How Do I Calibrate My Titan Extruder and Test Print?

[AaronGreen]

I tried finding an existing thread for this but wasn't able to...

The printer is printing above the established Z height calibration.

I calibrated the BLTouch offset to G31 X-63 Y0 Z2.6 P50. At this setting, the nozzle sits half a paper's thickness off the bed at G1 Z0.

However, when I print anything it prints [visibly at a guess] at Z1, or 1mm off the bed. Babystepping in down, even up to 5.5mm, does nothing. It's printed successfully in the past, usually half a mill higher than Z0 for some reason, and I'm able to babystep it down just to get the print away, but this is out of control.

I've included the Cura start code I set up a while ago, but haven't really discussed it with anyone here. I'm not sure if it's part of the issue. I thought maybe it has something to do with the executing, seeing that the Duet calibration tests are good, and the print data maybe wrong...

Config.g

; General preferences
M111 S0 ; Debugging off
G21 ; Work in millimetres
G90 ; Send absolute coordinates...
M82 ; Using Cura
M555 P2 ; Set firmware compatibility to look like Marlin
M208 X0 Y0 Z0 S1 ; Set axis minima
M208 X340 Y250 Z280 S0 ; Set axis maxima
M552 S1 ; Start WiFi Module

; Endstops
M574 X1 Y1 Z1 S0 ; Define active low and unused microswitches
M558 P5 X0 Y0 Z1 H10 F200 T6000 ; Set Z probe type to modulated, the axes for which it is used and the probe + travel speeds
G31 X-63 Y0 Z2.6 P50 ; Set Z probe trigger value, offset and trigger height

Cura Custom Start Code:

;FLAVOR:RepRap
;TIME:9941
;Filament used: 4.49025m
;Layer height: 0.1
;Generated with Cura_SteamEngine 3.0.4
T0
M190 S60
M104 S210
M109 S210
M82 ; absolute extrusion mode
Start Code
M280 P0 S160
M420 Z2 ; set fade leveling at 2mm for BLTouch (OK to send for Non BLTouch)
G28 ; home all
G29 ; bed probe
M82 ; relative extrusion
G92 E0 ;zero the extruded length
G1 X0.0 Y50.0 Z10.0 F3600 ; perform wipe and prime
G1 Z0.0 F1000
G1 Z0.0 Y70.0 E9.0 F1000.0 ; prime
G1 Y100.0 E12.5 F1000.0 ; prime
G92 E0 ; zero extruder again
;end start code
M82 ; relative extrusion mode
;LAYER_COUNT:398
;LAYER:0
M107
M204 T2500
M566 X1800 Y1800
G0 F3600 X176.318 Y114.382 Z0.3
M204 P1500
M566 X1200 Y1200
;TYPE:SKIRT
G1 F1500 X176.664 Y113.853 E0.03154
... and so on...

[dc42]

Your start GCode is running G28 to home the printer followed by G29 to generate and apply the height map. If you run those two commands manually, does G1 Z0 send the nozzle to half a paper's thickness, at any XY position?

Are you doing Z homing using the BLTouch or with an endstop switch?

[AaronGreen]

@dc42 said in How Do I Calibrate My Titan Extruder and Test Print?:

Your start GCode is running G28 to home the printer followed by G29 to generate and apply the height map. If you run those two commands manually, does G1 Z0 send the nozzle to half a paper's thickness, at any XY position?

Are you doing Z homing using the BLTouch or with an endstop switch?

Z homing includes the BLTouch.

A manual G28 gives me a half-paper thickness G1 Z0. However, doing a G29 THEN G1 Z0 at various places shows G1 Z0 at +1mm.

I had a look at the G29 readout and this is what I get:

However again, the bed is physically flat, as in I have placed the nozzle a half-paper thickness in 6 positions (4 corners and 2 internal) and physically measured the level of the bed. The readout I get is FAR from the physical level of the bed. Any ideas why it's very different?

BTW, I've seen this before (and didn't have a solution at the time), but if I levelled the bed according to the G29 readout until it was perfectly green the bed physically measured as obscure as this readout is.

I should also add that it's no longer doing a 12-point G29. It's doing a 9-point check, because according to the web control some of the far probe points are outside the bounds of the bed. I haven't set up the probe points using duet before, but I have the instruction thread on hand. I'd think that'd be something to fix next, but not the major contributor to the Z0 discrepancy.

Thanks David.

[dc42]

It looks like the bltouch trigger height depends on the XY location. This is something you can check, by measuring the trigger height at various XY positions.

One possible cause of this is that the print head doesn't stay level. A tilting print head will change the relative heights of the nozzle and the Z probe.

[AaronGreen]

@dc42 said in How Do I Calibrate My Titan Extruder and Test Print?:

It looks like the bltouch trigger height depends on the XY location. This is something you can check, by measuring the trigger height at various XY positions.

One possible cause of this is that the print head doesn't stay level. A tilting print head will change the relative heights of the nozzle and the Z probe.

I'll check the trigger height around the bed, the same as I did with the nozzle height, but it should be consistent, considering both are statically fixed to the gantry.

I'll also double-check the height of each end of the X axis profile, but I've also done this several times in the past with beer cans, and once again, the nozzle moves across the bed consistent 1mm above Z0. When it comes to print, it always adds height. Could the G29 throw the Z0 out as a result of it's mandate to establish bed height, whereas a G28 just establishes a true Z0 in the place it homes to?

[dc42]

@aarongreen said in How Do I Calibrate My Titan Extruder and Test Print?:

Could the G29 throw the Z0 out as a result of it's mandate to establish bed height, whereas a G28 just establishes a true Z0 in the place it homes to?

If the problem was a general Z offset, then possibly. But that's not the problem that your height map indicates.

[AaronGreen]

I've rebuilt the bed, and checked the squaring of the frame, and everything is tight.

The only discrepancy is the tensioning wheels on the bed chassis are tight enough to keep it steady, but increase the height of the chassis on the left side by half a mill. It's also opposite to the bed calibration results.

All I'm going to do next is calibrate the bed to the results screen, which I've laboriously done before, instead of a physical level.

I tried doing that tonight, but the Duet won't start the WiFi module for some reason, I have no idea why. It won't even do it using Pronterface. So I'm going to come back to it tomorrow.

[AaronGreen]

Got the WiFi module working again. I don't know why but it assigned itself a different IP address at some point. At the time I was trying to connect to it, I was using the same IP address that appeared in the M115.

Anyway, the next issue was adhesion, again. I tried plain, clean glass recent and it worked perfectly, and now it's back to not sticking at all. So back to the glue stick, which started working again. I have to wet it and reapply glue to areas waiting to be printed.

Then the extruder started failing to consistently extrude filament. It'd be fine, then thin out for about ten seconds, then back to full extrusion again. The printer is in a still environment, so I don't know what the issue is. It's been printing fine for the few test prints that I manage to squeeze out. Now it's completely unpredictable. Also, the fan now turns off between the G29 and the actual print. I have no idea why, as I haven't change the G-code, and the print file includes 100% fan use. It's caused blockages when I wasn't watching.

Next, having fixed the bed level, it's now showing that every single print the printer thinks the bed is on a lean. So it's been levelled very well physically, but when it starts printing, it goes from 0.1mm on the left to 0.0 or -0.1 on the right. I have to screw the bed down as it's printing to fix the issue. But then next print, it requires that I do the same thing again. At this rate, the bed is going to be on a crazy angle because the printer keeps telling itself it's on a lean that doesn't exist.

And finally, I figured that the official E3D hotend is probably heating up too much with the fan pull the air through the heatsink. So I've gone and reversed the wires at the board end connection to blow the air through the heatsink instead. This appears to be a lot more effective regarding airflow. However, the extruder is still failing to work consistently by the time it gets to the infill, if that.

1. Bed level constantly fails, after trying a physically level bed or Duet calibration.
2. E3D isn't extruding according to the print, even though it's calibrated within 1mm.
3. 18 months of calibrating the same printer, and still can't print a single thing to fix my motorcycle indicator mounts.
4. Sell the entire printer for cash, order prints on Thingiverse, cast the prints and fix the bike.
5. Return to spending my time reading and writing, and leave the Duet guys in peace.

[dc42]

@aarongreen said in How Do I Calibrate My Titan Extruder and Test Print?:

Got the WiFi module working again. I don't know why but it assigned itself a different IP address at some point. At the time I was trying to connect to it, I was using the same IP address that appeared in the M115.

The Duet doesn't assign itself an IP address, it asks your router for one and the router assigns it. To make sure that the IP address stays fixed, you can configure your router to always assign the same IP address to the Duet ("address reservation"). In the unlikely event that your router doesn't provide that facility, you can tell the Duet to use a fixed IP address in the M587 command - but then you run the risk that your router might assign that IP address to something else.

... Also, the fan now turns off between the G29 and the actual print. I have no idea why, as I haven't change the G-code, and the print file includes 100% fan use. It's caused blockages when I wasn't watching.

Do you mean that the heatsink fan turns off? Have you confused the heatsink fan with the print cooling fan? Which fan output is the heatsink fan connected to, and how have you configured it in config.g? It may help if you post your complete config.g file.

Next, having fixed the bed level, it's now showing that every single print the printer thinks the bed is on a lean. So it's been levelled very well physically, but when it starts printing, it goes from 0.1mm on the left to 0.0 or -0.1 on the right. I have to screw the bed down as it's printing to fix the issue. But then next print, it requires that I do the same thing again. At this rate, the bed is going to be on a crazy angle because the printer keeps telling itself it's on a lean that doesn't exist.

Try firmware 2.0beta3. There is an interaction between using a Z probe for Z homing and using it for mesh compensation that can lead to confusion. In 2.0 the way they interact is changed, which leads to fewer surprises.

Also, do you have one Z leadscrew, or more than one?

[AaronGreen]

@dc42 said in How Do I Calibrate My Titan Extruder and Test Print?:

@aarongreen said in How Do I Calibrate My Titan Extruder and Test Print?:

Got the WiFi module working again. I don't know why but it assigned itself a different IP address at some point. At the time I was trying to connect to it, I was using the same IP address that appeared in the M115.

The Duet doesn't assign itself an IP address, it asks your router for one and the router assigns it. To make sure that the IP address stays fixed, you can configure your router to always assign the same IP address to the Duet ("address reservation"). In the unlikely event that your router doesn't provide that facility, you can tell the Duet to use a fixed IP address in the M587 command - but then you run the risk that your router might assign that IP address to something else.

... Also, the fan now turns off between the G29 and the actual print. I have no idea why, as I haven't change the G-code, and the print file includes 100% fan use. It's caused blockages when I wasn't watching.

Do you mean that the heatsink fan turns off? Have you confused the heatsink fan with the print cooling fan? Which fan output is the heatsink fan connected to, and how have you configured it in config.g? It may help if you post your complete config.g file.

Next, having fixed the bed level, it's now showing that every single print the printer thinks the bed is on a lean. So it's been levelled very well physically, but when it starts printing, it goes from 0.1mm on the left to 0.0 or -0.1 on the right. I have to screw the bed down as it's printing to fix the issue. But then next print, it requires that I do the same thing again. At this rate, the bed is going to be on a crazy angle because the printer keeps telling itself it's on a lean that doesn't exist.

Try firmware 2.0beta3. There is an interaction between using a Z probe for Z homing and using it for mesh compensation that can lead to confusion. In 2.0 the way they interact is changed, which leads to fewer surprises.

Also, do you have one Z leadscrew, or more than one?

Thanks for all your time and help, I have really appreciated it over the past year.

I've decided to sell the printer and just get my designs printed professionally. Working with the Duet is a bottomless vortex of time and emotion that basically is its own endless project. Ten months of calibrating across 6 different versions of firmware was a fair enough go for me, but not what I was looking for in a design and print process. I much prefer spending some change on a print that I get a guarantee on delivery, rather than absolutely zero guarantee with a Duet.

Good luck. Bye.

[dc42]

@AaronGreen, I am sorry to hear you are giving up. Have you considered asking your local Makerspace/Hackspace for assistance, if there is one?

Building your own printer and getting it to work well does often require a lot of effort. It took me two redesigns and over £1000 of expenditure to get my Kossel to the state in which "it just prints" - all due to mechanical issues of various sorts - and even now the Titan extruder still gives me problems occasionally (Bondtech extruder on order).

If designing parts is where your real skill lies, then unless you have well over £1k to spend on a good ready-built printer, getting them printed by others instead may be a wise decision.

Good luck!