The 2sk246 2sj103's have worked very well in several amps Ive put together, an original (several) f5 circuits and an f5 turbo v2. They have had some uncalled for output mosfets, but overall, got them sounding good.
They biased fairly easily, very entertaining sound without much noticeable noise or annoyances.
I would not double them up personally,,,, although they might measure better, I think the transient interactions of doubling have more of an effect on the sound quality, versus having a minorly out of spec input stage.
Also, I have had the sk246 on an aca and that sounded great too. I think they are very good solid sounding jfets, I have enough for 5 more amplifiers and I intend on using the rest of them because they are worth it even though they arent the 170/74's.
Compared to some random supposedly reputable 163/44 nec's oh my gosh the nec ones have been kindof frustrating to get sounding right if that makes sense.
They biased fairly easily, very entertaining sound without much noticeable noise or annoyances.
I would not double them up personally,,,, although they might measure better, I think the transient interactions of doubling have more of an effect on the sound quality, versus having a minorly out of spec input stage.
Also, I have had the sk246 on an aca and that sounded great too. I think they are very good solid sounding jfets, I have enough for 5 more amplifiers and I intend on using the rest of them because they are worth it even though they arent the 170/74's.
Compared to some random supposedly reputable 163/44 nec's oh my gosh the nec ones have been kindof frustrating to get sounding right if that makes sense.
No, it does not make sense to me, but then subjective taste is impossible to compare.
What sounds right to me may not sounds right to you.
I subjectively prefer the sound of low distortion circuits.
And I have great results with 163/44, at least on par with 170/74, especially in circuits with no global NFB.
I have a small collection of curve traced ones, so I know they are genuine.
These I keep for my own use, my life-time stock.
I do agree that people underrate them.
They can deliver great results if you design circuits round them, rather than just use them as K170/J74 substitutes.
For example, we have built a John-Curl Blowtorch-like circuit with the SMD version, namely K208/J106.
The NEXEN both measures and sounds great, to me.
And it turns out to be much lower noise than many would expect.
https://www.diyaudio.com/community/...orch-preamplifier-part-ii.146693/post-5270193
Just my 2 cents,
Patrick
Some examples from well-known designers using K246/J103 :
http://jm.plantefeve.pagesperso-orange.fr/sche.html
http://www7b.biglobe.ne.jp/~konton/No-168 seisaku.html
https://www.diyaudio.com/community/threads/semisouth-goes-dodo-what-now.222098/post-7233757
http://la4.cocolog-nifty.com/blog/2010/10/fet3-f844.html
https://prosugo.web.fc2.com/yasuiamp.htm
https://plaza.rakuten.co.jp/shomarin/diary/200708110000/
https://forums.melaudia.net/attachment.php?aid=28755
Patrick
http://jm.plantefeve.pagesperso-orange.fr/sche.html
http://www7b.biglobe.ne.jp/~konton/No-168 seisaku.html
https://www.diyaudio.com/community/threads/semisouth-goes-dodo-what-now.222098/post-7233757
http://la4.cocolog-nifty.com/blog/2010/10/fet3-f844.html
https://prosugo.web.fc2.com/yasuiamp.htm
https://plaza.rakuten.co.jp/shomarin/diary/200708110000/
https://forums.melaudia.net/attachment.php?aid=28755
Patrick
Yeah, that design with leds makes no sense whatsoever...
When I have problems with imaging, where left and right phase shifts change transiently, thats when I Know its time to change parts. 246/103 pairs have reliably performed well on many occasions, which is why I would recommend them.
I can prefer more harmonic distortion if that tradeoff can offset the studio recording transients.
You will find many discrepancies with left right balance and imaging schemes with different recordings.
That's why its nice to have parts with wide "tolerances" and being less temper-mental.
Still, output capacitors and non audio specific jfets don't really bother me, when building a system around an amplifier with purposefully selected features it can add air and can cut out the harsh "bit per bit" amplified signals.
Makes for a house full of sound rather than sterile replication at one point in the room.
Ambience is universally preferable.
When I have problems with imaging, where left and right phase shifts change transiently, thats when I Know its time to change parts. 246/103 pairs have reliably performed well on many occasions, which is why I would recommend them.
I can prefer more harmonic distortion if that tradeoff can offset the studio recording transients.
You will find many discrepancies with left right balance and imaging schemes with different recordings.
That's why its nice to have parts with wide "tolerances" and being less temper-mental.
Still, output capacitors and non audio specific jfets don't really bother me, when building a system around an amplifier with purposefully selected features it can add air and can cut out the harsh "bit per bit" amplified signals.
Makes for a house full of sound rather than sterile replication at one point in the room.
Ambience is universally preferable.
Last edited:
One more to add to the above list.
This one I find interesting.
http://www.zpag.net/Electroniques/Audio2/40w_audio_amplifier_01.html
Patrick
This one I find interesting.
http://www.zpag.net/Electroniques/Audio2/40w_audio_amplifier_01.html
Patrick
I know that this will be off topic but I wish someone can answer my question here.
I have some troubles using 2SK246-j103 is a Vzaichenco project,Only, Distortion,Negative , Feedback Amplifier.ODNF.
In this project ,like the others that posted here,it isn't any clarification about 246-103 IDSS classification Y,GR or BL.
Why 🤔
I have some troubles using 2SK246-j103 is a Vzaichenco project,Only, Distortion,Negative , Feedback Amplifier.ODNF.
In this project ,like the others that posted here,it isn't any clarification about 246-103 IDSS classification Y,GR or BL.
Why 🤔
It seems like the matched 170/74 pairs are more desirable with a lower Idss.
Im not familiar with all of the specifications for these jfets, but I do know they are much more expensive with the lower idss. I think this has to do with the goal being an ideal jfet has less idle current. I think it is an idle specification....
When the chips are made, they all will come out slightly different, and the manufacturer measures the figure and sorts them out. Yields are lower for lower idss devices, so, I think that is why they are more expensive. And it costs manufactueres time and money to measure each of them and also keep track of inventory of each particular variation of one product in the product line.
You can measure your own devices, and they will vary, use the highest and lowest measurements, and if you see no improvment or difference, you gave it your best try with what parts you have on hand.
Im not familiar with all of the specifications for these jfets, but I do know they are much more expensive with the lower idss. I think this has to do with the goal being an ideal jfet has less idle current. I think it is an idle specification....
When the chips are made, they all will come out slightly different, and the manufacturer measures the figure and sorts them out. Yields are lower for lower idss devices, so, I think that is why they are more expensive. And it costs manufactueres time and money to measure each of them and also keep track of inventory of each particular variation of one product in the product line.
You can measure your own devices, and they will vary, use the highest and lowest measurements, and if you see no improvment or difference, you gave it your best try with what parts you have on hand.
I'm building an F5 V3 with a matched set of Toshiba 2SK170/2SJ74 JFETS measured at 9.9mA. Will I need to make any modifications to the circuit for them to work optimally? Sorry if this is a dumb question, but I read that I should use JFETS in the 6mA range for the F5, and now I'm worried.
