Direct Belt Drive Router

[keyne]

Zoners,
welcome to my build log
I will build a woodworking router, work envelope about 2100x800x100mm which will be driven by servos directly on toothed belts. The gantry will be driven from one side only (we had a separate discussion about this) and i want to use structural extrusion by MK (Maschinenbau Kitz).
The linear guides are BWC T4 dual-vee bearings riding on angle iron screwed to special profile from MK (see pic).
Z-Axis is also directly driven by a 200W servo with brake and helical rack/pinion.
The work surface is t-track made from OSB, i will also drill holes into it to hold bench dogs and "wonder dogs" (see lee valley catalog) for part fixturing.

Gantry beam is a 50x200mm extrusion. Plenty stiff... Rest of the structure 40x40 and 40x80 extrusion and 10mm aluminium plate.

Will use UHU Servo controllers. The Z motor has a 1024 steps encoder, want to get at least the same for the big motors (530W) but they have 10mm axles which makes it hard to find matching encoders.

Average cutting force will be 100-150N.

I made some 3D sketches to show you. Not every nut and bolt is shown..
All the yellow components are aluminium, light blue is steel.

[Karl_T]

>Will use UHU Servo controllers. The Z motor has a 1024 steps encoder, want to get >at least the same for the big motors (530W) but they have 10mm axles which makes >it hard to find matching encoders.

US Digital should have encoder kits that will work for you. Very large selection.

Karl

[ger21]

Jmo, but that gantry will rack a LOT if you only drive one side. No way will it be stiff enough for the undriven side to stay in position.

[keyne]

Thanks Karl, indeed US Digital has 10mm shaft sizes! somehow i overlooked this before...

Gerry, i value your opinion. The gantry beam is surely stiff enough, it will not bend much at my cutting loads. The beam is connected to the side plates by eight screws and two large brackets.
What exactly is the weak point in the design? Side plates too long & thin? vee bearings?

[ger21]

With the gantry off the machine, clamp one side to something solid. How much force does it take to move the other end? I'd bet not much, because it's a long lever.

[keyne]

How do you like this version?
(low table/high rails, like Shopbot)
I like it, although it comes with the downside of worse table access and a little less Y travel. The gantry is also wider than in the previous version, giving better support on the rails.

[bgolash]

Hi
When I started my project someone mentioned to space the rolling elements on one side far apart. The other side just was closely spaced or were not added. There are some commerical woodworking machines designed as I mentioned with one side driving the Y and Z. The drawback to this design is the overall length of the machine must be increased to account of any increase in the rolling element spacing. I believe with this design there are two parallel linear motion assemblies on the wide side to counteract all the cantilever loads from the overarm design. With this design only one lead screw or timing belt drive system is needed . There maybe a reason I`ve never seen a DIY project designed as I have described but it would be great to see it attempted.
To play it safe you could use a three point linear motion design but drive it from one side. That would save a second drive that is normally used on wide machines to minimize yaw.
Barry

[keyne]

Barry, is the machine you are showing a gantry or cantilevered design?
Lots of commercial machines are cantilevered to make table access easier.. they usually have one rail at the table corner and one near the ground. Like the machine in this thread (Epple wooden CNC) http://www.cnczone.com/forums/showthread.php?t=19124
Anyway, the rolling elements in my design are spaced about 300mm (12in) apart. With the design as it stands now (elevated x rails) i fear no racking. See pic for status quo.
The machine is all aluminium as planned now... a wooden table for support is still an option.

[ger21]

that's a cantilevered machine. We have a similar machine at work. The two rails are on the back one above the other, like on a typical gantry, with the screw between them.

I still don't think you can totally eliminate racking, but that's just a guess on my part.

I meant to mention before, why direct drive. With gearing around 10:1, you can use much smaller servos and get the same performance.

[keyne]

I already have the servos.. gearing is easier said than done, with belt reducers you're looking at a dual stage 60:20 transmission. I am not a home machinist, making a good reducer, even single stage, is not that easy for me. Therefore i opted for the simple way.
The next step in my build is making the servo controllers and testing the servos. that will allow me to judge whether direct drive power is sufficient or not.
Average cutting force of at least 100N (peak torque should be about 5x that) should be enough in my opinion. I may be lucky to get half of the force of a shopbot, but OTOH i have servos and i don't want to push a 5HP spindle. BTW, i think the rack and pinion systems can deal with 1Nm torque max (30T, Modul 1.25 or 20 pitch), which is about a third of my belt drive. This also means that they operate outside of the safe area at peak cutting force....

[bgolash]

Hi Keyne
"Barry, is the machine you are showing a gantry or cantilevered design? "
cantilevered design

Regards Barry

[keyne]

Managed to install ubuntu linux and EMC2 today. Didn't work quite as described on the linuxcnc page though, i had to manually add a package repository and install with the package manager, the script didn't work. Now it works, looks quite good. I did a dual boot with FreeDOS to run the UHU software on the same PC. The machine is a celeron 433, we'll see how it works.. i am not too concerned about max. step rate. Ubuntu seems a bit bloated, with gnome and office...
Also made a shopping list for my UHU servo controllers.

my next tasks are building the UHUs, connection to the differential encoders, parport connection, estop circuit... (all the electronics) and the servo power supply which will be three phase. Also need to test the little servo (200W, for Z-Axis, w/ brake).

Have attached a couple of photos of already aquired material, servo motors and uhu boards, big servo close up and a W4 Vee-roller w/ caliper to show the size (60mm diameter).