[Nut-upsuser] Battery test feature of APC Back-UPS RS series

Charles Lepple clepple at gmail.com
Sun Oct 4 16:03:53 UTC 2009


On Oct 4, 2009, at 8:16 AM, Jean Delvare wrote:

> Hi Arnaud,
>
> Sorry for the long delay.
>
> On Wed, 20 May 2009 16:01:20 +0200, Arnaud Quette wrote:
>> 2009/5/18 Jean Delvare <khali at linux-fr.org>
>>> The USB sniffer I am using [1] is not exactly user-friendly. I  
>>> have to
>>> restart the system for every new capture (otherwise I am unable to
>>> reset the capture file), and the capture doesn't end when I exit the
>>> application. It is also not possible to manually start or stop the
>>> capture. Is there any better USB sniffer application you would
>>> recommend?
>>
>> I don't know. I've not used windows for years...
>> maybe that one, but it's a shareware!
>> http://www.hhdsoftware.com/Products/home/usb-monitor.html
>
> OK, I've given it a try, it is quite powerful, unfortunately I don't
> know much about USB so I may not be using it in the best way.
>
> Anyway, I managed to capture some data and I hope it will help. I've
> asked for HID decoding and URB decoding. The files are here:
>
> http://khali.linux-fr.org/files/nut/
>
> idle-*.htm files are captures when I didn't do anything. test-*.htm
> files are captures when I triggered a battery test.
>
> I tried finding differences between the captures, and found that the
> test capture had the following entries not present in the idle  
> capture:

If I am interpreting this correctly, the software is sending a HID  
SetReport(0x16, 0x01) followed immediately by SetReport(0x16, 0x00).

 From your previous email:

Path: UPS.Battery.Test, Type: Input, ReportID: 0x16, Offset: 0, Size:  
8, Value: 6.000000
Path: UPS.Battery.Test, Type: Feature, ReportID: 0x16, Offset: 0,  
Size: 8, Value: 6.000000

You are correct that the Input is already mapped to ups.test.result,  
but it should be possible to specify that writing to UPS.Battery.Test  
will start a battery test. ("Feature" reports are writable as well as  
readable.)

I am not sure how that would send the 0x01 then 0x00 with this driver,  
though - maybe Arnaud or Arjen can help with that part.



More information about the Nut-upsuser mailing list