[Nut-upsuser] Shutdown after 5 minutes?

Julian H. Stacey jhs at berklix.org
Fri Sep 15 08:33:43 UTC 2006


midsummerdream.org> <450A4C01.3030908 at xs4all.nl>
Comments: In-reply-to Udo van den Heuvel <udovdh at xs4all.nl>
   message dated "Fri, 15 Sep 2006 08:45:21 +0200."
--------
Udo van den Heuvel wrote:
> Rob wrote:
> > Here's an updated set of info on my UPS, which includes what upsc
> > reports when the power is pulled by the computers are still on:
>
> We had a power outage this morning and I did some monitoring on the UPS
> which is connected to my small Linux box.
> Battery charge did not go below 70% when the box died after about 5
> minutes. No shutdown scripts, no warnings of shutdown, nothing.
>
> When you add the info that Rob has (still going down after 5 minutes
> although...) one might suspect a bug somewhere.

Start with basics: Suspect the battery. Even the microprocessor
inside the UPS that NUT talks to, neeeds good power, if the battery 
suddenly drops, the microprocessor may not have time to warn NUT
of an approaching or actual failure, & NUT might not have time or power.

As I wrote to someone else recently on this list: A battery can
appear to have a good charge state (= voltage) even when it's gone
bad & developed a high internal resistance.  That happened to me a
while back when I wasn't even running any software to control UPS.
UPS had been running for years, with charge state good (LEDs). I
pulled power, & it dropped out after maybe 10 secs ! 2 of the 4
batteries had good voltages but high internal resistance.

Test each of your batteries' internal resistances, run each battery
seperately into eg a few car 12V headlamp light bulbs, Use Ohms
law: V = IR.  Measure
  Voc = Open Circuit no load Voltage (about 13 or 14)
  I   = Current through bulb[s] (as many as you can in parallel)
  Vload = While bulb burns bright, measure voltage (but at the battery not
         at the bulb, to avoid voltage drops in thin wire)
Rinternal = (Voc - Vload) / I  

My really bad batteries had 2 & 3 ohms.  The less the better, I've
not measure a new one, but I guess not more than a few tenths of
an ohm (& a lot less than that for eg car starter batteries). 
(Battery manufacturers have copious info in PDF on web).

Also calculate capacity in Watt Hours, by running batteries some
hours till they grow dim. (You'll need to log current each half
hour, as it varies, as batt voltage drops, as resistance of light
bulbs depends on brightness.

--
Julian Stacey.  BSD Unix C Net Consultancy, Munich/Muenchen  http://berklix.com
Mail Ascii, not HTML.           Ihr Rauch = mein allergischer Kopfschmerz.
        Don't buy it ! Get it free !  http://berklix.org/free-software



More information about the Nut-upsuser mailing list