View Single Post
  #7  
Old April 2nd 07, 07:41 AM posted to alt.os.windows2000,alt.comp.mainboards.abit
Kyle
external usenet poster
 
Posts: 86
Default Computer freezes upon moving it

"Ron Gibson" wrote in message
news | On Sun, 01 Apr 2007 00:41:59 -0400, Kyle wrote:
|
| Add to the list poor electrical contact at the crimp locations for
the
| Molex HD power connector pins, which caused intermittent HD
problems with
| one system I worked on. It sure sounds like a poor electrical
connection
| somewhere, or even more nightmarish, a broken circuit board runner
that is
| making and breaking in response to mechanical movement causing
flexing of
| the circuit board.
|
| Boy you got the nightmarish part right as this type of failure can
be a
| real PITA.
|
| I have had the HD power plug do that more than once. Actually you
| shouldn't have this problem unless you are unlucky and got a lemon
| connector or in my case I'm plugging and unplugging drives all the
time.

In my case, the connector was fine, the contact pin (that slides into
the connector) crimped on the end of the wire was making good contact
with the insulator and intermittently touching the conductor.

|
| Most frustrating one I ever had to trace down ( the first time) the
| connection would only fail when I was using Drive Image to backup
about
| 10GB of data - Using the copy partition function. It would make it
about
| 3/4 of the way through and then lock up. Small backups were not a
problem
| and I could boot both drives without error.
|
| After a LOT of cussing I found that one of the wires (1 red, 1
orange
| and 2 blacks) was ever so slightly loose were it slipped into the
back
| of the plug. My guess is the high throughput under those conditions
made
| just enough heat to expand the fitting and open that connection.
|
| I never narrowed it down more than that as I just took some wire
cutters
| and cut it off and wired up a good plug. It took me a good 4 intense
| hours of work to find that.
|
| But I could have done it faster if I'd followed the suggested method
of
| removing every thing you can and still boot. Then add one device at
a
| time until the fault shows up - But heat of the moment and all
that...
|
| Another culprit to watch out for especially with 80 wire EIDE ribbon
| cables is a broken or failing connector where the wires attach to
the
| plug. Those are small wires and they can't take a lot of stress -
| remove by pulling the connector NOT the cable. Use needle nose
pliers
| if you have to.
|


Well, I didn't mention the other PITA electrical connection problem,
but cold solder joints can also give the intermittent failure
symptoms. I've spent plenty of time tapping on circuit boards in my
day. If one does try this, use a non-conductive tool for the
"tapping" and tap lightly. Seems that such problems are less common
with the vapor phase soldered SMD boards, but one can never be too
confident with any manufacturing process as things can fail anywhere.
I've seen dirt on a mobo interfere with the operation of the IDE
controller, simply brushing the dust/dirt of the mobo fixed the
problem.

And the comment on IDE cables is very pertinent, as I tend to pull on
the cables to remove them, which I know is a bad habit, but getting
needle nose pliers to grip the connectors is equally a pain. Too bad
the old connector ejectors (which add maybe a few cents to the cost of
the connector) are no longer in use, they DID server a valid purpose.
--
Best regards,
Kyle