Ball Screws :D, I wouldn’t settle for anything less. I haven’t tuned the machine yet but the tolerances are higher than my chinese digital caliper can read (perfectly acceptable for FRC use)
You sir, are a master
Our team will do anything, to have access to something like this. There have been several situation this past season where a cnc machine would have been a god send.
I have a cheap Chinese CNC router from Ebay at home. The electronics were done poorly and we had to re-work the stepper controllers, but mechanically, it is a fairly good machine. We use these router bits for aluminum and have had great results. They cut much cleaner and faster than a standard end mill, but still are not fast.
Very nice! Will look forward to the build log. I do not know of any sources for good multi-turn ball screws and ball anti-backlash nuts in a price range like that. Usually they resort to threaded rod or nylon blocks heated then pressed and sooner or later that bites you from wear.
Kind of suspect your CNC shield is a RAMPS 1.4?
Thanks for the advice, will keep it in mind if and when we hopefully get a cnc router 
What are the differences between an x-carve and a shapeko3? The shapeko 3 website shows videos of it milling into aluminum pretty easily…
A couple of things to note video is terrible at showing:
Accuracy, repeatability and grade of aluminium.
I have no doubt you can force aluminum into a drill press with an X/Y table as well and it will cut. Eventually the drill press will fail from that but I know folks do it. Some even toss the drill presses and replace them because the presses are -that- cheap they are literally disposable.
We actually used the X-Carve to machine our practice robot this year and it did a pretty good job. We put a LOT of work into learning how to properly use CAM software and feeds/speeds but eventually we found the sweet spot for both 5052 and 6061 with our setup. We machined our bellypan, chassis rails, superstructure gussets, and shooter components from components .050" to .1875" thick. You definitely will have better luck with with an actual CNC mill, but for a router, the X-Carve (with mods) can take you just about as far as you are willing to put into it.
What is the thickest block of aluminum do you think it can work with? (just wondering about the height of the z axis)
You can get about 2.25" of max height if you:
- Use the stock wasteboard
- Use the stock z-height plates
If you make your own brackets or sink a hole into the wasteboard, you can get a theoretical maximum of about 5-6". There’s guides on youtube; it is opensource after all.
Cool. What bits did you guys use?
These work wonders Carbide 3 flute with a special coating, $15-12 for the 1/8 and around $20-25 for the 1/4.
The seller, drillman, has great deals on some high quality bits; I recommend you read some of the FAQs about the Destiny line as it has a unique coating that more or less saves you from getting the bit stuck in 5052 and gooey materials.
GOFIRST is currently looking at getting a Routakit CNC router. Even with some upgrades, it comes out around $2000 if you get a standard version. Specifically we’re looking at getting the SD with the SDX upgrade kit and improved spindle. It looks like a pretty solid machine, and they have a pretty cool forum post documenting the process of designing their first one (I think it’s somewhere in the Shapeoko forums but can’t find it right now). They’re also super nice and responsive-- I’ve emailed them a few times with questions and am really hoping it makes sense for us to get one.
why is routakit better than xcarver or shapeko3?
I’d be interested to hear what settings, speeds/feeds you used. I really couldn’t get the one I had access to for a short while to do that much with with aluminum. I used Autodesk 360 to do my CAM and my guess is that I had some depth/plunge and speed/feed rates wrong.
Take a look at the construction and specs:
X-Carve:
DeWalt DW611 Router (or choice)
140 oz-in NEMA 23 steppers (or 60 oz-in NEMA 17 steppers)
Arduino with g-shield
Lead Screw or Threaded Rod z-axis
Max 31" x 31" x 2.55" work area
0.075-0.13 mm resolution
GT2 belts (pitch not specified, assumed 2mm)
~$1400 fully loaded
Shapeoko 3:
Router not included (DW611 recommended)
120 oz-in NEMA 23 steppers
Proprietary electronics
Belt-driven Z-axis
Max 16" x 16" x 3" work area
No apparent claims on resolution/accuracy
GT2 2mm pitch belts
$999
Routakit SD (no upgrades):
600W spindle
175 oz-in NEMA 23 steppers
PlanetCNC controller (which one not specified on page)
Ball Screw z-axis
Max 59.1" x 59.1" x 4" work area (30" x 30" x 4" default)
Repeatability 0.025 - 0.05 mm
Accuracy +/- 0.127 mm
9mm wide, 3mm pitch GT2 belting
$1799
Routakit SDX (what GOFIRST is looking at)
1500W spindle
30" x 15" x 4" work area
SDX upgrade for greater rigidity
$2179
We’re looking at it for somewhat similar uses to what I expect a team would want-- the most intense thing we expect to cut is maybe occasional 1/4" aluminum, but mostly 1/8" or polycarbonate. I haven’t heard good things about cutting aluminum on an X-carve, and the Shapeoko doesn’t have the work area we want. It might not be the right option for someone else, but it seems to be the closest to right for us, given our budget, without spending a lot of time designing our own (and, to be honest, we’re probably going to mod the crap out of anything we get anyways, because we’re all tinkerers).
Course you might not be able actually get those mechanical results without electronics that insure at full step enough power is supplied. Once you microstep all bets are off.
If 750W is about 1HP.
A 1500W spindle is 2HP.
Now the important questions about the spindle are:
- At what RPM is maximum mechanical power output? Does the motor driver effect that?
- Is that RPM a good match for the feed rates these NEMA23 steppers can deliver and honestly the answer in all 3 cases is likely no.
So the quality of cut and the tool life will diminish. Though obviously the results can always be good enough and someone could always get more tools.
I purchased a bare bones X-Carve 500mm last December for the team. Several of the parts and electronics were purchased separately. Total cost to put the machine in service about 690.00$. A stock X-carve will cut 6061-T6. However, it is not acceptable from a precision and speed perspective. The main problem is x and z axis torsional stiffness. I modded the x axis maker slide with a 40 x 4 500 mm Al plate to tie the 2 maker slides together. The torsional problem did not go away until I replaced the Delrin wheels with metal wheels. The machine can now cut with reasonable speed and precision, but not like a 12000$ commercial unit. It has been a great addition to the team and made many parts on our 2 robots. I should mention that about 1/2 the parts cut were 7075. With all the mods 7075 is not a problem. Stock it is. Looking back I could have built a better machine for the same money. However designing a machine from scratch takes allot of knowledge that I did not have last December. Be careful of the eBay Chinese stuff. Some of it is good and some terrible. Yes, a sub 1000$ router can be a great addition to a team. One thing I should note is that 1 requirement for our router was that it had to be portable. We do not have room for a permanent set up. The 500mm x-carve fits this requirement.
Has anyone worked with a router from Probotix? They seem to be significantly more rigid than most. They additionally use screw drives instead of belts.