Problems calibrating a 2m tall Morgan variant

[Daniel Aaron Bislip]

Hi everyone!

A colleague and I are building a 2m tall Morgan variant for the TU Delft faculty of architecture. As of yet everything’s experimental – We’ll make it pretty once the design works.
Here are some pictures of the printer: https://www.dropbox.com/sh/l7ezs9s5qijbmi4/AACqb-M7WdrEVNHNcYb_38w6a?dl=0

We were pretty happy with the hardware, so we got to calibrating yesterday.
At step 8 (the flowchart) after calibrating M360 and sending M360 again the arms do not move to the correct position, they try to move where they can’t go. video here: http://youtu.be/ZaJKwmvlmpc
note things: first, the arms have been combined with the drivewheels to create a more compact X/Y assembly. Secondly, the assembly was made a bit different so that the whole set up would fit (this is my fault): The theta arm is below the psi arm and it’s all just different from the original. I believe homing works as it should, as we took the differences of the arms into account.

Due to the modified design, the arms are unable to home the same way as the original design (elbows bent outwards instead of inwards: http://youtu.be/vSJ0i_VJ_wY). This also limits the bed size and movement in certain areas. Judging from moving the hot end assembly by hand we’d be able to print at least on a plate of 200*150mm, that’s if we can get it working. As of now, we’d be happy if we can just have it working, even with a tiny build plate.

I tried several home positions to prevent the arms from going to places they can’t reach, but this didn’t help.

Is there something we overlooked, or is this design not able to use the firmware as is? We’re considering going back to the original theta/psi assembly with the drivewheels and the arms being separate if this doesn’t work out.

Anyway, we look forward to hearing suggestions and we’d appreciate the help. Thanks!

[Quentin Harley]

Hi there!

You have found the reason why I did not attach the drivewheels to the arms 😉

Let’s see.

The first thing you should check is that the thick arm is Theta, connected to the X driver and X endstop. The thin arm to the right should be Psi, connected to the Y motor and endstop.

What firmware are you using? Does not do the usual Marlin jiggle…

Things you should check:
Home tower offset settings. Your Y setting will be much more than the usual 55 or 65mm

Also enter the home offset to the actual bed surface 0 point. Zero is to the top right if you look at the machine from the front.

Your steps/degrees might also be a bit low. the arms are moving slow during homing.

So many small things that could cause this.

Awaiting your feedback.
Q

[Daniel Aaron Bislip]

Hey Quentin!

Indeed, the old arms just didn’t work, so we decided to use the original design (but modified it so that they would be upside down)

The old design’s arms were placed incorrectly so theta was psi etc. The arms are now coherent with the motors. (though modified to be upside-down – they don’t seem to show any problems)

We may be running an old version of the firmware… We did have to change the pins.h for it to work with the megatronics board, but I did just take it out of an old folder we had. I’ll download the latest version and try again when I have time (most likely tomorrow)

Because of the new arm assembly, the nozzle can move *much* more and has a normal home position.

We are using 16tooth T2.5 pulleys so I used a value of 107.

I’ll keep you posted as the project progresses. In the meantime, here’s a video of the updated design homing:

By the way, will using mechanical end stops be a problem at all?

Regards,

Aaron

[Daniel Aaron Bislip]

Hello again Quentin,

I just downloaded and updated the firmware from here: https://github.com/qharley/Marlin, but the arms still do not jiggle into place. What could be causing this?

I proceeded to calibrate…
Long story short, after many headaches, it turns out I had the arms placed incorrectly. But I’m still having problems: When using M360 it does not go into the right position. After moving Y until it is in it’s right place, I change the value, home, and m360, and it moves perpendicular to the bed instead of parallel. The driveshaft also likes to rotate backwards 1-2 centimeters, breaking past the (theta) end stop while doing so.

What could be causing this? I did proceed to calibrate M361 and M364 despite these problems and they just go in their place perfectly! It seems that M360 just doesn’t want to go in its place after calibrating and using the value I’m guessing is correct?

I look forward to hearing, and thanks for the help!

Aaron

[Quentin Harley]

I am working on Smoothieware for Morgan at the moment, and will try to pass some of the improvements back to the original marlin codecase.

In the mean time, just run thought the prerequisites on the morgan calibration and startup page one by one, making sure you covered all your bases. Axial robots need absolutely no excuse not to work properly… Don’t give it any!

Q

[Daniel Aaron Bislip]

I ran through all the steps again and found out I didn’t set the stepper drivers up correctly. I immediately tried 1/32 steps but that made the motors go crazy. So instead I chose to keep the drivers at 1/16 and they seem to be working fine. I’ve gone through all the other steps of the calibration process thoroughly and I believe I’ve covered them all up to the flowchart.

M360 is finally working, but the value for SCARA Cal – Theta:-191.55 ; M206 X-191.55 I find strange. I think I remember reading in one of your posts that yours was around 5.5?

I’ve spent over 4 hours now trying to figure out why I can’t get m361 to work. After jogging it to the right position, changing M92 (it keeps going to a lower value), running and fixing M360, then running M361 again, it just doesn’t go to theta 90.
So once I’ve followed the procedure again and again until M92 gets a really low value and the arms only move a little and can not be jogged into place, I give up, start over and get the same issues again.

I really can’t figure out what I did/am doing wrong and would appreciate any insight. For now I’m going to have to wait until monday to continue.

Have a nice weekend!

Aaron

[RobertKuhlmann]

I guess not stepping was the fault, but the amperage your drivers give to the steppers (you loose steps if the current is too low or even too high). You’ll find several guides to calibrate the drivers throughout the internet.

[Daniel Aaron Bislip]

Thanks Robert, I’ll check it out on Wednesday!

[Quentin Harley]

Another thing to check is that you have the correct arm orientation. When viewed from the front, the arm to the left is Theta, and to the right Psi. It is possible to assemble the machine with the thick arms to the left, so be sure you know which is which. Also, when homing in Marlin, the Theta arm will make the jiggle first during homing.

Programming your initial home position is also vital to get the M206 values low.

Good luck!
Q

[Daniel Aaron Bislip]

I made sure the voltage was correct on the drivers, but that didn’t solve the problem. After that, things got hectic and I had to put the project on hold for a few weeks.

Make a long story short, we had a few hardware problems (the drivewheel wasn’t completely stuck to the shaft and would slip out a little, belts would untighten too much) which I fixed (screwed the shaft to the wheel, moved the motors slightly, made belt tighteners).

As for the electronics… The latest version of marlin isn’t compatible with the electronics we’re using right now, and I’m no good at programming, so I decided to just buy a RAMPS which should make everything go smoothly from now on. I’m free this Monday afternoon, so that’s when I can continue. I really look forward to seeing its first print – After swapping the electronics, I believe all our bases should be covered.

I’ll keep you posted sometime next week. I really appreciate the support!

[Daniel Aaron Bislip]

The RAMPS has finally been installed! It has the latest version of the firmware and it finally jiggles home. I quickly tried to M360 it but for some reason it wanted to go backwards. I’ll pay more attention and try to fix it tomorrow if I have time.

Aaron

[Quentin Harley]

Any progress?

[Daniel Aaron Bislip]

Not yet, I have (had) very little time to work on it these few days, but I made some time for tomorrow morning.

The arms do jiggle when home, but to get it to home I need to send G28 a few times (it moves a few centimeters then stops) and the arms want to move backwards with M360.
These are the only issues right now, and I’m sure they’re just firmware related.

I’ll add pictures tomorrow to explain the set-up better!

[Quentin Harley]

Check the initial steps per angle as a first step.

[Daniel Aaron Bislip]

I followed this guide step by step: https://www.morgan3dp.com/?page_id=1073 The only exception is the arm/drivewheel configuration, which you can see here: https://www.dropbox.com/s/nlyc4u3mbcz8yqy/2014-10-14%2010.59.04.jpg?dl=0 so X is connected to the rod and y to the tube.

I am using a T2.5 timing belt with a 16 tooth pulley on 1.8 degrees/step, so the value should be 107.

This is what I get with M503:
SENDING:M503
echo:Steps per unit:
echo: M92 X107.00 Y107.00 Z100.00 E450.00
echo:Scaling factors:
echo: M365 X1.00 Y1.00 Z1.00
echo:Maximum feedrates (mm/s):
echo: M203 X300.00 Y300.00 Z300.00 E45.00
echo:Maximum Acceleration (mm/s2):
echo: M201 X400 Y400 Z400 E8000
echo:Acceleration: S=acceleration, T=retract acceleration
echo: M204 S9000.00 T3000.00
echo:Advanced variables: S=Min feedrate (mm/s), T=Min travel feedrate (mm/s), B=minimum segment time (ms), X=maximum xY jerk (mm/s), Z=maximum Z jerk (mm/s)
echo: M205 S0.00 T0.00 B20000 X10.00 Z10.00 E5.00
echo:Home offset (mm):
echo: M206 X0.00 Y0.00 Z0.00
echo:PID settings:
echo: M301 P22.20 I1.08 D114.00

Here’s a video of it homing :
https://www.dropbox.com/s/f9tucgmzkzgpaka/2014-10-14%2009.58.29.mp4?dl=0
And M360:
https://www.dropbox.com/s/mq0yt0pivmms2po/2014-10-14%2009.55.12.mp4?dl=0

I’m going to redownload marlin from here https://github.com/qharley/Marlin and set the correct values in confguration.h and reupload. Will keep you posted!

[Daniel Aaron Bislip]

I just reuploaded the firmware and the same thing happens, the only values I changed are those in the guide. I’m using pronterface to control the printer. I have the firmware in dropbox if you want to take a look https://www.dropbox.com/sh/0e1ta84hpxseedb/AAB26rhpMfhgCiGY1kC_jbz2a?dl=0

I have to go now, I’ll probably have time again on thursday. Looking forward to your suggestions!

Aaron

[Author]

Posts