TLDR: Z Axis is drifting and Makermade is being unresponsive
Hello, I have a Makermade Original Maslow kit that I built in October and are still having issues with. Makermade has been unresponsive over the last two weeks, so I figured I would try here. My specific issue is that the Z Axis seems to be unpredictable. If I move it 5mm, it will end up somewhere in the range of actually going 4.5-5.5mm. Over time, this causes either parts to not cut out, or get cut too deep. I am using their upgraded Z-Axis kit, but I can find no documentation on what all of the settings should be for this. Is there anyone with a Makermade original Maslow, z-axis upgrade, and using either ground or web control that can help me out? I found that Makerverse does not play nice with the original MEGA board and firmware, otherwise, I would be using that. At this point, I suspect hardware issues, either with the controller shield, Z-axis encoder, or the motor itself, but I am not spending any more money on this project until I can get some usable results. TIA
Update: Makermade finally go in contact with me today, over 2 weeks from emailling them originally. They are going to replace my z motor in case something is wrong with it.
My z axis tolerance is off by .05 to .12 mm everytime it moves according to report sled position and gcode using webcontrol. Nothing as high as what you are experiencing. I’m interested in if the motor solves your issue.
If your issue is that the z axis is off it may be that you z pitch settings need to be fine tuned. Here is a link to a previous post with info taken from the meticulous z axis instructions.
Yeah, the bad thing is that it thinks it is in the correct location so after a few moves, it gets way off. If it knew it was off, i wouldnt have a problem. Then i know it would be some PID settings that need to be adjusted. I did confirm with Makermade that the pitch should be set to 16, which was actually what i figured, but because it would never move to the same spot twice, i would always calculate between 15 and 17. I have a feeling it will either be an encoder or a shield issue.
Its not how the router is mounted. Im using the upgraded z-axis kit, in other words, its the M2 design, with the spindle clamped onto the plate. And the problem is that its moving a different distance every time
That was the last thing that Makermade asked about before replacing the motor/encoder. They are all tight. I will probably have to keep an eye on them over time in case i start having issues later.
I figured you all checked that if they were replacing the motor. Is the motor housing gearbox all tight? I wonder 8f the encoder somehow came loose inside.
I find that the set screws can slip even when tight. I grind a flat onto the shafts going into couplers on every machine that uses them. Never had one slip or work loose again.
Other things to check:
The lead screw is not bent. Roll it on a flat surface. If it wobbles whilst rolling its bent.
The bearing rods are all parallel and movement is completely smooth with no resistance throughout the range when the lead screw is removed.
Belt tension is not too tight. You want it tight enough that there is no play but still easy to rotate.
If you still have problems after you get the new motor you could change the gearing. You will have to re calibrate, will lose speed of movement, but will gain torque and precision in movement.
So I got the new motor in. The drifting is better, but not gone. It was moving ± 0.5mm/move and now it is down to 0.1mm. Sounds like very little, but over many z moves, it adds up. One up then down is 0.2mm off of what WebControl says its at. My question at this point is if that is the best im going to get or should it at least keep track if it overshoots.
I think I have the same problem. For me, the gcode for z doesn’t match the sled position for the z axis reported by webcontrol. Additionally, there was visible evidence when pocketing where the depth was not always the same and caused an uneven pocket, more profound with more z movements.
I did not find a perfect fix. However, I found that slowing down the plunge speed while creating the gcode reduced the error.
We may have the same issue, but my problem is more that Web Control tells the arduino to move 10mm. Webcontrol says that it moved 10mm. The z axis only moved 9.9mm. Move it back 10mm, web controls says it moved back to the original spot, but the motor moved it 10.1mm this time, so its actually 0.2mm away from the spot. Do this 10 more times and now we arent even cutting the part. Ive already confirmed with Malermade what the pitch is for my machine, so I know its not that.
if it’s drifting (moving different amounts at different times), then the problem
is in either the encoders, or the wires.
Try not to run the wires to the motor directly adjacent to the power wire for
the router (just in case that’s causing interference that is being detected as
encoder pulses)
to test this, write some gcode that just does Z axis movements back and forth
and see what happens.
It is possible for there to be some slop, which combines with a tight router fit
to make movement in the two directions different, but that should be a
consistant error, not one that grows over time (to handle this, add a light
bungee across the top of the router to put tension towards the workpiece. if you
put too much it wears out the Z axis mechanism and the router can slip, digging
in to the workpiece and breaking the bit)
I am measuring the movement of the plate that moves up and down on the Z-screw. My setup is a original Maslow from Makermade, but with their upgrade kit for the Z axis, so it’s constructed like an M2. I will post some photos later today with the measurements off my digital calipers.
delta (off by .06 and then off by .25, then off by .37)
Does groundcontrol register the same numbers and do the same thing?
I’m betting it will do the same thing and the issue is in the controller firmware possibly PID settings and not webcontrol or groundcontrol. What version of firmware are you running (51.28 that makerverse requires?)?
Ground control doesnt make a difference. Makerverse doesnt run right on my maslow and makermade said that sometimes happens. They recommended just using webcontrol. I think it is most likely a PID issue, unless the controller isnt getting good encoder pulses back. The problem I see is that web control reports no error, yet there is error that should be registered.
