Rugged CoreXY

That force should be entirely contained by the hotend, provided the mounting for the bowden tube remains firm. The extruder pushes the filament relative to the bowden tube, and at the hotend the filament is trying to push the hotend off the bowden, which is why the mounts for the bowden at both ends have to be good. However much compression the filament is under, the bowden tube experiences the matching force in tension, but none of that force should be felt by the linear motion system. There is the issue of the elasticity of the bowden & filament putting lateral force on the print-head, and I've been ignoring that in the touchy-feely hope that it's relatively small compared to other forces on the system.

[edit: I had to struggle with that, it's exactly the opposite of normal bowden cables where the core is under tension and the casing under compression]

Edited 1 time(s). Last edit at 06/08/2016 09:03PM by JamesK.

Yes it's true the internet can be a pain, but I like to think that things can be clarified.

For example I sent this to the Hiwin Australia office:

Hi xxxxx,

I’m hoping to use a single long carriage (EG15) across the X axis, mounted on the “dog bone” shaped plate as in the picture. The mounting plate will be checked/machined for flatness, and the axis’ will all be trammed properly with a dial gauge.

The thick part of the green bracket will extend down the front side of the carriage into the work area and will mount the extruder/hot end assy, which I expect will weigh 600-700gms, but let’s allow 1.5kg.

I don’t know the moment loads, but I hear figures mentioned like 100-300mm/sec travel speed and 2000-8000mm/sec acceleration.

It does drop over one side, but I’ll be able to balance the load so it’s center of gravity will be directly under the center of the rail, and I’m expecting the load to be about 100mm below the bottom of the rail.

Can you please clear me up as to whether this is an acceptable way to mount it?

Cheers,

xxxxx.

And in return I received this reply:

Hello xxxxx,

Design looks solid. Worst case moments scenario is 0.1m x 1.5kg x 8m/sec/sec = 1.2Nm
This is around 1/100th of the 15mm block moments capacity.

So going on this:

Koko is correct in saying that my choice of linear is way overkill for this application, and totally more so if I'd used 2 rails. Also agree about the importance of reference edge and proper alignment for the best result.

I then went on to check the weight of both the rail and the block (MG vs EG, both 15mm), and got 37gms difference for the block and 156gms difference for 400mm of rail.

Total = 193 gms.

Considering Koko is recommending MG12 which will be lighter again, and the fact that we're 100 times the moment load (and even if the guy was miles off, we can assume 20 times) I'd say he's quite correct there too.

Right from the start I have said I wasn't interested in speed (I'll get an RC buggy or something to scratch that itch), and that I wanted the thing to be rugged and precise.

I also see the possibility of an engraving or milling attachment off in the far distant future, which would exert much more force than printing.

That and the fact that I'd to use the same rail all round, and I get a much higher spec rail for less money, I will try the EG15 rail first, and if the weight is problematic I'll then try the MG12 solution.

I believe I've taken advice on board and made my best efforts to make an informed and unbiased decision, which at this stage is as above until tested.

Edited 3 time(s). Last edit at 06/09/2016 01:08AM by hobbymods.

Sounds good to me, can't wait to see it come to life!

Is your dog bone part intended to impart some extra rigidity to the X axis? If so, you might get much more stiffness without a lot of increased weight by using a piece of rectangular tubing instead of a flat plate.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

Quote
the_digital_dentist
Is your dog bone part intended to impart some extra rigidity to the X axis? If so, you might get much more stiffness without a lot of increased weight by using a piece of rectangular tubing instead of a flat plate.

Thing with that is that it needs to have a flat surface with reference shoulder about 3mm deep ground into it.

I'll trim it up before the laser cut mockup.

I'm really basing my design on the old Zortrax M200 H bot clunker....heavy, slow speed, really good repeatable prints.

Just a bigger, open software corexy version...my needs are simple

But yes, I will look at ways to make it light, more rigid.

I'm just mocking up with the material I have at hand, which happens to be a few sheets of this 6mm acrylic.

Edited 1 time(s). Last edit at 06/09/2016 08:46AM by hobbymods.

For Aussies I used these guys for 20T pulleys and 10mm wide GT2 belt.

They are reasonably well known and have a reasonable reputation.

[www.ebay.com.au]

[www.ebay.com.au]

[www.ebay.com.au]

[www.ebay.com.au]

I'm considering configuring the pulleys with the staggered pulleys on the X axis, like the Fabotum design without the belts crossed.

[blog.fabtotum.com]

The Gates guy seemed to lose interest, but he did mention that even using different idler sizes for staggering pulley positions can cause variations in the torque in different parts of the belt circuit, which I have no idea if it is true.

Quote
hobbymods

The Gates guy seemed to lose interest, but he did mention that even using different idler sizes for staggering pulley positions can cause variations in the torque in different parts of the belt circuit, which I have no idea if it is true.

think about it. the top part of the belt will be a different length from the bottom part of the belt. so, the top of the belt will be under a different tension from the bottom. that's bad enough, but as the carriage approaches an idler that's angled, those tension variations are *going to change* dependent on the length.

i know someone did mention earlier that "everyone does this" (moves the idlers up so that the belts can cross while being in the same plane) but that does not make it a good idea. we're still (as a community) researching corexy.to work out what's best.

I think we're getting our wires crossed a little bit again.

He said it was important to use the same diameter idler pulleys, 90 deg angles, the same length of belt (obviously) and even tension in both belts (already discussed) for best balance of torque throughout the thing.

I also read that any kind of spring loaded tensioner might be an issue in a belt driven system that changes direction quickly (I was considering one for the Z axis) and shouldn't be used.

I don't claim to know anything at all about belts, pulleys, idlers etc. It's something I've never worked with car fanbelt aside, so I can't comment myself.

I'll go with the stacked belts, but yes I'm sure they add some kind of twist, they'd have to.

I've ordered a bunch of stuff now and might lay off with the theoreticals now until I've put something a little more substantial together.

Edited 1 time(s). Last edit at 06/10/2016 05:13PM by hobbymods.

Quote
lkcl

but as the carriage approaches an idler that's angled, those tension variations are *going to change* dependent on the length.

.

This quoted part is not accurate. There are no change in length. The 90 deg rule for belts that changes length is universal.

About strain in the belt;

As long as you are in spec for the belt you can twist and turn it as you like. What matters is that the center fibre remains the same length all around. It was discussed here some month back.

Changing plane for belts is a textbook thing and while you don't see it outside agriculture these days, it is still used .

In general, nobody here has ever reported a broken belt. Talking about wear on belts seem just a little far out.

Now, I think it is easier with two plane belts because getting the single plane belts crossing done right is hard.

On that subject ikcl, I think you just aimed this one at me:

"i know someone did mention earlier that "everyone does this" (moves the idlers up so that the belts can cross while being in the same plane) but that does not make it a good idea. we're still (as a community) researching corexy.to work out what's best."

Maybe you need to be a little careful in cutting in what people say and quoting without actual quoting and changing the text. If you think there is a consensus in the CoreXY builds that single plane belts with raised pulleys is what everyone does, then you say that, and then you say that you don't agree. But this is not the case. I think most builds are two plane these days. Which is good and right I think, I build them both. But crossed belt is not a disaster either and there are plenty of builds that prints very nicely with that...

Or maybe you were referring to twisted belts which I think is a much more relevant discussion these days. Here we can talk about something that splits the waters.

Hobbymod - as you said, all very theoretical. One very real question is if you are going to twist the belt coming off the motor so it is resting with the flat side against your pulleys or if you will let the teeth side run against plain pulleys or if you will use geared idlers. This impacts at least the noise but some people claim it also effects print quality.

If this is difficult to understand then try to think through the belt path on one motor and how the belt will interact with pulleys. Then you realize it.

I personally twist them, but it is an antique discussion also raging on Delta printers


/EDIT: Made the reply to Ikcl part more constructive.

Edited 2 time(s). Last edit at 06/10/2016 09:50PM by LarsK.

I wont be twisting belts or having them run hard against flanges.

I will extend the shaft on each stepper so that the drive pulley will be on the correct plane for each of the 2 circuits, and do some sort of pedestal bearing mount to support that extension.

I've got toothed idlers coming for where it turns on the toothed side and plain idlers for the other, I'm not going to all this trouble to ignore the basic rules of belt drives.

Interestingly I've read that smooth idlers should not be flanged for some reason, and I did note that the Gates guy expressed surprise at my request for flanged smooth idlers.

Quote
hobbymods
I wont be twisting belts or having them run hard against flanges.

I will extend the shaft on each stepper so that the drive pulley will be on the correct plane for each of the 2 circuits, and do some sort of pedestal bearing mount to support that extension.

I've got toothed idlers coming for where it turns on the toothed side and plain idlers for the other, I'm not going to all this trouble to ignore the basic rules of belt drives.

Interestingly I've read that smooth idlers should not be flanged for some reason, and I did note that the Gates guy expressed surprise at my request for flanged smooth idlers.

You could just design the motor mounts so that one will be higher than the other. That way you can use short shafts which I think is always best when you are applying sideways forces to them (less leverage).

I wonder why flanged idlers are a bad idea? The Open Builds idlers are all flanged (of course that doesn't mean it's the right thing to do). I suppose you wouldn't want the sides of the belts rubbing on the idler flanges but if the idler is a couple of mm bigger than the belt width, that shouldn't be an issue. If you have to rely on the flanges to keep the belt in the correct plain, then something has been built/designed wrong.

Quote
deckingman

Quote
hobbymods
I wont be twisting belts or having them run hard against flanges.

I will extend the shaft on each stepper so that the drive pulley will be on the correct plane for each of the 2 circuits, and do some sort of pedestal bearing mount to support that extension.

I've got toothed idlers coming for where it turns on the toothed side and plain idlers for the other, I'm not going to all this trouble to ignore the basic rules of belt drives.

Interestingly I've read that smooth idlers should not be flanged for some reason, and I did note that the Gates guy expressed surprise at my request for flanged smooth idlers.

You could just design the motor mounts so that one will be higher than the other. That way you can use short shafts which I think is always best when you are applying sideways forces to them (less leverage).

I wonder why flanged idlers are a bad idea? The Open Builds idlers are all flanged (of course that doesn't mean it's the right thing to do). I suppose you wouldn't want the sides of the belts rubbing on the idler flanges but if the idler is a couple of mm bigger than the belt width, that shouldn't be an issue. If you have to rely on the flanges to keep the belt in the correct plain, then something has been built/designed wrong.

My design will put the steppers out of the build enclosure, therefore out of the heat which Im happy to extend the shaft for.

Dont know about the smooth idler thing other than it came up twice in different places.

Longer shaft will be supported as mentioned above and I've got a trick or 2 up my sleeve for that bit that I'll try before I discuss

Edited 2 time(s). Last edit at 06/11/2016 05:20AM by hobbymods.

Quote
hobbymods
I wont be twisting belts or having them run hard against flanges.

I will extend the shaft on each stepper so that the drive pulley will be on the correct plane for each of the 2 circuits, and do some sort of pedestal bearing mount to support that extension.

I've got toothed idlers coming for where it turns on the toothed side and plain idlers for the other, I'm not going to all this trouble to ignore the basic rules of belt drives.

Interestingly I've read that smooth idlers should not be flanged for some reason, and I did note that the Gates guy expressed surprise at my request for flanged smooth idlers.

It is a misunderstanding that you cannot twist belt. If it is within spec you can do it. Refer to the below pages from Mechanical Components Handbook, and other design handbooks for belt designs.




Now the book also says there is a lifetime cost on the belt, but do we ever get near the lifetime for the belt? I don't think so. And even then, on the printers where people have toothed belt running against flat idlers, I also don't think that is very good.


You have decided to run with toothed idlers (which was what I meant to say when I said geared idlers) and that is the optimal solution I guess. Both problems solved. I am very interested to see what toothed idlers you have found, as it is something I myself have been looking for. I am also using toothed idlers at one point in my design and have ended up using those same we use on the motors. They are on a shaft that I have then supported by bearings but it is not optimal. In a different thread here somebody showed using miniature bearings to get the same but get the bearings inside the part itself.

Do those toothed idlers have an internal spacer for the bearings? Al the cheapo Chinese idlers I've seen have no spacer to keep the bearing hubs separated a specific distance. When you screw them down, the screw side-loads the hubs and it is easy to damage the bearings.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

Looks like Flat / Vee belt section of the handbook
Does it have a timing / Power Transmission belt section

HANDBOOK OF TIMING BELTS AND PULLEYS.
[www.sdp-si.com]

Not good enough?

You can twist a timing belt all you want
but
The common ones we use --- are NOT designed to be twisted

best to design for what the belt was made to do

Where did this can't run flanged pulley come from?

Well if you have a perfect design with perfect alignment great
but
a flange is going to help keep belt on the system
and with the back and forth motion we are using.
perhaps a ding is belt life.

DD -- had a good idea --- washers as flanges washer spins
instead of rubbing belt

Quote
LarsK
It is a misunderstanding that you cannot twist belt. If it is within spec you can do it. Refer to the below pages from Mechanical Components Handbook, and other design handbooks for belt designs.

Quote
cozmicray
Looks like Flat / Vee belt section of the handbook
Does it have a timing / Power Transmission belt section

HANDBOOK OF TIMING BELTS AND PULLEYS.
[www.sdp-si.com]

Not good enough?

You can twist a timing belt all you want
but
The common ones we use --- are NOT designed to be twisted

best to design for what the belt was made to do

[dpk3n3gg92jwt.cloudfront.net]

Page 199 give you the tolerances for angular misalignment. It is 1/6" per foot or 5.22 mm per m. I twist mine on a 430 mm length. I cant readily find how I can calculate the extra length in the outer strain but using simple pythagoras gives me 0.029mm difference - I am sure I am a magnitude below the tolerance.

The above reference is from Gates them-self.

I do agree that you will not find anywhere that recommend twisting timing belts. From the engineering perspective, it is correct is to get toothed idlers like Hobbymods is doing. I was under the impression that he might use plain idlers (don't know why) and I am of the opinion that it is better to twist the belts then have them run with the tooth against plain faced idlers. Also, I am still of the impression that twisting will only have an impact on lifetime.


I did not say anything about flanged or not flanged. I can tell you, that we agree on this. Also the sources that I have seen says that flanges are just fine. From the same reference as above page 199 again;

Quote
Any degree of sprocket misalignment will result in some reduction of belt life, which is not accounted for in the normal drive design procedure. Misalignment of all synchronous belt drives should not exceed 1/4° or 1/16" per foot of linear distance. Misalignment should be checked with a good straight edge or by using a laser alignment tool. The straight edge tool should be applied from driveR to driveN, and then from driveN to driveR so that the total effect of parallel and angular misalignment is made visible. Drive misalignment can also cause belt tracking problems. [b]However, light flange contact by the belt is normal and won’t affect performance[/b]

There's ebay links above for all the pulleys I'm using, and Im using both flat and toothed idlers depending on the side of the belt Im touching.

Im not twisting the belt because I dont need to and I would also just rather not.

Im not sure about the bearings as I havent seen them yet, but they are off ebay so seeing is believing.

As it goes Id like to mount the idlers on 8mm unbrako shoulder bolts (which have a 6mm thread) so I might need different bearings.

Either way I'll use spacers or whatever to make sure theres no issue when they are tightened.

Edited 2 time(s). Last edit at 06/11/2016 05:06PM by hobbymods.

I had some experience with standing pulleys on bolts before and found that the tension on the belt was enough to cause some flex in the bolt and/or the 1/4" aluminum plate it was screwed into and the belt would try to walk off the pulley. I'm still working final details of my CoreXY design out, but trying to avoid that problem which will be multiplied because I'm using two pulleys at different heights on the same bolt. I'm mounting the bolts/pulleys in pieces of rectangular aluminum tubing in hopes that it will prevent that sort of deflection of the bolt by supporting it at top and bottom. Of course, the tubing my flex. We'll see....

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

Quote
the_digital_dentist
I had some experience with standing pulleys on bolts before and found that the tension on the belt was enough to cause some flex in the bolt and/or the 1/4" aluminum plate it was screwed into and the belt would try to walk off the pulley. I'm still working final details of my CoreXY design out, but trying to avoid that problem which will be multiplied because I'm using two pulleys at different heights on the same bolt. I'm mounting the bolts/pulleys in pieces of rectangular aluminum tubing in hopes that it will prevent that sort of deflection of the bolt by supporting it at top and bottom. Of course, the tubing my flex. We'll see....

XY%20stage%201.png

Yes Im not surprised and had some concerns about free standing posts doing just that myself, more specifically I thought it might damage the thread in the aluminium.

Id hoped that going to 8mm shoulder bolts with 6mm threads and using helicoil thread inserts in the ally plate might have it covered, but maybe not.

Good heads up that one, I'll look at top and bottom supports as well.

I can't find any Timing /Power transmission belt information, analysis or design
for belts reversing direction frequently?

Until HP, Epson or some other large firm asks a belt manufacturer for a design we may never see it.
Could pull up old IBM selectronic typewriter data? (left / right moves and ball spin via timing belt)

Even then the belt life without an optimized design may be utilized
because it life is 10x of anything else that may go wrong.

[www.sdp-si.com]

13.1 Inside/Outside
Inside idlers are generally preferred over backside idlers from a belt fatigue standpoint. Both are commonly used with good success. Inside idlers should be pulleys, but can be flat, if the O.D.is equivalent to the pitch diameter of a 40-groove pulley. Backside idlers should be flat and uncrowned.

I think the toothed / smooth idler pulley discussion came from possible vibration/jogging induced by belt running over smooth pulley
maybe seen in print layers??

hiya larsk, sorry i am extremely busy, i'm no longer able to focus full-time like i have been for the past 3 months, so apologies i didn't have time to go back and find who it was, or exactly what was said - sorry about that. brief comment: relying on the centre fibre does not strike me as sensible, at all. the belt's specifications are almost certainly quoted for an evenly-distributed load, as well as it... it's too complicated to explain in a brief message, so sorry.

second thing (i know i took it out from the quote) it depends on which idlers you lift so as to be able to cross belts. if you lift the ones that change in length because of the x-end carriages, then you're in trouble: yes you'll get belt tension variations. so don't do that if you lifted the ones that have the longest (fixed-length) then you *might* get away with it. but personally i still would never do it.

Quote
LarsK

As long as you are in spec for the belt you can twist and turn it as you like. What matters is that the center fibre remains the same length all around. It was discussed here some month back.


On that subject ikcl, I think you just aimed this one at me:

"i know someone did mention earlier that "everyone does this" (moves the idlers up so that the belts can cross while being in the same plane) but that does not make it a good idea. we're still (as a community) researching corexy.to work out what's best."

Quote
cozmicray


I think the toothed / smooth idler pulley discussion came from possible vibration/jogging induced by belt running over smooth pulley
maybe seen in print layers??

that was me - apologies for not doing a follow-up: i found that the warbling effect changed depending on speed (i had a part where cura slowed down the speed due to layer cooling time). therefore, if the pulsing effect varies with speed, logically it can *not* be dependent on how the idlers are set up.

Quote
cozmicray

DD -- had a good idea --- washers as flanges washer spins
instead of rubbing belt

confirmed - doing that on the sandwich200v2 after i saw it on DD's somm. works well. used 20mm diameter M5 washers. got a whole bag for $EUR 2

Quote
lkcl
hiya larsk, sorry i am extremely busy, i'm no longer able to focus full-time like i have been for the past 3 months, so apologies i didn't have time to go back and find who it was, or exactly what was said - sorry about that. brief comment: relying on the centre fibre does not strike me as sensible, at all. the belt's specifications are almost certainly quoted for an evenly-distributed load, as well as it... it's too complicated to explain in a brief message, so sorry.

As I showed from the belt manufacturing own guidelines for engineering; You can have an angular misalignment of up-to 5.22mm per m or 1/4 deg on the belts. Any misalignment effects lifetime. Is that a problem? I don't think so. Maybe it also effects absolute load carrying capability of the belt, is that a problem? I don't think so. It is a debate about compromises really.

I posted a link to the belt specification above, around pages ~180 and forward is the engineering guidelines.

No problem with the comment. This is not a job and sometimes it is better to put a quick comment then not write nothing at all

Quote
LarsK

I posted a link to the belt specification above, around pages ~180 and forward is the engineering guidelines.

awesome. thanks larsk.

Quote
JamesK
That force should be entirely contained by the hotend, provided the mounting for the bowden tube remains firm. The extruder pushes the filament relative to the bowden tube, and at the hotend the filament is trying to push the hotend off the bowden, which is why the mounts for the bowden at both ends have to be good. However much compression the filament is under, the bowden tube experiences the matching force in tension, but none of that force should be felt by the linear motion system. There is the issue of the elasticity of the bowden & filament putting lateral force on the print-head, and I've been ignoring that in the touchy-feely hope that it's relatively small compared to other forces on the system.

[edit: I had to struggle with that, it's exactly the opposite of normal bowden cables where the core is under tension and the casing under compression]

there's also the fact that the ID of the bowden tube cannot ever be the same as the OD of the filament (otherwise it wouldn't move freely, even if it is PTFE): as a result the filament buckles (in unpredictable ways) a bit like a microwave down a wave guide, it bounces off the inside, but far less predictably. this and the other factors you mentioned is why you have to do 4 to 5mm retracts on bowden tubes. but, more than that: the amount of side-to-side buckling is going to *alter* depending on where the tube goes as the printhead moves... so the actual amount of extrusion that occurs *varies* with the position of the printhead... again in unpredictable ways. did one printer with a bowden tube, and in three very intense full-time weeks saw *every* single possible problem you could possibly have with bowden tubes (because i was designing from scratch), and won't ever do one ever again. saw the Flex3Drive, went "yippee!" and bought one immediately.

Quote
lkcl

there's also the fact that the ID of the bowden tube cannot ever be the same as the OD of the filament (otherwise it wouldn't move freely, even if it is PTFE): as a result the filament buckles (in unpredictable ways) a bit like a microwave down a wave guide, it bounces off the inside, but far less predictably. this and the other factors you mentioned is why you have to do 4 to 5mm retracts on bowden tubes. but, more than that: the amount of side-to-side buckling is going to *alter* depending on where the tube goes as the printhead moves... so the actual amount of extrusion that occurs *varies* with the position of the printhead... again in unpredictable ways. did one printer with a bowden tube, and in three very intense full-time weeks saw *every* single possible problem you could possibly have with bowden tubes (because i was designing from scratch), and won't ever do one ever again. saw the Flex3Drive, went "yippee!" and bought one immediately.

You make it sound like threading a piece floppy string down a 6 inch diameter drainpipe. In reality we have stiff, 1.75mm diameter filament being pushed against very little resistance (if the printer is set up correctly) down a 2.0mm ID tube. "Unpredictable side to side buckling" and "bouncing off the inside" - in your dreams maybe but not in reality. I appreciate that you once attempted and failed to design and make a bowden setup work, but many of us have been successful so please don't keep coming out with all this theoretical dross that has no bearing on real life practical applications.

It would be awesome if the sniping at each other could cease....seriously there's some clever cookies among you, let's match that with with some tolerance of opinions other than your own and some maturity? Please?

Back on track....

I realize my design here is massively overdone, and part of that was because I wanted the possibility of using it as an engraver etc off in the future.

Then I see this design (look down to the corexy part, halfway down) and really wonder if this is strong enough:

[blog.fabtotum.com]

I mean seriously, if that plate can be bent I'd assume it must be 3-4mm thick and relying on the bends for rigidity.

It uses only rods which is fair enough as they look like they're 8-10mm, but there's only one bearing on either side of the Y axis, and pulleys mounted at the front on very light bolts with no top support.

Then there's the printed extruder carriage.

I just don't see something here that would hold good repeatable lines.

I like the way they've run their belts though.

I think it's reasonable to be sceptical, the general consensus is that it's at best a very big compromise to combine printing with milling. It's an attractive idea to re-use the 3d motion platform, but anything but the lightest milling is so far removed from the requirements of printing that the compromises don't make sense. It's bad enough trying to design a printer only mechanism balancing the compromises between speed, cost, quality and features.

I guess the biggest problem may be simple practicalities. The biggest problem with 3d printing is long print times. The most direct approach to robust milling is an XY table on dovetail ways driven by ballscrews, with the tool post moving in Z. The very high moving mass would equate to low acceleration (unless money is no object I guess) which for printing would often result in slower than average print speeds. If a 12 hour print becomes 24 or more, do you really want to have your very expensive CNC mill tied up doing that, or does it make more sense to have a dedicated printer and a dedicated mill that can run in parallel?

Quote
JamesK
I think it's reasonable to be sceptical, the general consensus is that it's at best a very big compromise to combine printing with milling. It's an attractive idea to re-use the 3d motion platform, but anything but the lightest milling is so far removed from the requirements of printing that the compromises don't make sense. It's bad enough trying to design a printer only mechanism balancing the compromises between speed, cost, quality and features.

I guess the biggest problem may be simple practicalities. The biggest problem with 3d printing is long print times. The most direct approach to robust milling is an XY table on dovetail ways driven by ballscrews, with the tool post moving in Z. The very high moving mass would equate to low acceleration (unless money is no object I guess) which for printing would often result in slower than average print speeds. If a 12 hour print becomes 24 or more, do you really want to have your very expensive CNC mill tied up doing that, or does it make more sense to have a dedicated printer and a dedicated mill that can run in parallel?

Now that I've read some reviews and watched a video of it printing, I'd say it's very reasonable to be skeptical of this particular "multitool".