2sk170 and 389 spice models

[Werner]

Does anyone have an official or homebrew
Spice model for the 2sk170, sj74, and sk389
FETs?

Have been searching the net for ages, to no
avail.



Thanks,



W

 

[ftorres]

2SK389

Hi,

I've only got 3 models for 2SK389. Just make a little spice circuit to make a curve tracer, and see which one is the closest to mf's datasheet.

Enjoy :

.model J2sk389 NJF(Beta=51.76m Rs=8.008 Rd=8.008 Betatce=-.5 Lambda=11.22m
+ Vto=-.5275 Vtotc=-2.5m Cgd=18.28p M=.3367 Pb=.3905 Fc=.5
+ Cgs=20.07p Isr=112.8p Nr=2 Is=11.28p N=1 Xti=3 Alpha=10u Vk=100
+ Kf=92.85E-18 Af=1)

*SRC=2SK389BL;QSK389BL;JFETs N;Gen. Purpose;25V 20mA
.MODEL QSK389BL NJF (VTO=-2.5 BETA=6M LAMBDA=1.2M RD=4.95
+ RS=4.45 IS=6.32F PB=1 FC=.5 CGS=71.2P CGD=18.9P)
* 25 Volt 20M Amp 35.3 ohm Dep-Mode N-Channel J-FET 07-28-1995
* 2SK389BL, TOSHIBA, 1986 D.A.T.A.BOOK, TRANSISTOR EDITION, P.227, LINE 75

*SRC=2SK389;QSK389;JFETs N;Gen. Purpose;50V 10mA
.MODEL QSK389 NJF (VTO=-2 BETA=20M LAMBDA=600U RD=7
+ RS=6.3 IS=1.58F PB=1 FC=.5 CGS=19.5P CGD=5.5P)
* 50 Volt 10M Amp 50 ohm Dep-Mode N-Channel J-FET 07-28-1995
* 2SK389, TOSHIBA, 1993 JAPANESE FET MANUAL, P.46

I would be too interested in other models, if available

 

[maik]

A working model for the 2sk170 (basically one half of 2sk389) is available together with the
commercial simulation software HSPICE from Avant!. Unfortunately I cannot post it here because it is
proprietary information. But if you have a chance to get your hands on a HSPICE installation you
will find it in the library directory.

The device is made by Toshiba. But they do not have a spice model for it on the web page.

Maik

 

[ftorres]

Oups, forgot to say that the model is actually one half of the 389, so you've got to use two of them to get the complete device.

I get the models from usenet, but if there is any copyright infringement, let me now, I'll delete the post...

And... If someone has the spice model for 2SJ109, can it be shared ?

 

[flyingfader]

Here is a spice model for: 2SK170

MODEL 2sk170 NJF
+ VTO=-5.211e-001 BETA=3.683e-002 LAMBDA=4.829e-003
+ IS=1.000e-009, RD=0.000e+000 RS=0.000e+000
+ CGS=5.647e-011 CGD=2.562e-011
+ PB=4.860e+000 FC=0.5

Maybe it will help u.

flyingfader

 

[Werner]

Thanks a lot!

W

 

[bcarso]

There are some models floating around out there that seem to have a very small lambda, and can be extremely misleading. The models also make a very abrupt transition from the so-called triode region to the pinchoff region, and comparison to characteristic curves on the Toshiba datasheet show marked discrepancies.

Another difficulty with parameter measurements is the signal-induced self-heating of the parts. One can deal with this using low-duty-factor pulsed measurements, or a strategy that I'm exploring where data sets for a given temperature are acquired when the device has come to ~equilibrium, and sets are recorded keeping the sample isodissipational. This still takes time, but not as long as changing voltages/currents and waiting each time for stabilization.

Brad

 

[Nelson Pass]

I have relied on the models posted on Bob Cordell's site.

 

[bcarso]

Thanks Nelson! I guess I've never visited Cordelland --- I'll seek it out.

I did just create another model borrowing from the post above, and it still predicts a way-low output conductance (about 38uS) compared to a linear fit to the Toshiba datasheet curves for Vgs = 0 (closer to 94uS). But it gets closer at least.

I have enough parts that I could even get some decent statistics. Of course they would apply to that particular run/date code. When Harman was shutting down a project (the rule there rather than the exception!) they were allowing people to take away some of the obsolete inventory, and I got a whole bunch of 2SK364BL, which is the same chip as the 170 but characterized for analog switching.

Brad

 

[bcarso]

PS: found the site, found the models, but only see now 2 parts for JFETs: both Linear Integrated Systems ones, the LS844 and the LSK389C. Had there been an SK170 before (I know the 389 sections are close but not afaik identical)?