Discrepancies in readings on new APC BX1350M UPS

Teemsan

Distinguished
Dec 30, 2012
235
0
18,710
I just bought a new APC Pro BX1350M Uninterrupted Power Supply. I plugged it in, and cycled through the settings and everything seems to work.

When it's on and plugged in it reads the input and output voltage correctly as 123 volts. I have a voltmeter that actually reads 122.4 volts from the wall and 123.0 from the back of the unit, so that's all good.

It's max load is 810 watts. I put a load on it that read about 475 watts, and then I pulled the plug to simulate an outage. All the plugged in devices remained on, and it showed 7 minutes before shutdown time, which is about what I expected.

But unplugging the machine is where I noticed a potential problem.

When i unplugged it the wattage showed a drop to an output of 116 volts, which is not too bad, but when I measured with my voltmeter the output actually showed between 108.0 to 108.3 volts. 6 volts lower than what it was displaying, and a full 15 volts lower than plugged in. I thought it might be an anomaly so I tried it on each of the 5 outlets and all read around 108 volts when the unit is unplugged. I removed the load and tried it unplugged and it's the same - it shows an output of 116 volts but reads 108 on my voltmeter.

This is the first UPS I've owned. Is that margin of discrepancy acceptable for this kind of UPS?
 
Solution
The discrepancy between the UPS display and your meter is probably due to measuring the voltage at different points in the circuit using different measurement technologies.

You'd have to have a schematic of the UPS to determine where in the circuit it is doing its measurement. And then look at the accuracy tolerance of the measurement IC in the UPS vs your meter.

Check the UPS again after it is fully recharged.
 

Teemsan

Distinguished
Dec 30, 2012
235
0
18,710


I checked after it recharged again and it's the same - load or no load, when unplugged it shows 116 volts and reads 108 off the meter. I've tried two voltmeters even though I knew both meters are fine and reading to within a couple tenths of each other.

What you are saying makes sense, but when plugged in everything is in agreement. Input and output voltages show 123 on the unit and the voltmeter reads that from the unit and the wall. So if the accuracy of the readings were dependant on where in the circuit it's measured, wouldn't that accuracy tolerance be the same whether the unit is plugged in or unplugged? Or could it be different for each case?


 

No, it won't be the same plugged in or not. Plugged in, there is just a straight pass-through. Whatever is coming from the wall is what you'll get at the output. Unplugged, however, it will be running from the battery so it's a completely different power source. And the DC power from the battery must be run through an inverter to get AC power so at best it would be 120V.

The standard 120V voltage tolerance in the US and Canada is 114V to 126V. So you are within tolerances as shown by the UPS but 108V is out of spec but probably good enough to keep the machine running for some minutes to give you time for a proper shutdown.

IMO, the real test is whether or not it keeps the computer running long enough to do a proper shutdown - that is it's only function really. You're not going to be running from the battery for very long in any case.

I just tested my APC UPS and it goes down to 109V which is out of spec also but it has worked flawlessly for over 2 years.

BTW, you need to discharge the battery occasionally to prevent it from going bad. I boot up, close all running programs, and turn the power to the UPS off and let it run down to 60-75% and then turn the wall power back on about once a month. That also lets me know if the battery is going bad. If the battery isn't sufficiently charged the computer will shut down when I turn the wall power off. Doing this will help the battery last longer. It's the same thing as the battery in a car going bad after letting it sit too long. Starting the car (putting a load on the battery) prevents the battery going bad.
 
Solution

Teemsan

Distinguished
Dec 30, 2012
235
0
18,710


Thanks for the info and letting me know that yours reads 109V. Mine does work fine also, and provides the stated runtime. Thanks also for the tip on discharging the battery occasionally. Good to know.

Do you use the PowerChute software? If so, which Configuration > Runtime setting do you use? The Runtime setting options are too limited as I see it. I'm using PowerChute Personal 3.0.2. For Runtime settings it allows:
- Shutdown my computer when it has been on battery backup for > 1 min to 5 min are the choices
Or
- Shut down my computer when the time left on battery backup power is > 5 min to 8 min are the choices

Those settings are fine when you're away from the computer or you leave it on overnight. But if I'm at a load of even 60 to 65% and it goes to battery it shows about 7 minutes runtime remaining. So if I chose the max setting of 5 minutes for "Shut down my computer when the time left on battery backup power shows 5 minutes", then I have potentially 90 seconds to two minutes to shut down. At eight minutes it would shut down immediately because the 7 showing is already below that threshold. Even if they extended it to something like 3 minutes. That would provide 4 minutes to save what you're doing if necessary; especially helpful if it involves saving something online, then closing out your programs.

Despite the caveats I think it's a really nice unit.
 
I don't use PowerChute. I'm running Windows 7 and its power settings are good enough for me. I don't leave my machine running overnight and when I'm leaving for any time I close all applications and restart them when I return. We do have occasional power outages but I'm usually at home when the power goes off so a couple of minutes is all I need to shut it down. On occasion I'll manually shut it down if he have severe thunderstorms in the area just to be safe.