Overclock.net › Forums › AMD › AMD Motherboards › Overcoming the shortfalls of the M5A78L-M/USB3 (AM3+)
New Posts  All Forums:Forum Nav:

Overcoming the shortfalls of the M5A78L-M/USB3 (AM3+)

post #1 of 4
Thread Starter 
It's been an interesting experiment to say the least.

Getting a truly stable overclock (FX-6100) on this board has been quite the challenge. If you find yourself stuck with one of these, then it can be overcome, provided you lower your aspirations slightly.

Despite lasting through long prime sessions, I had the problem where the clock multiplier would drop down to x7. There is of course a lot of conjecture on what causes this. The obvious place to check would be the windows power scheme (best set to high performance). Then there's the bios settings, C6 State, CnQ, Turbo Core disabled. But where I actually found my gold ended up being the subject of TDP throttling.

Of note were posts I've read about the VRM (voltage regulator module) overheating.

On a whim I thought I would redesign the airflow of my system a little, instead of blowing through the cpu heatsink, I switched it to sucking air out of the heatsink, a very short distance to the exhaust/rear system fan.

airflow.jpg 18k .jpg file

My results were actually quite surprising.

For the cost of a couple extra centigrade at load (well within acceptable limits (48c), the throttling completely disappeared.

So it would seem, if you're stuck with a cheap board, it may be that you consider the board, before the cpu.

Later I plan to replace the fans, and get an extra one to try a push and pull scenario, but the system as it is, would probably survive without such an investment.
post #2 of 4

If the PSU has a top fan blowing in, the deflection in airflow must be reaching the VRMs.  That's a huge step up in terms of motherboard stability considering this is a low end 3+1 phase system.  It is still a low end 3+1 phase system not designed for overclocking and you should pretty much completely avoid that or ensure total VRM cooling with heatsinks that reach more of the airflow.  That protective throttling feature is there for a good reason.  It's there to prevent motherboard failure!

post #3 of 4
Thread Starter 
Thanks for the info.

I just stumbled upon your thread:- http://www.overclock.net/t/943109/about-vrms-mosfets-motherboard-safety-with-125w-tdp-processors whilst on a google search, some good reading there.

Looks like it'll be quite tricky to get some extra cooling in under there, but maybe. The idea of ghetto mounting the spare stock cpu fan/heatsink I have is probably the road to insanity.

Anyway, I'll get back to reading your thread, lest I let insanity run away with me.
post #4 of 4
Thread Starter 
Sorry for the double post, was hoping I could ask a question..

Pretty sure the board is 4+1, at least, there are 5 1R2 Mosfets next to the CPU socket.

The processor I'm using is the FX-6100 95w and the board is supposedly rated to 125w (I can bet that's over-reaching) so I wonder if that gives me any wiggle room. I've jury rigged a fan to blow a little more air down to the mobo, and under load the 'motherboard temp' (TMPIN0 is CPU and TMPIN1 is mobo on this board) stays <35c.

Having read around, I'm thinking some mosfet heatsinks and some thermal adhesive might be the way to go long term.

With all that in mind, is it anything but madness OCing on this board?

450

Kluge is the current vernacular I believe.
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: AMD Motherboards
Overclock.net › Forums › AMD › AMD Motherboards › Overcoming the shortfalls of the M5A78L-M/USB3 (AM3+)