Jump to content
Corsair Community

TX950W and Asus Sabertooth X58: Shut-down Problem


MaxiMax

Recommended Posts

My computer started shutting down randomly, as if power suddenly disappears or somebody unplugs the power cord.

When the system is on my PSU makes a squealing noise, which sometimes turns into a noticeable beep/beep sound.

 

I have tried the following things:

1. resetting CMOS on my motherboard

2. updating bios

3. resetting bios to the default settings

4. reinstalling motherboard drivers

5. reinstalling windows

5. disconnecting PCI express connectors and removing my graphics card

6. unplugging every cable that connects PSU to my motherboard and connecting everything back again.

 

None of the above helped me resolve my problem

 

Other thoughts:

+12V output constantly changes from 12.19 to 12.35. I do not know whether it is normal.

 

My mobo works well with another PSU. No shut-downs detected. It seems that these shut-downs are caused by the defective PSU rather than by my motherboard. Corsair PSU is obviously guilty.

 

Also, I found a couple of ASUS forums where the customers discuss incompatibility between Asus motherboards and some Corsair PSU models. As far as I know Asus and Corsair fixed this problem some time ago. However, it is disturbing that the symptoms discussed in the forums are exactly the same: a computer randomly shuts down and reboots. In any case, these sporadic shut-downs are quite dangerous for my motherboard as well as for other parts.

 

Any advice on how to fix this issue will be appreciated!

Link to comment
Share on other sites

  • Replies 72
  • Created
  • Last Reply
+12V output constantly changes from 12.19 to 12.35. I do not know whether it is normal.

Yeah , this is normal. You are allowed 5% varience in voltage. I would rather have that 5% in the plus side rather than the down . You are well within spec as far as the 12v goes. What about 5v, 3.3v?

 

 

Are you overclocking anything? Memory , cpu?

 

What are your RAM timings and voltages set at? Incorrect values can cause this too.

Link to comment
Share on other sites

Thank you for responding so quickly!

 

No overclocking.

 

Here are my bios settings:

Ai overclock tuner <auto>

DRAM frequency <ddr3 - 1600>.

I could set it to auto or to 1333 but that does not resolve my problem

UCLK Frequency <3200MHz> <must be at least twice as DRAM frequency>

 

DRAM 1st information 9-9-9-24-6-88-12-8-24-0

every value here is grayed out and set to auto

CAS# Latency: 9 [auto]

RAS# to CAS# Delay: 9

RAS# PRE Time: 9

RAS# ACT Time: 24

RAS# to RAS# delay: 6

REF Cycle Time: 88

Write Recovery Time: 12

Read to Pre Time: 8

FOUR act win time: 24

Back to Back CAS# Delay: 0

 

Timing mode: 1N

Round trip latency on CHA: 60

Round trip latency on CHB: 62

Round trip latency on CHC: 63

Write to read delay DD: 6 // DR: 6 // SR: 18

Read to write delay DD: 10 // DR: 10// SR: 10

Read to read delay DD: 7 // DR : 6 // SR : 4

Write to write delay DD : 7 // DR : 7 // SR :4

 

note: My DDR is corsair XMS 1600

 

Temperature and power settings

CPU 36.5C/97.5F || MB 32C/89.5F || power fan speed 740rpm

 

NB Fan speed n/a -- I do not have Northbridge fan installed.

 

CPU voltage: 1.200V

3.3V: 3.232V Stable

5v Voltage : varies from 5.022 to 5.046

12v: varies from 12.19 to 12.35

 

I am not sure what suspend mode in power options means. It is set to auto.

ACPI 2.0 support -- enabled. It is disabled by default. Both options have no effect on my issue.

ACPI APIC support enabled.

 

My MB had some problems waking up from the sleep mode. I set SATA to AHCI instead of IDE and that fixed the problem.

 

But I am still looking for a way to cure these random shut-downs.

 

Thanks again!

Link to comment
Share on other sites

It seems that incompatibility between TX950 and Asus MBs is a known issue.

 

I found a link to this thread, where RAM GUY says that "there was some problems reported between some ASUS MB's and our 950TX PSU and until the issue was isolated and a solution was found we offered anyone who had a problem to switch to an 850HX or 1000HX as a convenience. ASUS and Corsair worked to find and solve the issue."

 

The underlying message here is that the issue is already isolated. However, it is not the case.

 

Perhaps I got one of the "early" TX models from newegg. But is it possible to know whether I am buying the old or the new one? Both Asus and Corsair are the well-known manufacturers of MBs and PSUs and the fact that this problem still exists is upsetting.

Link to comment
Share on other sites

Ok, now I think that my problem is not caused by incorrect RAM timings and voltages.

I have it set at 9-9-9-24 -- and these are the correct timing values for Corsair XMS 1600 DDR3 memory. So it seems that the cause of the shut-down problem is nothing but PSU (in combination with Asus Mobo, of course).

 

It's time for the last test. I have already mentioned that my system runs without shutdowns with a different PSU. I wounder what is going to happen if I test my PSU in my old pc with a different mobo. If everything runs well then the problem is a matter of pure Asus/Corsair incompatibility.

 

Update: Another thread where the user MateoTTR describes a similar problem of "random restart" caused by mating of Asus Rampage III Extreme with HX1000 Corsair PSU. The models are different but the problem is the same.

 

I do not know what exactly these guys from A and C were fixing on both ends.

 

MateoTTR (like many others) asks for a full refund. I understand it. The danger here is to get a piece of crap replaced with a piece of scrap.

 

Update 2

Here is the same problem with Asus P5LD2SE and Corsair HX850!

 

And... bible hit me one more time. This time Asus P5Q deluxe and TX950. I could be criticized by RAM GUY (where is he?) that these cases are old. The last one is from Sep 2010. But I bought my PSU in November 2010. Ram Guy never mentioned when (in what month/year) the tech people from Asus and Corsair irreversibly fixed the issue.

 

My guess is that... they did not fix it.

 

In my case, the problem is between Asus Sabertooth x58 and Corsair TX950

Choose any PSU from Corsair and any Mobo from Asus and they are likely not to work.

 

Any advice on how to fix random restarts is appreciated!

Link to comment
Share on other sites

1. MateoTTR asked for a refund because he purchased a 2nd PSU instead of RMAing it.

 

2. Can random restarts be caused by a PSU? Sure. Does that mean that if you do have that issue that it's the PSU and/or every PSU ever made has problems? Of course not.

 

3. The Asus / TX950W issue was fixed around Spring 2010 IIRC.

 

4. When was the motherboard purchased and when was it manufactured?

Link to comment
Share on other sites

Yes random restart issues are quite annoying.

Very difficult to pinpoint the origin of the issue. If you search in different forums, you realize that either PSU is faulty or MB has an issue (in the latter case, shortcuts or a leaking capacitor can cause random reboots).

The only way around is to try a different PSU and/or to test the PSU on a different system.

This is why a purchased a different PSU model (realizing the core manufacturer for all AX models, Seasonic, had a great reputation, I went for the AX850). So I could finally make sure that my HX1000 was defective.

On a side note, I found a couple of thread (on this forum and others) where people would experience the same problem with the HX1000.

So maybe not such a common issue, but at least I am not the only one.

 

Also getting a refund or a voucher is not possible. So I decided to issue an RMA and I will sell second hand the replacement unit to pay for (part of) my AX850 purchase.

 

Best of luck with your PSU issues.

Link to comment
Share on other sites

Technical question: what is the best way to find my RAM voltages? Are they in Bios?

Should I look for it in my bios' AI (voltage) tweaker?

 

Again, I did everything I could to fix the problem. I formatted my HDD (completely) and reinstalled windows - so it should not be the software issue. No viruses detected.

 

My system worked well for 7 months. I have 5 year warranty for MB and PSU. I have already tried a different PSU on my system. No random reboots were detected. When I have some more time I will definitely try Corsair PSU on my old system, which, of course includes different everything (MB, memory, HDD) e.t.c. What should be my conclusion if it turns out that my new Corsair PSU works well with the old system and my old PSU works well with the new system?

 

The motherboard was purchased together with PSU and all other parts in November 2010 (from Newegg). Wired's question about MB's production date seems irrelevant because Sabertooth series were made available to the american public around august 2010. Please correct me if I am wrong about the release date.

According to Wired (and Ram Guy), by that time the compatibility problem between Asus MB and Corsair PSU should have been fixed. Let's assume that MB was indeed released after spring 2010, and people still have PSU shutdown problems that are so typical of older PSUs and Asus MB's. What conclusion is here to make?

 

 

 

Concerning the overall compatibility issue.

 

I am extremely unhappy that

(1) my system worked only for 7 months

(2) so many people have a similar problem with various psu models (including HX1000) and that

(3) the global presence of this problem is not being acknowledged by Corsair techs.

 

As for the latter, it would be merely ignorant to claim that people like me have their "individual" equipment problems that have to be treated exclusively on one-to-one level. If i did not hear other people having a similar problem I would worry less. Now there's no guarantee that by choosing Corsair and Asus (in combination) I am likely to find myself in the same deadlock. This predicament should be potentially undesirable for both manufacturers.

 

The desktop that I assembled 5 years ago works well without any random reboots or any other equipment failures. No overclocking. Everything is set to default. Why should I be interested in setting the voltages manually and thinking about what's wrong with the default settings of my Asus MB. Is it normal? No.

 

Is it the fault of Asus MB? Then why does it work with another PSU?

 

Now, a response to Wired: I never claimed that EVERY PSU has this problem. But how many faulty PSU would be enough in order to reconsider and the whole line of production? What about the history of Japanese tech industry - one unit is faulty the whole line goes down? There is no democracy in technology. I should not conduct an election and ask my PSU to do whatever pleases it.

I understand that I find myself in a waiting line of unsatisfied customers. But do you need to have too many of them to create some resonance? And what are you going to do to spare them a hassle?

 

If you ever build a system you want it to last, which would not be possible if either MB or PSU sing out of tune.

Link to comment
Share on other sites

Thank you for responding to my thread!

 

When did you purchase your PSU? Was it later than spring 2010? It looks like your system is new and powerful. That is also a cause of my disappointment - to invest so much money in a custom-made system with outstanding specs in order to fall victim to random reboots of an expensive PSU unit.

 

Also, it would be interesting to know when you RMA's your unit.

Link to comment
Share on other sites

I purchased my PSU in FEB 09, It started to shut down when I was running Quad SLI GTX295s, but I thought it was a one off.

 

In 2010 I upgraded and got 2 GTX480s, thats when all the random shutdowns started to be a problem.

 

I think that the HX1000 has not enough power to run 2 GTX480s as its 2 500w PSUs together, if it was 1 Rail it would be ok.

 

Either that or my HX1000 has been faulty from day 1.

Link to comment
Share on other sites

It's amazing that these shutdowns affect so many different units at once.

I have TX950 with one rail powering a single gtx460. That much power should be more than enough. I decided to get 950w thinking about future upgrades, but 750w would be enough for the system.

 

Is it normal to have these random shutdowns when there is not enough power? I guess the ideal protection mechanism here would be for the system not to power up at all.

Link to comment
Share on other sites

Everything is set to default. Why should I be interested in setting the voltages manually and thinking about what's wrong with the default settings of my Asus MB. Is it normal? No.

You are overclocking your memory. So yes you would have to set them up manually and adjust the bios voltages to reflect the OC on the memory. 1333mmhz is the max supported memory speed for your MB/CPU so anything faster than that will not be detected correctly with an "auto " setting.

 

Technical question: what is the best way to find my RAM voltages? Are they in Bios?

Yes they are in the BIOS

 

 

If you ever build a system you want it to last, which would not be possible if either MB or PSU sing out of tune.

Yes you do! But electronics by nature are not going to be 1000% trouble free. There is no way to judge how long any individual component may last. No matter the vendor or manufacturer they will have units fail. Is it disheartening? Sure! But thats the way it is. And there is no way to predict when or how a unit is going to fail.

 

If the unit works in another system then i would contact ASUS. If a Differnt PSU works in your system then RMA the PSU.

Link to comment
Share on other sites

(2) so many people have a similar problem with various psu models (including HX1000) and that

(3) the global presence of this problem is not being acknowledged by Corsair techs.

 

But how many faulty PSU would be enough in order to reconsider and the whole line of production?

Please stop trying to make this into a big issue that spans various PSU models. It isn't. Their various PSUs are made by different companies and as such it's impossible for them all to have a globally related issue like that.

 

 

As for the latter, it would be merely ignorant to claim that people like me have their "individual" equipment problems that have to be treated exclusively on one-to-one level.
Different people / different parts / different problems = troubleshooting one at a time. This is a support forum and as such you're largely limited to only seeing issues.

 

 

I understand that I find myself in a waiting line of unsatisfied customers. But do you need to have too many of them create some resonance? And what are you going to do in order to spare them a hassle?

You're looking at something like < .1% of the actual number of people that own their products.

 

 

Why should I be interested in setting the voltages manually and thinking about what's wrong with the default settings of my Asus MB. Is it normal? No.
Not every motherboard reports or sets voltages correctly. Could be a fault or just a sensor reading the wrong thing. Either way it's a troubleshooting step. You can certaingly skip it if you'd like.

 

 

Is it normal to have these random shutdowns when there is not enough power? I guess the ideal protection mechanism here would be for the system not to power up at all.
Depends on what's causing the issue. If the draw is low on boot, but it pulls more during a game, or a faulty card is pulling more voltage randomly in way too high amounts or just has a fault, then the PSU could be protecting itself, or the PSU could be at fault as well. Part swapping would be the easiest way to troubleshoot.

 

 

Guys, random shutdowns is one way a PSU can exhibit issues (that may or may not stem from itself). It doesn't mean they all have the same problem, or are even linked in any way.

Link to comment
Share on other sites

When the problem appeared I reset CMOS and returned bios to the default settings. Now my memory is set at 1333, which is a default. Regardless of what I choose (auto-1333 or manual-1600), I have the same shutdowns. My system worked with default settings (with memory set at auto-1333) for 7 months without any shutdowns.

 

When I set mem at 1600 and UCLK is 3200 should I necessarily tweak voltages or I could leave it at auto?

 

I do not understand why 1333mmhz is the max supported speed for my MB/CPU. My MB supports DDR3 1866/1600/1333/1066. The retailer says that my CPU supports 1600 and my memory is Corsair XMS 1600 triple kit. The timings are correct by default. I will post my default voltage a bit later.

 

The system shuts down without gaming. When I'm gaming it makes a louder buzzing noise. When it boots up it makes a squealing noise that does not go away. I have similar shutdowns when I leave my system in Bios.

 

I never argued for 1000% trouble free systems. Let's not reduce the shut-down problems to individualism. They are common and Corsair was aware of that. C claims that this problem is fixed and I am a not-so-rare example of the fact that it is not.

 

But, speaking of % let's do a bit of statistics. When I purchased my unit (from newegg) there were several negative reviews. All other reviews were positive. By now 125 customers left their reviews. There are 26 unsatisfied feedbacks about RMAd units. Many of them mention that their PSUs have a problem with Asus motherboards. Additionally, there are 9 people who gave the product average (three-star) reviews. All of them either RMAd the unit or replaced it with a different PSU. 4 satisfied customers RMAd the unit. that makes 39 customers. I am the 40th one. 32% of unsatisfied customers and defective units. Grossly calculated, every third unit is either defective or with some potential problem. In my calculation I try to exclude people who are dissatisfied with lots of cables of TX950. But if we take them into account and make a discount - then every 4th unit is defective. Even if it is every 5th unit. Does it sound like 1000% trouble free electronics?

 

Now, I had to wait for 7 months in order for the problem to appear. So it is unknown what is going to happen with the customers who purchased the unit in 2011 (or after spring 2010) and left positive reviews.

Link to comment
Share on other sites

When I set mem at 1600 and UCLK is 3200 should I necessarily tweak voltages or I could leave it at auto?

No you would have to adjust voltages manually to compensate for the overclock.

 

I never argued for 1000% trouble free systems. Let's not reduce the shut-down problem to individualism. They are common and Corsair was aware of that. C claims that they are fixed and I am a not-so-rare example of the fact that they are not.

Visit ANY other vendors site and you will find the same issue in their support forums as well. EVERY PSU manufacturer is going to have them.

 

But, speaking of % let's do a bit of statistics. When I purchased my unit (from newegg) there were several negative reviews. All other reviews were positive. By now 125 customers left their reviews. There are 26 unsatisfied feedbacks about RMAd units. Many of them mention that their PSUs have a problem with Asus motherboards. Additionally, there are 9 people who gave the product average (three-star) reviews. All of them either RMAd the unit or replaced it with a different PSU. 4 satisfied customers RMAd the unit. that makes 39 customers. I am the 40th one. 32% of unsatisfied customers and defective units. Grossly calculated, every third unit is either defective or with some potential problem. In my calculation I try to exclude people who are dissatisfied with lots of cables of TX950. But if we take them into account and make a discount - then every 4th unit is defective. Even if it is every 5th unit. Does it sound like 1000% trouble free electronics?

 

 

Out of how many thousands sold?

 

Now, I had to wait for 7 months in order for the problem to appear. So it is unknown what is going to happen with the customers who purchased the unit in 2011 (or after spring 2010) and left positive reviews
.

The same thing as everyone else who may potentially have an issue...They can have their unit replaced the same as everyone else if they need to.

Besides...You really trust the eggs reviews? They let people do reviews that have not even bought the product. How can you use figures that may very well be bias and bogus?

Link to comment
Share on other sites

And how can I trust the manufacturers, representatives administrators and genuine supporters? Aren't they biased? No, it's better to rely on eggs, at least there is some potentiality for the polyphony of the people's voices.

 

Besides, I am not talking about any PSU and any manufacturer. I refer to the concrete problem that makes me think about voltages and makes me wonder about the technical causes of the faulty equipment that I would not even think about otherwise. When something is broken you ask for a reason.

 

Asus MBs and Corsair TX950 seem not to work together - you see a pattern. I am not saying that it is a LAW or a RULE, but it is definitely a pattern. Now you might choose to ignore it and to RMA each faulty unit but it is not going to change a thing in a well established pattern of shutting down equipment. That is why the whole conversation is still here.

Link to comment
Share on other sites

Ok, I am reviewing my BIOS settings. Here are the things that seem strange:

 

Ai Tweaker Section

Ai Overclock Tuner: Auto (Should I set it to manual? Other options are D.O.C.P. and X.M.P)

Intel SpeedStep Tech: enabled. Is it a good idea to disable it?

Intel TurboMode: enabled. Should I disable it?

QPI link data rate: auto. (Other options: slow mode, 4800MT/s; 5866MT/s and 6400MT/s)

Asus/3rd Party UI Priority: set to "Asus Utility" another option is "3rd party utility". Not sure what it means.

 

Cpu Voltage control: manual <set to fixed voltage>. 2nd option is offset.

Cpu Voltage: auto

QPI/DRAM voltage: auto. And these are the options that I cannot change. Maybe I should set some value to manual before I am allowed to change these.

IOH voltage: auto

IOH PCIE Voltage: auto

and everything else is auto.

 

Now if I have an option to change QPI/DRAM voltage, what values should I enter? I already know that my RAM timing is correct (9-9-9-24). What voltage for my RAM (Corsair XMS 1600) would be normal?

Link to comment
Share on other sites

Another question. 1333 is MB default for the memory. But the memory is 1600 and MB supports it. If I change auto-1333 to 1600 and UCLK to 3200 would it be considered overclocking? Why?

 

I was sure that 1600 memory should be set to 1600 in order to work in its full capacity. I could have purchased cheaper XMS 1333. Now I do not quite understand why setting DDR to 1600 requires voltage compensation for overclocking. If QPI/DRAM voltage is auto, my mem voltage should be adjusted automatically, right? Thus, the problem here would be MB automatically choosing incorrect voltage for my DDR.

 

If I set my DDR to 1600 my timing values should not be below 9-9-9-24. Otherwise I will corrupt the memory modules.

But what voltage values should I enter for 1600 configuration?

Link to comment
Share on other sites

And how can I trust the manufacturers, representatives administrators and genuine supporters? Aren't they biased? No, it's better to rely on eggs, at least there is some potentiality for the polyphony of the people's voices.
I wouldn't put my full faith in NewEgg reviews for the simple fact that you don't need to own a product in order to review it. That in of itself makes their system wide open to invalid data. On the other hand, you're right in that there could certainly be biased information here. I personally believe that Corsair's guys are honest and on the up and up having known them for around a decade, but that's only my opinion. I'd recommend visiting various 3rd party forums. Ones that don't have official support forums within them. Nowadays you may not be able to find a major forum that doesn't have company reps visit them however, but that's a bit different from one having a permanent support area within it.

 

 

Asus MBs and Corsair TX950 seem not to work together - you see a pattern. I am not saying that it is a LAW or a RULE, but it is definitely a pattern.
That was an issue that Asus fixed on their end over a year ago and Corsair also put fixes in place to be safe. Look at the numbers of Asus and TX950W issues posted on the forum before the fix and after. I'd guess that you'd see the numbers drop dramatically. OTOH, it's not a perfect data set by any means.

 

You can "see a pattern" when checking out ANY product on this forum. Corsair's a major player, and so is Asus, as is a whole bunch of others. That means you'll see them in more systems. Therefore patterns may seem to appear just because they sell a lot. In reality the actual percentage of issues could be exceedingly low, and even more so when you only look at a certain motherboard manufacturer. It's called Apophenia.

 

Let's say 10% of all Corsair products have issues, and anything under 20% is considered an excellent percentage. Obviously made up percentages. So let's say 100 items are made. Therefore 10 items would have issues. 10's not much. Now let's say 1,000,000 are made. Now we're up to 100,000 items. Still fits within the "excellent" range, but because the volume of sales are not public, people may assume it's 100K out of 400K (25%). The point is, without knowing all the facts it's easy to assume anything. Could you be right? Sure, it's possible. Do I think there's enough evidence to support your theory? No.

 

 

As for 1333 MHz, that's what the CPU supports. 1600 MHz is technically overclocking, but it's exceedingly rare to see an X58 CPU that doesn't do it. Set Ai Overclock Tuner to XMP and you'll be good to go. That will set all other relevant settings for you. Should be 9-9-9-24 @ 1.65v (the rated specs basically).

Link to comment
Share on other sites

Thank you four your assistance, Wired. Should it be 1.65v for 1333 and for 1600 as well?

I set AI Overclock to XMP and now it shows grayed out parameters for DDR. 1600 then timing and then 1.65V-1.20V. Is this variance allowable? Another thing that is interesting - default (auto) setting changes my DDR to 1333 but when I choose X.M.P. DDR is set to 1600. Is it normal?

 

We will see whether this change influences the shutdowns in any way.

 

As for your point about my theory: I will never have enough evidence because I will never figure out the exact number of faulty units. Yes, all calculations are relative.

 

I had several devices that failed but I did not write negative reviews for all of them. So I imagine that there are people like this, just like other people who give five eggs for Corsair PSUs that were RMAd :)

 

I still do not believe that Asus/Corsair fixed this issue. My MB, if I am right, was released after the fix. And I still have a very similar symptom, if not the same one. When I go to newegg or to Corsair forums I see the customers who purchased their PSUs after the fix. Newegg's verified owners describe my problem in their own words. The manufacturer's default response is always the same ... do not worry! This issue has already been fixed. When the same phrase is impassively used to reassure 2011 buyers -- I do not believe it.

Link to comment
Share on other sites

without any doubt. The cause is what I am trying to figure out. And if the problem is not caused by PSU, I will be prompted to regard the problematic combination of Asus/Corsair as tragic magic, as a simple matter of coincidence. :)

Meanwhile I am testing X.M.P. and letting the germans kill me in World at War. :mad:

If that does not work, the only test that remains is to insert PSU into my old system.

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.


×
×
  • Create New...