Thanks Poynton,
that kinda confirms my fears.
W & W: you are right. Let's apply some power and load.
I have a genuine Toshiba SA1302 that I can compare BK to. Vce and Ic limits are easy but what about measuring Cob as a way of verifying die size?
Could a PSRR test rig measure Cob at bias?
that kinda confirms my fears.
W & W: you are right. Let's apply some power and load.
I have a genuine Toshiba SA1302 that I can compare BK to. Vce and Ic limits are easy but what about measuring Cob as a way of verifying die size?
Could a PSRR test rig measure Cob at bias?
Attachments
Your 2SA1302 is not real either. See Sakis' picture for what they really look like
They're from China. You can guarantee they will be nowhere near the originals. Smash one open and look at the die, they're probably just 2955/3055 types.
Go for something reputable, such as OnSemi's MJL3281/1302 or even Fairchilds 2SA1943/2SC5200 clones (also called FJL4215/4315)
It is just not worth getting ANY power transistors from eBay. Most of the sellers are China/Hong Kong and you can guarantee they're selling fakes. You can even tell from the pictures they have on the auction most of the time.
They're from China. You can guarantee they will be nowhere near the originals. Smash one open and look at the die, they're probably just 2955/3055 types.
Go for something reputable, such as OnSemi's MJL3281/1302 or even Fairchilds 2SA1943/2SC5200 clones (also called FJL4215/4315)
It is just not worth getting ANY power transistors from eBay. Most of the sellers are China/Hong Kong and you can guarantee they're selling fakes. You can even tell from the pictures they have on the auction most of the time.
JUST FOR FUN ....
one night i produced 3 complete amplifiers P3A .... all of them identical ....all was the same trafo boards caps bias .... when i say everything i mean everything ....except outs
one of them was original Toshiba 1302-3281
one of them was motorola MJL 1302-3281 ( had them matched given from ZECK as replacements parts)
and the other was 1943-5200 toshiba that i wasnt realy shure about even though i tryied them in 10pair per board delivering almost 550 w
well ....from all the toshibas 1302-3281 sounded better i spend the all night listening trying again and again listening to various types of music same speakers only changing amps .... in the end i think that they are really unbeatable
one night i produced 3 complete amplifiers P3A .... all of them identical ....all was the same trafo boards caps bias .... when i say everything i mean everything ....except outs
one of them was original Toshiba 1302-3281
one of them was motorola MJL 1302-3281 ( had them matched given from ZECK as replacements parts)
and the other was 1943-5200 toshiba that i wasnt realy shure about even though i tryied them in 10pair per board delivering almost 550 w
well ....from all the toshibas 1302-3281 sounded better i spend the all night listening trying again and again listening to various types of music same speakers only changing amps .... in the end i think that they are really unbeatable
Iain McNeill said:
They haven't made green PNPs since the early 90's. And even then , I wouldn't trust a green one unless it had the old logo on it.
Why do you need a special test rig to measure capacitance? Any DMM with a capacitance function will do - you don't need to measure at bias to tell whether it's fake or not. Cbe on real ones should be around 7 nF. Readings will differ a bit between meters because of different test signals, but a phony will measure half or less of the proper capacitance.
When you get a fake that is actually marked as a "Toshiba" you're guaranteed to be getting crap. They might be useable in a circuit that calls for a TIP2955/3055 pair (like your average 200 watt car amp). Devices from a second source may be legitimate attempts at cloning - but I have more confidence from ON or Fairchild than from some Chinese outfit. The only way to tell is to crank 'em up and see if they blow.
Thanks for the input guys.
Unfortunately, my DMM reads 1nF with nothing plugged in
When I try and measure Cbe on a selection of trannies I have here they all measure 1nF, including some vintage 2SB600's. Looking at the Toshiba spec they call out 400pF but with 10V across b-c. I'll need something a bit more sensitive.
Anyway, I ran them at 1A dc into 10ohms off a 36V supply to dissipate 26W in the package. Steady state on the heatsink after 30mins and I had a 37°C rise for a thermal R of 1.42°C/W. The Toshiba body had a 3° differential over the heatsink while the BK had 5° - suggests slightly smaller die?
For Ic=1A, Ib=5.9mA for Toshiba and 6.8mA for BK (hfe 180 vs 155) Not bad really - maybe they are a genuine attempt to copy.
Trying to think where I can get 20A from.......
Unfortunately, my DMM reads 1nF with nothing plugged in
When I try and measure Cbe on a selection of trannies I have here they all measure 1nF, including some vintage 2SB600's. Looking at the Toshiba spec they call out 400pF but with 10V across b-c. I'll need something a bit more sensitive.
Anyway, I ran them at 1A dc into 10ohms off a 36V supply to dissipate 26W in the package. Steady state on the heatsink after 30mins and I had a 37°C rise for a thermal R of 1.42°C/W. The Toshiba body had a 3° differential over the heatsink while the BK had 5° - suggests slightly smaller die?
For Ic=1A, Ib=5.9mA for Toshiba and 6.8mA for BK (hfe 180 vs 155) Not bad really - maybe they are a genuine attempt to copy.
Trying to think where I can get 20A from.......
Base to collector capacitances are not as telling as base to emitter. Cob's are all over the map and don't necessarily correlate to die size. Cbe's do.
And to really put it to the test you need to hit it with more than what a 2N3055 can handle. 1A at 36V is nothing. The test doesn't need to be long - 50V at 3A for one second at room temp will tell you quite a bit. If the die is too small or the attach sucks, it only takes a few tens or hundereds of milliseconds to run away. Use a big heat sink or keep the fins in an ice bath.
And to really put it to the test you need to hit it with more than what a 2N3055 can handle. 1A at 36V is nothing. The test doesn't need to be long - 50V at 3A for one second at room temp will tell you quite a bit. If the die is too small or the attach sucks, it only takes a few tens or hundereds of milliseconds to run away. Use a big heat sink or keep the fins in an ice bath.
personally signed by Toshiba-san himself. Very nice!
It is ironic that now I'm trying to destroy these devices I'm finding it very hard. Usually I throw this stuff together and it goes pop-pffft all by itself.
So I got a potentiometer with a driver tranny EF style to drive the TUT base. A selection of power loads and either 36V or 72V unless I want to wire up that 100lb toroid Nelson gave me last October. (yes, I really ought to invest in a bench PSU)
wg_ski: do you hit the TUT with a known width pulse on the base or is it more like a 1K base-collector, turn the power on and count?
Wavebourn: I'll try and collect some curves for the V-I I can provide but without 200V/20A drive capability I'm not sure it'll be too useful.
It is ironic that now I'm trying to destroy these devices I'm finding it very hard. Usually I throw this stuff together and it goes pop-pffft all by itself.
So I got a potentiometer with a driver tranny EF style to drive the TUT base. A selection of power loads and either 36V or 72V unless I want to wire up that 100lb toroid Nelson gave me last October. (yes, I really ought to invest in a bench PSU)
wg_ski: do you hit the TUT with a known width pulse on the base or is it more like a 1K base-collector, turn the power on and count?
Wavebourn: I'll try and collect some curves for the V-I I can provide but without 200V/20A drive capability I'm not sure it'll be too useful.
Iain McNeill said:
12V battery and two series diodes on the base, 60V supply on the collector, and calculate the emitter resistor for the desired collector current. Collector supply can just be a toroid, rectifier, and cap. Use a couple big 20W sand box units in series/parallel. for the emitter resistor - it has to sustain overload in event of a DUT failure. Use the base supply for control - turn on for a second or so. If it's well heat sinked it won't heat up above ambient and should take full rated 25C power. DUT will either blow up immediately or not. When it does, collector current rises but is still limited somewhat by emitter resistance so there should be no fire. The two diodes protect the base supply (battery) from reverse bias.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.