[Nut-upsdev] Krauler UP-D1200VA

Vamp admin at vampik.ru
Wed Sep 5 23:36:46 UTC 2007

Hello, Alexander!
You wrote to "Vamp" <admin at vampik.ru> on Wed, 05 Sep 2007 22:19:07 +0400:

>> The main problem is the unexpected behaviour of FL_LOAD_OFF flag. It's 1
>> in normal and battery modes and 0 in battery low and load off modes.

> This is a pure hardware bug. The only way to workaround it is to
> substitute status flags with the desirable ones. Not to mention that
> the conversion should be model-specific. It goes to be a really huge hack.

> Well, I'll look more careful what can be done.

I think the quick solution would be to ignore the FL_LOAD_OFF flag, as the
original Megatec software do.

>> shutdown.return, shutdown.stayoff, load.off, reset.watchdog doesn't work.
>> Reason is the driver error: set_data_krauler: output string too large

> These commands are not implemented in the "krauler" subdriver because
> my UPS don't seem to support them. If you figure out that your UPS
> supports these commands and how to run them I'll add them to the driver.

Yes, I already took a closer look at megatec and megatec_usb sources..

By the way, I found the comment:
 Still not implemented:
  0x6 T<n> (don't know how to pass the parameter)
and made some tests using UPSilon2000. One can choose a number of minutes
for the battery test to run, but it seems to be bogus, the command sent by
the software is always the same (0x06). Most funny, any of the three
commands (0x04, 0x05, 0x06) makes my hardware to run a 3 seconds battery

Some words about shutdown.. It seems there are commands similar to S<n>R0000
(UPS will never turn itself on when the utility is recovered after these
commands, it needs to be done manually by pressing the power button). The
settings window allows to choose <n> from the interval 0.2, 0.3, ..., 0.9,
1, 2, ..., 99 minutes. I'm confused by the relation between the value I
choose and the real time before the shutdown, may be the service or even
Windows restart is required for properly setting the value, because this
damn piece of software consists of three parts - rupsmon service, usbmate
helper and UPSilon2000 GUI client. Here are some observations:
command sent by UPSilon - time before UPS shutdown
0x18 - 20 sec
0x28 - 24 sec
0x38 - 30 sec
0x68 - 12 sec
0x78 - 52 sec
0x88 - 1 min 42 sec
Answer to any of these is "UPS No Ack". More testing is needed.

With best regards, Vamp.  E-mail: admin at vampik.ru 

More information about the Nut-upsdev mailing list