Open pylogger-html opened 2 years ago
Hey, Thanks! Good thoughts! POE is cool, but I'm currently not sure if it's worth the extra trouble right now. The 12W of the Ag9912 and AG9924 is (maybe) a bit on the low side of power delivery I guess. And I'm not sure how small the design can actually be for my power requirements. On top, I think it is a good thing to be able to run it from standard routers at home with no POE injectors etc. But I will take a look at that in the future. :)
A four-layer board is surely the way to go in the future. This board was somewhat of a learning project and I stumbled a bit into the current design. Thanks for the tip!
Actually... the POE module might just fit on the top board :D
I too am interested in POE :) Although i'm also familiar with feature creep in prototype designs and how it can distract for the core goal. Great work so far, looking forward to updates.
Another +1 for PoE. I'm working towards automating my house and I'm keen to be able to make blinds/curtains/shutters move. I can get PoE to the windows but finding PoE blinds in the UK is a nightmare.
Really finding PoE for most things smart home is awkward. My current plan would be to use a PoE splitter to extract the power and then something like this project to control a stepper. Obviously this won't be the smallest or cleanest solution. Otherwise my options are something like this olimex board and a stepper driver board.
Re crowdfunding/financing, if there was a nicer solution that could interface with home assistnant I'd probably be after 30 or so. Then I just have to find the blinds.
poe++ will deliver 70W and instead of doing DC-DC conversion for all power maybe you can use that poe voltage directly for motor and step down only control voltage. Naturally you will need high voltage driver chip for that.
Really nice project. I've been looking for something like this for a while.
I know you're short on space, but did you consider adding POE support? You can get modules that do most of the work for you (Ag9912 or Ag9924) and you might be able to squeeze one in between the two boards with some longer standoff connectors.
The other thing I was wondering is why you didn't use a 4-layer PCB. The cost increase is fairly minimal these days (looks like jlcpcb adds ~$2-$3 per board for qty=5, but most of your cost is in components) and your signal routing would be a lot cleaner. For signal integrity it's always helpful to have one continuous ground plane layer, although you ended up achieving almost that on the flooded back layer.