APC UPS 650

[NECRAT]

Has anyone ever just removed the Piezo Buzzer from one of these bad-boys. Does the UPS still work minus the buzzer?

I want to remove the buzzer and use the buzzer voltage as a logic trigger voltage, but without the schematic handy, I was curious if anyone else has done this.

[boiseengineer]

It won't care if the pizo's missing. Carefull though. "Ground" on the PC board where that is may not be AC ground. Some of those floated. Use an opto-isolator.

[kcbooboo]

I've removed the piezo noisemaker from SmartUPS units with no issue. I also removed one from a Liebert GTX-2 unit. They don't seem to care. I stow the removed component inside the ups by taping it down where it can't come loose, so the next owner can decide if it must be restored.

I think most APC UPSes have a DB9M on the back that will provide an indication when power has failed or the UPS is active.

Bob M.

[Baylink]

You don't say *which* 650 that is. If it's a SmartUPS, then, as the other commenter noted, it probably has an interface out of which it would be easier to get "on battery" and "full charge" and such more easily than reverse engineering the piezo driver circuit.

Though, admittedly, it might not be as much fun.

[NECRAT]

Done. Took me 10 minutes. Didnt have to remove the boards, just the battery. I removed it "dead" (disconnected, battery removed)

Also took me about 5 more to realize the battery is dead in it too. Oh well, for a free UPS, I can't complain about having to buy a new battery for it.

[NECRAT]

Well the damned thing shut itself down. No AC output at all, despite the fuse not being blown and the power switch being on.

And unfortunately, I am 200 miles away from it.

[kcbooboo]

I've found that after the APC units are about 5 years old, they start doing weird things. All you can do is replace the batteries, perhaps once. After that, it's almost better to just buy a new one.

Also, APCs might be the most popular units out there but they seem to have their share of problems. You can't fix them. The factory probably won't fix small ones either.

I've seen sites with desktop computers on APC battery-backup units of all shapes and sizes. When the power fails or glitches, some computers ride right through it, others reboot, and still others shut off and don't restart. Many rapid power interruptions and reconnections seem to go right through these units and into the loads. Yes, they do offer thousands of dollars in rewards, but who ever has the receipts?

Bob M.

[KC9AIC]

Well the damned thing shut itself down. No AC output at all, despite the fuse not being blown and the power switch being on.

And unfortunately, I am 200 miles away from it.

Is it 200 miles to the NNE of where you live? If so, that'd be right in my backyard, and I could run and go fix it. Or maybe you're 200 miles from home...

I have a number of mixed experiences with APC units. I've started buying Cyberpower UPS units as alternatives, as they seem (in my sample size of 3) to work as well as the APCs, at lower cost. Plus, they're readily available at the local Worst Buy.

[NECRAT]

I've found that after the APC units are about 5 years old, they start doing weird things. All you can do is replace the batteries, perhaps once. After that, it's almost better to just buy a new one.

Bob M.

At this point, that is the direction I am going to go.

[PID_Stop]

I stopped buying APC a few years ago: across our region (and especially here in the hub facility) we have -- well, had -- literally hundreds of SmartUPS units of various sizes, most rackmount, and about a dozen Matrix 5000s. We have found that after they approach 8-10 years (about halfway through their second battery) they start making a really nasty switch from commercial power to inverter, and back again... which is actually harder on some of our gear than brownouts in the raw commercial power. Also, many of the APCs have a nasty habit of overcharging their batteries, leading to the difficult chore of extracting swollen batteries without amputating fingers on the sharp sheet metal edges around the battery compartment.

We've switched over to the Liebert GXT2 (and now GXT3) series, in sizes ranging from 1500VA up to 10,000VA -- and the equipment they power, including an entire HD production control room and support racks, is rock solid. Unlike the APCs, these Lieberts are "on-line", meaning that the inverter always powers the load, and commercial power merely keeps the batteries charged. There's no switching, and unless you throw the unit into bypass mode, there's no direct connection between the commercial power and your equipment. It might be slightly less efficient than the APC line interactive types, but we don't see a noticeable difference in either power consumption or heat gain... but we no longer cringe when thunderstorms approach.

-- Jeff

[boiseengineer]

My various CyberPowers' at home and on some workstations at the office won't auto re-start after a long power failure & dead battery.
Found that out at a transmitter site. Had to go out there & press the power button to get the remote control back.
Plus they won't turn on unless there's good AC coming in even if the batteries are good.

[Baylink]

Yup; Lieberts are where you go from there.

Being continuous, they're what, 2-3 times as many $/kva as the standby ones, but it's worth it in a broadcast chain, particularly at remote sites.

[kcbooboo]

I tried a Liebert GTX-2 at a transmitter site with an old (>40 years) Onan 2-cylinder mechanical-governed generator whose frequency would vary about +/- 3 Hz with load (AM of course). It performed perfectly, riding through the switchover outage and continuously charging the battery while the connected equipment hummed merrily along. This unit auto-senses the AC line frequency when first plugged in (50/60 Hz) or it can be locked to either one, and the output voltage will match it. It doesn't care about the quality of the input voltage as it's just there to charge the batteries.

I also tried a TrippLite on-line model at the same site. It refused to charge the batteries at all when the generator was providing power. Apparently their models have a much tighter requirement on incoming power; if it's more than about 0.5 Hz away from 60 Hz, it thinks the power is of insufficient quality to use, so it just ignores it. Of course, 20 minutes later when the batteries die, the UPS has no choice but to either switch back to the unclean line power (egad), or just shut off completely. What good is either condition? That's why you bought a UPS in the first place.

I have a 20 year old Best Fortress standby UPS at home and it was rather picky about 60.5 Hz power coming from my own generator. A call to their tech support resolved the problem when they led me through a rather elaborate sequency of button pushes which seemed to desensitize the unit.

Seems to me the power quality necessary to charge the batteries in an on-line UPS is a moot point. How pure do you need the AC power to be, just to rectify it? Does the frequency really matter? Apparently some manufacturers think so, and they're the ones to avoid. Now I'm talking about the 0-1500VA units here; the larger units may operate differently.

I too have had to extract swollen batteries from APC SmartUPS units, usually after the 2nd set of batteries has been installed. I tink they count battery removals and go into "planned obsolescence" mode at that point.

True on-line UPSes are the way to go, but they do cost much more and all the ones I've run into have a continuously running fan in them, unlike the standby models that don't generate any heat unless they're running on battery for the 15-20 minute maximum run-time.

Bob M.

[PID_Stop]

The several comments about the cost differential between the "line interactive" versus "on line" UPSs spurred me to do some quick research, and I noticed several interesting things:

1) APC appears to have been bought out by Schnieder -- the same people who bought Square D. I have no idea what the practical effect of that change might be.

2) APC / Schneider now has two flavors of the SmartUPS: a traditional line-interactive series, and an on-line series that appears to be targeted directly at Liebert.

3) APC / Schneider seems to have learned a few things: I notice that many of their new models now have battery trays, instead of simply stuffing batteries into the finger-eating maw of sheet metal.

4) The price differential isn't as great as I had supposed: for a 1500VA model in a rack-mount profile (but without accessory rails), CDW gets $685 for a line-interactive APC. The newer on-line APC runs $1,107, and the Liebert GXT3 is $1,033. So the premium for an on-line unit seems to be on the order of 50%.

As a rule, the reason I deploy a UPS is because the cost to repair a piece of equipment, or the consequence of its downtime, is at least as high as the cost of the UPS. It's not only important that the UPS work dependably, but that it not itself cause the very sorts of problems we bought it to prevent. Our experience has been that in general, the APCs have a practical working life on the order of 8-10 years, and go through batteries every 2-5 years. Our oldest Lieberts are still in service after something like 15 years, and seem to get 4-6 years on a set of batteries. In fact, we have had only one Liebert fail (a casualty of a close lightning strike... but though it died, it protected the three $20K graphics generators). But we regularly have APCs spontaneously drop dead -- usually taking down power to whatever they were feeding. It's somewhat empirical, but I figure that the lower mortality rate and longer battery life means that we actually spend less overall on the Lieberts now.

-- Jeff

[Dale H. Cook]

I notice that many of their new models now have battery trays

Some of them already did, such as the two Smart-UPS 1400 units that I have in service. If the batteris swell up, though, the tray will usually jam, necessitating partial disassembly of the unit to change the batteries. I change mine every 3 years to prevent battery failure (I use Power-Sonic replacement gel cells), and I post the replacement date on each unit.