So that I can do some testing, I’ve placed some obstacles in the garden which I can leave in place and see how the modifications I’m making affect the performance.
I was finding that on certain obstacles the wheels were just spinning and no progress was being made. To try and overcome this I printed a knobbly tyre but this made things worse. However, taking inspiration from the rock crawler world, I printed off the existing tyres but with much thinner tyre walls. This has resulted in softer tyres and made a huge difference to the grip:
It has been suggested that the final wheelchair have a pump which can deflate/inflate the tyres as required. This would be extremely useful but it would require consideration of the benefits, weight, and power requirements.
All of the new “off-road” tyres and (as requested by Ada) purple wheels are on the car. I’ve also been able to get the motors to stop slipping which means their full power is now being transferred to the tyres. These changes have resulted in an improved ability to overcome obstacles and the prototype is now able to get over things it was unable to with the previous setup.
This video shows the car overcoming some relatively large obstacles in the wet.
I think this is pretty impressive although towards the end of the video, as the “wheelchair” comes down off the last obstacle, you can see that it’s in danger of tipping over. If this was full size it could result in the full weight of the wheelchair on top of Ada. Another problem which is happening quite often is shown in this video:
When the back bogie wheel (which is the middle wheel) has traction when all other wheels don’t, the bogie (or front end) turns upside down. I don’t know what the weight of the full size version will be and it might possible to turn it back around by hand (and obviously in a full size version of the wheelchair you’ll have more control over how much power is being transferred to each wheel) but I think it would be prudent to try and eliminate the problem completely.
Also, although the differential which keeps the seat level is working brilliantly, I think in a full size version the current method would be under substantial strain and would need a lot of maintenance. It also dictates where the seat and batteries are positioned as everything has to fit around the differential.
My to do list for the current prototype includes putting a limit on how far the bogie can pivot, investing in motors with more torque, exploring other mechanisms for the differential, and adjusting where the centre of gravity is positioned in relation to the wheels to see what effect this has.
In other news, parts to make a custom remote started appearing in the post today. More parts yet to come though.