toasted components

[mstechca]

It seems that now in my final designs, I get an occasional transistor to burn out. When I replace it with a newer one, and the same model, it works. Unfortunately, I don't have new ones now, and will buy some later on. In my circuit, I use 2N2222's, 2N4403's and one PN3563 (for the superregen detector).
It turns out that the PN3563 1/2 the time fails to work (even with the digital station circuit removed). I am starting to think that excessive heat could cause the problem.

I am using a 45W soldering iron (which is about 1/3 hotter than most pencil irons) because it melts solder quickly and easily.

Do you think I should go down to a 15 or 30W iron?
I'm starting to get concerned with that transistor.

 

[Nigel Goodwin]

Transistors are VERY hardy devices (at least silicon ones are!), if you're damaging them while soldering them in?, you most be applying the soldering iron for much, much, longer than required. It shouldn't take more than a second or two (if that?), but 10 or 20 seconds soldering shouldn't damage a silicon transistor.

Do they work at first?, then fail later?, if so it's probably not due to soldering, more likely it's exceeding it's specs in the circuit, and failing because of over heating.

 

[Dean Huster]

Yes, the big scare with soldering transistors and using heat sinks on the leads while soldering came from the germanium point-contact devices of the 1960s, which were pretty delicate. Devices these days, if you look at the spec sheet, can handle some pretty hot temps for a lot longer than needed to solder a joint (2 seconds is the recommended time to solder a PCB pad -- any faster and you need lower temp or a smaller iron; any slower, the opposite fixes prevail).

I agree with Nigel that the problem is probably caused by stressed circuit parameters.

Dean

 

[mstechca]

Transistors are VERY hardy devices (at least silicon ones are!), if you're damaging them while soldering them in?, you most be applying the soldering iron for much, much, longer than required.

What's the maximum time I can apply the soldering iron before the transistor suffers from heat?

Do they work at first?, then fail later?

In some cases, yes. Sometimes, some of them (same model) don't work at all, yet the other components remain the same.

Edit: when I tested the exact same circuit on the breadboard, the transistor works.

 

[zachtheterrible]

Cold solder joints :lol:

 

[audioguru]

Edit: when I tested the exact same circuit on the breadboard, the transistor works.

What does heat have to do with it then?
The circuit obviously needs the higher capacitance of the wiring of the breadboard on some of its connections in order for the circuit to work.

At RF frequencies, a pcb can't have anywhere near the high capacitance between wires or plug-in strips of a breadboard. Some jumper wires on a breadboard are inductors.
Didn't we say this before?

 

[mstechca]

I don't use jumper wires on the board.

I still think it is the transistor (Pn3563) , because the maximum ratings on it are low compared to the 2N3904.

and no, not cold solder joints. I made sure I have none of those. In fact, I bend all my leads before soldering them in, so that all components can hold themselves in place.

Then when I solder, the chances of having no connection between the copper and the lead will reduce drastically.