Unresolved Z axis drift?

My Z axis seems to "drift" from print to print. One run will get the first layer nicely "squished" so that the finished surface is smooth, the next will have lines visible, and another will fail because the nozzle is flying way above the bed. I'm sure I'm not doing something right in the setup. It senses the bed at nine points, and I can see the switches working.. What would cause this?

 

If the tip of your nozzle is encrusted with melted plastic it could be causing some measurements to be off. Also if the nozzle left a blob at one of the measurement points it could cause it to be off by a little bit.

Watch the switches when the print head is off to the right (X+) as having all of the weight on that side can cause one of the limit switches to trip much earlier than the other. To correct for this you can level everything using a dial indicator, or get it "close enough" by having the print head off to the right side and manually turn each of the z-axis rods until the switches engage evenly.

Lastly do not use the software based autolevel function in MatterControl. The software autolevel competes with the firmware one and weird things happen.

 

Pictures are always appreciated. Is your X axis level? Making sure you can slide the carriage back and forth a papers width above the bed will help reliability.

 

No gunk on the nozzle.
I have leveled out the two Z axis motors the best I can at this point. I do see it driving the Z motor as the X and Y motors run, so I take it that means I'm not completely level. I never see more than about 1/10 revolution.

I don't know how to not use the autolevel in matter control I haven't found where that lives yet.

 

Don't even bother looking for it. The autolevel in MatterControl is not needed. Autolevel is built into the firmware.

You just want to make sure that the start G-Code lines up.

...
G28 ;home all axis
M565 Z-0.9 ;Define a 0.9mm offset, this may need adjusting
G29 ;run autolevel routine
...

The "0.9" can be anything, the default is 1.0.

 

Under "Custom G code" I have:

Start G code:

G28 X0 Y0 Z0 ; Home all axes
G1 Z5 F5000 ; Lift Nozzle
M109 S[temperature]; Set the extruder temperature (The line literally says "[temperature]" not any numeric value

G28 Z0 ; Home Z again in case there was filament on nozzle
M565 Z0
G29 ; Probe the bed

I don't know Gcode but shouldn't the second to last line end with a semicolon? Or does it matter if no comment?

End G-Code:
M104 S0; Turn off temperature
G1 X10 Y200
M84; Disable motors

Not sure why the second to last line is there.

Every time I print, when starting a new print, I need to home the head, then manually (In Matter Control) take the head down to the deck, then back up 0.5mm. Then I hit Home again, and it seeks and finds the zero on Z. When I do this little dance before printing, it seems to work well.

Apparently my bed is also not level, no idea yet how to do this. While the printer is moving in X or Y I see the Z motors moving slowly. Not much, maybe 1/10 turn. I take that as my clue that the bed isn't level.

 

There are two ways to level the x axis relative to the bed. The preferred is to use a dial indicator and adjust each side until it reads the same on each side of the printer. Make sure the carriage is moved from side to side while making the measurement to account for its mass.

The low tech option is to put a piece of paper on the bed and manually turn each threaded rod until the paper barely has any resistance while the print head is on either side. Slide the print head to the left, turn the left threaded rod until the print head just touches the paper. Slide the print head to the right and repeat. Make sure you prevent the opposite head from turning when adjusting, they like to move in unison.

 

[temperature] is a software call. it fills in that value with whatever your starting temp is

doesn't need a semicolon,

G1 X10 Y200 is the park position

You shouldn't need to do any of that manual stuff. No idea what you mean by it works well, but it's completely unnecessary

The levelness of the bed is accounted for by the autolevel sensors. That's the whole point.

 

I would bet the left side z axis limit switch is tripping when the print head is off to the right.

 

Ok. that makes sense.

I agree, but if I don't do that ritual between prints, then the next print starts with the head about 3mm above the bed and it makes a stringy mess.

 

Hmm.. I don't know the internals, I was assuming the switches were in series. Possible. If the circuit would tolerate the leakage current, I'd be tempted to put an LED and series resistor across each switch so I know when they are open.

 

Keep terminal open and watch the G29 responses.

You're essentially putting a bandaid over a problem we could easily fix if we knew what was going on.

 

Yup. Agreed. I'm just so new to this that I don't much know where to look.

Is there a way to capture the terminal output to a file? It will be a while, I'm in the middle of a print now.

 

I don't think so. I believe you can copy and paste it from MC, might need to use the snipping tool to make a screenshot.

 

Here's the output at the start of a print:
Very reminiscent of Gerber photoplotter files.

->N1 M110 S1*96

<-ok
->N2 G21*24

<-ok
->N3 M107*38

<-ok
->N4 M190 S70*91

<-ok
->N5 M104 T0 S215*34

<-ok
->N6 M104 T0 S215*33

<-ok
->N7 T0*61

<-echo:Active Extruder: 0
<-ok
->N8 G28 X0 Y0 Z0*80

<-ok
->N9 M114*46

<-X:0.00 Y:0.00 Z:0.00 E:0.00 Count X: 0.00 Y:0.00 Z:0.00
<-ok
->N10 M105*22

<-ok T:212.0 /215.0 B:70.3 /70.0 T0:212.0 /215.0 @:85 B@:0
->N11 G1 Z5 F5000*52

<-ok
->N12 M109 S210*88

<-T:212.9 E:0 W:?
<-T:213.1 E:0 W:?
<-T:213.8 E:0 W:?
<-T:214.2 E:0 W:?
<-T:214.4 E:0 W:?
<-T:214.3 E:0 W:?
<-T:214.4 E:0 W:?
<-T:214.3 E:0 W:?
<-T:213.8 E:0 W:?
<-T:213.6 E:0 W:?
<-T:213.5 E:0 W:?
<-T:213.2 E:0 W:?
<-T:212.5 E:0 W:?
<-T:212.3 E:0 W:?
<-T:211.9 E:0 W:?
<-T:211.4 E:0 W:?
<-T:211.2 E:0 W:?
<-T:210.6 E:0 W:9
<-T:210.4 E:0 W:8
<-T:210.0 E:0 W:7
<-T:209.5 E:0 W:6
<-T:209.5 E:0 W:5
<-T:209.3 E:0 W:4
<-T:209.1 E:0 W:3
<-T:209.1 E:0 W:2
<-T:209.3 E:0 W:1
<-T:209.2 E:0 W:0
<-ok
->N13 M104 T0 S210*16

<-ok
->N14 M105*18

<-ok T:209.4 /210.0 B:70.6 /70.0 T0:209.4 /210.0 @:51 B@:0
->N15 G28 Z0*109

<-ok
->N16 M114*16

<-X:0.00 Y:0.00 Z:0.00 E:0.00 Count X: 0.00 Y:0.00 Z:0.00
<-ok
->N17 M105*17

<-ok T:209.3 /210.0 B:70.0 /70.0 T0:209.3 /210.0 @:54 B@:0
->N18 M565 Z0*86

<-ok
->N19 G29*42

<-Bed x: 15.00 y: 20.00 z: -0.06
<-Bed x: 107.00 y: 20.00 z: 0.06
<-Bed x: 199.00 y: 20.00 z: 0.34
<-Bed x: 200.00 y: 130.00 z: 0.14
<-Bed x: 108.00 y: 130.00 z: 0.02
<-Bed x: 16.00 y: 130.00 z: -0.06
<-Bed x: 15.00 y: 240.00 z: 0.28
<-Bed x: 107.00 y: 240.00 z: 0.26
<-Bed x: 199.00 y: 240.00 z: 0.49
<-Eqn coefficients: a: 0.00 b: 0.00 d: -0.13
<-planeNormal x: -0.00 y: -0.00 z: 1.00
<-ok
<-echo:endstops hit: Z:0.49
->N20 M105*21

<-ok T:209.8 /210.0 B:70.6 /70.0 T0:209.8 /210.0 @:51 B@:127
->N21 T0*9

<-echo:Active Extruder: 0
<-ok
->N22 G90*32

<-ok
->N23 G92 E0*118

<-ok
->N24 M82*47

<-ok
->N25 M107*18

<-ok
->N26 G0 F9000 X78.351 Y105.234 Z1.099*31

<-ok
->N27 G1 Z1.099*72

<-ok
->N28 G1 F1800 E0*40

<-ok
->N29 G1 F540 X79.089 Y104.746 E0.11035*55

<-ok
->N30 G1 X79.888 Y104.331 E0.22265*98

<-ok
->N31 G1 X80.706 Y104.01 E0.33225*88

 

Can you test the tightness of your Z axis helical couplers? Grab the threaded rod and see if you can twist the helical coupler.

 

You mean the part that the threaded rod attaches to? No slop there. The tops of the threaded rods wave a bit, I will be making a stabilizer for them shortly.