Topless Transistor

[N1BCG]

Remember when you saw your first topless transistor? I must be a late bloomer because I just experienced it a few days ago in a Radio Systems TR-20 b'cast transmitter.

One of the RSMQ01 transistors lost its top which exposed more than we should see in a semiconductor. Unable to replace the now unobtainium component, I *very* carefully applied heat sink compound to the plastic cover and equally carefully reseated it over the exposed entrails. Certainly not the strictly controlled environment of the original manufacturing process!

Incredibly, it ran fine at 20% power, 50%, and finally 100% for several hours. I'm hoping the compound will harden somewhat, or at least stay adhesive enough to keep the protective cover in place and restore the privacy of the RSMQ01.

After more than 36 years, much of which under nasty conditions in damp basements and high reflected power, this TR-20 is getting a reconditioning with new caps and other assorted components. It sounds amazingly good with phenomenal audio specs, generating a low level modulated RF signal (200 meters) and using linear amplification up to 20,000 mW.

[WB2EMS]

I think that's the first topless transistor I've seen.   

Back in my youth, one of the advisers in our Radio Club/Explorer post worked for a large copier company where they were working on some newfangled things called switching power supplies. They were lighter and more efficient, but at the time had a nasty habit of ceasing to oscillate, dumping all of the current into one set of transistors which weren't up to a sustained handling of the current. This state of affairs would cause the transistor die to heat up rapidly to incandescence and depart the area. He had a handful of steel TO-3 cases with 'exit wounds' on them where the die had fled the scene! 

[WBear2GCR]

heatsink compound?

That's not got any adhesive qualities. Probably that was some sort of high temp adhesive, perhaps a phenolic compound (not sure). Standard epoxy gets soft with temperature, but there are high temp compounds. Assuming that top is merely ceramic and not BeO, a toxic ceramic (if the dust is inhaled) that is also a very fine thermal conductor then it's actually mostly a dust cover. Non essential entirely to the operation of the transistor.

So unless that ceramic top is held on mechanically the heatsink compound won't buy you much unless the unit
sits undisturbed.

Perhaps I'm misunderstanding?

                       _-_-

[Opcom]

If it is passivated then it should take a long time to decay however moisture and unclean airs will eventually corrode it. The big problem may be to find the proper adhesive. regular epoxy emits fumes and those may harm the device. 

nitrogren can be used to flood the die so to clean away all the nasty things in the air while you replace the cap properly.

There are specific adhesives for the job but I do not know their names only seen them used in a lab.  I think you can use die attach adhesive.
http://www.masterbond.com/industries/die-attach-epoxy-adhesives
http://www.henkel-adhesives.com/die-attach-paste-adhesives-26785.htm
possibly if you contact support at one of those they might counsel you and offer you a small sample of adhesive.

[KD6VXI]

There's a guy in South America who claims he's pulled the caps off blown transistors, repaired the burns in the connecting wires,  replaces the caps and continues using the transistors.   He's well known,  but was called to the mat for that claim.

One of the larger CB amp builders 'super secrets'  was putting 2sc2879s on a cookie sheet and baking them.   He claimed it would 'getter' or boil off any water vapor in the atmosphere inside the cap.   I never did get from him how he prevented more from going in......   If steam can get out,  it's not hermetically sealed.   He thought this let his amps run at 21 volts (which they did,  in comp conditions).   Without doin this,  he stated the caps would blow off the tops of the transistors.   

This failure mode wasn't the uncommon in 11 meter amp failures,  fwiw.   Maybe there's something in common?

I found super glue worked to put the caps back on.  In a couple very small points,  replace the cap and let it set for a few minutes.

Then I applied it around the caps,  completely.  This way,  less fumes inside the transistor.

Super glue is no Bueno in high heat.   If you let your transistors get that hot,  maybe you have other problems.....?

I would NOT  trust heat sink compound for any length of time.   It gets dry,  brittle and some can become conductive over time.

Congrats on getting unobtanium going again!

--Shane
KD6VXI   

[W3RSW]

We could ask a manufacturer.  
A.) are they hermetically sealed and with what?
B. Is the atmosphere inside pure nitrogen or whatever and wouldn't it be funny if we find they're in a vacuum?

Well, not really. Probably nothing more than a dust and handling cap. Those little whiskers are delicate.  Actually for the short term being capless allows better heat dissapation.
Might even run a small fan over those whiskers. . CB minds next trick.

Actually unless absolutely sealed, the cap/seal has to breathe a/c various altitude operating conditions.  leads out of case sealed? Naw, I see breather holes on each side of leads

[W3RSW]

We could ask a manufacturer.  
A.) are they hermetically sealed and with what?
B. Is the atmosphere inside pure nitrogen or whatever and wouldn't it be funny if we find they're in a vacuum?

Well, not really. Probably nothing more than a dust and handling cap. Those little whiskers are delicate.  Actually for the short term being capless allows better heat dissapation.
Might even run a small fan over those whiskers. . CB minds next trick.

Actually unless absolutely sealed, the cap/seal has to breathe a/c various altitude operating conditions.  leads out of case sealed? Naw, I see breather holes on each side of leads

[N1BCG]

I see breather holes on each side of leads

I saw them as well, which was fortunate, as I began sketching out plans for a large nitrogen-filled bag within which the resealing task would be performed.

Additionally, the original brown adhesive didn't seem to have sealed those gaps. The cap is some kind of plastic (no need for a Beryllium Oxide scratch 'n sniff test) and seemed to be primarily a physical protective cover.

Admittedly, the heat sink compound was far from an ideal solution, but given what was on hand and the conviction that something had to be done asap, I went with it. Days later, and many hours of high temperature operation, it seems to be staying put. The cap has little mass and it's unlikely to get disturbed where it's located. Fingers crossed on this one.

Anyone have any cross-reference data for RSMQ01 transistors? Replacement would be the best solution, of course.

[WBear2GCR]

Next time bubble gum.
Bazooka, or Double Mint?

                _-_-

[Opcom]

Ha! no need for a bag, just a gentle breeze of N2 from a rented lecture bottle. But apparently even that is not worthwhile if the thing has vents.

I have never heard of a cap popping off from simple temperature but they can pop off if the die blows a chunk out. There are always lots of things I haven't seen. Got to love life.

Looks like you may be home free then. The modern RF parts I worked with are all hermetically sealed RF MOSFETs. Old pills I don't know from aspirin but ass-u-me d they were also sealed.

Just as an experiment, cover it with a short focal length lens so you can admire it.. then shine some light in there and see if you can 'adjust the bias'.