InspectorGadget

I have an APC SmartUPS SU1400XLRM-NET 5 1/2 years old. The batteries have gotten old and it started to die during a daily power dip (from a car dealer starting up a big compressor next door).

I replaced the battery, but the PowerChute Business Basic status panel didn't change, even when I forced a change in the "battery replacement date". Showed the same old status, same old voltage.

So I thought I'd reboot the UPS, and that's when the real trouble started. When I plugged it back in, it started beeping once a second (batteries or no batteries, load or no load) with no status lights. I can press the Off/"0" button to stop this. Then, when I press the On/"1" button, it starts its selftest. The "on line" light flashes and the lowest battery light is lit steadily for about 10-15 seconds. Then, it seems to try to go to battery power -- the "on line" light goes off, "on battery" light flashes for an instant, there's a click, the ouput power goes off, and the lowest battery light flashes.

If self-test fails, or the battery is too low, the user manual says this thing should just remain in on-line mode and act as a big on-off switch. So what's the deal?

I can call APC on Monday; I'm hoping someone on this forum might have an idea this weekend. My network servers are now unprotected. Thank you in advance.

- The Inspector

dai

did you give the batteries long enough to charge before using it

InspectorGadget

Thanks for your response.

Well, as I said earlier, it's not supposed to matter. If the batteries don't work or aren't there, it's supposed to go to line voltage (assuming it's present, which it is) and just light up the "bad battery" light. It even says in the manual that even without batteries (or with bad batteries), you can use the thing as a big on-off switch.

That being said, they were brand-new batteries with already a 27V terminal voltage on each pack. They've been sitting in the machine plugged into the wall for nearly two days now.

Behavior is the same. Incidentally, the "bad battery" light never comes on. Just the lowest battery-charge-level light blinks.

Ultimately, tomorrow morning, I think APC is just going to say, "uh, it's broke." I know the behavior means something and I'm trying to find out what it is. They try to railroad you into expensive UPS upgrades on EVERY phone call. I'm afraid that's all I'm going to get.

- DeviceNode

JamesO

Depending you where you are located, this ebay link may be of interest.

http://stores.ebay.com/InfoLaunch-Ba...eNameZl2QQtZkm

I have friends who have purchased from this guy and have been very happy.

He may even give some tech support?

JamesO

Volt-Schwibe

nevermind, what i was trying to say didn't apply.

InspectorGadget

FYI: I just got off the phone with APC. Their conclusion: "Internal Fault" In plain language, "It's broke."

I could trade up to the new SU1400 for $615, but it doesn't support my two giant SU48 battery packs. The cheapest upgrade that supports those external battery packs is $989.

AAAAARRRGH!

- DeviceNode

JamesO

Contact the guy on ebay and see what he may have on hand.

http://stores.ebay.com/InfoLaunch-Ba...eNameZl2QQtZkm

JamesO

batty_professor

Can you indicate the specs from the battery labels? or a link thereto? How 'bout a link to a .pdf of the manual for the unit? You indicate the battery voltage measures 27 volts? The unit must detect the battery voltage at a certain minimum voltage, or it's NO-GO. If they're labeled at 48 volts for instance, that 27 volts isn't enough and it's going to act just as you indicate, and will not even so much as trickle the batteries! We should be able to recover from this. Please advise.

InspectorGadget

I measured 27 volts on one pack. Each pack is made up of two standard 12V lead-acid batteries in series. There are two packs in the UPS, presumably also in series. 24V nominal x 2 = "48V". Sorry I wasn't clear about that.

All the old battery packs I've pulled out (from the main unit and from the external battery pack units) read 25-26V on the meter. So I don't know where PowerChute gets 27.02V. It keeps saying 27.02V even when all the packs are unplugged and removed. POS.

A no-go battery is supposed to light up an orange "Bad Battery" light with a battery and an "X" through it symbol. This never lit up. I doubt any of the batteries were "bad" per se... probably had much less capacity than when new but still enough to power the thing for a while.

The thing is definitely "broke," but the interesting question is how. That flashing lowest-battery-level light MEANS something, but they reference NOTHING in their documentation and the tech support guys were like, "Internal Fault, dude, no other choice but to fork over another $1000 and "upgrade" so we can do this to you all over again."

I have it nearly pulled apart now. I'm going to look for the obvious -- blown secret internal fuse that "never blows," self-disemboweled electrolytic caps, baked resistors or other charred giveaways, etc. Barring anything glaring like that, then I'm going need more intelligence about the thing.

That refusal to go into battery mode, complete shutdown of the output & refusal to go back into online mode, and the blinking bottom battery light are a DEFINITE SYMPTOM OF SOMETHING DETERMINISTIC IN THE FIRMWARE. I'm screwed by the typical tech-support shuffle. We live in a confederacy of dunces.

User Manuals: I can't get a direct link, but you can easily go to http://www.apcc.com/tools/mytools/in...&category=user, and then type in "SU1400RMXL".

RichardO: Thanks for the tip. Most of what he has are 24V units, but there were a couple of nice 2200VA RM units that may work. Shipping is $70 but it still may be a better way (and/or the only way) to go. I'm going to see if I can find something repairable in with a couple of hours of work before I give up. All this was unbudgeted and we're really in tight cash flow right now.

I really appreciate the effort that people are putting in trying to help. I'll continue to supply more information if you request it, and if I figure out anything I'll post it at least for posterity.

- The Inspector

InspectorGadget

Remember also that when you plug it in, the UPS beeps and clicks its relays at a regular 2Hz rate. Batteries or no batteries. Something's obviously wrong; it's just a matter of whether its salvageable or not.

If I had a circuit diagram I could fix it.

- The Inspector

JamesO

Do not forget to check for bad solder joints.

I find when something works fine for years and then starts acting up, it can usually be a bad solder joint.

Following is a list if items I have sorted bad solder joints on recently:
House gas fired power vent hot water heater controller, really did not want to spend $200 for a new controller on an 8 year old water heater that I plan to replace in the next few years anyway.
Fax machine, worked fine until I moved. Cold solder joints acted up after unit was moved.
Many audio components, DBX units, audio amps, processors.
5 year old factory Ford CD/Cassette unit with intermittent dimming, missing display (fixed this Sunday night!) Searched info on the net, pulled unit, re-soldered and re-installed in less than 1 hour, problem solved!
Circuit boards for pinball machines
All kinds of kid’s battery powered toys made in China. Talk about poor soldering work, these items are the Prize Winners!

And many other items I cannot think of at this moment.

Have you moved your gear lately? This is usually a good indication that you might have a bad solder joint somewhere.

When you open the unit, look around relays and power resistors on the circuit board for heat discoloration. Also look closely at the bulkhead connectors on the circuit boards. I usually find the bulkhead connectors are prime suspects for poor solder joints if I do not find any heat discoloration.

I usually find cold solder joints are for 2 main reasons:

1. Boards are now mostly surface mount, but larger components and connectors are still through hole design. The larger components need more heat to get a good solder joint as they tend to act as a heat sinks with too much mass for the wave soldering or IR machines. I almost always retouch all through hole items on mixed surface mount/through hole boards.

2. Inspect around all larger power resistors carefully. Many times they get very hot and transfer the heat to the solder joints, circuit board and surrounding components. The heating and cooling cause’s board expansion and contraction which can cause surface mount items in the near area to break solder joints. You will usually find power resistors in the power supply section, battery charging sections and sometime around relays.

I would also look for fuses as well.

Get a good light and magnifying glass, but when it doubt retouch everything if it is not too time consuming!

Good luck!

JamesO

InspectorGadget

Wow, JamesO, great advice. I had forgotten about cold solder joints. Yes, I've found a few in my time as well. As you suggest, I'll just go over it and retouch any joint that looks at all suspicious.

I haven't moved it lately, but I have had a lot of heat fluctuations. It got really hot in here last summer (had a hard disk fail) and then really cool in the winter. I can't afford to run A/C all the time; I'd have a $1000/mo electric bill.

The other thing to mention is that there is a power-dip event that happens every morning around (though not exactly at) the same time. I believe it's one of the car dealers near me starting up some huge piece of equipment. All the UPS's in my place fire off (I have a few other BackUPS on workstations in other rooms). My poor SU1400RMXL was "tested" every single day for the last year, sometimes twice a day. That had to have some stress on the system.

I've had PG&E (our local corrupt power company -- remember the FAKE CA energy crisis a few years ago?) out here a few times to document the power events. They left recording voltmeters on the line but claim to have never found anything. Sheesh, the first one they put on had a resolution of about 10 seconds! Then they came back and said they were putting on a "really high resolution" recording unit. It turned out to have a resolution of 100 ms! Many power events have durations in the microseconds. So they never found anything, and the culprit is still conducting their "evil" practices.

As for fuses -- I already found a couple of secret, internal "they never blow" fuses, soldered to the main board of all things. Unfortunately, they weren't the problem; they test out good.

I'll let you know what happens with the cold-solder-joint rework. It'll probably take me a few days -- it's going to be a big job to completely extract the PCB. It's got a ton of connectors and big wires bolted to big heat sink fans; I'm going to have to be very careful about extraction and documentation of identical connectors.

Thanks again.

- The Inspector

JamesO

Get out the digital camera!!!

Hey, this may be worth a look. Radio Shack has a PC controlled VOM that is fairly cheap, on sale $59.99. I have one, it works pretty good to log to a computer, the bad part is the mininum sample times is 1 second.

http://www.radioshack.com/product.as...%5Fid=22%2D812

JamesO

batty_professor

I'm leaning toward a corroded battery connector somewhere. We are afterall dealing with lead-acid batteries here. And though they may be sealed, they still may have "gassed" at some time. That's corrosive. Do the batteries all connect to one central pair of terminals on the mainboard? I'd pay close attention to that area too. Do you have an analog meter available to you? and can you use such to watch the voltage at that point when attempting startup/selftest? I would say "careful", but you sound confident inside this thing.