So, let me first state that this question is entirely work related. I’ve run across different things on these DeltaMakers which have baffled me to no end. I’m only just now “giving up” and asking the community for help. I have a couple of different issues I’ve run into on a couple machines, and they have stumped me for a couple weeks. To the point of simply putting the machines in a corner and letting them sit there.
First issue, which has plagued us for a while - is an issue of the auto-leveling algorithm. What issues with build surface can the auto-leveling procedure not work through? On a bunch of machines, close to the outer edge, there seems to be a point where the nozzle (while printing) just drops into the build plate and scrapes along. Opposite of that, it lifts and prints do not stick to the bed. It’s the same with every machine it feels like. Low spot in the back left, high spot in the front and right. In the middle of the bed, about 130x130, everything is fine. What can the auto leveling algorithm of jcrocholl’s branch not compensate for?
Second issue, is one of “wavy” prints. I’ve resorted to printing a square test tower in order to check for printers which exhibit this behavior and we’ve only shipped ones that are arrow-straight. The print is a 20x20x80 square tower with 10% infill, 2 perimeters. There seems to be no rhyme or reason to the direction it goes. I have one printer that will output a tower where the left and right walls of the print are completely straight. The front and back of the print move smoothly, randomly, back and forth as the tower is built up. The vertical towers have been realigned, and they’re completely perpendicular to the X/Y axis. The tower carriages have been checked, and they are snug, with no play and no catching. I’m stumped. Could these issues be inter-related?
I don’t know the details of the auto-leveling algorithm, so I’m just talking very generically for thought – but when trying to “curve fit” a bunch of data points, the fit curves are only good within a certain window - and outside of the window, the curve diverges quickly. Does the auto-leveling you do include probing data from the area where you’re having problems? If not, then the fit isn’t being tested at that point, and the resulting polynominal may simply just be blowing up outside of where they were fitted for.
We have an industrial customer that has a modified MAX. Prints nylon cams for carpet looms. They print about 4" dia and up to 6 feet tall. They had a similar issue with vertical wave. They found that once the print started, there was just enough increase in bowden tube pressure that it would start to push/pull the nozzle a bit. They added an additional ten feet of bowden to go up and hang straight down. A bit much, but they said that solved it. On a note, their unit uses 3mm line.
First issue could be related to what sounds like the Y tower (front right side) being slightly IN as far as the geometry. That causes the low back/left issue. Delta’s always “point” to the tower that has some kind of issue going on
@J_O11 the low/high spots are actually happening on every machine. Is there a chance it could be the leveling algo? If we tilt the bed, run the auto-leveling algo again, the machine auto levels correctly with the pitched bed, but continues to have those high/low spots, in the same spot, on every machine. :[
Are you running marlin fw? It really sounds like an error in the dimensions in the fw or the actual machines y tower position. We’ve had our share of oops with one of our lasers some time ago, and it was doing exactly what you described, and the Y tower was inwards only by .5-.8mm Repetier fw has compensation for the towers positions not sure about marlin
It’s jcrocholls branch of Marlin. This is a machine all put together with T-slot and custom metal extrusions, the dimensions are great. In fact the towers are positioned via reference points on the extrusion itself.
Did you try slapping a straight edge on your structure pieces? Playing with the issues in my head it seems like bowed materials might produce some of it.
Also I had a little of the wavy things. I tracked it to bad bearings replaced them and resolved it. Never figured out what it was but I suspect misformed balls in the bearing. This led to a seemingly random wave dependant on how the funny ball rolled around the corner of the bearing onto the rod.
@Addidis_no , we found we were actually able to introduce the effect by different tensions of the V-wheels on the makerslide, so you are probably onto something.
One other thing that produced random waves, loose set screws on pulleys. This was a nightmare to find as it was barely loose but it would sometimes get the waves when it moved fast enough to slip it. Id also consider turning the trimpots on the drivers if there are any. Those things are touchy and its worth turning them down and resetting them just to be sure.
Check the gears/teeth the arms ride up for uniformity. Seems like the most likely place for something to be off. If the teeth skew it should present like this. Particularly important if tensions on the v wheels effect mesh on this gearing.
@Tom_Martz Were they able to measure that the error was not there while not printing? Sounds like an issue with the bend radius of the tube, which shouldn’t be affected much by the pressure of the filament in the tube.
Ive had this issue with a bad pulley, well actually belt if I put it correctly. I had some debris cathing into one of the belts and it produced exactly the same issue.
The other thing to check is pulley concetricity. If you have a bad pulley, this just might be the cause of this. Note, all this said after reading everithing else is mechanically ok and in mind that printed with PLA.
About auto leveling I really havent got a clue and really see no need to have it. I just do it after changing my hardware (maintanece, ect…), knok knok, havent done it for 6 month now.
