We use cookies to personalize content, interact with our analytics companies, advertising networks and cooperatives, and demographic companies, provide social media features, and to analyze our traffic. Our social media, advertising and analytics partners may combine it with other information that you’ve provided to them or that they’ve collected from your use of their services. Learn more.
1 - It’s possible, the X axis was somewhere inbetween the tension of the Y axis before & after I tightened it up today. It did have some snap to it though.
2 - Checked this, none spin by hand
3 - Is there an established way of checking this? The potentiometer is changing the amount of available current to the stepper motor? So this should be measurable during operation with a multimeter…
4 - you are referring to the grub screws which secure the ribbed wheel - or something else? I just rechecked all of the grub screws holding the short end; they’re torqued so that the allen wrench just begins to deflect (with the long end inserted into the grub screw).
My depth of cut is set to 0.5mm, I’ll try a DOC of 0.2mm after I re-calibrate the steps/mm. Hadn’t considered the depth of cut since wood is relatively soft, but this is a really valid point.
Having the V-wheels too tight can cause problems too. Right now I can move the V-wheels by hand if I really try hard, and it made everything a bit better. If it’s too tight, then it may be putting extra torque on the stepper motor since the bearings are far from frictionless.
Check out this forum post for some good pot information: New drifting
As for the loose pulleys, it’s very possible for the little screws to come loose (or even come out). If that happens, it can cause the shaft to spin in the pulley without actually turning the pulley. Double check that they are tight, and throw some loc-tite in them as well if you notice they start to loosen up. Mine are still solid, but the first time one loosens up they’re all getting locked.
You’ll likely have trouble measuring the current accurately with a multimeter. Using a single mulitmeter while the motor is stationary, you can only read one of the two coils, and don’t necessarily know if you have all of the current going through one winding or not.
When microstepping is enabled, the total current supplied to the motor is split up depending on the step differently between the 2 different sets of coils in the motor. Temporarily turning off microstepping can make the measurement easier for testing purposes.
Another thing to consider when tightening the belts (Y axis) is the length of each belt after you have it tight.
For example, if you tighten both to the exact “tone”, as a lot of people do, the left belt might be two or three teeth shorter than the right belt. (or more)
What I did, was get one belt roughly where I needed it to be, mark a start tooth and end tooth with a short copper wire (one strand out of the stepper wire). I would then match the other belt, tooth by tooth, and mark the exact same teeth on that belt.
Once I installed the belts and got the right tension, I measured the distance between the two marks. By ensuring these were the identical distances, even though one belt might be slightly tighter/looser than the other, it ensures that both the left and right side move the exact same distance. As long as you get it tight enough not to slip or cause backlash, this would make your X-Carve move in perfect sync along the Y axis, even if one is not as tight as the other.
Before doing this, my left belt was almost 1/8 - 3/16th shorter, so over the distance of the cut, the left side had moved 1/8 - 3/16 further than the right.
To me, this sounds like either the acceleration or max rate might be set to high, or voltage to low. With short movements, it does not get a chance to accelerate, however with long movements it tries to accelerate to its top speed.
Check your max speed and accel settings. ($120-$122 and $110-$112)
I believe I’m getting closer to a real circle now.
The second cut from the right is the result of last evenings work with re tensioning the X axis belt, eccentric nuts and ziptying all belts at the clip,
The first cut from the right is the result of modifying the steps per mm as measured over a 500mm distance. Both $100 & $101 were changed.
As you can see I’ve added a square (10mm X 10mm), and changed the depth of cut to 0.2mm from the default 0.7mm.
The square is cutting at a width of 9.40 to 9.45mm and a height of 9.87 to 9.90mm. It’s difficult to measure in hundredths of a millimeter with soft woods…
The circle still has an oblong/egg appearance, possibly it would be helpful to record a short video of the X-Carve cutting one of the circles. It’s visible when the bit is cutting from the center of the circle outward spiraling circles, especially at the bottom right quadrant that the bit is not traveling as far in the X direction as it is in the Y direction.
I am going to investigate microstepping / potentiometer adjustment as I’ve noticed a movement such as Y500, will result in a machine location of something like 500.015 instead of 500.000. The same happens when trying to bring it back to the home position. X/Y are within +/- .020 of 0.000.
Tomorrow I will have the 600mm ruler in 1/2mm increments that @RobertA_Rieke recommended earlier in this thread, which should further increase the steps per mm accuracy.
I tried increasing the X potentiometer in two increments of clockwise 1/8th turns. Didn’t seem to notice much of a change in roundness of the circle.
I did notice during the cutting of the circle that there is a hesitation when the bit is going to the bottom right quadrant, the bit travels downward more than it travels to the right. Possibly it’s a missed x axis step? I’ve included a short video of the circle being cut.
Currently the 10mm diameter circle is being cut oblong so that:
Try loosening up the X- and Y-axes belts a little bit. If the belt is too tight, it could be requiring the stepper motors to put more torque in to move and therefore losing the microsteps.
I had too much left over belt doubling back after the clip.
I was watching the X-Carve cut from the side and noticed the leading static wheel was rolling on and off of the excess belt. I really do feel a bit stupid for not catching this until now. In any event I pushed the excess belt down into the extruded aluminum and the 10mm circles & squares now look like this:
Both the squares & circles have almost no variation between cuts. The squares are 9.45mm wide (X), and 9.60mm tall (Y). The circles are still subtly oblong, ranging from 9.4mm to 9.7mm in diameter depending on the rotation of the caliper.
I am hopeful that upon re-calibrating the number of steps per mm in both directions that the squares will be closer to their prescribed 10mm size; and the circles will then lose that minor variation in their diameter.
Also, for anyone who wants to investigate tuning their potentiometers, the helpful staff @ Inventables forwarded me the GrblShield’s readme on setting motor current.
I am kicking myself for not suggesting you look there. When I did mine and heatshrunk it I cut it down to about an inch overlap (didn’t want too little since the teeth meshing is part of what holds it in place) and made sure to set my limit endstop to omit the overlapped / shrunk area too. From your test images I assumed each cut was in a different location on the table…not that you were moving the workpiece.
Sorry for not helping shorten your calibration efforts!
Also FWIW when I built my machine I didn’t see any clear reason why the ‘adjustable’ end of the Y belts needed to be on the back, aside from perhaps more overlap meaning you were sacrificing a little usable area on the front. I put mine in front for more easy access to adjustment. (Adjustable end of X is also to the ‘left’ when facing the machine for me.)
the reason is cuz when the adj for the belt is in the front it can block the idler pulleys from reaching “0,0”. i found out the hard way. Also when its in the back u ca use the front y plates as a hardstop square.
Did you happen to resolve your problem with circles? I have followed all the tuning , tightening and adjustments listed on the forum, and yet I still get circles that are slightly out of round. Your description of the oblong circle pattern is spot on with what I’m dealing with.
