[Jakeller449]

I just finished up my extruded aluminum style CNC router. A little background on this machine that I designed:

V bearing guides with Fixed and Eccentric bushings.
1/2"-10 Acme rods (48" Y, 36" X, 12" Z).
Anti-backlash delrin ACME nut purchased from McMaster.
Machined steel plates for motor mounts, ACME nut mount, gantry and bearing mounts
10 Series 80/20 Extrusion. 2"x4" Gantry beam and bearing rails with 2"x1" table supports.
Gecko G540 w/ 4 NEMA 23 381oz/in motors from Automation Technologies
UC100 Motion Controller
UCCNC Software
No homing or limit switches installed
Shielded wire (I believe they are grounded to the plug end at the G540, I will be checking tonight. I do not believe I have the motor case grounded)


If you want more details on my machine I would be happy to provide it but my design is based on the thousands of other extruded aluminum machines out there. I have my A-Axis slaved to my X-Axis to move the Gantry.

Setting up the machine, I have squared the gantry by measuring the extended length of each Gantry carriage and moving each axis individually until they are equal. This method seems to work fairly well for squaring the Gantry. However, I have noticed that if I jog the machine at a higher feedrate the X-Axis seems to stick and sometimes even cause the motor to stall. I have slowed down the feed rate and that has improved this problem but it has not completely resolved the issue. After running a program, I find that the gantry can be out of square by 1/2"-3/4". The slower the feed rate, the smaller this gap tends to be. I have slowed the feedrate down to a point now where I am starting to burn my wood with brand new bits and I still have this square issue.

This is causing major problems in trying to achieve accurate results and I am no where close to repeatable. I am trying to figure out what all can be contributing to this step loss and out of square condition that results. I am still learning so there are probably a few basic steps I am missing. I have tried using the Trim function on the G540 but I dont really understand what everyone on the forums and youtube are saying about where this should be set. I have also checked my UCCNC axis settings to make sure my X and A have the same configurations. My guess is the issue is either in the mechanical side or on the wiring/motor side but I still have no idea where to begin to troubleshoot those issues. Any advice or direction would be much appreciated.

Thanks.

[ger21]

Are the screws really 1" diameter?

What are your velocity and acceleration settings?

What voltage are you running the motors at?

If you're using 1" 10 tpi screws, the machine is not going to be able to reliably run very fast.

[justicedavid65]

I am having the same issues with my build, so I'm interested in what answers you get. I can jog around and get no lost steps, but when I run g code, my gantry will be out of square when it's finished running. I can't actually see the lost steps (I have tried), so I don't know which side is lagging?

[peteeng]

Hi Jak - You have a similar system to me but I use belts not screws. As you have the A-Y axis slaved I see no reason why this could be a software or firmware issue. The UC100,UCCNC and Gecko put out the correct steps. So I suspect its mechanical. You need to have the gantry very well connected to the guides and the screw working perfectly. So check how this mechanical system is working. If one side is sticky vs the other then this is the answer. So check the leadscrew are smooth and that the gantry is stiff transversely so it can't walk. Peter

[Jakeller449]

Sorry, they are 1/2" - 10 screws. I believe velocity has been set down to around 30units/min with acceleration around 10units/min. This is for X and A and it has given be the best results. Y axis I can run to around 60units/min with acceleration at 15units/min with pretty good results.

My antibacklash nuts were super tight when I got them. Running them in and out with a drill a bunch of times didnt make the fit any looser so I did what another forum recommended and cut a slot in the end of the screw to act as a tap. This worked well and I was able to get the screws to run fairly free from that point forward. If one screw is tighter than the other, but still able to move freely with just two fingers, would that be enough resistance to affect freemovement?

I am running 48V 7.3A. Motors are NEMA23 381oz/in @ 3.5A each

[ger21]

What "units" are you using? I'm guessing inches?

would that be enough resistance to affect freemovement?

Yes, it takes very little friction to cause stalling or lost steps.

Lubricate the screws/nuts with a lightweight oil. It will help a lot.

With 1/2-10 single start screws, 75-100ipm is probably about the max you can expect to get reliably.

[Jakeller449]

Yes I am running it in inchs. I have tried to run this machine faster but once I go past 70-80ipm then the motors will stall and I will lose steps. I had that issue with my Z-axis when it would try to rapid out so I slowed the Z-axis way down (like 30ipm) to make sure I have enough torque to raise the head. I am guessing the slide on my z-axis is a little too stiff which is why I am having the problem there but I havent noticed any major issues with the slides on X and A. But it sounds like that is where I need to take a good hard look.