2 caps ripped off board but still boots. Need help

[Flying-Q]

I have been given an HP mATX 1366 board and i7 920 chip with 3x 2GB from out of a Pavilion Elite e9180t. It boots to BIOS and HP diagnostics, but it has 2 of the 3x 560uF solid caps ripped off next to the furthest memory slot. The diagnostics shows no errors (apart from no bootable drive of course). Also what little info the BIOS gives regarding voltages they look OK. For a couple of pounds I can get a pair of replacement caps from mouser electronics to solder on.

What part of the board are these caps associated with? I'm pretty sure they are not part of the VRM circuitry as those are way over the other side of the board, between the socket and the I/O. I thought they might be memory power supply, but surely that would show as an error in diagnostics?

Is it worth trying to repair the board? Any thoughts please?

 

[XxMongermaniacxX]

hahaha that is so wierd. I think if it is working fine through testing and everything, then there is no point

 

[makkem]

Hi
Let us assume those are capacitors for the RAM.
The purpose of capacitors for this purpose is to remove any AC component in the power supply to achieve a noise free supply and so allow the RAM to run perfectly and also they store power so as the load on the RAM increases so its power draw increase causing the supply voltage to drop as the VRM cannot compensate rapidly enough, as the voltage drops below the voltage in the capacitor it will discharge into the line holding the voltage constant.
Note that these are very raid transient effects and are too fast to show up on voltage monitoring software.
The effect without the capacitors ?
If you have a decent PSU probably none though it could have the effect of very slightly reducing the speed of the RAM,in effect a very slight underclock.
If it was me I would not replace the capacitors before trying it with software that uses a lot of RAM and if there was crashes caused by the RAM then I would fit the capacitors because of course there is a slight risk of damaging the motherboard when soldering.
And if it is not the RAM power then they will be doing something similar on something else and the likelyhood is that them not being there is not going to give a noticeable performance hit.

 

[mctylr]

[...] 2 of the 3x 560uF solid caps ripped off next to the furthest memory slot. [...] I thought they might be memory power supply, but surely that would show as an error in diagnostics?

If the 3 capacitors are in parallel, then the one remaining capacitor is most likely trying to do the work previously done by all three, left as-is it may prematurely fail, but not necessarily instantly. If it is part of the voltage regulation for the memory, then the ripple voltage from the DC-DC converter will be increased, possibly damaging the RAM modules, or generating memory faults.

The BIOS/UEFI self-test on boot is not comprehensive, and often fails at detecting minor fault. It is designed for testing to ensure major components or subsystems are installed and working at the most basic level.

Is it worth trying to repair the board? Any thoughts please?

I would say it would be a crap shoot, but if you are experienced with a soldering iron and SMT, then it is pretty inexpensive and not terribly time consuming to try.

If it was my primary computer, I would replace the motherboard to be cautious against component failure and data lost.

If the capacitors are through hole, ensure you get a replacement with the same (or smaller) case diameter, and same (or very close) lead spacing. With surface mount, ensure you get the correct package so the solder tabs align to the pads on the motherboard.

 

[Flying-Q]

I can get the exact model capacitors (link in top post) for literally pennies; postage is likely to cost more so I think I shall go ahead and fix. I'd like to give best answer to both makkem and mctylr but I have to toss a coin.




Edit: makkem got it, sorry mctylr.