RMA 5579722 - Corsair AX1200 PSU

Hi Folks,

I have just submitted this RMA, request, However I suspect that more details may be needed and indeed I have a couple of queries, so here goes. This is a PSU related issue, however by necessity I also mention HDD's and RAM here. Sorry for the length of the post!

Overview:

It is believed that a PSU issue is (has) damaged multiple (3) HDD's and is a root cause for current system instability.

Spec:

<removed and added into profile!>

Background: The PC has recently become unstable (memory *originally* suspected) after 12 months of heavy use (am both a Gamer & Webdesigner). Instability is in the form of random lockups, unforseen reboots and specific instability in some games

Originally an RMA was/is arranged for the memory due to intermittent errors in memtest x86 (RMA 5553291). This has not yet been sent in (I had wanted to try test in a 3rd party system and arrange for an advance replacement) - however the following info now changes things.

Details:

At the same point in time I needed to add more storage to the system (only the original SSD was in machine to start with, 'work' files are on an external NAS).

A new 2 TB HDD was ordered.

The First HDD arrives (WD Black, 2TB). Appears dead on install. Try an alternate power and data cable, followed by changing the controller (marvell/intell) controller, no changes, returned under RMA to supplier (with agreement) as a DOA.

Second HDD arrives. Same symptoms. Very suspicious of this at this point! I specifically try the known good cables from the working SSD. Nada. Install drive in an older backup system. Nada. Speak to the suppliers tech support - neither of us can believe that we have two DoA's, but can not think of a likely alternate reason.

Both drives are returned - and a third drive is sent out. With my agreement the supplier confirms the drive is working prior to shipping.

At this point a PSU issue has occurred to me - so the first thing I do is install it in my older system - and its seen by BIOS with no issues.

I then install it in my main system - dead drive. I do however notice a (faint) electrical 'crack' when the system is booted first time with the drive in. A second attempt to power up (after checking all cables, everything was seated correctly) does not generate the 'crack'. Drive is confirmed dead on multiple sata and power cables, and both sata interfaces on the board.

I move the drive to mu older system - still dead.

I believe at this point we have confirmed (by the elimination of all else) that we have a PSU overvolting the HDD ?

I should also note, that the supplier has received back HDD's one and two, and has confirmed them both as entirely dead.

So...

I am still running the system (I rely on it for work), and am now of course concerned by the SSD (I guess it draws a different current or is less susceptible ?), and the possibility of any other components being damaged (past, present or future tense!) ?

As already mentioned I am currently suffering from a perceived instability from ram and possibly the gpu vram (directx errors, lockups etc). This was up to this point an extensively tested and very stable system! it now appears likely that the PSU could be the root cause of all of the issues. Other than the addition of HDD's no other hardware, bios, firmware changes (etc) have taken place.

Conclusions

How do I stand with regards to an advance RMA on the PSU ? In addition I think at this point with the ability to monitor power (and my concern for this happening again and effecting the rest of the system) I would happily pay the difference for the 'i' variant (which now appears to to be the only one according to the website) - Is this something I can resolve with yourselves or with customer services ?

Where do I stand with regards to the (third!) damaged HDD ? I clearly wont be able to send back a third HDD to the supplier! :roll:

I will now also look to fulfill the existing RMA on the RAM especially if its been overvolted / spiking (this would explain the intermittent nature of the memtest errors). Can I send everything back in one package to make courier costs more palatable and handle both via the advance RMA system ?

Finally...

Thanks in advance for your help and time - I appreciate that occasionally things go wrong with hardware - at least now I guess I wont have to panic too much as I am dealing with you guys (Corsair) as opposed to many other PC manufactures and there customer service 'efforts'.

Regards,

Mizpah / Martin

Mitzpah, what are the voltages being reported in your BIOS? I guess that would be the first question to look at .

How do I stand with regards to an advance RMA on the PSU ? In addition I think at this point with the ability to monitor power (and my concern for this happening again and effecting the rest of the system) I would happily pay the difference for the 'i' variant (which now appears to to be the only one according to the website) - Is this something I can resolve with yourselves or with customer services ?

You would need to place a call to CS to arrange an advanced RMA. They can not be arranged here in the forums because of credit card security reasons. Also , unless the PSU was directly purchased from the Corsair Store they can not refund you money on a unit that was not purchased from them to begin with. If you did purchase it from them, then it may be a possibility but in general a unit can only be RMA'd for the same exact model.

Where do I stand with regards to the (third!) damaged HDD ? I clearly wont be able to send back a third HDD to the supplier!

Corsair would have to test that PSU and determine if it was at fault first. If it is, then you can fill out a damage claim form. But again, it would take a call to CS to get that sorted.

I will now also look to fulfill the existing RMA on the RAM especially if its been overvolted / spiking (this would explain the intermittent nature of the memtest errors). Can I send everything back in one package to make courier costs more palatable and handle both via the advance RMA system ?

The RAM will take up to 2.0v before damage would occur, so i don't believe that is a real danger there. Besides that in order for the RAM to be overvolted to the point of damage that would be a MB fault and not a PSU issue since the RAM receives it's power from the RAM slots and not directly from the PSU. Once power reaches the MB then it is the boards own voltage regulating circuits that determines those voltages.

I really hate to say this, but i believe your issues stem from a bad Mb rather than a PSU. HDD's run of a combination of the 5v &12v rail but have different wattage requirements. With SSD's using considerably less wattage than a a mechanical HDD, and you would really need to be out of spec to damage drives. Add that t your other issues with system stability and it really ppoints towards a bad board than it does a PSU.

Thats just my 2 cents worth though.

Hi Peanutz,

Thanks for taking the time to reply ^^

Lets see how we go with the responses :)

1) Re: replacing with ax 1200 vs ax1200i - ultimately replacing with either is fine, I would just be a little more comfortable if I could pay the difference and upgrade at this point in time. I Have seen evidence on the forums of other people who have been able to achieve the same courtesy of CS. The main point is of course getting it ultimately resolved with out downtime! I have no interest in a refund, just a PSU that works well that I trust :)

2) re: the HDD - Of course more than reasonable - my only issue is playing pass the parcel between ASUS and corsair, something I am keen to avoid at all costs. An initial enquiry to the mobo reseller redirected me firmly back towards the PSU - there logic was that intermittent over-voltage would cause the issues I experienced, and that if it were the mobo, my symptoms woudl be different and the SSD woudl also be an issue...

I of course am at the mercy of other peoples opinions!

3) Re RAM - Okk no problem there, currently it apeeras stable in games (this was not the case a week ago)( however booting into memtest from a CD gives me errors at around the 3rd pass. Could an intermittant adjustment in voltage cauise these lock up and crashes ? (of course if it does so without borking the memory there is no reason to replace it :)

4) Re Board vs PSU: Ultimately I am just keen to get everything resolved (and quite frankly need to with the investment and reliance in the machine. A few thoughts/questions however:

The Board is the Asus Maximus IV Extreme-Z. It should be a top end piece of kit, and on a par quality with the AX1200. I note it has 'overvoltage / antisurge protection' -- However I dont have the specs as to the methods or efficency of the same. I woudl hope however that this would have prevented any issues to the HDD via the SATA Connector ?

Obviously the HDD's direct connection to the PSU will not benefit from the same :P

I don't have a multi-meter to test volts on the board live, in the bios however the 5v rail is reported at 5.040. Do you know of anything I can run when its under load to test the same ? (Clearly there is no HDD on it to draw the load in question however, so I am not sure how much value that will add ?!) By the same token is there a reliable way to log any fluctuations in power over time ?

Can you think of anything else that I should or could be testing ?

Finally do you know of anything technical that would be a detailed primer on all the electronics involved that would help to educate me technically on the power interfaces, loads, watts etc ? I have no wish to chase down a blind alley :)

With SSD's using considerably less wattage than a a mechanical HDD, and you would really need to be out of spec to damage drives.

Note that the SSD runs fine on the same sata cable, interface and PSUU cable - any (3 now tested) 'instadie' on any connection.

I look forwards to any more insight!

OK Jumping in and answering one of my own questions:

http://www.tomshardware.com/reviews/power-supply-specifications-atx-reference,3061-2.html

Gives me a good reference to all things power related.

I note that the SATA power connector has a 3v, 5v and 12v cable . A little research indicates that the SDD will only use the 5v. The HDD will use the 12v feed for the motor, and the 5v for electronics.

Does this raise the prospect of the 12v feed being the issue ? Is this used anywhere else, or is monitored in some way ? What else currently uses the 12v rail that is testable ? (and if its the mobo, will that be protected by the VRM and over-volt protection it has ?)

Cheers again!

Mizpah

Does this raise the prospect of the 12v feed being the issue ? Is this used anywhere else, or is monitored in some way ? What else currently uses the 12v rail that is testable ? (and if its the mobo, will that be protected by the VRM and over-volt protection it has ?)

No it doesn't....Your allowed 5% variance across all voltages to be within ATX specs. So your 12v rail would have to be sitting above 12.6v to be in any way shape or form considered over volting. To keep destroying just HDD's seems kind of unlikely to me. You CPU uses 12v, the MB uses 12v and also your GPU's use 12v.

Drives not being recognized would mean that the interface is malfunctioning, not that your burning up the motors in those drives. So that would actually point towards the 5v rail which would also affect your SSD

But with out a volt meter to watch your voltages there is really no other way to keep an eye on them accurately because software is often wrong.

1) Re: replacing with ax 1200 vs ax1200i - ultimately replacing with either is fine, I would just be a little more comfortable if I could pay the difference and upgrade at this point in time. I Have seen evidence on the forums of other people who have been able to achieve the same courtesy of CS. The main point is of course getting it ultimately resolved with out downtime! I have no interest in a refund, just a PSU that works well that I trust :)

If you purchased it from somewhere else other than the Corsair store , then all you can do is RMA for the same model. But again, you would need to contact CS by phone and discuss this with them.

2) re: the HDD - Of course more than reasonable - my only issue is playing pass the parcel between ASUS and corsair, something I am keen to avoid at all costs. An initial enquiry to the mobo reseller redirected me firmly back towards the PSU - there logic was that intermittent over-voltage would cause the issues I experienced, and that if it were the mobo, my symptoms would be different and the SSD would also be an issue..

You always have the option to RMA the PSU...But i just don't see a big red flag that says your PSU is malfunctioning. Just use the link on the left to request an RMA but if you get a second unit that does the same thing then you can b sure the problem is elsewhere

peanutz your forgetting the possibility of a short in the molex connector feed,

yes it comes out of the same 12v rail but that doesnt preclude the idea of one of the molex connector wires being backwards inside the psu or indeed from an insulator problem causing a short to occur only when connected to a device,

your correct in that the 12v rail feeds mobo + gfx so it's unlikely to be an overvolt situation on the 12v line,

far more likely as my psu (hx1000w) has just done this exact same thing, its undervolting the 12v rail under load when warm due to something wearing out

hence the instability and random lockups to the system in general (the cd drive uses 12v rail so runing memtest from a cd will likely through up errors on random segments even though thers no problems with it in this situation)

i'll assume due to the lack of info that you used the same power connection each time instead of getting a differnt molex power cable out for the 3rd drive before killing it

since your other items are running off the 12v rail we can assume there fine and there connectors are working as intended, which leaves a few options to do specifically wiht the molex connector used ot kill the hdd's with in terms of having a loose pin, having had some loose connectors that got reseated in the incorrect holes or some other similar problem hence the "crack" on powerup as they short out through the drive

so its either a dodgy psu or a borked cable (most likely a combination of both by now)

as the rma process required the return of all the psu cables + the psu i'd go with getting it replaced tbh at least then you can return the last hdd as DOA request a differnt item or your money back and buy from another supplier as your unhappy with there product shipping.......

peanutz your forgetting the possibility of a short in the molex connector feed,

No i didn't...that was stated in Mitzpahs first post....;)::p:

The First HDD arrives (WD Black, 2TB). Appears dead on install. Try an alternate power and data cable, followed by changing the controller (marvell/intell) controller, no changes, returned under RMA to supplier (with agreement) as a DOA.

 

Second HDD arrives. Same symptoms. Very suspicious of this at this point! I specifically try the known good cables from the working SSD. Nada. Install drive in an older backup system. Nada. Speak to the suppliers tech support - neither of us can believe that we have two DoA's, but can not think of a likely alternate reason.

I could possibly see one bad molex , but not multiples. Besides a low 5v rail isn't going to kill a drive. It may not show up in the BIOS or fail to operate, it to be confirmed dead at his supplier with a totally different PSU is slim to none.

No i didn't...that was stated in Mitzpahs first post....;)::p:

 

 

I could possibly see one bad molex , but not multiples. Besides a low 5v rail isn't going to kill a drive. It may not show up in the BIOS or fail to operate, it to be confirmed dead at his supplier with a totally different PSU is slim to none.

noo he stated that he tried multiple connectors after he fried it, not before

hdd's run of the 12v rail floppy's and ssd's run off the 5v rail

all it takes is one bad connector and mitzpah plugging into it each time.......

also the "crack" noise that makes it not show up will be damageing the psu

hdd's run of the 12v rail floppy's and ssd's run off the 5v rail

Sorry, this is wrong, and a quick google search will confirm that. They run off both 12 and 5v. 12v for the motors and 5v for the interface. as i stated earlier.

http://www.tomshardware.com/forum/249934-32-hard-drive-minumum-voltage-requirements

If any of those rails would be that far out of spec he would have other symptoms besides the drives....He also stated that he fixed the stability issues using the same PSU.

Also, as I stated above, he always has the option to replace the PSU. I just don't see any big red flags that say this PSU is malfunctioning.

3) Re RAM - Okk no problem there, currently it apeeras stable in games (this was not the case a week ago)( however booting into memtest from a CD gives me errors at around the 3rd pass. Could an intermittant adjustment in voltage cauise these lock up and crashes ? (of course if it does so without borking the memory there is no reason to replace it :)

In this case RAm stability would be caused by the MB own voltage regulating circuits because RAM doen't get it's power directly from the PSU.

4) Re Board vs PSU: Ultimately I am just keen to get everything resolved (and quite frankly need to with the investment and reliance in the machine. A few thoughts/questions however:

 

The Board is the Asus Maximus IV Extreme-Z. It should be a top end piece of kit, and on a par quality with the AX1200. I note it has 'overvoltage / antisurge protection' -- However I dont have the specs as to the methods or efficency of the same. I woudl hope however that this would have prevented any issues to the HDD via the SATA Connector ?

The board , no matter how high end it is is subject to failures just like any other component. How many times have you seen or read of brand new board no matter the make or model show up DOA?

unless things have changed drasticly since i was last involved in pcb manufacture the 1.3v (ish) thats used ot power the ram is still transformed via a dc-dc converter taken off the 12v line on the mobo,

this is why random memtest modules appearing faulty on the same set of ram along with no signal lockups when more juice is required (for gaming etc from the gfx card) often are good indicators of a faulty power supply unit

that added ot the fact the op specifically states he started having stability issues after he inadvertantly shorted out the 12v rail while attempting ot connect a hdd (some molex quick release connectors are able ot be plugged in backwards as they dont have the standard molex form on the plug)

this all together would indicate a psu problem over other hardwear problems,

not forgetting of course that if the psu is going down theres a 50/50 chance it will take out everythign else connected to it (a typical example would be pouring water into a powered on system at the psu and discoverign afterwards nothign works)

now i might be way off but since shorting out the power feed at the device(consumer end) bypasses most psu circuit protection systems theres a high chance he damaged it anyway, weather he caused a +and - 12v line to be crossed over or not, the level of active eletrical energy required to kill a hdd is substantially higher than the 1 amp (max) there designed to pull down the wire from the psu, and clearly if the 5v connected device is still operating fine his mobo uses the 12v line and its the 12v line he caused to become shorted out (bleedin lucky he didnt fry his gfx card truth be told)

when attempting to find somethign to help me explain what i mean i found this :

Voltage Regulators

The power supply must deliver a good, steady supply of DC power so the system can operate properly. Devices that run on voltages other than these directly must then be indirectly powered through on-board voltage regulators, which take the 5 V or 12 V from the power supply and convert that to the lower voltages required by various components. For example, older DDR (double data rate) dual inline memory modules (DIMMs) and Rambus inline memory modules (RIMMs) require 2.5 V, whereas DDR2 and DDR3 DIMMs require 1.8 V and 1.5 V, legacy AGP 4x/8x cards require 1.5 V, and current PCI Express cards use only 0.8 V differential signaling—all of which are supplied by simple on-board regulators. Processors also require a variety of voltages (as low as 1.3 V or less) that are supplied by a sophisticated voltage regulator module (VRM) that is either built into or plugged into the motherboard. You’ll commonly find three or more different voltage regulator circuits on a modern motherboard.

i took the liberty of hyperlinking to it directly by copy the part that explains the most,

however from what i understand it was always taken off the 12v rail in the past and although intel did change things on some versions of there chipsets to 5v "most" manufacturers are still takign it wholly from the 12v line input ot the mobo

also noteworthey some manufacturers are moving into soley useing 12v dc output psu's with all the transforming done onboard for devices with other power requirements (to me this is a step backwards ot the early 80's but i suppose it has benefits for crossover between portable and desktop based systems given the ever changing digital landscape)

and no i'm not trying to be condescending and for the most part i also agree that his mobo is probably on the blink (there extreamly prone ot bad capacitors in the last 5 years)

in fact if it were me personally i would be getting that replaced as well,

as for the ram i would go ahead and rma that as well because if you overamp'd /shorted the 12v line chances are you powersurged the mobo and caused some damage to that as well

to note most rma terms dont cover user error however i would probably be contacting a pc professional and then my insurance company with a view to replacing everythign because faulty manufacture or failure ot display the correct warnings about none standard parts is also something actionable if my rma's didnt produce new parts

:brick:/sigh

Mitzpah, as I stated earlier, I really don't think the PSu is the issue but rather the MB. But if you want to have the PSU replaced, just use the link on the left to request your RMA

I'm not going to respond to this thread anymore because it's just going in circles and not doing you any good. Just keep in mind that if you get a new unit that does the same thing get the MB replaced.

If you can,test the RAM in another system , a friends or whatever. If it test out okay there would be no need to have it replaced at this time. Memory has a lifetime warranty on it so that could be replaced whenever or if it ever fails. But with a questionable MB, i would wait...thats just my 2 cents worth though.

/yells for HOODEDAssault...another maximus!

Hmm I had a longer forum post, but the forums must be hungry, they just ate it!

After a phone conversation with Justin (a response to my RMA request), Corsair have (as usual) stepped up to the plate and provided a new PSU and RAM under Advance RMA.

I can now report after swapping everything over that:

• The system is now entirely stable again (long mem test and 36 hours stress testing).
  
• The (fourth!) hdd worked perfectly first time.
  

With all manufacturers things can and will go wrong a percentage of the time - its unavoidable. The measure of a companys service however is how they deal with the issues as presented. Corsair have been a pleasure to work with in a frustrating situation - I can't say the same for other manufacturers.

I will continue to ensure that wherever I have the option I buy Corsair!

To all that took the time to reply to the thread (Peanut, Nano, Ram guy et all), many thanks for your time and effort.

Mizpah