SU2200INET trim LED always on after battery replacement

[Anonymous user]

I have two SU2200INET SmartUPSs which have slightly different firmware.
* The test UPS has serial number YS0321110227, firmware 100.14.M
* The control UPS (where nothing is wrong) has serial number YS0336222267, firmware 80.14.I
The test UPS's batteries were over seven years old and failed suddenly. They overheated and warped slightly, and a very small amount of battery gel escaped and corroded the terminal bolts. The batteries were replaced by an authorized APC service company.

The problem: Ever since the battery replacement, the amber "trim" LED is permanently on.

To our knowledge nothing has changed about utility power. The control UPS is behaving fine.

Apart from this LED behavior and low output voltage, the test UPS is doing fine, including self-tests.

h2. Test UPS

Upper transfer voltage HITRANSFER 239 229 234 239 224
Lower transfer voltage LOTRANSFER 177 177 172 168 182
Return threshold RETURNCHARGE 15 00 15 50 90
Output voltage on batts OUTPUTVOLTS 208 208
{color:#ff0000}MODEL : SMART-UPS *1400*{color}
STATUS : TRIM ONLINE
LINEV : 231.4 Volts
MAXLINEV : 232.7 Volts
{color:#ff0000}OUTPUTV : 205.4 Volts{color}
SENSE : Low
LOTRANS : 177.0 Volts
HITRANS : 239.0 Volts

h2. Control UPS

Upper transfer voltage HITRANSFER 253 253 264 271 280
Lower transfer voltage LOTRANSFER 196 196 188 208 204
Return threshold RETURNCHARGE 0 00 15 50 90
Output voltage on batts OUTPUTVOLTS 230 230 240 220 225
MODEL : SMART-UPS 2200
STATUS : ONLINE
LINEV : 231.4 Volts
MAXLINEV : 232.7 Volts
OUTPUTV : 231.4 Volts
SENSE : High
LOTRANS : 196.0 Volts
HITRANS : 253.0 Volts


Questions:
* Why is the UPS model wrong on the test UPS?
* Why are the programmable values for low and high transfers different?
* Is there something we can do to make the test UPS behave like the control UPS?
Thanks for any advice.

Message was edited by: soumen@cse.iitb.ac.in

[Anonymous user]

Was the printed circuit board maybe bumped when the batteries were installed?

Very unlikely. The terminals and fuses on the kaput batteries were slightly corroded. No gel had contacted any part of the UPS. The old batteries were removed easily. The technician cleaned the terminal lugs and used new bolts and fuses on the new batteries.

Was someone issuing commands to the UPS via the console cable?

Yes. The technician said he reset the battery runtime estimate so that the UPS can calibrate it again. After recalibration over the next few hours, the number reported was very plausible given the load.

Supposing the console commands messed up something else in EPROM (such as telling the UPS it is a US model), how can I fix it?

Thanks a lot.

[Anonymous user]

Turns out the technician had accidentally changed the model code while resetting the battery calibration. I managed to set things right using information on this link:

http://osdir.com/ml/monitoring.nut.user/2007-01/msg00120.html

The steps are (voids warranties!) ---

Open a serial console to the UPS.
Press *1*, wait 2 seconds, press *1* again.
You should see response PROG.
Now press b.
You will see some firmware code.
Keep pressing +* followed by a 1 second pause followed by b until you see the correct firmware code.
That was *80.14.I* for my UPS.
Finally press R to get back to dumb mode.
You should see response BYE or some such.
Quit the console without resetting the serial port in any way.

Once again, do the above at your own risk. No warranties, no indemnity.

[Javik_apc]

Well,. that was a very interesting read. Sort of like learning how to make a phone "blue box", I guess.

So you can make one UPS act like another by merely changing the model version, and the circuitry actually has the capability of complying with these code changes? Hmm, I suppose I could save some money by buying a 700 and making it think it is a 3000, and tacking on a few marine batteries.. 😉

Next questions:

* Is it bad form to link to an article on hacking the Smart Signalling protocol, on APC's own website? Might as well post jailbreak articles on Apple's forums.

* Why would this "tech" have needed to fiddle with such deep voodoo to reset the battery replacement date? You don't need to go to such extremes for this. Both the normal serial console and web console would work fine for it.

[Anonymous user · Accepted solution]

The technician did not (just) modify the battery replacement date. As a battery ages, the UPS changes its estimate of the drain and charge capacities and rates. When the battery is at the end of its service life, the UPS has a fairly poor impression of it. When you connect new batteries, unless some of these stale calibrations are reset, the UPS continues to have a very pessimistic view of the new batteries, even after a runtime calibration. I have witnessed change of batteries for a dozen APC UPSs in my organization, and in each case the technician used a serial console to tell the UPS that it has brand new batteries. Then they charge the new battery to 100% and finally conduct a drain runtime calibration.

I don't see why techniques to fix problems in out-of-warranty APC UPSs cannot be posted on an APC Website. I would very strongly advise anyone against trying to use the same technique to change their model number from the correct one to an incorrect one. There is a grave risk of fire and battery meltdown. I am lucky the mistake made by our technician did not fry our UPS in the first place. So in principle, my postings could prevent a fire, save property and even lives.