Replacement For Toshiba 2SK170/2SJ74

It shocked me quite a bit when i saw those Jfets in my Marantz CD player, which i bought brand new since the early 90's, because most people claims only curvy 7 printings is original Toshiba's.

So now i think the lettering printed on the Jfets with straight 7 instead of the curvy 7 might not be fakes, maybe Toshiba licenced out to some vendors to produce them when they switch their plant to VLSI ic making themselves, which was where the money is, in the 90's.

Out of curiosity, i switch out one channel of my M2X daughter card and replaced with my stash of straight 7 Nfet with similar IDSS. The M2X sings just as beautifully, i can't tell the difference when i switch between left and right channels.

Hope i can find some time to scope both channels to see any difference.
 
According to this post..

Digit 7 should be curvy.

...Marantz used fakes with straight 7.

:confused:

But according to this post... ;)

That typeface looks genuine - pls refer to my post here: My Transistors, original or copy?

Or to make it simpler:

It's been confirmed by multiple sources, e.g. the trusted eBay seller alweit, that Toshiba's very old production runs used a different, printed (not laser etched) font - one that, AFAIK, matches what you've got there...
 
no curved 7

Hi,

These fets were purchased in 1999 from mr. Borbely himself, no curved 7, but I can't assume that these are fake!
Gerard

Z
 
They look like they have 6-8 mA Idss, which is perfect. Slope looks ok too. How do I calculate gm? Its a curve, not a line.

gain = delta_Ids / delta_Vgs = (Ids_2 - Ids_1) / (Vgs_2 - Vgs_1)

For a spot measurement, make Vgs = -0.01 and then +0.01, measure Ids at both points and plug into the equation above.

Using a sourcemeter, I write a simple script that forces Vgs of -0.01, then 0, then +0.01, measuring Ids at each of the three points. From these measurements out falls Idss and gain.

What you choose for Vds will have an effect on the results. You can use Vds from the datasheet or Vds from the circuit that you plan to build. Your choice.
 
For the grade A LSK170 I calculated delta I over delta V as 1.85 mA/mV. Does this look ok?

For the grade B LSK170 I calculated two values 3.44 and 3.54 mA/mV respectively. Ok?

For the 2SK170 in question, I calculated 2.19, 2.23 and 2.45 mA/mV. Are these values acceptable?

One ampere per volt (A/V) is the derived SI unit of electrical conductance equal to one siemens.

Your 1.85 mA / mv is 1.85 siemens. The typical gain of the J74/K170 is 0.030-ish from my measurements. You can also check the datasheet.

Double-check your measurements. If you cannot find the error, post your test methodology/circuit. I think you are too high by a factor of something like 100.
 
Last edited:
In your graph with the lines circled in red, I look at the top green trace and I see that a change in gate voltage from 100 mV to 150 mV (negative) results in a change from about 12.5 mA to about 11 mA. That's about 50 mV of dV resulting in about 1.5 mA of dI, which works out to 0.030 A/V or 30 mS. Pick points closer together, the estimate gets closer.