'PCH Voltage: The only thing that works' A testing report

Hey Corsair Forums,

 

This is purely a record of my somewhat exhaustive testing regarding the Force GT 120GB (and SF-2281 controller driven SSDs in general) in compatibility with P67 series chipsets.

 

If you are having problems with P67 chipset, please attempt raising PCH voltage in your BIOS. Throughout all of my testing this is the only method that stops the disappearing drive symptom.

 

My testing began with known working system, optimized defaults P67 motherboard. I first tested using Firmware 1.2 with Intel RST OROM 10.5 from ASUS website. Result was very frequent freezes almost every 10 minutes. The drive exhibits this behavior with Firmware 1.2 until LPM registry fix is applied. This is separate from infamous "sandforce BSOD" problem and this issue may have been resolved with Firmware update 1.3.

 

The next stage of my testing involved Firmware 1.3, with the same configuration and Intel RST OROM 10.5 with LPM registry fix applied. The result was blue screens approximately every 16-24 hours. I then attempted loading Intel RST OROM 10.6, and the result was the same with BSOD every 16-24 hours and the drive disappears from BIOS. Firmware 1.3 tried with MSAHCI driver and the result is the same. Firmware 1.3 does not resolve the infamous Sandforce BSOD, at least not in any my testing.

 

Final stage of testing involved BIOS setting adjustment recommended in this forum tried one at a time for log purposes. Using Firmware 1.3, adjusted BIOS setting to Hotplug enabled on Port 0. No effect on issue, BSOD persists every 16-24 hours. Next I adjust PCH chipset voltage to 1.150V. The result is that BSOD no longer occurs on that interval infact not again for 17 straight days. PCH voltage is THE ONLY ADJUSTMENT that works to affect stability on the P67 chipset. Whether or not this is an Intel or Sandforce problem is unknown, but there is clearly a lack of compatibility between the devices at optimized default setting. Sandforce and Intel have claimed they are working in cooperation to solve the problem. UNTIL THEN IF YOU HAVE A P67 CHIPSET THEN RAISING PCH VOLTAGE IS YOUR ONLY SOLUTION (ALONG WITH LPM REG FIX IF YOU USE INTEL RST)

 

There are still some quirks and such possibly related to compatibility as within those 17 days explorer.exe did crash once while copying a file. I am unsure what to attribute this to other than the aforementioned issues between P67 and Sandforce.

 

I hope my research has been helpful. Please I would appreciate comments on your thoughts and the results of your individual testing. Did adjusting PCH voltage work for you?

Good to know. Thanks. For me, disabling TRIM is the only thing that stabilized the drive. Would like to hear other feedback first before trying this + reenabling TRIM.

Good to know. Thanks. For me, disabling TRIM is the only thing that stabilized the drive. Would like to hear other feedback first before trying this + reenabling TRIM.

 

Try PCH voltage if you use a similar configuration.. trust me. Disabling TRIM sucks you don't want to do that man

This was suggested on the force 3 threads. it worked for some and didn't work for others.

This was suggested on the force 3 threads. it worked for some and didn't work for others.

 

For those with P67/Z68 chipset I strongly suggest trying it. I was actually able to reproduce the issue on a second P67 board today.

Oddly enough, I have a P67 Sabertooth system here that has been flawless with every drive I put into it, single Force 3 120s, RAID-0 Force 3 120s, GTs, etc.

Oddly enough, I have a P67 Sabertooth system here that has been flawless with every drive I put into it, single Force 3 120s, RAID-0 Force 3 120s, GTs, etc.

 

Hmm interesting.. I read that Anand reproduced the problem on a Z68 board and then he replaced the power supply and the problem disappeared. My PSU is an HX850 and it runs great no issues, but maybe some small fluctuation in power is causing these problems?

If I have some more time this week I will test it with a different power supply and see..maybe it is some power fluctuation causes all of this with many users. In any case it seems SF-2281 is extremely sensitive to this

Hmm interesting.. I read that Anand reproduced the problem on a Z68 board and then he replaced the power supply and the problem disappeared. My PSU is an HX850 and it runs great no issues, but maybe some small fluctuation in power is causing these problems?

 

I agree with you it should be either the controller is not getting enough power for operation or the Sata signal dies off when the SSD is on high operation for example booting or installing things on the SSD. But what yellowbeard said interests me, he must have the best motherboard ever!. I'm also wondering if he overclocked it as it can sometimes help with BSODS and crashes.

I agree with you it should be either the controller is not getting enough power for operation or the Sata signal dies off when the SSD is on high operation for example booting or installing things on the SSD. But what yellowbeard said interests me, he must have the best motherboard ever!. I'm also wondering if he overclocked it as it can sometimes help with BSODS and crashes.

 

I've run every SSD I have here at both stock and OCed on a Sabertooth and a Z68 MIVE-Z. GTs single, GTs in RAID-0, Force 3 single, Force 3 RAID-0, etc. Both systems have AX series PSUs.

I've run every SSD I have here at both stock and OCed on a Sabertooth and a Z68 MIVE-Z. GTs single, GTs in RAID-0, Force 3 single, Force 3 RAID-0, etc. Both systems have AX series PSUs.

 

I'm not sure but the PSU might also cause the issue. If possible, i would recommend you grab a few old or new PSUS and give them a go. If they do crash then it might be possible that the PSU is also a contributing factor to the SSD crashing.

 

Example Template

PSU| No BIOS Modifications | Short Cable : CRASH/NO CRASH

PSU| No BIOS Modifications | Long Cable : CRASH/NO CRASH

PSU| Overclock (Frequency,Core Voltage,Etc) | Short Cable : CRASH/NO CRASH

PSU| Overclock (Frequency,Core Voltage,Etc) | Long Cable : CRASH/NO CRASH

PSU| Overclock (DRAM Frequency, etc) | Long Cable : CRASH/NO CRASH

PSU| Overclock (DRAM Frequency, etc) | Short Cable : CRASH/NO CRASH

PSU| Stock, PCH | Short Cable :CRASH/NO CRASH

PSU| Stock, PCH | Long Cable :CRASH/NO CRASH

And so on...

 

If you understand it, you would know you would need to be extremely patient as this is a long time consuming template. But if you want to reproduce those issues, i would follow this type of template. but guess what.... thats only for the PSU, If you are also going to test the motherboard.... :eek:

I'll give the cable test a run through.. I actually have a 10cm cable lol shouldn't be any issue with length there..

 

I actually just the other day applied an overclock and the result is no different.. same as stock if I take the PCH voltage down I get the BSOD in about one day.

 

If there is no difference using a 10cm cable from an 18inch cable then I'm starting to believe it is power related

 

The PSU requires some more time to test and I don't have the time right now lol have to work on other stuff..

 

This drive must be a nightmare headache for the end user lol

May I ask the logic behind cables/cable length relation when BSODs occur almost in all cases when idle, rather than load?

Sorry but this is the blind leading the blind. Complete hogwash.

Sorry but this is the blind leading the blind. Complete hogwash.

 

No, it isn't. Our internal testing how shown on some boards that you do in fact get some stabily improvements in "some" cases from adjusting this voltage. We've even shown some performance increases with it.

 

Please don't try to deter others from useful troubleshooting when you are uninformed.

the issues are with the drives not people setups. I have tried every chipset and these problems happen on all of then. And I am using a Brand new AX1200 psu with brand new cables even RMA'd and got a replacement direct from Corsair and guess what its even worse than the first one.

I hope my research has been helpful. Please I would appreciate comments on your thoughts and the results of your individual testing. Did adjusting PCH voltage work for you?

Would an accessory SATA controller card resolve the issue in an affected system?

The problem is the SSD its self, as so many people with different setups are all having the same problem and the RMA new one was brand new with the 1.3 firmware and that become unbootable after 2 weeks it was even worse. if go back to the c300 or the inferno ssds i have no problem or mechanical all my problems do not happen so its the drive if it wasn't I would have the same problems with the other drives.

No, it isn't. Our internal testing how shown on some boards that you do in fact get some stabily improvements in "some" cases from adjusting this voltage. We've even shown some performance increases with it.

 

Please don't try to deter others from useful troubleshooting when you are uninformed.

 

How are things going with a firmware upgrade yellowbeard?

I'm asking cause that you are correct that these solutions fix some people's problems (or at least improve their experiences), BUT there are many people who still expect a permanent solution.

How are things going with a firmware upgrade yellowbeard?

I'm asking cause that you are correct that these solutions fix some people's problems (or at least improve their experiences), BUT there are many people who still expect a permanent solution.

 

I agree...the ones still having problems, even after jumping through all the troubleshooting hoops, are anxiously awaiting a PERMANENT fix. Its really a big letdown when your system runs fine 24/7 for almost a month and out of nowhere the SSD just decides to drop offline, and wont show up in BIOS until you power down completely, then start back up.

 

Its been 3 months since the original "recall" of these Force Series 3 drives, and only 1 FW update that really doesnt solve anything. Im really starting to doubt that we will see a new FW at all. I can understand that SOME people are having luck with 1.3 fw, but I seriously doubt the issues are completely gone. Give it time, the problems will show up again, just like it did for me. Like I said above..my SSD was working flawlessly after 1.3 FW. I left my PC on 24/7 and had no issues until tonight. So I WAS one of the lucky ones, until now. Im finished jumping through the troubleshooting hoops, hoping my SSD doesnt crap out on me. We(consumers), shouldn't have to try all these things just to get a drive to work. It should work right out of the box. Plain and simple.

 

If Corsair/Sandforce are still scratching their heads about this, then maybe they should just do another recall so we can all get our money back. If the companies involved can't even figure out whats wrong with their own product, then just say so and stop wasting all of our time. Its not our job to be your R&D team. I've been following these forums since the recall.Even when my system was working, just to see how things are going. And I keep seeing the same crap over and over. And don't tell me that some people are happy and the FW solved all their problems. Saying that to me is like saying...well...I guess your just SOL! THE PROBLEM IS THE FREAKIN' DRIVE! ADMIT IT!! Go back to the drawing board and fix this thing! Its a really bad *** drive......when it works! UGH..I dont even know what to say anymore...everything is just repeated over and over. Im tired of being the patient customer, holding on to dear life on this drive in hopes of a permanent fix,or new firmware. I got better things to do with my life than come on here, hoping to see a new fix for something that OBVIOUSLY shouldn't have been released in the first place. :mad::mad:

Ι've tried all the solutions mentioned and I indeed managed to achieve a significant drop at the BSOD's, freezes frequency. Other users are luckier than me, others aren't.

 

Regardless of that it is really frustrating for me to still see my system hang or/and reboot for no reason needing a cold reset to recover.

 

The reason that I frequently remind the need of a permanent solution here is because I have the impression (through their employees' participation here) that corsair does not work on a permanent solution. I don't see anyone talking about a working firmware fix at progress but I constantly see people from corsair proposing the same workarounds (which don't work for everybody) hoping that users will be lucky enough to see an improvement.

 

I will be content for now with my current system state if someone assures me that the permanent solution is in progress (updating info about the stage of these works). This doesn't happen though and as a result I don't have a clue if there is anything in progress at all.

Well said Philharmonik. And as screwing around with settings just to get a drop in BSOD's when there should be none at all this is a situation that is clearly not good enough!

I've noticed am improvement as well with this fix. Although I only went up to 1.10 for the PCH. It was previously set to auto which was 1.05.

Yes it's true there is improvement with this. unfortunately, my "testing period" with newegg is over and I am returning this thing. It has been fun to experiment and see the speeds and try different things.. but this thing is far too unreliable. even after PCH fix I have had BSOD and unresponsiveness..

 

This is the OP signing out... good luck everyone and try PCH voltage if you want to keep the drive... I hope corsair resolve this

I've noticed am improvement as well with this fix. Although I only went up to 1.10 for the PCH. It was previously set to auto which was 1.05.

 

You might wan't to forgo that and go the full 1.15, I went to 1.10 initially and although I got a few days more of stability, BSOD still occured. So far after another increase, no issues *fingers crossed*