Potential thermal issue with RM750 and RM850

[jAMBAZZ]

Room temperatures are typically much lower than internal PC temperatures, so there shouldn't be a problem unless the room temperature is unusually high (say, over 35°C). The fault was discovered when ambient temps were much higher (45°C) and without airflow, thus creating a smaller delta between ambient and operating temperature. Since the fan operates on both temperature and time delay, the temperatures raised too quickly for the fan controller to operate the fan. The ambient temperature of the room alone should be enough to maintain a lower operating temperature, so even prolonged operation at higher loads should give the fan controller adequate time to turn on the fan.

 

 

 

Actually, no. While some fanless units have orientation requirements, the RM Series can be mounted in any position just like any other actively cooled power supply.

 

Some fanless PSUs require a certain mounting orientation because they use the housing as a heatsink that needs to be exposed to the aspiration within the case or have vent holes that need to circulate air to/from the outside of the chassis. But keep in mind, these PSUs are made to run at 100% load without a fan. The RM cannot and should not operate anywhere near 100% without the fan turning on.

 

Jonny I have some questions/assumptions I hope you could look at before I return my 750RM as my SN is 1333. :[pouts:

In all these assumptions, I assume that the PSU FAN is not spinning, because in my system, that draws 360W, I have not yet seen the fan spin.

 

1. 
   My first understanding is that the PSU will shut down if the PSU temp is too high. This can happen if the ambient (room) temp is > 45C. But if the airflow in the PSU is good, because of case architecture, PSU should be OK.
   PSU could also shut down when the ambient (room) temp is fine, but the case temp is very high (>45C), this could happen in "traditional" cases with very bad airflow, specifically where the PSU is top and rear mounted, because the hot case air convects up and into the PSU, adding more heat to the PSU which it self is already warm because its fans are not spinning.
   

• 
  Can you clarify when the FAN will turn on in PSUs prior to SN=1345?
  As I understand, it will only spin if load >= 40%, regardless of PSU temperature, correct?
  

• 
  If the PSU is bottom-mounted in a case where it intakes air from the floor and the PSU section of the case is cut off from the rest of the case, then the PSU sucks in room temp air, and as long as this is not >= 45c this situation will be dandy, because the case airflow cannot impact the PSU temp.
  

 

I my assumptions are true, I cannot see why I should RMA my 750RM.

I just have to either:

• Monitor the case temp so that is not >45C
  
• Use more power in which case the PSU fan will spin and remove air thus reducing temps to safe levels.
  
• OR get a new case with bottom mounted PSU that intakes air from floor.
  

 

My specs:

Antec Sonata I, PSU rear and top mounted.

Intake fan: 120 mm 36/74 CFM (low/high)

Exhaust fan: 120 mm 36/74 CFM (low/high)

i5 2500K overclocked

GeForce 670GTX overclocked

Combined wattage at max load 375 W (3 threads of small FFT Prime95 + Furmark burn-in)

[jonnyguru]

Jonny I have some questions/assumptions I hope you could look at before I return my 750RM as my SN is 1333. :[pouts:

In all these assumptions, I assume that the PSU FAN is not spinning, because in my system, that draws 360W, I have not yet seen the fan spin.

 

How are you measuring this 360W? Also, if this is measured at the wall, keep in mind that this is AC, not DC. So if the PSU is 90% efficient, that means 90% of this AC is turned into DC, so the output is in the neighborhood of 324W.

 

1. 
   My first understanding is that the PSU will shut down if the PSU temp is too high. This can happen if the ambient (room) temp is > 45C. But if the airflow in the PSU is good, because of case architecture, PSU should be OK.
   PSU could also shut down when the ambient (room) temp is fine, but the case temp is very high (> 45C), this could happen in "traditional" cases with very bad airflow, specifically where the PSU is top and rear mounted, because the hot case air convects up and into the PSU, adding more heat to the PSU which it self is already warm because its fans are not spinning.
   

 

Correct.

 

• 
  Can you clarify when the FAN will turn on in PSUs prior to SN=1345?
  As I understand, it will only spin if load >= 40%, regardless of PSU temperature, correct?
  

 

No. The fan turns on based on temperature, but there is a delay. The delay was added to prevent the constant on - off - on -off condition we got to see with earlier versions of the ZeroRPM (for example: the GS Series). Using the load in the marketing just makes the point in simpilar terms. I'm afraid I don't know the exact internal temperature the fan is supposed to turn on at, but if the room is 25°C and you're putting out 300W, your fan should turn on after about 15 minutes.

 

• 
  If the PSU is bottom-mounted in a case where it intakes air from the floor and the PSU section of the case is cut off from the rest of the case, then the PSU sucks in room temp air, and as long as this is not >= 45c this situation will be dandy, because the case airflow cannot impact the PSU temp.
  

 

Correct.

 

If my assumptions are true, I cannot see why I should RMA my 750RM.

 

I just have to either:

• Monitor the case temp so that is not >45C
  
• Use more power in which case the PSU fan will spin and remove air thus reducing temps to safe levels.
  
• OR get a new case with bottom mounted PSU that intakes air from floor.
  

 

My specs:

Antec Sonata I, PSU rear and top mounted.

Intake fan: 120 mm 36/74 CFM (low/high)

Exhaust fan: 120 mm 36/74 CFM (low/high)

i5 2500K overclocked

GeForce 670GTX overclocked

Combined wattage at max load 375 W (3 threads of small FFT Prime95 + Furmark burn-in)

 

How long did you run Prime95 and Furmark together? If it's cold in your room right now, the fan may not trigger for a while. Like I said, it has to reach a particular temperature and maintain it for a period of time.

 

I'd say fire it up again and let it run for at least half an hour. Either the fan should turn on or the PC should shut off.

[jAMBAZZ]

How are you measuring this 360W? Also, if this is measured at the wall, keep in mind that this is AC, not DC. So if the PSU is 90% efficient, that means 90% of this AC is turned into DC, so the output is in the neighborhood of 324W.

 

I measured it at the wall so you are totally right about the efficiency and AC/DC thing. I get 220V btw.

 

 

No. The fan turns on based on temperature, but there is a delay. The delay was added to prevent the constant on - off - on -off condition we got to see with earlier versions of the ZeroRPM (for example: the GS Series). Using the load in the marketing just makes the point in simpilar terms. I'm afraid I don't know the exact internal temperature the fan is supposed to turn on at, but if the room is 25°C and you're putting out 300W, your fan should turn on after about 15 minutes.

I think we need more than 1 factor such as "room temp is 25°C" because if the PSU is bottom mounted in a room temp of 25° and uses 300W I would not expect it to start spinnings its fan at all, ever. On the other hand if room temp is 25° and the case temp is bad b/c of bad airflow and 300W output I would expect the fan to turn on.

In either situation: I thought the problem was that the fan does not turn on @ low loads in these early batches so why are we discussing this? I am still not sure I understand the conditions that need be met before this PSU shuts down. Also I don't understand what OTP means :confused:

 

 

 

How long did you run Prime95 and Furmark together? If it's cold in your room right now, the fan may not trigger for a while. Like I said, it has to reach a particular temperature and maintain it for a period of time.

 

I'd say fire it up again and let it run for at least half an hour. Either the fan should turn on or the PC should shut off.

 

I think I left it on for 30 minutes, but I will test some more in the weekend!

 

 

PS: I found that I have a higher max wattage if I Prime95 1 less thread than available, because this leaves enough CPU resources to saturate furmark, which otherwise does not use 99% of the GPU.

[jonnyguru]

OTP = Over Temperature Protection. OTP is what causes the PSU to shut down PRIOR to the fan coming on.

 

And you're correct. I should have said "ambient" instead of "room". It doesn't matter if the room is 25°C if the PSU is open to the inside of a hotter chassis. So to correct what I said: "If the ambints are 25°C and you're putting out 300W, your fan should turn on after about 15 minutes."

[abn505]

I recieved my RM850 directly from Corsair on 11-15-2013. I hope it was the newer model. I will check when I get home. It came directly from Corsair and took a week, so I am hopeing.

[abn505]

Well, the number is 1338, just 5 from the newer lot number. Called Corsair, they said if you are a normal user with a case that is well ventilated, my case is the 750D. We both agree that I should not see the temps we are talking about. At least I have 5 years if it does act up.

[abn505]

I have had my RM850 now for two weeks. I have not had any issues, although the fan has never kicked on yet. I have run Futermak 11 and PCmark7 with no issues. Fan still never came on. I did not run them at the same time. I didn't think that was possible. I have my RM850 installed with the fan up. What is the best way, fan up or down? I would think either way is fine, I know up will pull in case air, but I have a 750d case with plenty of air flow. Fan down will pull in air from the bottom, but the small gap betwwen the bottom of the case and the table wouldn't allow for actual room temp air to come in I would think.

[jonnyguru]

I have had my RM850 now for two weeks. I have not had any issues, although the fan has never kicked on yet. I have run Futermak 11 and PCmark7 with no issues. Fan still never came on. I did not run them at the same time. I didn't think that was possible.

 

I don't think it is. That's why I suggest running Furmark and Prime95 at the same time.

 

I have my RM850 installed with the fan up. What is the best way, fan up or down? I would think either way is fine, I know up will pull in case air, but I have a 750d case with plenty of air flow. Fan down will pull in air from the bottom, but the small gap betwwen the bottom of the case and the table wouldn't allow for actual room temp air to come in I would think.

 

Either way is fine.

 

Looking at your system specs, I'd say your power requirements are so low that your PSU fan may never turn on. Not unless you put a second graphics card in there.

[abn505]

I thought about SLI, but don't know how much I would gain. I suspect not a lot to justify 300 more dollars.

[spqr]

Damn, my RM 750 just arrived, and it is affected by this problem (LOT 1336) :(

Now waiting for answer from Corsair to get a replacement...

[peanutz94]

Vuk, have you used the PSU yet and had any trouble with it? There is no need to have it replaced unless it is going to meet the criteria in the OP.

 

As long as you have a well ventilated case you shouldn't have any problems.

[ICDP]

Well, actually 1355 is greater than 1341, so if that were your serial number, you'd be ok. ;)

 

But also, that can't be your serial number. Can you take a pic? Reason being: The first two digits are the year (13 = 2013) and the next two are the week (52 = last week of the year), so 1355 is impossible. ;)

 

Hi, I opened a ticket and called USA tech support and they opened an advance replacement returns for the PSU. Unfortunately the RMA they sent for the original PSU is not prepaid freight costs as mentioned by the OP.

 

My ticket number is 6261275 and I left another note saying the RMA should be prepaid. Is there anything else I should do to ensure my freight is paid for. The girl on the phone said they would take $120 from my credit card I don't return the original PSU. I don't want to be out of pocket for returning a PSU that has a design flaw.

 

Incidentally I purchased an extra GTX780 for SLI.

[Zapee995]

Hello Corsair community! My pc is:

 

Z77 extreme 6

3770k@~4.6Ghz

Gtx780

XMS3 8Gb 2000Mhz

Samsung 840Pro 128GB

WD 500GB

Corsair TX750M

 

My computer is under custom watercool with one d5 pump and my pc case corsair air 540.

 

I purchased the RM850 but didn't tested it yet...The lot number is 1335 and at this point i have to ask..Every RM750/850 below 1341 has this problem or some specific montels??

 

My room temp is ~22oC an i will test the psu soon. :P

 

Another thing i want to know is how long does the corsair RMA take?? Thanks in advance.:)

[spqr]

Hi!

 

My first RM750 LOT was 1336.

I got the replacement, the LOT is 1333 :eek:

 

It was told, that under 1341, it is factory faulty, now what??

Replacement LOT is 1333, but it's fixed? :question:

 

+ the nylon was open, which contained the cables... Corsair sending used/repaired PSUs as replacement for brand new? SHAME.

[ICDP]

Hi!

 

My first RM750 LOT was 1336.

I got the replacement, the LOT is 1333 :eek:

 

It was told, that under 1341, it is factory faulty, now what??

Replacement LOT is 1333, but it's fixed? :question:

 

+ the nylon was open, which contained the cables... Corsair sending used/repaired PSUs as replacement for brand new? SHAME.

 

I did an advance RMA as well. 1st PSU was from lot 1335, the second (replacement) is from the same lot number. Does this mean Corsair sent a similarly affected PSU?

 

I did get one power off problems (GTX780 SLI) and have not seen it with the new one. Is there way we can tell if our replacement PSU are OK?

[jonnyguru]

+ the nylon was open, which contained the cables... Corsair sending used/repaired PSUs as replacement for brand new? SHAME.

 

Can't explain why the cable bag was open, but the PSU has only been out for two months so I doubt it's "used". It would take longer than that to process, re-certify, ship, etc.

 

My best guess? Engineering probably tested the unit before they shipped it and used the cables to hook it up to the Chroma.

[spqr]

I did an advance RMA as well. 1st PSU was from lot 1335, the second (replacement) is from the same lot number. Does this mean Corsair sent a similarly affected PSU?

 

I did get one power off problems (GTX780 SLI) and have not seen it with the new one. Is there way we can tell if our replacement PSU are OK?

 

Eh, what a shame...

You applied for another replacement?

 

I think I will contact corsair again and ask for a replacement with LOT number >1341, this is totally ridiculous....

If they are not willing to replace it, then I am gonna sell this crap as new and buy Seasonic, which is the best anyway.

 

@ICDP: mine makes a clicking sound when it turns on/off. Is it so with yours too? Weird, coz I have never heard any PSU doing so... I am totally fed up with corsair...

[jonnyguru]

A lot of PSUs click when they turn on and off. You have a relay in there that switches power from the +5VSB to +5V when turned on.

 

As for the lot codes you guys are seeing: When the problem was first discovered, a new run of units were not yet in production. I believe they reworked existing production units by adding the second thermistor, so they would not have new serial numbers (although I personally think it would have been wise to somehow mark the units differently).

 

We're going to confirm this with engineering, but as you may or may not know; it is Thanksgiving holiday in the U.S. so nobody is available in the Fremont office to answer these kinds of questions until 12/2.

[spqr]

A lot of PSUs click when they turn on and off. You have a relay in there that switches power from the +5VSB to +5V when turned on.

 

As for the lot codes you guys are seeing: When the problem was first discovered, a new run of units were not yet in production. I believe they reworked existing production units by adding the second thermistor, so they would not have new serial numbers (although I personally think it would have been wise to somehow mark the units differently).

 

We're going to confirm this with engineering, but as you may or may not know; it is Thanksgiving holiday in the U.S. so nobody is available in the Fremont office to answer these kinds of questions until 12/2.

 

Thanks jonnyguru!

I hope so, but the problem is, that I can not even sell this unit with this LOT... Everybody thinks this is faulty one, so I would need another!

[ICDP]

A lot of PSUs click when they turn on and off. You have a relay in there that switches power from the +5VSB to +5V when turned on.

 

As for the lot codes you guys are seeing: When the problem was first discovered, a new run of units were not yet in production. I believe they reworked existing production units by adding the second thermistor, so they would not have new serial numbers (although I personally think it would have been wise to somehow mark the units differently).

 

We're going to confirm this with engineering, but as you may or may not know; it is Thanksgiving holiday in the U.S. so nobody is available in the Fremont office to answer these kinds of questions until 12/2.

 

Thank you for the feedback. I assumed this would be the case with existing stock modified. I will check back after the holiday period to see if there aer further updates.

[Zurs]

My RM850 just came today and I stumbled across this thread. My sr # is 1335.

I got it thru Amazon. They do a very high volume of business and I'm hoping the fix has been made to this PSU already but have no way of knowing I guess.

 

After I install and test it I will post back with my results. I really don't want to RMA it.

 

Thanks and Cheers!

[Technobeard]

My RM850 just came today and I stumbled across this thread. My sr # is 1335.

I got it thru Amazon. They do a very high volume of business and I'm hoping the fix has been made to this PSU already but have no way of knowing I guess.

 

After I install and test it I will post back with my results. I really don't want to RMA it.

 

Thanks and Cheers!

The fix has not been made to your PSU.

If you have no way to test it because your load is so small that it would never trigger the fan or trip the OTP, then the best advice I can give you is to RMA it regardless just so you don't have to deal with it once you do get the second card and start overclocking.

 

The affected units are in lot codes < 1341 (the first four digits of the S/N.)

 

Having said that, just because you have one of the PSUs doesn't mean you'll ever see the issue.

[spqr]

Now I am really fed up with corsair...

LOT of my first PSU was 1336. I required for RMA, got a new one with LOT 1333... totally brainless step from corsair.

I asked for RMA again, I was told to get a new PSU with LOT 1341+, I have proof of this answer.

The replacement PSU of the REPLACEMENT has just arrived, the LOT is 1336...

 

:...:

 

WORST. CUSTOMER SUPPORT. EVER. :veryangry

 

I can not even sell this unit, because ppl know this is faulty, because of the LOT.

 

Now I just want my money back, and get SEASONIC, won't buy corsair product ever again...

[jonnyguru]

So... you want a PSU with a lot code higher than 1341 just so you can sell it?

 

I believe I explained in my previous post that the units were reworked and that they would have the same lot codes as before.

 

The lot codes represent the date of manufacturer: 13 = 2013. 36 = 36th week of 2013. Anything newer than 1341 would have to be a completely new production run and I don't believe those have even shipped out of China yet.

 

Also, you JUST bought this thing not even a month ago. If you're so intent on getting a Seasonic, why didn't you return the PSU to the shop you bought it from and ask them to send you something different in return?

[spqr]

So... you want a PSU with a lot code higher than 1341 just so you can sell it?

 

I believe I explained in my previous post that the units were reworked and that they would have the same lot codes as before.

 

The lot codes represent the date of manufacturer: 13 = 2013. 36 = 36th week of 2013. Anything newer than 1341 would have to be a completely new production run and I don't believe those have even shipped out of China yet.

 

Also, you JUST bought this thing not even a month ago. If you're so intent on getting a Seasonic, why didn't you return the PSU to the shop you bought it from and ask them to send you something different in return?

 

Man..., I was told to get a PSU with LOT code 1341:

 

"Created By: ******** TS (12/2/2013 4:58 PM)

I've talked with customer service and they'll be arranging another replacement with the correct lot code. As soon as I have the updated information I'll let you know."

 

This was a lie.

And yes! I am going to sell the PSU right when it arrives.

 

Why I didn't send it back to the shop? Because I ordered it from abroad it would cost quite a lot to send back 4kg package, and asked corsair support, if they are going to replace it, if it is affected, and they told YES.

 

I don't really care if PSUs with LOT 1341+ are out from China or not... Then why didn't they tell me that? And send the replacement, when it is possible.

 

You shouldn't really be upset, I think, I paid for a faulty product.... I don't want something brand new reworked... If so, it is called refurbished, and it costs less or so. On the top of it, it is announced, if something is refurbished... With this "problem" I can not even sell it, noone gonna buy this.

 

What a crap service.... :mad:

 

Oh yeah, and few days ago you couldn't even tell, if it is reworked or what....