On Wednesday, March 07, 2012 02:43:56 PM, Kristoffer Walker wrote:
> I'm trying to patch together a machine with parts I've been stashing
> in the attic. I have a case, but the motherboard doesn't fit inside
> it. The power supply and disks are in the case, but the mother board
> is sitting outside of it, on a block of wood. So, it should be no
> surprise, really, that when I turn on the power switch it doesn't
> start. No spinning fans, no spinning disks, nothing except a barely
> audible electronic sound, and I can't even tell where that is coming
> from. The sound fades away after a few seconds.
> 
> Of course there could be a multitude of problems preventing this
> machine from ever starting, but I was wondering if the motherboard
> needs to be grounded to the case in order to work? Anybody know?

I agree with Thor.

The motherboard gets DC ground from the power supply.  I think some computer 
cases are grounded, but that need not be the case for the motherboard to 
operate.  I've occasionally run motherboards on a bench to make sure they work 
before putting them into a case, and that has always worked fine.

Here are some ideas from things I've run into in the past.

Years ago with a 600 MHz AMD Athelon I ran into the problem you're currently 
having, where the motherboard wouldn't start on the bench.  The problem turned 
out to be that I had attached a CPU fan that didn't have an RPM sensor lead, 
so the motherboard refused to start.  I seriously thought the motherboard was 
broken until I remembered seeing fan RPM readings in the BIOS before and had 
the idea to change the fan for one that had an RPM sensor -- then the 
motherboard started right up.

I've also seen motherboards (some old PIII's) that intermittently refused to 
start because they came with a buggy BIOS (which is presumably why the owners 
of these boards got rid of them).  When I finally did get it to start I 
flashed a new BIOS version and that fixed the problem in that case.

The jumper headers to attach the power button or reset buttons are 
occasionally odd.  I don't remember if the reset button leads are normally 
shorted and pressing the reset button breaks the connection (i.e. "mate before 
break"), or if it's the opposite such that you might need a jumper in place of 
the reset button leads.  [I think it's been too long since I've had a 
motherboard on a bench.]  I also seem to recall occasionally having to place 
certain jumper headers in reverse of what the label says, such as in the case 
of the speaker connector or the HDD LED connector.  ATX motherboards have to 
be commanded on, usually via the power button attached to the jumper header.  
[I believe that's just a momentary switch.]

Switch-mode power supplies (which is what all computers use) don't like the 
"underloaded" condition.  As such, some switching power supplies refuse to 
start unless they have enough of a load attached to them.  My experience 
thusfar has been that the motherboard + CPU + video card was enough, but I've 
occasionally heard stories otherwise.

  -- Chris

--
Chris Knadle
[email protected]
_______________________________________________
Mid-Hudson Valley Linux Users Group                  http://mhvlug.org
http://mhvlug.org/cgi-bin/mailman/listinfo/mhvlug

Upcoming Meetings (6pm - 8pm)                         Vassar College
  Mar 7 - Desktop Shootout - 9th Anniversary of MHVLUG
  Apr 4 - An Intro to Chef
  May 2 - May 2012 Meeting

Reply via email to