Hi, this is an alert to users who recently ordered OpenSprinkler Pi: we’ve found an annoying problem with the 8-pin ribbon / rainbow cable that comes with the package. Apparently many of the cables have disconnected crimp wires, causing the OSPi to not respond to RPi, or even worse, causing all stations to turn on at the same time. I estimated that there are potentially up to 60 orders that might be affected. We’ve proactively taken actions to send a replacement cable to everyone who ordered OSPi recently (within the last 2 weeks). If you are seeing an issue of your OSPi not responding to RPi, try to use a multimeter to measure the connection on each of the 8 wires of the cable. If any wire is not connected, please email to [email protected] and we will send you a replacement cable right away.

Here is a closer look at the cause of the problem. When I take the wire out of the crimp connector, you can clearly see that the good ones (shown on the left below) have metal wires tightly secured by the crimps, while the bad ones (shown on the right below) have the metal wires cut from the cable, resulting in disconnection. This is probably due to a defective wire stripper or crimp tool, or incorrect operation that causes the metal wires to be completely cut through. In any case, I’ve informed the supplier and complained about the the quality, and will never order from them again.

So the lesson to learn here is that when testing OSPi, we really should use the individual cable that goes with each board, instead of using a common testing cable. Another lesson to learn is to never overlook any part, even something as simple as a cable!

I became aware of the cable issue when a couple of users reported that their OSPi did not work. This is very puzzling because every single board has been tested and verified, so the chance of DOA is very small. I’ve asked two of them to send their boards back to me, and after testing, I couldn’t find any obvious problem — the board works fine, and all demo programs work fine. However, until yesterday, I have apparently overlooked the problem with the cable: when testing, we have always used a common testing cable, instead of the user’s cable. Partly it’s because I never thought the cable would have any problem at all. Now I’ve learned a lesson, and it’s a good lesson to learn!