Also think of a place for a single or 2 mosfet switch. It is useful for switching on auxiliary valves, and in standard relays the contacts burn out and noise occurs.
I use a mosfet to switch the 6-way valve, and if it is an OC system, also an additional section to redirect the stream from the pump
I see what you mean. It could be useful indeed.
Could this idea be extended to the section control, in place of relays ? Is there people who thought about that ? Are there sprayers or seeders or other tools that can’t work with MOSFET switchs for section control ?
1 Like
It’s not easy with precision air seder. In the factory controller, switching off a section is a single button that subtracts or adds a section with another press. I don’t know if AOG could control these buttons ??
Disconnecting the motor of the section, the controller, not seeing the pulses from the encoder, displays an error and restoration of the power supply does not = start the engine.
I was thinking to simulate the encoder during AOG engine shutdown and thus cheat the factory driver. Maybe someone has a better idea?
Connecting the EHR switch is also not easy in my MF. Lowering, Raising, Neutral - Each of these switch positions gives a specific resistance pattern on the switch pins. A simple relay does not work.
1 Like
All right everyone, as promised, here are the results of the poll I made a few weeks ago. I had 24 answers, which a nice sample I think. You can still answer though.
Based on your answers, and also on what I need from this design, here are the colclusions:
- Most people think the idea of an all-in-one PCB enclosed in a rugged enclosure is a good idea. Phew!
- Most people think it would be at least nice to reduce the amount of wires needed to setup the autosteer. By grouping functions on a single board, this issue would be at least partially addressed (there is not much to do about wires across the tractor though)
- People are more divided about the digital connection means. I must admit that the question was not really clear, as it partially mixed considerations about connection and about processors. In conclusion, I thing I will focus on USB compatibility with an embedded USB hub. Ethernet is nice and has a lot of advantages (see here and here), but more complicated to implement and makes less sense on a one-board setup, in my opinion. Also, I can’t really keep both options available, I’m afraid it would be a nightmare to design.
- Almost perfect split between people who want one or tho gnss antennas! I will do my best to make these two solutions available on this PCB, using the code from MTZ8302 on ESP32.
- Most people think antennas should be removable, probably in a cost reduction approach. It doesn’t change much for the PCB, but good to know.
- Most people don’t know what’s the best position for MMA and BNO. From a physics point of view though:
- MMA can’t be impacted by metals. But it should be horizontal and maybe not too high to avoid being tossed from left to right in the field.
- BNO is used as a magnetometer (or compass), so it makes a lot of sense to put it away from metallic surfaces and outside of metallic enclosures. Kaupoi tried this anyway and, against all odds, reported it worked fine. So at this point I don’t know nothing no more! All I can tell is that i already put it on the roof once, together with MMA, both connected through 3.3V I2C with a 3m long twisted pairs cable (details here). Works fine.
I’ll output I2C on the conenctor, probably use an I2C extender like P82B96, and anyone will put MMA and BNO wherever he wants if he chooses to build a single antenna setup.
- Most people are not afraid by SMD components. It doesn’t mean they will solder them by themselves though. I’ll do what I can to source components from the JLCPCB parts library, but their catalog is not that thick so it may not be entirely possible. It was also reported that they had issues with their stock inventory. Other PCB manufacturers / assemblers can work with components you send them (like PCBWay), but the price will be higher (I didn’t do the math though).
- As said above, I will embed a USB hub on the PCB, which should satisfy a lot of people. Sorry for Ethernet enthusiasts though…
- I will output a 5V to recharge a phone, and also an adjustable DC power supply for a tablet. I personally find these useful, and a lot of people believe the same.
- Most people think the relays for sections control should be outside of the box. I didn’t expect that! But that’s understandable, as you probably are looking for modularity here. It would be a shame to have to carry a bunch of unused wired for section control when ploughing or mowing. But what if I told you that it’s possible to have 8 relays on a separate 18 pins connector, but on the same ECU box ? Damn, I don’t know about you, but that looks awesome to me! Also I’m pretty sure i can fit them inside the enclosure. I will also output USB on the other connector for people who want their own separate box, with more than 8 relays for example.
- Hydraulic lift control shouldn’t be an issue. I may use a solid state relay or MOSFETs though, for their longevity, as discussed just above with baraki.
Now, let’s design!
8 Likes
Does someone have at hand a Cytron MD13S driver ? I’d like to know the references of the driver chip and of the power MOSFETs, as they seem to work very well. Thank you !
If the driver would be integrated to the PCB, could we also think of adding current sensors? This way you could run propo valves “correctly” in current control instead of direct PWM.
1 Like
@redalert11, yes it looks like it. Thanks
@nut could you tell me more ? I don’t intend to modify neither the Arduino code, nor AOG, but your feature looks like it needs that. Do you know more precisely how it should be implemented ?
The idea I had was more add the current sensing capability as an option, so you could still use the standard AOG and ino, but if the driver is integrated to the board you could just use current sensing mosfets and add the possibility to measure the output current with an opamp from the sense line. You don’t need any high power resistors to do that. Then if you want to build current control for a solenoids you’d have the sensed current (or a voltage proportional to that) connected to the analogue inputs on the arduino. Just on “should work in theory” level, though 
Some schematic I found with google:
1 Like
@Valentin: More and more devices come with USB-C for power and data. You may combine both to one USB-C cable, if you integrate something like this:
I would use preassembled boards for the charger and the DC/DC. You cannot buy the components for the price you get the whole PCBA…
Examples:
Car charger USB-C
DC/DC
Mind that I’ve shown a USB-C plug with reduced pin count (easier to solder…)
Hi,
Maybe include a possibility of CANBus?
Even if at the moment it is not widely used, surely more and more people will want to use it in the future.
1 Like
@Tooki57 I read about CANbus somewhere else in the forum, so yes people may be interested, but I don’t know much about it. Do you know what it could be useful for ? Also, did you already design something with it, related to AOG ?
@GoRoNb lately I stepped away a bit from the project. I’m getting back to it, and I’ll keep you posted.
What you suggest is to use the USB C capability to recharge the tablet and connect it to the USB hub with only one wire, is that right ? Provided that the tablet bought is compatible with USB C recharge of course. I agree that it could be useful, even if I don’t personally have a need for it (I use a Surface Pro 4 in the cab). I’ll think about it.
Where I don’t see things your way is about pre-built modules. They may be a bit cheaper when you scavenge them form Chinese products, and quick to work with, but this comes with a great deal of drawbacks in my opinion. It’s more bulky than discrete components on the main PCB. It’s also harder to interface, the bottom side may not be flat, and you’ll certainly need flying wires. All in all, the system would end up less reliable, less solid and less professional looking, especially if I follow the same reasoning for other features.
Hello ,
Yes I use CanBus on 2 tractors. I use it in the Uturn functions instead of the relays to control the tractors.
I use MCP2515 modules, but it shouldn’t be complicated to integrate it into a pcb.
I also see a benefit in putting things together on a PCBA. So integrating CAN is interesting (you need an SPI + one chip select per CAN). Müller/Trimble has a system, that steers just with GNSS and CAN. Back to the PCB: There are components like 12V => USB-C PD, where you have no chance to integrate that for an acceptable price. That’s why I suggested to use such a converter and feed it to the PCBA. Does all the CC line handshaking and power conversation for you. The concrete reason, why I asked about the project, is that @Aortner is trying to build a group of people pushing AOG right now
1 idea … :
Maybe an serial out from nano to go to HC06 bluetooth ?
@GoRoNb being able to power the computer/tablet through USB C would indeed be really useful. If scavenged boards can do it and cost pennies, I guess it must be possible to implement it ourselves for a reasonable cost. I don’t know enough about the USB C universe to be sure about that statement though, but there seems to be many branded chips able to take care of these things. Anyway, it all comes to weighing the benefits and drawbacks of each solution, and compare them to what we are trying to build with this PCB : do we want something compact and reliable, something useful with all the features we can imagine, or something as low cost as possible ? Certainly a balance between these. I’d be opened to discuss of these topics with you @GoRoNb, @Aortner and anyone else that want to be involved.
@Tooki57 like with CAN bus, do you also have ideas and examples ow use cases with the HC06 bluetooth module ? How useful is it ? How reliable is it ?