Active Phase Switching(APS) help

[purple stain]

My mobo has APS(Active Phase Switching).Whenever the CPU goes under load the APS throttle's the voltage depending on how much the CPU is being stressing.My stock volt is 1.4 and under load it rises to 1.46,with no O.C.
How am i suppoed to O.C. if i cannot turn off APS?

Phenom ii x4 965 @3.4ghz
MSI 890FXA-GD65
4G DDR3 1600mhz
XFX6870
Seasonic 520watt.

 

[purple stain]

BUMP

 

[purple stain]

no one knows?

 

[Leaps-from-Shadows]

APS doesn't do anything to the voltage. What it does is activate extra phases when the CPU is under stress. Extra phases smooth out the power going to the CPU, theoretically allowing for a higher overclock.

It's normal for a CPU under stress to be fed more voltage when the voltage is set to Auto in the BIOS. If you set it manually, it shouldn't fluctuate as much. Note: 0.06v isn't much of a fluctuation anyway.

 

[purple stain]

Thats a huge flucuation...If i had the stock cooler on i bet it would overheat it because of the flucuation.

I've set it to manual before but the same thing happens.I feel like i'm missing something in the BIOS but i've tryed everything and i've sent it back before.

 

[Leaps-from-Shadows]

No it is not a huge fluctuation. Many voltages on the board routinely fluctuate -- the +3.3v, +5v, and +12v rails likely do as well. You may think 4% is big, but it's not.

1.4v not overheating versus 1.46v overheating? lol If you were that close, then 1.4v wouldn't be a safe voltage anyway.

Try overclocking it. It's entirely possible that at higher clocks it won't fluctuate as much. I know this happens sometimes, as it has happened to me on one particular Gigabyte board.

 

[purple stain]

The mobo i had before this was a gigabyte 770t-usb3 and it never had flucuating voltages it always remained constant.
Here's my thing.Say i'm going for an extreme O.C. and the max volt for my CPU is 1.55.Now i lets say i have it set at 1.55 but with the APS it will go up by .06 which means it will actually be at 1.6volts not 1.55.Thats where i see the problem is.
I've seen the rails flucuating and i think thats normal but not the cpucore.I think that should stay constant.

I did try O.C.ing and got to 3.9ghz with the muti at x19.5 then i used the O.C. Genie to get another 200mhz out but it failed the stress test.O.C. Genie tool is kinda nice tho it auto changes the voltages on the NB and HT.Although i don't think it does it for the CPU.

 

[Leaps-from-Shadows]

What I'm saying is that the "fluctuation" may not be constant -- they usually aren't. At a higher clock it may smooth itself out.

If you think it's such a huge problem, get a different board. And by different, I mean something other than MSI as there isn't a better board from them. I haven't heard of any overclocking problems with that particular board, but it's certainly possible. You can go with ASUS or Gigabyte.

 

[purple stain]

I went looking around for the exact same problem and i haven't found anyone with it.I already RMA'd it thinking it was just a voltage error but it seems all their boards have this "thing".I must be the only person that finds it to be a problem.

I have tryed it at higher speeds and at higher volts and it still comes back with the same result.It will remain stable at whatever volt you set it to and then when it's beening stressed it will jump up .6 or depening on how hard it's being stressed.Since APS delievers volt depending on how stressed the CPU is i can only assume thats whats doing it.

It's a real shame because this mobo is the best.8+1 power phaseing and x2 pci-express x16 slots running at x16.

 

[bardacuda]

I went looking around for the exact same problem and i haven't found anyone with it.I already RMA'd it thinking it was just a voltage error but it seems all their boards have this "thing".I must be the only person that finds it to be a problem.

I have tryed it at higher speeds and at higher volts and it still comes back with the same result.It will remain stable at whatever volt you set it to and then when it's beening stressed it will jump up .6 or depening on how hard it's being stressed.Since APS delievers volt depending on how stressed the CPU is i can only assume thats whats doing it.

It's a real shame because this mobo is the best.8+1 power phaseing and x2 pci-express x16 slots running at x16.

that really is a shame....so maybe the mosfet's don't fry but it wants to fry your CPU? I got their NF980-G65 and an Athlon II X4 640. Had to put the voltage to 1.488 to OC to 3.6GHz and when stressing it went as high as 1.536V. I have since backed off my OC to 3.5GHz @ 1.432V since learning about their poor VRMs/MOSFETs/heat-sink mounting and it only goes to 1.488V or 1.456V when stressing. It's like you have to decide what you feel is a safe voltage for your CPU and set the voltage .05V lower in the BIOS, and then figure out what clocks are stable with that setting.

 

[purple stain]

Yah thats why i'm thinking of sending it back.Thats exactly what i did.I set it back .06 because when it's under stress it will rise that amount.Just really wierd that this happens and no one else has said anything.I'm going to refund my mobo and just get a SSD instead.Then i'll wait until the new AM3+ mobos come out.I know the mobo i have is capable of AM3+ with a BIOS update but i'm sure the new AM3+ mobo's will be more capable and they will start the 9xx series.

 

[bardacuda]

It may be too that that board is more advanced when it comes to voltage regulation and it can detect what CPU you have and it's temperature so if it feels like it can handle .06V it'll give it to it whereas with a stock cooler it might only give it .01 or .02V because the temps are higher. That's just pure conjecture though 😛

 

[purple stain]

That doesn't make sense.How does the mobo know which cooler i have on it?lol

 

[bardacuda]

I'm not saying it knows what cooler you have but it does know what your idle temperature and your temperature curve looks like when adding voltage and could theoretically extrapolate from that the combined influence of your cooler and your ambient temperature. Not saying it does that just saying maybe who knows?

EDIT: I'd say you are better waiting for a 9x0 board though. Who knows how well the AM3 ones will work with Bulldozer or what features/RAM clocks/HTT clocks you will get on one? And you don't want to put in a nice shiny new $400 (or however much) CPU and see your voltages all over the place and be worried about frying it when AMD does not support it and won't observe the warranty.

 

[purple stain]

But still even if it did detect my temps...Lets say i have my voltage set to 1.55(that is the max volt for my processor) and my temps are below the max yet the mobo jump it up another .06.That would make my voltage go to 1.6+ and totally destroying my CPU.

And yes thats what i'm thinking of doing.This board is very good and at a very good price but i guess i'll wait till the new boards come out and have a stronger NB/SB and many other features i'm sure.

 

[bardacuda]

Well it wouldn't necessarily destroy it but I definitely agree it's not good and would accelerate it's lifespan considerably. I actually had this athlon at 1.58ish for 5 or 10 minutes just to see if I could get a stable OC higher than 3.6GHz but it wouldn't take. Did you try setting it at like 1.52 or 1.53 and stressing it to see if it levels off more at higher voltages? Not saying you should risk it just wondering if you already did before.

EDIT: Oh nvm I see in your post just before my first post that you did. Just out of curiosity though what's the highest you had it set in BIOS and how much did it go up to while stressing before you shut 'er down?

 

[purple stain]

The highest i've set it is 1.5.And it rose by .06 so it was at 1.56,thats the highest that i'm comfortably going.It always rasies by .06.I've never had experience with MSI before so i guess this is how all of their mobos must be.DrMos and APS is their "speciality" they say.But those are the things that make the voltage do that yet their is no way to turn it off in the BIOS.

 

[bardacuda]

Hmm I had hoped MSI turned things around with this board, and although the board itself appears to be more robust, it's not good if it's going to do things like that to your CPU voltage. 🙁 Hopefully MSI fixes this issue as well with their new boards.