Incorrect Home Position Y+392.77 MK3.2

[Anonymous]

Dan, when I initially power on the MK3.2 with the pen positioned on the calibration.pdf origin the coordinates read X0 Y0 Z90 F100%. Once I use the controller to set “This is home”, the coordinates now read X0 Y392.77 Z10 F100%. This it does whether I include or not include the additional step of disabling the motors then enabling them. I have uploaded the latest firmware 7.15.5 and commented the robot_polargraph.h line to define this is a 3+ rather than 5+. Machine_style is polargraph in the configure.h. On another nonrelated issue, I also do not get a response when I type D2 as you suggested in another post. Why does the MK3.2 believe it is at Y392.77 rather than Y203.2 as it should per the calibration.pdf? Thank you for any help you can provide.
Allen
fastal77

[Dan]

Have you run the robot connected through the software at least once? When the settings change in the software it sends a gcode command that updates the robot’s memory, after which home should be correct. If it is not correct then that’s an issue I will fix.

There is no D2 command. There is a G2 for drawing curves. What thread (link please) are you talking about?

[Anonymous]

Dan, additionally when I use the Manual Driving within the MK3_2 7.15.5 software, I get inconsistent movements of the pen holder. Also, when I click the ‘Set home’ button underneath Manual Driving, the output is G92 X0.0 Y392.5989990234375 D6 X0 Y392.599 with the Arduino display showing X0 Y392.77 Z10 F100%. Within settings the MK3+ is selected. Thank you for your help resolving this.
Allen

[Dan]

Are your belts slipping on the pulleys?

Are the pulleys tight on the motors?

What do you mean when you say “inconsistent”?

That numerical discrepancy can’t be a rounding error. very strange.

Do the pulleys ACTUALLY have 20 teeth? One time we had an order where some units had 18 and were marked as 20, but we got rid of the whole batch a long time ago.

[Anonymous]

The pulleys do have 20 teeth, and the belts are not slipping. ‘Pen up’ moves the pen holder close to the board surface whereas ‘Pen down’ moves the holder away from the surface. Is this correct? When I test run the drawing.svg w/o a pen, the holder goes through what appears to be the correct motions, only shifted to the right on the board approximately 5″ down and 6″ right of the ‘origin’ of the calibration pattern. Also, the pen holder hugs the board surface when not drawing and away from the surface when drawing.

[Anonymous]

After reading the thread “Recommended Makelangelo 5 settings, calibration”, I typed D7 and pasted the result below:
D7
> D8 L0.01 R1313996.75
Mine is a MK3.2 so I don’t expect the same values as a MK5+. However, it is obvious that the D8 belt length returned is erroneous. I measured my belt length(s) and listed them below:
28 inches – pulley to bottle
27.5 inches – anchor bolt to bottle
19.5 inches – pulley to pen holder bolt
75 inches Total = 190.5 cm
I typed: D7 R101 L101 and D8 returned with: D8 L101.00 R101.00
I used the values someone else posted, but how do I determine “how much belt to pull to get the pen holder back to the home position after hitting the limit switches”? Then I can adjust accordingly using the D7 command.
Also, the pen up and pen down servo position is still backwards. When I reversed the red and black wire from the servo, the Arduino wouldn’t power on. Once I returned them to their original position it powers on. Is the black wire the voltage source wire? Thank you Dan for your help resolving this.
Allen

[Dan]

A Makelangelo 3 can’t auto-home (it does not have homing switches south of the motors), so the D7-D9 & D11 commands are meaningless to M3.

Pen up and pen down is completely different from “does not turn on at all”. separate issues. I’m glad you got the machine turning on and (i infer) servo moving.

In Settings > Pen you can change the up and down angles, as well as test them before you save your new settings. I doubt these settings are saved into the robot’s memory, it’s only in the software. it was done this way long ago on the idea that one software could run many robots and the up/down value would never need to be useful stored in the robot’s brain. change my mind?

[Dan]

> only shifted to the right on the board approximately 5″ down and 6″ right of the ‘origin’ of the calibration pattern.

Photo? That sounds very strange.

> Also, the pen holder hugs the board surface when not drawing and away from the surface when drawing.

Yes, that’s backwards.

[Anonymous]

Upon startup of the MK3_2 7.15.5 software, the following is output. In particular, the dimensions at the end grab my attention as erroneous, perhaps since I don’t understand their meaning. I do want to understand what this output means. Also, I am emailing an mp4 video and some photos of what I am working with in a zip file photos_
mk3_2.zip. Hopefully, the wiring images will shed some light on why the pen behaves backwards.
Allen

HELLO WORLD! I AM POLARGRAPH #17998
Firmware v8
== http://www.marginallyclever.com/ ==
M100 – display this message
M101 [Tx.xx] [Bx.xx] [Rx.xx] [Lx.xx]
– display/update board dimensions.
As well as the following G-codes (http://en.wikipedia.org/wiki/G-code):
G00,G01,G02,G03,G04,G28,G90,G91,G92,M18,M114

> D10 V3
M101 A0 T457.2 B-457.2
M101 A1 T609.6 B-609.6
M101 A2 T170 B10
G92 X0.0 Y392.5989990234375
D6 X0 Y392.599
G0 F6500 A2500
M17

> (-457.20,-609.60,10.00) – (457.20,609.59,170.00)

> Saving dimensions.
(-457.20,-609.60,10.00) – (457.20,609.60,170.00)

> (-457.20,-609.60,10.00) – (457.20,609.60,170.00)

> Saving home.

[Dan]

the connection for the right hand motor is backwards. do not unplug it while the robot has power.

[Anonymous]

The color of the wires matches the burnt-in descriptor on the board above.

[Dan]

the firmware was later changed to make it faster and simpler. i can’t edit your laser etched wood.

[Anonymous]

Dan,
I reversed the right motor connection. Since the MK3.2 believes the home position is at Y=392.77, I positioned it on the board as such as you’ll see The manual driving commands seem to respond appropriately. I’m including some photos and video of a dry run of the thankyou.svg file. It appears to be drawing the thank you upside down: the pen is above the calibration pattern while drawing the lower part of the thank you, and below the pattern when drawing the top side.
https://www.dropbox.com/s/t57m3vnrxtz5x3p/PhotosMK3_2_042618.zip?dl=0

HELLO WORLD! I AM POLARGRAPH #17998
Firmware v8
D10
== http://www.marginallyclever.com/ ==
M100 – display this message
M101 [Tx.xx] [Bx.xx] [Rx.xx] [Lx.xx]
– display/update board dimensions.
As well as the following G-codes (http://en.wikipedia.org/wiki/G-code):
G00,G01,G02,G03,G04,G28,G90,G91,G92,M18,M114
HELLO WORLD! I AM POLARGRAPH #17998
Firmware v8
== http://www.marginallyclever.com/ ==
M100 – display this message
M101 [Tx.xx] [Bx.xx] [Rx.xx] [Lx.xx]
– display/update board dimensions.
As well as the following G-codes (http://en.wikipedia.org/wiki/G-code):
G00,G01,G02,G03,G04,G28,G90,G91,G92,M18,M114
> D10 V3
M101 A0 T457.2 B-457.2
M101 A1 T609.6 B-609.6
M101 A2 T170 B10
G92 X0.0 Y392.5989990234375
D6 X0 Y392.599
G0 F6500 A2500
M17
> (-457.20,-609.60,10.00) – (457.20,609.59,170.00)
> Saving dimensions.
(-457.20,-609.60,10.00) – (457.20,609.60,170.00)
> (-457.20,-609.60,10.00) – (457.20,609.60,170.00)
> Saving home.
G92 X0.0 Y392.5989990234375
D6 X0 Y392.599
G92 X0.0 Y392.5989990234375
D6 X0 Y392.599

[Anonymous]

Dan,

I reversed the pen servo connector at EXP3 on the Arduino 2650 Rumba from black, red, white to white, red, black. This was an attempt to correct the pen ‘up and down’ issue that is presently reversed: pen is up when actually it is down, and conversely so. Accordingly, the MK3.2 is unable to draw the thankyou.svg due to the pen ‘up and down’ reversed, even though it does go through the motions. This you can witness in the https://www.dropbox.com/s/t57m3vnrxtz5x3p/PhotosMK3_2_042618.zip?dl=0 which includes MOV_0340.mp4 demonstrating a dry run of the thankyou.svg. Please help me with this issue. Thank you.

Allen Bowden

[Dan]

reversing power through the circuit will probably cook the servo. that’s not a workable solution.

* Connect to the robot
* go to settings > pen
* test up and down angle
* probably swap values
* test until up and down angle are satisfactory
* regenerate thankyou gcode
* run drawing again

[Anonymous]

That worked for the pen ‘up and down’. I also downloaded the latest firmware from https://github.com/MarginallyClever/Makelangelo-firmware/tree/dev. Making progress… The thankyou.svg draws in a compressed format and upside down. Here is a link to the photos I took: https://www.dropbox.com/s/hi6ab8kgn9snr5r/MK3_2_050418.zip?dl=0

[Anonymous]

Dan,
How do I correct the squashed/compressed and upside down drawing w/ my MK3.2?

[Dan]

* home robot
* use manual driving

does Y-100 go down 10cm?
does X+100 go right 10cm?

Did you adjust motor steps in your firmware to 200 (for your 1.8 degree steppers)? Did you adjust for microstepping?

[Anonymous]

Y-100 moves up 5.72 cm, X+100 moves left 9.1 cm
In the Makelangelo-firmware I find “Translate the XYZ through the IK to get the number of motor steps and move the motors.” However, I do not see where to specify a value, eg., 200. Also, where do I adjust for microstepping?

[Anonymous]

In the configure.h: Is this correct?

// motor details
#define MICROSTEPS (16.0) // change this. microstepping on this microcontroller
#define DEGREES_PER_STEP ( 0.9) // change this. as advertised by the stepper motor maker

#define NORMAL_MOTOR_STEPS (360.0/DEGREES_PER_STEP) // 360/0.9=400. 360/1.8=200.
#define STEPS_PER_TURN (NORMAL_MOTOR_STEPS * MICROSTEPS) // default number of steps per turn * microsteps
#define PULLEY_PITCH (2*20.0) // 2mm per tooth, 20 teeth.
#define THREAD_PER_STEP (PULLEY_PITCH/STEPS_PER_TURN)
#define MICROSTEP_PER_DEGREE (STEPS_PER_TURN/360.0)

[Dan]

#define DEGREES_PER_STEP ( 0.9) // is a 400 step motor
#define DEGREES_PER_STEP ( 1.8) // is a 200 step motor

y-100 moves up? that’s backwards. one or more of your motors is plugged in incorrectly.

* connect robot
* open settings
* jog left motor “in”. pen should move up and left.
* jog left motor “out” pen should move down and right.
* jog right motor “in”. pen should move up and right.
* jog right motor “out” pen should move down and left.

This will tell you which motor is wrong. Do not unplug motors while power is on! It can damage the electronics.

[Anonymous]

Dan,
Every one of the jog commands execute exactly opposite. However, once I edited the configure.h file to “#define DEGREES_PER_STEP ( 1.8) // is a 200 step motor” and uploaded the Makeangelo-firmware, the thankyou.svg now plots proportionately, yet reduced in size approx. 50%, and inverted. This is indeed progress… we’re on the right track. Please tell me what to try next. Thank you.
Allen
fastal77

[Dan]

eactly backwards? your motors are wired wrong. if jog in goes away from motor then

* disconnect power
* unconnect motor plug from RUMBA
* reconnect with plug flipped over (wire colors reversed).

if you fix both of these and left in makes the right motor move, then you have to

* disconnect power
* unconnect both motors
* reconnect left motor into the socket where right motor used to be and vice versa. do NOT flip over plugs.

[Author]

Posts