Question MSI B85M-IE35 Motherboard, random power losses?

Jan 14, 2021
1
0
10
0
(EDIT) Okay everyone. I have found the issue and have since fixed it and the board is working a real treat. I will include the solution and the reasoning that lead me to find it in case others happen across similar issues.

I came across videos and forums about X79 boards including M.2 slots and remembered that X79 doesn't support M.2 and NVME slots. So This lead me to popping the heat sink off my chipset to see if I had a rebranded chipset that wasn't actually B85M. The chipset is legit on my board however I noticed that the heatsink was held in place with just a very cheap adhesive pad apart from the two cheap plastic clips at opposite corners. It was a real pain getting this adhesive off the die and heat sink but I did it. The adhesive was heavily browned and the die itself had signatures of very high heating.

I applied some silver thermal paste (Phoronix brand) and reapplied the heatsink and low and behold the power losses no longer occur and the board works great. It is unusual however, that I did not get any anomalous temperature readings from my many hardware monitors but it may be that it was just reporting inaccurately.

So, if anyone buys an MSI B85M-IE35 motherboard, be sure to apply thermal paste to the chipset and be extremely careful not to damage the exposed silicon die. Hope this helps anyone that got this board and may be experiencing the same issues I describe below in the original post.

(ORIGINAL POST) Hi all! I have recently taken to building budget gaming rigs using outdated xeon cpus and older chipsets that can still be leveraged for decent mid-range gaming performance. I have come across an issue that I cannot seem to get to the bottom of.

First here are my hardware specs:

mobo: MSI B85M-IE35

CPU: Xeon E3-1220 V3

GPU:
Gigabyte GTX 1650 super
Nvidia 8800 GT
Nvidia GT 610

RAM:
1333MHz 2x 4Gb Crucial
1600MHz 2x 8GB Crucial

PSU:
Raidmax rx750ss 750W 80+ Bronze
Cooler Master rs-460-pmsr-a3 400W

HDD: Seagate Barracuda 1TB 7200rpm
SSD:
Samsung evo 1TB
TeamGroup 500GB
SanDisk 240GB

Control Hardware for testing ~

HP Proliant ML110 G6 workstation mobo
Xeon x3430 cpu

So I am currently using the above listed HP mobo with the x3430 cpu. The ram I am using is the 1333MHz 2x 4GB crucial dimms. The PSU I am using is the Raidmax rx750ss. The storage devices I am using are the TeamGroup SSD as a boot drive and the seagate barracuda for mass storage. I do not get any of the issues I listed below with the current setup on the HP mobo and I have used it as a workstation for the last 5 years with no issues.

The MSI board, when paired with all of the above hardware, except for the x3430 cpu since the chipset doesn't support it nor does the socket LGA1150 support x series xeons, consistently and very randomly losses power and in about 2 seconds powers back up.

This phenomenon occurs while in BIOS, while booted in safe mode with only core drivers, with alternative OS's such as Linux and freebsd, and during normal windows 10 operation. The only factor I have not been able to test with different hardware is the cpu for the MSI LGA1150 board. All of the hardware listed, except for the MSi mobo and xeon e3-1220 v3 I have been using for the past 4-5 years without issue and I know them to be in good working condition.

I have connected both psu's to an oscilloscope to see the continuous voltage and current draw during both idle and max cpu/gpu usage scenarios and I have seen no unusual numbers whatsoever. I have swapped out between both power supplies, all GPUs. RAM sticks in both single and dual channel (no xmp profiles) as well as single dimms in single channel mode. The power loss event happens only when using the MSI board and the e3-1220 v3 cpu. Both PSUs will randomly power off when using the MSI board and e3-1220 v3 cpu.

I have probed the power rails to the cpu, the VRMs, the PCI express lanes, the PCI power lanes, tested every individual capacitor on the board, tested all resistors, and I have constantly monitored thermals of all heat-prone components on the board both using thermal cameras judging by delta over ambient temperatures as well as software monitoring using CPU-Z, Speedfan, MSI afterburner, Intel Extreme tuning utility, and HWmonitor. I have not recorded any unusual heat spikes and at max cpu load my recorded temps never get higher than 65c. GPU never exceeds 60c at full load. I have recorded no anomalous current surges on my scope from GPU or cpu during idle and max load scenarios with both benchmarking and stress tests for as long as will go before power loss.

My last ditch effort has been the most up-to-date BIOS for the mobo as well as 5 previous BIOS versions for this board. The same event occurs no matter the BIOS version I use.

The only thing I can think of is the CPU may be causing this due to possible incompatibility with EIST so I did two things: 1.) I disabled intel speed step so that the cpu state was always C0. Power loss still occurred. Then I made sure the PSUs were compatible with EIST (since they are slightly dated PSUs) and they are both compatible with cpu state C6 and C7. So EIST was not the culprit.

My last option is to obtain another LGA1150 E3 xeon and a 4th gen i3,5,7 to see if this still happens with different cpus. Before I waste money on another cpu, do you guys have anything I can try that I may not have thought of yet? I am completely stumped.
 
Last edited:

ASK THE COMMUNITY