Simplest CNC router?

[Guy Mathews]

Rob,

I have to agree with your statement about belt drives. My big yellow came with a belt drive on the 4th axis. I was told at the time of purchase that the belt would not break. I insisted that it would. My experience with cutter torque on a 4th axis stepper with both my Shopbot and my 10 Spindle Lyndex were on my side. I was given extra belts to appease my sense of impending doom.

After breaking 2 belts a few months later I changed my belt drive on my 4th axis to gear drive. It is also my understanding that the company is following my design for this on future 4th axis set ups. This was verbal to me on the phone so we will see what happens.

This link will show the change to gear drive. You will have to follow it down until you come to it.

http://www.sawmillcreek.org/showthread.php?t=103562

Belt drive is not the end all in the world of CNC. It has a place. My X and Y are belt drive from the servo to the reduction gear, on my big yellow CAMaster and to date have given me no problems. However, experience has taught me one thing...

There's always room for problems!!!

[Rob Wright]

Guy - I followed that thread on the 4th axis work - neat

You are right - I am not knocking belt drives when it is a short belt from a stepper/servo to a pinion or lead screw. I am just concerned over a few of the designs that have a long belt that is physically attached to the gantry or carriage and in a long loop. I have read a few success stories of this impelmentation on CNCzone and a few more failures with the belt stretch. At JoesCNC, Joe is working on a belt system that is psychically attached to piece of 8020, and "loose"belt is then aligned with the timing groves - there is only a few inches of belt in tension this way and was avoiding the belt stretch/elastic rebound issue. Even with the kevlar belts there is still some found. That drive was found to give 1700+ipm in tests (limited by mach3 kernel speed and 10 microstep gecko's). There is one builder currently implementing this solution for a 4x8 table. It really is an interesting design - but is being held pretty close to the chest by the Joescnc.com forum members.

[Jim Kountz]

Jim - I must respectfully disagree that the Joe's CNC is slow. I run mine at 300ipm rapids, cut 3D at 225 ipm. This is with the "standard" 2 turns per inch lead screws. There is an option now for 1 turn per inch lead screws that will double the speed. Since these systems use steppers that have much better torque at lower speed, these new lead screws work out very well. There is also a few builders on the Joescnc.com forum that have rack and pinion drive that are 4x8 machines that have approached 1200 ipm. Also belt drive is not the end all - there are some issues with belt stretch. I followed the build of automatedwoodworks log on cnczone.com and it is a great machine - but there is a fair amount of welding, and the linear slides that he uses are a couple dollars each. All of this adds up quickly in the world of CNC

Rob, thanks for chiming in, always good to hear from those who have been there and done that. Those are some amazing numbers and kudos to you for achieving that kind of speed. If I may, how did you deal with the whip of the lead screws that seem to plague these systems? I have read many a posts over at cnczone where this was a major problem. And dont get me wrong, Im not saying that belt drive is the end all of CNC, in fact I didnt mean that at all. I was just saying that I have come across more success with them than failures.
Im very into this thread and look forward to hearing some more about your machine and any others. Its very cool stuff!!

[Rob Wright]

Jim,

With the standard 2 turn lead screws at 300+ipm, I have little to no whipping. That is done by adding a second set of flange block bearings to each bearing mount at each end of the gantry( X beam) and two Y rails. This results in a two bearings mounted 5-inches apart on each end of the 5-ft lead screw. This is shown here with a before and after video on a 4x4 hybrid.
http://www.youtube.com/watch?v=dDk88vqzuQY

another member has mounted dual anti-backlash nuts on each side of the 6" wide Y axis carriage

A third member has placed a bronze bushing in place of the second AB nut as outlined above.

Again - the high lead screws (1 or 2 tpi) really eliminates it from the start, that along with spending an hour or so to properly break-in and adjust the alignment goes a long way to solving those problems.

People using 8 or 16 tpi lead screws do have a big problem with whip and backlash on anything above 100ipm like you said - I know I had a solsyva that couldn't do 50ipm without shaking itself into a fault condition.