Toshiba 2SC3281/2SA1302 - Are these fakes?

[CrazzyAbtTubes]

For reference here are the 2SD845/2SB755 Toshiba outputs, these are not fakes, but the real thing, from a really good sounding LXI amplifier.

[wg_ski]

Originally posted by CrazzyAbtTubes
I have a bunch of Toshiba 2SC3280's and 2SA1301's that I Plan on paralleling in an amplifier, I just have no idea if they will all agree with each other when I do so...

Nextly.... are these Toshiba 2SC3281's and 2SA1302 Fakes? They look genuine, will they work with the others in parallel?

Don't parallel unless they're from the same batch - even if genuine.

That 1302 is suspect. Not only did they discontinue the green cases when they discontinued that script "Toshiba-san" logo, the mounting hole is too small.

[wg_ski]

Originally posted by CrazzyAbtTubes
For reference here are the 2SD845/2SB755 Toshiba outputs, these are not fakes, but the real thing, from a really good sounding LXI amplifier.

I really liked those old transistors. It's a shame you can't get them anymore. I still have a few pair of C2565/A1095 in the same case. Highest freaking fT's I've ever seen in an audio power transistor - which is very good when using them above the audio range.

[Lumba Ogir]

wg_ski,

That 1302 is suspect.

It is, 2SC3281 as well.

[unclejed613]

Originally posted by wg_ski



Don't parallel unless they're from the same batch - even if genuine.

That 1302 is suspect. Not only did they discontinue the green cases when they discontinued that script "Toshiba-san" logo, the mounting hole is too small.

the digits rather than the "JAPAN" under the date code is suspect too. for some reason, it's more common to find PNP fakes than NPN.

there's a document you can download from the Toshiba web site. the subject is their semiconductor reliability tracking program. in there is detailed drawings of exactly what and where every inked or laser etched marking on the case belongs and what the date codes mean. unfortunately, just as with the sanken devices, it takes careful comparison to figure out the mold marks (raised letters or digits inside the dimples in the plastic) and which ones belong where. i am beginning to see a pattern with some mold marks where at least the last digit for the year is usually in one of them, and should be the same as the year mark in the printed date code.

[CrazzyAbtTubes]

Thanks for the reply, I am more used to working with tube equipment (as the name describes ) but I have decided to rebuild one of my SS amps using the Toshiba 2SA1301 and 2SC3280.

I think that I can parallel the output using the ones that have matching numbers, so the "R" going in one channel ( these were in parallel formation in another amplifier along with the fakes) And the "O" will go in the other channel ( from a cheap but working Fisher amp)

If there is anything wrong with this idea let me know, or make a seggestion...

As for the suspected fakes, they can go rott on a shelf for a while.... Maybe later I will split one open since I don't have a full set.

[CrazzyAbtTubes]

Oooooooo, this is interesting, one chip is smaller than the other, and there is some white stuff on the small one! That ain't a 100 Watt BJT!!!!!

The other one appears to have some copper pad, and has a much larger piece of silicon on it, I killed it!

Well that solves that mistery, I think when I need to find new Toshiba outputs.... I will just go to Ebay and buy some cheap SS amp such as the Fisher CA860 that the really good "O" types were found in.

[EWorkshop1708]

I cracked open some ON-Semi 4281/4302 samples I had, after I blew two up in a test.

The chips inside both look similar in size to the Genuine Toshiba in your picture.

I notice the older Toshiba are rated 150W
The newer On-Semi are 230W

I wonder why they use the same size piece of silicon? Is the Toshiba underrated, or the OnSemi overrated, or does the die size matter that much with power dissipation? I couldn't help but notice how big that Toshiba transistor silicon die was.

[unclejed613]

the die size matters mostly for current density. the thermal tab (material, mass, and shape) and bonding method of the silicon die matters for heat dissipation ratings. also some dissipation ratings are calculated for a specific test heat sink, others are calculated for an "infinite" heat sink, to make it easier to calculate heat dissipation with a known finite one.

that was a nice picture. how did you accomplish separating the epoxy from the tab and die so cleanly? i made some pics where i work showing the die size of some MN2488 transistors, but i ended up with just the bottom layer of the die on the heat tab, and chunks of epoxy.

[Lumba Ogir]

unclejed613,
nice explanation, as usual.

that was a nice picture. how did you accomplish separating the epoxy from the tab and die so cleanly?

Skilfully done for demonstration purposes. Well, those fellows have obviously not much in common, it`s not just about power handling.