I'd really appreciate some guidance as I'm about to start a unique and interesting project.
Last year, I built a custom canopy for my car. It sits in front of a window, so it must be low. The car is made of fibreglass and must be fully covered in the winter - car covers are bad for paintwork. So, I built a lightweight canopy which wraps around 70% of the car. The canopy is raised and lowered using 2 linear actuators - https://www.ebay.co.uk/itm/303571732977. The linear actuators are controlled by wireless remote control - Linear Actuator Controller DC 12V Switch Forward/Reverse Wireless Remote Control | eBay
The problem is the two linear actuators run at slightly different speeds, around 50mm over the 1m extension. The canopy still works, but the extra load put on the gearbox of the slower actuator is shorting its life. You can get synchronised actuators, but they are costly at this size, so I need to find another solution.
My idea is to slow down the speed of the DC motor when one actuator is longer than the other and speed up again once they are both level. I've seen plenty of projects online to control DC motor speeds, so I don't think this will be too difficult.
The big question is how to detect when one linear actuator is ahead of the other? Digital spirit level, distance from the ground, other options?
You can purchase actuators that have built in feedback, but since you already have them, you could probably use an accelerometer to sense if the canopy is level, or two laser distance sensors to measure the extension of each actuator.
It's only a 50mm max difference though. Over the width of the canopy, I don't know if a tilt sensor will detect that.
[edit]
Assuming the canopy is 1500 mm wide, that 50 mm height difference is only 2 degrees of rotation. Actually seems odd that it would affect the gearbox.
Great idea! Unfortunately, I can't find a suitable sensor. Ideally, I'd pay around £50 max for the materials. I already have a bunch of Arduino boards and power adapters. Is this reasonable?
yeah the tilt switch would not capture 1.5° or would just be too sensitive as this is almost flat. I suspect the accelerometer would also struggle to be precise enough.
Your actuators seem to be on both sides of the canopy (I guess the poll at the back of the car is "passive" just to make the structure more robust. It indeed feels weird that a 1.5° would put so much stress on the gearbox.
could you install a piece between the actuator and the canopy so that it would absorb the angle along the way up or down? (ie something that can tilt a bit but still push/pull the canopy)
How about just adding some weight to the more lightly loaded actuator (presuming rear one pictured) so each actuator sees the same load? Then they should run at the same speed.
Took the words right off my keyboard
I suspect that the system is overconstrained, leading to one of the actuators binding. Adding some compliance like a stiff rubber bushing would probably solve this.
I used 4 linear rails mounted on very stiff posts to control the verticle movement. The bearing blocks (the bit that moves) are mounted on 2mm mild steel which provides a degree of flexibility to the system allowing for movement.
I've just replaced the broken actuator and the new one is considerably slower (120mm over 600mm). This is also the point where the resistance to too much causes the actuators to stop
I've just replaced the faulty actuator but unfortunately, the new one is considerably slower. Approx 120mm over 600mm. There isn't enough 'play' in the structure so I need to find a way to slow the faster one down.
