Kossel Calibration Problem

This was probably posted this originally in the wrong forum.

I picked a Mini Kossel as my first printer, but the openbeam that arrived was for a Legacy Kossel, so I built a Mini with the 360 mm base beams and 600mm tall. The rods are 312mm hole center to hole center and the base is 250mm diameter. It uses wheels instead of rods (quiet and smooth). My probe is offset to the right instead of the back and I intend to move to a servo, but right now it is the long allen wrench with a very small micro switch on the end, so the probe offset is x 15, y 15 and z -9.

When I issue a G28 it homes and with a G29 it calibrates with 29 samples (it seems a bit skewed to the front, but it stays on the glass). If I home it, the z says 185 which is what I entered for the . When I issue a G1 X0 Y0 Z0, the hot end stops 5 mm above the glass. I have changed both the MANUAL_Z_HOME_POSITION which was 180 (and that is what I measure, it was 175 before that) and the probe offset which was -5 (I have tried 0, 5, -5 and now -9). The position of the head after giving it the G1 X0 Y0 Z0 is always 5 mm above the glass. I have looked at the code and don't see anything obvious. The startup and calibration output follows:

07:17:14.841 : start
07:17:14.841 : echo: External Reset
07:17:14.841 : Marlin1.0.0
07:17:14.841 : echo: Last Updated: Feb 24 2014 07:15:39 | Author: (lreinhart, Mini Kossel)
07:17:14.841 : Compiled: Feb 24 2014
07:17:14.841 : echo: Free Memory: 2965 PlannerBufferBytes: 1232
07:17:14.841 : echo:Hardcoded Default Settings Loaded
07:17:14.841 : echoteps per unit:
07:17:14.841 : echo: M92 X100.00 Y100.00 Z100.00 E100.00
07:17:14.841 : echo:Maximum feedrates (mm/s):
07:17:14.841 : echo: M203 X200.00 Y200.00 Z200.00 E200.00
07:17:14.841 : echo:Maximum Acceleration (mm/s2):
07:17:14.841 : echo: M201 X9000 Y9000 Z9000 E9000
07:17:14.841 : echo:Acceleration: S=acceleratm/s), B=minimu acceleration
07:17:14.841 : echo:M204 S3000.00 T3000.00
07:17:14.841 : 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), E=maximum E jerk (mm/s)
07:17:14.841 : echo: M205 S0.00 T0.00 20000 X20.00 Z20.00 E20.00
07:17:14.841 : echo:Home offset (mm):
07:17:14.841 : echo: M206 X0.00 Y0.00 Z0.00
07:17:14.841 : echo:Endstop adjustement (mm):
07:17:14.841 : echo: M666 X0.00 Y0.00 Z0.00
07:17:14.841 : echoID settings:
07:17:14.841 : echo: M301 P7.00 I0.10 D12.00
07:17:18.569 : echoD init fail
07:17:18.632 : echo:Unknown command: ""
07:17:18.647 : FIRMWARE_NAME:Marlin V1; Sprinter/grbl mashup for gen6 FIRMWARE_URL:[github.com] PROTOCOL_VERSION:1.0 MACHINE_TYPE:Len's Kossel EXTRUDER_COUNT:1 UUID:00000000-0000-0000-0000-000000000000
07:17:53.482 : Bed x: 0.00 y: -120.00 z: 11.90
07:17:54.746 : Bed x: -80.00 y: -80.00 z: 12.70
07:17:55.791 : Bed x: -40.00 y: -80.00 z: 11.62
07:17:56.696 : Bed x: 0.00 y: -80.00 z: 10.80
07:17:57.538 : Bed x: 40.00 y: -80.00 z: 10.43
07:17:58.412 : Bed x: 80.00 y: -80.00 z: 10.62
07:17:59.457 : Bed x: 80.00 y: -40.00 z: 9.89
07:18:00.331 : Bed x: 40.00 y: -40.00 z: 9.70
07:18:01.095 : Bed x: 0.00 y: -40.00 z: 10.08
07:18:01.859 : Bed x: -40.00 y: -40.00 z: 10.82
07:18:02.686 : Bed x: -80.00 y: -40.00 z: 11.94
07:18:03.778 : Bed x: -120.00 y: 0.00 z: 13.08
07:18:05.042 : Bed x: -80.00 y: 0.00 z: 11.61
07:18:06.056 : Bed x: -40.00 y: 0.00 z: 10.50
07:18:06.992 : Bed x: 0.00 y: 0.00 z: 9.74
07:18:07.881 : Bed x: 40.00 y: 0.00 z: 9.45
07:18:08.708 : Bed x: 80.00 y: 0.00 z: 9.61
07:18:09.613 : Bed x: 120.00 y: 0.00 z: 10.38
07:18:10.642 : Bed x: 80.00 y: 40.00 z: 9.83
07:18:11.563 : Bed x: 40.00 y: 40.00 z: 9.58
07:18:12.358 : Bed x: 0.00 y: 40.00 z: 9.79
07:18:13.154 : Bed x: -40.00 y: 40.00 z: 10.45
07:18:14.027 : Bed x: -80.00 y: 40.00 z: 11.45
07:18:14.917 : Bed x: -80.00 y: 80.00 z: 11.77
07:18:16.009 : Bed x: -40.00 y: 80.00 z: 10.90
07:18:16.945 : Bed x: 0.00 y: 80.00 z: 10.29
07:18:17.803 : Bed x: 40.00 y: 80.00 z: 10.11
07:18:18.676 : Bed x: 80.00 y: 80.00 z: 10.45
07:18:19.831 : Bed x: 0.00 y: 120.00 z: 11.13
07:18:19.846 : 13.900 13.160 12.420 11.900 11.160 10.780 10.580
07:18:19.846 : 13.380 12.700 11.620 10.800 10.430 10.620 10.330
07:18:19.846 : 13.060 11.940 10.820 10.080 9.700 9.890 10.080
07:18:19.862 : 13.080 11.610 10.500 9.740 9.450 9.610 10.380
07:18:19.862 : 12.450 11.450 10.450 9.790 9.580 9.830 10.080
07:18:19.862 : 12.400 11.770 10.900 10.290 10.110 10.450 10.210
07:18:19.862 : 12.670 12.010 11.350 11.130 10.640 10.430 10.340

I was really excited to see it home and do the calibration. Does anyone have an idea why it is stopping 5 mm above the glass? Any suggestions as to how to proceed?

edit- another symptom, it reports that it is at x0 y0 z0 when 5mm off the glass, but when I tell it to go to x0 y0 z174 it is still a mm off the end stops. I changed MANUAL_Z_HOME_POSITION back to 175.

Edited 2 time(s). Last edit at 02/24/2014 01:39PM by Len.

when it says z=0 and it is 5mm above the bed, the z-height in the firmware is 5mm too low. Try changing it to +5mm, but approach slowly to z=0 then ;-)

As I said, I changed the z probe offset to a number of different settings and it made no difference, Same with the MANUAL_Z_HOME_POSITION. Is there another z-height? I tried going with a manual calibration and determined the correct MANUAL_Z_HOME_POSITION is 181.2, but it is still 5 mm off the glass if I try to do an auto calibration.

Thanks for the reply though.

that's probably a wrong configuration of the probe dimensions. It is found somewhere in the firmware (yes - sorry ;-)), something like z_probe_offset or the like (i can't check at the moment). It defines the position of the probe towards the hotend, if this value is wrong, the hotend will be positioned at the wrong place - in your case, 5mm too high.

Anyone found a solution? I have the same problem and i can't get trough it. I have changed manual home pos as well as the probe offset but nothing works for me.
The only solution i came up with is G92 Z4.7 command which changes Z0 to Z4.7 and it's not very elegant.

Quote
maciek117
Anyone found a solution? I have the same problem and i can't get trough it. I have changed manual home pos as well as the probe offset but nothing works for me.
The only solution i came up with is G92 Z4.7 command which changes Z0 to Z4.7 and it's not very elegant.

Z height.

Not probe.

Threw the z-Probe away, after it engraved the Heatbed. (Self-Deploying while printing...)
Did an extra round on manual calibration --> perfect. (Deviation of heatbed <0,0... --> smaller than I can measure...)
If - somewhere in the future - object does not stick to heatbed anymore, I will recalibrate...

Quote
Sir_Death
Threw the z-Probe away, after it engraved the Heatbed. (Self-Deploying while printing...)
Did an extra round on manual calibration --> perfect. (Deviation of heatbed <0,0... --> smaller than I can measure...)
If - somewhere in the future - object does not stick to heatbed anymore, I will recalibrate...

what method did you use to calibrate the printer?
i hear of people doing it without a probe i just dont see how it can get everything.

Papertest.

Get a Piece of 80g/m² Paper, travel around on the plate, do several Testpoints
For each testpoint
step down with pronterface, until the nozzle grips the paper. --> Paper is 0.1mm
Enter G-Code M114 to get current Position.
gently try to get paper out, while it is beeing gripped by nozzle -->
if it is easy --> Distance is about 0.10mm
if it is hard --> Distance is about 0.05mm
if paper is destroyed --> Distance is about 0.00mm


Adjust "Delta_Diagonal_Rod" and Endstops, until paper is gripped with same force and at same height on every testpoint.

You could even redo test with 40g/m² paper --> Distances are half the above mentioned (0,000/0,025/0,050)

Quote
xile6

Quote
Sir_Death
Threw the z-Probe away, after it engraved the Heatbed. (Self-Deploying while printing...)
Did an extra round on manual calibration --> perfect. (Deviation of heatbed <0,0... --> smaller than I can measure...)
If - somewhere in the future - object does not stick to heatbed anymore, I will recalibrate...

what method did you use to calibrate the printer?
i hear of people doing it without a probe i just dont see how it can get everything.

Not that much to get assuming you have a flat bed that is perpendicular to the towers + having some clue as to how long your rods are (can be calibrated rather precisely later).

Z-height, endstop ajdustments and delta radius.

If your build is not so correct then you might want to add angles or/and tower offsets.

Quote
Koenig
Not that much to get assuming you have a flat bed that is perpendicular to the towers + having some clue as to how long your rods are (can be calibrated rather precisely later).

Z-height, endstop ajdustments and delta radius.

If your build is not so correct then you might want to add angles or/and tower offsets.

For me, this is Standard calibration. Z-Probing and/or Papertest are useless, if you did not do Standard calibration.

Quote
Sir_Death

Quote
Koenig
Not that much to get assuming you have a flat bed that is perpendicular to the towers + having some clue as to how long your rods are (can be calibrated rather precisely later).

Z-height, endstop ajdustments and delta radius.

If your build is not so correct then you might want to add angles or/and tower offsets.

For me, this is Standard calibration. Z-Probing and/or Papertest are useless, if you did not do Standard calibration.

Exacly, this is the calibration, a Z-probe does not add anything under the pretenses I wrote.
That was my point.

It just takes a little bit longer doing it manually than having a probe, but it's mainly a one time thing. (If you don't tinker with your printer all the time, then it is very convenient with a probe)

Edited 1 time(s). Last edit at 11/04/2015 05:09AM by Koenig.

Quote
xile6
what method did you use to calibrate the printer?
i hear of people doing it without a probe i just dont see how it can get everything.

If you are good at math then you can use the paper method with this maxima notebook: [github.com]
Not that I would recomend it. A probe is easier: [forums.reprap.org]

Did not know this maxima Notebook - did everything by hand, and retry / Change / retry / Change / retry..... - tok me about 3 Hours, but then you know the neccesarry G-Codes while you are sleeping

I started like that too. Then got pissed and wrote the math down