No, buy from Mouser, Farnell, or any other reputable seller.
And in any case you have no guarantee about matching.
I use tons of them, buy them by the stick (30 units each) and match them myself *if* needed, usually all same *original* Factory package/lot are incredibly close .
The risk is not worth "saving" a few bucks.
And in any case you have no guarantee about matching.
I use tons of them, buy them by the stick (30 units each) and match them myself *if* needed, usually all same *original* Factory package/lot are incredibly close .
The risk is not worth "saving" a few bucks.
What do you mean by "6+6"? You should certainly be able to get six matched pairs from a stick of 30 devices. Getting a matched sextet to match another sextet is probably asking a bit much.
Years ago I created matched pairs of BJTs with a friend of mine using a curve tracer. We started with a bag of about 100 devices. We ended up with 6-7 devices that couldn't be matched to any other device in the bag. Granted, those were BJTs, which tend to match more tightly than MOS, but it was also many years ago and the semiconductor processes are much tighter today than they were then. Anyway. I hope this gives you an idea of the yield to expect.
Buy your devices through Mouser/Farnell/Newark/Digikey rather than eBay. By buying from a distributor you know you're getting good devices.
Another relevant question to ask would be whether the devices need to be matched at all.
Tom
Years ago I created matched pairs of BJTs with a friend of mine using a curve tracer. We started with a bag of about 100 devices. We ended up with 6-7 devices that couldn't be matched to any other device in the bag. Granted, those were BJTs, which tend to match more tightly than MOS, but it was also many years ago and the semiconductor processes are much tighter today than they were then. Anyway. I hope this gives you an idea of the yield to expect.
Buy your devices through Mouser/Farnell/Newark/Digikey rather than eBay. By buying from a distributor you know you're getting good devices.
Another relevant question to ask would be whether the devices need to be matched at all.
Tom
I make Guitar/Bass/PA amplifiers so I match at a reasonable bias current , I chose 11 mA because I prefer my amps cold (Musicians abuse them anyway) but you can scale down current setting resistor to increse idle test current to any value you need, just use same matching current for all devices used in the same amplifier.
Then I stick small labels to each and when needed choose close Vgs values.
That said, same lot MosFets from any major manufacturer (think Fairchild/ST/etc.) are incredibly consistent.
An externally hosted image should be here but it was not working when we last tested it.
Texas Instruments erased history by omitting the author of this application note (hint, he is sometimes on DIYAUDIO):
http://www.ti.com/lit/an/snaa045a/snaa045a.pdf
Although it does not specifically deal with the devices in question, it gives some great practical advice relating THD and biasing.
For IRFP240/9240 my guess is that you should be matching in the low hundreds of mA.
Several years ago I regressed Vgs(th) and gm -- there's a pretty good correlation when you are using devices from the same batch.
Just my 2 pfennigs.
First thing I would buy from China would be a good cheap
knockoff of the 328Mega Transistor Tester, for under $20.
With one of them in hand, even if not the latest version,
could easily sort and match your own.
I got one in Dallas from Tanner for $14 that maybe isn't
latest firmware, doesn't have an encoder knob or wave
output features, won't tell me body diode forward drop
without having to disconnect the gate and make pretend
like its just a diode... It will still tell you the pinout, gate
threshold, and capacitance of just about anything with
3 pins, HFE, etc... Does SCR's and inductors too...
knockoff of the 328Mega Transistor Tester, for under $20.
With one of them in hand, even if not the latest version,
could easily sort and match your own.
I got one in Dallas from Tanner for $14 that maybe isn't
latest firmware, doesn't have an encoder knob or wave
output features, won't tell me body diode forward drop
without having to disconnect the gate and make pretend
like its just a diode... It will still tell you the pinout, gate
threshold, and capacitance of just about anything with
3 pins, HFE, etc... Does SCR's and inductors too...
Last edited:
Hello,
I think it's hard to be sure that parts are genuite when you buy in china.
For mesure, i use a DCA55 with mosfet with same temperature ( Peak Electronic Design Limited - Atlas DCA - Semiconductor Analyser - Model DCA55 ) great for a lot of mesures.
I also try to swap one mosfet directly in a amp, and mesure accros a 0.47 Ohms resistor for exemple, but others values ( for other mosfets ) are not stable in this case.
I think it's hard to be sure that parts are genuite when you buy in china.
For mesure, i use a DCA55 with mosfet with same temperature ( Peak Electronic Design Limited - Atlas DCA - Semiconductor Analyser - Model DCA55 ) great for a lot of mesures.
I also try to swap one mosfet directly in a amp, and mesure accros a 0.47 Ohms resistor for exemple, but others values ( for other mosfets ) are not stable in this case.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.