I got my TM1637 display today. 😃

I wired it up to my Arduino Uno, checking with a continuity meter that the connections were correct. I have learnt not to trust looking at the pin positions because it's easy to see off-by-one. I then ran the test sketch in the blog article which implements a 4-digit counter. It works great. This provided a baseline so I knew the display worked and the serial protocol was the one published.

I wired it up to my breadboard, again checking the connections with a continuity meter. A mistake like swapping the power lines, would blow the display and set me back a few weeks, the time required to receive another display from eBay. And I would be kicking myself.

Turned it on and got scrambled segments. ☹️ So I went back to the bit-banging code and compared my assembler code to the C original. I tried simplifying the code by commenting out sections. One thing I tried was to set all 7 segments on. However the result was 6 lit up and this was different for each digit. Aha, I must be missing one data bit in the serial transmission. Looked at the code, and sure enough, I had missed one final clock transition in my code, all of two lines of assembler. This is always a hazard when translating code from one language to another.

I took the usual celebratory photo which you can see in the Details section, showing 12:35, not long after switch on, as the initial time is 12:34:56 (seconds not displayed). 😃

The time setting switches don't work yet. I have made some large changes here from the plasma clock project so that's not surprising. I'll have to go back to the simulator to debug this. But enough for the day, better stop while I'm ahead.