Originally Posted by MightEMatt
Is it line interactive? I can't seem to access the powercom website to get any specs. The narrowest range I can set on my UPS for under and overvoltage is 88V to 136V. This means that if the brownout is still within that value, the UPS will not do anything. It wouldn't be unreasonable for a PSU to trip off at 90V for safety. Try pulling the plug completely to see if the computer stays on, this could at least narrow your investigation.
Yes, it is line interactive. Powercom site is indeed down (and was extremely outdated even several years ago, so I'm not surprised).
I have no idea what the range is, but in the past, it would catch brownouts with 100% consistency and there was never an issue. I can try pulling it from the wall as you said, but that's a pain, both physically and due to the huge RAID rebuild times.
Originally Posted by Pawelr98
Basing on reviews the battery lifespan is rather short in this UPS.
It uses 4x12V battery set ?
If you want to test them under load then use a car light bulb(standard 55W) or something similar.
Check for voltage dips on each battery.
Theory wise you could also upgrade the PSU to be able to withstand short voltage dips.
Larger primary side capacitors should do the trick providing that there's no protective circuitry which monitors AC mains input voltage.
The only UPS which is 100% sure is an online UPS.
AC mains into DC, battery in the middle and then conversion back to AC.
It's inefficient but provides a stable AC output which doesn't change no matter what happens on grid AC.
You could convert your current UPS into "online" by providing the power to batteries instead of UPS AC input.
The main problem is that an offline UPS in not really rated for continuous usage (transformer and inverter will overheat), if the load is a fraction of rated power(say 50%) then it should be fine.
yes, 4 12v batteries. As stated above it always handled voltage dips just fine in the past. In fact, there was a period when the infrastructure in this area was so bad that it was boosting the voltage for at least half the day, every day, and also occasionally bucking it as well. It has a beefy cooling fan which keeps the unit itself from overheating under those conditions... unfortunately it overheated all the batteries to the point that they swelled up and couldnt be removed from the unit without sawing off a tiny bit of the chassis, but thats another story.
It seems like the best thing I could do is try to test the batteries under load. My system typically draws about 600w. Would I need a similar load to test the batteries, or would a smaller load like a lightbulb as you suggested still show meaningful results?