I've attached some photos. A couple are a bit out of focus.
These show:
- The prototype mounting brackets for the sensor that I 3D printed
- The sensor and magnet installed on the steps and the body panel of the RV.
- The PCB being tested on the bench.
- The PCB mounted in the box with the relay attached and the plugs.
- The box installed in the battery compartment. The insulation is there to minimize any noise from the cables hitting the side of the box. Probably not needed but I haven't removed the insulation yet to test.
I also milled a bracket out of aluminum, but the 3D printed one is holding up quite well, so there is likely no need to use a metal one.
I'm still kicking around the idea of putting together a kit with detailed instructions on how to install. The challenge is that it is hard to know how much variance there is in the spacing between the step arm and the body of a given RV, so it may not be practical to have a common bracket, and that might be something that each owner would need to fabricate.
The principle is pretty simple. Use a limit switch to detect when the step is fully extended and locked.
The controller is powered from the motor power lead that is +12V when extending the step. The relay is inserted in the motor lead after the +12V to my controller using the normally closed contacts.
When the sensor detects the step is fully extended the relay is energized which opens the circuit removing +12 to the motor while leaving my controller powered.
When the Lipper controller times out it removes power from the motor circuit which turns off my controller. The relay drops out which will then apply power to the motor on the next cycle.
When the step retracts the previous +12V motor lead is not ground, so my controller is not powered up and the step retracts normally.
This solution only works with the case where the step has an issue on extension. In all the motor assemblies I tried that was where I had the problem. I have some theories on why it is only on extension and that is based on tearing down the motor and gear assys.
I also have a design that can work with both extension and retraction but I haven't tested that yet.
Based on discussions with Lippert support, and assuming they were correct, the issue appears to be that they have a specific stall current that they are looking for that they use to determine when the step is extended or retracted. In typical automotive window applications the controller is looking for an increase in current draw from normal and not a specific threshold. Because of tolerances and the way the plastic gear is formed when the steps extend it is possible for the gear to skip so the motor never hits the specific threshold current.
The approach I took resolves the issue while leaving the existing controller intact. And since everything is plug in it can be easily removed if needed.