Late 90's Peavey Transtube Supreme repair (power transistor questions)...

[J M Fahey]

These are a couple of my fakes.

On the left you see the fake Toshiba that is being sold even by reputable suppliers.
Fact is, it *does* work ... sort of.
The chip is twice the size of a regular 2N3055, is covered by a drop of white epoxy (quite common on fakes), but they skimp on the intermediate "coin" that spreads the chip heat into a wider surface, because the iron case base is a relatively bad conductor.
These transistors are not bad electrically, beta is normal and they stand good voltage, some up to 105 or 110V, but dissipate less.
They are happy in home stereo equipment and the like, but on hot sweaty guitar amps, they live a short life.
On the center you have a real 2N3055 ST, the chip is the normal size for that power, it's soldered to a full size copper "coin", which of course has 2 holes punched on its edge to allow for the emitter and base glass insulators.
All "coins" I've seen also have a small, almost centered hole, which has no visible purpose (for me).
Maybe it's a guide hole for some operation.
The chip position usually seems to be somewhat at random.
On the right you see the cheapest (not in $$$ but in a despective way) fakes of them all: they have a *thick* case base, almost 1/8", instead of the normal 0.8 to 1 mm, but use *no* copper coin at all, the chip is soldered straight to the iron base.
These burn more than firecrackers on a 4th of July.
EDIT: this Forum downsizes the images somewhat, to see it full size:
http://img534.imageshack.us/img534/5...hibafakes1.gif
Select "view full size"

[nashvillebill]

I'm pretty sure I had some fake 2N3055's a few years ago. Bought 'em from a reputable supplier too. I tried them in an old Kustom--I had the brilliant idea of pulling the old perfectly good ones and putting in these new ones. What WAS I thinking?

Didn't make it through my first gig before one of them fried, taking out the driver transistor and emittor resistor, which went up in flames--I could see the light from the flames through the unused guitar input jack, and of course I smelled the magic smoke.... Replaced the fried parts, put in the original 2N3055's and the amp works fine.

[tube power]

Here are the ones from Hong Kong...



That's the 24, I'm about to operate on the 25, which has a thin case...

[tube power]

25...



Crazy, I've seen all sorts of mixed up features???

[tube power]

I opened one more

I opened one more, seems to be genuine...



Yep, works just like it should.

[J M Fahey]

Well, now you know it, and have added very useful information to our common knowledge pool, thanks.
To begin with, the closest to "ON" those transistors ever were, was when the seller put the package he was about to send, close to his electrical rice maker.
The "24" is similar to my "Toshiba", mounted on a sub-sized copper "coin" heat spreader.
You have the correct size and shape on my "ST".
The "25" is *much* worse, the die uses no copper spreader, is soldered straight to the base, no copper coin.
It's covered in the infamous white epoxy and maybe there are two small chips under that goop, because there are two emitter wires.
The third is by far the best, although it looks, let's say .... sort of empty.
You should send 6 of them to my Lab for all kinds of experiments. Scientific, of course !!
Curiosity: a silicon chip seems to be cheaper than a 16 mm copper disk

[tube power]

I guess it's going to get a chip either way, but not adding the coin saves a step, that's time and money.
And copper isn't cheap. The amount they use adds up.
Hundreds of thousands of pieces later, they save a significant amount of cash.

Let the buyer beware...

[tube power]

I'm ordering the right parts from Mouser.
I would like to know what replacement fuses to get.
The original was white ceramic, 4 amp 250 volt.
They have various timed fuses, slow, fast, very fast?

[Steve Conner]

The speed of a fuse is identified by a code letter after the current rating. Slow = T, regular = no letter, fast = F, very fast = FF. (The silver foil or bolt-based ones are "WTF".)

So look at your original fuse. It should also say next to the fuse holder on the rear panel: 5A T or suchlike.

AC line fuses on amplifiers are usually slow-blow (T). DC rail and speaker fuses are usually F.

BTW: I got my shipment of fake transistors, and tested the Vceo to start with. They pass. The NPN one is over 300V, the PNP is 275. Will do some more measurements soon.

[tube power]

The speed of a fuse is identified by a code letter after the current rating. Slow = T, regular = no letter, fast = F, very fast = FF. (The silver foil or bolt-based ones are "WTF".)

So look at your original fuse. It should also say next to the fuse holder on the rear panel: 5A T or suchlike.

AC line fuses on amplifiers are usually slow-blow (T). DC rail and speaker fuses are usually F.

BTW: I got my shipment of fake transistors, and tested the Vceo to start with. They pass. The NPN one is over 300V, the PNP is 275. Will do some more measurements soon.

Did they look like the ones I got?
Even if they pass all the bench tests, doesn't the fact that they are improperly constructed still render them "unusable"?

[Steve Conner]

Yes, they have the same outward appearance as the ones you got. I won't open them until I've done all the tests.

They aren't MJ15024/5s, that's for sure. But they may turn out to be usable audio power transistors. Then again they may not, I expect them to fail the current gain and second breakdown tests, and explode when installed in an amp, producing some cool Youtube footage.

[nashvillebill]

In a similar vein, I also have some 25 amp bridge rectifiers that I purchased a few years ago from the same supplier that had the questionable 2N3055's. The terminals are incorrect internally, the ones marked AC to AC is a "short". They work (if I identify and use the correct terminals) but I wouldn't try to put 25 amps on them...

[J M Fahey]

tested the Vceo to start with. They pass. The NPN one is over 300V, the PNP is 275.

That matches my (unfortunately extense) experience with fakes.
Thay usually not only match but exceed voltage specs.
Those widely available "Toshiba" 2N3055 I mentioned stand up to 120 VCEO ("Normal" is 80 to the rare 100 to meet and exceed rated 70 VCEO) so that does not surprise me.
They *almost* meet Beta specs, and fail miserably the dissipation part, either by use of small copper coin or straight lack of it. (S.O.B).
I had to increase my short protection system and downgrade my amps because, as an Importer told me, "in the old days Factories competed on quality, "buy mine which is better than the other"; today they cutthroat each other and shave cents, so they meet published specs, but not exceed them".
In my classic breadwinner, the 100W into 4 ohms power amp, +/-42V, I went from 2N3055 (Motorola/RCA) 1969 to selected ones (EI, ST) 1998 to TIP142/7 2004 (ST, ON) to IRFP250 (IR, Fairchild) 2008 (so far, no fakes).
After 2008 I had also to lower rails to +/-39V to use up my huge stash of TIP142/7.
Problem is not voltage by itself, which they surpass, but second breakdown.
I do not think they designed a new, inferior manufacturing process (that would be ridiculous) but that they now tune it for higher speed and are slightly mor relaxed on the QC department. (Meeting but not exceeding specs).
During 3 years I had used Indian 2N3055H by USHA/UR , they were incredibly robust, even more than the ON ones, but now the supply dried up.
Oh well.

[tube power]

Got these from Mouser, should go in tonight...

[J M Fahey]

Coming from Mouser they *should* be good. "Should"
I find odd that on the right one the labelling *seems* to be way off center, the right "margin" being about 3 times the one on the left, although it might be due to the odd angle at which the picture was taken.
Can't you take another picture with the transistor centered and the camera perpendicular to it?
I also can't understand why the right one looks somewhat larger, the image is blurrier (shaken) than the other and the lighting hue seems to be warmer.
Were you holding it in your hand?