Unwinding Rotary Axis for Homing

More
1 year 3 months ago #332539 by jschulze
I've considered picking up an absolute encoder for it. I already had the servos, so I was just trying to use them up. Each axis has a 100:1 harmonic reducer that's belt driven. The belts made the package a bit smaller and allowed me to route the cables through the center of the gear reducers. I'm not concerned about unwinding during a program. I just don't want to shut down the machine at say 400deg and then add another 320deg by always homing in the same direction.

I've got a half baked solution in mind. A while back I made a tool changer and spindle homing procedure from the home all button by calling a halui mdi command and then using g code and I think I might be able to do something similar for this.

Please Log in or Create an account to join the conversation.

More
1 year 3 months ago #332542 by jimmyrig
Few ways of doing this.

The way I handled it on a 4th axis plasma build was with a magnet on the rotary and a hall effect sensor sensor to read it. Then sending the rotary back to 0 at the end of the gcode via the post processor.

It's not fool proof (power failures) but it works well enough as I just drive past the home switch then home again the one in 1000 times I have a power failure during running

Could use a 3d printed gear reducer to get a 1:4 rotation and put the sensor on that....so 720 turns into 180.

Please Log in or Create an account to join the conversation.

More
1 year 2 months ago #334151 by jschulze
If anyone stumbles on this looking for a solution I was able to work it out in hal using wcomp and some and, or, not components. Basically I used the wcomp to check the startup position from the last_position.txt file, then using a bit of funny logic I made it so after the joints hits the home switch in the positive direction, the wcomp keeps the home switch closed for the remainder of the positive direction travel, then in the negative direction I used the min wcomp output to disable the home switch input for the remainder of the negative travel. So the home switch only properly works in a narrow joint position range. With this setup the normal homing routine does exactly what I need. When the joint is in the positive direction it backs off until it's in the proper range and in the negative direction it ignores the physical switch until it's in the proper range. There's some silly logic with some and/or/not comps which I'm sure could be made more elegant, but it does work.
The following user(s) said Thank You: jpg

Please Log in or Create an account to join the conversation.

Time to create page: 0.337 seconds
Powered by Kunena Forum