My CNC Router Build Adventure

[peteeng]

Hi Jayne - The BST brand has the same physical specs as Hiwin. Get medium preload. BST has always been prompt to sort any issues I have had with their product. I have used various brands and the BST feels the same and as far as I can tell they are good for the price. A light preload car will have motion hysteresis and not be as stiff as a higher preloaded car. Light is fine for lasers and low contact force applications. But routers need preload and mills need heavy preload. Unfortunately BST can only provide medium and they are the same price as std, but you have to specify the preload required in the order.

The relationship between the bed and nose and the Z top and bottom height is always a juggle. Its good time spent sorting that out for your intended jobs. You can have 200mm travel but not be able to use it if you make the gantry the wrong height or the bearings in the wrong place. A vertically adjustable or placement spindle is a good thing. Some jobs will need it to be up a bit some down to get the tool in the right place or the top position of the spindle in the right place eg to allow the tool to have safe clearance over fixtures.... or to get deep enough in part. Getting to nitty gritty stuff now.... Peter

and model the grease nipples in and the bolts etc near the ends of the rails. Or at least remember that they can rob you of 10-20mm travel if your not aware of them....

[peteeng]

Hi Jayne - Cleaning out some files thought you may like this. Only comment I don't like is the Golden Rectangle. Its an interesting number but its just a number has no real significance. Just make bearings as far apart as possible.... Peter

[mactec54]

Hi Jayne - The BST brand has the same physical specs as Hiwin. Get medium preload. BST has always been prompt to sort any issues I have had with their product. I have used various brands and the BST feels the same and as far as I can tell they are good for the price. A light preload car will have motion hysteresis and not be as stiff as a higher preloaded car. Light is fine for lasers and low contact force applications. But routers need preload and mills need heavy preload. Unfortunately BST can only provide medium and they are the same price as std, but you have to specify the preload required in the order.

The relationship between the bed and nose and the Z top and bottom height is always a juggle. Its good time spent sorting that out for your intended jobs. You can have 200mm travel but not be able to use it if you make the gantry the wrong height or the bearings in the wrong place. A vertically adjustable or placement spindle is a good thing. Some jobs will need it to be up a bit some down to get the tool in the right place or the top position of the spindle in the right place eg to allow the tool to have safe clearance over fixtures.... or to get deep enough in part. Getting to nitty gritty stuff now.... Peter

and model the grease nipples in and the bolts etc near the ends of the rails. Or at least remember that they can rob you of 10-20mm travel if your not aware of them....

Grease nipples don't take up any travel at all, they are not part of the linear bearing and can be placed on either side of the Bearing block, even joining them together with tubing, and having ( 1 ) central grease point is the best way to take care of Linear Bearing lubrication

A medium Preload is a bad Idea unless everything is machined, even a light pre-load can be a problem if mounted on unmachined surfaces

[peteeng]

Hi Mactec - grease nipples can take up travel if you use the rail holes for the hard end stops. Agreed central greasing is great but most hobbyists don't do this. I have used medium and high preload on my machines and they don't have machined surfaces. My first machine I used std preload and replaced them with high preload after 700hrs of use as they developed clearance. The machines do perform better with preload. Peter

[mactec54]

Hi Mactec - grease nipples can take up travel if you use the rail holes for the hard end stops. Agreed central greasing is great but most hobbyists don't do this. I have used medium and high preload on my machines and they don't have machined surfaces. My first machine I used std preload and replaced them with high preload after 700hrs of use as they developed clearance. The machines do perform better with preload. Peter

You should not be using the Rail mounting holes as a stop, the Linear Bearing should not have to hit anything or it will damage the end seals

So the medium preload you have now, won't last as long as the std linear Bearings you had before, they will soon be just as loose as a standard Linear Bearing when they are not mounted on a machined surface, this is the problem everyone gets. sure they will be good for a little while.

They need clearance when they are not mounted on a machined surface, this clearance allows for the poor mounting surfaces.

[JayneV]

Hi everyone,

This is a long post (TLDR - CAD model has been tweaked, nearly finished)

Mactec, thank you for the pivot pin and grub screws idea for the left/right tramming adjustment. It didn’t occur to me to make the forward/back adjustment at the linear bearings either side of the gantry. I was thinking making the adjustments more locally around the z plate. I will think about your suggestions and how to implement them into my design.

Peter, thank you for the Fundamental Principles file. It was an interesting read, introducing me to several principles I can now learn more about. My all time favourite is KISS, which can be be applied to so many aspects of life.

I also, understood what you meant about the grease nipples on the carriages potentially robbing some travel. I downloaded the HIWIN models for the linear rails and carriages, since they have the same dimensions as the BST parts. The models included the grease nipples and I discovered the Y and Z axes range of travel are not affected by the grease nipples, but the X axis is. A couple minor adjustments will sort this out.

The last couple of days I have been busy tweaking the model and repeating several hours worth of work after Fusion crashed with lots of unsaved changes. It was a rookie mistake where I got carried away with making the changes and forgot to save along the way. Oops!! The changes were mostly related to swapping out the generic linear guides and ballscrew assemblies with more accurate models, and then tweaking the surrounding structure so everything fit together nicely. The entire gantry assembly is now about 95% complete. Some mounting holes still need to be created for mating parts. I still need to finish swapping the ballscrew assemblies on the base for the Y axis, but it is partially done and have simplified how the ballnut attaches to the gantry be removing one element from the load path. The ballnut now bolts directly to the gantry mounting plate and two of the bolts also capture the lower plate of the gantry beam itself.

I just noticed the constraint that mates the gantry beam to the mounting plate is slightly misaligned in the attached photos. I’ll fix that tomorrow so ignore how the right side bearings are not aligned with the rail. Looking at the second attached image (with the blue bottom plate of the gantry beam) it seems to make sense to extend that plate out to the edge of the mounting plate so all 4 bolts of the ballnut bracket pass through the plate, as was suggested some time back in this thread.

Would there be any benefit to adding a top plate to the saddle assembly as shown by the red markup in the last image to offer some added stiffness?

Our lockdown restrictions were eased today, which means we are now permitted to freely travel within the city limits if we have had both vaccine shots, so tomorrow I will be going to my brother’s shop to make a part or two. The parts will just be some spacers for the linear bearings on the x axis so I can finally hold a physical part in my hand. Hopefully that will boost my motivation because I feel like I am getting stuck in the CAD model suffering from analysis paralysis.

I will also have a chance to go over the model with my brother to figure out how to machine the contact surfaces for the linear rails. The z axis should be easy because it is not that big, but the X and Y axis might be larger than the working envelope of his machines. I’ll know more tomorrow.

Have a great evening/day.
Jayne

PS. The first image, in full colour, was a screen shot captured from my computer, the rest are screenshots from the Fusion 360 app on my phone. There doesn’t seem to be an option in the phone app to turn on colours or edge visibility.

[mactec54]

Hi everyone,

This is a long post (TLDR - CAD model has been tweaked, nearly finished)

Mactec, thank you for the pivot pin and grub screws idea for the left/right tramming adjustment. It didn’t occur to me to make the forward/back adjustment at the linear bearings either side of the gantry. I was thinking making the adjustments more locally around the z plate. I will think about your suggestions and how to implement them into my design.

I also, understood what you meant about the grease nipples on the carriages potentially robbing some travel. I downloaded the HIWIN models for the linear rails and carriages, since they have the same dimensions as the BST parts. The models included the grease nipples and I discovered the Y and Z axes range of travel are not affected by the grease nipples, but the X axis is. A couple minor adjustments will sort this out.

The last couple of days I have been busy tweaking the model and repeating several hours worth of work after Fusion crashed with lots of unsaved changes. It was a rookie mistake where I got carried away with making the changes and forgot to save along the way. Oops!! The changes were mostly related to swapping out the generic linear guides and ballscrew assemblies with more accurate models, and then tweaking the surrounding structure so everything fit together nicely. The entire gantry assembly is now about 95% complete. Some mounting holes still need to be created for mating parts. I still need to finish swapping the ballscrew assemblies on the base for the Y axis, but it is partially done and have simplified how the ballnut attaches to the gantry be removing one element from the load path. The ballnut now bolts directly to the gantry mounting plate and two of the bolts also capture the lower plate of the gantry beam itself.

I just noticed the constraint that mates the gantry beam to the mounting plate is slightly misaligned in the attached photos. I’ll fix that tomorrow so ignore how the right side bearings are not aligned with the rail. Looking at the second attached image (with the blue bottom plate of the gantry beam) it seems to make sense to extend that plate out to the edge of the mounting plate so all 4 bolts of the ballnut bracket pass through the plate, as was suggested some time back in this thread.

Would there be any benefit to adding a top plate to the saddle assembly as shown by the red markup in the last image to offer some added stiffness?

Our lockdown restrictions were eased today, which means we are now permitted to freely travel within the city limits if we have had both vaccine shots, so tomorrow I will be going to my brother’s shop to make a part or two. The parts will just be some spacers for the linear bearings on the x axis so I can finally hold a physical part in my hand. Hopefully that will boost my motivation because I feel like I am getting stuck in the CAD model suffering from analysis paralysis.

I will also have a chance to go over the model with my brother to figure out how to machine the contact surfaces for the linear rails. The z axis should be easy because it is not that big, but the X and Y axis might be larger than the working envelope of his machines. I’ll know more tomorrow.

Have a great evening/day.
Jayne

PS. The first image, in full colour, was a screen shot captured from my computer, the rest are screenshots from the Fusion 360 app on my phone. There doesn’t seem to be an option in the phone app to turn on colours or edge visibility.

Peter, thank you for the Fundamental Principles file. It was an interesting read, introducing me to several principles I can now learn more about. My all time favorite is KISS, which can be be applied to so many aspects of life.

My thoughts also.

Just add a hole in the end plates or a cutout, where the Grease Nipples may hit no travel lost, don't move the Bearings in because of this, there are more than one way to solve this, if you see this problem.

[JayneV]

My thoughts also.

Just add a hole in the end plates or a cutout, where the Grease Nipples may hit no travel lost, don't move the Bearings in because of this, there are more than one way to solve this, if you see this problem.

That is what I intend to do. The bearings are about as far apart as they can be without increasing the dimensions of their respective mounting surfaces. It was only the X axis bearings that have a slight interference problem with the grease nipples but I think that will be solved with a notch in the gantry mounting plates. Will confirm when I fix the constraint error in the gantry assembly.

[JayneV]

Hi everyone,

I’ve decided to use 2010 ballscrews for the x and y axes, mainly due to the length of the Y at 1500mm and the X at 1050mm. For the Z axis I have gone with a 1605 screw at 400mm length. With a 16mm screw, the assembly is a little more compact, reducing the distance from the spindle to the gantry beam by a small amount. Can I reduce the screw size further, by going down to 12mm or will that cause problems?

What are the important factors to consider when selecting screw size? I know of two things, whipping and buckling, but I don’t know how to do the calculations so I’ve gone with what I think is oversized as a precaution. I’m guessing a larger diameter screw will have more inertia which may affect acceleration but I could limit the acceleration in the controller settings.

Jayne

[peteeng]

Hi Jayne - The saddle does not need the extra plate, better to put more material into the Z axis tool plate. Looking at some of my calcs (by the way the Hiwin manual has the equations for whirl and buckling) a 1400mm long Dia1610 screws first vibration is at 1228 rpm so this is above the speed a stepper will run at. It also buckles at 287kgf which is unlikely to be achieved by an N23 stepper as well (maybe in a crash) . Dia20mm screws are big maybe consider 16mm... The CAD development side can be a slog but worth it. Once you get to a final design have a break do something else you like to do then review with a clear head. It is worth doing a drawing set and that IS a slog... It is often quicker to rebuild a model from scratch then try to sort it. Its the developed geometry that is the hard bit and due to the constant changing of the geometry models can become flaky. When I manage engineering work one step before releasing a design is to check and rationalise every part so it becomes robust and simple. Then the same applies to assemblies. Makes future work so much easier but that's a commercial strategy. Re: accelerations. The drive system is self limiting. You start with a small value and slowly bring it up until the motor stalls then back it off. Keep at it and enjoy. Peter


https://knowledge.autodesk.com/suppo...341a3a2f5.html

[peteeng]

Hi Jayne - On a CAD thought. There is a modelling strategy called Resilient Modeling Strategy (RMS) and its worth a look. Using parametric modellers is usually not taught you are thrown into it and modellers tend to have lots of bad habits. RMS is useful to becoming a power user. Peter

[mactec54]

Hi everyone,

I’ve decided to use 2010 ballscrews for the x and y axes, mainly due to the length of the Y at 1500mm and the X at 1050mm. For the Z axis I have gone with a 1605 screw at 400mm length. With a 16mm screw, the assembly is a little more compact, reducing the distance from the spindle to the gantry beam by a small amount. Can I reduce the screw size further, by going down to 12mm or will that cause problems?

What are the important factors to consider when selecting screw size? I know of two things, whipping and buckling, but I don’t know how to do the calculations so I’ve gone with what I think is oversized as a precaution. I’m guessing a larger diameter screw will have more inertia which may affect acceleration but I could limit the acceleration in the controller settings.

Jayne

16mm for the Z axis would be the norm for a build like this, X and Y Axis 20mm would be a good choice, 12mm the core diameter is too small for what you are building, if inertia is a problem you can make a rotating Nut that takes away any of the screw inertia problems.

[JayneV]

Good morning Peter,
Thank you for the screw calculations. I was unsure about selecting the size, especially for the long Y axis. My gut feeling was that 16mm may or may not be sufficient so I upsized to 20mm to be sure. I'll work through your calcs to try and get an understanding of the results and consider using 16mm.

Regarding CAD, I have heard the term Resilient Modeling Strategy (RMS) before, but didn't know what it meant. As for parametric modelling, it's something I have attempted to use efficiently but continue to fall short because I am always eager to get on with whatever model I'm designing rather than follow some tutorials to learn how to do it properly. My CNC, Carlos (yes, I named it Carlos ) is too far along to try and remove the "bad habits", but I'll watch the videos you linked and any others I find before starting future projects.

I have started making a drawing set (only 2 sheets so far) and intend to make a complete set along with a full parts list. Some drawings will be to send to laser/waterjet cutters and the rest will be for me while building Carlos.

Have a great day,
Jayne

PS: Why is the machine called Carlos? I got tired of referring to it as "the machine" or "the CNC" and Carlos was the first name that popped into my head.

[JayneV]

16mm for the Z axis would be the norm for a build like this, X and Y Axis 20mm would be a good choice, 12mm the core diameter is too small for what you are building, if inertia is a problem you can make a rotating Nut that takes away any of the screw inertia problems.

Hi Mactec,
As the model is right now, the X and Y have a 20mm screw and the Z has a 16mm screw.If I downsize X and Y, I will need to adjust some bits and pieces to account for the different size mounts, not a big job. I haven't seen photos of a rotating nut setup so would need to do some research and figure out how the motor and nut are connected. It's probably overkill for Carlos but you have piqued my interest and will do a search to see how what a rotating nut setup looks like.
Jayne

[JayneV]

Just found this video, looks much simpler than I imagined it would be. Would the belt drive introduce some backlash or would the backlash be negligible for such a short belt?

[mactec54]

Hi Jayne - On a CAD thought. There is a modelling strategy called Resilient Modeling Strategy (RMS) and its worth a look. Using parametric modellers is usually not taught you are thrown into it and modellers tend to have lots of bad habits. RMS is useful to becoming a power user. Peter

Parametric modeling why would you want to learn this, Direct modeling is superior, if someone is just starting in cad, they should be doing it, with Direct modeling cad software. Files are better to work with and easy to make changes on the fly, and the cad to cam is also much better.

You have 3 types of modeling packages

Parametric cad modeling

Direct cad modeling

Synchronous modeling, Being Parametric and Direct cad modeling in the same package

[JayneV]

Hi Mactec,
I have been been working in direct modelling mode quite a bit with Carlos, partially because the model wasn’t properly set up in parametric mode and partially because sometimes a quick change was just easier to do directly. However I have found that sometimes things go a bit haywire after making a direct change which is most likely due to the model currently being in edited in both direct and parametric mode without being properly optimised for either mode. It would be good to learn how to use both methods properly rather than figuring it out as I go.

[JayneV]

Hi everyone,

Part manufacturing for Carlos officially commenced today. Nothing too fancy, just a couple spacers so I can see a CNC mill in action. I found it very exciting and can't wait to get Carlos up and running to start playing. I'm trying to decide what finish to have. I could try a light rub with scotchbrite or bead blasting. Anodising is another option, either clear or a colour.....decisions decisions.

Anyway, these parts are just a small teaser to keep momentum going until I finish the CAD model.

Jayne

[mactec54]

Just found this video, looks much simpler than I imagined it would be. Would the belt drive introduce some backlash or would the backlash be negligible for such a short belt?

Not if you use the GT3 series Timing Belts

[JayneV]

Not if you use the GT3 series Timing Belts

That’s good to know. I’d like to do a project using timing belts some day.