Grrr
There's nothing more annoying than finding out that you've managed to bridge /RAS and A3 on the bank of 4116s while putting in sockets. Naturally, the bloody solder bridge is on the component side where you can't get at it without removing all the sockets again.
Argh.
And sockets tend to turn to molten goo with the hot air gun removal method.
Argh.
And sockets tend to turn to molten goo with the hot air gun removal method.
Post edited by Winston on
Comments
When you refer to "hot air gun" (as others have) are you talking about a hair dryer (level) device or something a bit more potent? With my very limited understanding of electronics, I would have throught you would destroy a chip with hot hair long before melting solder (assuming a paint stripper level gun).
Don't forget these chips were soldered in the first place by wave soldering. Modern PCBs are assembled in an oven that gets up to 260 celcius, and bake there for a few minutes! (I do, however, remove the ULA - that's been consistently in a socket at least to Issue 4S which suggests that it's not at all heat tolerant or they'd have wave soldered it with the rest of the chips).
I've harvested everything from ADCs, 74HC* parts and static memory from PCBs using my trusty B&Q gun, and these chips end up in my projects (and they work fine). The only trouble is the typical IC socket has plastic with a very low melting point (well under the melting point of solder), so I'll have to remove them the slow way.
The trouble is the bridge is underneath a socket (which is flush with the PCB) on the component side rather than the solder side, so I can't even find out where it is let alone scrape at it without removing the sockets. My fault for trying to do a late night repair job, really.
a low ohm shorts locator comes in helpful too - I've a little unit from Polar that has an audible tone indication that helps to narrow down a s/c to a small area very quickly.
Heat the sockets for a short while and drop the board to the ground when the is has melted. Repeat a few times, but not too long.
No guarantee, but I got the sockets out of the PCB this way with my paint heater.
The result - another repaired Spectrum.
I also now have some 4116 chips which have failed in interesting ways, rather than simply not working at all, which will help in the development of my Spectrum testing peripheral board.
(The only other problem was that while I was testing the repair, I came back to the work bench to find my Z80 project had crashed - it was showing the RST 8 crash dump - and then it wouldn't boot up at all. This sucked because I'm also using it as a flash ROM programmer! That turned out to be the bus breakout board having gone faulty, how I don't know because all it is is an IDC connector and some pins and pull up resistors so I can plug it into a breadboard! Well, that's a debugging job for tomorrow... I programmed a new ROM 'blind', since the LCD is on the breadboard and I couldn't use it. Fortunately it worked).