Nut upsuser - Jun 2013 - several problems with Powercom BNT-500AP

OS name and version:

pi at raspberrypi ~ $ cat /etc/*release

PRETTY_NAME="Debian GNU/Linux 7.0 (wheezy)"

 

uname -a

Linux raspberrypi 3.6.11+ #393 PREEMPT Fri Mar 8 16:36:28 GMT 2013 armv6l
GNU/Linux

 

exact NUT version

pi at raspberrypi ~ $ dpkg -s nut

Package: nut

...

Version: 2.6.4-2.3

 

NUT installation method: package

 

exact device name and related information:

BNT-500AP

http://pcmups.com.tw/eA/html/product/show.php?num=229
<http://pcmups.com.tw/eA/html/product/show.php?num=229&root=13&kind=113&page
=1&keyword> &root=13&kind=113&page=1&keyword
 

I have several problems with my UPS:

1. Wrong battery and UPS date reported.

According to serial number 40438471204 UPS is manufactured on April 2012,
but NUT reports:

battery.date: 2010/12/20

ups.date: 2010/12/20

2. NUT occasionally (about once a day) reports that battery should be
replaced, but I have tested it by removing the power and it can supply about
170 W for about 20 minutes, which, according to specifications, is OK.

3. NUT very frequently (about 10 times a day) goes on battery followed by
immediate return to utility power, though there are no brownouts. My PC that
isn?t connected to UPS doesn?t reboot.

4. This is the most bothering. I have seen it only once. NUT reported ?on
battery? (actually there was no brownout) followed by immediate ?battery is
low?, that triggered shutdown of pi. 

Broadcast Message from nut at rasp

        (somewhere) at 20:45 ...

UPS BNT500AP at localhost on battery

Broadcast Message from nut at rasp

        (somewhere) at 20:45 ...

UPS BNT500AP at localhost battery is low

Broadcast Message from nut at rasp

        (somewhere) at 20:45 ...

 

Immediately after shutdown I turned the Raspberry pi on and checked the
status. It showed:

pi at raspberrypi ~ $ upsc BNT500AP at localhost

battery.charge: 100

battery.charge.low: 10

battery.charge.warning: 30

 

There is only pi and two low power embedded routers connected, so the total
power consumption from UPS is < 50 W.

 

Driver debug output attached.

-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://lists.alioth.debian.org/pipermail/nut-upsuser/attachments/20130605/6a708ac5/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: nut_driver_debug2.zip
Type: application/octet-stream
Size: 4879 bytes
Desc: not available
URL:
<http://lists.alioth.debian.org/pipermail/nut-upsuser/attachments/20130605/6a708ac5/attachment.obj>

On Jun 5, 2013, at 11:49 AM, M?rti?? Pu??tis wrote:
> Driver debug output attached.
The debug output looks normal (many devices do not implement the interrupt pipe;
hence, the timeouts). Unfortunately, we might need the log over a longer period
of time, including your observations as to when various events occurred (such as
transitions to battery).

-- 
Charles Lepple
clepple at gmail

Thanks for the support.
Here's list of when transition to battery with return to utility power
within a few seconds occurred in last few days:
Wednesday
1:41
11:48
11:54
12:59
13:42
14:42
18:13
18:14
Thursday
5:23
Friday
5:12
7:39
7:43
Saturday
7:34

Here's list of when messages that battery needs replacement occurred:
Wednesday
20:34
Thursday
9:07
23:09
Friday
18:25
Saturday
7:32
19:47

I'll take a longer driver log. This page
(http://www.networkupstools.org/support.html) says that to take a driver log NUT
should be stopped, so I won't be able to tell when these events occurred
within a driver log.

-----Original Message-----
From: Charles Lepple [mailto:clepple at gmail.com] 
Sent: Saturday, June 08, 2013 6:07 AM
To: M?rti?? Pu??tis
Cc: nut-upsuser lists.alioth.debian.org
Subject: Re: [Nut-upsuser] several problems with Powercom BNT-500AP

On Jun 5, 2013, at 11:49 AM, M?rti?? Pu??tis wrote:
> Driver debug output attached.
The debug output looks normal (many devices do not implement the interrupt pipe;
hence, the timeouts). Unfortunately, we might need the log over a longer period
of time, including your observations as to when various events occurred (such as
transitions to battery).

-- 
Charles Lepple
clepple at gmail