Welcome to my new Shapeoko :)))

Hello one and all. I am delighted to report that my Shapeoko SO3 has finally arrived. It was delivered at around 12:50 this afternoon. The weather was lashing with rain and the package left by a gentleman who was much smaller than the box. He said his goodbyes and shouted “it’s heavy mate”. I brought the box inside my home and set about uncovering the contents and putting them into the place they would eventually reside.

First impressions are what a really nice piece of engineering the Shapeoko SO3 is. I set to work by laying out and arranging parts in the order they would be required, while taking some time out to admire the quality of the workmanship. Eating some home made banana cake and drinking a hot 500ml mug of tea, I set to work. Pictures usually help to tell a story so I have some pix for you to look at. First, that package. I am sure that many people will remember having to haul one of these packages into their home.

I transferred everything from the box to the (workshop) shed and laid out the individual boxes and printed the Shapeoko assembly guide.

The first job was to unpack the frame and the baseboard and put them together. It was a relatively easy task to line up the screws with the holes in the baseboard and then loosely screw the items together for later adjustment. I figured the task would be dropping the screws through the pre-drilled holes and lining up the screw holes with the holes in the baseboard. I managed that well enough but some of the screws were difficult to turn.

I took the metal frames apart from the baseboard and on inspection, I noticed that some holes had paint on the thread of the screw holes. Out of 12 screw threads, 3 were as expected, 9 offered some resistance to the screw and out of that 9 threads which resisted the screw turning, 4 were able to stop the screw turning after it had been started.

I was a little surprised but I could understand how the issue arose. I selected an intermediate tap and cleaned each of the threads. Once the paint was removed, there was no difficulty in inserting the screws and turning them. Pix follow… in the first image, the highlighted area depicts a thread which is painted.

The next image shows one of the holes with an intermediate tap in a tap wrench, just sitting in the hole after cleaning the paint from it.

This image shows the paint removed from a hole by the tap. The removed paint is resting on a cardboard box that Carbide 3D pack their goods in. You will see that some of the fragments are quite large and the fragment at the bottom left of the frame has adopted the shape of the thread.

Another issue which I discovered early on in the assembly process was the electrical plug. I cannot be certain but I believe it is an American pattern of plug for wall sockets. I had to find a UK standard plug to fit the power supply. The fitting on the power brick is commonly issued here too. We refer to that as a three pin ‘kettle’ lead fitting.

The UK plug is fused and earthed and the line and neutral pins are also insulated in a way that prevents an unintentional contact with the pins. Finally, the wall socket has a sprung plastic shield keeping the socket closed until the longer earth pin is inserted.

The image is of the plug which was supplied in the Carbide 3D box. It cannot be used here in the UK because it is not fused and the pin configuration and shape will not fit UK wall socket outlets.

The next images show the standard UK plug from the front and top. These are clear images of theplug and its fuse compartment and standard approved UK pin configuration.

Pin configuration and fuse.

Insulated pins and long earth pin to open the wall socket.

Once the frame was assembled, it was time to move on to the axes carriages starting with the Z/X axis. Here there was a need to attach a tram plate, stepper motor, the router mount and open the V wheels. I felt that it was a simple task because the illustrated assembly guide was a great step by step set of instructions and I found it very helpful.

Trying to find the smallest hole in the tram plate was not quite as easy a task as it should have been. I was reasonably sure that I had the tram plate oriented correctly (semicircular cutout at the bottom) to match the same shape in the carriage. The plate was a different design to the illustration in the assembly guide and could usefully point to the cutout for orientation.

The small hole to be placed on the left top of the tram plate was a different matter. A choice of two should be an easy task (a 50% chance of being wrong). Both holes were roughly the same size. I used the unthreaded portion of the 55mm long screw to try and see how much play there was. I eventually settled on one and taped it so I would know how to position the tram plate later.

I would suggest that all of the holes except the top left fixed normal sized hole, are created similar to an oval fishplate on a railway line. The very small movements currently offered for tramming correction may not be sufficient to permit easy adjustment. (just a thought…) Failing that idea, it should be possible to nick the corner of the hole which was intended to be non adjustable. The nick would be covered by the head of the fixing screw. This would eliminate searching for the correct orientation.

Fitting the four tram plate retaining screws was problematic but I just retapped the holes and the screws went in easily. The image shows the swarf that came from the tram plate fixation holes, after tapping.It appeared to be metal to my eyes.

The Z/X axis carriage was not done with me yet. I was about to add the stepper motor and found that the 30mm stand-off hollow pins were all loose. A slot cut across the top of each stand-off cylinder would have allowed them to be tightened. As it was, I applied a screw to the thread and tightened both the stand-off and the screw. I was careful not to jam the screw into the stand-off and the items have stayed put. the stepper motor was added. It reminds me to mention that the illustration in the assembly guide might well have been used to confirm the orientation of the stepper motor.

Attaching the router mounting and opening the V wheels was completed without fault. Time to move on and open the V wheels on the two Y axes. Once again, this proved to be an easy and faultless operation. Both Y axes had two screw holes drilled in the outside aspect. The assembly guide had suggested that only Y1 required two screw holes. I understood that Y2 would be left undrilled.

Assembling the rails was the next task. There was a warning note about pre-screwing into the extrusions because of the possibility of swarf in the screw thread, making cross-threading easy. This was the most frustrating part of assembly thus far… The extrusions have a thread on the inside of a tunnel that is not a complete 360 degrees in shape. This probably gives the inserted screw an opportunity to move outside the boundaries of the metal which are described by the thread.

My curiosity has been raised. I would like to ask a question out of my technical curiosity (providing the answer is not commercially sensitive) about why the threaded material of the extrusions is not complete. It is likely that the female thread is not as efficient as a holding device for the male thread, when 10% or more of the female thread is missing. The extrusions are well made with a substantial 5mm wall thickness. The next image clearly shows swarf in the threaded section on the upper left of the image.

What was the technical rationale behind the decision to not offer a fully threaded section to join each of the extrusions to the other components? My supplementary question is the natural corollary to any assessment of the technical rationality for not using a complete threaded section in the extrusions. Would some cross threading be prevented by the addition of a significant chamfer and lead-in to the thread?

Final images are the assembly showing the free running gantry. All that remains is applying the belts, squaring again, tramming and tightening all loose screws. Additionally, there will be the installation of the electrics and the trim router. Once I am happy that everything is working as it should, I will add the JTech laser to the mix.

The bed was evelled in 9 different places in a cartesian X-Y pattern and then a further 6 places in an isometric pattern including the centre from each direction. Spirit level was reading level at all 15 points.

A work in progress…

Side view…

Well, that’s my day’s work for now and I will start again tomorrow. Please do not think I am complaining. I have really enjoyed assembling my Shapeoko SO3 and I am glad I made the choice to buy some great engineering and fantastic support.

One final question for the science buffs. I have placed the Shapeoko on a very dense 12mm thick rubber mat. I was hoping to dampen or remove any resonant frequencies and have a quieter and more stable machine. I wondered if using a seismographic phone app will detect much in the way of difference between measurements made with the feet included and measurements made with the whole frame resting on the rubber mat, which is incredibly dense and weighs almost 40lbs. Thanks for any assistance. Goodnight all.


You did all of that in one day?.. oh man am I slow.


Half a day! :grin: I thought that I was taking it particularly slowly because I was not wanting to get anything wrong. I really loved working on the machine though. I was in my man cave with some of the finest Mississippi Hill Blues playing while I worked. RL Burnside is one of the truly great bluesmen to come out of the USA. Heaven. :smiley:

1 Like

If you read the small print on robotshop.co.uk really really carefully they’ll tell you to buy their EU to UK plug adapter which is an appalling crime against wiring you would have thrown away immediately anyway.

The lead for the Shapeoko power brick is an EU 220V cable.

As for the rubber mat, I’d suggest lowering the adjustable feet until the metal frame of the machine just sinks into the mat a little.


My understanding is that it makes the profile easier to extrude — you’d need to check in w/ @edwardrford and maybe @Luke

1 Like

Yes, indeed, Liam. I don’t do crap electrical anything. Adapter use invariably means something is amiss with the installation anyway. Your picture adequately illustrates the notion that junk electronics are going to provide junk safety. I am not impressed by the CE marking, which it is up to the manufacturer to apply.

The UK site of RobotSHop is an online storefront only. There is no physical place to visit. I was surprised to discover that my Shapeoko originated in Heerlen NL. To my mind, trying to use a US plug top through any kind of adapter is just asking for trouble.

Thank you very much Liam for the suggestion about the frame and the rubber mat. I think I will try and measure the vibrations with a seismometer. I want an indication of the vibration saving made by this heavy and expensive mat .

Thank you for your insight Will. I hope that either @edwardrford or @Luke will be able to expand on your answer. I know a good deal less than nothing about extrusion manufacturing. I had considered the issue because I know that I can usually wield a manual screwdriver without cross-threading a screw.

I found that when I was starting to insert the screws into the extrusion ends, that I had begun to cross-thread about 50% of the screws. It is not a deal-breaker by any manner of means but it certainly caused me to question what I was doing wrong.

In another time and place, I would have been tempted to chamfer the hole to provide a better lead-in. Without being able to guarantee an absolutely perpendicular entry for a countersink (which would likely have drifted towards the split under any sort of light pressure) I chose not to do that.

I guess that my sense of engineering is such… I can understand that the design does work with the proviso that all four screws are located and locked home. There is probably not much chance that the screws will let go in use. My brain asks me how do you hold one thread against another when there is a significant and deliberate space within the normal 360 degree contact patch between the male and female threads.

Fully threaded sections (in the line of flow) in extrusions introduce two challenges: cost (extra material, and extra threading operation), and straightness spec (thermal mass and cooling across sections of differing thickness). It is prohibitively expensive to tool ‘extruded in’ full perimeter holes (to improve the straightness issue), and most producers (from my experience) refuse to do so - I think (straying from what I know here) it also substantially increases the machine tonnage required to operate the tool.


Extrusions must be smooth, you can’t extrude features like threads. Every extrusion is machined on one of our Haas VF4’s, adding features to the prominent sides, then each end is machined parallel. Once it’s out the ends are tapped whilst in a jig.The length of the thread is determined by the tapping tool, but is more than sufficient for this application.

Lots of people ask us why we haven’t made an XXXL and it’s mainly down to the length of the extrusion and logistical. We would need to invest in something bigger than the VF4 which is the best part of 200k. Joining extrusions wouldn’t be something we would do.


Hi Andy. Many thanks for this explanation. I now have some idea as to why the extrusion profiles are produced in the manner which we receive them. I had no clue about the limitations imposed by features which were additions to the extrusion being say… just a single bar of metal with a uniform length, width and height. I have never seen any kind of extrusion operation hence the dumb question.

Do you have any opinion about whether a chamfer might improve the lead-in to the screw holes?

Hello Luke, thank you for taking the time to improve my understanding of the issue which I raised. It was a question born out of my apparent lack of ability to drive the provided screws in a straight line in about half of my attempts. The carriages are substantial in weight and that did not help me to locate the screws.

Yes. This is understood! Would adding a chamfer by way of a lead-in to the thread be helpful?

Yes, that is apparent now that I have everything screwed together.

I barely have enough space to house the standard sized model (I really wanted to own a Shapeoko though) so I will not be asking you for a much larger machine. :grin:

That is good to know.

1 Like

@jepho chamfers would assist initial thread engagement, but there is a cost/benefit to ‘with chamfer’ tooling or further operations. As said, the outcome is good so additional cost is questionable.


@AndyC Thanks again Andy. I understand the point now. The cost/benefit ratio makes it a low priority and bean counting would probably negate the need for spending more on production because the current solution works. Good to know that the initial thread engagement would be assisted by a chamfer. My brain is now satisfied. :+1:

1 Like

Extruding Aluminium is an expensive and complex process, here’s a quick overview of the really expensive toys you need;


Is this entirely true?

If you search for “extruded aluminium screw holes” on google and look at the image results, there are plenty of extrusions available with completely enclosed screw holes.

1 Like

Enclosed holes are a) drilled in, or b) extruded in - but at the expense of flatness/trueness tolerance, or reduced extrusion rate to control cooling whilst the ejection from the tool can be controlled as straight. It’s all possible, but it’s not all the same cost.
There are layers of tooling/speed/flatness/stiffness/cost trade-offs with all such subjects, but ‘in the main’ people tend to go for controlling costs and accepting the trade-offs. Short, wide flatness tolerance, low stiffness extrusions is one such set of trade-offs in which closed holes are possible…

1 Like

Okay… I’ll have to take your word for it. I can’t find any readily available aluminium extrusions that do not have fully enclosed screw holes. So I have to assume that they are all sub-par with respect to flatness tolerance, or that the cost trade-off is that enclosed ones are cheaper.

1 Like

One example of this is MakerSlide — if memory serves, the profile actually changed — the initial version had an additional web which was subsequently removed.

Interestingly “Wide Makerslide” has open screw holes:

Interesting - the X-Carve pictures have the extruded holes isolated with webs, presumably to reduce thermal stresses on the main body and thus improve flatness for a given machine speed. They’re also quite fine walled compared to the ShapeOKO ones next to them. Someone with structural engineering knowledge might want to comment on the likely overall stiffness from their webs vs the thicker walled alternative…?

I think the holes are offset because each of the black extrusions have t-slots on two or more faces… maybe that layout is more common that the Shapeoko one since ones like that below are pretty much the default sort of extrusion that comes up (at least for my simple searches):