CoreXYNZ build

Thanks for the input zelogik. Your work was a major inspiration for this project. I agree that the single cable chain is a good way to go however I will carry on with what I have until I do an upgrade or major service. I don't want to run the filament through the chain as I am wanting to use the brittle carbon filled filament.

Glad to hear that my project was a major inspiration for you :-D

Ok you use direct extruder, so yes brittle filament can be a pain. so the bowden extruder have an another advantage (except for the retractation).
And with problem for brittle filament with direct extruder i have an idea that i can patenting!
- Just add a very small "torque" motor to push the filament with a kind of spring system to let back pressure for retractation.

No seriously I think that Stratasys or pro 3D printer have a system like that, as you put your brand new cardbridge and your filament goes until entering into the extruder.

#RodF
Great job!

Are you satisfied with the 3 screws for the Bed?

Quote
zelogik
And with problem for brittle filament with direct extruder i have an idea that i can patenting!
- Just add a very small "torque" motor to push the filament with a kind of spring system to let back pressure for retractation.

No seriously I think that Stratasys or pro 3D printer have a system like that, as you put your brand new cardbridge and your filament goes until entering into the extruder.

Adding a second motor to "push" the filament through the tube won't help with brittle filament. Brittle filament breaks because of the bends in the feed tube. A second motor doesn't eliminate the bends. If you want to feed brittle filament (not really a good idea), just feed the filament straight down from above the extruder by mounting the spool over the machine.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

Unfortunately work is getting in the way of my project at the moment. However a break is coming up so I can give it the final push. I am hoping the 3 screw Z system will tend to cancel out any wobble and that the 2 linear rails will also hold it steady.
Yes DD Feeding the filament from above is the way I am going to go. Perhaps not the most attractive setup but simple and effective.

Edited 1 time(s). Last edit at 11/20/2016 04:42PM by RodF.

Wired the X and Y motors and the end stops. Powered it up and all 3 axis are working.



I'm trying out Fusion 360 and made my first part a pad for the Y end stop.

Stuck the heater and Print Bite onto the build plate and installed it. The plate has a fixed point at the rear and 2 sprung adjustable points near the front. I have the XY carriage moving correctly and homing. The XY stepper power settings are still on .5 amp and the Z is .8 amp. Working fine at these power settings so it should be pretty solid at 1.9 amps.


Edited 3 time(s). Last edit at 12/04/2016 06:56AM by RodF.

Why are you measuring stepper driver current in capacity?

Quote
Mikk36
Why are you measuring stepper driver current in capacity?

My mistake I meant amps.

Got the wiring finished. IR Zprobe seems to be working but I still need to sort out the firmware for it.

Had an issue with the extruder heater faulting. dc42 sorted it out. It was induction in the wiring harness. I bypassed the harness to test it. That allowed me to tune the extruder. I will probably rewire with twisted multi core cable.

I have been working through the commissioning process and have got to the printing stage. I haven't printed PLA for quite a while and had to resort to tape to get it to stick. I hadn't followed the instructions and heat soaked the Printbite for a couple of hours so I am doing that at the moment. The little flat extruder motor seems capable of handling a high extrusion rate. Duetwifi runs very quietly and I really like the web interface. Early days yet, there is a lot to learn still.

I've been doing more testing. I really like the IR probe, it produces a height map of the bed and then uses it to compensate for inaccuracies. I've been checking its accuracy with the dial guage and it seems pretty good. Repeatably seems good too. Here is an example of a 50mm interval height map it produces, it also saves a numerical file.

I produced a test part to check accuracy. It was designed as X100 Y100 Z101.5 It printed at X99.04 Y99.3 Z101.37 I expect the discrepancy is shrinkage rather than step/mm issue, however I might adjust to allow for it. I am getting a bit of banding on the Y axis corresponding to the 4mm Z scew pitch. That would make sense as the bed is less stiff in the X axis. (if that makes sense).



Edited 1 time(s). Last edit at 12/27/2016 06:18PM by RodF.

Possible to extend the bed support to reach the other side of the printer so you can install a 3rd vertical linear rail and move those 2 lead screw to that side?

Edited 1 time(s). Last edit at 12/27/2016 07:12PM by FA-MAS.

Yes that would work. However it looks worse than it is as I am holding it at an oblique angle to the light. Also I have yet to install the ACP sheet on the back which will stiffen it considerably. Still got the enclosure and filament reel to do.

I am happy with the squareness in all axis.

Check squareness by measuring diagonals of the squares in each plane. If they match, the axes are square (for that plane).

Is the bed heated, and are you using PID control or bang-bang? Those Z artifacts could be caused by bed temperature variations.

Edited 1 time(s). Last edit at 12/27/2016 11:08PM by the_digital_dentist.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

Quote
the_digital_dentist
Check squareness by measuring diagonals of the squares in each plane. If they match, the axes are square (for that plane).

Is the bed heated, and are you using PID control or bang-bang? Those Z artifacts could be caused by bed temperature variations.

Unfortunately due to warpage on this part measuring the diagonals is not a reliable test so I am using this dial angle guage. A bit hard to hold the camera and measure at the same time

The bed has a mains heater with SSR. At the moment it is bang-bang but I will do an auto-tune shortly and set up PID.

Hi,

The z banding you have look like to have a regular undulation, is the same was the pitch of the z screws?
If yes them is probably a problem with some of the ball screw is not rotating correctly.
One suggestion that maybe solve this, you could design and print 3 pieces that secure the top of each ball screw to the main structure.

Quote
filipeCampos
Hi,

The z banding you have look like to have a regular undulation, is the same was the pitch of the z screws?
If yes them is probably a problem with some of the ball screw is not rotating correctly.
One suggestion that maybe solve this, you could design and print 3 pieces that secure the top of each ball screw to the main structure.

Yes Filipe that's a good idea. Even one in the right place might be enough. I will wait to see how much the ACP sheet on the back works out. There is still plenty to do on this machine.

Edited 1 time(s). Last edit at 12/29/2016 05:28AM by RodF.

Quote
filipeCampos
Hi,

The z banding you have look like to have a regular undulation, is the same was the pitch of the z screws?
If yes them is probably a problem with some of the ball screw is not rotating correctly.
One suggestion that maybe solve this, you could design and print 3 pieces that secure the top of each ball screw to the main structure.

I'd think you'd want to constrain the bed from moving by adding another linear rail on that side rather than constraining the lead screws. Even if you constrain the lead screws, if the screws are even slightly not straight, it would still introduce wobble to the bed. If the bed is restricted from wobbling, but the screws are allowed a little wiggle room, that would be better.

Think about adding another linear rail on the opposite side of those 2. You may also "decouple" your ball screw nuts from the bed supports by removing the screws holding them. Gravity would keep the bed in contact with ball screw nut, and since the holes they fit into are slightly oversized, it'd allow a little axial movement that shouldn't translate to the bed. Though, you'd want to make sure they're at the center of gravity for your bed.

Edited 1 time(s). Last edit at 12/29/2016 11:53AM by FA-MAS.

Quote
FA-MAS

Quote
filipeCampos
Hi,

The z banding you have look like to have a regular undulation, is the same was the pitch of the z screws?
If yes them is probably a problem with some of the ball screw is not rotating correctly.
One suggestion that maybe solve this, you could design and print 3 pieces that secure the top of each ball screw to the main structure.

I'd think you'd want to constrain the bed from moving by adding another linear rail on that side rather than constraining the lead screws. Even if you constrain the lead screws, if the screws are even slightly not straight, it would still introduce wobble to the bed. If the bed is restricted from wobbling, but the screws are allowed a little wiggle room, that would be better.

Think about adding another linear rail on the opposite side of those 2. You may also "decouple" your ball screw nuts from the bed supports by removing the screws holding them. Gravity would keep the bed in contact with ball screw nut, and since the holes they fit into are slightly oversized, it'd allow a little axial movement that shouldn't translate to the bed. Though, you'd want to make sure they're at the center of gravity for your bed.

Well that makes sense. I do have some spare rail too. However I will wait and see how the ACP stiffens the 2020 that the Z rails are attached to. I can see them flexing if I push the bed side to side.

Instead of the expensive route of running another linear guide, I'd suggest using them as they were intended to be used- fully supported instead of end supported. By using them end supported you've got improved bearings compared to round rails, but they're still going to flex and let the bed move laterally, following any bend in the screws. Cut a metal plate, maybe 1/4" aluminum or a couple pieces of the t-slot material and screw it to the frame and the linear guides. That should improve things considerably.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

He has extrusions behind the linear rails on the Z axis. Looking at the parts, I'd say they're about 480mm long.
Looks like the only place he has them end supported are on the X

Using Misumi's deflection calculations, a single 2020 extrusion that's 500mm long, will deflect up to 0.5mm with a 10kg load
Using 2 like he has looks to deflect half as much.

Plus he's actually putting most of the weight on the lead screws.
I wouldn't suspect the extrusions/rails to be bowing outwards.

I'd expect them to accept any kind of twisting motion though.
Not so much at the tops or bottoms as he's got corner brackets there that should keep them from twisting there.
But more in the center I'd expect.

I'd second the plate idea Dentist advised, see if your ACP helps it.

If you have it, I'd say still put a 3rd linear rail on the other side and let those ball screw nuts float and wobble.

Edited 1 time(s). Last edit at 12/30/2016 01:26PM by FA-MAS.

Ummm on closer examination it seems I had forgotten to tighten the screws holding the rails to the 2020 frame. So the bed had been essentially floating on the screws. Much better now, but it will still get the ACP on the back.

RodF,

Not sure why you placed the limit switch for the Y on the gantry. Doesn't this require an extra cable carrier?

I agree the screws need float supports on the top, maybe bolted on the underside of the top alum extrusion? Heck, there just a block with a bearing to keep the screw centered, cheesy.
Problem with unsupported screws is they tend to whip, centrifugal force is easy to under estimate and will cause vibration (detaching your print from the bed) and torque your build plate even though it is attached to linear trucks.

Regarding the X-rails. While sure they're more than adequate for your needs, it's good you have two. First, they help keep the gantry square, and second, a single rail mounted vertical like you have will only allow one race in contact during Y changes, this way you have more a more centric X carriage. If they were mounted on their sides they would provide even better Y performance and still more than ample weight capacity with 4, however complicate your design some.

On the subject of the gantry, offsetting the idlers also offsets belt forces imparting some racking. Though should not be significant to be an issue, and will be masked by the use of 2 trucks per rail well spaced on the Y axis. I'm surprised you choose this approach over the proven collinear belts on the inside of the Y. From a design perspective, see why you used that belt layout, very clean lines, straight runs.

Otherwise, a really nice all metal machine. Should serve you well, for a very long time.

Not much work on the machine as I have been printing with it. I want to get the machine enclosed so I can try printing ABS-PC. I put a sheet of clear poly-carbonate on the base along with some rubber feet.

Hi!

Ive been reading thru this thread.
Very nice build and skills from you!
I got alot of input from it. For my own coming build.
Don't be afraid to make more posts here!

Pictures avd videoclips.

How big is the bed. May I ask how much is warps when you heat it up? (if you have measured it.

I like the duetwifi board. That probe looks very promising.

Hi!

Ive been reading thru this thread.
Very nice build and skills from you!
I got alot of input from it. For my own coming build.
Don't be afraid to make more posts here!

Pictures and videoclips.

May I ask how much is warps when you heat it up? (if you have measured it when heated?)

I like the duetwifi board. That probe looks very promising.