Gigabyte 990fxa-ud5 vs SABERTOOTH 990FX R2.0

[DatDirtyDawG]

Yep I remember feeling somewhat frustrated at the absurdly dumb responses I got in the beginning but thanks to the rest of the group it really turned around quite nicely. In fact, as suggested I ended up getting the Sabertooth and am very very happy with it (for the most part...) barring this one annoying issue where it sometimes hangs in the post screen (one in every two or three boot ups requires power down and do it again) Changed the battery, went to stock default speeds, swapped out memory sticks and nothing. I think its the motherboard (which is covered by warranty) but *ugh I'm not in the mood for swapping that out .

Anyway like i said, other than that the board is quite fantastic. My 8350 is running oh so nicely at 4.5, encodes my gaming videos like a champ and with my new GTX770 it plays every game to date without a hitch. Couldn't be happier

 

[RGone]

DDDG, have you got another power supply to try and see if you still have booting issues?

I have seen P/S, ram or amount and number of sticks and just an odd setting in bios cause that multiple reboots and then fine.

The board is seeing something at first boot that it does not like and then goes ahead. P/S gave me fits once with a heavily overclocked board.
RGone...

 

[DatDirtyDawG]

DDDG, have you got another power supply to try and see if you still have booting issues?

I have seen P/S, ram or amount and number of sticks and just an odd setting in bios cause that multiple reboots and then fine.

The board is seeing something at first boot that it does not like and then goes ahead. P/S gave me fits once with a heavily overclocked board.
RGone...

Actually I do, I can give that a shot. I have an old OCZ 850W SLI certified that works. That would suck because the one I'm using is a nice 1200W heavy duty Raidmax. I'll give that a try

 

[RGone]

Actually I do, I'll give that a try

Good luck man.
RGone...

 

[Johan45]

DDDG, have you got another power supply to try and see if you still have booting issues?

I have seen P/S, ram or amount and number of sticks and just an odd setting in bios cause that multiple reboots and then fine.

The board is seeing something at first boot that it does not like and then goes ahead. P/S gave me fits once with a heavily overclocked board.
RGone...

+1^^
I had a Raidmax "ONCE" never again these words shouldn't go in a sentence together.

a nice 1200W heavy duty Raidmax

It probably has 4 small +12v rails on it. I had a 730w that drove me around the bend trying to figure out why my computer kept giving me the Nvida driver stopped working error. It couldn't run 2 9800's together in SLI at idle let alone under load.
I don't mean any offence, they look nice but it's only skin deep.

 

[DatDirtyDawG]

DDDG, have you got another power supply to try and see if you still have booting issues?

Well I tried another PSU and same thing. So annoying...it's a small issue in that windows will launch on the second try almost every time and I have zero issues once I'm up and running. It's just one of those little things that nags at me lol

+1^^
I had a Raidmax "ONCE" never again these words shouldn't go in a sentence together.
It probably has 4 small +12v rails on it. I had a 730w that drove me around the bend trying to figure out why my computer kept giving me the Nvida driver stopped working error. It couldn't run 2 9800's together in SLI at idle let alone under load.
I don't mean any offence, they look nice but it's only skin deep.

No offense taken and sorry to hear about your experience. My experience however has been nothing less than stellar since the day I bought it. Had it for a good year with 2 GTX 560 TI in SLI (those bad boys draw quite a bit of juice...) and everything ran flawlessly. Lot's of GPU Stress tests, tons of gaming and the like all without a hitch and my 1200W Raidmax purring like a kitten. My model: RaidMax RX-1200AE

This whole boot up thing started when I got my new Sabertooth/FX8350. In fact I even swapped out the 8350 and put my old 1090T back in (just to make sure it wasnt the processor) and again, same crap.

 

[RGone]

Sabertooths don't normally do that shett. I have had boards that with certain power supplies did that crap. Is why I suggested you check another power supply. I think they have a bad "power good" circuit response with the power supply.

I would try and REflash the same bios over the top of itself and afterwards power down completely remove power supply plug from wall outlet and remove battery for about 3 mins, in effect a good CMOS clearing, then battery back and REapply the power to power supply and see what happens. Maybe that will give a different symptom of fix the crap.
RGone...

 

[Johan45]

This is what I use now. After my experience with the Raidmax I vowed to never get a multirail PSU again. Half way down the page it reads 83a on the 12v rail. Yours has 40a. The PSU is also trying to start a much heavier load now with the 8350 4x4 sticks of ram etc. and like RGone says the sabertooth is a smart board and if it doesn't like something it trips the self preservation mode.

 

[DatDirtyDawG]

This is what I use now. After my experience with the Raidmax I vowed to never get a multirail PSU again. Half way down the page it reads 83a on the 12v rail. Yours has 40a. The PSU is also trying to start a much heavier load now with the 8350 4x4 sticks of ram etc. and like RGone says the sabertooth is a smart board and if it doesn't like something it trips the self preservation mode.

What does that mean 83a on the rail vs 40a? The difference between this system and the last is the board and processor, that's when this boot issue started. Since then I got rid of the the two GTX560TI's and went with a single GTX770 SC. I also swapped out a mem stick after I discovered one was bad (very subtle..I only discovered it via memtest) otherwise there was no indication of bad memory as the board recognized them all. I used memtest more so to eliminate bad ram as the culprit but even after removing the bad ram the system still hangs every other or every third attempt. Ive since replaced and retested all the ram they are perfect as well. I'm going to try what Rgone suggested (reflash the bios, unplug the PSU from the board, remove battery etc)

 

[RGone]

Power in watts is Voltage times the Amps

MODEL RX-1200AE
AC INPUT: AC 115V-230V, 15A , 50-60HZ

DC OUTPUT: and the MAX OUTPUT CURRENT: for that voltage and rail.
+3.3V = 24A

+5V = 30A

+12V1 = 40A

+12V2 = 40A

+12V3 = 40A

+12V4 = 40A

-12V = 0.5A

+5VSB = 3.0A

MAX COMBINED POWER:
170W = total power output in watts for +3.3V and the +5V rails/output.

1020W = total power output in watts for all 4 +12V rails added together.

6W = total power output in watts for the -12V rail.

15W = total power output in watts for the +5VSB rail.

TOTAL POWER: = 1200 watts rating.

Now remembering that Power in watts is Voltage times the Amps, let us do a little math.
#1 +12V rail is rated at 40 amps so its watts output would be 40 x 12 = 480 watts.
#2 +12V rail is rated at 40 amps so its watts output would be 40 x 12 = 480 watts.
#3 +12V rail is rated at 40 amps so its watts output would be 40 x 12 = 480 watts.
#4 +12V rail is rated at 40 amps so its watts output would be 40 x 12 = 480 watts.

So looking at the individual +12V rails and their watt rating per rail we would think we could add the watts rating of each +12V rail together for the 12V watts output. If we do that we would have a total of 1920 watts of power available on the +12V circuit. NOT so. They only rate the 4 x +12V rails to output a total of 1020 watts.

One (1) of the +12V rails might have a draw of 480 watts on it. That would only leave 540 available watts between the other three (3) +12V rails. WHEN I used to use power supplies with multiple +12v rails; I practiced "load balancing" in order to draw more or less equal power from the 2, 3 or 4 rails of the +12V power output.

In order to do "load balancing" one has to know how the power supply has set the various +12V rails to output the power for the rail. Which +12V rail is mostly powering the cpu. Which +12V rail is mostly powering the video cards and so forth so that each +12V rail is loaded about the same. Good power supplies make this "load balancing" process simple. They make judicous splitting of the connectors so that the balancing is more or less done for the end user by which connector does what in supplying power. Plus good power supplies will have the information available as to which +12V rail is supplying which type output connector so the end user can ensure he is practicing "load balancing'.

That is the long version of what "johan" meant by his +12V being 83Amps since he has only a single +12V rail type power supply and you have a four (4) +12V rail type power supply each of which says it can output 40 amps, which the math shows is not precisely so at least surely not all four (4) +12V rails supplying 40 amps at the same time.

This is not an indictment of one type power supply against another, but rather a how it is.
RGone...ster.

 

[DatDirtyDawG]

That is the long version of what "johan" meant by his +12V being 83Amps since he has only a single +12V rail type power supply and you have a four (4) +12V rail type power supply each of which says it can output 40 amps, which the math shows is not precisely so at least surely not all four (4) +12V rails supplying 40 amps at the same time.

This is not an indictment of one type power supply against another, but rather a how it is.
RGone...ster.

Well I can't thank you enough for the thorough breakdown (and in a way I can wrap my knucklehead brain around)

Why would a manufacturer design one with 4 rails vs another one with one? Is there an advantage to using one over the other? As per Johan he mentions this in his post as an advantage. Is the 4 rails at less amperage causing a deficiency triggered by the board/processor upgrade that just wasn't there before when I ran that PSU with a M4N98TD EVO/1090T? I had no idea PSU's were this involved.

 

[Johan45]

Well I can't thank you enough for the thorough breakdown (and in a way I can wrap my knucklehead brain around)

Why would a manufacturer design one with 4 rails vs another one with one? Is there an advantage to using one over the other? As per Johan he mentions this in his post as an advantage. Is the 4 rails at less amperage causing a deficiency triggered by the board/processor upgrade that just wasn't there before when I ran that PSU with a M4N98TD EVO/1090T? I had no idea PSU's were this involved.

Taken from this article

The power supply completes internal checks and tests before allowing the system to start. If the tests are successful, the power supply sends a special signal to the motherboard called Power_Good. This signal must be continuously present for the system to run. Therefore, when the AC voltage dips and the power supply can’t maintain outputs within regulation tolerance, the Power_Good signal is withdrawn (goes low) and forces the system to reset. The system does not restart until the Power_Good signal returns.

You ask why a manufacturer would make multiple rails. High wattage PSU for low cost is the answer. The internals don't need to be as robust to handle 40A as it does 80A , therefore the internals are cheaper.
The advantage is there is no need on my part to balance anything and if need be any section it supplies can pull as much amperage as it needs, for example my two 580s can pull over 200w each under load. That's 33A which is already close to your 40A rating. Bump that up to 240W each and you're at your limit. If the PSU has difficulty with that load it shuts down.

So the next point at 4.5 on the new CPU, considering they rate the 9370 at 220W with a base clock of 4.4G We'll assume that the CPU itself is is pulling about 20A on the one rail just by itself. If the PSU can't supply an instant, quality power signal when you flip the switch it'll reset, after it's had a chance to catch up then the PC will run.

With PSUs it's not all about the Watts or Amps but also the quality and consistency of the power as well. I've read some reviews on the raidmax and they do struggle there. The one I read on Johnny Guru couldn't maintain it 80 plus bronze reating under load either.
I'm not saying for certain the the PSU can't run your system but it can cause the issues you're having. Especially if this is happening from a cold boot before the PSU has a chance to "warm up"

 

[DatDirtyDawG]

Taken from this article

You ask why a manufacturer would make multiple rails. High wattage PSU for low cost is the answer. The internals don't need to be as robust to handle 40A as it does 80A , therefore the internals are cheaper.
The advantage is there is no need on my part to balance anything and if need be any section it supplies can pull as much amperage as it needs, for example my two 580s can pull over 200w each under load. That's 33A which is already close to your 40A rating. Bump that up to 240W each and you're at your limit. If the PSU has difficulty with that load it shuts down.

So the next point at 4.5 on the new CPU, considering they rate the 9370 at 220W with a base clock of 4.4G We'll assume that the CPU itself is is pulling about 20A on the one rail just by itself. If the PSU can't supply an instant, quality power signal when you flip the switch it'll reset, after it's had a chance to catch up then the PC will run.

With PSUs it's not all about the Watts or Amps but also the quality and consistency of the power as well. I've read some reviews on the raidmax and they do struggle there. The one I read on Johnny Guru couldn't maintain it 80 plus bronze reating under load either.
I'm not saying for certain the the PSU can't run your system but it can cause the issues you're having. Especially if this is happening from a cold boot before the PSU has a chance to "warm up"

That's interesting because the other PSU that I tried (that yielded the same result) is an old OCX 850W but this one also uses multiple 12v rails (2 @16a each and 1 @30a) so I'm guessing this one's even worse lol.

Yes, this happens most often from a cold boot. Not always but if I were to guess I'd say more often than not the cold boot is a 50/50 on whether the system will hang at the post screen, on a standard reboot after it's been running awhile its about 70/30 (still happens, just not as much). Almost always it fires right up on the second try. One more bit of info that I didn't think to mention, and I only thought of it after reading:

"If the PSU can't supply an instant, quality power signal when you flip the switch it'll reset, after it's had a chance to catch up then the PC will run."

On two occasions I decided to leave the system in it's seemingly frozen post screen state just to see what happens....both times the system ultimately did boot up it just took a good solid 5 or so minutes.....

So wow, what an education on PSUs. Until now I've only ever looked for wattage, ATX compatiblity, SLI/Crossfire and 80Plus ratings. I'm just not in a position to replace my PSU at the moment (not to mention I still have to confirm that it's the culprit)

Still, I can't say this enough. Other than this sh*tty boot issue (more annoying than anything else) My machine, once up and running runs beautifully with this PSU. Many many GPU stress tests when I had my two 560TI 2GB cards, hours and hours of BF3, BF4, BL2, Hours and hours of video encoding, P95 during the OC process, you name it as the list goes on and all the while no BOSD ever during any of these processes nor random game crashes. Hell even had one faulty stick of ram (oddly so as it showed up fine in bios and in windows, only found the stick to be bad via +memtest troubleshooting) and yet still no system instability or anything bad stick and all (again, once the system was up and running)

Either way I will figure this out (i hope) but if it turns out to be my PSU I guess I'll just have to live with it for awhile. If it's the board I have the option to have it replaced under warranty. Thanks again to everyone (RGone, Johan) for your valuable input, suggestions and education. I can tell you I see PSU's in a completely different light and will certainly make a more educated choice next time

I will post my results after unplugging everyting/bios reflash/battery removal-replacement as suggested

 

[RGone]

http://www.jonnyguru.com/forums/showthread.php?t=3990
Reload this Page Single vs. Multiple +12V rails: The splitting of the +12V rail

Generally we don't get caught by iffy power supplies; at least normally with newer brand name units.

That link above is actually a forum post by johnnyGURU. No you do not have to like him or his forum but he does very good work overall. He has post #1 because he started the thread. Look at POST #2. The very first person posting after JGuru has a power supply not well designed according to everything just written by JGuru. So there are possible issues unless one really does the homework thing even with a newer powersupply. It is up to the individual to do the legwork required to get a good power supply. The high watts rated ones are no longer cheap if they ever were. Spend your own money wisely. Do the looking for specs. Heck check to see how what you are looking for was rated after a test over at JGuru's place. He puts enough results in a thorough breakdown sheet for taking a long hard look at what is what. Know what +12V rails power what and most especially so if you have dual or triple video cards in mind.

From a stictly engineering stand-point we can cuss and discuss this to high hale. It is not necessary at all. JGuru laid out a few simple guidelines and to follow them is to mostly ensure you, yourself, do not fall into some power supply trap.
RGone...ster.

 

[Johan45]

Just another thought, I don't know which BIOSversion you're using but that could also have a bit of a glitch. The re-flash should straighten that out, what I was going to say is I have tried quite a few Versions and my preference is the ver.803, not saying that you should use it I'm just saying that for my pieces parts it works very well.

 

[RGone]

DDDG, you say >> is an old OCX 850W >> did you mean OCZ 850 watt and which series? There is more than one model even of the OCZ 850 watt power supply based on when bought.
RGone...

 

[DatDirtyDawG]

DDDG, you say >> is an old OCX 850W >> did you mean OCZ 850 watt and which series? There is more than one model even of the OCZ 850 watt power supply based on when bought.
RGone...

Yeah that was a typo. It's an old Turbo-Cool 850 SSI It was a system I had when I was running with an Athlon 64 X2 5200 Brisbane 2.7GHz and a pair of GT9800 in SLI

 

[RGone]

Took that actual power supply model >> Turbo-Cool 850 SSI >> went straight to specs and they are shown in image below.
RGone...

 

[Johan45]

Took that actual power supply model >> Turbo-Cool 850 SSI >> went straight to specs and they are shown in image below.
RGone...

View attachment 135325

Yes that could definately be an issue. Nice detective work. SherGonster Holmes

 

[DatDirtyDawG]

Sabertooths don't normally do that shett. ...

Ok finally, this absurd mystery has been solved. After re-flashing, removing and reinserting the battery, unplug/re plug the PSU this dumb problem persisted.

I'm quoting this particular statement because it put me in a mode to address all things connected (fans, everything) and whaddya know....lol

I disconnected one of my two DVD ROM (which appeared to work fine...so I thought) and voila, no more hangs. Like ZERO, never again. I've rebooted umpteen times and no hangs since removing this one particular DVD ROM. Hook it back up and once again we have hangs.

Anyway BIG thanks to Rgone, Johan, Blaylock and everyone else who took a crack at this. Who knew that despite looking like it worked fine the DVD rom was conflicting with my board and my Sabertooth simply wasn't having it..