Stepper motor mounts? /efficient machine build strategy

[catahoula]

Hi,

Compiling a BOM for my first CNC build. I've been able to answer most questions via google, but stepper motor mounts eludes me. Are most people machining their own mounts, as there's so much variability in ballscrew dia, mounting location, etc? Or are there some stock motor mounts that work with typical 1605 chinese ballscrews and I'm just not finding them?

Background: It's not that machining is a big problem, as I have access to a mill and a machinist friend, but this is a tool for a side business I'm trying to launch into my main gig, so I'm really focused on minimizing the amount of time futzing and fabricating. I need to make aluminum molds, so 16x20" x-y travels, good accuracy, rigidity for surface finish, and a high-speed spindle because I also need to do some carbon fiber trimming. So that rules out a benchtop vertical mill conversion, and while the Avid Benchtop Pro router probably has enough stiffness and resolution, instead of spending my time checking and re-machining the 80/20 extrusion so the frame is guaranteed square, I'd just as soon spend that time building a simple gantry and bolting rails and ballscrews to a plate, and getting true rigidity and accuracy while I'm at it.

The general plan for the machine is to do a fixed, bolted steel gantry on a granite or iron surface plate, and use the Avid CNC nema 23 kit for electronics and motors. All that stuff is pretty straightforward, but I feel like the motion components is where I'm going to get lost in the weeds and lose days or weeks of evenings and weekends getting things right. I would love to sip this build like a fine wine, but time spent that way is time not doing the actual thing I'm trying to do!

Also, definitely tell me if I'm crazy

Cheers,

Brian

Similar Threads:

• Need Help!- Most efficient strategy for milling many small features?
• Build a linear stepper motor..
• Most efficient electric motor?
• Problem- First Build stepper motor problems
• Sherline Mill Machine + CNC Motor and Mounts

[machinedude]

i have always made my own motor mounts. when it comes to the ball screw drives you want it accurate for alignment. the less run out you have on your screw the better off your going to be. these machines don't get tossed together quickly and do what your wanting to do. so it boils down to how accurate do you really need? your talking about molds so your going to get into 3D tool paths. your machine size is small so it's easier to get things rigid and accurate on something that size but it's also going to be light weight so cutting aluminum could be a problem if your trying to cut any amount of material off your molds. carbon fiber is not so bad but don't be surprises if HSS gets dull on you fast. carbon and graphite is hard on tooling and carbide tools work better and are more cost effective.

[maxspongebob]

In designing your own system you will usually wind up fabricating several parts. I would recommend that you use a CAD package, Fusion 360 and Inventor are what I use, to build a model of your system before you start buying hardware. Build a virtual model and then you will be able to figure out what parts you need to fabricate and which one's you can purchase.

How you mount the motors is most often dictated by how you mount the ballscrews. The motor has to be concentric with the screw and also provide enough space for the coupler that you will use. When I am designing a system I usually integrate both the screw bearing mount with the motor mount. Then I use a simple NEMA spacer to place the motor in the correct position to accomodate the coupler.

If you build your own, you can make it just right. You can see below my latest creation. This is the gantry end (X Axis) motor mount. The screw bearing mount is what the motor attaches to via a spacer. The mount also has grease galleries included to ease lubrication. Also I like using flanged bearings to simplify fabrication and preloading.

[catahoula]

Thanks guys. That all makes sense. I do like the motors mounted directly to the ballscrew end bearing mount. I know there will be some fabrication, just trying to make sure I know exactly how much in order to minimize surprises.

On that note- how do you guys feel about rigid vs flexible motor couplings? Are flexible couplings just a crutch for poor design, or are they a good idea in general to remove the possibility of radial loads on the motor and bearings? I was leaning toward using the spiral type steel couplers for good measure.

I left a lot of details out of there- I do all my part modeling in Fusion already so I will definitely be modeling the machine completely before ordering anything. I've been using 3d printing, both FDM in house and outsourced SLS, epoxy casting, and some outsourced machining and waterjet cutting so far, but I've realized that to make any money I need to do volume, which means a lot of molds (fortunately they are small, at about 12" x 8"). Between three stages of molding and some fixturing, each part needs 6 to 7 different mold halves, so when you multiply that by 5 or 10 for production, means that being able to mill them in house is going to save a lot. I would love to outsource all moldmaking, but that adds up to like 40k sunk just into molds to get rolling.

3D toolpaths are definitely what I'm after, so I'm focusing on rigidity and matching the kinematics of each axis. Hence moving bed/fixed gantry and equal lead and motor size across all axes. And cutting carbon fiber is absolutely a different animal; I will be using mist or possibly flood coolant to deal with the dust and improve tool life, and have talked to Onsrud directly for help choosing appropriate cutters, which pretty much means special geometry and diamond coatings.

I also appreciate the encouragement. I guess that's the nature of the compromises involved in machines- each application is a little different. A $40,000 machine would definitely solve the problem, but so far it looks like I can do this with high precision and not too much fussing for about $4200. I'm basing my design largely on Wade O's router build (https://www.wadeodesign.com/design-details.html), but scaled down to a 24"x36"x4" granite Grade B toolroom surface plate and an 8" x 6" x 3/8" rectangular steel tube for the gantry. Cast iron would be easier for drilling, but granite is both cheaper/easier to come by and easier for maintenance with the straight water coolant I plan to use. I'm guessing the machine should weigh about 600-700 lbs, maybe 800 with the stand and enclosure, but with relatively lightweight moving components, so I think that'll be pretty dang solid for such a small work envelope and yet compatible with the 420 oz-in NEMA 23 motors.

[maxspongebob]

I use spyder couplers. The spiral couplers have too much spring to them.

https://www.cnczone.com/forums/bench...ml#post2122768

What do you need for Z travel?

You need the machine I just built. Maybe. It is more mill than router, but the x,y,z travels are 12x12x5. So it is not as large as you were looking at but very sturdy.
https://www.cnczone.com/forums/uncat...-software.html

[peteeng]

Hi - Several companies supply motor mounts that also have the support bearing mount integrated with it. They run at $25USD or a bit more from asia and about $70AUD local. So its not worth machining something if the std motor mount works for you. SYK is a well known brand. I've attached a page from their catalogue, but other brands are out there. Peter

https://www.syk.tw/

[machinedude]

that kind of design is probably your best bet at keeping this solid. only down side is the longer rails needed on the moving table end of it. i personally would not skimp on the table thickness if it were me. so finding a good balance between table thickness and bearing block spacing is going to be a key area on that kind of machine. depending on what you end up with for a weight on the z axis is probably what i would focus on for a motor to lift the weight. your ball screw pitch will also factor into things as well. stepper motors are not great for high speed tool paths at a high feed rate so some time spent doing some drive calculations will probably save you some headaches. a direct drive with a stepper motor with a 5 mm lead is probably going to fall in the 100 to 200 IMP traverse range. study the torque speed curve charts for the motors you have in mind that should give you a ballpark idea of where you will land on that side of things. screw whip is not as critical on the smaller machines so with steppers and a direct drive you should probably be ok there. the thing to remember with stepper motors is as you spin them faster to get your higher feed rates you loose a lot of the rated torque on the motor.

[catahoula]

maxsponge- the spyder couplers look good.

Pete- Awesome. I also just circled my way back to Automation Overstock and noticed that they also provide motor mounts for their house line of chinese ballscrews. That vendor seems like a good way to go, still much cheaper than Hiwin but not such a gamble as Amazon/Ebay.

machinedude - As far as table thickness, are you thinking that 4" is on the thin side? I'm a little nervous about drilling that many holes through 6" of granite, honestly. Even 4 is going to be a pain. I am definitely not an engineer but 4" seems ok for 36" length compared to other machines out there.

Regarding motors and lead screws- I am new at this and not trying to reinvent the wheel, so was looking at the Probotix Unity2 controller and noticing how they recommend their 550 oz-in nema 23 steppers for benchtop mill conversions with dovetail ways. So I figured with a linear rail application and a lot less moving weight, the Avid nema 23 steppers at 420 oz-in would be plenty up to the task. Also, I was initially figuring 5mm lead ballscrews to maximize resolution and driving torque with the relatively small motors. The steppers that Probotix and Avid are spec'ing all have torque dropping off around 400 rpm, some more sharply than others.

I did a bunch of speed/feed calcs based on manufacturer specs of tools I'd like to use (1/8 and 3/16 ball end mills for finishing, 1/4" end mills for roughing, 1/4" composite bits for trimming), and if I can comfortably hit 150 IPM that should suffice. Maybe I should consider 10mm lead on the ballscrews to make sure the motors stay within the torque curve? At 5mm lead, 400rpm is ~80 IPM, at 10mm it's ~160 IPM. I guess the main question is is a 400-500 oz-in stepper motor strong enough to push a 10mm lead ballscrew through aluminum at 150IPM? I'm focused on the Avid and Probotix options because I would like to have an all-in-one electronics kit with a little company support behind it. Obviously the Avid nema 34 option exists but it's another $1000. It's not going to be a full-time production machine, so going a bit slow isn't an issue as long as it isn't so slow as to get rubbing. The composite trimming cutters are all ok with fairly high rpm and slow feeds.

This is definitely constituting thread drift but you guys have been super helpful. Awesome forum!

[peteeng]

Hi Catahoula - 1) I think the urethane element couplers are too compliant for accuracy on a mill. Look at disc couplers 2) For a mill power is always preferable to speed so 5mm lead is the candidate 3) check out Buildbotics controller. All in one many many features and it works out of the box. Peter

Tell us more about the granite stuff... and why drill holes all the way through?

https://buildbotics.com/

[machinedude]

machinedude - As far as table thickness, are you thinking that 4" is on the thin side? I'm a little nervous about drilling that many holes through 6" of granite, honestly. Even 4 is going to be a pain. I am definitely not an engineer but 4" seems ok for 36" length compared to other machines out there.

the surface plate you plan to attach the linear rails to is not what i was referring to. the table top that mounts to the linear rails and sits on top of the bearings on those rails is what i would not skimp on. you will run in to deflection if the spacing on the bearings and work table thickness are not right.

[catahoula]

The Buildbotix control unit is really interesting, and leaves more room in the budget to go with NEMA 34 motors. I'm a little surprised the drivers are capped at 36V, as it seems stepper motors across the board perform much better at 48V or 72V. I can't find a torque chart for the 1200 oz-in nema 34's they recommend, but the 800 oz-in recommendation at 32V actually has the same torque as the Avid nema 23 420 oz-in at 42V when rpm's get much over 600 or so.
.
I am more comfortable sticking with 5mm lead. The critical purpose of the machine is accurate milling of 3d organic contours, so resolution and power is key. If I get motors with decent driving torque at 7-900 rpm, it solves the feedrate issue completely.

As far the granite surface plate, my aim is to use it as the base for the bed rails. Maybe it's just from having spent way too long without actual machining equipment in my shop, but having a known flat, stable surface to build from is deeply appealing. And more corrosion resistant for water mist than going with a cast iron or steel surface plate. My thought is to drill all the way through and use through-bolted connections for gantry, linear rail, and ballscrew mounts. Just seems a bit more straightforward, and if it doesn't go well, the pilot holes are already drilled to make larger pockets and epoxy in some stainless or aluminum inserts to drill and tap.

I have to say that while the coronavirus situation is generally unfortunate, it is good for making some headway on projects like these.

[catahoula]

machinedude- gotcha. I was thinking of going with 0.75" aluminum plate for that very reason

[machinedude]

machinedude- gotcha. I was thinking of going with 0.75" aluminum plate for that very reason

.750 would probably be a min. thickness. it just depends on how big the work table has to be for your needs? it's just an area of the machine that requires some planning since your getting into higher cutting force needs for the target materials you want to work with. i have another build i started and have the X axis built for a VMC style build and i used 1.5 thick square for the skeleton and have a 1.25 thick steel top that build and its rock solid. i think the travel was around 29" for that axis. the table is 12" x 30"

[catahoula]

Wow that looks VMC worthy for sure. Just out of curiosity, what's your ballscrew lead and motor size for that machine?

My minimum size worktable is 14x18 with 5" Z, or I guess I could do a 9x28, but thats kind of a weird shape. I'm shooting for about 16x22 table and 8" Z for a little flexibility. Trying to resist the temptation to just marginally increase all the sizes on this machine though. If the business goes well, I'm going to need a regular 4x8' router also, for high volume fabric cutting and part trimming, so I'd rather keep this machine fairly specialized to mold making.

[machinedude]

i have 750 watt DMM servo's for that build. the rails and screw are just your typical cheap import stuff. the rails are 20mm and the screws are just a 16 mm x 5 mm lead nothing fancy. the size of the rails you use depend on how well you support them. i see all kinds of build where people use huge rails but they hang half way off the machine with no support under them?

one other area people over look until they start making chips is covering the ways and ball screw. i started making way covers on that build and still need to finish the other side up but you should have something for covers. here is another picture with the way covers on the one side.if you watch the Haas service videos on Youtube they will give you some insight on how they do them on the commercial machines.

[dazp1976]

Hi,

Compiling a BOM for my first CNC build. I've been able to answer most questions via google, but stepper motor mounts eludes me. Are most people machining their own mounts, as there's so much variability in ballscrew dia, mounting location, etc? Or are there some stock motor mounts that work with typical 1605 chinese ballscrews and I'm just not finding them?

Background: It's not that machining is a big problem, as I have access to a mill and a machinist friend, but this is a tool for a side business I'm trying to launch into my main gig, so I'm really focused on minimizing the amount of time futzing and fabricating. I need to make aluminum molds, so 16x20" x-y travels, good accuracy, rigidity for surface finish, and a high-speed spindle because I also need to do some carbon fiber trimming. So that rules out a benchtop vertical mill conversion, and while the Avid Benchtop Pro router probably has enough stiffness and resolution, instead of spending my time checking and re-machining the 80/20 extrusion so the frame is guaranteed square, I'd just as soon spend that time building a simple gantry and bolting rails and ballscrews to a plate, and getting true rigidity and accuracy while I'm at it.

The general plan for the machine is to do a fixed, bolted steel gantry on a granite or iron surface plate, and use the Avid CNC nema 23 kit for electronics and motors. All that stuff is pretty straightforward, but I feel like the motion components is where I'm going to get lost in the weeds and lose days or weeks of evenings and weekends getting things right. I would love to sip this build like a fine wine, but time spent that way is time not doing the actual thing I'm trying to do!

Also, definitely tell me if I'm crazy

Cheers,

Brian

I aslo made my own for the PM25 type machine. They're kind of similar style to the ones in the images in the earlier posts.
My bearing blocks differ slightly. Using 24mm thick material, milled 2*32mm holes 10.5mm deep either side with 27mm hole through the middle 3mm to leave a 2.5mm ledge for the bearings to seat on. These hold 2* AC bearings back to back of sizes 12*32*10mm for 1605DFU screws.
I was fortunate to have my dying Sieg X2 able to complete the job.

Y axis. I made it wider and added 4 extra fixing points tapping the casting.

X axis. Also added 3 more table fixing points.

Z axis

[vger]

Catahola,
I hear you about having a little extra time with the current pandemic situation. Let’s hope we all make it through this safely.

Talking about couplings, motor to ball screw, rigid is going to be best but a real pain to align properly. First thing is the coupling itself needing to have the through hole machined such that the hole (probably step sized for two different shaft sizes) has to be dead on concentric. Having worked for many years doing laser shaft alignment and industrial balancing taught me a few things. With the close quarters of a stepper motor/ball screw mount, conventional laser alignment or dial indicator alignment is going to be out of the question. With a good coupling in place and tight on the shafts, leaving the motor mount screws loose but in place and a tiny space between the motor flange and the mounts, put the assembly in a vertical orientation motor on top. Use a dial indicator setup clamped to the ball screw bearing mount and indicator at the tail end of the motor perpendicular to the axis of rotation, slowly rotate the ball screw by hand in one direction. Any deflection indicates either bent shaft (hopefully not) or non concentricity in the coupling. If that is good, loosen the coupling allowing the motor to contact the mounts and feeler gauge for any gap. If there is a gap, shim it as close as you can and then tighten the motor down.

If you decide on disk couplings, make sure they are torque rated much higher than the holding torque of the motor. I made that mistake on a build and had to order heavier duty couplings after a couple of weeks of running the machine.

Steve
Trapped on a small Caribbean island

PS. I like the granite surface plate idea!

[catahoula]

Steve- that does indeed sound like a pain! Good technique though. I think I'll start out with some type of flexible coupling to simplify the initial build and get the machine running, and then maybe go there later.

Machinedude- on the covers note, how much do I need to worry about corrosion on the linear rails? Part of the reason I don't want covers is to make regular lubrication of the rail carriages easy. With using water mist coolant and some type of enclosure there will be a lot of humidity, especially under the moving table.

dazp- nice integrated solution to the motor mounting.

Now it looks like I'm going to lose my day job as soon as next week for at least 2 months and possibly permanently, so I guess I need to take all this great information and look optimizing machine size and components to get the rigidity and basic performance at the very minimum cost. This will be either the perfect forced springboard into a new venture, or a complete disaster!