Day 24 (3rd July 2019)

Today I fixed the 6th ZipCPU tutorial so that it now passes formal as well as simulation, one thing that’s missing is to try it out on real hardware. I did some work on getting it to work on hardware today - however the issue is that I don’t have a USB-UART converter handy with me, and I consider it a good thing because it’ll now turn me into the direction of implementing USB-UART on the FPGA directly :) . Turns out there do exist USB-Serial implementations on the TinyFPGA BX board that I am using - see [1] and [2].

Implementation [1] seems to be well documented but it runs off a single clock domain of 48MHz. [2] has two clock domains, one for the CPU and rest of the logic and the other is the 48MHz domain which runs the USB protocol engine. From a design perspective my feeling is that [2] is better but [1] seems to document that it works really well with continuous data flow, and I seem to incline towards [1] personally. It also seems like a good way to be introduced to clock domain crossing, let’s see where this detour takes me!

Next I have the debounce tutorial by ZipCPU (the 7th one in the series) lined up, and after that I have an idea to combine it with a rotary encoder. This should be fun and interesting to implement, and I think this is enough stuff in the pipeline to cover this week. Let’s see.