DC42 firmware, multiple topics after first run

Hi Dave, printed on my big delta for the first time tonight and I have some questions and observations:
First of all, I have not changed any of your code from before when I was running the duet on my old mini kossel. The version I was using ok before and on the current setup is 1.09g-rc1-dc42 (2015-08-22).

1. My scale is way off, like 5%. 68mm prints as 72mm. Since autocal does all the work I'm not sure what is wrong. The only inputs I have are the rod lengths and the bed diameter and the rod lengths are super accurate. (of course the steppers counts for the motors are good)
2. The heater keeps turning off after I do an autocal run.
3. I can't find the reference about how to calculate the H parameter for the z probing. Please provide it as an equation, ie H=center probe-spot probe.
4. Autocal is getting about .1 mean dev, with spots that are as large as .25mm difference (using the H values as calculated above). If I reverse the sign of the H values, I get more like .16 mean dev.
   I know this is a reflection of the mechanical accuracy; is this a reasonable number for this size of system ie errors are within measurable limits for a system this size. It can be difficult to measure distances and squarenesses with just sticks and rulers, since I don't have a humongous digital calipers.
5. The custom effector platform and carriages have 80mm rod spacing. It probably is magnifying dimensional errors although it supposed to make it easier to tighten tolerances and reduce the impact of play.
   Here are the nozzle/probe offsets as manually measured for each spot. This is what I am basing my H values on.
   G1 X0 Y0 Z3 F5000 ;2.47
   G1 X150 Y0 Z3 F5000 ;2.68
   G1 X106 Y106 Z3 F5000 ;2.47
   G1 X0 Y150 Z3 F5000 ;2.41
   G1 X-100 Y106 Z3 F5000 ;2.34
   G1 X-150 Y0 Z3 F5000 ;2.30
   G1 X-106 Y-106 Z3 F5000 ;2.24
   G1 X0 Y-145 Z3 F5000 ;2.35
   G1 X106 Y-106 Z3 F5000 ;2.63
   G1 X100 Y-40 Z3 F5000 ;2.46
   G1 X66 Y40 Z3 F5000 ;2.40
   G1 X0 Y80 Z3 F5000 ;2.51
   G1 X-75 Y25 Z3 F5000 ;2.37
   G1 X-60 Y-50 Z3 F5000 ;2.38
   G1 X0 Y-90 Z3 F5000 ;2.28
   G1 X35 Y-50 Z3 F5000 ;2.40

Thanks

Edited 3 time(s). Last edit at 10/15/2015 03:16AM by shadowphile.

Quote
shadowphile
First of all, I have not changed any of your code from before when I was running the duet on my old mini kossel. The version I was using ok before and on the current setup is 1.09g-rc1-dc42 (2015-08-22).

I suggest you upgrade to 1.09k-dc42.

Quote
shadowphile
[*] My scale is way off, like 5%. 68mm prints as 72mm. Since autocal does all the work I'm not sure what is wrong. The only inputs I have are the rod lengths and the bed diameter and the rod lengths are super accurate. (of course the steppers counts for the motors are good)

Are you using 7 factor autocal? If you are, then switch to 6-factor so that it leaves the diagonal rod length alone. Also, check that the diagonal rod bearing spacing is exactly the same at the upper and lower ends. If it isn't, that will cause the effector to tilt as it translates, which will affect the scale.

Quote
shadowphile
[*] The heater keeps turning off after I do an autocal run.

Sounds like something in your bed.g file is wrong. Please post it.

Quote
shadowphile
[*] I can't find the reference about how to calculate the H parameter for the z probing. Please provide it as an equation, ie H=center probe-spot probe.

H = actual Z probe trigger height at that probe point, minus the trigger height you set in the Z parameter of the G31 command. Typically you set the G31 Z parameter to the trigger height at the centre of the bed, then use the H parameter to make corrections if needed at other points.

Quote
shadowphile
[*] Autocal is getting about .1 mean dev, with spots that are as large as .25mm difference (using the H values as calculated above). If I reverse the sign of the H values, I get more like .16 mean dev.
I know this is a reflection of the mechanical accuracy; is this a reasonable number for this size of system ie errors are within measurable limits for a system this size. It can be difficult to measure distances and squarenesses with just sticks and rulers, since I don't have a humongous digital calipers.

That's about double the errors I see on my printer. I use 13-point 6-factor calibration. I have a Check Bed macro, which uses the same code as bed.g (including same H corrections) except that it doesn't home at the start and it uses S-1 instead of S6 on the last G30 command. It reports about 0.065mm deviation (about the same as autocal does) and the worst spot error is about 0.13mm.

Quote
shadowphile
[*] The custom effector platform and carriages have 80mm rod spacing. It probably is magnifying dimensional errors although it supposed to make it easier to tighten tolerances and reduce the impact of play.

I think a large rod spacing will reduce dimension errors rather than magnify them.

Quote
shadowphile
Here are the nozzle/probe offsets as manually measured for each spot. This is what I am basing my H values on.
G1 X0 Y0 Z3 F5000 ;2.47
G1 X150 Y0 Z3 F5000 ;2.68
G1 X106 Y106 Z3 F5000 ;2.47
G1 X0 Y150 Z3 F5000 ;2.41
G1 X-100 Y106 Z3 F5000 ;2.34
G1 X-150 Y0 Z3 F5000 ;2.30
G1 X-106 Y-106 Z3 F5000 ;2.24
G1 X0 Y-145 Z3 F5000 ;2.35
G1 X106 Y-106 Z3 F5000 ;2.63
G1 X100 Y-40 Z3 F5000 ;2.46
G1 X66 Y40 Z3 F5000 ;2.40
G1 X0 Y80 Z3 F5000 ;2.51
G1 X-75 Y25 Z3 F5000 ;2.37
G1 X-60 Y-50 Z3 F5000 ;2.38
G1 X0 Y-90 Z3 F5000 ;2.28
G1 X35 Y-50 Z3 F5000 ;2.40

Looks to me like you probably have effector tilt varying with X position.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

I am using 6-factor because I am sure my rods are within .1 total (better actually; I used a precision jig with zero play to assemble them).
(printer is running right now but I will check, maybe I entered a bad number)

I was also very careful to ensure the rod spacing was spot on 80mm at all 6 locations, for this very reason. However, that doesn't account for all the other ways the carriages can be misaligned. An 80mm spaced carriage is sensitive to torque when mounted on a narrow 20x20 extrusion, plus the OpenBuilds v-wheel/v-rail parts turn out to not like axial loads, to my chagrin. I am kind of stuck with that for now until I can redesign something better. I will have to find a way to check the other alignments of those carriages.

Your equation for the H parameters is the one I ended up using, so good.
Here is my bed file. I left a lot of commented stuff in for historical reasons and what is left isn't much.
-------------
; Auto calibration routine for delta printers
; Before running this, you should have set up your zprobe Z offset to suit your build, in the G31 command in config.g.

M561 ; clear any bed transform, otherwise homing may be at the wrong height
G31 X0 Y0 ; don't want any probe offset for this
G28 ; home the printer

;*** Remove the following line if your Z probe does not need to be deployed
;M98 Pdeployprobe.g ; deploy the mechanical Z probe

; The first time the mechanical probe is used after deployment, it gives slightly different results.
; So do an extra dummy probe here. The value stored gets overwritten later. You can remove this if you use an IR probe.
;G30 P0 X0 Y0 Z-99999

; Probe the bed and do 6- or 7-factor auto calibration
;G30 P0 X-52 Y-30 Z-99999 ;H-.09 ; X tower (center is 1.28, X tower is 1.19, so make H -.09 to correct)
;G30 P1 X0 Y-75 Z-99999 ;H-.04 ; between X and Y towers
;G30 P2 X62 Y-42 Z-99999 ;H.03 ; Y tower
;G30 P3 X62 Y35 Z-99999 ;H-.01 ; between Y and Z towers
;G30 P4 X0 Y60 Z-99999 ;H-.04 ; Z tower
;G30 P5 X-52 Y30 Z-99999 ;H-.07 ; between Z and X towers
;G30 P6 X-11 Y-44 Z-99999 ;H-.02 ; half way to X tower (actually about halfway to X-Y mid-tower position because of IR pcb)
;G30 P7 X36 Y-15 Z-99999 ;H.02; half way to Y tower
;G30 P8 X0 Y30 Z-99999 ;H0 ; half way to Z tower
;G30 P9 X0 Y0 Z-99999 S6 ; centre, and auto-calibrate 6 factors


; This is for the 300mm bed (H offsets = probe-spot height - center height) Center height is set in config.g file.
G30 P0 X150 Y0 Z-99999 H0.21
G30 P1 X106 Y106 Z-99999 H0
G30 P2 X0 Y150 Z-99999 H-0.06
G30 P3 X-100 Y106 Z-99999 H-0.13
G30 P4 X-150 Y0 Z-99999 H-0.17
G30 P5 X-106 Y-106 Z-99999 H-0.23
G30 P6 X0 Y-145 Z-99999 H-0.12
G30 P7 X106 Y-106 Z-99999 H0.16
G30 P8 X100 Y-40 Z-99999 H-0.01
G30 P9 X66 Y40 Z-99999 H-0.07
G30 P10 X0 Y80 Z-99999 H0.04
G30 P11 X-75 Y25 Z-99999 H-0.1
G30 P12 X-60 Y-50 Z-99999 H-0.11
G30 P13 X0 Y-90 Z-99999 H-0.19
G30 P14 X35 Y-50 Z-99999 H-0.07
G30 P15 X0 Y0 Z-99999 S6 ;centre, and auto-calibrate 6 factors


;*** Remove the following line if your Z probe does not need to be retracted
;M98 Pretractprobe.g ; retract the mechanical Z probe

G1 X0 Y0 Z150 F5000 ; get the head out of the way of the bed
____________________________
I have a 'just report bed heights' macro exactly like yours.

It is good to know how accurate your system is considering the size is similar to mine, gives me a target to shoot for. I ended up with a 350mm glass bed and heat plate, with a 300mm silicone heater (500 watts!).

I ran into another new problem last night.
I decided to leave a print going to get some sleep and tried out the Sleep check box (M1 command).
I got up this morning and the printing had froze in place with the nozzle in a solid pile of ooze, seemingly right after I went to bed. The heat was turned off and the interface said the job was 100% complete.

I don't know what the Sleep check box on the web interface is supposed to do. I will check it out when I am back at my printer.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

never mind about sleep button, the model has stopped at the same spot 3 times now. Still trying to figure out why. Been messing with the slicer settings but no go. It pulls up fine in the online gcode viewer. The web print status says it is 100% finished. Is there a log somewhere?
On a slightly different tact, the model imported into the interface is missing info like estimated filament usage. This has happened many times with other models although I can't recall if it is 100%.

I suggest you attach your gcode file to a post (you may need to zip it up) or make it available on Dropbox etc., and also tell us at what height the print stops.

Are you using the latest (version 1.09k) firmware?

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].