Jeshua Lacock wrote: >On Nov 12, 2009, at 4:55 AM, John Thornton wrote: > > >>>If I am using a servo with 512 PPR, is a step 1/512 of a revolution? >>> >>> >>If your servo drives take step and direction signals then they are >>just like steppers to the >>software. If it takes 512 steps to go one revolution and it takes >>two revolutions to move one >>inch then your scale is 1024 = the number of steps to move one inch >>(or mm). >> >> > >Thanks for confirming that John. > > > >>>I >>>guess I do not know how step/microsteps apply to servos... >>> >>> >>It would not apply to servos only to drives that take step and >>direction signals. Steps and >>micro steps and gear ratios and pulleys all add up to one thing "how >>many pulses does it >>take to go one unit". >> >>If you have servo drives that take PWM or 0-10v signals and feed >>back position info to EMC >>via an encoder then that is a whole different beast. >> >> > >Got it. I thought I read somewhere that servos are treated like a step >motor from EMC. > > No, that's only when you use step/dir servo drives, such as the Geckos. It is better to treat servos like servos, by using a drive that takes an analog command as input, and makes the motor go a certain speed (velocity mode) or apply a certain torque (torque mode) based on that input.
>My encoders send the signal back to the Gecko 320's. So I assume that >the Gecko "knows" the position of the servo and EMC instructs it where >to move. > Sort of. The geckos have an internal counter. The counter is changed by the computer (from the step/dir signals) and by the encoder feedback signals. The electronics in the gecko will basically apply power to the motor to make the counter stay at zero, so when the PC issues a step (changing the counter to 1, for example), the motor will be "pushed" until the encoder feedback makes the counter zero again. > This is just speculation on my part, so I could be wrong. > > Yep, you are, sort of :) >>>Is it possible with EMC to command the servo to move 1 revolution >>>(or >>>a specific number of steps) instead of specifying a distance? >>> >>> >>If you set your scale correctly I assume you could... dunno why you >>would want to >> >> >I was thinking it would allow me to set the scale. I could tell it to >do 512 steps and measure, or 25,600 steps and measure the distance the >belt moves.. > > Set the scale to something (like 1000), then tell EMC2 to move 1 inch. Measure the actual travel, and then re-set the scale to 1000*(expected travel)/(actual travel). Make sure the number looks sane - you should be able to verify it if you know the encoder resolution, gear ratio, and screw pitch. Use a longer test travel distance once you're close - the difference between 1000 count encoders and 1024 count encoders, or the difference between 5mm and 0.2 inches per screw turn can be pretty subtle. - Steve ------------------------------------------------------------------------------ Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day trial. Simplify your report design, integration and deployment - and focus on what you do best, core application coding. Discover what's new with Crystal Reports now. http://p.sf.net/sfu/bobj-july _______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
