There was a fresh load of robots delivered last week for the next few upcoming projects and I spent a little time today looking through them for ideas. I had in mind an idea for linkage like that is used on backhoes and front end loaders to keep the bucket from tipping from it's position when raised and lowered and found the same general idea on one robot. A few quick measurements and it looks simple enough...

As far as encoders go the only thing I could find is what is built in to the motors. The connectors and pinouts seem to support this from what I could find out. I looked through Fanuc's website and couldn't find any encoder listings in the parts sections for the models in the shop. All of the motors were listed. I'm thinking the positions are based on a home position and motor revolutions. It's possible there are some with absolute encoders, but I don't think we deal with any. They do list a pretty impressive .2mm repeatability though which is probably better than I will be able to model.

 

Fanuc have traditionally used a mixture of both absolute and incremental, for their robots it has always been absolute variety as far as my experience.
Their commutation pulses for their AC servo motors is a unique method producing a sine wave extrapolated from a 4 bit binary signal.

 

Fanuc have traditionally used a mixture of both absolute and incremental, for their robots it has always been absolute variety as far as my experience.

I may be mixing terminology, but I thnk I understand what you are getting at. I was taking absolute as an encoder at the joint which would then report the exact position of the joint itself.

It seems like you are saying absolute and incremental in terms of programming instead of an encoder that outputs say a 32 bit value for it's actual position.

 

Incremental encoder is generally indicated to be the 90° quadrature version. dual pulse type.
I should have elaborated.

 

So after blowing off housework that needs done and scouring the internet I decided I like this idea for the gearing It's a bit thinner than most harmonic gearsets

I may still try the brushed DC motors with an encoder wheel on the input side of things. I'm trying to stay low budget and I already have a handful in my parts box. It won't be hard to design a coupler that can be replaced to fit a stepper if things don't work out.

So now it's time to define some specs, dust off my high school algebra, trig, geometry, and physics memories as well as some kinematics research, and figure out what I need to do to get it done.

I thought this was going to be simple until Irving mentioned backlash... one very critical problem I never thought of that changed everything.

Hope to be back in a few weeks with something.