[Nut-upsuser] UPS Battery internal resistance matters.

Arjen de Korte nut+users at de-korte.org
Fri Dec 14 09:56:08 UTC 2007


> What suprised me was this table entry from APC ASTE-6YWRZE_R0_EN.pdf
> 	Smart-UPS 2200/3000 VA 100/120/230 VAC Tower User Manual Page 9
> ] Function:		Automatic Self-Test
> ] Factory Default:	Every 14 days (336 hours)
> ] User Selectable Choices: Every 7 days (168 hours);
> ] 			 On start up only;
> ] 			 No self-test
> ] Description:	Set the interval at which the UPS will execute a self-test.
> APC are less conservative than we are !
> I agree NUT should not do self test as shipped by default.
> I think NUT by default should change the APC default,
> 	(I'll look later to see what NUT does, as I'm new back to NUT).

'Correcting' things automatically is not a good idea. This tends to
frustrate the heck out of users, when NUT is doing something they have not
asked for. The only exception being changing the settings of the UPS to
make monitoring possible in the first place. Most UPS'es will run a quick
test automatically once ever week or two, this is not uncommon at all.

> I agree UPS users should do periodic battery tests, at times they
> peronally decide appropriate for their installation.
>    However, rather than doing it by front button,
> 	( which merely shows pretty but misleading voltage LEDS, &
> 	one may not get scared enough by eg my button test just
> 	now, where one 60 Watt light bulb after a few seconds self
> 	test took the batteries down drom 5 LEDs to 3 LEDs ...)
>    Instead NUT could offer a template script, callable by human or
>    crontab, which can measure Voltage before during & after And
>    load, & thus resistance.

Not at all. It makes no sense to measure internal resistance if you don't
know what the cut-off value is. An acceptable internal resistance for a
small battery, may be totally unacceptable for a larger installation. It
is the ability to provide enough power to the load from the batteries.
Besides this, the accuracy and resolution of the 'measuring' circuits in
most commercial grade UPS'es will be a far cry from what is needed to
really measure internal resistance. It is not uncommon to have resolutions
of 0.1 or even 1 Volt for the voltage measuring and many UPS'es have no
means of measuring the battery current with any kind of accuracy anyway.

If you monitor the health of your batteries from the moment of installing
them (this is vitally important), it is not too difficult to determine
when they have worn out if you do weekly quick- and quarterly deep
discharge tests. It is the trend that matters here, not absolute values. I
would recommend changing the batteries regardless of the outcome after 3
to 5 years anyway (of maybe even sooner if the ambient temperature is
above 25°C). Replacing batteries is cheap compared to the cost of failing
when they are really needed.

Best regards, Arjen

More information about the Nut-upsuser mailing list