Gallery, Projects and General > How to's
get a grip
vtsteam:
That reminds me. I once had need of a very large gear to mesh with a toothed belt, and didn't have a dividing head to make one.
I just turned an aluminum blank to a tight fit with a toothed belt turned inside out. A small amount of adhesive (though the friction fit was more than sufficient) and I had my "gear". Another toothed belt meshed with it perfectly. This was driven by a small purchased belt gear mounted on a stepper motor. It was part of an auto-sampler I built for a Cary spectrophotometer.
dsquire:
--- Quote from: vtsteam on October 11, 2013, 06:38:02 PM ---That reminds me. I once had need of a very large gear to mesh with a toothed belt, and didn't have a dividing head to make one.
I just turned an aluminum blank to a tight fit with a toothed belt turned inside out. A small amount of adhesive (though the friction fit was more than sufficient) and I had my "gear". Another toothed belt meshed with it perfectly. This was driven by a small purchased belt gear mounted on a stepper motor. It was part of an auto-sampler I built for a Cary spectrophotometer.
--- End quote ---
Steve
That is a super idea. I can see all kinds of possibilities with that idea. I have seen a variation of it used as a rack and pinion on CNC router tables being built on the CNC forum. Built and adjusted properly there would be no backlash. :thumbup:
Cheers :beer:
Don
vtsteam:
It worked very well. I had that ring mounted on an aluminum plate mounted on two drawer slides. That was also driven by a stepper. So I had a linear and a rotary translation. A plate of optical filters under test was placed on top of the rotary ring and a light source on top was triggered when the table translated to each filter position. It was very simple. And very cheap. I built it in 2 weeks. I used turbocnc to drive it, using G code as a program to drive the sampler. That ran off a dumpster computer with a bad hard drive. I ran the whole thing off a floppy.
This amazed the company that had just hired me. They expected to spend $20K and several months to develop this -- they had a backlog of filter orders, and this automated testing to help them catch up. I think that with the controller (hobbyCNC kit) the total materials cost was about $250.
They made me engineering manager after that. A mistake I later had to correct.
Navigation
[0] Message Index
[*] Previous page
Go to full version