Norco hotswap controller causes drive death

[micron98]

Following is my configuration when hdds fried.

 

i7 3770

2 x G.Skill 8GB

Antec 550 Platinum

Onboard Video

2 x IBM M1015

3 x 12V 0.65A 120mm FANs at 50% duty cycle

3 x WD EARX 2TB

1 x Intel 510 120GB SSD

1 x Seagate 500GB

 

According to the spec there are 2 rails and I am assuming one goes to the board and CPU and the other for GPU but in my case I dont have GPU installed so all 1 rail is connected to rest of components other than CPU,RAM,Onboard Video

 

While it was running in above configuration, I hot added additional drives which ended up frying.

A few brancs of dead drives :

hitachi desk star 2tb

WD earx green 2tb

seagate 500GB

Maxtor 73GB sas drive

 

so not really related to 3TB drives at all. It managed to burn green drives as well.

 

You  might want to take a multimeter and test the molex plugs on your PSU. make sure that it is within specs.

 

I took current values using above configuration and it reads

 

12V : 12.33~34

5V : 5.05~5.06

 

and according to wikipedia http://en.wikipedia.org/wiki/ATX

 

tolerance :

12V : 0.60V which is +11.40 V to +12.60 V so 12.34 is within spec

5V : 0.25V which is +4.75 V to +5.25 V so 5.06 is within spec

 

I am still running everything without issues except that I am afraid to add more drives until backplanes are all sorted.

 

@johnm if you dont mind, would you share your backplane version? and whether it is mini-sas version?

[BLKMGK]

Just curious, for those that have seen drive failures - were you powering both Molex connectors? I'm currently running more and larger drives than the last poster and at least 6 were hot added. I'm currently only powering one set of Molex but intend to power both soon...

[Johnm]

@micron98

 

If you look at the actual specs for that PSU, it looks to be a bad choice for a storage server. That is a gaming PSU. It is a 550watt 4 rail system (according to the output specs, it has 4 rails).

 

I would defiantly not rule out that your PSU could be the problem. especially since you had no visible backplane damage.. you might have overloaded a single rail (although you usually see drive go offline, not poof). It is also possible you have some slop in one (or more) drive bays and a plug misaligned and shorted the whole chain (12v on negative would not be good). it could have also be a spotty molex y-adapter (again not very likely, but happens).

 

I'll have to find my backplane models I have several, I have 3x 4224's and one has had a SAS backplane swapped to SATA one. I think they all have different generations of backplanes.

[micron98]

@BLKMGK

Just curious, for those that have seen drive failures - were you powering both Molex connectors?

 

I only had one of Molex connected to the PSU

 

If you look at the actual specs for that PSU, it looks to be a bad choice for a storage server. That is a gaming PSU. It is a 550watt 4 rail system (according to the output specs, it has 4 rails).

 

I was looking at this link http://www.antec.com/product.php?id=705285&fid=5022027 which says "2 fully-protected High Current +12V rails with high load capabilities ensure maximum CPU & GPU compatibility"

 

but on the output table there are 4 12V and this site http://www.pccasegear.com/index.php?main_page=product_info&cPath=15_226&products_id=19330 says 4 fully-protected high current rails

 

different versions? or just a mistake?

 

anyhow according to my calc,

3 x WD earx 2TB consume 3.3W idle so : 0.275Ax3 = 0.825A

1 x Segate 500GB consumes 8W idle so : 0.66A

1 x Intel SSD consumes 100mW idle so : 0.008A

3 x Fans consume .65 @ 50% duty cycle : 0.325A x 3 = 0.975

All up 2.468A

 

even I had everything in one of the 4 rails ( supports 30A each ) it shouldn't overload the PSU and that is what I have seen.

 

When I hot added drives, existing drives idle and were functioning and still continue to function even hot added drives burned.

 

I think this proves that the PSU was not overloaded and drives were fried due to bad backplanes.

 

 

 

[BLKMGK]

It would be helpful if Norco could clarify what the two Molex connectors are for. If it's for load sharing to take load off of some component that has marginal specs then they need to document it!

[micron98]

Can anyone else confirm your backplane version?

[BLKMGK]

Mine is up and running with all backplanes in use but next it's down I'll try to pull one and take some pics etc.

[micron98]

Thanks @BLKMGK

[micron98]

Contacted retail again about this and came back with a reply that Norco replied that out of 400 sold I am the second to report similar issue and added that this is not a common issue.

 

If you however experience the issue, it is deadly. Recommend new buyer to test with a spare crappy drive first then implement the rest.

 

I have asked retailer to replace all 6 backplanes since two of 6 failed on me and I dont know whether rest 4 are bad or not and this  may be due to a bad batch.

 

Will give you more updates on how things progress

 

 

 

[bcann]

Contacted retail again about this and came back with a reply that Norco replied that out of 400 sold I am the second to report similar issue and added that this is not a common issue.

 

If you however experience the issue, it is deadly. Recommend new buyer to test with a spare crappy drive first then implement the rest.

 

I have asked retailer to replace all 6 backplanes since two of 6 failed on me and I dont know whether rest 4 are bad or not and this  may be due to a bad batch.

 

Will give you more updates on how things progress

 

Funnily enough, i got told the exact same thing. I am only one of 2, guess that means im "special"

 

Yay for me.

 

They emailed me out of the blue and said that they would take the whole lot back and give me a full refund. i said they can take the whole order back, as it was all bought for the case, so they can take the sas cables and the fans and the fan mount as well as far as i am concerned.

[micron98]

I've ordered two drives with the replacement backplanes and I will use these two drives for testing hotswap. If these die like the others, retail store said that they will reverse the whole order including the two drives

 

hope it was just a bad batch. 

 

@bcann I also feel very special...

[guyonphone]

I have yet to have any issues, here is the backplane I am running, it appears to be older than the ones currently posted on the thread.

 

One or more drives burned : no

Used hotswap : no

Type : mini-sas

Case : 4224

Version :

94V-0

RoHS

 

Fuhe BP-001MA

6 Gb/s SAS/SATA

/2011.10.03

[micron98]

Two WD EARX 2TB drives + 6 new backplanes arrived.

 

Replaced all 6 backplanes and emptied all drives.

 

Turned on and went into the SAS bios to monitor drive activities.

 

Started sliding in the first drive into the top left corner.

 

and a few seconds later added another into the second left of the top row.

 

SAS card couldnt see any drives so popped them out and tried them on the

 

USB docking station and couldnt feel drives spinning.

 

Well in summary : two more drives dead and I will be requesting for refund of the case and two drives.

 

I think this is as far I can go with this case.  Either I am very very very special or it wasnt the backplane.

 

perhaps this is an issue with the case itself where something is touching the backplane where it shouldnt.

 

Hopefully @bcann you can find your peace soon

 

 

[bcann]

UPDATE:

 

The OEM in australia issued me a FULL refund with all anciliary parts refunded as well, as well as also refunding the postage costs AND paying in full for the courier to come and pick it all up.

 

Having said that i'm still out for the cost of a PSU, and 4 * SAS HDD's which admittedly i got cheap enough, but its still a $400 odd dollar lesson.

[Johnm]

That makes me think it is something else like the psu as I had suggested originally. The chances of swapping the backplane and popping the drives again sounds suspicious. That's the common denominator each time.

[micron98]

That makes me think it is something else like the psu as I had suggested originally. The chances of swapping the backplane and popping the drives again sounds suspicious. That's the common denominator each time.

 

Well I have migrated all into an old desktop using the same PSU, MB, RAM, CPU etc. with 8 drives connected to the same sas controller and no issues over two months.

As far as I can tell, no issues with my PSU.