Thank you Kasey197 for the link.
I read it but didn't understand the point :-(
Well, as you know design choices are all about compromises, and NOT using
de-generative resistors must have a clear benefit in order to be left out.
As I mentioned earlier, the only places that leaving out de-generation would be ok is the
Input-Pair Stage (JFET couple) and the Output Pair Stage (the big FETS), but not the middle VAS stage with the Toshiba MosFets (I think there are laterals).
The only advantage in leaving out the degeneration resistors in the MosFet VAS stage would be the slightly
higher output voltage swing.
But that little headroom will come at a high cost and I do not believe it has been implemented.
HOWEVER, since Pass Labs is using Lateral MosFets here (I dont have much experience
with them), I guess - as I said earlier, it might be possible.... I dont know.
It would be very educational to hear Mr. Pass, or Mr. Colburn to respond on this.
Cheers
I read it but didn't understand the point :-(
Well, as you know design choices are all about compromises, and NOT using
de-generative resistors must have a clear benefit in order to be left out.
As I mentioned earlier, the only places that leaving out de-generation would be ok is the
Input-Pair Stage (JFET couple) and the Output Pair Stage (the big FETS), but not the middle VAS stage with the Toshiba MosFets (I think there are laterals).
The only advantage in leaving out the degeneration resistors in the MosFet VAS stage would be the slightly
higher output voltage swing.
But that little headroom will come at a high cost and I do not believe it has been implemented.
HOWEVER, since Pass Labs is using Lateral MosFets here (I dont have much experience
with them), I guess - as I said earlier, it might be possible.... I dont know.
It would be very educational to hear Mr. Pass, or Mr. Colburn to respond on this.
Cheers
----------------------------------------------------
Thank you Mr Pass for the response.
I agree with you on the output stage envelope.
Since the Lateral MosFets are used in the second, VAS stage, then they probably have a more civilized behavior than the vertical ones in that very arrangement.
Ok, I thank Mr. Pass, Zen Mod and Kasey197 for your help.
Then the conceptual schematics becomes something like this:
Last edited:
...................................
Hmmm...
Good question, honestly I do not know. They might be verticals.
The high trans-conductance of the MosFet pair in that arrangement (in case they are verticals) would make them difficult to use without degeneration.
Not to mention the gain difference between device samples, so one
might get a device with high gain and another one with low gain.
Degeneration makes this variation much less sensitive by linearizing the transconductance of each polarity.
Mr. Pass claims this (no degeneration) sounds better, however I would not pursue this arrangement.
There are other ways of disposing of the degeneration resistors and side-stepping the disadvantages of leaving them out.
-----------------------------------------
I took a slightly closer look at the datasheet of these Q's.
They certainly appear to be vertical Fets (thanx to member: Radule).
They appear to be - electrically at least - on par with IRF610 and IRF9610,
with the IR parts being slightly superior.
However, P-channel, IR parts did have some transfer curve idiosyncrasies in the past, I do not know
how the new Vishay parts measure today.
I have used IR in the past, and they are truly excellent parts. Very consistent in gain and cap
variations and extremely robust.
PL was extensively using IRF pinkies in the past , and as it seems - lately completely switched to 2SJ313/2SK2013 in all positions involved in audio path ( it would be certainly waste to use them in switching positions)
that seems somewhat contradicting to your claim of IR parts being slightly superior, at least for audio
that seems somewhat contradicting to your claim of IR parts being slightly superior, at least for audio
I am sure you are right. Otherwise there would be no other reason to do it.
I did however comment on the "electrical characteristics" (not the sound) of the IR parts compared to the Toshiba parts.
Look here:
Last edited:
I'm playing music,right now, on a amp with vertcal mosfet for gain stage,without degeneration, why this should be difficult?
It is not.
You do have to pay attention to adequate heat sinking, limiting the drift to
local phenomenon, and carefully adjusting against temperature rise.
The last part is something that DIYers can excel at.
---------------------------------------------------------------------
Yes, you are right.
I guess it is always more fun to drive a car with no brakes as long as it can be argued that the
engine is not that powerful ;-)
The current through the MosFets are undefined at the transient start, and there are no accommodations
to control it during operation - maybe there are - but only the designer and people with schematics know.
As you mentioned ealier, we need larger heat sinking and CAREFUL adjustments to arrive at the proper bias.
Can we guess why?
It is simple:
Because the circuit will have a volatile behavior, and always walks on the edge of the cliff, ready to jump off of it at any time.
Depending on the devices used, a transient event can launch a very high current through the MosFet
stage, and it can get so high as to damage or even destroy the devices.
As you know, Power MosFets still work (poorly) even if there are slight damages on the substrate.
As mentioned above, you are absolutely right about the fellow diy:ers.
We have all the time in the world to make adjustments and our amps do not go anywhere.
I guess: I would do anything for sound, but I won't do that ;-)
Mr Pass, I am curious:
Is Pass Labs planning to apply this arrangement (no degeneration) to other productcs as well, or is it unique to the XA25?
Thank you all for your comments.
Last edited:
Are you using a two-stage, complementary Fet based VAS as in the
schematics posted earlier?
I think with everyones help we will get very close to the "conceptual" schematics.
However, there will always be small details that will remain unknown.
And this is not nitpicking.
Every design decision is about compromise.
The designer gives up something to gain something else.
There are other solutions to getting rid of the effects of degeneration in a linear
amplifier circuit than just eliminating it.
The no-degeneration decision in the 2nd MosFet stage is very surprising by such seasoned engineers of Pass Labs
since the current through the channel may be undefined under transient conditions, and can be arbitrarily
so high as to damage the devices, or at least lead to unreliable operation.
However there may be limiter circuits that we don't know about.
But I respect Pass Labs designers and we are just discussing these decisions in order to understand and learn more about audio design.
However, there will always be small details that will remain unknown.
And this is not nitpicking.
Every design decision is about compromise.
The designer gives up something to gain something else.
There are other solutions to getting rid of the effects of degeneration in a linear
amplifier circuit than just eliminating it.
The no-degeneration decision in the 2nd MosFet stage is very surprising by such seasoned engineers of Pass Labs
since the current through the channel may be undefined under transient conditions, and can be arbitrarily
so high as to damage the devices, or at least lead to unreliable operation.
However there may be limiter circuits that we don't know about.
But I respect Pass Labs designers and we are just discussing these decisions in order to understand and learn more about audio design.
Last edited:
There is no undefined transient condition.
J fet drain loads is a pull down (that is definition) for 2nd stage Vgs and jfet self bias mechanism keeping things under control.
Yes, you are right, but there are two problems:
JFets can source and sink (in a Push/Pull config) more than Idss at transient conditions, since the
current through the feedback network can under some circumstances exceed bias Id (current feedback).
2nd stage MosFets have high enough gain (when un-degenerated) to pass an undefined amount of current through the channel under peak.
Last edited:
so , let us see your XA25 .
or you just mixed up things , thinking that XA25 is made by , say , Musical Fidelity ?
Zen Mod,
I sincerely do not appreciate your comments anymore, and the offensive way you express yourself.
I am asking you to please be nice.
We are merely trying to understand the XA25 design and learn more about audio.
We are not trying to step on any toes or build and sell copies on ebay.
The goal is try to understand the design decisions.
You seem to know a lot about Pass Labs products, but you should be patient with the rest of us who do not.
We would appreciate all your comments and help.
I sincerely do not appreciate your comments anymore, and the offensive way you express yourself.
I am asking you to please be nice.
We are merely trying to understand the XA25 design and learn more about audio.
We are not trying to step on any toes or build and sell copies on ebay.
The goal is try to understand the design decisions.
You seem to know a lot about Pass Labs products, but you should be patient with the rest of us who do not.
We would appreciate all your comments and help.
well ...........
I'm finding your approach rude .
More precisely - your attitude towards (technical and business choices of ) Pass Labs , especially NP , who is most responsible for amplifiers development in house.
You're seeking knowledge ;
Conduct the same, asking politely for reasoning behind not-easy choices , not constantly disputing the same .
If you know better , why you are here ?
And , whatever situation is , you ought to respect gray hair ..... that's at least how I'm raised .
"We are not trying to step on any toes ......."
So , you can't say a man : "now you made it all wrong , why you did it that way ? "
now , I'm Roger and Over from this thread , have some time to waste in more pleasant way .
I'm finding your approach rude .
More precisely - your attitude towards (technical and business choices of ) Pass Labs , especially NP , who is most responsible for amplifiers development in house.
You're seeking knowledge ;
Conduct the same, asking politely for reasoning behind not-easy choices , not constantly disputing the same .
If you know better , why you are here ?
And , whatever situation is , you ought to respect gray hair ..... that's at least how I'm raised .
"We are not trying to step on any toes ......."
So , you can't say a man : "now you made it all wrong , why you did it that way ? "
now , I'm Roger and Over from this thread , have some time to waste in more pleasant way .