vFET / SIT data sheets

Jump to Latest
#84
StevenOH said:
I will go back to Ditratherm story in post #63.
I saw that info in one document originated in Asia, which looked reasonably credible,
and, of course, requiring investigation.
Recently, I have clarified the origins of the V### V-FETs - it is from NEC - their development
parts numbers (for the registered 2SJ19_20_21 / 2SK69_70_71).
I will attach the comments from someone in Japan, who quoted one document from the summer
of 1974, (clearly from Japanese magazine).
Click to expand...
Took a few stabs in the dark, i.e. purchasing "random" Japanese magazines from 197x 😀, but...
 

Attachments

  • 143.jpg
    143.jpg
    411.5 KB · Views: 334
  • 144.jpg
    144.jpg
    147.1 KB · Views: 278
  • Like
Reactions: V FETish, Nashou66 and RainfallSky
#92
Can you measure it? (the Ciss)
I tried to simulate 2 stages single ended class A, a voltage gain stage with 2SK79, and a source follower
with 2SK82, loaded with constant current source with IRFP240, the result was much worse than the other
variant with LMOS ECX10N20 (2 in parallel). I also tried with 3, even 4 2SK82, but without any significant
success. VFETs (and triodes) will work much better when their load is connected to anode/drain, instead of
cathode/source. Their triode characteristics are not usable when they work as a follower, any good linear
transistor performs better, as its inner resistance is much lower and its linearity is not so important when
it works as a follower.
 
Last edited:
  • Like
Reactions: V FETish
#97
The very crude measurements that I took checked out. The skirt measures 55. Drain measures 30. The source measures 43.

1724723924479.png


As far as testing, I have someone willing to curve-trace them for me. I'm just trying to ensure that they are real before I send them his way, not to waste his or mine.

The only thing that I am unsure about at this point is the lettering. Usually, most I have seen only have Tokin on one side and the part number on the other, this one has it together. It looks like a newer part...
 
#98
bigger circle plate is Drain, smaller is Source, wire is gate

though - hoping they're real deal, good luck with heatsinking arrangement

using Keratherm 86/82 as pad for drain, same for source, I should go with squeezing it between two heatsinks

or if luck is on your/our side, maybe it's possible to use just one side heatsink - for Drain; then it can be clamped with - say - proper plate made of aluminum sheet (at least 3mm thick) isolated with piece of bare pcb material

that, not going above, say, 100W of heat per device

for more, I would go with non-isolated direct contact, heatsinking on both sides ..... which means heatsinks isolated from chassis


all that, trying to recall in memory few pics of same ones in some industrial assembly ......... I believe it was Soundhappy who posted those
 
#99
Oh boy... instead of getting my Furious amplifier builder badge.. more like a foolish amplifier builder badge.
Zen Mod said:
for more, I would go with non-isolated direct contact, heatsinking on both sides ..... which means heatsinks isolated from chassis
Click to expand...
So essentially I have two live heatsinks, one for the source and the drain?

It should be fun, If I don't screw it up somehow.... :clown:
 
#100
Zen Mod said:
or if luck is on your/our side, maybe it's possible to use just one side heatsink .......

that, not going above, say, 100W of heat per device
Click to expand...


that I meant - with Keratherm sheet as isolator; need to be confirmed with proper measurement, allowing no more than 10C of difference between drain plate and heatsink ditto to drain body

and .......... I wouldn't go without hefty follower as driver, say 30mA of Iq or more
 
You must log in or register to reply here.
Top Bottom