Would like some thoughts and discussion on this board design. It would be the new v4 pcb. 16 channels of relay line level control.
1 Like
Is the Cytron still on the left, just over top of the board wiring that is shown
Yes, above the MCP23017 which converts I2C to 16 channels. Had to make the board longer 
but it does a lot now
What do you mean by line level control of relays? Still another relay board and nano or some other setup
1 Like
2 ways to drive a relay, with a line level input or drive high current to pull the coil in.
Could add the software for section control and hydraulic lift to the autosteer board
1 board to rule them all
2 Likes
I really like the idea of one board and software all in one to run relays. A larger board does not bother me at all, I actually like that
1 Like
I think single board design sounds like the way to go. 
Guess it will still be an option to use the machine controller to have the section relays on the implement instead just in the cab.
Is the : “To Auto Steer: Machine PGN” button in the module configuration like a switch one or the other or will both work?
Or another option to have a separate switch box for the physical switches and use the machine controller.
IT just decides whether or not the PGN for machine control is sent to the autosteer port. For UDP it doesn’t matter since everything is already everywhere
1 Like
Thanks, i know the udp network, should express myself better.
When on usb its either autosteer or machine and not both? Like could we use the autosteer on usb and forward it from autosteer to machine over ethernet if desired when we need longer cables than usb allows.
Maybe an other idea, a separate relayboard, stackable on top of the autosteer PCB, for those who already have PCB v2
Example: These are some pcb’s made by krone. The relay part (solid state) is also stacked on top and replaceable
If you see in the PCB folder there are 2 boards, and that was kind of the idea, stack them. But buying prebuilt relay boards is far cheaper then making your own so that makes it a bit tough to decide what is best.
Hence the idea of just the drive on autosteer board, relays can be on their own board.
When its usb you can send machine and autosteer pgn’s to autosteer. by checking the box. Therefore only 1 usb port to do everything
1 Like
Ok so it works well. I have the RelayPCB that’s in the support files, and its I2C is connected to the autosteer board. Super easy to add code to control relays. Just need to decide what the best way to go is. All separate, MCP on autosteer, nothing on autosteer but just a connector to I2C, not sure what is best.
My preferred option would always be separate module, with UDP.
My thinking is that the autosteer board belongs with the tractor…and the section relays belong with the implement (wether it be on the implement or in the implements oem control box)… as many different implements require section control, and imo for the cost involved, its better to just add it to each implement, so its just plug and play, UDP in, choose implement in AOG and go…
I guess it can also work by having it in the tractor in the autosteer board, with a universal plug that interfaces with different implements, but for me, I prefer the idea of a single UDP cable (ethernet) or even wifi for each implement…much like ISObus / canbus in the OEM world?
Thoughts?
2 Likes
I have the impression that the condensators pull a lot of power when switching on. The i2c-bus so near to this “wirelines” isn’t a problem? I have no much knowing in this stuff, but this was my first thougt when I saw the pcb. So this is more a question than an reply.
Another thing is, I know Brian does not like long i2c wires, but I would need a screw conector for the i2c bus. Because on the search for good and relativ stable roll values with mma I tried three different positions in tractor cabin at my last setup. This is not doable without a cable. There, where I mounted the mma now, I can’t mount the whole pcb.
I hadn’t thought of having multiple implements with different relay setups so not a bad idea to have a relay board per implement. It’s the hookup, programming and implementation of each different one that causes a rookie like the problems
1 Like
Hi,
Why not change the game and using a single arduino mega board, making a pcb to fit a Mega. With a Mega there will be plenty of pins for in the future, those can be used for switches and LED or so. Also three uarts, for example a wireless communication to a module on the planter for monitoring it and so on.
but at the other end its maybe better to have two boards talking to each other for reliability.
I don’t know, just my toughts 
1 Like
Well i don;t mind long wires, the problem with I2C is capacitance. The more you put 2 conductors side by side the more capacitance there is. In the devices, the pcb, connectors, it all adds up. The limit then is 400 pf at 100khz. Pico is 4x10 to the -12. Pretty darn small. If the I2C locks up, it stops the nano dead. so that is the concern.
However, been playing around with lowering the i2c frequency, the latest video i did at 25Khz i2c speed.