Battery Temps on SUA1500RM2U

[Anonymous user]

I've seen a lot of variation in the temperature for the battery in my 1500VA rackmount lately. The UPS is only under 65 - 75% load at all times. The server room operates at 65-70 F. Any reason why this started varying so much lately? I've attached a graph from MRTG (The OID I'm capturing is 1.3.6.1.4.1.318.1.1.1.2.2.2.0).

[Erasmus_apc]

The OID .1.3.6.1.4.1.318.1.1.1.2.2.2 (.0) is actually the upsAdvBatteryTemperature which is the current internal UPS temperature, not a temperature for the battery in the sense of the sealed lead-acid battery inside the UPS. If your load has been fluctuating along with the A/C temp this is a possible explanation for temperature change. Other issues could be the UPS being on battery or bad power causing the UPS to enable it's AVR. Have you added or removed any load recently? How old is this UPS and has the battery been replaced before? Any recent power issues?

Edit: a bad, hot battery could also cause an issue, in theory.

Message was edited by: Erasmus

Message was edited by: Erasmus

[Anonymous user]

If it is a power problem, wouldn't I see the issue with the other 2 UPSes in the same room? This is the only one that is exhibiting this. We did rerun all the wires in the back of the rack, but we did that Saturday night / Sunday morning (The temperature drop-off isn't until well into Monday). There was more of a load on this UPS after we cleaned up the rack than before, which doesn't explain its sudden drop-off and variation of temperatures on Monday. I have a temperature sensor on the intake side of the rack is always between 66 and 70 F (And it is higher than this UPS). We also have a SUA750RM2U right above the 1500, and it isn't experiencing these changes in temperature.

The UPS was purchased in June of 2006, and the batteries haven't been changed since I started here in Feb. of 2007, so I'm guessing they've never been changed.

I've also attached the load graph for the UPS as well.

Message was edited by: jblack

[Erasmus_apc]

We can't rule out power issues occurring just because there are multiple UPS' in one room unfortunately. Do you have any PowerChute or Network Management Card logs that we can look at? Having both the data and event logs would be helpful.

Any other physical changes to the rack setup or the room environment at all?

[Anonymous user]

The UPS log looks pretty clean. That Saturday I used SNMP to shut the UPS down to test something custom I had written to shut the servers off. I did plug one of the servers into this UPS that was supposed to be in one of the other UPSes, but it had little bearing on load (6% to 8% and I just plugged it back into the other UPS). I wouldn't think it would cause variations like I've been seeing in temperatures. Other than that, nothing has changed in the room.

Message was edited by: jblack

[Anonymous user]

Here is a weeks worth of data from the UPS itself.

[Erasmus_apc]

Can I have: the exact model and serial number of the UPS?

Can I also have: The model, serial number and application/AOS firmware version and name on the network card (will be under Administration -> general-> about on newer cards, OR Help->About on cards with older firmware).

I have reviewed and compared the data logs / event logs / temp. graph and as of right now I will be honest and say that I do not see a pattern of decrease/increase in TupsC (temp) compared to in/out AC voltage, batt. capacity, or watts/load%. Edit: and of course there is no event info in the logs that is relevant to the date ranges of your concern.

However...there is a pattern when comparing battery voltage to the temperature. When the battery voltage = 27.40, the temperature is at it's highest readings of 30.1C. When the batt goes to 27.67vdc this is where we see nearly every single range of low temp. values. This could have to do with the charging cycle and how many of our UPS' start their internal cooling fan upon initiation of battery charging...but I will need to check on this, since technically this pattern should have been present before, unless the fans were blocked and no heat was able to dissipate. In the meantime please get me the pieces of information I requested above.

Message was edited by: Erasmus

Message was edited by: Erasmus

[Anonymous user]

Since I removed the other server, the UPS may have gone back to what looks normal (About 30C) and may be sitting there (Time will tell). Is there a magic load % or temperature where the fan starts running? --- Maybe the extra server was just enough to put it over the top.

UPS:

White Label on back says: SUA1500RM2U

Web Interface says:
Model: Smart-UPS 1500 RM
Serial Number: AS0603297919
Firmware Revision: 617.3.D
Manufacture Date: 04/04/05


Network Card:
Hardware Factory

Model Number: AP9617
Serial Number: NA0733010880
Hardware Revision: A10
Manufacture Date: 08/18/2007
MAC Address: 00 C0 B7 73 B5 92
Management Uptime: 5 Days 13 Hours 28 Minutes



Application Module

Name: sumx
Version: v3.5.5
Date: 05/24/2007
Time: 16:14:52



APC OS (AOS)

Name: aos
Version: v3.5.5
Date: 05/24/2007
Time: 15:44:22

[BillP]

the fan kicks in if any one of the following conditions are true:

• the load >75%
•the UPS is on battery
•the UPS is charging the battery
•the internal temperature goes above 104ºF in the UPS

[Anonymous user · Accepted solution]

The UPS must have been right at the load where it kicks the fan on. I've swapped that server (Which changed the load like 5%) and now it is hovering around 30 C (Which is what it used to do). Thanks!