For the average TO247 devices, one should not go much beyond 30W for reliability alone.
That is assuming you have sufficient heat sink, which should not exceed 60°C.
And you should use a modern insulation like Kerafol.
What is proposed here are alternatives which allow people without access to macthed quads and octets to still be able to build.
Nothing more, nothing less.
Cheers,
Patrick
That is assuming you have sufficient heat sink, which should not exceed 60°C.
And you should use a modern insulation like Kerafol.
What is proposed here are alternatives which allow people without access to macthed quads and octets to still be able to build.
Nothing more, nothing less.
Cheers,
Patrick
As Patrick mentioned, my earlier posts were playing around with the idea of a "bigger" J2 in SPICE. In the real world, the MOSFETs and JFETs would have had a very short life (the input pair as well, which were seeing well over their spec'ed voltage).
I did later draw up one for the real world, but I haven't built it. Note that it's now strayed rather far from the original -- much farther than Patrick's posts.
Which brings me to one other note: I believe Nelson uses the 2SJ109s to reduce distortion a bit. Maybe it doesn't show up in SPICE because the two JFETs have identical models?
I did later draw up one for the real world, but I haven't built it. Note that it's now strayed rather far from the original -- much farther than Patrick's posts.
Which brings me to one other note: I believe Nelson uses the 2SJ109s to reduce distortion a bit. Maybe it doesn't show up in SPICE because the two JFETs have identical models?
Attachments
> I believe Nelson uses the 2SJ109s to reduce distortion a bit.
There is no reason why 2SJ109 is superior to 2 well matched 2SJ74, except for thermal coupling.
2SJ109 is only guaranteed for Idss to 3%).
xen-audio
And this is a power amp with a gain of 10, not a MC phono pre.
So thermal coupling is not an issue.
I have used 4x 2SK372 for the Wurcer phono pre, and drift is still no issue.
https://www.diyaudio.com/forums/att...-phono-preamp-180729-wurcer-phono-modules-jpg
> still tricky biz,
I don't believe in magic. But what do I know ....
Cheers,
Patrick
There is no reason why 2SJ109 is superior to 2 well matched 2SJ74, except for thermal coupling.
2SJ109 is only guaranteed for Idss to 3%).
xen-audio
And this is a power amp with a gain of 10, not a MC phono pre.
So thermal coupling is not an issue.
I have used 4x 2SK372 for the Wurcer phono pre, and drift is still no issue.
https://www.diyaudio.com/forums/att...-phono-preamp-180729-wurcer-phono-modules-jpg
> still tricky biz,
I don't believe in magic. But what do I know ....
Cheers,
Patrick
There is no reason why 2SJ109 is superior to 2 well matched 2SJ74, except for thermal coupling.
Sure, but I thought the question was why parallel at all? (Would 2 well-matched 2SJ74 have lower distortion than a single?)
As for why 2SJ109 instead of 2 x 2JS74: probably because Nelson had a drawer full of them to get rid of....
Cheers,
Jeff.
> why parallel at all?
Probably because of gate leakage of some Semisouth's, especially the audio grade (>10µA).
Patrick
Probably because of gate leakage of some Semisouth's, especially the audio grade (>10µA).
Patrick
I wasn't talking about one's ability to decipher even exact circuit (even Peasant ZM succeeded in Babelfishing it)..... I was thinking of his decision of not revealing much details ..... there must be some reason, eh?
EUVL's schematics above look fine to me if you want to build a J2.
And if you want to DIY your own, be my guest.
Dual J109's gave me more front end bias and greater transconductance.
The R100 Jfets were selected for low leakage, but it's not a big issue.
There was a good build with IRFP240's on the mu follower, but we elected
to stay with the Jfets for all devices.

And if you want to DIY your own, be my guest.
Dual J109's gave me more front end bias and greater transconductance.
The R100 Jfets were selected for low leakage, but it's not a big issue.
There was a good build with IRFP240's on the mu follower, but we elected
to stay with the Jfets for all devices.

> Dual J109's gave me more front end bias and greater transconductance.
I think I got it now. 🙂
My simulation indicates that 2SK3497 gives somewhat lower distortion than IRFP240.
And almost identical to Semisouth (with comparable Yfs).
Hence my preference.
Thanks,
Patrick
I think I got it now. 🙂
My simulation indicates that 2SK3497 gives somewhat lower distortion than IRFP240.
And almost identical to Semisouth (with comparable Yfs).
Hence my preference.
Thanks,
Patrick
The 2SK3497's appear to be nearly as unobtainable as the SemiSouths. Do you have a stash of those you'd be willing to whittle down, Patrick?
And, as ever, thanks for supporting the community, Nelson!
Cheers,
Jeff.
And, as ever, thanks for supporting the community, Nelson!
Cheers,
Jeff.
> The 2SK3497's appear to be nearly as unobtainable as the SemiSouths.
> Do you have a stash of those you'd be willing to whittle down, Patrick?
We have plenty, and are willing to supply them if sufficient interest.
Here attached the schematics in pdf.
Patrick
.
> Do you have a stash of those you'd be willing to whittle down, Patrick?
We have plenty, and are willing to supply them if sufficient interest.
Here attached the schematics in pdf.
Patrick
.
Attachments
A question then - are we just looking at Yfs as the determining factor in the 'upper' fet?
As I understand it, the 2SK3497 is a hight Yfs part than the '240, but what are the other factors that have gone into that part choice?
As I understand it, the 2SK3497 is a hight Yfs part than the '240, but what are the other factors that have gone into that part choice?
I ask as the irfp150 has a Yfs of ...11S? Off the top of my head, as well as twice the capacitance of the 240. But my knowledge to work out what that practically means is nil 
I'd hazard, reduced distortion and reduced bandwidth?
edit: turns out I've a pile of the 3497, as well as the '150's, but I'm still curious as to the answer here

I'd hazard, reduced distortion and reduced bandwidth?
edit: turns out I've a pile of the 3497, as well as the '150's, but I'm still curious as to the answer here
150 is good enough for follower duty
Cheers ZM, although I was more trying to understand the why's of it and how I should think about these things for myself. I actually bought the '150's because you'd previously described them as 2 x 240's in one package and then used them in an Aleph, biased the h%#l out of them (rather than trying to match a pair of 240's) and then enjoyed the end result. 😀
For this purpose, as I understand it, the higher capacitance won't be an issue particularly, and the increased Yfs is a positive in terms of distortion?
Does that sound about right?
- Home
- Amplifiers
- Pass Labs
- FirstWatt J2