Sign Extension for 13 bit to 16 bit number

[bigjoepops]

I am reading a 13 bit signed number out of an ADXL345.  It is contained in a high byte and a low byte.  I joined the two bytes into a UINT16, but for some reason I can't seem to find a way to get this number converted properly into a INT16.

 

I was trying the logical shift but that doesn't take care of it.  After I shift the number to the left by 3 bits, is there something else I need to use to shift it back but fill in the sign bit?  I searched this forum and the internet, but I couldn't seem to find a definite solution that worked.

 

Thanks for your help.

[ShaunR]

I am reading a 13 bit signed number out of an ADXL345.  It is contained in a high byte and a low byte.  I joined the two bytes into a UINT16, but for some reason I can't seem to find a way to get this number converted properly into a INT16.

 

I was trying the logical shift but that doesn't take care of it.  After I shift the number to the left by 3 bits, is there something else I need to use to shift it back but fill in the sign bit?  I searched this forum and the internet, but I couldn't seem to find a definite solution that worked.

 

Thanks for your help.

 

According to the spec

 

 

(13-bit) measurement at up to ±16 g. Digital output data is formatted as 16-bit twos complement

 

13 bit is the resolution, not the data format.

 

If you want to get sense out of the raw 13 bit register, I think you will need to calculate it via the following formula if its not already done for you.

 

g = (register_value-4096) * 0.00390625

 

where the "register_value" is the UINT16 provided by your high and low bytes.

 

That should map the 13 bit register onto the +16 to -16 g range..

Edited September 4, 2014 by ShaunR

[ned]

Shaun's probably right, but in case you do need to sign-extend, one approach is this: http://forums.ni.com/t5/LabVIEW/24bit-hex-to-2-s-complement/m-p/1686050#M599007

[bigjoepops]

I was expecting the sign extension to work on the part but for some reason it isn't.  I have a call into Analog Devices to try to figure out why this isn't working right.  I was hoping to find a work around while I wait to see if Analog Devices comes up with a reason why it isn't working.  Ned thanks for your post.  It worked.