advertisement


Interest check GB 2SK369 V

B

Bemused

What's this all about then?
As the title,
A few of us could do with some matched Idss Toshiba 2SK369 V.

We would have to buy a whole bundle of 500+ this should throw out some matched sets with highish Idss plus most probably some low Idss sets and some singles.

I understand 500 off could be around $500 plus taxes and carriage.
We have someone keen to do the matching :D

So no commitment just show interest if you have any.

Qty, matched, unmatched.

For myself

32 x matched at around 20mA+
10 x unmatched strugs.
 
OK, matched, sufficient for SENs for two Buffalo DACs, both SE, one dual-mono? Plus a couple of spares? I'll stump up for more if necessary to get to minimum quantity.
 
Definitely in for a matched set of 32! I believe the "V" grading by specification gives a minimum Idss of 14mA up to a maximum of 30. It should be possible to create groups with an "average" Idss of at least 20mA.:)
 
Hi Tony

I'd be up for a handful if they work out at a reasonable price. Got a few off fleabay but always good to have in the drawer.

Mark
 
I'd have some too - might be an interesting input device for getting a phono circuit back on the go.

Might not be quiet enough for the very lowest output MC devices (say below 200uV), but I think might be workable above that.
 
Yea me to just ordered a pair of Buffalo SE's so I guess 32 matched plus another 4 spare matched if available
 
PigletsDad said:
I'd have some too - might be an interesting input device for getting a phono circuit back on the go.

Might not be quiet enough for the very lowest output MC devices (say below 200uV), but I think might be workable above that.
Click to expand...

Let's say 8 pieces, unmatched.
 
So were about half way, this is a funny one as we don't actually know what values will pop out.
I would guess the matched sets of eight quads for dual mono sens will be middle ground and some nice strugs should popup with nice high Idss.
Any statisticians amongs us?
 
I have done some stats in the past; let me think about this one.

The worst possible case is that they are uniformly distributed between 14 and 30 - in reality you would expect a more tightly clustered distribution. The worst you could possibly have is for all 500 parts to differ by exactly equal amounts from 14 to 30mA. So the increment between them would be 16mA/500 = 32uA. So the worst match for a pair, for this incredibly unlikely distribution, is about 0.2%. Following this conspiracy, you couldn't get a group of four worse than about 1%. In reality, by picking groups around the randomness you should do much better.

If the overall probability distribution is peaked, then the differences will get squeezed, and the groups will be tighter.

Somebody will need to make a test rig with a transistor socket, a stable PSU and a decent DVM, to measure them all. You will also need to handle them with tweezers - at this level the heat from fingers will upset the measured values. Ideally the measurement rig should be at constant temperature, but I expect you will get away without.
 
PigletsDad said:
Somebody will need to make a test rig with a transistor socket, a stable PSU and a decent DVM, to measure them all. You will also need to handle them with tweezers - at this level the heat from fingers will upset the measured values. Ideally the measurement rig should be at constant temperature, but I expect you will get away without.
Click to expand...

Thanks for the fast calculations.
Since a few people also want unmatched Fets it is even easier to find tight groups of idss values.
I guess no need for tweezers, because they are probably on a reel, which also makes labeling them easier.
 
I have some half decent meters including a fluke or two and the fluke system calibrator, and spare fleas, salas shunts, placids, etc.
I would volunteer either some kit or do the matching myself. The system calibrator is errm, out of cal cert but never historically needed correcting when sent for cal. There again we dont need absolute measurements?
 
PD, thanks for input on stats.
I wonder if the manufacturer has some distribution curves.
I was hoping to get the fets measured whilst still on the reel / tape advantages of handling and numbering etc.
 
If the FETs you get are like mine the Idss values will be in the 16-26 mA range.
The Idss values are fairly well distributed and with 500 devices the matching yield will be very good (>90%).
In the SEN the devices are running at Idss so it would be best if the Idss is measured when the FET have reached equilibrium temperature (about 3 minutes).
Have fun:cool:
 
NicMac said:
If the FETs you get are like mine the Idss values will be in the 16-26 mA range.
The Idss values are fairly well distributed and with 500 devices the matching yield will be very good (>90%).
Click to expand...

Thanks for the info

NicMac said:
In the SEN the devices are running at Idss so it would be best if the Idss is measured when the FET have reached equilibrium temperature (about 3 minutes).
Have fun:cool:
Click to expand...

Yikes
 
You must log in or register to reply here.


advertisement


Back
Top