Hello to all!

I have created FIR filter in Matlab (FDATool), and i have simulated how 
the filter is working.

In the first time i have used floating-point, and i have taken right 
result in output signal. But in the next time i have used fixed-point, 
and it is the result of simulation .

What is it? Errors of quantization? And how i can fix it?

But i need fixed-point in realisation of FIR on FPGA.