I’ve now got three creators, one which I bought three weeks ago (creator #1), and two that I bought last week (creators (#2 and #3). The first board has been working fine (though there were some issues running the compass demo when I first got it, but that seems to have resolved itself).
The second two boards have been misbehaving alot. I’ve done some systematic tests today to try and get to the bottom of the issues. I’ve used two raspberry pi boards, each with Raspbian 2016-05-10 Jessie, and the latest git pull of Matrix Creator HAL.
Having read the post on potential power supply issues with creator behaviour, I have kept my power supply the same as it been working fine with the first board. Please see the table below outlining my tests and results. Rows are test results, and columns indicate which demos have worked properly.
For experiment controls I swapped Raspberry Pis’, the SD cards in the Pi’s and the matrix creators across the Pi’s.
Looks like there is a difference across the SD cards. The PI’s were all bought at the same time.
Just before I wrote this post, board #2 has begun to work properly, without my doing anything different other than checking the tests again. I actually noticed that there is different behaviour of the D45, 48 and 50 LEDs on the top of the board. Can a techie please clarify what these indicate as they appear to give symptoms to the state of the board (perhaps the FPGA programming?). How can one tell whether the board has been flashed properly (at startup during the process, and afterwards)?
My first thought was that there is a potential issue with the manufacturing batch. As board #2 is now working, I am less convinced of this. I also wondered whether there might be a difference in the software that is installed, specifically matrix-creator-malos matrix-creator-openocd matrix-creator-init (I was following the instructions on the Wiki, which seems to have disappeared now).
It seems to me that there is a “breaking in” period for the boards. It seems odd, but I have noticed it with both boards #1 and #2 as initially some if not all of the demos didn’t function properly, and now they all do. Board #3 is still to work properly. I’ll leave it as a test case if you guys can provide some guidance.