Something different to test out

Very smooth - nice work!

Linuxcnc must have some limits when it comes to planning moves? that part is still pretty new to me. it capped the velocity at 860 IMP. I’m not sure if it’s in my settings or if it just maxed out in the space of the 10" diameter of the cut and that was all it could accelerate up to? i have my soft limits set at 14" on X and 18" on Y and the most i could get was around 1500 IPM on X and 1700 IPM on Y so not matter how fast it works out to be you are limited by axis length and acceleration.

all in all i am pretty happy with the rates and moveomet quality :slight_smile:

Could you post your .ini file?

i don’t have internet on that computer but since you brought it up i know exactly what file you are talking about. Is there a value that i can edit besides my acceleration? i think i seen some other stuff that might be for the motion planning but did not mess with it since i was not sure.

i think it’s just a room thing in the tool path. i have my MAX_LINEAR_VELOCITY @35 & MAX_ACCELERATION @ 50 on the X & Y axis. the @35 setting came from the MAX_VELOCITY setting for the display settings which is set at 2,100 IMP

yeah it’s the acceleration. i was able to bump that up to 60 and anything above that kicks an alarm. i ran the same program and it now tops out at 942IPM so i was able to gain around 80IPM

1 Like

this machine needs an enclosure period!! :slight_smile: it has no issues at all cutting aluminum getting peppered is an understatement :slight_smile: i had to stop testing it was that good of a mess :slight_smile: you will never see super glue and tape setups with this thing. i made a quick and simple low profile vise just to get started with so i can make better work holding solutions but i need to box this thing in. i had it getting 250 IPM while cutting and was nowhere near any limits. i have to dial in accuracy yet but got a head of myself with a little bit of cutting tests.

since i have linear scales around and a spare DRO i wanted to try this opposed to a dial indicator reading over a small distance. it worked really well. i was really curious about how accurate a belt is over the rolled ball screws everyone uses. over a 10" distance i was able to dial things in really close and test to see what is going on in the middle as well. i think if memory serves me well a rolled ball screw are only accurate to .003 to ,007 per foot as far as lead accuracy goes. these belts float around between 0.000 to .002 over 10" and i don’t need any backlash comp either. the glass scales worked well for all this.




Really cool machine you got going there!

some test cutting.https://youtu.be/nixNujgF4pM?si=471w8nH7I-MJYK8e

That is awesome… Nice work. What rpm were you running?

24,000 RPM wide open :slight_smile:

I must say this made me chuckle! Very overkill, but love the engineering.

the scales were to just calibrate the step scale in Linuxcnc, and over a large distance they give you a better idea how accurate you can get things. i wanted to see how good backlash was and see how the lead accuracy was with belts. rolled ball screws only have a lead accuracy of say .003 to .007 per foot and i wanted to compare the two so i don’t think it is overkill. the belts floated around between .000 to.002 over 10 inches. 1 other thing they let you see is how acceleration may cause the machine to over shoot the target distance. So the fine details all add up for a better and more accurate machine in the end that produces consistant results :slight_smile:

here is something i think is pretty cool and a good test for this machine, i have never used this small of a ball nose mill before but have been pushing it to see where the limits are. i am surprised i have not boke one yet honestly :slight_smile:

CNC labor of love (youtube.com)

Very cool and amazing detail!

with a tool that small you don’t have much of a radius so finish is not as good as you can get with tool with a bigger radius, but you need the small diameter to get details like that. i started with a .001 stepover going in one direction but going both ways did just a well with the finish. using a .002 for a stepover going both ways cut a lot of cycle time out of the program when i got to .003 for a step over the finish started to fade a little but still not to bad. when i got to .004 it really dropped off in quality. maybe feeding a little bit slower might help but i did not test that far.