My Transistors, original or copy?

[EchoWars]

/OT

If you need more than 10, you better have an armed friend, or an escape route.

Back on topic: No, I certainly don't recommend doing business with Match-a-Knob. I got perhaps a dozen original, good transistors, and several dozen bad fakes (some with two dies...had pics around here at one time). I gave up on sourcing out-of-production transistors right about then.

[wojtek5001]

Are they real or fakes? I bought them in local shop couple months ago. I measured them with simple multimeter, and beta for one of them was about 40 - little too low. I've destroyed one, and die is 4x4.5mm (need daylight to make photo, so i'll try post it tomorrow).

[unclejed613]

a few pages back, there's a pic of a pair of C5200's which look real. i'm beginning to make a list of the date code styles different manufacturers use. i'm getting my info either from data sheets when available, or from known original devices when i can find them. Toshiba uses a dot and two alphanumerics for their date code on the right side under the part number. so since it doesn't have a toshiba date code, the devices in your pic are fakes based on the date code characteristics. toshiba also has two beta ranges in their data sheet, R and O. since the beta range designator is not on your devices, or even the ones in the pic a few pages back, they are most likely fakes (yours and the ones a few pages back.)

[wojtek5001]

Another picture of 2SC3280

[unclejed613]

that does look like a correctly sized die, but still with the beta code missing from the package, i would tend to think they're not original devices.

[wg_ski]

Usually the die doesn't make it intact when taking apart a real TO-247/TO-264. Bash the case apart and the die comes out in pieces if the passivation is worth anything. And I haven't seen real PNPs in green packages since since the early days of white silk-screened logos (with "Toshiba" in the old script font).

[unclejed613]

i just downloaded Toshiba's Semiconductor Reliability Handbook. on pg 20 of chapter 1, they show the two date coding schemes they have in use, and i will fill in a small tidbit of info from actually looking at real toshiba devices. the two date coding schemes are the standard 4 digit one YYWW (year and week) and an abbreviated one using YM (last digit of the year, and a letter from A to L for the month). now for the undocumented tidbit.... they put a dot before the abbreviated date code (one up high like a degree(temp) mark). the date code is on the right hand side below the part number. the devices shown above have an S at the beginning of the date code, which is a bogus date code.

like i said, i have begun compiling (direct from manufacturer's publications where possible) the various date and lot code styles that they use to track reliability data. some of these markings are also logos, factory codes and such imprinted in the mold mark dimples (with sanken devices, the presence of an SK in the dimple above the collector lead actually indicates that it is a fake). from known good and original samples of sanken devices, i have found that sanken hasn't used an SK logo in the dimple at least as far back as 2005.

[powerflux]

i recently bought a batch of 2n3055's from ST, price was pretty normal, but they keep blowing/shorting or acting strange.

last night i was so pissed i trashed a new one open, but i have no material to compare it with.

any opinions?

[unclejed613]

looks a little small, even for a 3055 which is a 15a device. the bond wires even seem thin for a 15A device. ST publishes the same specs as MOT, 60V, 15A, hfe=20-70, Ft 2Mhz. what was your applied voltage?

[powerflux]

somewhere between 43 and 50, but certainly not more.

the second channel is built up from old RCA 3055's and performs perfect, even under full load for hours