Counterfeit pwr x-istors

[Gtr_tech]

Just got the last parts order from MCM and in that batch had some 2n3772s for an amplifier repair. Thought it was odd that Toshiba would make a 2N type, but put them in anyway. After getting the unit put back together enough to test I powered the unit up to take some voltage readings at the driver section. All seemed well so I fully assembled it to let it cook for a while before returning it to the customer. Stuck a meter on the output and powered it up to check for DC before I put a load on it...seconds later I hear the pwr xfmr grunt and the fuse pops. Took the damn thing back apart....one driver shorted, one output x-istor shorted and the rest of 'em extremely leaky. The Toshiba name attached to an American deginated part # got me suspicious and I looked at one closer. The mounting base is not completely flat...it rocks on a straightedge. The "cap" that covers the die inside is not attached in a uniform manner, one end sticks up a little so it was not likely sealed. I need to get with MCM to see what they are willing to do about the situation, but if they won't take them back, I'll pop the cap off one and check out the die itself.

I was going to post a picture of the part but my camera says otherwise.

Be sure to check out this page and watch it out there kids.
http://sound.westhost.com/counterfeit.htm

[Enzo]

That is unfortunate. MCM has been one place I have had good results on power transistors over the years. All the Sankens I have purchased there so far have been OK.

Allied has them in stock for $3.57 from ON Semi.

[Mark Black]

I just wish the counterfeits were always so apparent (non-flat base, etc.).

I got skunked on a complete channel rebuild some years back on one channel of a Crest 8001, with something like 11 each of 2SA1302 & 2SC3281 TO-3P outputs. Bench test went OK but then our part-timer wanted to hear what an 8001 would do for some JBL double-18" cabs we had in the front room. So I say great - hook it up and crank it! Lasted about 20-30 seconds before the smoke got out.

So I'm thinking man - I suck! Tried to figure out what I had missed or where I had gone wrong. In the course of my research I came across the Westhost site and after reading his page on fakes I decided to crack open the transistors & have a look. Sure enough - they were all phonies. I'm pretty sure I got those from MCM too, although most stuff I get from them seems OK.

That was my intro to counterfeit transistors.

I've heard from a co-worker that even Crest themselves were victimized and used/provided fakes unknowingly (he got some from them).

I'd like to get the parts counterfeiters, spammers, 419-scammers, and maybe a few lawyers together on an island for a cocktail party (Molotov-cocktails, that is...).

[Enzo]

I have also so far had good results buying Sankens from B&D Enterprises

www.bdent.com

[jrfrond]

I had this problem about two years ago with the 2SA1302 & 2SC3281 which we specifically order for Crest amps. The metal collector plate looked like someone had taken sandpaper to it. It was not the usual smooth finish. Every one we used shorted or leaked. I purchased these from Electro Sound in NJ, a company that imports and supplies OTHER distributors, including MCM. Now, it is probably NOT the supplier's fault, but nowadays, when purchasing from the Far East, you can get counterfeit ANYTHING. Our company owner's son live in China for a year and gave me the lowdown on how the Chinese government turns a blind eye to the rampant counterfeiting of almost anything under the sun, because it is just so lucrative. We are definitely living in a Caveat Emptor society currently.

The best thing you can do is examine your new stock and look for any anomalies. Toshiba making a 2N3772??? Not bloody likely!!!

[loudthud]

The problem seems to be that buyers of parts at the distributor will buy anything that has the right numbers on it. I just spent an hour on the phone with one of my contacts in the semi industry. The real shame is that on most plastic power transistors, the die costs $.05 and the package is $.50 to $1.00. No wonder most counterfeits have cheap looking packages. The package on a TO-3 is about $2.50 to $5.00. Most parts seem to be made in China in the same factory that makes the real parts, they are just made at night after the boss goes home.

[imaradiostar]

Hah...I have a friend with contacts overseas that says that they'll make a "fluke" during the day and a "fleke" after hours. Sounds like the same story.

jamie

[Gtr_tech]

Most parts seem to be made in China in the same factory that makes the real parts, they are just made at night after the boss goes home.

Rotten bastards then. I'm not in the business of semiconductor quality control nor do I have the setup (or time) to test each device before its time to use it just to make sure it lives up to its part#. I've heard of the counterfeit thing years ago when I worked for an electronics parts distributor but thought the better dealers could guarantee thier stock to be authentic.
Bean counters outnumbering the technical guys now I guess. Country.....hell....handbasket.......

[jrfrond]

The problem seems to be that buyers of parts at the distributor will buy anything that has the right numbers on it. I just spent an hour on the phone with one of my contacts in the semi industry. The real shame is that on most plastic power transistors, the die costs $.05 and the package is $.50 to $1.00. No wonder most counterfeits have cheap looking packages. The package on a TO-3 is about $2.50 to $5.00. Most parts seem to be made in China in the same factory that makes the real parts, they are just made at night after the boss goes home.

You are almost on the money. The scenario is worse than that. China is so good a counterfeiting that there is an unwritten grading system: A, B, C, and D. The A grade stuff is made on the same assembly line, "at night" as you stated. The B grade is the same materials, possibly assembled in some worker's home in their spare time. C grade is usually same materials, farmed-out to people who have no clue with assembly instructions, and D is a total bogus knockoff of the worst quality.

Yes, it is sad. We get f**ked by China and the Middle East every day, and our government just lets it happen. Semiconductor fraud won't be an issue until it crashes a space shuttle, or screws with Dubya's GI Joe Walkie-Talkies.