Shapeoko Machine capabilities and limitations

Talk about all things CNC
cvoinescu
Posts: 4442
Joined: Thu Jul 19, 2012 6:50 pm
Location: Camberley, UK
Contact:

Re: Shapeoko Machine capabilities and limitations

Post by cvoinescu » Tue Dec 15, 2015 2:50 pm

sgth0mas wrote:Having such high tolerances is a game killer for me which is why i asked the question. Seeing an elastomeric element used to control position in a force loaded application worried me, but i wanted to confirm my worries before writing it off. Hopefully there will be a rack and pinion or ballscrew option with the SO 4.
It's the fiberglass reinforcement of the belt, not by the rubber, that transmits the force. Belts are surprisingly good, and not the limiting factor (at least not for the stock Shapeoko 3 size -- they fare less well in larger machines). Without also upgrading the V-wheels, it does not make sense to use ballscrews -- certainly not on X and Y.
Proud owner of ShapeOko #709, eShapeOko #0, and of store.amberspyglass.co.uk

Fablicator
Posts: 113
Joined: Thu Jul 02, 2015 7:59 pm

Re: Shapeoko Machine capabilities and limitations

Post by Fablicator » Tue Dec 15, 2015 4:07 pm

I have a very different experience than some others here.

I went into this with an extremely strong 3D printing background, a decade of CAD skills, and a love of tinkering. I have no fear when it comes to abusing equipment to make it do my bidding.

The stock SO3 is lackluster for everyday use, and in my opinion, needs a little over 1.1K in added investment to make it into a decent everyday tool.

You will need:

$150 Dewalt 611
$100 Precide bits collet kit (1/8 and 1/4)
$120 Homing, probe and enclosure kit
$450 Aluminum Bed
$50 USB Isolator (except for sparkfun version which may not need it)
$100 Dust Shoe
$100 Clamp kit
$50 9mm Pulleys and Belts


And a good bit of time to learn the new system's idiosyncrasies.

I'm reliably getting better than +- .005 on mine right now in aluminum, using 2 spring passes for tight tolerances. I'm sure I could get tighter tolerances than that, but haven't had the need so far.

I trust this setup to run pretty reliably, and have been running it about an hr every day. I would get some extra V wheels if downtime can't be tolerated.

There are some strong annoyances I have experienced with the system so far:

*It is very very bad at plunging in Z into hard materials.
*Excessive vibrations when cutting in the Y direction due to V wheel/XZ plate flex.
*Occasional disconnects when turning off the router, or using heavy duty power tools on the same circuit. (even when using the usb isolator and following all other suggestions to fix the issue)
*Can be extremely hard to shim and get perfectly square/level alignment.
*The carbide motion software only works for the most basic of tasks (switching to Universal Gcode sender solves this issue)


Overall I love my machine, and have more than recouped my initial expenditure through making parts. I learned a lot about the process, and what to look for in the future in an upgrade machine.
For the price class, it is unmatched if you want to learn and have a capable machine.


If you want timely support, turnkey accuracy, and a much easier time getting into CNC, go for the shop bot.

If you really enjoy tinkering and have some free time, get the SO3.

Both machines seem to have excellent resale value, so even if you decide the Shapeoko isn't what you really needed, you can always get most of your investment back.

twforeman
Posts: 1347
Joined: Tue Jan 29, 2013 4:51 pm
Location: Minneapolis, MN
Contact:

Re: Shapeoko Machine capabilities and limitations

Post by twforeman » Tue Dec 15, 2015 5:04 pm

Fablicator wrote:I have a very different experience than some others here.

If you want timely support, turnkey accuracy, and a much easier time getting into CNC, go for the shop bot.

If you really enjoy tinkering and have some free time, get the SO3.
I agree 100% with this assessment. If you want turn-key, you need to spend a bunch more money.

If you are willing to put the time into upgrading and learning how the S3 works, then get an S3.
Ender 3 3D Printer
ShapeOko v3 serial #0004 - upgrade thread
All of my ShapeOko related blog posts

sgth0mas
Posts: 101
Joined: Sat Dec 12, 2015 5:57 pm

Re: Shapeoko Machine capabilities and limitations

Post by sgth0mas » Wed Dec 16, 2015 2:23 am

Thank you everyone for the honest replies, it’s been enough for me to make a truly informed decision. Most of y’all were right, if I’m after a plug and play cnc with the ability to cut somewhat tight tolerances at a decent rate without a lot of attention…then I need to pony up the cash. Since I have been using a CNC for a while now, machining isn’t as much fun. My enjoyment is in the design, analysis and testing…not watching a cutter fly.

I finally looked at this in a more detailed approach. My concern with the belts is the sensitivity to cutter loading. Just to do some quick back of the envelope calcs, let’s look at the elongation up to the max torque range of the motor (these are quick calcs so checking this math is much welcomed). The motor is a 120 oz-in stepper, which comes in at 7.5 in-lbs. The pulley is 0.6” I believe, so if we divide the 7.5in-lbs by the moment arm of 0.3”, we get a maximum tangential load of 25lbs at max torque
.
Now, let’s look at the belt properties. The Gates data sheet lists a 1” GT2 belt to have a modulus of 18,000, and specifies proportional reduction for smaller sizes with a further 10% between ¼” and ½”. That means the 18,000 modulus becomes 4,050.

So taking our max load of 25 lbs, and calculating belt elongation at the far end of the machine, we get max belt elongation to be .0988”. This is the elongation induced at max torque and max travel in one direction. Now if the belt was always loaded to this value it wouldn’t matter, but that is almost never the case. So when the .0988” starts to relax, we will blow our tolerances.

To combat this, one could increase the width of the belt to 9mm. This cuts the max deflection at max torque and max travel to .063”. If we reduce feed rate to use only a quarter of the torque, we are now at .016” elongation. Another improvement is to run at the center of the table to cut the effective belt length in half, so we’re closer to .008” of belt elongation.

Now, without looking at the deflection of the cutter, the router bearings and the rest of the structure, one can see that to hit the +/-.010 tolerances I’m after, I would be limited to a pretty reduced feed rate and I would need to limit my work area. My toolpaths would have the be very carefully planned to ensure consistent cutter loading. In saying this, I would also Imagine the big xcarve would perform very poorly because of the belt length (belt elongation is proportional to belt length so all displacements are almost doubled).

By switching from belts to rack and pinion, we eliminate this deflection and only deal with backlash. Remember…belt elongation is in addition to all other deflections encountered during the machining operation. That's why I indicated it would be such a big improvement.

Looking at these numbers, I believe +/- .010” is possible, but my patience with slow feed rates is not. This has been a good study, and I’ve definitely learned more about what I’m truly after.

sgth0mas
Posts: 101
Joined: Sat Dec 12, 2015 5:57 pm

Re: Shapeoko Machine capabilities and limitations

Post by sgth0mas » Wed Dec 16, 2015 2:31 am

WillAdams wrote:
sgth0mas wrote:at least +/- .010 tolerances...preferably half that.
http://www.shapeoko.com/wiki/index.php/FAQ
What is a reasonable expectation for accuracy? --- If tuned properly you should be able to get to a resolution of 0.003″ –- 0.005″ (~0.075mm -- 0.127mm) --- 0.001″ (0.025mm) is achievable with careful setup and an upgraded machine.
http://www.shapeoko.com/wiki/index.php/Parts#Motion
Delrin V-Wheels
AIUI, linear bearings would more than double the cost of the machine. I'd be interested to see a parts list / build which would clarify that. AFAICT there aren't any readily available / easily findable machine designs which use linear slides.

I don't see how including some software (for the CNC Piranha) offsets the real world expense of physical hardware. A touch screen is a pretty inexpensive part, esp. in quantity.

I've searched for a comparison of linear motion systems, but haven't found any --- do you have some good references?

Tried to assemble a list of all the DIY / hobby level CNC machines on the reddit hobbycnc wiki: http://www.reddit.com/r/hobbycnc/wiki/index but it's hard to do comparisons w/o a sortable table.

The issue w/ the aluminum extrusions is the manufacturing tolerances --- usually one gets a straight piece, but there's a possibility of not, and there's also the issue of how squarely the ends are cut.
Resolution is just the minimum movement that the machine can move based on the constraints of the electronics, so it's driven by the number of steps and the pulley geometry. It doesn't take into account belt elongation, structural flex, cutter deflection, router bearing movement or anything else. So a resolution of .001" doesn't mean I will get anywhere close to my tolerances.

When I speak of the piranha, I'm talking about the copy of vCarve Pro that comes with it. That software is over $600. Now I don't like the plastic supports they use, but the lead screws and linear bearings are going to be more capable. In all honesty, I would bet that 1/4" aluminum plate would be cheaper than the 1" HDPE they use....I have no clue why they have gone that route. Maybe it's to save on manufacturing cost and not material cost?

But based on the cost of the Piranha, the linear bearings would not double the cost of the SO3. And simply replacing the belts with rack and pinion while using the existing guide system would be even cheaper. It won't beat a laguana or axiom, but it'll be a huge improvement over the current state.

Gadgetman!
Posts: 1061
Joined: Sun Aug 25, 2013 12:44 pm
Location: Norway
Contact:

Re: Shapeoko Machine capabilities and limitations

Post by Gadgetman! » Wed Dec 16, 2015 8:50 am

Max load of 25lbs sound a bit much, really. (I don't like pushing stuff to the max.)
What about using 12mm wide belts?
and limiting yourself to say, 15lbs?

Rack and Pinion has some issues with wear, of course, but I assume that you can handle that without too much trouble.

As for the 'added investments', not all need those.
The Homing probe and enclosure? Sure, homing probe and all that can be nice, particularly when doing jobs that require toolchanges, or when cutting multiple parts out of blanks mounted in a fixture. It's not needed for simple work, though. Enclosure kits are for those who can't make one themselves. We're fabricators here...
Dust shoes? sounds like a job for a SO...
Clamps? We don't need no steenkin clamps! ;-) tie downs can be made out of scraps of plywood and bolts.
Aluminium bed? That's some right fancy stuff you got there. ;-) I'd love to have one myself, but then again, my SO won't fit on it.
It's probably not all that much more stable than my torsion box, though.
(or heavier... The bl**dy box weighs in at 40Kg, at least... )

You don't NEED that Dewalt 611 either. No, really. It's just the recommended spindle for this machine at this point of time.
If you have another spindle you want to use, such as a DC spindle, and you can mount it onto the machine, go ahead.
(DC spindles tends to be quieter. You may not even need that enclosure then.)

The SO series of machines are wonderfully customizable, and can be adapted to suit lots of different tasks.

Whatever you decide, keep us posted.
Weird guy...
Shapeoko 2014F: 1000mm X/Y, 300W Quiet Cut spindle, Arduino Uno/G-shield and GRBL 0.9i
15x30 drag chains, custom spindle mount, 9mm belts, 8mm endplates, 6mm motor plates.

WillAdams
Posts: 8488
Joined: Mon Apr 09, 2012 6:11 pm
Location: Pennsylvania --- south of the Turnpike, East of US-15
Contact:

Re: Shapeoko Machine capabilities and limitations

Post by WillAdams » Wed Dec 16, 2015 11:59 am

Tim Foreman was able to get repeatable 0.001″ cutting movement: http://timf.anansi-web.com/wp/shapeoko-accuracy-really/ With an upgraded ShapeOko 1.

There seems to be a disconnect between Gates specifications (overly conservative?) and what is experienced in actual usage (most people tension the belts more than is recommended) — using a suitable roughing clearance and counting on the finishing path for precise and accurate cutting seems to be a reasonable tradeoff. Before I got my touch plate set up and my Z-axis dialed in, I’d often lower the Z and repeat a cut, w/ no mis registration. I am using 9mm wide belts (and suggested that in my first message in this thread --- I really think they're the biggest bang for the buck upgrade for the SO3).

Pressing a CD/DVD or providing a download doesn’t add much to the costs of a product, and I doubt it changes the total of their B.O.M. much. This started as an opensource project, which is why I’m involved, and claims about software valuation don’t hold any weight w/ me. If you can provide a tear down w/ parts costs of the machine, I’ll yield the point.

I’d be much more swayed by a documented B.O.M. w/ purchase links which proves that a machine such as you describe can come in at a price which will allow it to be sold for $1,000 — note that most business accounting models demand that the raw cost of parts be half or at most two-thirds of the list price.

Or, a plan for such an upgrade to the SO3 which includes a B.O.M. and purchase links.

I’ve been looking for a good overview and discussion of the tradeoffs and expenses of the various linear motion systems w/o much luck — links and resources on this would be welcome.

FWIW, I’ve always been a budget-oriented guy and my costs have been significantly lower thus far:

- Makita RT0701 and Elaire collet
- aluminum shim for the mount (need to design and cut a new one)
- HDPE Wasteboard w/ T-nuts
- clamps were made from the leftover HDPE
- HDPE dust shoe w/ a couple of rare earth magnets, a hose clamp and misc. hardware
- Tim Foreman’s nifty limit switch kit — made a touch plate out of scraps and odds and ends laying around my workshop.

need to 3D print a cyclonic dust separator and get a better vacuum for dust collection and make (or buy) a better spindle mount.
Shapeoko 3XL #0006 w/Makita RT0701 Router w/0.125″ and ¼″ Elaire precision collets
Nomad 883 Pro #596 (bamboo)

northbear
Posts: 252
Joined: Mon Apr 09, 2012 9:21 pm

Re: Shapeoko Machine capabilities and limitations

Post by northbear » Wed Dec 16, 2015 3:28 pm

sgth0mas wrote: Now, let’s look at the belt properties. The Gates data sheet lists a 1” GT2 belt to have a modulus of 18,000, and specifies proportional reduction for smaller sizes with a further 10% between ¼” and ½”. That means the 18,000 modulus becomes 4,050.

To combat this, one could increase the width of the belt to 9mm.
Another way to improve the belt is going with a 3mm pitch belt rather than using a 2mm pitch belt. 3mm pitch belt has a modules of 30,000 (for 1" wide belt). It also has a greater ultimate strength (2200 lbs vs 1200 lbs) and working tension (114 lbs vs 25 lbs) again for for 1" wide belt. As sgth0mas mentions, these values are proportionally smaller and derated for smaller width belts (see second page of second link below).

I have been happy with my 3mm pitch belts, although I have not pushed to see what the best tolerances I can get out of them would be. On the other hand, I do believe I still hold the current speed record http://www.shapeoko.com/forum/viewtopic ... =30#p21023. Of course this is probably because all of the new shapeoko 3 owners didn't know there was a shapeoko racing league :lol:

Sources of information (mostly for my own reference so I don't have to search to find them again later):
http://www.sdp-si.com/D265/PDF/D265T015.pdf
http://www.bbman.com/assets/files/pdf-l ... erties.pdf
My buildlog is here

WillAdams
Posts: 8488
Joined: Mon Apr 09, 2012 6:11 pm
Location: Pennsylvania --- south of the Turnpike, East of US-15
Contact:

Re: Shapeoko Machine capabilities and limitations

Post by WillAdams » Wed Dec 16, 2015 4:51 pm

Added both of those to: http://www.shapeoko.com/wiki/index.php/ ... References (along w/ footnote links back to your post).

There's also a link there to a previous discussion on this sort of thing which presaged the switch to GT2 from MXL. http://www.shapeoko.com/forum/viewtopic.php?f=11&t=543

And Northbear's post is esp. interesting: http://www.shapeoko.com/forum/viewtopic ... =543#p4512 where a diagram is posted showing an up to 0.001 positioning error --- cancels out sometimes or something?

One further addition on the upgrades to improve accuracy / precision --- 400 step motors. I have a set on my SO1 and am definitely going to have to consider a set when I upgrade my SO3 to a larger physical size.
Shapeoko 3XL #0006 w/Makita RT0701 Router w/0.125″ and ¼″ Elaire precision collets
Nomad 883 Pro #596 (bamboo)

krtwood
Posts: 26
Joined: Tue Jun 16, 2015 4:01 am

Re: Shapeoko Machine capabilities and limitations

Post by krtwood » Wed Dec 16, 2015 7:54 pm

I wouldn't necessarily expect any better from a Shark. I have had the lowest end Shark for a number of years and while it has linear bearings and screws, the spindle is really far out from those bearings and everything in between the two is plastic. There's quite a bit of flex in there. I made a dust shoe for it but had to take it off because whenever the brushes hit the edge of the work piece it would tilt the spindle. I don't know how much is different on the "HD" but it just looks like black plastic instead of white plastic in the spindle mount. I've got a SO3 sitting in the box waiting for me to have time to put it together so I'll be able to directly compare the two at the end of the month. One of the things I'm thinking about doing with it once I have it running reliably is to make new parts out of aluminum for the Shark to see if I can improve it. As long as I have one working I can tinker with the other one.

Post Reply