Why it occur overflow??

I modeled simulink for hdl coder conversion and i used a fixed point tool box. But it occur overflow like a below figure. You can see the below figure, fixed point tool has simulation result from 0 minimum to 32.958984375 maximum value. But i applyed data fixdt(0,16,9). It can cover sufficient range from 0 to 32.958984375. But it occured overflow when i run for pressing button of 'simulate with embedded types'. So i applied it on a trial about range fixdt(0,20,10) (I thinked that it very sufficient range). However, it more increased overflow. why it occur overflow??

Answer 1

[Kiran Kintali]

The overflow behavior is based on what kinds of operations are happening on the variable. Is it in a feedback look like a for-loop where there error is accumulating and causing overflows etc., Please share your code (design and testbench) if possible for better guidance. It would hard to diagnose from just a screenshot.

[Ki Hyeon Jang]

Always very thanks ~

This is my simulink code for hdl conversion. I just applied a data type of fixed point tool as a default word length 16 and proposed fraction length. And test bench data is using 'image from file' block. Because 'from multimedia file' is not supported fixed point. But i want a various input image for simulation range.

So i wonder another thing. That is 'is there method of various input image for simulation range?'

[Kiran Kintali]

Hi Ki, bmp file is missing; it will be hard to figure out ranges without a proper testbench. Thanks

[Ki Hyeon Jang]

oh~ sorry. I upload a bmp file. Ok I agree your opinion. So i tried various things because i want to figure out range(i think that image should input 'all white', 'all black', and 'general image'), but i don't find proper method. you have any idea of testbench for extract various range??