Results 1 to 12 of 12

Thread: So, anyway, I was talking to this bloke in the pub....

  1. #1
    Join Date
    Jun 2003
    Location
    UK
    Posts
    1,317

    Default So, anyway, I was talking to this bloke in the pub....

    ... after my G5 and 23"ACD died, described here.

    As yet, I don't know what caused the failure. I do know that the ACD's psu failed. I have not yet found out what has failed in the G5. I'm hoping that it's the G5 psu.

    The bloke in the pub is a friend of mine who deals professionally with big computer installations. I told him the tale, and his first reaction was that it might be a UPS malfunction. Come Wednesday I'll have him run his practiced eye and DMM over my APC Back-UPS 800.

    This bloke pointed out that a UPS malfunction might be the common point in the failure.

    He did go into some detail about what the malfunction might be, but like I said, we were in the pub!

    So.... I'm laying this one out aforehand to see if any Gurus have an opinion on it. I haven't done any research about known failures of APC Back-UPS 800's.

    It would be ironic should it turn out to be that the failures were caused by the UPS!
    "illegitimis non carborundum"

  2. #2
    Join Date
    Aug 2001
    Location
    Grangeville, ID USA
    Posts
    9,119

    Default

    gotta be something in there about a pint meter to measure your success.


    bad power in = dead power supplies. no doubt about that.
    molṑn labe'
    "I am a mortal enemy to arbitrary government and unlimited power. I am naturally very jealous for the rights and liberties of my country, and the least encroachment of those invaluable privileges is apt to make my blood boil."
--Ben Franklin

  3. #3
    Join Date
    Jun 2003
    Location
    UK
    Posts
    1,317

    Default

    Well, said bloke in the pub and I had a quick look at this APC Back-UPS 800 with a multimeter. All was done with no load since we neither of us have handy a disposable load with an IEC plug... hehe!

    AC input was 240vac.

    With input connected, output was 240vac.

    When the input was disconnected, output voltage fell to 200vac.

    There is/was no indication on the APC UPS that the battery required replacement.

    That voltage drop would indicate that a current spike to the connected loads should occur if the input voltage dropped. The very thing that a UPS should not do! I suspect that that is what fried the ACD's psi, and probably the G5's too.

    The UPS should never allow that. So, I think the damage was caused by the UPS.

    I have yet to have the discussion with APC.

    My Back-UPS 800 is out of warranty, I accept that.

    But APC have a lifetime policy as regards damage to equipment.

    I am prepared to go for a straight swap for a new UPS from them and eat the cost of a new psu for the G5 if that's all the damage to the G5.

    I'll let you know.....
    "illegitimis non carborundum"

  4. #4
    Join Date
    Feb 2001
    Location
    on the landline, Mr. Smith
    Posts
    7,787

    Default

    Get a quote to repair the G5, give them the bill, and put the $ in your pocket.

    Keep your eyes peeled for a used G5 if you really need a Mac that old....should be cheaper than the repair.

    UPSes that do harm drive me nuts. Had a big rack mounted job shut down about 4 servers once. No power event, just a UPS failure that shut off power. Like they are never supposed to do.
    "Imagine if every Thursday your shoes exploded if you tied them the usual way. This happens to us all the time with computers, and nobody thinks of complaining." -- Jef Raskin

  5. #5
    Join Date
    Jul 2003
    Posts
    1,818

    Default

    I think the first problem was you're talking to some bloke rather than some dude.

    Leaving that aside, if you get a replacement from APC, make sure they give you one with Sine Wave as the wave type since you have Intel Mac.

    Been doing some research myself on a new UPS and it seems there are some issues with Intel Macs and those that don't have sine wave. Anyone else want to confirm or deny that. With APC that's found on the Smart-UPS units, i.e. the ones that cost more than ones they market for home use.

  6. #6
    Join Date
    Jun 2003
    Location
    UK
    Posts
    1,317

    Default

    Quote Originally Posted by eric View Post
    ...if you get a replacement from APC, make sure they give you one with Sine Wave as the wave type since you have Intel Mac...

    With APC that's found on the Smart-UPS units, i.e. the ones that cost more than ones they market for home use.
    ...but not the APC Smart-UPS SC 1000VA apparently! That one still says "stepped approximation...."! And the next one up... not available here.

    This is all starting to get very messy. I'm about ready to walk away from APC.

    In a word, I'm gutted! I went with APC in the first place because they came well recommended and are a major player.

    I will try very hard not to get all wound up about this. Perhaps I worry about nothing. APC may yet do the honourable thing. Who knows. But I'm not holding my breath.
    "illegitimis non carborundum"

  7. #7
    Join Date
    Jun 2003
    Location
    UK
    Posts
    1,317

    Default

    Ok, I'll try to make sense, but don't hold your breath....

    The bloke in the pub knows his onions.

    I gave him the dead PSU out of my G5 and asked him to have a look inside.

    He did. Two 12V regulators were blown clean off the board.

    What is the difference between an APC UPS and a Smart-UPS?

    My cheap APC UPS failed and delivered a low voltage to the G5's PSU. The G5's PSU did it's best to supply the demand at the lower input voltage and promptly blew components off the board. So yeah, the G5's PSU did act as a fuse!

    But it should never have come to that!

    A Smart-UPS would not have delivered that low-voltage, according to my friend who deals with UPS's for a living. A Smart-UPS would have done all it could to supply the demand even unto it's own destruction, but it would not have delivered that low voltage which fried the PSU's in the G5 and the ACD.

    Discuss.
    "illegitimis non carborundum"

  8. #8
    Join Date
    Aug 2001
    Location
    Grangeville, ID USA
    Posts
    9,119

    Default

    Yeah... minimum voltage allowed for a G5 is 100 V. Below that is starts to crap itself.

    I would rather a UPC powered off than went off spec.


    R
    molṑn labe'
    "I am a mortal enemy to arbitrary government and unlimited power. I am naturally very jealous for the rights and liberties of my country, and the least encroachment of those invaluable privileges is apt to make my blood boil."
--Ben Franklin

  9. #9
    Join Date
    Feb 2001
    Location
    on the landline, Mr. Smith
    Posts
    7,787

    Default

    He's correct. Any good UPS is supposed to maintain minimum voltage.

    My experience: UPSs are good, except when they are not. When they fail—not often, but it does happen—bad things can happen. If you are lucky, they simply fail, and power off.

    Had a rack full of servers go down once when a high end UPS failed. I had it on good authority that this does not happen. If the UPS fails, it simply passes unconditioned power through to the devices.......same as no UPS. But there I was staring at about 5 servers that all shut off at once. All plugged into one massive UPS that was also mysteriously dead, not to be revived.

    Had a workstation UPS under a desk that the user plugged a space heater into. After a short period, smoke was boiling out, and everybody panicked and raced out of the room. Leave it to the IT guys to brave the smoke and unplug the !@#$% heater. Point is, the UPS did not shut off (the heater and the PC were still on....even as the UPS plastic case was starting to melt, and flames were exiting the rear vents). It did not simply shut down from excessive load as one would expect. I lovingly refer to this as the UPS China Syndrome.

    We all expect a UPS to take the hippocratic oath: Do no harm.

    So how often do they do harm? I can't say. 1 in 100 end in destructive failures? 1 in 1000? Less than that?
    "Imagine if every Thursday your shoes exploded if you tied them the usual way. This happens to us all the time with computers, and nobody thinks of complaining." -- Jef Raskin

  10. #10
    Join Date
    Aug 2001
    Location
    Grangeville, ID USA
    Posts
    9,119

    Default

    We have several mongo APC units. They can run 3 or 4 MacPros each. Had them for 8 years now. Last year, the one I am hooked to started blasting the fan on and off, as they will when the power goes down and the UPS start supply off the batteries. But the power was up. Put my hand on the thing and too hot to touch.

    Replaced the batteries and all is good again - BUT, I will never feel like they are safe, as in from fire. The battery cases were melted pretty badly, must be a dead short inside. I only got 6 or 7 years out of that set of batteries, so definitely my own fault. Bt you would think there would be some sort of protection from that much heat.
    molṑn labe'
    "I am a mortal enemy to arbitrary government and unlimited power. I am naturally very jealous for the rights and liberties of my country, and the least encroachment of those invaluable privileges is apt to make my blood boil."
--Ben Franklin

  11. #11
    Join Date
    Mar 2001
    Location
    Virginia... where one Democrat CAN make a difference
    Posts
    2,929

    Default

    Quote Originally Posted by eric View Post
    seems there are some issues with Intel Macs and those that don't have sine wave.

    All power supplies… well I should say all power period has a sine wave. The difference here is whether you have a stepped (modified) sine or a pure sine waveform. Stepped sine power (UPS, inverter, generator etc) are usually cheaper priced and work for many things. What they don't work for are cell phone (mp3 player, tablets, notebooks) wall chargers and any other wall wart inverter style charger they will burn out your charger on these things and can damage your device's internal battery and other things. I am not sure how that would affect a computer PSU but it is an inverter… Power company power is pure sine. Chargers like the iPhone wall charger use a fixed point on the sine wave as a timer for determining how long to charge. If your stepped sine steps right over that point your charger will not know when to shut off and dangerously over charge. Here's a pic



    Blue line is the pure sine
    red shows the stepped sine

    A/C motors run just fine with a stepped sine power source so things like refrigerators, freezers, power tools all work fine. As do incandescent lights. I do not know one way or the other whether LED's or CFL's have any issues with a stepped sine though I wouldn't expect CFL's to have any issue. I know nothing about LED's
    Damien,

  12. #12
    Join Date
    Feb 2001
    Location
    on the landline, Mr. Smith
    Posts
    7,787

    Default

    Good to know. Thanx!
    "Imagine if every Thursday your shoes exploded if you tied them the usual way. This happens to us all the time with computers, and nobody thinks of complaining." -- Jef Raskin

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •