Replacement For Toshiba 2SK170/2SJ74

[Oneminde]

Well, one can design with SE or Differential, no complimentary, and use Tosh 2SK209 and be completely satisfied. Take the Yamaha M-5000. Very musical amp, uses SK209 all day long.

 

[EUVL]

2SK209 Diff pair ?
Done that before.
https://www.diyaudio.com/community/threads/udness-or-you-dont-need-semisouths.343164/


Patrick

 

[jean-paul]

The world is a lot less fun without 2SK170 / 2SJ74.
At least my world.
And probably also that of Nelson's.


Cheers,
Patrick

If your fun in life depends on obsolete JFETs then send me a PM. I will donate my last 2SK170 to you.

 

[Oneminde]

And... worked fine ?

What I am wondering about, and this is a serious question. Since many N-chn are symmetrical, meaning drain can work as source and vice versa. Assuming the K209 is, would replacing the P-chn in a complimentary diff pair work if we flip the K209 such that the upper and lower half are source-to-source ??

- edit -
That would mean we end up with a quasi-complimentary discrete JFet differential pair with complimentary cascode BJT and MOSFET voltage stage

 

[EUVL]

You can never make a PJFET from an NJFET.
Semiconductor physics.

Of course a 2SK209 diff pair works fine.
Many examples built.
And better than J113 diff pair in same application.

Oh, and I still have 2000+ 2SK170BL on hand, in addition to V's and GR's.
Plus matching numbers of 2SJ74's, plus many others.
So I shall have more than enough fun for my lifetime.
But most generous of you for your donation, maybe to others.
I am sure there are many takers.


Cheers,
Patrick

 

[audiosteve]

https://www.aikenamps.com/index.php/what-is-miller-capacitance
https://audioxpress.com/article/JFETs-The-New-Frontier-Part-1

Thanks Oneminde! That article on JFETs by Borbley is great!

 

[Oneminde]

EUVL.... there is a hole in your theory... get it... LMAO

 

[MEPER]

For normal DIY's amp use (not the amp with 1000 JFETs) it is possible to get 2SK/2SJ for the rest of an amp builders' life.
Some quads or octets should do it. Then why not just get them?
It is good practice to get some when you buy other stuff from DIY Store. I have added some many times to the order. I am relative sure I have enough for the future. This will not stop me to get some more if good offers show up Inflation proof investment?

 

[Oneminde]

For normal DIY's amp use (not the amp with 1000 JFETs) it is possible to get 2SK/2SJ for the rest of an amp builders' life. Some quads or octets should do it. Then why not just get them?

The plans I have demand more than octet, but I will start with that since that covers my need for a pre- and power amp. One can always get more

 

[JCN]

Hope your are right and they will last for the rest of my life:
One Box K170, one J74 - pre selected in 0.5mA steps

Regards, J-C

 

[exojam]

nope
music is still and always here

'nuff for me

and I have crate of toobz

I only have a small drawer of toobz, so any donations are appreciated.

 

[EUVL]

DIYA Store has plenty.


Patrick

 

[robert8]

DIYA Store has plenty.


Patrick

Can someone verify if these are real Toshiba 2SK170 or fake?

 

[EUVL]

Measure and compare with datasheet,
https://www.diyaudio.com/community/threads/fet-measurements-with-just-a-dmm.308861/

You should download LTSpice (free) to read the .asc files.


Patrick

 

[JCN]

Never seen K170/J74 with such bended outer legs. And the print (no straight „1“ and no curved „7“) looks like fake.
You can zoom my picture above and compare the print.

But safest is to measure Idss and Vgs. The fakes I had received 2 times before I found a reliable source were with much lower Transconductance, resulting in a much too high Vgs for a given Idss.
K170-BL are usually in the range of 0.5V.

Regards, Jean-Claude

 

[jean-paul]

The gift that keeps on giving.

 

[EUVL]

I can confirm that Toshiba 2SK170 with 2.5mm pitch legs exists.
They used to be mounted on tapes for automatic insertion.
We know, because we bought a 3k-reel from local Toshiba distributor.

That is not to say whether those posted above is genuine or not.
You need to test, at least Idss and Vgs, preferrably also Yfs.
https://pearl-hifi.com/06_Lit_Archive/14_Books_Tech_Papers/Borbely_Erno/JFETS_The_New_Frontier.pdf
Fig. 5


Patrick

 

[Extreme_Boky]

I published a while ago some matching data on LSK170/LSJ74 JFETs.
These are being hailed as worthy equivalent of the Toshiba 2SK170/2SJ74.
https://www.diyaudio.com/community/...lsk170-jfet-pairs-grade-b.321649/post-5706027

Yesterday, I was playing with a complementary pair of the LS JFETs.
To my great surprise, totally unexpected results as follows :

The pair is wired as a simple complementary source follower with no source resistance.
They were matched to better than 0.1mA Idss at +/-12V individually beforehand.
In the source follower circuit, if the Idss is perfectly matched, output DC should be 0V.

Here are the measurement results :

Supply DC_offset
+/-12V -0.9mV
+/-9V_ -6.4mV
+/-7V_ -10.6mV

Idss of both JFETs drop with Vds, but the NJFET a lot more than the PJFET.
This means that in many applications, a pair matched at 12V rails are unuseable at even 9V rails !!!


How does Toshiba compare ?
I picked a matched pair from my private collection, with Idss at 8.05mA and matched at +/-9V to 0.02mA.

Supply DC_offset
+/-12V 0.8mV
+/-9V_ 0.0mV
+/-7V_ -0.6mV

So when to using LSK/LSJ matched pairs, make sure that the matching conditions is the same as in the application.
Unless you are happy with 10mV DC offset.



Patrick

I think this post deserves a repost. Thanks for sharing

 

[EUVL]

The most surprising thing is that the LSK170 drops more than the LSJ74.
i.e. Drain current Id depends more on Vds, or less ideal "penthode" characteristics.

My own conclusion :
Only use Toshiba, especially as N-P pairs.


Patrick

 

[Mark Johnson]

Or use circuit designs which are insensitive to channel length modulation (spice parameter LAMBDA). If d(Ids)/d(Vds) is a problem, arrange to operate the JFET at constant Vds; solved. Here's one possibility among many.