SP1614N Temp Reporting

jamesbeam · Post by **jamesbeam** » Wed Jan 26, 2005 9:16 pm

Greetings SPCR'ers.

I have been trying to stress test a Samsung SP1614N (standard mounting at bottom of cage) by running jobs such as Defrag, Precompact and Sandra File System Benchmark.

Here is the Speedfan graph:

Ignore the left hand side.

Temps were crosschecked with DTemp and HDD Temperature.

Despite polling intervals of as little as 20 seconds, it seems that the HDD is only updating the S.M.A.R.T. temperature information every 10 minutes.

The questions:

Is the 10 minute update interval normal? Is there anything that can be done to increase the frequency?

Temps are reported in increments of 2 degrees. Is this normal?

The temperature changes between readings are as much as 14 degrees. I can understand that different parts of the tests would cause different temp changes, but the swings seem excessive. Is there a test program that will cause a steady rate of heating?

The Rig:

Recently put together after lurking in SPCR. Thanks contributors

A64 3000+ CG underclocked at 1.5Ghz / 1.11v. 7000B-AlCu @ 1150 rpm. Asus K8V SE Deluxe. Antec Phantom. ATI 9550. SP1614N. 1.5GB DDR.

The generic case was too hot, so mesh covered windows were cut into the side panels.

Pics of modified case here.

I am very pleased with this rig. With only 1 fan it is almost as quiet as my VIA Eden 533 / Seagate Barracuda IV, but much faster, and expandable.

CPU temp at full load is 49C (ambient 33) with the fan at 1150 rpm.

Now I want to make sure that all the other components remain at safe temps with ambients up to 35C.

Cheers,
James

burcakb · Post by **burcakb** » Thu Jan 27, 2005 12:06 am

James,

WELCOME TO SPCR !!!

You could try looking at SMART temps with MBM. However, MBM SMART monitoring option has a warning that says don't poll at intervals lower than 90 seconds. From that, I'm guessing it's not a good idea to poll constantly.

I've never personally tried a defrag on my spinpoint but 51C seems way too high for such a cool harddisk

eLekTRiK · Post by **eLekTRiK** » Thu Jan 27, 2005 12:53 am

I have 2 SP1614C 160GB drives in SmartDrive 2002C enclosures. They idle at 31-33C.

I also noticed that the drive temps don't update very often, and couldn't find a workaround.

Not only that, but I've found using an external probe that the drive temperature (measured on the side of the drive) is about 5 degrees higher than what is reported through S.M.A.R.T.

FWIW, the highest temp that I've seen in my setup is 55 degrees while rebuilding a RAID 1 mirror (prolonged continuous writes - interestingly, the source drive didn't get hot). This is with a 120mm intake fan blowing over the enclosures. I haven't had any trouble with the drives yet, and I don't think they fared that much better outside the enclosures.

rtsai · Post by **rtsai** » Thu Jan 27, 2005 8:51 am

I can confirm what you've seen. I have a background process monitoring my SP1614N temperature, and it's only ever reported temperatures of 14, 16, 18, 25, 27, 28, 29, 31, and 33 C, going back to August 2004.

If anything, I've kind of been worried that the drive is under-reporting temperatures, but (a) with 33C, I have a lot of headroom, and (b) this is Boston in winter, so it certainly feels pretty durn cold indoors at times.

Jan Kivar · Post by **Jan Kivar** » Thu Jan 27, 2005 9:14 am

Wow, that's really one oddball for a drive.

First of all: Were you using more than one temp monitoring program at a time? Because I've never seen a SpinPoint with a temp over 50°C. What is the firmware revision of the drive (you can check it from the SMART page in SpeedFan)?

One "easy" way is to use IBM's/Hitachi's Drive Fitness Test. There is a feature called Exerciser, which makes the drive seek constantly. IIRC it should show the temperature while the test runs, but there is no screencaps on Hitachi's site to prove me right or wrong.

I haven't seen any 10-minute update interval. I haven't monitored my SP1614N's that closely, but my SP1213N does change every minute (when there is a change - 0 to 30 mins after cold boot, until the temperature stabilizes). Again, the firmware revision could be the key, both my drives have "TM100-24" (don't remember what revision for SP1213N).

2-degree increments are what I'm seeing in all my three drives. The value is always odd (opposite of even, that is), but if the drive is power-cycled (shutdown computer, then start again), there is a small period when a even value is shown. Search up my thread about SP1213N, there's more detail about the temperature obscurities.

Cheers,

Jan

EDIT: The small text contains false information. The correct utility to use is the (Drive) Feature Tool found here.

jamesbeam · Post by **jamesbeam** » Thu Jan 27, 2005 7:03 pm

Thanks for the helpful replies everyone.

Since posting I have taped a thermometer to the top of the HDD and rerun the Sandra test. SMART ranged from 35 to 43 but the thermometer only ranged 37.4 to 38.3. Worse, there didnt seem to be any correlation between the two sets of readings.

Next step is to try the Hitachi Drive Fitness Test. Probably wont get a chance to run it for another 24 hours.

As others have mentioned, SMART only reports odd numbers except for a brief period after a reboot.

Firmware is TM100-24.

jamesbeam · Post by **jamesbeam** » Sat Jan 29, 2005 11:59 pm

Next step is to try the Hitachi Drive Fitness Test

Alas, cant watch temps as it runs.

One interesting finding - the SMART temps are updated once a minute for approx 30 minutes after a reboot then revert to 10 minute updates.

Anyway, the highest HDD temp seen in the tests was 20.3 above ambient. That compares well to the 18.9 degrees quoted at Storage Review.

Design Ambient of 35C + 20.3C = 55.3C

I think I can live with that. Decoupling the drive might be pushing things a bit though.

Cheers,
James

Jan Kivar · Post by **Jan Kivar** » Mon Jan 31, 2005 6:46 am

jamesbeam wrote:
Next step is to try the Hitachi Drive Fitness Test
Alas, cant watch temps as it runs.

Now that mention it... It was actually the (Drive) Feature Tool found here. See the User Guide page 18. Sorry about the false info. My bad.

jamesbeam wrote:One interesting finding - the SMART temps are updated once a minute for approx 30 minutes after a reboot then revert to 10 minute updates.

That sounds interesting. I've noticed that after shutting down completely and then (almost) immediately powering up again the drive's temp is reported higher than before the shutdown.

Cheers,

Jan

Jan Kivar · Post by **Jan Kivar** » Mon Jan 31, 2005 7:09 am

jamesbeam wrote:One interesting finding - the SMART temps are updated once a minute for approx 30 minutes after a reboot then revert to 10 minute updates.

Seems to hold water after brief defrag tests.

Jan