Talk:SensorTag User Guide

From Texas Instruments Wiki
Jump to: navigation, search
For technical support please post your questions at Please post only comments about the article SensorTag User Guide here.

Comments on SensorTag User Guide

Slon said ...

IT der acc Sensor an kxt9 insted of an ktxj9

--Slon 15:53, 6 November 2012 (CST)

Jboe said ...

Thanks for pointing out, it is fixed now

--Jboe 08:13, 14 November 2012 (CST)

Kickingvegas said ...

Somewhat confused in reconciling this statement with what I see in the example iOS source code, where the offset is put in the high byte.

The TI Base 128-bit UUID is: F0000000-0451-4000-B000-000000000000. Alle sensor services use 128-bit UUIDs, but for practical reasons only the 16-bit part is listed below. This is embedded in the lower 16 bits of the 128-bit UUID. Example: 0xAA01 maps as F0000000-0451-4000-B000-00000000AA01."

Looking at deviceSelector.m, the address I see for the equivalent example above is set to

   [d setValue:@"f000aa01-0451-4000 b000-000000000000" forKey:@"IR temperature data UUID"];

where the "aa01" is set higher in the bit field.

Please clarify.

--Kickingvegas 22:32, 11 December 2012 (CST)

Capiman said ...

I think i have found 3 bugs in the description: (BTW: Very helpful, many thanks!) 1) Gyroscope <Configuration> is not at 0x44, but at 0x5B 2) Bit number in Test Service for Gyroscope test is not 6 but 5. 3) Search for "<3.4 ;Hz" (it is in Sensor overview). Think it shall be "<3.4 MHz"

--Capiman 11:48, 1 February 2013 (CST)

IronSheep said ...

It appears that there is a lower limit for both "Magnetometer Period" [10-255] and "Accelerometer Period", also [10-255]. Lower values 0-9, for either, return an error if attempt to write value to characteristic.

--IronSheep 04:56, 18 April 2013 (CDT)

IronSheep said ...

Experimentation with the Magnetometer Period settings appear to show that the Description for this Characteristic F000AA33* should read the same as the Accelerometer Period Characteristic F000AA13*. That is it appears that both descriptions should be "Period = [Input*10]ms where Input is [10-255]".

--IronSheep 05:04, 18 April 2013 (CDT)

Lnanek said ...

Running into the same issue as Kickingvegas, but on an Android device. Seeing "f000aa00-0451-4000-b000-000000000000" returned by the SensorTag device instead of the "F0000000-0451-4000-B000-00000000AA00" suggested by the documentation. Same for the other service UUIDs.

--Lnanek 18:39, 19 April 2013 (CDT)

Inder said ...

IR Temperature bad handle for enable sensor: 0x2A right value is 0x29

--Inder 05:53, 12 June 2013 (CDT)

Inder said ...

uint16 Pr; int16 Tr;

its reversed! (signed is Pr - pressure) uint16 Tr; int16 Pr;

--Inder 04:42, 17 June 2013 (CDT)

Ziggles said ...

I'm not sure the conversion factors for the MAG3110 Magnetometer are correct in the code snippets. The calculation assumes a 2000uT Full scale deflection for a full 16bit count range (65536). However, looking at the MAG3110 data sheet tables, the conversion rate for sensitivity should be 0.1uT per count. This is confirmed in the section (5.1.2) about the "OUT" data registers which says that the raw count is limited to +/-10000 max and min, the factory config count can be +/-10000 and the user config can be +/-10000 giving a max/min range of +/-20000 or +/-30000 depending on the mode. The SensorTag firmware doesn't allow the use of the user config, so this can be ignored. We end up with a max +/-2000uT reading from a max/min count of +/-20000, which equates to the 0.1uT per count in the table. So I think that the calculation "....* (2000f / 65536f) ...", should read ".... * (0.1f)...". I changed this in the Android App code, and I get readings much more in line with the total magnetic field densities that I would expect in this region. Comments anyone ?

--Ziggles 15:48, 15 Nov 2013 (GMT)

There are no threads on this page yet.