Dual extruders on a printer can have the tips at the same Z plane, but the router bit tips have to plunge to some extent into the workpiece, and if they move together, then cutting with one is likely to crash the other.
Or like David said, it would be sufficient to have a mechanism like a solenoid for a fixed lift distance to bring one tool up (relative to the other) while they both move on a common Z. That could be accomplished by some kind of toggle mechanism and then it wouldn’t need a fully independent second Z. And therefore it wouldn’t need all the firmware that dual-Z would entail.
A single X core has a working space 180 mm narrower than the rail length. (From here)
Having two X cores with one just sitting there (taking up space) means the other one has a working space 360 mm narrower than the rail length. And since the reachable working space of X1 is offset from the reachable working space of X2 by 180 mm, the overlapped region where both tools can operate is 540 mm narrower than the rail length.
For a 48" wide zone accessible to both tools you will need rails at least 1220+540=1760 mm.
Is there any chance you can list the changes you had to make to firmware?
Or a link to the thread where you configured the firmware?
I know you used an OCTO board which made it so you didn’t need to play with the PINS file, but could you maybe add in the changes that you would need to make if we are going to have to use a “standard” 6 driver board?
I know that I am going to do this mod, sooner rather than later and there seems to be great interest in it from other people. I am not sure of the feelings of people using the OCTO or not, but i will most likely reuse my existing 6 driver board.
SKR 1.2 was more straight forward to setup thanks to the V1E docs, precompiled firmware, help from this forum, and it’s archive. SKR worked great until I killed my SKR. Personally, my feelings for Octopus haven’t changed.
Reusing SKR 1.2 and utilizing that 6th driver for IDEX extension would be neat.
I will have a look at that link you shared, but as i am using a different board (SKR EZ V3) it was a bit of fun to get going and the standard firmware is a great place to start.
Sigh, now to go and read that thread. While not looking too stupid.
I see a 2 y tubes for rails in one of your builds. Would you say is better to track the y axis on both sides, i don’t like the y axis to be like that (also didnt liked the skate wheels on my lr2 i used 2 supported linear rails and bearings
The right side of the design normally calls for a rail, and I used a rail on the left side primarily so I have something hard and flat, instead of soft wood. The rail I can shim to be flat, whereas wood is hard for me to cut straight and soft pine can get smashed and any flat surface can collect chips.
In my design I am not trying to constrain the rollers in the X direction on the left side. Only the right side rollers constrain X. This is important because if both are constrained then any errors between the rails will cause the two YZ plates to fight each other and it can produce poor stiffness in the X direction as the whole gantry can slide/rock in the X/Z plane. If each side were a linear rail instead of rollers then it might not rock but it would have a permanent stress on the gantry and the Z rails might not slide so easily.
So even though I’m using two metal rails, I don’t think it’s a good idea to constrain both in the X direction.
Nice. Round tube sounds precarious when not square, are you using larger more forgiving EMT on left side, e.g. 1"?
Didn’t understand the benefit of round tube vs slat strap or box metal until you mentioned this is self clearing. Very cool. Your round tube approach seems less precarious than box metal on 45deg tilt. Guessing EMT is cheaper too?
I’m using regular 3/4" EMT. Larger EMT or 45 degree square angle would all have the same tolerance, which is that the highest point must be within the width of the 608 roller bearing. The brackets for holding the tube (made for the right side) have room for some adjustment in the X direction, so the size of the frame doesn’t have to be perfect within 1 mm. (And a modified bracket can be made too if it’s still too far off.)
I am also thinking maybe next time the tube mounts could be mounted to the face of the ribs in this type of design:
The attachments for the ends of the belts will still need support, so the side slats can’t go away completely, unless the belt ends were mounted to the ends of the metal tubes somehow…
I had a roof fall on it and it sustained very minor damage. It should be easy to fix, but I don’t have a good place to put it at the moment so I haven’t been diligent about it.
But I do want to measure its stiffness, so I need to get it back to working condition even if I need to put it in a poor location temporarily.
Ugh so sad to read that man. I keep dreaming about a full sheet capable lr3 using 2 cores. I keep reading every 2 days the configuration you used to marlin to see if i can learn it and take the challenge
Just read the whole topic and… Wow… Impressive work, and a truly innovative mind!
I particularly love that this is a flexible solution, you can do multitool as well as running two cuts in parallel with the same setup
One Idea comes to mind though…
I’ve seen a lot of questions regarding the loss of working area…
Wouldn’t it be a lot easier (altough more pricey) to duplicate the whole LR assembly and have two lowriders sharing the Y axis?
It’s often easier to have a longer table than a wider table and X axis
Home/Garage workshops are often not that wide and, anyway reaching to the back of a 2meters wide table is no fun
Plus the Y axis won’t suffer from being longer stiffness-wise
You could even build a “lowrider train”, quickly adding/removing modules depending on the current work
Maybe it’s not as cost effective approaching the cost of two lowriders unless you factor in the cost of the table, but I think it may be easier for some setups
Control board would be tricky though… Maybe some sort of klipper stuff with an sbc that controls both “machines” through USB…
You could even imagine a single 3-5meters long table so that you have two CNC back to back for batch production, or a single multitool cnc, depending on your needs
Or, if you find a way to get the front belt out if the way (eg: belts inside unistruts on the side of the table, or belts running flat in top of the table), you’d get a multitool CNC and a workbench back to back that can transform into two cncs)
Dual core is surely the simplest/cheapest way, but if you want to go all-in or need to maximize production rate, this “dual Y idex” setup might be worth a try
I agree, it is pretty simple to put a second machine on a single set of Y belts, although it would require some attention to the homing routine. Most likely you would home one machine toward -Y and the other toward +Y, but it would be doable.
The LR machine typically has two Y, two Z, and one X, and two machines wouldn’t share any motors so you would need 10 drivers in total. Maybe klipper can control all these motors from one controller but more likely you will have two controllers so that you can home and square Z and Y for each one. This would probably mean splitting the job into two gcode files and the coordination starts to get complicated if you want machine 2 to begin after machine 1 finishes.
You will most likely end up with two fully independent machines sharing a table, and manually start jobs on each of the machines at the appropriate time. This is by no means a bad setup, and you could cut two independent smaller jobs on separate parts of the table simultaneously, which IDEX can’t do. So in some ways this is a better setup, the downside being that you can’t do T0/T1 to switch tools, which is not that big a deal.
I don’t think it is easier, because the IDEX approach is so darn simple hardware-wise to get working. A second carriage sharing the X belt and an extra endstop is in some ways very similar to sharing two Y belts and two extra Y endstops. The firmware for IDEX is fairly straightforward in that the code doesn’t need any hacking and it works with just configuration updates.
I agree that an increase in width is generally worse than an increase in length, especially if you have an existing LR3 and want to add a second tool. Shrinking usable X space has a greater chance of becoming too small to get the job done, compared to shrinking Y, and it’s a lot more work to change the X machine size.
The other option, which is expensive but should be doable, is to get a true ATC and line up a bunch of tools along the end of the table. The crossed-gantry MPCNC and the tubes of the LR2 are somewhat tricky to work around, but with the open geometry of the LR3, it should be relatively easy. The cost is higher than IDEX and higher than a second machine, but you get to keep a narrow table and as a bonus you get more than 2 tools.
That’s what I had in mind
Two separates controllers coordinated by a single klipper board
I think they’re seen as a single “printer” afterward
It’s initually meant to keep the cost down when you need a lot of axis, as you can but multiple cheaper controllers (an skr pico with 4 tmc2209 costs 20-25$, you can get 8 outputs for less than 50$)
so sad to read that man. I keep dreaming about a full sheet capable lr3 using 2 cores. I keep reading every 2 days the configuration you used to marlin to see if i can learn it and take the challenge