SU3000RUMXL3U doesn't switch to battery when the power goes off

[Anonymous user]

We have a SU3000RUMXL3U unit with two batteries. It's eight years old. One of the batteries was changed six years ago. All the LEDs in the front look normal. The load on the unit is only about 40%. The battery LEDs show it is 100% charged. Run time remaining is 1 hour 18 minutes. There is no under or over current from the utility power supply as indicated by the LEDs. But if the power goes off for even a split second, the UPS unit completely shuts down and drops all the loads immediately. When this happens, all the LEDs are off in the front and no beep sounds. Just to make sure this is not just a fluke, we unplugged the utility power supply. Immediately all the systems powered by the unit were shut down. When plugged back in, all the LEDs came back to the aforementioned state. We have the network management card (NMC) in the unit but the logs aren't much of a help. The only thing we see is the cold start entries indicating the power up and power down status of the NMC.

**UPS: Restored the local network management interface-to-UPS communication.**
**System: Coldstart.**

As you can see the load on the unit is only 40%. So it's not overloaded. Going by the LEDs, the batteries are 100% charged. The status is always on-line. So what's going on with our unit? What could cause it to completely shutdown everything and itself as soon as the power supply is lost? Are the batteries bad even though it says 100% charged? Since this is a complete UPS shutdown, we are unable to get any meaningful logs.

Any help is greatly appreciated.

[Anonymous user]

The last self-test was done in 2007. So the unit couldn't let me know the batteries are going bad after the last successful test.

I just did a manual self-test. It took down the entire UPS. I am planning to reset the UPS (power off, push the "0" for 10 seconds and power on). Also, measure the voltage on the batteries. Do you know the required battery voltage (minimum) for the SU3000RUMXL series?

Thanks

[Anonymous user]

The last self-test was done in 2007. The schedule was disabled. So that's why I am thinking the unit never had a chance to test the battery and alert when it was going bad. Now I am assume, at least, one of the batteries is dead. What's the best way to measure the battery voltage? I see the voltage reading for the batteries in the NMC as 55. That seems to be bit low.

[voidstar_apc]

>
One of the batteries was changed six years ago. [...] Are the batteries bad even though it says 100% charged?
>

Batteries last three to five years. When dead, a fully charged battery will not support the load. If the battery isn't replaced, the UPS may drop the load while attempting to perform a self-test or when it reacts to a power anomaly.

[voidstar_apc]

>
The last self-test was done in 2007. So the unit couldn't let me know the batteries are going bad after the last successful test.
>

By default, the UPS does a weekly self-test, although it can be disabled through software or NMC.

>
I just did a manual self-test. It took down the entire UPS. I am planning to reset the UPS (power off, push the "0" for 10 seconds and power on). Also, measure the voltage on the batteries. Do you know the required battery voltage (minimum) for the SU3000RUMXL series?
>

I don't know. Perhaps someone else knows, like the tech support fellows who lurk around here.

[BillP]

The SU3000RMXL3U uses two of the [RBC27|http://www.apc.com/products/resource/include/techspec_index.cfm?base_sku=RBC27&tab=models] which the nominal string voltage is 48Vdc. Batteries typically last 3-5 years depending on operating temperature, charging cycles, etc. and as they age they lose their capacity to hold a charge, so the NMC may measure the correct floating charge but the batteries themselves may not be able to hold it's capacity to hodl the load compared to before.

[BillP]

Also The Su3000 had a problem where some resisters got cooked and the ups wont switch to battery for nothing

[Anonymous user]

I just moved everything off of this unit (zero load) and unplugged the power supply. Immediately the UPS was shutdown. Tried the unit reset by holding the "0" button while unplugged with no success.

Can I do other tests to confirm battery as the culprit? I don't want to buy two new batteries if the problem is with the unit itself. In that case, I should probably look into buying a whole new UPS.

Also, I thought the UPS itself has an internal battery. I can't imagine the internal and two external batteries all went bad but it is possible.

Thanks

[voidstar_apc]

>
Can I do other tests to confirm battery as the culprit? I don't want to buy two new batteries if the problem is with the unit itself. In that case, I should probably look into buying a whole new UPS.
>

The only official way is to test with known-good batteries, perhaps from another UPS.

Another forum poster here, KVAr, has a [test using a wall-wart|http://forums.apc.com/spaces/5/smart-ups-symmetra-lx-rm/forums/general/4334/smartups-3000-repairable...]. I've never done it myself, and it's used to diagnose UPSes whose batteries are so dead the UPS won't even turn on.

The good news is we expect all your batteries to be bad or going bad after six years, so that makes other failure modes less likely... the so-called [explaining away|http://www.eecs.qmul.ac.uk/~norman/BBNs/The_notion_of__explaining_away__evidence.htm] effect. Though it doesn't make it any less annoying if you buy new batteries and discover the UPS needs to be replaced.

[Anonymous user · Accepted solution]

The 55v battery voltage reading you are getting from the NMC is the charger voltage, the UPS hasnt identified a bad battery (1) because self test was disabled and (2) beause the cell blocks havn't completely open circuited causing a resistance change, which the smart UPS will usually identify. I am more than confident in diagnosing the issue as bad battery due to the age of the batteries and the lack of self testing. Replacing only one battery module in that style of unit is a complete waste of money, always replace all together, as uneven charging and discharging occurs, ruining the good one much earlier than expected, not to mention putting high current draws on the one module, the one good battery will also mask the bad ones and the UPS wont realise a resistance drop. Replacing them together will result in the end of life being together and the UPS diagnosing an overall battery resistance drop.

Via the NMC, reconfigure the unit to have no external battery (no of batteries 1) and completely disconnect the extended run frame via plugs on rear. Unplug all batteries from UPS (external and internal) and varify the battery fauly LED indicator (red) and beeping initiates.Tthis will confirm that the UPS still has its ability to diagnose a bad battery. Then plug each individual battery module into the internal battery slot of the unit (one at at time, one after another) and see if the bad battery alarm/led extinguishes. This is basically using the UPS as your battery tester, to identify / confirm all the packs are dead, all three batteries connected at once (even bad ones) may have just enough resistance across them to "trick" the UPS into thinking the battery is still good. If unplugging all batteries from the unit doesn't trigger an alarm then the unit is faulty and requires replacement. You could also just put a multimeter on the battery connectors (disconnected from ups) to check the block voltage, buth this still doesn't confirm their ability to hold a decent charge, ideally 48v is what you are looking for.

I have seen this many times before, what was your reasoning for disabled self test? and why dont you conduct regular runtime calibrations on the unit to varify battery operation? I'm surprised that you are surprised by this outcome, resultant from lack of maintenance and testing.

The other safe option would be to buy 3 new battery packs and if they dont work in this unit, then order another new unit without batteries and utilise the new batteries in the new unit. You can't really go wrong that way.