[Nut-upsuser] The dreaded Tripp Lite SMART500RT1U and NUT

[Charles Lepple]

On Tue, Jul 1, 2014 at 3:12 PM, Steve Ballantyne
<steve.ballantyne at gmail.com> wrote:
> On Tue, Jul 1, 2014 at 9:17 AM, Steve Ballantyne
> <steve.ballantyne at gmail.com> wrote:
>> How would I go about providing you this stuff?  Can you point me to
>> something that would help trace out the USB data?

Tracing the data on Windows is not my speciality, but there are a few
messages in the NUT archives (can't remember whether nut-upsdev or
-upsuser, but the search form at
http://www.networkupstools.org/support.html hits both). The basic idea
is to configure the USB sniffer software to monitor your UPS, then you
would take notes on what the Windows software reports as you do a
similar test to what you posted.

A few people have done this by running Windows in a VM, and monitoring
on the Linux host side. Others have installed the sniffer software
into Windows itself. The details are going to depend on what you find
to be a more palatable challenge. The free Windows sniffer software is
mostly from the XP era, so with a newer OS, YMMV.

I don't know if the serial port will be easier or harder to deal with.

> I thought I would try to make sense of the ups.debug data.  All I have
> really figured out is that the fourth hex segment for S is a 00 for
> plugged in and a 01 for on-battery. Which the driver already knows.
> The rest is a mystery to me.  But I played around with unplugging and
> running the load down, and then recharging, etc.  Here are my notes
> and observations.  NC = "no change".  I am assuming that those fields
> are the make/model/serial number, etc.
>
> S fourth segment = 00 for on-power, 01 for on-battery

How far down did you let the battery go? I don't usually recommend
running it down all the way very often, but it would be good to do
that at least once to verify the LB flag. On yours, I think that would
look like "S: 00 ..."

The "01 04" seems to be the state before a self-test on yours as well
as on the ASCII version of the SMARTPRO protocol. See comments around
"switch(s_value_2) {" in the source code.

> UNPLUGD a while
> ups.debug.D: 00 00 00 7d 0d 00 00 '.......'  <---- 4th segment = 125

Looks like the first pairs of digits are the input voltage, and "00
7d" = 125V out.

> ups.debug.L: 06 00 00 07 58 58 0d '....XX.'  <---- 1st segment = 6

Not sure, but I have a feeling that one of 06 or 07 is a percent load.

Lots of good progress here! I tried to enable a few more commands for
your model. For instance, the "M" values seem to be min/max voltage,
so there is a "reset.input.minmax" command to use with upscmd.  Also,
you can trigger a battery test with "test.battery.start". I assume the
work the same way as on Protocol 3003 :-)

The attached patch is meant to be applied on top of the previous
patch. If it's easier, I can diff it against Git master.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: tripplite_3005_v0.25.patch
Type: text/x-patch
Size: 5835 bytes
Desc: not available
URL: <http://lists.alioth.debian.org/pipermail/nut-upsuser/attachments/20140701/8ef0bd26/attachment-0001.bin>