This seems close to what I have been describing. What I added to this was:
If the second signal arrives within this delay (of 1 second), then the relay can extend the winching time by a second from the moment of reception on. So, under good conditions, repeating this logic, this would lead to a continuous winching till the valve is raised.
It would be good to have some verification that the time between the UI button pressing and the relay switching is checked. It cannot be that the command is received 10 seconds later and is still accepted and the valve is raised. (seems to me)
So I’m currently looking at the functionality of software on Sonoff style remote relays, to find out what functionality can be implemented on it. Then the UI on Android and Apple. That seems like a harder problem.
It’s to stop you guillotining something or burning out the motor if something gets pinched in the end seal. Too much mechanical stress, the weak link breaks. It’s a last desperate precaution, like an electrical fuse.
If your winch is designed to be a winch, it probably already has something like a weak link / shear pin in the drive path already, something that snaps at ridiculous overload. That may or may not be too high a load for your use.
Ross, I don’t know about the electrical safety features of the winch. As you point out, there may be an electrical protection against overload. On the mechanical side, the current installation already has the following protection. Suppose I park with the wheels of my car on the valve. In this case, the winch will start winching. The valve won’t move. Instead, a weight below the winch gets lifted and when it reaches the winch, it activates the mechanical off switch. This is the same weight-switch combination that in normal conditions ends the winching when the valve has reached its vertical position. Then the weight is also lifted and the mechanical off switch activated.
To me, this system looks close to what you suggested.
No exactly. In your describtion you press a button once, and the system is repeating the signal until you release the button.
My idea is that you have to press the button each time to move the winch another second. The loss in comfort would be a win in security. Because the winch only will move if all communications work as implemented.
And you can use any smart switch, because you only need one channel (UP).
To put the winch down again, you are onsite and can press the regular physical buttons.
Hi,
The way I would do that system is to use any roller shutter controler that does have proper end switches support and can take % as position as input ( I use fibaro z-wave for my rollers) but that could be something wired. Preferably taking MQTT as input.
Then have a separate system like small RPI running motion detection on the camera with well designed zones that would supervise the operation of opening and would be sending a stop command to OH in case any movement (except the barrier itself) would be in the frame.
Like one extra variable inthe OH dedicated to the status “ok to continiue operating” when no motion around the barrier is detecetd for long enough.
Maciej, thank you. Both your ideas look totally applicable to my setup. I already ordered a Foscam 4mp camera. I am going to try out motion detection. But also human presence detection. I have no idea yet how reliable this can work… I’ll see.
