[Nut-upsuser] Incorrect Values From usbhid-ups

[Bill S]

On Fri, Mar 28, 2014 at 9:39 AM, Charles Lepple <clepple at gmail.com> wrote:

<<Snip>>
>>>
>>> I'll be honest, I haven't really done much with the CGI scripts lately. For debugging values, it is probably better to be closer to the driver. I'd recommend using "upsc" or the driver logs.
>>>
>>> To that end, the .ac.txt and .bat.txt files both show 24.0 V:
>>>
>>>   0.081935     Entering libusb_get_report
>>>   0.082018     Report[get]: (2 bytes) => 09 f0
>>>   0.082029     Path: UPS.PowerSummary.ConfigVoltage, Type: Feature, ReportID: 0x09, Offset: 0, Size: 8, Value: 24
>>>   0.082035     Entering libusb_get_report
>>>   0.082115     Report[get]: (2 bytes) => 0a f0
>>>   0.082124     Path: UPS.PowerSummary.Voltage, Type: Feature, ReportID: 0x0a, Offset: 0, Size: 8, Value: 24
>>
>> I know there are two 12V batteries in series that provide 24V which is
>> what is being shown, but an actual 24V for new fully charged batteries
>> seems low to me.  Between that and the fact that there are no tenths
>> of a volt shown and the battery is shown as 24.0 both on AC and on
>> battery makes me think we are not seeing the actual battery battery
>> voltage but some nominal or default value.  I do not think the battery
>> voltage is actually being tracked.  ???
>
> "24.0" is slightly optimistic on my part ;-) but the value "f0" is hex for 240, so the UPS is capable of reporting increments of 0.1V. We used to print the floating point values with "%.6f" or similar, which is IMHO slightly more misleading since nobody is measuring with microvolt resolution. That's been on my list to fix for some time now.


I agree.  Three significant figures is the right number.  More as you
say would be rather meaningless but only two significant figures
really does not tell you much.

>
> ConfigVoltage (battery.voltage.nominal) is often an even multiple of 12V, but you're right, the actual voltage should be higher for fully-charged batteries. That voltage may be hiding somewhere else in the report.
>
> If there is a battery test command, that (plus the -DDD logs) would be a good way to find where the actual battery voltage is hiding.

In effort to find the actual battery voltage I ran
/usr/local/ups/bin/upsc from the command line (results attached) and
it also reports a battery of voltage of 16.0. I also tried unloading
NUT and running CyberPower's latest PowerPanel Linux software but the
only thing I could see to do with that software is run "pwrstat -test"
and "pwrstat -status" and neither one resulted in any output of
battery voltage.  If there is something else I should be doing along
those lines that I missed, please let me know.

I guess if there is any good news it is that it appears that the
problem is not with the CGI scripts for both the output AC voltage and
battery voltage.

I also noted that when running the PowerPanel software that the
reported output AC voltage matched the input AC voltage and that the
input AC voltage went to zero when on battery. I would take that to
mean that the UPS has those values available somewhere and so perhaps
it has the right battery voltage as well if they can be accessed.



>
> --
> Charles Lepple
> clepple at gmail
>
>
>
>
> _______________________________________________
> Nut-upsuser mailing list
> Nut-upsuser at lists.alioth.debian.org
> http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/nut-upsuser
-------------- next part --------------
A non-text attachment was scrubbed...
Name: upsc_out.txt.gz
Type: application/x-gzip
Size: 454 bytes
Desc: not available
URL: <http://lists.alioth.debian.org/pipermail/nut-upsuser/attachments/20140329/0478a98d/attachment.bin>