[TriLUG] OT: Diagnosting APC Back-UPS Misbehavior

Pete Soper pete at soper.us
Wed Mar 6 17:18:19 EST 2013


Are the new batteries charged? What's the voltage when they're in the 
unit? With no load it wouldn't be unusual for them to put out a 
believable voltage.

Lead acid batteries are basically wearing out when they're not 
discharging and don't have a minimum "float charge" voltage on them. In 
fact this process can happen even when the charger is connected but is 
providing a wee skoshin too little voltage, as the voltage that arrests 
the chemical reaction is very definite and temperature sensitive. (The 
proper float charge voltage is also different for gel cell batteries vs 
typical car batteries.) That is, left by themselves with no charger in 
place lead-acid batteries are both self-discharging and developing 
sulphate coatings on the electrodes that degrade the capacity of the 
battery and will eventually kill it.

Is it possible you got the extra specially discounted batteries that sat 
in a warehouse and that are DOA? It sure seems unlikely both UPS units 
would have suffered a mysterious failure and coincidentally misbehaved 
with the "new" batteries. If the UPS units have absolute crap capacitors 
that's the one mechanism I know of that could cause them both to have 
died. But 8-9 years isn't that long with modern capacitors. I would 
think a bit of searching would indicate whether there is a problem with 
the circuitry for these units.

If you can borrow a standalone charger it may give you a definite 
indication the batteries are dead or never indicate they're charged no 
matter how long they're connected. That's what I've used to discriminate 
between dead batteries vs dead UPS units. I personally haven't found a 
dead UPS unit yet.

-Pete

On 03/06/2013 03:49 PM, Scott Chilcote wrote:
> Hi Folks,
>
> Sorry to wander OT, but I've seen good advice on this list related
> to UPS problems and I'm hoping someone has some insight.
>
> I have two APC Back-UPS XS 800 units that I took offline last year when
> they showed the "Replace Battery" light.  I ordered new batteries last
> week, and put them in one of the units.  It did not come back to life as
> expected.  It showed the "On Battery" light instead, and started beeping
> four times.  This is what they do when the power goes off.
>
> I tried resetting the unit (disconnecting the battery and the AC plug,
> holding down the "On" switch for 60 seconds).  No change.
>
> Then I tried setting the unit's sensitivity to the lowest setting.  This
> caused it to power up normally (green light).  However, it soon started
> making a relay chattering sound, which is quite loud.  I shut it off.
>
> Since I had another Back-UPS XS 800 that needed batteries, I moved the
> batteries over to that unit.  I even changed out the cable and spacer
> for the battery pair.  When I plugged it in - lo and behold - I had the
> identical misbehavior.  Same "On Battery" state.  Same relay chatter
> with the sensitivity set to low.
>
> Next step: Read the manual.  It says to check the AC voltage level in
> the house.  I tried the UPS on several different outlets around the
> house, same problem.  I measured the AC voltage, and got 113VAC on my
> meter.  We have four other UPSes of varying kinds, and they are working
> fine at the default sensitivity level.  The manual also says to
> check the circuit breaker on the back of the unit.  I did that
> earlier and found that neither one had tripped.
>
> I checked the voltage on both of the new batteries (disconnected).  They
> make the meter swing to at least 12 VDC.
>
> These UPSes were both made in late 2004 according to their serial
> numbers.  Did they die at the same time, while unplugged?  Do they count
> the number of battery replacements and self destruct (humor)?
>
> I spent some time looking at APC's forums, this does not appear to be a
> frequent problem.  I tried their online support chat, but the trees
> outside my window were visibly growing while I waited (both of their
> level ones may be off this week).
>
> Thanks for any cluefulness!
>
>      Scott C.
>




More information about the TriLUG mailing list