I slipped the new tansistors in and started the unit up... seemed ok. I hooked up a 1khz test tone and turned the volume up a bit. The wave form begain to distort and zap... a small puff of smoke and the right channel went slient.
well... I could not figure this one out. On investigation I found substantial damage to the channel... It was toasted
On a hunch, I cut the Trasistors open. I am no expert on counterfeits, but the white glue that held the die directly to to the back plate seemed funny to me. I cut open a like part number form another source to compare. The differece was dramatic. I took some photo's of the pair and sent them to a couple of highly regarded Audio engineers. They both came to the same conclusion (independently)... The dies were too small.... Both were counterfeit!!! Rod Elliott has done a lot or research on the counterfeit issue. His website covers the issue in depth (http://sound.westhost.com and http://sound.westhost.com/counterfeit.htm).
Well, even the one I thougt was real seemed to be suspect. The next morning I spent a couple of hours going through my inventory of transistors. I keep my trasistors separated by source as well as type. So I took one example from each source that I had any doubt about (I consider Mouser and Newark safe) and cut it open. Wow, was I suprised. Of six, I only found two real transistors.
The transistors picured above were all purchased from on-line electronics dealers. Sources I thought were safe. I sent the dealer images of the transistors he sold me that blew up my Marantz, just to warn him he may have counterfeits in his inventory. I never received a reply.... which I find strange. Either he knows and does not care, or he just does not what to believe it. I am not going to name sources here as I am no expert on this matter. But be warned... the proplem of counterfeit transistors is far more wide spread then I had ever imagined. I am getting counterfeits from sources that are fairly large distributors of electronics.
Needless to say I trashed about a quarter of my output transistor inventory.
1) Die is too small
2) The connecting leads are thin and flimsy
3) Lettering is often different (orientation, or marking order, etc...)
4) Sometimes (but not always) the lettering will wipe off easily with a solvent (know as the "acetone test")
