Machsupport Forum
Mach Discussion => General Mach Discussion => Topic started by: woffler on January 19, 2010, 10:54:15 PM
-
I have been thinking of going to servos and i am a little confused or just stupid one of the two. Anyway here goes as i understand it mach is not a closed loop system am i correct ?second if this is so then what would be the advantage to servos over steppers is there a way to update position for accuracy or not ?
I know you would gain speed in rapids that is fine but what i am really looking for is finer accuracy in repeatable positioning .I have ball screw and know backlash that i can measure with a indicator anyway could some one please kind of explain this to me it would be really appreciated .
Thank You Woffler
-
I have been thinking of going to servos and i am a little confused or just stupid one of the two. Anyway here goes as i understand it mach is not a closed loop system am i correct ?second if this is so then what would be the advantage to servos over steppers is there a way to update position for accuracy or not ?
I know you would gain speed in rapids that is fine but what i am really looking for is finer accuracy in repeatable positioning .I have ball screw and know backlash that i can measure with a indicator anyway could some one please kind of explain this to me it would be really appreciated .
Thank You Woffler
Servos will give you no better accuracy than a properly designed stepper drive. In both cases, when the motors are properly sized, and operated within their limits, positioning will be accurate 100% of the time. Steppers are inaccurate only when under-sized, under-driven, or operated outside their capabilities.
Regards,
Ray L.
-
I think i may have a case of under driving it maybe, i have kept them low in order to not loose steps .i have 960 oz rs stepper motors with v203 geckos for drivers the problem is when cutting in profiling i get the the cutter moving into the part when it plunges on each pass leaving a line on the part that is unacceptable down the side of the part.The rest of the part will be fine .Can not figure this one out?
-
clear yes, a servomotor can have more precision to a motor pap, if this has a good control and several encoder grooves.
-
I think i may have a case of under driving it maybe, i have kept them low in order to not loose steps .i have 960 oz rs stepper motors with v203 geckos for drivers the problem is when cutting in profiling i get the the cutter moving into the part when it plunges on each pass leaving a line on the part that is unacceptable down the side of the part.The rest of the part will be fine .Can not figure this one out?
this is not due to your motors but just a fact (anyway we have the same "problem")
solving this would need a lead-in/lead-out
-
Your problem sounds like you have play somewhere, either in your axis or spindle bearings or could even be due to flex in the tool itself. Lead in will probably help as spunk has mentioned.
As to which is more accurate, that all depends, the servos/drives normally used on hobby type machines have low count encoders and can probably not be tuned so accurately and may actually be less accurate than steppers but better servos/drives with high count encoders can be much more accurate than a stepper. High count encoders will however require that you use electronic gearing in the servo drive when using the parallel port which may not give you any gains, the best option is to use an external motion controller such as the smoothstepper, dspmc, kflop, galil etc , you can then take full advantage of servos with high encoder counts.
Hood
-
i think even with very little play or flex you'll get a mark where the tool enters...
if you'd stop the tool on the contour with the spindle still activated you'll also get a mark
just a matter of the mill "rubbing" against the surface on that one spot for several depths thus leaving a mark
correct?
-
I am really looking for is finer accuracy in repeatable positioning
This will can be had via mechanical gearing. ie resolution ( 1/steps per unit ) provided the rest of the system
is up to par with it. Even if all was perfect, like spunk said, you still will get a mark because of the spindle/ tool runout.
The end of a mill will shear the metal differently than the side of the mill. Heck, you have a little from the spindle +the collet+at the end of the tool= a run out far greater than the positioning accuracy. Add to that just how tight the spindle is in the head , etc. You can experiment by comparing a single plunge cut to say multiple finishing passes ( where the finish pass is just about cutting air).
It's a matter of degree and don't think you will ever make that mark go away completely.
Put an indicator on the end of say a ground stock in place of a cutter. Just push lightly with your finger ( about 5- 10 lbs of force ) then try again pushing a little harder and note the difference in indicator readings. That gives you a flavor of how the cutting force will affect the surface finish somewhat.
RICH
-
I almost always get steps on multi pass profile cuts and my mill is an older very well built mill with good quality ball screws and huge over the top steppers.
In fact when cutting profiles out of flat plate sometimes the steps go down both sides of the cutter, on the part and the stock.
About the only way to get rid of these is to have the first cuts slightly bigger than the final finish cut.
Steve
-
It's good practice to have a finish pass after roughing anyhow...Better yet, two. Rough to .040" over then two finish passes - one .010 over then finish to size.
But then I'm not working to a clock either....
-
By far the best way to avoid entry and exit marks is to plunge to depth OFF the part, then do an arced or tangential entry to the part for the finish pass. If you're careful about *where* the tool contacts and leaves the part, you can eliminate the mark altogether. Plunging and retracting on the cut line will pretty much always leave a mark. To do otherwise would require a machine with absolutely *perfect* tram, a tool with zero runout, etc. etc. IOW, not the real world....
Regards,
Ray L.
-
Thanks i am working on it as i write i an going to use another program that allows for finish passes and see what we get.
Thanks Woffler