Certain variants of the complementary differential input stage topologies require tightly tolerated jFETs because the quiescent current through all four jFETs can be adjusted with just one variable resistor between the source terminals of both pairs.
It is in many cases not possible to obtain the same value for quiescent current on each of the jFETs without selected ones.
This is also the case with the power amplifier unit from Kaisaya HiFi Audio (Accuphase clone) - to find under
https://www.diyaudio.com/community/...cuphase-clone-and-music-fax-power-amp.336281/
go also to the circuit diagram in the first attachment.
Only due the fact, that an offset servo unit is present, the Accuphase power amp clone from Kaisaya HiFi Audio is in working condition - so I guess.
Basically the same input topology is to find in the Pass-Labs X250, but here are two variable resistors in use (in the rails to the drain terminals) - go to the circuit diagram (created from me according the original PCB) of second (and third) attachment.
Interesting here is the different connection of this variable resistor in the positive and negative half.
Now the questions in case of using the topology from circuit in first attachment:
1) According to which parameters should a selection for two pairs be made when I want to avoid a circuit modification ?
Possible is for both N-CH and P-CH pairs (i. e. all four pcs.) this:
a. according nearly identical Idss and
b. nearly identical Yfs and
c. both nearly identical Idss and Yfs at the same time
2) Modifying the circuit of input differential amplifier -
a. in such way, that the complementary character is eliminated - i. e. either the upper half or lower half of the front end operates only as a current source - e. g. by adding electrolytic capacitor of around 100-470uF from base Q10 resp. R9 (1k) to neg. rail
b. in such a way, that various value of Idss doesn't matter due the use of variable resistors at the right place (maybe in a similar kind on Passlabs X250)
in both cases it would be sufficient to carry out the selection only within one genus (i. e. N-Ch and P-ch separately, which would mean using double jFET's like the 2SK389/2SJ109 and which of course would require considerably less effort and time)
Thank you very much for appropriate hints therefore.
This threads don't provide the wanted information:
https://www.diyaudio.com/community/threads/2n-series-fets-like-2sk246-2sj103.80546/
https://www.diyaudio.com/community/threads/2sk246-bias-problem.254493/
https://www.diyaudio.com/community/...jc-2-preamplifier.120243/page-20#post-6142525
https://www.diyaudio.com/community/threads/jfet-matching-jig.302205/
https://www.diyaudio.com/community/threads/jfet-matching-sorting.228897/
https://www.diyaudio.com/community/threads/jfets-replacing-and-matching-questions.261618/
https://www.diyaudio.com/community/threads/question-about-the-effects-of-matching-jfets.404818/
https://www.diyaudio.com/community/threads/2sk389-2sj109-matching.384352/
P.S.: Under
https://pdf1.alldatasheet.com/datasheet-pdf/view/30636/TOSHIBA/2SK246.html
https://pdf1.alldatasheet.com/datasheet-pdf/view/212911/TOSHIBA/2SJ103.html
https://pdf1.alldatasheet.com/datasheet-pdf/view/30581/TOSHIBA/2SK170.html
https://www.alldatasheet.com/datasheet-pdf/pdf/30559/TOSHIBA/2SJ74.html
are the associated datasheets.
offers concerning matches pairs:
https://www.ebay.at/itm/402248021541 (matched N-CH and P-CH dual jFET)
https://www.ebay.de/itm/281748054275 (matched N-CH and P-CH single jFET)
https://www.ebay.de/itm/185864174433 (2SK880-GR/2SK880-BL as 2SK170 replacement)
It is in many cases not possible to obtain the same value for quiescent current on each of the jFETs without selected ones.
This is also the case with the power amplifier unit from Kaisaya HiFi Audio (Accuphase clone) - to find under
https://www.diyaudio.com/community/...cuphase-clone-and-music-fax-power-amp.336281/
go also to the circuit diagram in the first attachment.
Only due the fact, that an offset servo unit is present, the Accuphase power amp clone from Kaisaya HiFi Audio is in working condition - so I guess.
Basically the same input topology is to find in the Pass-Labs X250, but here are two variable resistors in use (in the rails to the drain terminals) - go to the circuit diagram (created from me according the original PCB) of second (and third) attachment.
Interesting here is the different connection of this variable resistor in the positive and negative half.
Now the questions in case of using the topology from circuit in first attachment:
1) According to which parameters should a selection for two pairs be made when I want to avoid a circuit modification ?
Possible is for both N-CH and P-CH pairs (i. e. all four pcs.) this:
a. according nearly identical Idss and
b. nearly identical Yfs and
c. both nearly identical Idss and Yfs at the same time
2) Modifying the circuit of input differential amplifier -
a. in such way, that the complementary character is eliminated - i. e. either the upper half or lower half of the front end operates only as a current source - e. g. by adding electrolytic capacitor of around 100-470uF from base Q10 resp. R9 (1k) to neg. rail
b. in such a way, that various value of Idss doesn't matter due the use of variable resistors at the right place (maybe in a similar kind on Passlabs X250)
in both cases it would be sufficient to carry out the selection only within one genus (i. e. N-Ch and P-ch separately, which would mean using double jFET's like the 2SK389/2SJ109 and which of course would require considerably less effort and time)
Thank you very much for appropriate hints therefore.
This threads don't provide the wanted information:
https://www.diyaudio.com/community/threads/2n-series-fets-like-2sk246-2sj103.80546/
https://www.diyaudio.com/community/threads/2sk246-bias-problem.254493/
https://www.diyaudio.com/community/...jc-2-preamplifier.120243/page-20#post-6142525
https://www.diyaudio.com/community/threads/jfet-matching-jig.302205/
https://www.diyaudio.com/community/threads/jfet-matching-sorting.228897/
https://www.diyaudio.com/community/threads/jfets-replacing-and-matching-questions.261618/
https://www.diyaudio.com/community/threads/question-about-the-effects-of-matching-jfets.404818/
https://www.diyaudio.com/community/threads/2sk389-2sj109-matching.384352/
P.S.: Under
https://pdf1.alldatasheet.com/datasheet-pdf/view/30636/TOSHIBA/2SK246.html
https://pdf1.alldatasheet.com/datasheet-pdf/view/212911/TOSHIBA/2SJ103.html
https://pdf1.alldatasheet.com/datasheet-pdf/view/30581/TOSHIBA/2SK170.html
https://www.alldatasheet.com/datasheet-pdf/pdf/30559/TOSHIBA/2SJ74.html
are the associated datasheets.
offers concerning matches pairs:
https://www.ebay.at/itm/402248021541 (matched N-CH and P-CH dual jFET)
https://www.ebay.de/itm/281748054275 (matched N-CH and P-CH single jFET)
https://www.ebay.de/itm/185864174433 (2SK880-GR/2SK880-BL as 2SK170 replacement)
Attachments
Last edited:
Not an answer to your question, but on topic: For the best possible complementarity in the 4Q differential I would use EUVL's method.
https://www.diyaudio.com/community/attachments/jfet-degeneration-pdf.564854/
https://www.diyaudio.com/community/attachments/jfet-degeneration-pdf.564854/