Z-lift issues

[Xenocrates]

So, I've been printing a good bit, most recently on brackets for a heated chamber. I found that after the first lower bracket, for whatever reason, I ended up with the hotend catching on a portion of the raft and pulling the whole raft out of alignment. I started with a Z lift of .3MMs, which had been working fine previously. Then I switched it to 1MM Z lift, which worked for the next two. Now it's doing it again, and it seems the easiest solution other than increasing Z-lift again to an even more ridiculous number is slowing the printer when it does the raft, especially on the parts where it's traversing over itself.

I'm fairly certain this isn't a bed leveling issue, since my bed was checked to .04MMs consistency earlier, and I don't think it's an extrusion issue, since the strands it grabs are flat to the bed and level before it catches. It may be that the printer is over-extruding some on the move and thus the filament coming out is catching, but it doesn't really seem like that happens as it's slowed down. It seems like the printer is not respecting Z lift when running off the SD card.

Does anyone else have experience with this, and if so, have you found a better solution?

[teoman]

Usually catching happens when your parts start to curl up.

So increase it for now, then you can go to extensive trouble shooting when you have your enclosure setup.

[Earthbound]

What slicer?

Retraction and/or wipe settings are critical, too. A Z-lift with a full nozzle just creates high spots.

[Xenocrates]

Slicer is matterslice, retraction is 2MM with an E3D V6, which is on the edge of too much for some people, and teo, I was looking at the part as it occurred, I had good bed adhesion. This is a first layer only problem it appears. Curling I would hope would not occur on the first layer. I'll be looking into it more, but it may just be gremlins. We'll see once the enclosure is done. And I haven't got any wipe set as yet, just the retract, since I figured wipe was largely meant for dual extrusion. I'll test it with wipe later, but for now I'll increase the Z lift further and watch the prints as they go so I can potentially slow it down to 50% on the move. it seems like if the raft was printed somewhat slower, such as if it followed the first layer settings, that it would be utterly fine, but I suppose that the raft isn't considered the first layer.

[geneb]

If your first layer is curling, that means the bed adhesion is poor. Fix THAT before you worry about z-lift.

g.

[Xenocrates]

If your first layer is curling, that means the bed adhesion is poor. Fix THAT before you worry about z-lift.

g.

That's the thing. The first layer is not curling. It is bending when caught by the hotend. Looking at it, the nozzle appears to leave a very slight bump on the strands of the raft it passes over, so it appears to be a combination of not respecting the Z-lift in the raft, and not enough retraction. Cause there is no way the thing is moving the whole 1.5MMs up and still leaving that trace. Then again, weirdly, matterslice appears to treat the raft entirely separately from the part, as it also isn't bothering with the 12% first layer speed, reserving that for the first layer of the actual part.

[teoman]

Hnmm.

Check if your endstop screws are loose.

The ones that hold the switches in place

[Xenocrates]

Teoman, this appears to be a slicing issue, as when I ran the G code through a parser, it appears matter control just doesn't bother with any of my settings on the raft. The endstops are secure, and besides which, if those were lose, it would not be such a consistent problem. It occurs in the same place each time, in the same way, at the same time. Honestly it looks like I need to bother the matter-control folks about their slicer doing silly things. Yes, if the endstops were lose it would present problems, but mostly in the system not having a good zero reference to work from, which might cause tilted parts or just plain failures of bed adhesion. But this is not an adhesion issue. It's seems somehow my matttercontrol install is not taking into account my settings for the print when it creates rafts. Which is rather bizarre.

[geneb]

I don't have MC in front of me, but doesn't it have speed settings specifically for rafts? Make sure you've got the configuration level set to Advanced(?) in order to see _everything_.

g.

[Xenocrates]

Raft has fan speed settings, and the air gap, expand and extruder settings. No independent speed. It appears that Z lift isn't the issue at all, but rather that the raft refuses to use it. I've looked at the Gcode by hand, in mattercontrol's viewer, and in editNC, and none of them show the raft using any Z lift at all when it moves. Compared to the now easily visible Z lift on moves in the main part. Time to go register a gripe with matterhackers, I suppose.

[Right to Make]

I think I may chime in here. You might be experiencing what I am, and that is during movement of the axes, doing lateral movements, or movements that are parallel to the print bed, the head can dip. I have not found the source for mine as of yet but I take off my hot end to mount a dial indicator in its place. I then run a tower calibration macro to get a feel for the radius, and to check bed level. This macro homes then comes down to the center and stops, then i set to zero on the indicator, then it does not lift but it moves to each tower in a straight line. (This is the same macro you can find in the manual. ) During the movement I notice while the macro runs at a normal printing speed, the dial gauge creeping downward towards the glass. It is instant and variable. This happens at the start, stop, or at random parts of any particular move. Mine creeps around.003", to .007" to .010" which is enough to knock a tall part over. Mine, it will come down to the center and i can set that to zero and it will be repeatable all day long. Then when the movement stops at each tower I can get that to zero as well and it is repeatable. Its the movement from a stop to start is where i see a dip or a raise in the head. I have been trying to find a way to compensate for this but its just to random. Please if this did not make sense let me know i can try this another way. I should just make a video. But I think this is what you are running into. I had to make a new lathe handle for my lathe, and it was tall, around 4 inches tall, and went through 3 attempts and each one was near the end when it failed and it was printing while it tipped, not on a rapid move. For the most part it is not a problem as my machine is producing some good prints out of PETG, ABS, and Bridge Nylon. It is a good printer but I am contemplating making new effector rails from aluminum and new joints out of C-954 Aluminum Bronze. Just from a rigidity standpoint and see how stiff I can get things using a belt driven system. I was thinking that it is flexing when its moving from binding from one of the rails, maybe one is twisting or something, it probably wouldn't take much for things to bind and that might cause a dip or a raise when moving. Maybe go aluminum with the effector as well just to see if i can get rid of all the shaking. I think it would need to be slower movements then because of the extra weight, but there i go again rambling. Sorry. Let me know what you think. This has been just my three cents.


Side note. I noticed on my stock rostock max v2 that there was a slight amount of play if i were to put my finger on the bottom side of the effector and push straight upwards. This play was from one of the connections of the rail and the joint. The hole in the rail was bigger than the pin in the joint creating some play. This was causing some minor print issues as i placed some of my wife's hair bands from the effector to the cheapskate on each rail. This improved overall print quality very much but did not solve any dipping movements. I then had to go to walmart to buy her some new hair bands.

[Right to Make]

Not trying to hog the board but I have noticed that using matterslice there is no retraction made from a support/raft to the print. No retraction, no z lift. you can have z lift with 0 retraction but the retraction will make the z lift a lot more apparent to our eyes especially if you set your z lift very low like around .1mm of lift. I cant tell it lifts. I end up with strings around the support layers or rafts. Matterslice doesn't retract doing supports then going to the print but slic3r will. I end up using slic3r for things with support for this reason, but matterslice is usually my go to slicer.

[Xenocrates]

sounds like a H-radius issue tyler. I've gotten a mail from matterhackers support saying they'll work on Z lift and similar in the raft. Carbon fiber is likely preferable to aluminum and aluminum bronze, both from a weight and rigidity standpoint, as the lower weight will put less stress on the entire system. You might also look at the new ball joints, although I'm not sure if you have them or not, given your join date. They have much less play. You may also want to look at your belt tension. With regards to the H-radius adjustments, you may want to check out the spreadsheet here: http://forum.seemecnc.com/viewtopic.php?f=77&t=5906

[Right to Make]

Thanks for replying. The H radius seems ok or locked in, because the towers are at the same level as the center. and its not uniform diping. its just the movement dips when accelerating. I have played with the jerk and accel settings but its all reverted to stock numbers. My bed is level and i can run a print and use the dial gauge and watch the movement dip when the head changes direction. I think it is enough to leave marks on the top of a flat part because of the amount of dip. That is what the initial post was about, it was about markings or possible dragging marks. And if you have a tall part it hits or drags slightly because of this dipping it can potentialy knock over parts. In my case i did that lathe handle that was tall but did not have a big enough base and just didnt want a raft on it that is why mine toppled because of the dip in the head movement. I have excellent bed adheasion. Maybe ill post a video later this weekend of this so its clearer. But thanks for the reply, and maybe this will help someone else as well.

[Kaidos33]

I had a similar issue. To isolate the issue on the lift, I printed rectangles 120 mm X 10 mm X 5 mm. I aligned it perpendicular with each tower. I notice when it was perpendicular with Z it had more lift. Turn part a 45 degree, CW then CCW, then notice the difference. The further away from center the worse the lift was. All this lead me back to Calibration of towers, squaring them up with bed. I found that the square I started with wasn`t square. Purchased a roofing square that has T foot or lip on one end. Z-Tower was out of Square at about 12 inches up. The smaller square looked good at 6-8 inches. But at 12 inches up I had a 1/32 inch gap.

Rostock Max v2

[Right to Make]

That sounds like a great way to trouble shoot towers. But are you talking about squaring the towers to the glass build plate that is on the onyx? Or are you going off the machine itself. I cant square anything on the onyx as there is a low spot between X and Z tower. Adding the glass build plate makes it flat but with the dip in the onyx, the plate is not square, so I take the onyx and glass off and use the snowflake as I know it is flat and the main board is flat as well. And I use a 12 X 24 square that reaches all the way to the top of the tower, and they seem square. I did not like to square it off the glass plate as I know the onyx is not flat and if its not flat you cant count on the glass to sit square on it. What are your thoughts of this? I was always just going to shim the glass off the onyx with stainless steel heat wrap foil to get the dip out but haven't made it that far, and I haven't needed to use that area of the bed.