-32763.0 dBm

[Aleksander Morgado]

On 27/09/12 13:28, Bjørn Mork wrote:
> Aleksander Morgado <aleksander at lanedo.com> writes:
> 
>> Ec/Io is supposed to be given as negative 0.5dBm increments. This seems
>> to be true for Get Signal Strength, where the value is given as a uint8;
>> but is probably wrong for the newer Get Signal Info where the value is a
>> 16bit integer, so probably is signed; e.g. it should be a 'gint16'
>> instead of a 'guint16'. What I don't know is whether it is still given
>> in 0.5dBm increments or not.
>>
>> Can you try the attached patch and see if it gives more appropriate values?
> 
> I still have no idea what the appropriate values are, but the patch puts
> the value inside a 0 to -100 dBm scale at least:
> 
> bjorn at nemi:/usr/local/src/git/libqmi$ cli/qmicli -d /dev/cdc-wdm0   --nas-get-signal-info
> [/dev/cdc-wdm0] Successfully got signal info
> WCDMA:
>         RSSI: '-84 dBm'
>         ECIO: '-2.0 dBm'
> 

Your output with Get Signal Strength was before:

ECIO:
        Network 'umts': '-126.0 dBm'

That was given from a raw read value of '252' treated as unsigned. If we
treat that value as signed we would get '-4', which multiplied by (-0.5)
gives us +2dBm.

So, if I put back the multiplication with (-1) in Get Signal Info and I
make all EcIo values reported as signed, we should then get both values
equal.

Can you test the attached patch?


-- 
Aleksander
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 0001-nas-Ec-Io-given-as-a-signed-integer-in-newer-Signal-.patch
Type: text/x-patch
Size: 5787 bytes
Desc: not available
URL: <http://lists.freedesktop.org/archives/libqmi-devel/attachments/20120927/c5d11190/attachment-0001.bin>