[1000mp] AW: one last request for suggestions

Uwe Fleischer uwe.fleischer at asco-online.com
Thu Dec 7 10:06:32 EST 2006


Hello,

> It ususally
> takes a few hours of contesting to make it happen. I had it open several
> times, and have been through the reseating thing, and it always came back
> to life before I could track down the problem.

Both facts
a) the problem occurs after a longer period of heavy duty use
b) the problem disappears after opening the cover

lead to the conclusion, that the problem has any relation to temperature. As
already mentioned in another thread, two decades ago, I've repaired far over
thousand computers and peripheral devices. One knowledge remaining in my
mind since this time is, that there is an efficient pair of weapons to fight
against problems like these:
a) A hair dryer which can produce hot air (attention: do not use another
"hot air equipment", which can produce more heat than about 60 or 70°C,
otherwise you may severely damage your components or solder them out of the
board)
b) A special (CO²) spray to produce very cold temperatures (up to -50°C)
when spraying it out of the tin. The spray head can be equipped with a very
thin and flexible tube (which is part of the product), so that you can aim
exactly at one small point when spraying.

Best effects can be gained, by waiting until the problem occurs, with the
screws already removed from the cover. Quickly remove the cover and try to
cool down distinct components by spraying on them and watch what happens.
Start with obviously critical circuits first, like CPU-chips, PLL- or
DDS-ICs, or other "big" custom-chips.
If you assume, that you've discovered the bad component, try to reproduce
the problem and the termination of the problem two or three times. If the
effect is always the same any time you test it, then after cancelling the
problem by cooling the component down, you can carefully try to produce the
negative effect, by heating up the component with the hair dryer.

With a little patience, this method is very effective in isolating
components that produce temperature related malfunction. Although I must
admit it is a little more complicated, if the problem resides on a board,
deeply nested between other boards  :-(

55 and pse give a feedback if u tried it out and it worked.

73 de Uwe, DL8UF




More information about the 1000mp mailing list