Help with APC SmartUPS 1500

[signal]

I have a SmartUPS 1500.  I just put new batteries in it.  There is a lot of corrosion around the battery compartment.  I put the new batteries in, they charged up.  The UPS acts like its working ok.  But when I kill the power input either by unplugging from the wall or running a self test, it shows as if it drains battery real quick (the Battery Charge LED's drop quickly) and it powers down, because it thinks its out of juice.  But actually moving that battery into another UPS of the same model, shows the battery still shows fully charged.  So the batterys are actually fine.  I even swapped in some batteries from the other UPS (same model) and the same behavior is exhibited.  I did a recalibration on the UPS, and tried a "reset" which APC tech support had me do (unplug from wall, power off UPS, unplug battery from UPS, and then plug it all back in).  Still same behavior.

My initial question is how can I safely work on this?  I took the lid off, and put a meter to the two large caps, they showed about 8.47V each.  I was hoping the meter would discharge the caps but it just held 8.47V.  So I used a screwdriver and there were a lot of sparks from the one cap, not much from the other, now they show 0V.  There are some smaller caps, I ran the screwdriver over those and didn't see anything and got no meter readings on them.  Then there are some other large devices like transformer which I assume is likely not to worry about.  I want to disconnect some of the boards to have a look.

[bitseeker]

It's good that you discharge the capacitors before poking around on the PCB. However, don't use a screwdriver like that. It stresses the caps with the sudden discharge (high current) and damages your screwdriver. Make yourself a capacitor discharge tool (Google it or YouTube it). It's basically a power resistor with a lead on each end. The resistor reduces the current flow to avoid the huge sparks -- it just takes a few seconds to discharge as a result, rather than instantly.

Since the same problem is exhibited with known good batteries, do some continuity and/or resistance checks from the ends of the cables that attach to the battery, back toward the PCB and onward. It sounds like there may be some connections with high corrosion that aren't allowing sufficient current flow. You should be able to see it as resistance on the DMM.

[signal]

bitskeer, thanks for the tips.  Looking at the board, nothing looks amiss.  There is no smells or any visual signs of something cooked.  There was the corrosion in the battery area, and as I said these are now fresh batteries.  I looked at the cable like you said.  I measured end to end and both sides measure about .2 Ohms.  However upon visual inspection, it looks like the negative lead has corrosion on the terminal that would plug into the battery harness. You can see in the pictures attached.  I assume I can probably just pop that cable out of the connector and clean it up with some brush and IPA, does that sound right?  I can't seem to get it out of the connector, I need to figure out how.

[wraper]

https://conetrix.com/Blog/manually-esetting-an-apc-smart-ups-battery-constant-after-new-battery-replacement

[bitseeker]

However upon visual inspection, it looks like the negative lead has corrosion on the terminal that would plug into the battery harness. You can see in the pictures attached.  I assume I can probably just pop that cable out of the connector and clean it up with some brush and IPA, does that sound right?  I can't seem to get it out of the connector, I need to figure out how.

It doesn't look like a lot (not enough to cause the issue you describe), but doesn't hurt to clean it. Then, try resetting according to the blog post that wraper linked.

[signal]

I did figure how to get the wires out of the connector from this post http://www.ebay.com/gds/How-To-Remove-Anderson-Plug-Contacts-/10000000178442598/g.html

Turns out the wires look fine, the negative is cast in a more silver color and the red a more gold and so thats what I was seeing as a difference.

I will try the reset procedure and report back, thanks everyone, much appreciated.

[signal]

Unfortunately in putting it back together, i must have shorted something or done something wrong, because now the display shows all lights on (like a christmas tree).  It doesn't have a wiring fault, and appears to be working as it did before, but I may have pinched the ribbon cable or something because all the lights are on 

[signal]

I wonder if I fried the LED display driver
http://www.datasheets360.com/part/detail/mbi5026cd/3982016178066627845/?alternatePartManufacturerId=0

The cable going from the display to the main board looks fine.