• UPS Replacement

    From Barry Martin@454:1/1 to All on Sun Oct 24 18:21:00 2021

    Hi Folks!

    Found out something interesting today: it's suggested UPSs be replaced
    every six to ten years max. The UPS itself. Looked into it because I
    had a UPS fail -- well, one of the batteries. Guess had a power glitch
    which shut down the UPS but I didn't notice until several hours later
    (since have written a small script to monitor). Powered up, plugged the
    plug to test -- UPS worked fine.

    About a month later heard a pole transformer fuse blow -- was a few
    blocks away and this time heard the other UPSs click over. The one in question died again. Batteries were only about 18 months old, so should
    be fine but they felt a lot warmer than seemed normal. Thinking either
    faulty battery or maybe the UPS was starting to fail - I've had it for
    quite a while.

    Today tested the removed batteries using an automotive bulb. First
    battery - nice and bright. Second battery -- hmm, yup: connection is
    good. Deader 'n a door nail! Started cutting the duct tape holding
    them together: saw a big split in the case! That one goes to recycle,
    the other is probably good but will be used in a lesser-critical device
    set.

    Have a slight suspicion on this UPS: seems like it had failed the same
    way some time back (the 18 months?) but at the time didn't think much
    about it other than the batteries didn't last as long as normal and I
    use this brand and rating (9AH) in just about all the UPSs here --
    easier and cheaper to keep spares in stock!

    So did a bit of research and found the life of a consumer-grade UPS is
    six years, ten maximum. This particular UPS is from 2007, so 14 years!!
    Found a couple of articles which may be of interest:

    https://info.hummingbirdnetworks.com/blog/three-signs-its-time-to-replac e-your-ups


    https://tldp.org/HOWTO/UPS-HOWTO/maintaining.html Also: http://www.tldp.org/HOWTO/UPS-HOWTO.html

    Section 5.5 of the latter has a section re: extending the capacity
    using 35 ampere-hour deep cycle marine batteries. I read some years
    back about using 9 AH batteries in UPSs designed for 7 AH -- some
    discussion on may damage the charging circuit, some discussion on the
    system may shut down at the same time because it doesn't know how to
    interpret the larger capacity battery, etc. I've used 9AH batteries for
    years and have not had any bad issues (going to exclude the one UPS I
    talked about above because I'm thinking more age-related.) Here the use
    of 9AH over 7AH has extended the run time -- over the years have had
    some lengthy power outages. I don't know how comfortable I'd be putting
    a 4x-rated battery in: the draw is only going to be whatever the UPS
    uses normally, just able to run on battery power longer. My concern
    would be the recharging circuit: the discussion on using 9AH batteries mentioned potential issues with overheating/breaking down of the
    charging circuit. OTOH there is constant use of small capacity chargers
    to recharge huge capacity batteries: have one here for the battery
    back-up on the sump pump.




    ¯ BarryMartin3@ ®
    ¯ @MyMetronet.NET ®

    ... I finally got eight hours of sleep. It took me three days but whatever.
    --- MultiMail/Win32 v0.47
    þ wcECHO 4.2 ÷ ILink: The Safe BBS þ Bettendorf, IA

    --- QScan/PCB v1.20a / 01-0462
    * Origin: ILink: CFBBS | cfbbs.no-ip.com | 856-933-7096 (454:1/1)
  • From Ky Moffet@454:1/1 to Barry Martin on Tue Oct 26 00:48:00 2021
    BARRY MARTIN wrote:
    Hi Folks!

    Found out something interesting today: it's suggested UPSs be replaced
    every six to ten years max. The UPS itself. Looked into it because I

    Well, consider the source... they sell UPSs.

    The longest I've seen one last was about 25 years. (Battery replaced
    once. Circuits finally died after a lightning strike one pole away.) The shortest... pretty much any current off-the-shelf unit only lasts about
    2 years. The shell may still function but the battery will be kaput,
    because all consumer UPSs overcharge the battery, which prematurely
    kills it.

    CyberPower (what they sell at Costco, otherwise a decent unit) has this
    down to a fine art; their batteries reliably die a few days out of
    warranty. Since the unit is $95 and a new battery is $88 plus I have to
    drive clear up to the other side of town to get one, I see no point in fighting with it to get the damn thing apart and extracted; I just get a
    new unit, tho when I think of it in time I get Costco to swap 'em in
    warranty. :D

    https://tldp.org/HOWTO/UPS-HOWTO/maintaining.html Also: http://www.tldp.org/HOWTO/UPS-HOWTO.html

    This overcharging problem, which ESR (the guy who wrote the above HOWTO) documented elsewhere, is apparently by design as it's entirely
    unnecessary (AND in the olden days, they didn't do this). ESR was
    working on a consumer-level UPS with a circuit that fixes the problem,
    but far as I know it hasn't come to fruition.

    In chron order:

    http://esr.ibiblio.org/?p=7839
    http://esr.ibiblio.org/?p=7848
    http://esr.ibiblio.org/?p=7881
    http://esr.ibiblio.org/?p=7918
    http://esr.ibiblio.org/?p=7923
    http://esr.ibiblio.org/?p=7930
    http://esr.ibiblio.org/?p=7932
    http://esr.ibiblio.org/?p=7948
    http://esr.ibiblio.org/?p=7954

    And that's the last post on the subject. (A year ago the blog server
    went KABOOM, plus he had some health issues, and that probably put a
    crimp in things.)

    You can tell when overcharging starts to kill it, even if you can't tell
    from runtime issues, by the stuffy metallic smell they emit ... that's hydrogen being vented by the overcharged battery. Also, they'll often
    get warm or hot even when the power is on and they're not discharging.
    One that's operating normally (not overcharging) stays cool. However,
    staying cool doesn't guarantee it's not on its way out.

    If you find one that lasts several years... when the battery dies, I
    would replace the battery and keep the shell, as that's one that is NOT overcharging the battery. Replacing it might be tossing good hardware in
    the dumpster in favor of hardware not so good.
    þ RNET 2.10U: ILink: Techware BBS þ Hollywood, Ca þ www.techware2k.com

    --- QScan/PCB v1.20a / 01-0462
    * Origin: ILink: CFBBS | cfbbs.no-ip.com | 856-933-7096 (454:1/1)
  • From Barry Martin@454:1/1 to Ky Moffet on Tue Oct 26 08:20:00 2021

    Hi Ky!

    Found out something interesting today: it's suggested UPSs be replaced
    every six to ten years max. The UPS itself. Looked into it because I
    Well, consider the source... they sell UPSs.

    Yes, noted that, If last forever hard to make the replacement game so
    they went the green and new features way.


    The longest I've seen one last was about 25 years. (Battery
    replaced once. Circuits finally died after a lightning strike one
    pole away.) The shortest... pretty much any current off-the-shelf
    unit only lasts about 2 years. The shell may still function but
    the battery will be kaput, because all consumer UPSs overcharge
    the battery, which prematurely kills it.

    One problem with current electronics is they design too close to the
    edge: 150v surge and pfft! (Maybe not that number but seems close.) I
    did install a whole-house surge suppressor which does seem to work:
    since then I've noticed the lack of brightening of the light bulbs (but
    then also have mostly switched to LED), and maybe more split-second
    power outages, so the supressor is shunting to ground (?).

    Here also have some old UPSs: one Belkin (~425VA) died and seems to be
    due to a fuse. Can see it on the schematic but no idea where it is on
    the board. ...It's since been gutted and used as a battery pack.
    Another Belkin (same - I bought the last three on clearance) did have an
    issue with the battery overheating during normal power -- the battery
    was past normal life and no problems since.

    As for the other UPSs here, I have not noticed a 'running hot' issue, so hopefully means their design is such they are not overcharging. (Made
    note to check the one I just replaced the batteries in.)



    CyberPower (what they sell at Costco, otherwise a decent unit)
    has this down to a fine art; their batteries reliably die a few
    days out of warranty. Since the unit is $95 and a new battery is
    $88 plus I have to drive clear up to the other side of town to
    get one, I see no point in fighting with it to get the damn thing
    apart and extracted; I just get a new unit, tho when I think of
    it in time I get Costco to swap 'em in warranty. :D

    Yes, I'd swear there's a timer in some devices! For the $7 I'd get a
    new unit also. The batteries most of the UPSs here use cost $25-30,
    which is a lot less than a new UPS. As for fighting to replace the
    battery: yup -- some are not too happy to be opened up! One UPS was
    easy to open but the batteries are in almost a prison cell: have to
    unscrew the PC board, unscrew supports, do something else... I had
    looked on-line and pretty much everyone said it was a great UPS but the batteries were essentially impossible to replace. Send in for the
    manufacturer to replace (probably swap out) -- what do I do in the
    meantime, hope the power doesn't go out?? I ended up drilling a hole in
    the side of the case, two heavy-gauge wires come out -- inside attached
    to the wires which went to the original batteries (still inside),
    outside connect to two batteries. Have to admit it's the easiest one to replace batteries!!



    https://tldp.org/HOWTO/UPS-HOWTO/maintaining.html Also: http://www.tldp.org/HOWTO/UPS-HOWTO.html

    This overcharging problem, which ESR (the guy who wrote the above
    HOWTO) documented elsewhere, is apparently by design as it's
    entirely unnecessary (AND in the olden days, they didn't do
    this). ESR was working on a consumer-level UPS with a circuit
    that fixes the problem, but far as I know it hasn't come to
    fruition.

    Probably not. I don't follow in-depth but probably would have noticed
    someone advertising this feature. Due to 'frequent' power outages here
    (mostly birds or squirrels getting too friendly with the pole transformer
    in the back yard (utility lines back there instead of along the street)
    have been on the look-out for long-lasting, good quality, relatively inexpensive UPSs to run "critical" stuff.



    http://esr.ibiblio.org/?p=7839

    I'll take a look at these later: I'll probably snag some notes.

    And that's the last post on the subject. (A year ago the blog
    server went KABOOM, plus he had some health issues, and that
    probably put a crimp in things.)

    Both can create problems. :/


    You can tell when overcharging starts to kill it, even if you
    can't tell from runtime issues, by the stuffy metallic smell they
    emit ... that's hydrogen being vented by the overcharged battery.
    Also, they'll often get warm or hot even when the power is on and
    they're not discharging. One that's operating normally (not
    overcharging) stays cool. However, staying cool doesn't guarantee
    it's not on its way out.

    I have not noticed any smell like that, though probably due to the other
    person here cigarette smoke. Have noticed, but after-the-fact, the
    occasional over-heated battery. And as noted above, that seems due more
    to the battery. I have stickers with date battery installed on the UPSs
    and most are lasting several years while this one UPS seems to be
    lasting only 18 months.


    If you find one that lasts several years... when the battery
    dies, I would replace the battery and keep the shell, as that's
    one that is NOT overcharging the battery. Replacing it might be
    tossing good hardware in the dumpster in favor of hardware not so
    good.

    Right. I tend to frugal so try to buy decent stuff that will last. One indicator seems to be the output: if pure square wave I don't even
    consider it. Go by degree of stepped sinewave, though now seems most
    I've taken a look at are doing pretty good at approximating the sine
    wave.

    ...Think I'll play with 'Network UPS Tools' later: re-found it the other
    day when checking the information. (apcupsd is for APC UPSs, NUT for
    just about everything else.)




    ¯ BarryMartin3@ ®
    ¯ @MyMetronet.NET ®

    ... If number 666 is considered evil, then 25.8069758 is root of all evil.
    --- MultiMail/Win32 v0.47
    þ wcECHO 4.2 ÷ ILink: The Safe BBS þ Bettendorf, IA

    --- QScan/PCB v1.20a / 01-0462
    * Origin: ILink: CFBBS | cfbbs.no-ip.com | 856-933-7096 (454:1/1)