conversion of phase from radian to deg

[Maitry07]

Hello support team,

I am capturing the absolute phase in radians with the 2QN format( FIX16_13 format) .
1 sign bit, 1 integer bit , 13 fraction bits

Now I want to convert the absolute phase from radians to deg. i.e
So, I am using multiplier IP core - xilinx to multiply the phase (2QN format) with 57 integer value( 180/Pi).
but I am not understanding on how to understand the multiplier output ? so, my question is weather I can directly multiply 57 with 2QN format and it will provide correct result or do I need to do some conversion from 2QN format to any other format?

 

[FvM]

You'll first define the target number format, e.g. 8Q7, then the respective fixed point multiply.

Fixed point libraries can perform the required shift and truncate operation for you. A bit more accuracy than integer factor 57 would be appropriate for 16 bit resolution.

I wonder what's the purpose of converting angle to degrees in FPGA?

--- Updated Jan 13, 2023 ---

Is your FPGA providing human interface, e.g. driving a display or printing numbers to an UART?

 

[Maitry07]

Hello,
Thank you for your response.
Actually My final agenda is to measure the relative phase in deg. between the 2 outputs. and this relative phase I will further provide to DAC to get the equivalent analog voltage for the measured relative phase.

 

[FvM]

If you send the output to DAC, why scale to degree? Instead you'll scale directly to DAC range.

 

[Maitry07]

ok, I got your point.
You mean to say, I directly figure out the relative phase by subtraction between 2 absolute phase ( FIX16_13 Format) and get the relative phase in radians.

And this radians result further I give it to DAC without any conversion to deg.

Is my understanding correct?

 

[FvM]

Xilinx cordic IP has unusual phase range of +/- 4 instead of +/- pi. Therefore subtraction result lays possibly outside normal range. I would normalize first to +/- pi and then substract. Difference can be send to DAC without further scaling, presuming a binary DAC input format.

 

[Maitry07]

Yes, I have already set the Coarse rotation parameter to extend the CORDIC range to full circle instead of +/-4.
Thank you for the response.

 

[FvM]

Why, expect you want normalized phase difference, but don't know your specification. Suggest to test your math with pencil and paper method.

 

[Maitry07]

Hello,
Let me give you a detailed brief.

I am using the DDC algorithm with 2 RF-ADC that take CW RF input( No modulation). DDC converts both RF inputs to I, and Q samples which I am further proving it to CORDIC vector translation mode to generate amp and phase.
CORDIC output phase is in radians. then I will subtract both phases to get the relative phase.

As my RF input is a pure sine wave. my phase range will be -pi to pi. ( full circle). both RF inputs will be having a phase difference which I want to measure. that's why I have set the coarse rotation parameter of the CORDIC.

I have verified the individual amplitude at the CORDIC output, which is accurate. now I am working on relative phase.

 

[FvM]

Coarse rotation option is in fact neccessary. In addition, you need to take care, that substraction is executed as modulo 2*pi operation. For example, if you have A = pi, B = -0.5 pi, you get A - B = pi + 0.5 pi = 1.5 pi, but the expected normalized result is -0.5 pi. By upscaling phase range to +/- pi, you get automatic modulo operation during substraction.

 

[Maitry07]

YES, I am using coarse rotation.