APC NS 40 reports wrong load percentage

[Anonymous user]

Last week i purchased APC BackUPS NS 40 to replace 8 years old APC UPS that has failed. There are three devices on the UPS: cable modem, unmanaged 8 port gigabit switch and linux machine that acts as firewall and a router between our LAN and the rest of the world.

These three devices together draw between 49 and 67 watts of power depending on the network traffic, but the UPS reports only 5% load and 67 minutes of battery runtime remaining. According to the specs sheet for the UPS, the output power capacity is 257W/450VA, so i'd expect the load to be roughly between 19% and 27% and the battery runtime to be at most 16 minutes. I have double checked that the devices are indeed plugged into the UPS, and that they are plugged into the power backup side of it. I have also put load on the linux machine (recompiling kernel) to see if the load percentage changes, and it did not.

While this is not a critical problem, i'd like to get it sorted out eventually. If anyone has any ideas on how to fix this, I'd appreciate them.

Thanks.

I'm including the output of 'apcaccess status' from the linux machine:

APC      : 001,034,0839
DATE     : 2015-07-06 11:03:34 -0400
HOSTNAME : xxxxxxxxxxxxxx
VERSION  : 3.1
UPSNAME  : Back-UPS NS 4001
CABLE    : USB Cable
DRIVER   : USB UPS Driver
UPSMODE  : Stand Alone
STARTTIME: 2015-07-06 10:22:53 -0400
MODEL    : Back-UPS NS 4001
STATUS   : ONLINE
LINEV    : 121.0 Volts
LOADPCT  : 5.0 Percent
BCHARGE  : 101.0 Percent
TIMELEFT : 67.7 Minutes
MBATTCHG : 10 Percent
MINTIMEL : 3 Minutes
MAXTIME  : 0 Seconds
SENSE    : High
LOTRANS  : 92.0 Volts
HITRANS  : 139.0 Volts
ALARMDEL : 30 Seconds
BATTV    : 13.4 Volts
LASTXFER : No transfers since turnon
NUMXFERS : 0
TONBATT  : 0 Seconds
CUMONBATT: 0 Seconds
XOFFBATT : N/A
STATFLAG : 0x05000008
SERIALNO : 4B1507P56665
BATTDATE : 2015-02-15
NOMINV   : 120 Volts
NOMBATTV : 12.0 Volts
FIRMWARE : 903.K3 .D USB FW:K3
END APC  : 2015-07-06 11:03:39 -0400

[BillP]

Just curious - what is telling you the load is 49-67 watts that you're comparing against?

I was also thinking, can you pull the plug on the UPS and see if the reading changes while it is on battery, for kicks?

[Anonymous user]

On 7/6/2015 2:55 PM, Angela said:

Just curious - what is telling you the load is 49-67 watts that you're comparing against?

I used kill-a-watt to measure the wattage. About 6 months ago when I upgraded the linux machine from P3 to G645, I was curious about how much the power usage would drop, and the old P3 was 100W higher than the G645. That's a nice chunk in savings, that machine is always running.

On 7/6/2015 2:55 PM, Angela said:

I was also thinking, can you pull the plug on the UPS and see if the reading changes while it is on battery, for kicks?

That was a good catch, the UPS seems to only report the load correctly when on battery power, after I plugged it back in, it just stays at whatever percentage it was at when the power was out.

[BillP]

Ahh, ok. I asked for you to pull the plug because this type of UPS (it is similar to our Back-UPS ES) uses a standby topology. Someone told/reminded me a little bit ago these UPSs can only detect an overload while on battery power due to the Standby topology design. Thus, I figured that means it can only accurate detect the load %/W while on battery power.

(Here is a white paper on UPS systems if you're interested in further detail regarding standby design -> http://www.apc.com/whitepaper/?wp=1 )

[Anonymous user]

Thanks, makes sense, i didn't even consider active vs passive UPS when i bought the darn thing, guess i'll have to put this one on less important equipment and go get a SmartUPS instead.

[BillP]

I think that Line Interactive topology units would be an option for you. We have line interactive design Back-UPS like Back-UPS XS or RS/Pro. Or, next higher up would be Smart-UPS SMC depending on what your other requirements are.

[Anonymous user · Accepted solution]

Thank you, I will read up on them. And thank you for Your help. 🙂