New build RDX200, odd problem.

[Radiohead57]

A friend of mine asked me to buld him a computer and he already had the mobo CPU RAM etc so I undertook the task today.
CPU 939 3200 DH-E6
2X152MB Hynix RAM
6600gt PCI-E vid card
Mushkin 550watt power supply (must say I'm extremely impressed with this unit)
LG DVD-RW
Samsung Spinpoint ATA100 80GB HD (new)
Assembled all parts, proceed to initial boot using one stick of RAM with all mem settings at "auto" in the bios. Machine booted straight away without issue and proceeded to install OS. Got to about 8 minutes remaining to finalize machine reboots with 0A error. Attempted OS install second time and oddly enough even though I hadn't changed anything the OS went on without issue. Keep in mind I'm very proficient with Nvidia chipsets but this is my second go with the RDX (purchased mine a month or two back for next to nothing on eBay which surprised me considering they were still fetching a good price retail) my first install on my machine was smooth as silk and it continues to run very nicely and I must say I was pleasently surprised, I actually figured it might not even work considering the price. Anyhow back to my tail of woe. The box was up and running downloaded a few apps drivers etc after first installing the chipset drivers from the install CD which included the Marvell lan drivers both installed without issue. Accessed net without issue through my cable modem/router (all in one Efficient Networks).
Heres where this gets interesting, I had used identicle sticks of this memory in a couple of my own builds and they ran well in DC mode and actually recommended them to him when he was making his RAM purchase. It was at this point that I inserted the second stick which appeared to be OK. I had previously popped both sticks in one of my DFI NF4 boxes to run memtest on them beforehand and they ran 4 passes at 220mhz 8-3-3-2.5 (sounds like they should be fine right?). I began getting reboots and reporting the errors via error reporting to MS, they returned driver issue several times. The only driver I had installed outside of the chipset were the vid drivers for the EVGA 6600GT so I uninstalled the vid drivers through control panel, went to the Nvidia folder which MS neglects to remove, deleted the folder and contents and emptied the recycle bin. Then used driver cleaner pro to remove any remnants. Still getting errors even without the vid drivers.
At this point I started over with a single stick of RAM again, reformatted the drive and reinstalled the OS and met with better results results after loading a few apps etc, this is before loading any vid drivers. Inserted the second stick only to turn up errors again this time looks like RAM errors, removed second stick issue gone ie no more reboots etc so I went ahead and loaded a different set of vid card drivers. All was well benched the vidcard without issue, ran demo for sound in 3dmark all remains well. Loaded the second stick of RAM again then errors start popping up again. Ah so its the RAM I'm thinking. Swap out the Hynix sticks for a pair of Corsair Value select and the system appears to be solid as a rock now except for one issue. I have loaded speedfan, everest home edition and every time i try to access any monitoring features for temps voltages etc in Everest the machine simply reboots. I can look at the system summary etc and all appears well but the second I hit the sensor icon in everest down we go, the second I load speedfan down she goes. I'm kinda at an impasse here, do I reinstall windows hoping that the hynix may have been at fault when doing this OS install (still loaded the OS with the hynix but began getting errors as explained earlier which prompted me to swap ot the hynix for the Corsair). I guess I'll give that a shot perhaps the "giveio.sys" file is somehow messed up causing the monitoring apps to crap ot but at this point I'm unsure.

[vaioman]

Seems a clean reinstall with the Corsair ram is the best starting point......would rule out any ??? about the os install at least.

[Radiohead57]

Went ahead and reformatted and reinstalled with the Corsair in place (both sticks). The box seems just fine now and after doing a bit of research found out this is very common and has to do with how the 3rd party monitoring proggies scan the SMBus of the system. Brings them down consistently. The packaged smartguardian software works as will speed fan with a little fiddling to tell it not to scan the SMBus.

Alls well.