2sj109 model
.model 2SJ109 PJF(Beta=39.21m Rs=0 Rd=0 Betatce=-.5 Lambda=4.338m Vto=-.5762
+ Vtotc=-2.5m Cgd=67.64p M=.2562 Pb=.3905 Fc=.5 Cgs=61.12p
+ Isr=158.7p Nr=2 Is=15.87p N=1 Xti=3 Alpha=10u Vk=100
+ Kf=109.9E-18 Af=1)
.model 2SJ109 PJF(Beta=39.21m Rs=0 Rd=0 Betatce=-.5 Lambda=4.338m Vto=-.5762
+ Vtotc=-2.5m Cgd=67.64p M=.2562 Pb=.3905 Fc=.5 Cgs=61.12p
+ Isr=158.7p Nr=2 Is=15.87p N=1 Xti=3 Alpha=10u Vk=100
+ Kf=109.9E-18 Af=1)
My apologies for not having done the bill of materials, yet. I am finishing my prototype AX and I am handling the Mills resistor purchase. Chad may already have it on his computer, so he may get to it before I do.
Aleph-XJ
rtirion,
Two questions:
1. In the jfet-front-end, why not scrap C5 and C6 completely ? It clearly improves the linearity of the high frequency response.
2. Why 1K and not 221 for R13, R31 ?
Best Regards
rtirion,
Two questions:
1. In the jfet-front-end, why not scrap C5 and C6 completely ? It clearly improves the linearity of the high frequency response.
2. Why 1K and not 221 for R13, R31 ?
Best Regards
...and if the 100k feedback-reistor will be changed to 87k we get a flat frequency response with the 2sj109 up to 30,000hz (at least in the simulation). So the potential benefit of the irf9610 regarding better high-frequency response should not longer be an argument not to use the better sounding jfets...
c5 and c6 => stability
c5 and c6 are there for stability reasons.
Please try and run a sim with 10Khz square wave as input.
Once with and once without the c's.
Without you will see overshoot. This is indication of too little
phase margin.
In a real circuit you can experiment (use for instance 5 pF) and
look at the square wave output on a scope. Mr. Pass wrote that
he designs his XA pcb's is such a way that the c's are no longer needed.
Regards
c5 and c6 are there for stability reasons.
Please try and run a sim with 10Khz square wave as input.
Once with and once without the c's.
Without you will see overshoot. This is indication of too little
phase margin.
In a real circuit you can experiment (use for instance 5 pF) and
look at the square wave output on a scope. Mr. Pass wrote that
he designs his XA pcb's is such a way that the c's are no longer needed.
Regards
Today, while modding my Aleph X, I disconnected 100 ohm output to ground resistors. Didn't observe any instability or increased offset. It is still 0.5 mV steady. As to sound, report later.
Peter Daniel said:
I have understood that it takes only one role given for DC return (or bleed off) in case capacitance exists in the signal path. If my understanding is misunderstanding, pls kindly let me know.
JH
Tweaking the 220u caps
Peter,
I would be very interested to learn what the difference in sound would be when replacing the 220uF caps with some MKP. From the simulation I saw that values down to 47uF would not harm the music down to 15hz.
THX
Peter,
I would be very interested to learn what the difference in sound would be when replacing the 220uF caps with some MKP. From the simulation I saw that values down to 47uF would not harm the music down to 15hz.
THX
Re: Tweaking the 220u caps
What about some N type Black Gate instead? 😉 I am getting a quote directly from Japan. Anyone interested?
There is a spot for the PS decoupling caps on the Version 1.0 board that would be nice to fill.
Peter, what mods are you applying to your AX? Fets in the input?
I did try to lowering the output resistors down to 10 ohms and it didn't make any difference. I was never brave enough to remove them altogether.
What I am not using are input caps. It seems to be doing very well when driven balanced from the BOSOZ directly.
Blitz said:
What about some N type Black Gate instead? 😉 I am getting a quote directly from Japan. Anyone interested?
There is a spot for the PS decoupling caps on the Version 1.0 board that would be nice to fill.
Peter, what mods are you applying to your AX? Fets in the input?
I did try to lowering the output resistors down to 10 ohms and it didn't make any difference. I was never brave enough to remove them altogether.
What I am not using are input caps. It seems to be doing very well when driven balanced from the BOSOZ directly.
I'm not changing for Fets. I'm too conservative to do a change like this.😉
I removed 100ohm output to ground resistors. In my setup it didn't change anything in output offset behaviour and the amp is very stable. I also removed all the parts associated with active current gain. It didn't make day and night difference. What I noticed is improvement in low level detail and slightly more open sound, something like a bit of a curtain effect gone. But nothing really to write home about.😉
I removed 100ohm output to ground resistors. In my setup it didn't change anything in output offset behaviour and the amp is very stable. I also removed all the parts associated with active current gain. It didn't make day and night difference. What I noticed is improvement in low level detail and slightly more open sound, something like a bit of a curtain effect gone. But nothing really to write home about.😉
The resistance to ground at both sides of the output is tojh6you said:
help minimize absolute (not differential) drift. The feedback
from outputs to the Sources of the diff pair do the same thing.
It is, of course, possible that one or the other will do the
job alone.
pass/ - wears a belt and suspenders.
I noticed that in my setup the absolute DC offset starts with 6V and slowly, after maybe 2 hours, goes down to 0V. It stays stable then. Is it normal behaviour? Thgis didn't change, when I removed output to ground resistors. I'm also using 16 output devices per channel in my X.
Smarty pants, that was a picture of XA200 Rev 0.
The owners of this rev will be treated to a free update, BTW.
Believe what I say, not the pictures I post.
The owners of this rev will be treated to a free update, BTW.
Believe what I say, not the pictures I post.
Let's think about this:Peter Daniel said:
If the offset is already at 0 when you remove the resistors,
there will be no difference in the current draw, and thus the
voltage will stay the same, at least for a while.
If you remove the resistors, I predict a wider range of
variation from cold to warm, even if the ultimate result
is near 0.
This reminds me of a friend of mine who used to own Jadis amps. He said that sometimes he had to wait 8 hours for the amp to open up and present all its glory ( or magic). But usually he got tired first and went to sleep before..😉
It takes about 2 hours for my amps to get absolute offset to 0, but then, I keep them on all day long.😉
It takes about 2 hours for my amps to get absolute offset to 0, but then, I keep them on all day long.😉
Member
Joined 2002
Just thought that i would throw this in. Peter them transformers look awsome. the last one that i recieved from anthony came with the instructions and the metal plate that bolts the tranny down. Thanks : O ) they look awsome.