Research

Lawna has been a stalwart machine on which I tried several different motors. I generally like to keep old projects running while I build the replacement but in this case I'll make an exception. Lawna has several good motors and wheels which I will cannibalize and use on the new machine to keep costs down. The prototype has served its purpose. RIP Lawna, June 2014.

I had a fairly good idea of the cutting motors to use based on experience with Lawna. I will use the same wheels so the bottom of the mower will be 50mm from the ground, giving a cutting height of about 30mm.

Mower Blades

So here I am experimenting with blades again. Ideally, we want to use the same geared motor that was used on Lawna as this gave good torque with a low power consumption and reduced noise. The downside was the speed of rotation was lower, about 2100 rpm whereas the raw motor can run at 13000 rpm or more.

This mower needs to cut a 30 cm strip of grass. Initially, I tried to use the same blade arrangement I used for lawna, only extending to be a 30 cm blade. I tried some mowing using this arrangement. It did not work out well. See the results below.

As can be seen, the blade snatched on the grass, stalled and bent. I could use a more substantial blade but the fact that it stalled means that I would need a higher torque motor which will mean more power consumption.

Next I decided to experiment with a smaller blade, with the idea of using two or three of them. I want sharp blades on this mower so made up a prototype using Stanley knife blades attached to a plywood disc. The disk was 120mm and the blades protrude 20mm each which give a cutting width of 16cm so I would only need two such blades to achieve my 30cm target.

A made a small jig up using the intended cutting motor that was mounted at the same height above the grass as the robot. I ran wires to the jig, placed it on the grass and powered it from a small 12 volt battery. Results were impressive, it cut the grass easily without stalling or clogging. I may need to iron out some of the physical mounting issues, but I believe this looks like a good solution.

June 2014