Core-XY based on Rat Rig V-Core and Duet 2 Ethernet

snoozer

The think is that I am not really clear how to "adjust" anything. The angles are set by the extrusions and the corner plates. The extrusions length is a perfect match, there is no visual difference. I will try if I can somehow make the double-t nut fix at different angles I suppose. Are those double-t nuts a good way to do it actually ? From feeling while assembly it felt much better to work with the drop in single t-nuts that came with the kit.

Jan

PS: I wonder, would it be better to leave 0.5 - 1mm room in the corners for the upright posts or is squeezing the horizontal extrusions agains the uprights (what I did) the way to go ? I am thinking leaving room could give maybe room for adjustments, on the other hand assuming the extrusions profile are the same fixed measurement all around squeezing it should result in a square.

mrehorstdmd

The angles should be set by the square cut ends of the t-slots and the plates should only reinforce the corners.

You can test the t-slot for square cuts by standing the t-slot pieces up on their ends on a table. Stand them next to each other. It will be obvious if they aren't square because the pieces won't stand parallel to each other.

If the ends are square, you can tap the center hole at each end of each piece, drill tool access holes, and bolt the t-slot directly together using button head cap screws whose heads fit in the t-slot. If any corner reinforcement is required, add plates later. If the cuts aren't square you might be able to shim the ends where they bolt together using strips of metal from a beer can.

alt text

If the plates have to be used to force square joints I'd probably use a rafter square to adjust for squareness of each joint as I proceed with the assembly.

alt text

I tried to include a couple images, but they don't seem to be working today.
Here are links:
https://cdn.instructables.com/FYM/776V/IA57RRJB/FYM776VIA57RRJB.LARGE.jpg
https://www.amazon.ca/Empire-Level-2990-Magnum-Rafter/dp/B0002YSAX8

snoozer

@mrehorstdmd

Thanks for explaining that so well. I did not stand them all up individual, but I have stood them on the cut all side by side and they looked straight. I put a metal ruler across to see if they are all same length and they where. If that's 100% true now IDK but nothing was obviously out of line. The V-Core does not really work the way you showed in the picture. Here how they intend it to go together:

Sure I could drill the parts and put them together as you suggested. I have had it on my steel table now and it stands straight, no wobble nothing. Another thing that came to mind is that I did not check the end mounts of the X axis if they are square to the actual X extrusion. Those are 3D printed, there is obviously a change they are a bit off.

Jan

Alexander Mundy

Maybe a little late, but to square the ends of extrusion I use a table style disc sander. A cheap one will work as long as you get the table and slide 90 degrees to the sanding disc which isn't tough with a spare piece of extrusion and turning it 90 degrees at a time and lightly touching it then observing the sanding marks and adjusting according. Once all extension ends are square I take a spare piece of long extrusion and put 2 set of t nut/ washer/round spacer/washer/bolt and space them apart so the shortest piece of extrusion in the same length group barely fits to make a jig. Then lightly sand the rest in the group to the same length checking often for go/no go in the jig. This will get you extremely close to the same length extrusions without a caliper long enough to accurately measure them.

mrehorstdmd

There's probably enough slop in the fit of the nut-strips in the t-slot to allow the joints to be assembled squarely. Pick up either a machinist's square or a rafter square and have another go at it. You can test the square for squareness by drawing a vertical line with it placed on a board, then flip the square over and check that the line you drew follows the edge of the square. Even cheap plastic rafter squares are usually square. Machinist's squares are "adjustable" and should be checked before first use.

What are those corners made of?

snoozer

@mrehorstdmd

The corner plates are made of aluminium.

http://www.ratrig.com/hardware2/openbuilds/plates/90-degree-joining-plate-black.html

I don't have a square of any kind here, I checked with the frame on a piece of cardboard and drew a line, turn it around and the line matched. Sure it may not be a 100% solution but I move forward from here now with the belt installation. I have a square next week and check again then.

PS: I do actually have the print bed which is square, yes I need to do a bit of adjusting. I better do that now before moving to the belts.....

snoozer

That was a lot easier than I thought, I have no experience with these extrusions, there is indeed enough play with those double t-nuts to square it up. Also the holes in those corner plates are bigger then the bolts, so there is some play there as well. It was only out by approx 0.5mm on a distance of about 320mm.

snoozer

Hello,

so the build is at a stage where I have wired up my Duet2 Ethernet to have the following connected:

Steppers
Endstops (NC)
Hotend Fan
Hotend
Extruder

And I am running into trouble already. When I try homing the X or Y axis it moves to the endstop, and then instead of backing off a little it keeps ramming into the switch. I am pretty confident I have the switches configured right.

; Endstops
M574 X1 Y1 Z1 S1

In the machine properties I can see the state of the endstops correct as hit when it should be. Running the homing of for example the X axis and pressing the switch manually I can observe how it tries to home fast 1st, and the speed reduces BUT it does not reverse direction. Here the homex.g content:

G91 ; relative positioning
G1 Z5 F6000 S2 ; lift Z relative to current position
G1 S1 X-305 F1800 ; move quickly to X axis endstop and stop there (first pass)
G1 X5 F600 ; go back a few mm
G1 S1 X-305 F360 ; move slowly to X axis endstop once more (second pass)
G1 Z-5 F6000 S2 ; lower Z again
G90 ; absolute positioning

Where it says go back a few mm it does actually not, it keeps moving to drive against the switch.

Physically left should be home and that's where the switch is. So G1X5 should move the head 5mm to the right but it does not. Have I maybe overlooked something elsewhere that would explain this ?

Thanks
Jan

snoozer

OK, IDK what went wrong but I have reconfigured it from scratch now with the RFF Config Tool and it now works.

Endstops in config.g
; Endstops
M574 X1 Y1 Z1 S1 ; Set active high endstops

Homing in homex.g
; homex.g
; called to home the X axis
;
; generated by RepRapFirmware Configuration Tool v2 on Mon May 06 2019 14:46:11 GMT+0100 (Irish Standard Time)
G91 ; relative positioning
G1 Z5 F6000 S2 ; lift Z relative to current position
G1 S1 X-305 F1800 ; move quickly to X axis endstop and stop there (first pass)
G1 X5 F6000 ; go back a few mm
G1 S1 X-305 F360 ; move slowly to X axis endstop once more (second pass)
G1 Z-5 F6000 S2 ; lower Z again
G90 ; absolute positioning

Jan

snoozer

I have been playing around with speeds, acceleration and currents. My concern was that I would not get the speed out of my 0.9 degree steppers. I was so wrong, the head if flying around REAL fast, way faster than ever needed for a print. Even the bad moves pretty quick. Light at home is pretty crap, did not bother taking a video yet.

The print bed is a bit of a worry however. I have 2 motors for the Z axis which are mechanically independent. When playing around with speeds etc I would ofc come to the point where the motors just stall. Unfortunately not both at exact the same time. While that is only a matter of adjustment I am concerned that if I have build up of material on the nozzle and the head does catch on a blob of plastic that my motors ger out of "sync" and the bed is no longer level. Also I dislike the idea of flexing couplers between lead screw and stepper shaft. They are provided with the kit and I have them put in for now. Is my concern about the two motors getting out of sync justified or am I just to worried ? Would you put a belt between the two ?

My bed is only temporary and I dislike a 4 point mount anyway. The kit did not come with any means of bed adjusting. I just got a few springs and printed brackets for them to rest in and push against the bed from below.

I am not 100% happy with that at all. The bolts are supposed to go into the small slot, they are M2 or something small. I will modify that to a 3-point mounting and countersink M5 bolts in the print bed. I wont bother with magnets on this. The final setup would be a 230V AC heated bed from E3D or a machined aluminium plate, not sure yet.

Jan

dc42

Motors will stall if you set the maximum speed (M203) and/or maximum acceleration (M201) too high. If you set the Z parameters in those commands to values that your printer can do reliably, then it's most unlikely that you will get one Z motor stalling and not the other.

So with the proper settings, you needn't worry about the Z motors getting out of sync during a print. OTOH they may get slightly out of sync when you power the printer off and on again. If your printer has a Z probe, you can resync them before printing as described at https://duet3d.dozuki.com/Wiki/Bed_levelling_using_multiple_independent_Z_motors,

snoozer

@dc42

Thanks for the reply. A Z probe is on my bucket list for this build. I have an inductive probe but I wont bother mounting it cause I want to be able to print on glass and prefer to give a piezo probe a shot. My speed tests have been VERY much on the edge, not at all settings for real printer set-up. I have reduced the speeds a good bit to have lower motor currents and reasonable practical speeds instead of show-room effect

Thanks
Jan

snoozer

One more thought about the bed and end stops. I have one end stop at the low logical end of the z axis for now which means homed the bed is lifted all the way up. That seems to be what I have seen on most or all printers in fact. its not really very practical from a operational perspective. To overcome this I can see two options.

A: Modify the homeall.g to have the bed drop all the way after homing to allow easy access and do the same after every print. Not clear yet how that would work when using a Z probe.

B: Having the end stop at the bottom (logical high end) and home the bed there. Again I have not yet looked into Z probing, no clue if this would be an option.

Which is the better approach, anyone has practical experience to share related to this ? It only really come into play where the bed moves for Z instead of the tool I guess.

Jan

BlueDust

@alexander-mundy this maybe a little late, but still how I did it with my V-Core. I used long vice/clamps to hold it together, from corner to corner diagonally. So when I was squaring one side/square I had an X made with two vice/clamps. I with then measure to make sure the corners diagonally we're each the same distance from each other and tighten down all the screws. For good measure (), I would also measure the sides of the squares distance from each other to make sure they were also all the same, but they have to be after the corner to corner diagonally were already the same (just checking my work I guess). Once I had one side done moved on to the next.

snoozer

@bluedust
Thanks for the tip, I am short of "large tools" at home that why I used my print bed as square. I have to bring a tape measure from work to measure the diagonal to be sure its identical both ways. With my "square" is have it done so I have no visible gap which should be OK. My X axis front and back have the same distance from the edge of the frame on both ends which is also an indication for the lot being square. Have not yet printed anything as I am still in the process and the bed is not yet done. I am sure I will take it apart again anyway, 2 of the double t-nuts have lost their thread because I was to generous with the force tightening the bolts. The slot in nuts seem to be a lot stronger and I think I prefer them. I have a bunch of them ordered.

Jan

snoozer

My current Duet 2 Mount can be found here:

https://www.thingiverse.com/thing:3613809

The Fusion 360 file is here:
https://a360.co/2Vm7KSC

Its my temporary solution until I have a final setup with proper print bed. Its out of the way on the upper rear 2040 extrusion, easy access and view of limit switch LEDs and so on. I have no active cooling as its exposed and my motor currents are below 1A.

Picture was taken before wiring was complete, so a bunch of cables are not shown here. Printed on the Prusa MK3 with Prusament PLA Galaxy Black.

Have FuN!
Jan

Phaedrux

@snoozer said in Core-XY based on Rat Rig V-Core and Duet 2 Ethernet:

One more thought about the bed and end stops. I have one end stop at the low logical end of the z axis for now which means homed the bed is lifted all the way up. That seems to be what I have seen on most or all printers in fact. its not really very practical from a operational perspective. To overcome this I can see two options.

A: Modify the homeall.g to have the bed drop all the way after homing to allow easy access and do the same after every print. Not clear yet how that would work when using a Z probe.

B: Having the end stop at the bottom (logical high end) and home the bed there. Again I have not yet looked into Z probing, no clue if this would be an option.

Which is the better approach, anyone has practical experience to share related to this ? It only really come into play where the bed moves for Z instead of the tool I guess.

Jan

Homing to Zmin is more common I think because it's far more important to have that set accurately than the Zmax. It's just a lot easier to adjust the distance between the nozzle and the bed. Of course, regardless of how you home the Z axis, where it finally rests is up to you. You could home to Z0 and then move it to the bottom. But that's a lot of long travel moves on a slow axis.

Using a Z probe lets you automate that Z0 procedure and opens up the door to using other bed leveling and compensation mechanisms, which philosophical opposition aside, do have their place. At the very least, having an accurate way of finding and setting the Z0 position is a requirement for a good first layer. That's just easier to do when homing to Z0.

It's definitely possible to do both though, and in fact on a CoreXY with a moving bed, having a way to reliably home to Zmax is very useful for resume after power loss functionality. In my case, I use a z probe for homing and have an optical endstop on Zmax. The optical flag is adjustable with a screw, so I find Z0, move the bed down 300mm, and then adjust the flag until it triggers the endstop. Then I have a macro to homeZMax that does a simple homing move to the max end of travel and sets position to 300mm.

There's going to be some variation over time due to differences between each homing trigger height being slightly different, or thermal expansion and drift loosening things up, but it seems to work already for resuming the print on layer heights I use. You can tell there is a subtle layer line if you look at it, but it's better than a failed print.

Personally, I wouldn't want to home to Zmax every time because of how long it takes and since it's going to have to just come all the way back up at the start of the print. If I need access I have a macro that places the bed half way down and moves the gantry to the middle. Or I use the homeZmax macro to get it all the way down.

snoozer

@phaedrux

I understand what you say re proper positioning for the 1st layer. I think for now I probably go with the end stop at Z 0. You mentioned using a macro for homing to z max. That's something I have not looked into yet at all. Using marcros would ofc make things very easy once I have a handle on that. I wont be home now till Thursday so I am not hands on to look into it.

Regarding speed, my Z moves pretty fast if needed, IDK the value of the top of my head now but that was one of the things that drove me mad on the Prusa how slow Z moves.

Does your setup have two mechanically independent Z motors ? How do you make sure your bed is level in the geometry of the printer. I have seen some printers especially with ball screws and rails drop the bed once steppers are powered down. Mine does not do that but if it does you intervene by hand to level it somehow ? I was thinking of a mechanical strong end point like adjustable threaded spacer under the nut for the lead screw. that is level to the geometry, or in other words parallel to the gantry above to have a known point to make sure i am definitely level. Is it possible to define stall detection for the z max and end stop by NC switch or Z probe for z 0 ?

Thanks for your input, much appreciated.
Jan

deckingman

@snoozer Just to say that macros are nothing more than a sequence of commands that get run one after the other, and that homex.g, homey.g, homez.g and homeall.g are in fact macros.

snoozer

@deckingman If I remember right (not at the printer now) on the right in DWC was a list of shortcuts or buttons to click to call macros, I just have not yet changed the defaults to what makes sense for me. I am sure I figure it out.

Regarding homing z to 0 or max, my main concern is that I have not yet figured out how I safely move z to z max without previously homing to z 0. was thinking of a hard mechanical stop at z max but i would have to have stall detection enabled for only z max, not z 0 or x and y 0, looks as if that is not a possibility. I was then thinking of using E0 or E1 as end stop for z max but I am not sure such configuration is even possible. The RFF Config Tool does not seem to allow for that.

I am just thinking out loud here. Sounds all a bit weird I guess but in practical reality once I have the printer enclosed I would need to be able to move the bed to z max to access a print that may be on the bed after the printer was switched off. I like to have a button for that which does not require home z to 0 1st with a print on the bed and no access from above.

Jan