I have been measuring different JFET´s and found nice matching curves with a 2SK170 of 8.2 mA and a 2SJ74 of 9.0 mA and around 2.4 Ohm extra resistance for the 2SJ74, 5 Ohm was a bit too much. The higher Idss the JFETs are, the lower the difference between the two resistors can be.
But like NicMac said you have to measure them individually to get good results.
I think I posted it here, if you extract the beta parameter over a reasonable set of operating conditions the basic FET equations can give you good results even with equal degeneration resistors. I recently sorted 170/74 pairs for 0 offset with 25 Ohms in both sources, and still got 0.003% THD at line level. It's a no trim solution. By no trim I mean that zero offset and (near) minimum THD can be achieved without pots.
As for using trim pots I posted years ago the complimentary JFET input stage where the current ratios could be trimmed. You have two degrees of freedom you can trim offset and drift to 0 or offset and 2nd harmonic to 0 all with totally unmatched FET's.
Last edited:
I dream Linear Systems jfet's are available at Mouser , DigiKey , RS etc.
http://www.linearsystems.com/lsdata/datasheets/201175 - LSJ74 Rev A6 dated 2017 10 06.pdf
http://www.linearsystems.com/lsdata/datasheets/201110 -LSK170 Rev A17 dated 2017 08 28.pdf
and indeed his duals version. Metal can package are handy
http://www.linearsystems.com/lsdata/datasheets/201175 - LSJ74 Rev A6 dated 2017 10 06.pdf
http://www.linearsystems.com/lsdata/datasheets/201110 -LSK170 Rev A17 dated 2017 08 28.pdf
and indeed his duals version. Metal can package are handy
Attachments
different question gents...those of you who received fake 2sk170, any one figured what actual jfets were used as donor to be relabeled?
even they are lower transconductance, they may be used with some mods to the circuits in active crossovers or phono preamps
anyone? are they easily identifiable by some parameters?
even they are lower transconductance, they may be used with some mods to the circuits in active crossovers or phono preamps
anyone? are they easily identifiable by some parameters?
Very high noise compared to the real thing, and nowhere near as linear, leading to much higher distortion.
This seller is known by Nelson to have genuine Toshiba Jfets.
Toshiba 2SJ74 + 2SK170 LAB MATCHED QUAD to 0.03mA AND 4mV (7ma range) | eBay
This seller is known by Nelson to have genuine Toshiba Jfets.
Toshiba 2SJ74 + 2SK170 LAB MATCHED QUAD to 0.03mA AND 4mV (7ma range) | eBay
If interesting jfet's remplacement is hard to find
new vendors is available on line with semi's resonable price
Genuine Toshiba 2SK170BL/2SJ74BL high transconductance JFET complimentary pair | eBay
Have fun
new vendors is available on line with semi's resonable price
Genuine Toshiba 2SK170BL/2SJ74BL high transconductance JFET complimentary pair | eBay
Have fun
Attachments
OK
"As of Jan. 2018, we have yet to begin sorting the parts we have in stock, so give us a few weeks to get enough stock built up."
So they start selling before they have measured all there 4k stock.
You can guess how well the match could be from a smaller pool.
And as we have mentioned before, claiming sub 1% match has no meaning if you do not apply temperature control.
Just try gently breathing on a JFET running at Idss, and see how much the current changes.
Patrick
So they start selling before they have measured all there 4k stock.
You can guess how well the match could be from a smaller pool.
And as we have mentioned before, claiming sub 1% match has no meaning if you do not apply temperature control.
Just try gently breathing on a JFET running at Idss, and see how much the current changes.
Patrick
Attachments
Vendor jfet's Toshiba stock source was HI-FI audio manufacturer
After being bitten a few times on ebay I just stick to trustworthy sources.
Maybe not quite a problem with lower power devices but I have bought some rubbish power devices on ebay. They seem to blow up at the slightest fart near them.
Just try gently breathing on a JFET running at Idss, and see how much the current changes.
Patrick
Please Patrick send to jfet's vendor message with proper match DUT circuit and procedure with all important parameters
