Hello,
It turns out I was measuring the wrong pin.. Haha the clk_out is fine.
My input clock is fine, I have created a very stable clock on it previous to this experiment.
I will make sure to change my ports to STD_VECTOR.
I had removed the when 21 case to wrap back to 0 thinking it might help so I'll put it back in.
Could you please expand on your third suggestion about signals? I think this means using temporary variables inside PROCESS, and then assigning the variable to a port at the end of PROCESS.
Other than that I think I need to keep on re-reading the data sheet and simulating, instead of burning.
Thanks