On Fri, Feb 15, 2013, at 06:41 AM, Erik Christiansen wrote: > On 15.02.13 10:40, andy pugh wrote:
> > > I have a number of axes without even any tensioner on the belts (just > > holes bored at the exact centres) and they seem to work OK. Even the > > ones with tensioners are not "twang" tight. > > I might give it a go then. > > > There are belt designs with reduced backlash, and even zero backlash. > > http://www.contitech.de/pages/produkte/antriebsriemen/antrieb-industrie/download/td_synchroflex_en.pdf > > Page 23. > > Suggested pre-tension is on page 26, and appears to be approx half of > > the drive-tension. > > Thanks. That looks interesting. If I could go straight off the stepper > shaft, it'd save a bit of work, and the couplings. But the tiny radial > thrust rating on the stepper seems to kill that. > > Erik I have to agree with Andy, simply based on industry practice. You pretty much never see "motor coupled to shaft carrying timing belt pulley". You always see "timing belt pulley directly attached to motor shaft". The only time flexible couplings show up is when the motor is in line with the ballscrew, and no belt or gear is used. You do want to keep the pulley as close to the motor bearings as possible. Hanging it way out on the end of the shaft will increase the bearing load. Make the pulleys as big as you have room for - that helps you three ways: 1) the belt tension needed to transmit a given torque goes down 2) there are more teeth engaged between belt and pulley. That both increases the ability of the belt to transfer torque, and it reduces the possibility of slip (which could show up as backlash) 3) if there is any movement (backlash) between belt and pulley, a given amount linear movement results in less angular rotation of the shaft. Regarding backlash, it might help to work an example with numbers: Suppose you have a 5 thread per inch ballscrew, and are using a 3" diameter pulley to drive it. The circumference of the pulley is 3 x pi = 9.42 inches. It takes five revolutions of the pulley for the axis to move one inch. That means 5 x 9.42 = 47.12 inches of belt movement to cause one inch of axis movement. So any backlash between the pulley and the belt is reduced by a factor of 47 before it gets to the axis. If a belt tooth is 0.005" smaller than the groove in the pulley, then in theory the belt could slip by 0,005". Divide that by 47 and you get 0.000106" at the ballscrew. You mentioned "the tiny radial thrust rating". Usually, "thrust" means axial thrust. I haven't seen the stepper data sheet, are you sure you are looking at the right rating? Belts produce only radial loading, no axial thrust (assuming they are aligned properly). One more thing - keep in mind that the rated torque of a stepper is at zero speed, and you can't load it to even half that level and expect to get much speed out of it. If you are calculating your belt tension and bearing loads based on stepper rated torque you are being very conservative. Most stepper machines run at one-fourth of the rated torque, or even less. -- John Kasunich [email protected] ------------------------------------------------------------------------------ Free Next-Gen Firewall Hardware Offer Buy your Sophos next-gen firewall before the end March 2013 and get the hardware for free! Learn more. http://p.sf.net/sfu/sophos-d2d-feb _______________________________________________ Emc-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-users
