New DIY build - design suggestions/ideas welcome

[zeeflyboy]

Thought I'd experiment a little with the seals for the Z-axis. I have some red material on order so we can stick with the colour scheme but for now to dial it in I'm just using what I've got which is a nice orange!

So this is the part, I've designed it now so that it has a locking flange to prevent it from going walkies... it should just push into the top/bottom plates and wedge in place.






So off to the printer to try it out... I'm using a material called Cheetah by ninjaflex. It's a fairly firm but flexible rubber material (shore hardness 95A) which is abrasion resistant, resistant to chemicals and pretty tough. These are printed at 0.2mm, I will try 0.1mm as well to see if we can get them a bit smoother.








Here you can see the flexible rubbery nature:




Fit is excellent, but maybe just a smidge tight. I need to check it with the part actually held in the plates, but it may just need a very slightly enlargement of the interior cut out.

[zeeflyboy]

So I was going to show you some lovely Z-axis end plates, but I goofed them up (I know, shock horror right?). I must have put in the incorrect stock dimensions as it just missed the back edges by a fraction... something also went a bit wrong when doing some interpolated drilling and my X axis shifted by 1mm. Something I've never seen happen before on this machine.... Can't have been the motors skipping steps as they are closed loop and would have faulted so something perhaps happened on the PC end. As it happened it wasn't a big deal, just resulted in some messy and over sized internal screw holes but combined with the other mess up and I'm just going to bin them and start over... I can take the opportunity of having seen them to do a few little design tweaks at least.

So close but no cigar:

Roughing - 8mm DoC, 1mm WoC, 1500mm/min




Finish pass done




Flipping for reverse chamfer on the seal holders:




Shame I need to re-do them!


Anyway, in happier news my RoverCNC spindle mount arrived today and it's a beauty... I ended up paying almost as much in shipping as in cost of the part sadly, but it was worth it - so much nicer than the extruded type mounts you see everywhere, this one has been machined from a solid block. You can see the difference in how smooth the bore is and just laying down the spindle in the bottom half gives a super snug fit with no wobble. Should be a great improvement compared to the one on my current machine.





Mmmmm... chunky.


I also had a bit of a brain wave (or rather realised I was being stupid as usual), there was absolutely no need to separate the rail wiper seals from the central seal given that I was 3D printing them. I joined them up into one big unified seal instead which is a much better idea.... Nailed the printing, I made the error of not accounting for thermal shrinkage which is why they were a touch tight on the rails. A scale factor increase of 0.7% in X and Y put the dimensions spot on the money once the material had cooled and they are now a beautiful fit.









Quick test with the rails:




And top plate

[mactec54]

Anyway, in happier news my RoverCNC spindle mount arrived today and it's a beauty... I ended up paying almost as much in shipping as in cost of the part sadly, but it was worth it - so much nicer than the extruded type mounts you see everywhere, this one has been machined from a solid block. You can see the difference in how smooth the bore is and just laying down the spindle in the bottom half gives a super snug fit with no wobble. Should be a great improvement compared to the one on my current machine.

I also had a bit of a brain wave (or rather realised I was being stupid as usual), there was absolutely no need to separate the rail wiper seals from the central seal given that I was 3D printing them. I joined them up into one big unified seal instead which is a much better idea.... Nailed the printing, I made the error of not accounting for thermal shrinkage which is why they were a touch tight on the rails. A scale factor increase of 0.7% in X and Y put the dimensions spot on the money once the material had cooled and they are now a beautiful fit

That's the best seals I have ever seen done for a Router Build, just a little bit thick, but anything like that is adjustable, a slight change in the design and add a flexible V shaped wiper around the front edges, would be beneficial, I sure a lot will follow that have a printer,will do them similar

What probe are you using

[zeeflyboy]

Cheers.

It's the TPA-2 probe by kurukesu... nice compact little design that uses carbide rods and balls and submerses them in a di-electric oil to prevent oxidation.

Sometimes I wonder if I should do fewer but less incremental updates - not sure which you guys would prefer?

Anyway, for now a little extra work done today.

I've done a few tweaks to the Z-axis design after having a long think about how it's going to be assembled and aligned on the machine. To that end I've added an eccentric bushing to the top of each carriage mount and a 6mm dowel pin at the base to act as a pivot point. The idea being that I can install the backplate on the machine, lining it up square to the gantry arm. One dowel pin will provide a hangar while installing and ensure its central on the slider.

Then I can use a dial indicator to align the right carriages to a machinist square based off the bed, which will ensure the rails are aligned 90 degrees to the bed in X, then use those to dial in the left carriages.

Only had time to do one side of the carriage mounts today but It came out very nicely and for once I managed to make a part without messing anything up. When it cam to reaming the dowel pin hole I just had a very low feed rate and span the spindle by hand... I'm thinking there must be some sort of cunning arrangement I can come up with in future to spin the spindle slowly using a separate motor and some HTD belt/pulleys for things that need super low rpm like reaming/thread milling.

Anyhoo:

Top side finished




Flipped and bottom side done (used dowels to align with Y, then probed the reamed hole to determine zero)










I also designed a top seal for the motor mount, took a few prints to get it perfect as the CAD model of the motor mount isn't entirely correct in the non-critical dimensions:

[zeeflyboy]

So re-visted my end plates with the correct size stock this time...

I also tweaked the designs a little, adding a port for the limit switch wire to the top plate and changed the bottom plate to accept a standard FF12 bearing mount (along with some room for adjustment).

(correct size!!!) stock:




First op done and bolted down




Top side ops done




Bottom sides:








And the back section of the Z-axis is looking a bit more complete! Still needs limit switch installing, side plates and seals machining up, then I can move on to the main front plate and tramming plate.








Good view of top seals:

[handlewanker]

Hi.....I expect this question has been asked before bit I couldn't find it in the whole thread......why did you opt to have the rails moving instead of the bearing blocks on the Z axis?

I'm asking as I'm thinking of doing a mini CNC router and the aspect of the Z axis configuration has me puzzled........I see it in a lot of builds with the blocks moving and the rails attached to the X axis saddle.

I've attached a sketch of what I think is preferable, but would like to know if there is a reason for the "other" way.

The slides in both are shown at their lowest level.

The rails are coloured in blue and the bearing blocks are in red.........my preference is type A in the sketch.
Ian.

[mactec54]

Hi.....I expect this question has been asked before bit I couldn't find it in the whole thread......why did you opt to have the rails moving instead of the bearing blocks on the Z axis?

I'm asking as I'm thinking of doing a mini CNC router and the aspect of the Z axis configuration has me puzzled........I see it in a lot of builds with the blocks moving and the rails attached to the X axis saddle.

I've attached a sketch of what I think is preferable, but would like to know if there is a reason for the "other" way.

The slides in both are shown at their lowest level.

The rails are coloured in blue and the bearing blocks are in red.........my preference is type A in the sketch.
Ian.

That is easy to answer, the moving rail is by far the best design, there are ( 2 ) things that come into play, better Z axes support over the whole travel, it is supported by the bottom linear Bearings, so any loading does not change,as it moves up and down, it also adds more support to the Z axes moving plate

This is not always the case though, where one way is better than the other, this Design limits the up travel, to the bottom of the Linear Bearings, where as the moving Linear Bearings can move higher, but does not support the Z axes moving plate as well, so it is all about how rigid that you want the most important part of your machine the Z axes

Your drawing is not a reality, ideal when both are at there lowest travel, the Linear Bearings would be at the same level

[ger21]

my preference is type A in the sketch.

If you move the bearings down to the bottom of the gantry in Sketch B, then there's virtually no difference.

Mounting the rails to the Z plate is a better way if you need a lot of gantry clearance. But if the gantry clearance is less than 5-6", there's very little difference.

[SSN Vet]

Late to the party here, but wanted to chime in to say this is a very stout looking build.

If you'd permit me to ask one question... when you slave two steppers to run the Y-axis, does that require a 4th driver? or is it possible to run two steppers off of one driver? I think I know the answer (4th driver required) but I'm planning my own build and want to make sure.

[handlewanker]

If you move the bearings down to the bottom of the gantry in Sketch B, then there's virtually no difference.

Mounting the rails to the Z plate is a better way if you need a lot of gantry clearance. But if the gantry clearance is less than 5-6", there's very little difference.

Hi....if you move the bearings down to the bottom in sketch B you won't get any travel at all....the slide is already at the bottom of it's travel in both sketches and shows the unsupported hang down you will have in sketch B..

My observation, hence the query, is that you can have a larger bearing spread in sketch A as the X axis saddle carrying the rails can be as high as you want.

This ensures the Z axis slide is always completely on the rails at it's lowest level at all times.

In sketch B it moves off of the rails more and more as it goes down......the cantilevering forces will increase therefor as the slide gets to it's lowest level.

I cannot see any advantage with the sketch B design configuration.
Ian.

.

[handlewanker]

That is easy to answer, the moving rail is by far the best design, there are ( 2 ) things that come into play, better Z axes support over the whole travel, it is supported by the bottom linear Bearings, so any loading does not change,as it moves up and down, it also adds more support to the Z axes moving plate

This is not always the case though, where one way is better than the other, this Design limits the up travel, to the bottom of the Linear Bearings, where as the moving Linear Bearings can move higher, but does not support the Z axes moving plate as well, so it is all about how rigid that you want the most important part of your machine the Z axes

Your drawing is not a reality, ideal when both are at there lowest travel, the Linear Bearings would be at the same level

Mac....the linear bearings cannot be at the same level....to be so would mean you have no down travel in the B sketch configuration I posted.

It's a total reality sketch..... if you apply a scale of 100mm for the gantry clearance.....in this case the lower of the 2 round bars.

To get the same amount of 100mm down travel in sketch B you need to have the bearing blocks raised to the level shown in the B sketch.......which means you have double the cantilevering forces applied to the lower bearing block as compared to sketch A.

The other reality is that if you needed extra travel in the Z axis slide to clear a long tool such as a drill etc, the design in sketch A will just need longer rails above the X axis saddle......the configuration in sketch B cannot do this.
Ian.

[zeeflyboy]

Your drawing is slightly off on sketch B which is maybe why it doesn't make sense. In reality in sketch B your spindle would be higher up on the plate so that the ER collet was just below the bottom of the plate.

How high from the bottom of the travel you place the lower carriage dictates how much retract you can achieve, which becomes a balancing act between having the carriages closer to the cutting, thus lowering the moment arm, and having them high enough to allow the tool to be retracted sufficiently.

Sketch A is a good general purpose arrangement, but the "back to front" arrangement is better for rigidity. It is generally more restricted in the ability to retract the bottom of the tool above the bottom of the gantry which can limit your ability to work with long tools and tall stock so it becomes a question of what design better suits your requirements.

[zeeflyboy]

If you'd permit me to ask one question... when you slave two steppers to run the Y-axis, does that require a 4th driver? or is it possible to run two steppers off of one driver? I think I know the answer (4th driver required) but I'm planning my own build and want to make sure.

In theory you could run both off a single driver, people do it in the 3D printing world sometimes. Never seen it done on a CNC and I assume the more advanced digital drivers wouldn't like it.

Much better to run it's own driver - you can either run it as a slave, or potentially run the step/dir from one axis output to both drivers.

[mactec54]

Mac....the linear bearings cannot be at the same level....to be so would mean you have no down travel in the B sketch configuration I posted.

It's a total reality sketch..... if you apply a scale of 100mm for the gantry clearance.....in this case the lower of the 2 round bars.

To get the same amount of 100mm down travel in sketch B you need to have the bearing blocks raised to the level shown in the B sketch.......which means you have double the cantilevering forces applied to the lower bearing block as compared to sketch A.

The other reality is that if you needed extra travel in the Z axis slide to clear a long tool such as a drill etc, the design in sketch A will just need longer rails above the X axis saddle......the configuration in sketch B cannot do this.
Ian.

I have built them both ways, and the moving rail is more rigid, both ways have a different problem, if you want more rigidity then you use moving rail, if you want more clearance the you go with the linear Bearing as the moving element, yes the linear Bearing for the moving rail are mounted at the bottom of the Z axes mounting plate

[handlewanker]

I have built them both ways, and the moving rail is more rigid, both ways have a different problem, if you want more rigidity then you use moving rail, if you want more clearance the you go with the linear Bearing as the moving element, yes the linear Bearing for the moving rail are mounted at the bottom of the Z axes mounting plate

Hi, I understand fully what you mean, I needed to know WHY one was preferred to the other...... but I still see that the A sketch has more going for it as a design concept especially when it comes to mounting the stepper motor too.

In 30 + illustrations of CNC routers both on EBAY and Pinterest amongst others.......ranging form a simple $250 model to a $30K + job, the configuration of Sketch A was 100%.......in no rendition of CNC routers was the Sketch B type apparent.

I would have to conclude that if it's general practice and good enough for industry, it's plenty good for me.

BTW.......the Z axis plate that the rails are mounted on are supposed to support the rails, not the rails having to add support to the plate to prevent cantilevering forces from flexing it due to the downhang of the Z axis at it's lowest level.....that is a ludicrous solution.

I won't pursue the point further as it's hijacking the build thread......I'll concede that there are more ways to peel a banana etc.
Ian.

[mactec54]

Hi, I understand fully what you mean, I needed to know WHY one was preferred to the other...... but I still see that the A sketch has more going for it as a design concept especially when it comes to mounting the stepper motor too.

One simple answer Rigidity which we have stated more than one time, As for the motor mounting for the Z axes there is no difference, both would mount the same way

the Z axis plate that the rails are mounted on are supposed to support the rails, not the rails having to add support to the plate to prevent cantilevering forces from flexing it due to the down hang of the Z axis at it's lowest level.....that is a ludicrous solution. Ian.

The rails are not supporting the moving plate, but they do add more strength to the plate as an added benefit

[handlewanker]

That's ludicrous.......the Z axis plate is SUPPOSED to ensure the rails are maintained flat and supported, not the other way round............although being of high tensile and case hardened steel on an aluminium background I suppose it could be an allowable spin off for that purpose.

Rigidity will always be a factor of the stickout distance the Z axis slide is away from the X axis slides coupled with the hang down distance at the maximum down shift position.

From the foregoing analysis of the different configurations I think it must be a case of agreeing to disagree amicably and choose whatever is thought to be the best option.

It's not cast in stone either way, although Industry does overwhelmingly prefer to go the way I sketched in A.

Of 60 Z axis design examples I "researched", approx.95% were as in sketch A......that must mean something.

BTW.....ALL column mills have the rails attached to the face gf the machine column and the bearing blocks on the head......why would a CNC router, moving or fixed gantry, be different?
Ian.

[zeeflyboy]

Rigidity will always be a factor of the stickout distance the Z axis slide is away from the X axis slides coupled with the hang down distance at the maximum down shift position.

I guess one way to think of it is to imagine you have the fixed design with the carriage right at the bottom. When at the bottom of travel the "hang down" distance of cutting tip to bottom carriage would be the same, carriages being at the bottom either way in either design. At any height above the lowest travel, the "hang down" distance will be lower on the fixed carriage design as the carriage is always at the bottom and the stick out distance reduces as the spindle plate moves up in Z.

Also as mentioned your drawing isn't really accurate, as the spindle would be mounted higher up the plate with the ER collet just below the bottom of the plate. This allows the spindle to be clamped lower down near the neck, further increasing rigidity. With the "traditional" design in sketch A that I have on my current machine, I have to clamp the spindle over half way up to be able to reach the bottom of the bed.... the only solution to which is to have a plate that hangs down, which then doesn't benefit from also having two large steel rails stiffening it up as it would in B.

Horses for courses.... both ways work so I would say go with whatever floats your boat. I haven't had a Z-axis of this design before so thought it would be interesting to try, I have never seen any quantitive testing of one vs the other (and it would be quite difficult to do) but it does seem to be generally accepted "knowledge" that this design results in a stiffer Z-axis.

I would just say design it whichever way suits your build and makes you happy!

[mactec54]

BTW.....ALL column mills have the rails attached to the face gf the machine column and the bearing blocks on the head......why would a CNC router, moving or fixed gantry, be different?
Ian.

And the Head is supported over it's entire travel, a router type machine does not do this so there is no comparison in it's Design, look at the X axes of the same machines and you will see the rails mounted on the table, this is the same principle as what the moving rail design is doing for a router type machine when used for the Z axes

Nothing is cast in stone, we said that from the start, but you don't get it, you have a choice to build what ever you like, if it does not work how you would like it to, or expect it to, you just do it over again

In back to back tests, the moving rail Design for Router type machines worked better, have you ever built anything close to any type of machine to cut metal if so post us the photos

Build something show us something that you have built, that works then you can criticize other designs, start your own thread when you are going to build something, I have 3 different designs of moving rail Design, that all work very well, think outside the box, don't be a follower, do something different

[handlewanker]

Now Now Mac.....I haven't been to the Moon either but I know it can only be done with a rocket that has boosters to get it to achieve escape velocity......that being the case I do not have to build a rocket vehicle to suggest it is the best way....others have been there before me.

As I said, 95% of the CNC routers I researched use the sketch A type of configuration.........if that doesn't mean anything to you perhaps you are a go it alone type and prefer swimming against the stream by choice.

At any rate, it's a matter of personal preference.....if you use one or the other configurations I doubt whether the World will come to an end abruptly.

I dispute the design concept of sketch B as being less practical or suitable....but that is based on my experience of machinery etc.

BTW.....experience is the history you base all your concepts on.....going against the stream is just a personal choice not a better solution.....as Zeeflyboy has said.

We could digress from the thread further and debate the concept of whether it's better to have linear rails as moving objects or fixed............but applied to a mill or router table and that would be a matter of conjecture as both designs are used throughout the industry from the cheapest to the most expensive.

On the other hand, the Z axis design in sketch A is not in that category as it's the most preferred design concept you will find......more a personal choice than horses for courses etc.
Ian.