Upgrading to Shapeoko Pro XXL or go for a completely different machine?

I’m using Solidworks for CAD and Fusion 360 for CAM.

So the only reason for using a spindle instead of a router would be sound? It doesn’t run any faster and doesn’t cut any faster?

Thanks. I’ll take a look as soon as I’m allowed in the group.

1 Like

Got ya approved. :slightly_smiling_face:

Lots of good info here. I’m just too lazy to upload to multiple types of social media. :smiley:

The reasons I see to use a spindle are these:

  1. You can get them in much higher power ratings than routers. This doesn’t really matter for SO3 users because the power rating of the router is not the limiting factor of the machine.
  2. They tend to have less runout than a router so you can achieve better tolerances. Doesn’t matter for wood.
  3. They can be quieter if you go with a water cooled style.
  4. They usually have an ER style collet system which is more flexible than the non standard collects routers use. You can still get collects for routers, there are just less options.
  5. They can be controlled via software if you set them up that way. Routers can be as well with some modifications like the SuperPID.
  6. The RPM controll is generally better so you can achieve better finishes because your RPM doesn’t vary as much when you engage material.

Also, If you would like I can take a look at your Fusion CAM and see if I can find any quick changes that may help you.

There was a lengthy thread comparing the options at:

Daaamn… you’re cutting through that walnut as if it was butter. I’m guessing walnut is the most similar to European oak when it comes to hardness and such? Poplar and birch are softer right?

You’re even using climb milling from what I can see? In my experience, the router has a much easier time with conventional milling, but it creates a lot more splinters.

Are you using the Carbide router in all 3 videos? You’re using the SO Pro on all 3 runs, right? Do you feel a big difference between the SO3 and the Pro?

@SLCJedi @nwallace If you want to take a stab at improving the cut time, you can find the file for the top here:

This is how things look now:

As you can see it’s cutting the big pocket that’s the culprit now.

I love your design. Well done indeed.

With respect to your questions, I am probably unusual in that I’ve never run a job on my Shapeoko with a trim router attached. I bought it second hand, and when I turned on the router and heard the scream it made, I decided it was not for me and ordered a water-cooled spindle immediately. Spindles are much quieter - a dentist-drill compared to a chain-saw.

It might be worth considering that a spindle probably makes tool changes quicker.

Since the spindle is turned on and off via GCode, it has already quiesced and is still by the time it comes to the front ready for the new tool. So you don’t have to turn it off, wait for it to stop, and then change the tool. A BitRunner would do the same, of course.

The other efficiency is that ER-20 collets and nuts are readily available, and they hold a tool even when fully unscrewed. So if you have frequently swapped tools, you can leave them in their collet and just fully unscrew the complete assembly and then re-screw in the new one. This also gives you the chance to blow out the dust in the idle collets while the machine is cutting.

I don’t find a tool change takes more than a minute (okay, that’s an estimate… I haven’t timed it). It’s also a useful break to do a quick check of the stock and cuts, with the dust collector not covering it.

1 Like

Would another option be to avoid such a large area cut, and optimise your design another step? Cut a ‘fence’ piece and match it into a shallow groove around the lid - two operations, but nothing like as much cutting time…?

1 Like

The design is not going to be changed. That’s a no-go.

1 Like

I forgot to mention that a ER-20 collet will take larger bits, such as a 12mm endmill. That might change the time taken for your main pocket clearance.

Is a spindle strong enough to handle a 12mm endmill? What if I buy the Mafell spindle? That seems like a smaller spindle?

To be honest, I’m not sure. I’ve tried 8mm and that works just fine.

My spindle is 2.2Kw since it was best bang-for-£ and worth trying for €285. It certainly looks big enough to handle it (it’s 80mm in diameter and about 300-350mm tall), but I have not tried a 12mm endmill.

The Mafell spindle is indeed smaller from the pictures I’ve seen. @WillAdams might be able to comment on its power ratings. From watching his progress attaching it, it seems there is a trade-off for the rapid tool change in that each tool change requires the attachment of a more complicated variety of dust collection offset rings, pre-calculated per tool. As I mentioned before, I personally don’t find tool change is a lengthy or arduous operation, but I can see how it might be troublesome under some circumstances.

I think your correct on the hardness. But the walnut seems harder than the red and white oak I’ve cut in the past. Poplar is about the softest hard wood, and the birch is softer still.

I went with the default direction in VCarve. :smiley:

All 3 videos are the C3D router that shipped with my SOPro. The difference between my modified SO3 and the SOPro is fairly dramatic in my opinion. The router is the limitation. The SOPro is well matched to a 65mm, 1.5kw spindle I think. Less weight than an 80mm one, and should be equally capable as the machine.

According to the Janka scale…walnuts are about 1010 in hardness and Oaks are 1220-1335 … so, technically, Oaks are “harder”. I think it’s more about the impact the open grain of an Oak will have on the cutting quality (fuzzier)…Walnut is a great wood to work, as it almost always cuts cleanly with just about any tool.


Not worried about the power on the Mafell unit (and hoping for the sake of the bearings to not bump up against any limits — I’ll be using it w/ the same sort of feeds and speeds I’d use for the Carbide Compact Router (so I don’t have to track which machine I’m running a file on) — I just wanted the more convenient tool change).

That said, I’m pretty sure that there’s a lot of headroom and a person who had only the one machine w/ this spindle could greatly increase their material removal rates, esp. on a Pro.

The dust collection adjustment hasn’t been that bad — I select a dust shoe when I begin a job and usually it’s possible to stick w/ it for the duration.

1 Like

I still run a Makita, and the loudest parts of the system are the cut and the disturbed air from dust collection, I’d rate the air disturbance as the least desirable sound, and a definite motivator to wear hearing protection. My 2 C3D, and sizeable collection of Makita routers, all have minor runout. I wouldn’t think this even recognizable on a wood product. Simplicity and cost, on a simple lightweight machine, my opinion is the router wins.

As far as machines go, its a matter of cash flow. If you spend more, you will get a better machine. The Shapoko(3,4,pro) is forgiving and cheap, if you are at least some what aware of your processes, you can crash it with very little repercussion. It’s a low investment tool, great for learning and low production. Adding a functional ATC to a machine will cost more than the base price of a Shapeoko Pro, and that’s going about in in a semi DIY fashion(Assuming you invest in new and reputable components). You could buy 2 Pro’s for a little more than the cost of just the (semi diy)ATC system.

The non standard size of the Shapeoko(and most inexpensive cnc routers), in my use case, is it’s true labor cost and productivity downfall. I cannot load a full sheet, half sheet, or quarter sheet. All sheet stock has to be processed. Add in increased machine time and reduced finish quality(additional post processing).

Your options are many, with budget and priorities to consider. You could build a large light duty machine for less. Could buy multiple Shapeoko’s. Can buy a “desktop” machine from a reputable domestic manufacture, $7-10K US with basic options and no ATC. Could step up to a domestic entry level rack and pinion industrial machine, 15-25K US with basic options, no ATC. Chinese machines are always an option, and widely used in Woodshops here. Cost less, but more time on your part researching, communicating, waiting, and dealing with potential issues. Or you can just roll with what ya got. I don’t know what a jewelry box costs, how many you sell, or the value of your time.

Me - My minimum product wholesale cost to customer is roughly $600. This is something that would barely fit on a Shapeoko3 XXL. 4 tool changes. Not counting manual tool change time, approximately one hour of machine time. Add stock processing, machine loading, unloading, and post processing. It’s not practical to machine singularly. I contract out the work, and take my cut. The rest doesn’t fit anyway. Cost analysis suggests I am not capable of servicing the volume required to justify the addional floor space, power requirement, and industrial machine and tooling cost - All for minimal capital gain. Sure, it would be cool, and maybe someday soon, just because.


This is pretty interesting. Personally I’d not look to do anything drastic. A full ATC setup costs a couple of thousand $, you need air, tooling/collets, a different controller, expensive spindle, a VFD and a host of other kit. By the time you’ve dropped that on a machine it will take quite allot of sales to repay the initial outlay.

I’d say if the issue is heat, jump onto a water cooler or even air cooled spindle.

Having a read through you also mention about machining at high rpms, I’d usually opt for around 18-20k on wood. 30k is usually a bit high which could be adding to the heat issue?


I am running my SO3 making coasters and other products for my company, Kalos. The machine often runs 6-8 hours with rests for tool changes and stock flips. I use 4-5 tools on average and find that the BitSetter and BitRunner were the only upgrades needed to basically eliminate tool changes as a bottleneck.

My router got hot the first few times I used the machine. Then bought Gwizard, watched some videos, and cranked up the DOC and feed…still run most cuts at 18,000 RPM in walnut, maple, and cherry. The router and cutter are barely warm after a long cut. I have been running the machine like this for months now. I am running as fast as I can in a production environment. My products are a lot smaller than yours, so I can run multiples much easier… that’s a big difference. Still, I think you may be a bit too conservative on your feeds and speeds. I have cut a pocket nearly the size of the one in your design (for a jig) and I didn’t have any heat issues. I am using Amana tools almost exclusively now, but I still use my C3D stuff from time to time.

The only things I have broken are belts and tools from crashing the machine or drops on the floor. Before you upgrade I would look into Gwizard and see what you get for feeds and speeds. The program tends to be aggressive, so I have had to dial it back a bit as I have learned the limits, bit my speeds and feeds have been vetted thoroughly now and I am quite happy.

I still wish the machine was faster, but I can’t justify the expense of an upgrade with my business not making a profit.

Hope this helps. I love the process of building manufacturing processes that are scalable, reliable, and repeatable. I am REALLY happy you have a product that is selling! Congrats!


That software looks like it’s really worthwhile looking into. I’ll take a look at it :slightly_smiling_face:.