a few minutes. There are a couple dozen guys here who could pull the thing together over the course of a few days and post a schematic.
(Jeez, by the time you're reading this line there are already fifty people simulating my suggestion...)
I just...I dunno...unless I can come up with something interesting to add to the mix, I think I'll sit this one out. I'm certain that Nelson has thrown something kinky into the topology, but anything that I'd build would be pretty predictable given what I outlined above. The booger would be getting hold of the JFETs, but enough has been written about JFET parts availability that I'm not going to add to it here. Suffice it to say that the "UGS" I'm currently using is based on a complementary differential with a 2SK389 opposed to a 2SJ109, both of those cascoded with the J310/J271 for lower voltages, or MOSFETs for higher rails. That runs sideways into a normal-ish folded cascode using the J310/J271 cascoded with MOSFETs (remember we're swinging rail to rail now or fairly close to it, so the J310/J271 won't cut it, voltage-wise). No, I'm not using current mirrors of any sort. The one thing that might strike some of you as odd is my choice to fold into the J310/J2171. Why not just drive the MOSFET directly? Two primary reasons: The J310/J271 are easier to drive in terms of capacitance, and I am specifically holding the gain down. You can build something like I'm describing or you can build something like one of the circuits in the UGS thread. The trick is that you'll need the output drive to end up at ground potential, not at V- plus the Vgs of the output devices. Once you've got that requirement in mind, you can actually use any number of front ends; doesn't even have to be a "UGS."
Hint: I've mentioned in the past that I look at "X" circuits as either Same-Side-Feedback (SSF) or Opposite-Side-Feedback (OSF). If you use an SSF front end, you won't have all the DC offset weirdness that drove so many people crazy at the beginning of this thread. However, if you want to use OSF, that's fine. There are enough ideas outlined in this thread that you should be able to get along okay. Oh, and you might want to have a look at the F1 schematic, too, while you're at it.
I'll keep thinking about it. If I come up with something kinky, I'll post it, but for the time being it just doesn't light my fire. Besides, Nelson has already said he's going to post something, so if no one cobbles a circuit together in the next little bit, he'll prod you.
Grey
The thing is a bit counterintuitive in that the N-ch devices on top run harder than the P-ch devices on the bottom. The comparatively light bias of the P-ch devices is made up for by ('made up for by'...ugh, what an ungainly concatenation of words, mea culpa) a current source running in parallel with the P-ch outputs. Given that lower wattage signals will be handled primarily by the heavier bias from up top working against the current source, that part is presumed to be kinda sorta class A. Since the current source is defined (unlike the "Aleph current source"), once the output demand goes beyond its contribution, the P-ch devices kick in and you're in something that's arguably class B, but with the current source still contributing. Want more class A? Increase the current source contribution. Want less? Haul back on the leash. Wanna get really funky? Use an Aleph current source instead of a fixed one. Okay, that suggestion might take a little work (ask yourself what happens when the current source runs out of steam), but it's not that bizarre.
Nelson suggested earlier in the thread that I make a run at a Ver. 2.0 of the Aleph-X. I sat down and sketched out a couple of ideas that I thought might be interesting to pursue. I managed to get a few hours of time in over the next week or two and a couple of things were looking promising, but then Nelson said that the XA .5s were using the older Aleph 0 output stage...which kinda unraveled the stuff I'd been working on, as it was based on the "Aleph current source."
Bummer.
Okay, so I went back and looked at the Aleph 0...and looked...and looked...and looked...and (yes, I know this is heresy) I just couldn't get excited. It would be the work of an afternoon to throw together two of the output stages and hang an X front end between them, but no matter how I looked at it, I couldn't really see anything interesting that I could bring to the table. I mean, really, it's a no-brainer to toss the MOSFET front end and replace it with some sort of JFET-based "UGS" front end, but there's already a thread chock-full of UGSs--any of which will work just peachy--and the output stage is fairly straight forward once you sit down and stare at it for a few minutes. There are a couple dozen guys here who could pull the thing together over the course of a few days and post a schematic.
(Jeez, by the time you're reading this line there are already fifty people simulating my suggestion...)
I just...I dunno...unless I can come up with something interesting to add to the mix, I think I'll sit this one out. I'm certain that Nelson has thrown something kinky into the topology, but anything that I'd build would be pretty predictable given what I outlined above. The booger would be getting hold of the JFETs, but enough has been written about JFET parts availability that I'm not going to add to it here. Suffice it to say that the "UGS" I'm currently using is based on a complementary differential with a 2SK389 opposed to a 2SJ109, both of those cascoded with the J310/J271 for lower voltages, or MOSFETs for higher rails. That runs sideways into a normal-ish folded cascode using the J310/J271 cascoded with MOSFETs (remember we're swinging rail to rail now or fairly close to it, so the J310/J271 won't cut it, voltage-wise). No, I'm not using current mirrors of any sort. The one thing that might strike some of you as odd is my choice to fold into the J310/J2171. Why not just drive the MOSFET directly? Two primary reasons: The J310/J271 are easier to drive in terms of capacitance, and I am specifically holding the gain down. You can build something like I'm describing or you can build something like one of the circuits in the UGS thread. The trick is that you'll need the output drive to end up at ground potential, not at V- plus the Vgs of the output devices. Once you've got that requirement in mind, you can actually use any number of front ends; doesn't even have to be a "UGS."
Hint: I've mentioned in the past that I look at "X" circuits as either Same-Side-Feedback (SSF) or Opposite-Side-Feedback (OSF). If you use an SSF front end, you won't have all the DC offset weirdness that drove so many people crazy at the beginning of this thread. However, if you want to use OSF, that's fine. There are enough ideas outlined in this thread that you should be able to get along okay. Oh, and you might want to have a look at the F1 schematic, too, while you're at it.
I'll keep thinking about it. If I come up with something kinky, I'll post it, but for the time being it just doesn't light my fire. Besides, Nelson has already said he's going to post something, so if no one cobbles a circuit together in the next little bit, he'll prod you.
Grey
Hi Blues and Ian,
There is an A0 service manual on the Passlabs web site. Look at page 24 for a better idea of what is going on in the X.5 and XA.5 output stage. I think this is what NP is referring to. There is an even clearer example in the a0s manual if you have one of those from the "old" days.
Graeme
There is an A0 service manual on the Passlabs web site. Look at page 24 for a better idea of what is going on in the X.5 and XA.5 output stage. I think this is what NP is referring to. There is an even clearer example in the a0s manual if you have one of those from the "old" days.
Graeme
GRollins said:
Paragraph 1: The P channels are also forward biased at idle, so
what you see is a mix of SE A, and PP A at idle. For each half, you
see a slightly asymmetric load line, and the transition in
proportion from SE to PP to "PP AB" occurs at different points
in the two halves of an X circuit.
Paragraph 2: No, it's a variation on the same old stuff - look to
the original Aleph 0 in this case, not the later.
Paragraph 3: Actually the SSF and OSF have the same issues
with absolute offset. That's because there's no common mode
feedback in an X amp. Anything that's identical (in phase) on
both sides is not processed by the feedback loop. X amps
process asymmetrical error.
😎
Nelson Pass said:
Yeah, that's why I said primarily. The P-ch MOSFETs are taking up some of the load, depending on the relative bias between the current source and the rest of the down-under devices. It's user-defined, depending on how much A you want vs. AB. I have no earthly idea what to call that operating class--I guess AB describes it as well as anything, given that it's A, then B, but it also depends on where you're looking, as the current source is there all the time, so it's class A.
I'm pretty sure I remember toying with an SSF circuit a couple of years ago and it was stable DC-wise. It'll be the very devil to find my notes on that circuit (if any...I don't always write things down because my time is so limited that if it comes down to a choice of trying something or writing, I'll nearly always chose the circuit over the notebook). I didn't pursue it because there was something limited about it. I don't remember exactly what. The topology would be interesting, though, if I could remember what I did.
Grey