How reliable are SMART drive temp sensor readings?

smilingcrow · Post by **smilingcrow** » Fri Jul 14, 2006 8:43 am

I have a Samsung P120S 250GB and a WD3200KS drive in my system and have been monitoring the temps. According to Speedfan & DTemp the WD seems to sit around 50C, whilst the Samsung is usually 4 or 5 degrees cooler. But when I touch the drives, the WD feels warm whilst the Samsung is hot. Is this an error with the accuracy of the SMART data or does it mean that the WD is cooler on the surface of the drive and hotter inside where the temp sensor is?

Both drives are temporarily sitting on acoustic isolation foam in a P180 with the case open. The power supply is currently sitting outside of the case whilst I evaluate a number of components. The power supply is a passive Silvestone and it feels slightly cooler than the Samsung which is strange as it is dissipating more heat and over a larger area. The WD is in the lower chamber of the P180 whereas the Samsung is sitting on the bottom of the main chamber. In effect, the WD is in a cooler environment than the Samsung.

Iâ€™ve not evaluated the noise of the WD drive yet as I have too many other more important issues to deal with first. However, it does seem to make a loud clicking sound once in a while as it possibly resets itself. Odd!

If I do decide that the WD is quiet in my system, am I safe to rely on manual temp sensing? Itâ€™s obviously a lot easier to use SMART temp readings when locating a drive and determining whether it has enough cooling.
Both drives are connected to the same SATA controller, which is part of the Intel chipset.

Bluefront · Post by **Bluefront** » Fri Jul 14, 2006 3:32 pm

Apparently there is no set location for Smart temp sensors across different brands. If your smart temp is only 1-2C over ambient by the time the system boots after sitting off for a few hours, at least you know the start point is fairly accurate.

Take a manual temp reading off the edge of a drive.....that's a fairly good way to judge temps, and the accuracy of the smart reading. (IMHO)

wwenze · Post by **wwenze** » Fri Jul 14, 2006 4:24 pm

The Smart readings are within 0-2 degrees (higher) than what I get from the underside of the case nearest to the platter for all my drives.

arrikhan · Post by **arrikhan** » Sat Jul 15, 2006 12:30 am

I would assume that the outside temp, is different from the inside temp and not a good reading to take unless one understands the ratio between the two due to the constant heat disipation off of the HDD casing.

I'm assuming the probe for HD's that SMART reads from is actually internal. Would that be right? ...thus the temp is only as good as the probe

Arrikhan

smilingcrow · Post by **smilingcrow** » Sun Jul 16, 2006 2:07 am

arrikhan wrote:I would assume that the outside temp, is different from the inside temp and not a good reading to take unless one understands the ratio between the two due to the constant heat disipation off of the HDD casing.
I'm assuming the probe for HD's that SMART reads from is actually internal. Would that be right? ...thus the temp is only as good as the probe

You make a very good point. So itâ€™s possible that the WD is not dissipating heat as well as the Samsung, which is why the internal temp sensor is reading hotter whilst it feels cooler to the touch. If anyone else has one of these WD models Iâ€™d be grateful if you would touch the drive and see how hot it feels compared to the reported temperature.
I touched the side of the drive and that felt hotter than the top surface. Maybe itâ€™s designed to dissipate heat more through the side of the drive which is typically in contact with a metal cage! Sounds a bit unlikely and risky also as you canâ€™t guarantee this contact.

The very occasional load clicking sound that the drive makes made me wonder if it was damaged. I ran the WD diagnostic and it reported errors and that I should contact tech support. Hard to say if this fault is affecting the drive temp.

I havenâ€™t been able to test the acoustics unfortunately, due to the system being in an open state with the power supply outside the case. I tried touching the WD & Samsung and the WD felt as if it was producing a bit more vibration.

barth2k · Post by **barth2k** » Mon Jul 24, 2006 7:11 am

I have a WD2500KS and smart gives 53C!! It's warm to the touch but not esp. warm or hot. Of course I don't know what the diff. between internal and external temp is.

smilingcrow · Post by **smilingcrow** » Mon Jul 24, 2006 7:26 am

My WD3200KS turned out to be faulty and Iâ€™ve returned it for a refund. I didnâ€™t get the chance to fully test the acoustics unfortunately. It didnâ€™t jump out as being louder than my P120S under the non standard conditions that I had it installed in.

Chrissicom · Post by **Chrissicom** » Thu Jul 27, 2006 12:04 pm

My Maxtor 6V300F0 drives both show 38 Â°C right now with HDDLife Pro and my temp sensor (Aerocool Gate Watch) attached to the above disk near the top left corner of the drive shows 39.3 Â°C. Since the inside temp should be lower than case temp with good heat dissipation I think the Smart readings are pretty accurate.

Spod · Post by **Spod** » Thu Jul 27, 2006 10:08 pm

In general, drives do dissipate more heat from the sides than the top or bottom. People mounting heatsinks on drives have found that mounting them on the sides results in better cooling than mounting them on the top or bottom.

The difference in readings may be due to the drive's position in case airflow, but it's unlikely in this case. It's probably just a difference in the positioning and/or calibration of the respective sensors. You don't know whether it's measuring temperature at the motor, the casing or the hottest chip on the PCB.

If the samsung feels hotter, I'd say it's hotter! Dissipation characteristics aside, once both drives reach a steady state, the drive that dissipates more power will likely end up hotter, unless the airflow or exposed surface area is dramatically different.