tag:blogger.com,1999:blog-5001794558945659136.post634449988301605814..comments2024-03-23T06:32:36.547-05:00Comments on Mark Rehorst's Tech Topics: Here's What Happened When I Swapped Servomotors For Steppers In My 3D PrinterUnknownnoreply@blogger.comBlogger6125tag:blogger.com,1999:blog-5001794558945659136.post-47889955808135537382023-10-05T13:17:32.487-05:002023-10-05T13:17:32.487-05:00Yes, servomotors may be usable for 3D printing, bu...Yes, servomotors may be usable for 3D printing, but not these motors unless you're willing to accept reduced print quality compared to steppers. The servomotors were VERY quiet compared to steppers, but if you use high jerk/acceleration there will still be noise when the motion changes direction (with any servomotor). <br /><br />I did try them again with 3:1 pulley reduction and the print quality did improve, but it was still not quite equal to the 400 step/rev motors I originally used in UMMD.<br /><br />These cheap servos work great for sand tables like Arrakis without any tuning of the parameters.Mark Rehorst DMDhttps://www.blogger.com/profile/13616282129430791353noreply@blogger.comtag:blogger.com,1999:blog-5001794558945659136.post-19223750175651634912023-09-30T05:54:41.431-05:002023-09-30T05:54:41.431-05:00Old post but I came here just searching for experi...Old post but I came here just searching for experiences with servos on core x y printers, so I'll leave a comment anyway...<br />- "Real time automatic adjustment" from JMC is *highly* dubious, pretty sure it just sets a bunch of parameters based on p01-03 stiffness - which you still need to set to the maximum possible rigidity without introducing issues. I tried a JMC servo with "auto tuning" enabled, driving a small belt on an A axis, and was not able to get it to be anything other than incredibly floppy using this mode - belt resonance quickly became a problem (and I couldnt get anti-resonance settings to work). Which is not too surprising, Yaskawa Sigma 7 servos also have a "tuning-less function" which is supposed to do the same thing, only it doesnt completely suck like JMC, but it's still not able to get the same performance as if you properly tune it. JMC dont have a real auto tuning function, which seeks out the correct parameters before you actually use the motor, the naming of this setting is misleading - you really need to manually tune it.<br />- Unfortunately JMC manuals are trash, like most cheap Chinese servo manuals, I found contradictory parameter descriptions. So, yeah, that's fun.<br />- Buy second hand high quality servos over JMC etc. Yaskawa S5 100W servos and packs for about $200 + cables. Totally different experience, but takes time to figure out all the part numbers (e.g. can take stp/dir, no brake on the servo, ideally a low inertia model, etc etc). Would love to make a core X Y based on Yaskawa s7 servos, too many other projects on at the moment though!!Robhttps://www.blanch.orgnoreply@blogger.comtag:blogger.com,1999:blog-5001794558945659136.post-86474548172975919702021-02-23T10:49:05.597-06:002021-02-23T10:49:05.597-06:00I don't know if it's the encoder or the st...I don't know if it's the encoder or the stepping control, but someone on the Duet forums used the same motors with a 3:1 reduction (via belts and pulleys) in a printer and the prints it produces look perfect. See:<br />https://forum.duet3d.com/topic/21503/nema-17-intergrated-servo-motor-possible/8Mark Rehorst DMDhttps://www.blogger.com/profile/13616282129430791353noreply@blogger.comtag:blogger.com,1999:blog-5001794558945659136.post-349591712500602822021-02-23T09:04:26.312-06:002021-02-23T09:04:26.312-06:00If I get right you right you think that the build ...If I get right you right you think that the build in encoder of the motor isn't accurate enougth. Do you think switching to the odrive controller in combination with a high resolution incremental encoder could bring better results?Alexandernoreply@blogger.comtag:blogger.com,1999:blog-5001794558945659136.post-26007944564533285812020-12-01T10:17:31.254-06:002020-12-01T10:17:31.254-06:00The RS-232 connection is only for adjusting the dr...The RS-232 connection is only for adjusting the driver parameters, not for actually driving the motors. The 9600 rate was default for the "modbus", whatever that is, but the normal comms were done at 57,600. In operation, the motors are driven by step/direction/enable signals that come from the Duet expansion board, so as far as the Duet is concerned, they are stepper motors.Mark Rehorst DMDhttps://www.blogger.com/profile/13616282129430791353noreply@blogger.comtag:blogger.com,1999:blog-5001794558945659136.post-86290384245150784572020-11-30T18:37:29.343-06:002020-11-30T18:37:29.343-06:00Sorry, not quite following. What RS-232 baud rate ...Sorry, not quite following. What RS-232 baud rate did you end up with to the servomotors? Strongly suspect 9600 is far too low, and would yield the results you saw. Are the servomotors doing point-to-point movement, or do they accept arcs?<br /><br />As 9600 baud ==> ~1000 characters/second. Guessing if average command is 20 characters, then ~~50 updates/second. Would explain the results you saw.<br /><br />(Also, not at all expert in this domain.)Preston L. Bannisterhttps://www.blogger.com/profile/03201415630192091526noreply@blogger.com