Data cable for Back-UPS 550?

[Anonymous user]

At one of the water districts where I am Operator I found the desktop computer plugged into a very dusty Back-UPS 550.  The yellow sticker was still over the UPS outlets, the computer was plugged into the surge protected outlets, and I have to suspect the battery was never connected.  If possible I want to remedy that situation, and if we are lucky all I have to do is connect the battery and plug things into the right outlets.  At this point I do not know whether the data cable is stowed away somewhere in the office, or has gone missing.  If the latter I will want to replace it, so question arising, is that a standard USB A male to USB B male, or is it something special?

The data cable question is of course related to whether the current version of PowerChute Personal Edition will work with that not very new unit.  Thanks to Bill, the PCPE on my own system is running perfectly and of course I am wondering whether the same trick will work, if necessary, on this other installation.  The computer in that water office is an HP 20 all-in-one, on which I am about to perform a memory upgrade, running Windows 10 version 1703, the Creators Update.  Nobody in that office had noticed that the UPS was never properly set up . . .

Thanks for enlightenment,

[Anonymous user]

At a very minimum, you'll need to make an investment in a new battery. Lead acid batteries don't do well in storage. Even if it was hooked up, with as much dust as you indicate being present, I wouldn't trust it to be any good any longer. (The average lifetime in regular use is about 3-5 years.)

If you look at the model/serial number tag on the unit, that can tell you when it was manufactured. The date (year/week format) is present within the first part of the serial number, after any initial letters.

As far as communications go, most of these models utilize a proprietary cable with a USB type A connection at one end and an RJ-45 male connector at the other. I don't know if APC sells this cable as  separate replacement part, but they are plentiful and cheap on the secondhand market.

PowerChute Personal Edition will support this UPS, as will certain other third party monitoring tools (like apcupsd).

[Anonymous user]

I'll check out the serial number and year of manufacture later today, when I'm at the water office in question.  I'm aware that lead-acid batteries do not prosper when stored unused.

The data cable that came with my own recently installed Back-UPS BN900M (s/n 4B1643P23365, thus manufactured just less than one year ago) is USB type A at the computer end, and USB type B at the UPS end.  The data cable receptacle on the Back-UPS 550 in that water office is also USB type B (rather than RJ-45).  Unless the pin-out is non-standard, that type of cable is readily available and cheap (usually used to connect printers).  Given that a standard (non-super-speed) USB cable has only 4 conductors, I don't see much room or reason for alternative pin-outs?

As you are no doubt aware, there has been quite a bit of online discussion (some of it heated) about the proprietary cables with RJ45 (or RJ50) terminations.  This Back-UPS 550 has, however, a data cable receptacle (marked for PowerChute, no less) of USB type B format.

[Anonymous user]

Update to preceding: the battery turns out to have been connected, so it has presumably been kept on full charge throughout.

The unit is model BE550G, s/n 4B1234P32260, so just over 5 years old.  Can't imagine why they left the yellow label pasted over the battery backup outlets, using only the surge protected outlets.  I may have time to fiddle with it later today, but not right now.

[Anonymous user]

Another update: I bought a standard 6 foot USB-2.0 cable, USB-A at one end, USB-B at the other end, and installed it between the computer and the APC  Back-UPS  550.  The computer thereupon went through the usual "installing device" routine and then displayed the usual expandable battery charge status icon in the system tray.  It reports the battery charge as 100% (full charge).

Conclusion: this was the correct cable for this application.  This cable retailed for about $20, but price matching with Amazon brought the price down to a few cents over $5, including tax.

Question 1: does the battery charge indicator reflect reduced battery capacity as a result of age?  My answer: evidently not.  It is showing me 100%, but the critical question is, 100% of what?

Question 2: will the battery time estimate (tomorrow when I install the PCPE software) adjust for reduced battery capacity?  My answer: i have no idea, that must depend on the algorithm the PCPE software uses to estimate available battery time.  On my own system the estimated battery time is around 60 minutes, going up or (usually) down depending on what else I have running on the computer.  So the "battery time" function is evidently sensing the momentary power consumption of the system.  Question arising: does it also sense the available battery capacity? or is the "battery time" estimate computed on the basis of a fixed value for the battery capacity?  In the latter case one would have to expect any estimate of "battery time" of a 5 year old battery to be wildly optimistic.

I suppose one way to sense the "battery time" would be by sensing the voltage drop caused by the existing load, whatever that might be at any given moment.  In that case the reported "battery time" might well be sensitive to the condition of the battery.

Sorry if I am being a pest, but I do like to know how things work .

[Anonymous user]

I wasn't sure when APC switched to standard USB connectors on these. Evidently it was earlier than I thought.

These units are very simple creatures internally and they don't keep statistics on much of anything. Even if they fail a self test, they're likely to behave as though nothing happened on the next power up (until the brief self test at power on fails again, if it does). Whatever estimate it comes up with will in fact be wildly optimistic, even if it has been previously demonstrated that the battery is in no shape to last as it should.

Near as I can tell, PCPE only reports what the UPS tells it about remaining runtime.

If you're willing to use a generic battery, a replacement is about $20 shipped.

[Anonymous user]

This morning I installed the PCPE software on the water district computer.  I used the standard installer rather than the command-line version.  The installation was uneventful but I noticed several differences between this and the version installed (via command line) on my own system:

1. The install on the water office computer disabled the Windows native power management system, so the charge status icon disappeared from the system tray.  This did not happen on my system, which still has Windows power management and the charge status icon.

2. The version on my system appears to be newer (though both carry the same version number, 3.0.2) and has a more detailed display with a good many options and information displays that don't appear at all on the water office system - including the self-test option.

Both machines are running the same Windows version (1703, the Creators Update).  Mine is Windows 10 Pro, the waterworks one is Win 10 Home.

The waterworks machine shows a battery run time of 476 minutes, utterly ridiculous.  That number did not change when I moved the computer plug from a surge-protected outlet to a battery-backed-up one.  Could that perhaps mean that the "100% charged" battery is in fact 100% dead?

So I am wondering whether I should not uninstall PCPE from the waterworks computer, and reinstall using the command-line files.

Question arising: will the uninstall automatically re-enable Windows native power management, or must I do that by hand?

[Anonymous user]

I'm not sure what you mean by the "command line" version of a PowerChute installer. Every version of PowerChute Personal Edition I'm familiar with has only a graphical installer. PCPE definitely does (and should) disable Windows' built in power management.

The difference in what's displayed probably has to do with the exact model of UPS you have. Some models have more functionality than others do.

As far as the ridiculous runtime figure, it could indicate that the unit needs to be reset. APC calls this a "braindead" and it's done by first safely powering down the attached load. From there, turn off the UPS and disconnect it from the outlet. Turn it over, disconnect the battery, and depress the power button for a few seconds. You may hear the unit react briefly.

Before you reconnect the load to be protected, I'd try plugging in something like an incandescent lamp and testing it with that. Reconnect the battery, plug the UPS back in and turn it on. See if it will conduct (and pass) a self test with the lamp attached and turned on. If it does, the battery is probably "good enough" to get by while you get a replacement.

[Anonymous user]

Out of curiosity, does PCPE disablement of Windows native power management have a direct effect on the "percent charge" icon that shows in the system tray of battery-powered computers (including desktops connected to a UPS)?  I do understand the necessity for disablement or at least modification of the Windows power management scheme, since PCPE obviously needs to control the power switching during power outage.

Difference in what is displayed, point taken, though I am surprised at the missing self-test option, since the Help screen refers to it directly.

Reset and self-test, aha, I will try that next time I am there.  Since the water office in question is a public agency, their Board will have to approve the purchase of a replacement battery.  Plenty of them available, I notice.  Question arising: the unit says somewhere (i think on the label) that it is a BE 550G.  The PCPE software reports it (under "Technical Data") as ES 550G.  Are those the same (I thought not?) and do they use the same replacement battery?

[Anonymous user]

Any time I've installed PCPE, it has disabled Windows' own battery level indication in the process of disabling Windows Power Management. So I'm going to say that yes, it would.

I don't know why PCPE doesn't show a self test option when it's being used with a lower end UPS. The hardware itself will respond to a software self test command. (Other UPS management software can do this.) The help file itself is generic, and thus reflects features your UPS may not have, like voltage trim and boost.

The Back-UPS "ES" series has been produced for quite a while and as such it has gone through a lot of revisions. The 550 VA model uses a 12V, 7AH battery. (You can also use the 12V 7.2, 8 and 9AH batteries for a bit more runtime.)

[Anonymous user]

OK, dumb question, how to perform a self test (as in, reset the UPS, per instructions in previous post) if it's not on the menu?

[BillP]

Hi,

Self-test for that unit can be run by pressing and holding the power on button while the unit is powered on. The unit will beep and you should then release the power on button. The button will flash to indicate the test is running.

The unit also runs a self-test each time it is power on.

BTW: Running a self-test is available in PowerChute based on the UPS model. Each model Back-UPS offers a set of features and benefits. That model does not offer the option of running a self-test via PowerChute. If you would like that feature and want to stay with the BE / ES still Back-UPS upgrade to a BE750G.  

[Anonymous user]

Thanks Bill, that is very clear.  If I understood correctly, the blinking green light when the unit is switched on indicates a self test in progress.  In that case the unit in question has passed the self test several times in the last week or so.  Am I correct in thinking that if it did not pass, the red "Replace battery" light would come on?

I'll report further when I've had a chance to try the "reset" procedure as described by William.

Whether to upgrade this particular unit is up to the Board of the water district in question.  At least, we can slightly upgrade the battery capacity.  I'm sure I will be able to persuade them that it's high time to replace the 5 year old battery in this unit.

[BillP]

Hi,

On 9/1/2017 2:31 PM, John E said:

 If I understood correctly, the blinking green light when the unit is switched on indicates a self test in progress.

Correct.

On 9/1/2017 2:31 PM, John E said:

Am I correct in thinking that if it did not pass, the red "Replace battery" light would come on?

Correct.

[Anonymous user]

I had an earlier thread about difficulty installing PCPE on my desktop computer. That included some discussion of battery runtime estimates. I can't find the old thread, and have some new information, so I'm posting it here hoping someone will put it in the right place if this is not it. The question arose (also in the present thread about a different UPS unit) of how accurate the battery runtime estimates might be. My desktop computer is a Dell XPS 8700 running Win 10 Pro, 64 bit, version 1703. Also plugged into the UPS are a Samsung 24 inch LED monitor running off a 4 GB NVIDIA GTX 745 graphics card, a 2 TB WD MyBook external HDD and a Nikon 4000 dpi slide scanner. The UPS is a Back-UPS NS 900M. With the battery at full charge it shows available runtime as usually 56-67 minutes and power consumption as 48-52 watts. Last night it was put to the test. We had a terrific wind storm, and the power went out. The UPS carried the above described load for over 38 minutes before the power came back on. At that point the battery charge indicator in the system tray showed 60% charge remaining. So did PCPE, also indicating about 28 minutes remaining runtime. I conclude that the battery runtime estimate on this unit is right on the money, not exaggerated at all. After power was restored, it took about 3 hours for the battery to reach full charge again. Our local Internet service is by microwave relay from a Comcast backbone 20 miles distant. The storm knocked some of our hardware out of alignment and it took till mid-morning today to get everyone back on line. I observed that with no Internet signal (but the router and receiving dish powered up) the runtime estimate was 5 to 10 minutes LESS than when the Internet signal became available at my house. Go figure (at that time, the router and the dish were not plugged into the battery-supported receptacles). That oversight has now been corrected (and the Nikon scanner removed from battery backup, I won't be scanning slides in a darkened room).  With the added load, the estimated runtime is around 50+ minutes and the power usage around 65 watts.

[Anonymous user]

I thought I had already posted this, but maybe I got mixed up?

Anyway, I finally got a chance to try the reset procedure described by William.  Didn't have a lamp, so used a small table fan (13 watts) as test load.  Shut down and disconneccted computer, shut off and disconnected UPS, disconnected battery, held down power button for several seconds (unit made some kind of small sound).  Reconnected battery, plugged in UPS, connected test load (13 watt fan), switched on UPS which self-tested and ran the fan (green light OK), ran another self test for good measure (same result).  Shut off fan, unplugged it, plugged in computer and turned it on, did another self test, checked PowerChute on computer: it stll showed improbable number (476 hours runtime).  Plugged in 13 watt test-load fan (expecting it to make some change in the runtime number).  No change in the runtime number.  So maybe a) the UPS is not sensing the power draw of its load on the battery-supported outlets?  or b) it doesn't nave access to whatever other number it uses to compare with the current load to arrive at a runtime estimate? or c) something else?

I suppose the next step is to install a new battery and see what happens?

[Anonymous user]

On 9/8/2017 11:43 AM, John said:

 Didn't have a lamp, so used a small table fan (13 watts) as test load.

There are two problems with this: the first being that it's a very small load and the second being that it's a highly inductive load. Any load you use to test a UPS should run to at least 30% of its rated capacity, remain fairly constant over time and should be resistive in nature (as an incandescent lamp is).

If I'm remembering things correctly, the battery is over five years old. The lifespan of these batteries is from 3-5 years. Beyond that, they're on heavily borrowed time.

The runtime estimate isn't very accurate at the low end of the scale, and I suppose it should be considered that maybe the attached computer doesn't use a lot of electricity.

[Anonymous user · Accepted solution]

I can only say "yes to all of the above" .  Points about the load very well taken.

The battery is just about exactly five years old.  It's apparently been kept fully charged for most or all of that time.  I want to replace it, but first need authorization from the Board.

The attached computer is an HP Pavilion 20 all-in-one so yes, it probably doesn't use much electricity.  PowerChute on that machine does not display the power draw.  I suppose I can find a way  of measuring that.  My lamps at home are all LED so I'll have to scour around for an incandescent one - or put together a resistor pack (I'm so old I can remember when Radio Shack was actually a radio parts store).  Time to trot out some LONG unused memory circuits! .