X, Y, Z Axes are "slipping" during cutting

Discussion about the CNC Shark Pro Plus

Moderators: al wolford, sbk, Bob, Kayvon

Post Reply
glassdale
Posts: 6
Joined: Wed Feb 29, 2012 7:51 pm

X, Y, Z Axes are "slipping" during cutting

Post by glassdale »

I have been having problems when cutting circular frames. I cut circles from 6" to 16" inch diameter, each with a 3/4" frame. I do this by cutting successive circles, starting at 6", then in 2" increments (6, 8, 10, 12, 14, 16, and 18" circles). I cut these out of 3/16" (0.188") hardboard (the material used to make pegboards), using a 1/4" diameter end mill bit. I have been doing this for the last 2-3 years with few problems. However, lately I have been running into the problem that the X and Y axes are slipping a little during each cutting, and more rarely, the Z axis also slips.

For instance, when cutting a circle, the coordinates at the 90 degree points should be (using a 1/4" bit):

Original (12 O'clock): X=0, Y=radius of circle-radius of bit (for a 6" circle, this is X=0, Y=2.875)
90 degrees clockwise (3 O'clock): X=radius of circle-radius of bit, Y=0 (for a 6" circle, this is X=2.875, Y=0)
180 degrees (6 O'clock): X=0, Y=-radius of circle+radius of bit (for a 6" circle, this is X=0, Y=-2.875)
270 degrees (9 O'clock): X=-radius of circle+radius of bit, Y=0 (for a 6" circle, this is X=-2.875, Y=0).

For each subsequent circle, the radius at each point will increase by 1". When I start cutting, the router now doesn't move to the correct coordinates, but will be offset in the Y direction by some amount, with the X coordinate still at 0. During the cutting of the circles, however, both X and Y will slip so that the above coordinates (as shown on the controller program screen) will not be achieved. If the slippage is small during the cutting of each frame, I will still get an acceptable frame, but now the slippages tend to be too large. I may get a few good frames from each set, but not all, as I should be. Also, the problem is getting worse lately and I'm wasting a lot of stock.

The Z axis doesn't slip like the X and Y axes do, but frequently it will initially go to a wrong value. It will then be offset by that amount for the entire run and not slip further. For instance I typically cut at depths of increments of 0.060", making either 4 passes to achieve 0.240" depth or 5 passes for 0.300" depth. Lately, it will go to say 0.040", and then the increments will be 0.040, 0.100, 0.160 and 0.220 depths instead of 0.060, 0.120, 0.180 and 0.240 depths. Again, if the amounts are small, it's no big deal, but sometimes the bit goes to very large depths (up to 0.800") on the initial cut, which has to be stopped before going very far.

In the past, cycling the controller box's power and shutting down/restarting the control program would cure the problems, but that doesn't always work now. I started doing those things and shutting down/restarting the computer, but again that is no longer reliable in fixing the problem. Finally, I started taking the controller box apart and taking out re-inserting the micro card, but today that didn't even work. I don't know what's going on and really need some help.

The room I work in is very warm this time of year (about 90 degrees), but I have worked in the same room in past summers without the problems. Plus I have the controller box on a USB-powered fan that's made for laptops, which should help some if temperature is the problem.

Is it possible that the controller box itself is going bad, or maybe the chip is failing? I can't think of anything else that could cause the problems I'm having. I don't understand how the controller can know that it's not hitting the correct coordinates and not do anything about it. During the run, the controller screen shows that the coordinates aren't being achieved, and when the run is done, the coordinates will be offset by the amount that was hit during the cutting of the final (18") circle (see attached file).

I'm attaching the .crv and .tap files for reference. Does anyone have any thoughts?
Attachments
Screen Shot - Circles.jpg
6-16 Inch Circles.tap
(7.51 KiB) Downloaded 429 times
6-16 Inch Frames.crv
(849.5 KiB) Downloaded 452 times

joewist
Posts: 5
Joined: Sun Jun 28, 2015 2:48 pm

Re: X, Y, Z Axes are "slipping" during cutting

Post by joewist »

I hade a similar problem on my Z axis my machine was a demo model. I took apart the lead screw and found that material from the Bakelite hade built-up on the lead screw and lead nut. cleaned and all is well.
Hope this works for you. cheap fix.

joewist

rungemach
Posts: 460
Joined: Mon Aug 02, 2010 8:24 am
Location: Sarasota, Florida

Re: X, Y, Z Axes are "slipping" during cutting

Post by rungemach »

Hello Glassdale

There can be several problems here.

It may also help to understand that the controller has no idea where the router actually is. There is no "feedback" from the machine to the controller like a servo system may have. The controller just sits there spitting out a stream pulses to the motors and expects that everything downstream will be fine, and the motors always turn as they are told to. If something goes wrong, like binding, or hitting an obstacle, or just trying to go too fast for conditions, the motor will not turn (skips a step) but the controller will not realize it and just keep on going as if nothing happened.

IMHO, The design of the control box does not allow for a lot of heat removal from the heat generating components inside the box. When electronics get too hot, they can become erratic or intermittent. They may be fine when they get back to a safe temperature.

The "usual suspects" for problems like this would be.

Running cut feed speeds too fast for the motors to power through the material. (the vectric defaults are gererally too fast for a shark), so slow down the feed speed and see if the problem persists.

Physical binding of an axis due to bearings or lead screws getting clogged with dust.

Slipping of the couplers that connect the motors to the leads screws. Using cap screws in place of the set screws can help keep the connection tight. Use locktite on the screws when they have been tightened. The motor and lead screw shafts should have flats on them where the screw is supposed to engage the shaft.

Overheating of the controller. There is no forced air circulation in the controller itself. If the control box is getting warm, it needs help.
Having a fan blow air across the controller case can help a little. Personally, I would have a small fan blowing air through the controller box.

Hope this helps you find the root causes of your problem.

Bob

glassdale
Posts: 6
Joined: Wed Feb 29, 2012 7:51 pm

Re: X, Y, Z Axes are "slipping" during cutting

Post by glassdale »

Joe and Bob,

Thank you both for your suggestions. Bob's first point is exactly what makes this so hard to understand. The controller box (and program) don't know where the router really is at any given time, they just assume that the router is where it is supposed to be, based on the signals that have been sent to the machine. So why doesn't the display show the values that it should instead of something else? At each quadrant, the values shown should be exactly as calculated in my original message. It's as if the control program is sending incorrect signals to the box, which then passes them to the machine (if it were the box, then the display would still show the correct values). But why is it doing this?

Also, slippage of the lead screws or any other component, or feeding too quickly for the router to keep up, would not affect the display of the coordinates that the router hits. Given my problems, it's as if the control program does know where the router really is, since I know that the router becomes offset from the correct coordinates, and the program displays these incorrect coordinates, although I'm just assuming that the coordinates displayed are the actual coordinates that the router hits - I have no way to measure how far off the router actually is at any given time.

I could accept that the controller really does know where the router is, and I'm sure many others would be happy to have this capability, but it's not supposed be that way with this machine. I'm beginning to think that temperature plays a role. This is because usually (although not always) when I start cutting, things may not be perfect, but the slippages are smaller and within my tolerances. Over time, and as the box theoretically heats up, the errors seem to grow. But this still doesn't explain the discrepancy between the theoretical coordinates and the actual ones displayed.

Dale

rungemach
Posts: 460
Joined: Mon Aug 02, 2010 8:24 am
Location: Sarasota, Florida

Re: X, Y, Z Axes are "slipping" during cutting

Post by rungemach »

Hello Dale,

I ran your tap file on my Mach3 machine and job ends at 0,0,0.8 as it is supposed to. Your feed rates and cut depth should not be a problem.

Based on your description that mentions the display is not returning to zero, I suspect controller software or electrical component issues (possibly heat related). If you would have said your display returns to zero, but the router is not at zero, I would suspect issues downstream of the controller, like the motor drivers, motors, and mechanical resistance issues, or driving too fast for conditions.

I would suggest that you start with a "cold" machine and controler. Try running the job in air, and see if you can duplicate the incorrect ending position. If it is off and you re-zero and run again, does it consistently produce the same error at ending or different values each time? Repeat the jobs several times to see if heat build up in the controller is affecting things. You may have a weakened component that is beginning to fail or a software issue..

Inside the shark controller box is the computer that does the G-code to step and direction calculations, the motor driver circuits that control the power to the motors, the power supply that provides the dc power for everything, as well as the solid state relay that controls the router on and off. ALL these devices generate heat, and high temperatures are generally the enemy of electronics. The control box has some air vents, but no active air circulation. The box must try and shed the heat by transferring the heat through the metal box to the outside air. As computers and electronics get too hot, they can start behaving erratically while not failing completely. You may have a weak component that behaves erratically above a certain temperature.

Some Sharks have also exhibited the "plunge of death" where the z goes straight down in to the table for no known reason. I am not sure if this has ever been solved definitively. I would regard that behavior as a different but "known" issue.

I hope this helps.

Bob

Post Reply