[ger21]

A Low inductance Nema 23, with a good drive, and high voltage, can be usable at 1500+RPM.
But 1610 screws are a much better choice.

And 100ipm is quite slow for a 48" machine. Ideally you'd want it to be capable of at least 250-300ipm rapids.

[Jandyman]

This is great. I love learning new stuff. FWIW, the whip calculator says I could run the 16xx at 1033 rpm, so if the motor could run 1000 rpm, that would give me 400 ipm which is more than enough for rapids. Questions I need to answer - a) are my Nema 23s low inductance and b) what about my drive? It's a Geckodrive 450, but IIRC, the power supply isn't super high voltage. But that could be fixed. I will research. Ball screws looking good again.

[routalot]

The topic seems to have moved on a bit from DOC to attaining rapids.Might I suggest that rate of material removal be considered a factor?With the rigidity and feed available with a ballscrew you will be able to increase the DOC while holding a more accurate position and should be able to complete the cutting routine in less time.With the cutting process described thus far as producing the outline of a guitar,I don't see the machine needing to spend much time at rapid feed rate,other than moving from the home location to the start of the cut and then back again at the end of the job.

Whatever you decide to do,I suspect the comment earlier in the thread about constantly bumping against the next week point of the machine will be proved correct.The good thing is that they will involve ever smaller deviations from the intended design and you may reach the point where the current machine is good enough to make some parts for the next machine.

[Jandyman]

Everything you say makes sense, Sir Routalot. I just need to get to "good enough". The stuff I've learned on cnczone in the last two weeks, along with concomitant internet research, leaves me so much better prepared than I was coming in. That will serve me well. It's going to be a process, experimentation is unavoidable. I don't think it will be too hard to get to a point where I can make parts to help upgrade my current machine, given its design. And maybe THAT will get me to a point where I could make parts for a more rigid machine and one matched more exactly to my application.

[wizard]

I already did that. The belt stretch is so dominant, it is difficult to tell what else is going on. That's true in the X direction as well (gantry moving along the rails).

I’m having real problems with this post. I can’t imagine any human stretching timing belts by applying force to them. At least not timing belts of a size reasonable for a router. I work on devices with timing belts as narrow as 5 mm to larger than 150 mm in width. The general failure mode is for the belts to break. You can get some stretch in a newly installed belt but that is dealt with by retensioning the belt.

Understand that I’m not attempting to argue just that I’ve simply have never seen belt stretch to the extent you describe.

[Jandyman]

It surprised me too. But a couple days into examining design specifications for belts, I concluded that my eyes were not deceiving me. Remember that the amount of belt stretch is proportional to the belt length. And in this case, the situation is exacerbated by the design. By using an idler design with the motor on one side, the distance from the carriage to the anchor point is 45 inches when pushing the router to one side (with the carriage in the middle of the gantry), but only 15 inches when pulling the router the other direction. And guess what? There is far more flex when pushing than pulling, further verifying the diagnosis. The problem basically seems not to exist in the better direction. Still not convinced? Then why did the router vendor double the belt width on later models? I looked into an upgrade but is a major undertaking, not any easier than going to a ball screw, and no cheaper.

[cyclestart]

Kind of off topic, thinking about belt tension leads to wondering how much radial force a stepper shaft can reasonably handle before an outrigger bearing is necessary. Maybe a zenbot has a support bearing. I'm not familar with those machines. I've always avoided making motor belts piano string tight.

[Jandyman]

That's a good point that I thought about briefly and then forgot about. One thing I do know is that at this point my belt is above the recommended tension. That helps a lot, as the modulus of elasticity goes up with increasing tension. A friend tells me not to worry about this, I'll just see faster belt wear, and belts are cheap. But if I'm headed in the direction of a ball screw upgrade, it will be a moot point.

[ger21]

It won't take long before the shaft snaps off, if you have a decent amount of force on it.