gMax Design & Assembly Problems and Work-Arounds

#1
Being fairly new to the gMax design, I’m still discovering design and assembly challenges that I’ve had to hack around. I want to put these issues on the table and also convey how I’ve worked around them on my unit — with varying degrees of success.

Let me say at the onset that I don’t necessarily know the correct names for some of the parts. Please bear with me on that front. I’d appreciate any corrections for misnamed parts below.

If you decide to make any of the modifications I’ve described below, be certain to read through this posting at least once in its entirety before you begin.


Keeping the Extruder Nozzle(s) Stable and Level

I have a dual extruder unit. While performing the initial “certification process” that I put the unit through when it first arrived, I noticed that the height of the nozzles relative to the bed was readily changing. On inspection, I discovered that the extruder motor bracket, and all the hardware attached to this bracket, was attached to the X-axis back plate via 3 screws. These screws actually pass through the 3D printed extruder X-axis plate and screw into nuts placed during assembly in an inset cavity in the X-axis plate. These nuts remain in place because these cavities are covered from behind by a metal plate behind the X-axis plate. (I’ll call this metal plate “the final back plate” until somebody tells me its correct name.) One of these 3 screws is near the bottom of the assembly, near the extruder wiring. The other 2 screws are to the left and right of the extruder(s), near the top of the assembly.

I've found that there are 2 problems in the current design that cause the nozzle heights to be unstable.

* All 3 of the aforementioned screws tend to vibrate loose during printer operations. This results in the relative nozzle heights varying significantly. This variability in the nozzle height results in poor surface quality and has caused many a job to fail due to weak bed adhesion. This variability is particularly troublesome during dual jobs when both extruders are heated, causing the nozzle that happens to be lower to ravage the surface of the current layer.



Bob Fix: After much experimentation, that included Locktite Blue (which didn’t fix the problem), I’ve added a lock-nut to both of the upper screws behind the metal “final plate.” Adding and tightening these lock-nuts was complicated. Each of these long (40mm?) screws need to pass through the holes in the motor bracket, through the X-axis plate, through the nut in the inset cavity of the X-axis plate, through the metal final plate, and now screwed into the nylon locknuts — all with just the right amount of tensions. The addition of these lock-nuts is even further complicated by the very limited space between the screws behind the metal final plate and the Teflon spacers of the upper roller bearings that ride along the X-axis beam.

* The area of contact between the extruder motor bracket and the extruder X-axis plate is fairly smooth on both parts. The area of contact for the X-axis plate is printed as the top layer of the part. The contact area of the motor bracket is the bottom layer of that printed part — resulting in a particularly smooth surface when printed before assembly on Kapton, Buildtek or other smooth bed surfaces. These 2 surfaces are so smooth that they vibrate out of position relative to each other almost without regard to how tight the 3 aforementioned screws are.

Bob Fix: Without increasing the stability of the grip between these two parts, they could still vibrate out of proper position over time. After some experimentation with various double-sided adhesives and cleverly folded fine grid sandpaper, (neither of which worked) the solution that has worked for me is a short strip of a 120 grit 1-inch wide sanding belt. The vertical spacing between the single lower screw and the 2 upper screws works out to be just over an inch. So I cut a strip of this sanding belt about 3 inches long. The front side of the belt is obviously rough. And yet, the canvas-like back of the belt also has sufficient grip for my purposes. I slipped this cut strip between the back of the motor bracket and the front of the X-axis plate before leveling the nozzles and tightening the screws.



Bob Fix: I had to make sure that the height of both nozzles of my dual-extruder rig were level — and stayed level — until the screws were permanently tightened. The process I used for leveling the nozzles begins with the assembly together (with the sanding belt strip in place) but with all 3 screws slightly loose. I knew that the assembly is loose enough because I could pivot the motors clockwise and counter-clockwise a degree or two along the X-axis. Then I used the Move Axis function of the firmware to lower the gantry in 0.1mm increments. I placed a single PostIt Note on the bed below the nozzles so that both of the nozzles came to rest on the PostIt as it was lowered. Similar to the old-school bed leveling process, I lowered the Z-axis slowly as I rocked the extruder assembly back and forth. — The objective being to allow both nozzles to rest with equal pressure on the PostIt, thus indicating that the nozzle heights are the same. I needed to continue to test for levelness as I incrementally tightened all 3 screws. I knew that the nozzles were permanently level once all 3 screws were fully tightened and the nozzles still exhibited the same pressure on the PostIt — even as I raised the Z-axis an increment or two.


Z-Axis Roller Bearing Tension Doesn’t Allow X-Axis Beam To Stay Level

There are 3 roller bearings running up each of the Z-axis beams. These roller bearings are screwed into a metal plate on each side. The X-axis beam is also attached to both of these metal side plates. The entirety of the Z-Axis beams, the roller bearings, the metal plates and the X-axis beam are supposed to true. “True” means that the 2 Z-axis beams should be parallel vertically and at the top and bottom along the X-axis. “True” also means that the X-axis beam should be vertically level.

But on my unit the X-axis beam “wants” to be about 4mm lower on the left side than on the right. Sure, I can twist the left hand Z-axis lead screw to raise it. But the tension caused by the frame actually being assembled out of true results in the Z-axis lead screws twisting back to where they want to be — with the left-hand side 4mm lower than the right.

Bob Fix: I need to manually adjust the height of the left-hand lead screw 3mm or so about once every day so that the software doesn’t have to work overtime trying to adjust to the untrueness (untruthfulness?) of the frame.

Little Known Fact: The Merlin firmware is able to raise and lower the Z-axis beam as it prints across above the build surface. But the final retraction move to the X and Y Homes does not use this auto-leveling capability. If the left side of my Z-axis is lower than the height of the last layer, then the still-hot extruder nozzle will scrap along the top of the freshly printed top surface — and even scrap the bed surface when the object just printed is only a mm or 2 high. Ouch!


PLA Dual Extruder Fan Shroud Melts If Right Extruder Is Hotter Than 225C

The extruder shroud on my dual rig was printed using PLA. The design for dual extruder configurations has an outlet that blows below both nozzles. So the part of the shroud that blows below the left nozzle passes very close to the right extruder. This will result in the shroud being destroyed from melting if the temperature of the right extruder is 225C or higher. So the current shroud doesn’t allow any material that requires an extrusion temperature of 225C or higher to be used in the right extruder.



Bob Fix: Don’t print any material requiring a temperature higher than 225C. Bummer.

Better Bob Fix: Beg somebody with a high-temperature extruder to print me another dual extruder fan shroud using polycarbonate. Any volunteers? (Please, please, please!)


Extruder Teflon Inserts Too Short For Ninjaflex

As the enclosed photo shows, the Teflon insert in my left-hand extruder is shorter than the insert in the right-hand extruder. Careful examination of the left hand extruder also shows that the filament drive feed is slightly to the left of the center of the Teflon insert. The combination of the feeder being misaligned and the insert being too short results in my not being able to print Ninjaflex using the left extruder.



Extra Credit: The careful reader will notice that I can’t use Ninjaflex in the right extruder because the high temperature required will cause the PLA fan shroud to melt. Yet I can’t print Ninjaflex in the left extruder because the Teflon insert is too short and the feed mechanism off-center.

- Bob