Big Bamboo - New Machine Project Started

I started my CNC adventure last March with an off-the-shelf CNC Shark Pro+ (from Nextwaveautomation/Rockler) and learned a lot. But I found quickly that this machine (while fun for light duty woodworking and carving) did not fit the bill for precise mechanical prototyping in plastic and aluminum due to lack of rigidity and control software features.

So, I decided to build my own machine with the focus on rigidity, also a bit bigger. The intent was to use the Shark Pro+ to cut the critical components of the new machine and to sell it when finished as to recover most of the cost.

Now, I started buying some stuff before I ran into this forum, so some may be sub-optimal (i.e. overkill) as I learned but at least I got some good design advice from Ger21 and others.

The design is as in the picture below, with a 36" x 31" x 8" motion range, 80/20 T-slot table and all structural parts made from 3-ply 3/4" Bamboo plywood. I mentioned that in another thread and there was some doubt about the suitability of bamboo but i did a lot of research and found it:
- has the same weight/stiffness ratio as aluminum (slightly more than Birch). The lower absolute modulus does however require larger cross-sections compared to aluminum.
- it is unfortunately also a bit more expensive than Birch
- is very nice to machine precisely without much fuzz
- is very hard, i.e. metal parts bolted to it won't sink in easily
- well, it also looks interesting..... and I can get high quality flat sheets locally

I just started the build. More info in following posts.

https://lh3.googleusercontent.com/-N...gBambooCNC.jpg

Here some pictures of my first build activities....

Table prepared with mounting holes for 80/20 profiles (type 1545)
https://lh6.googleusercontent.com/-f...0/P1020745.JPG

Started cutting parts:
https://lh6.googleusercontent.com/-T...2/P1020746.JPG

How to make most of an 8x4' sheet:
https://lh6.googleusercontent.com/-T...o/P1020755.JPG

Cart is ready (made from 2x4 lumber) and preparing the stringers for the table:
https://lh3.googleusercontent.com/--...2/P1020748.JPG

Bolting the rails to the 8/20 profiles:
https://lh4.googleusercontent.com/-g...2/P1020744.JPG

Bearing mount for x-axis thrust bearings (double tapered roller bearings):
https://lh5.googleusercontent.com/-E...0/P1020752.JPG

As I learned the NEMA34 steppers that I bought are seriously oversize for this kind of machine (1600 in-oz) but I decided to keep them anyway. I will run them a bit under nominal current and they will surely not overheat....
Another advantage is the massive 14mm motor shaft with substantial bearings which should allow me to run the z-axis without separate thrust bearings (I hope). Both x-axis steppers will be force-synchronized with a timing belt.

All lead screws are surplus Kuroda ground ballscrews 15mm dia x 15mm pitch.

Otherwise: 20mm supported rails mounted in opposing direction; the lower rail of the y-axis and the lower end of the z-axis will have 4 instead of 2 trucks to help with the higher forces.

The z-axis has the trucks fixed and the rails moving. Just came out easier that way.

how accurate is the setup with two x screws and a belt drive?
can it lead to racking?

If you can somehow add a panel to close in the gantry, it could then be a torsion box, and would be much stronger. With the spindle in the center of the gantry, the gantry looks like it might be susceptible to twisting.

Quote:

---

Originally Posted by technodream

how accurate is the setup with two x screws and a belt drive?
can it lead to racking?

---

Obviously I don't know yet ;-)

But the intent is in normal operation for the belt to pretty much idle along, that means I will not give it much tension and hope it will not have much detent effect (if that is what you mean with racking).

The purpose is more as a safety device in case of electrical malfunction. The worst case would be that one drive loses the pulse and is held firmly in place while the other one wants to go ahead full torque. In this case I suspect serious gantry damage that could be avoided with the belt. But maybe I am just paranoid and in a way, the belt is optional.

Quote:

---

Originally Posted by ger21

If you can somehow add a panel to close in the gantry, it could then be a torsion box, and would be much stronger. With the spindle in the center of the gantry, the gantry looks like it might be susceptible to twisting.

---

Thanks for the hint, I am probably going to do that.

Jerry,

It's great to see someone trying out a new material.

Although the NEMA 34 bearings might handle heavy radial loads, I'd be very tempted incorporate a thrust bearing to handle the axial loads on the Z.

Also, I'm trying to understand the timing belt. Just thinking out loud here.

If one motor misses a step (or thirty), it seems unlikely the other motor will be able to "right" the system. A moving motor will always have less torque than a stopped motor, so it should never win that battle.

In that case, maybe what happens is that the active motor being unable to budge the stopped motor will also lose steps and the two become synchronized again, but the error is not mechanically corrected. The job is probably ruined, but perhaps no damage to the machine has occurred. (Maybe my thinking is wrong about this.)

Anyway, since you've got those honking motors, why not drive both screws with a single motor and use the timing belt to prevent racking due to lost steps?

Steve

Quote:

---

Originally Posted by stevespo

Although the NEMA 34 bearings might handle heavy radial loads, I'd be very tempted incorporate a thrust bearing to handle the axial loads on the Z.

---

Bearings of that size are normally specified with 300-500 pounds static load radial rating and e.g. SKF recommends not to exceed 50% of that as axial load but there are 2 bearings in the motor. That would leave me with several hundred pounds of permissible axial load and I believe I am not going to exceed that with my 1/4" shaft end mills.

Quote:

---

A moving motor will always have less torque than a stopped motor, so it should never win that battle.

---

Well, depends whether the failing motor has no pulses (worst case stall solid) or if it has no power altogether (it may be dragged along). But that worst case scenario the active motor (when moving slow) may still try to push one gantry leg forward with 400 pound force (assuming 50% nominal current, 800 in-oz effective torque with the 15mm pitch screws at 80% efficiency). I believe I am building a pretty strong machine but I am afraid it may be a bit too much.

Quote:

---

Anyway, since you've got those honking motors, why not drive both screws with a single motor and use the timing belt to prevent racking due to lost steps?

Steve

---

Well, I could but then I depend on the belt to be backlash-free for accuracy and maintain it always well tensioned. And I would have a nice motor and drive sitting unused....

Ah, yes those load ratings should do nicely.

Your chances of losing steps with those monsters pushing things around seems pretty unlikely. A complete driver/motor failure is also pretty unlikely, but the belt should help avoid racking in either case. Makes sense.

Steve

Quote:

---

Originally Posted by JerryBurks

Obviously I don't know yet ;-)

But the intent is in normal operation for the belt to pretty much idle along, that means I will not give it much tension and hope it will not have much detent effect (if that is what you mean with racking).

The purpose is more as a safety device in case of electrical malfunction. The worst case would be that one drive loses the pulse and is held firmly in place while the other one wants to go ahead full torque. In this case I suspect serious gantry damage that could be avoided with the belt. But maybe I am just paranoid and in a way, the belt is optional.

---

Since you have these powerful steppers, have you considered just running one stepper for the x axis, sing the belt to drive the two screws?

Cool looking bamboo panels!

Quote:

---

Originally Posted by JerryBurks

Bearings of that size are normally specified with 300-500 pounds static load radial rating and e.g. SKF recommends not to exceed 50% of that as axial load but there are 2 bearings in the motor. That would leave me with several hundred pounds of permissible axial load and I believe I am not going to exceed that with my 1/4" shaft end mills.
...

---

That sounded a little extravagant to me, I quickly checked one size34 stepper motor specs in a Pacific Scientific motors PDF;
"max axial load 135kg, max radial load 49kg (radial load measured half way along the output shaft)".

I'm sure I have seen motors specced for MUCH less axial load, I think some motors have a springy wave style washer removing axial play, that was more common on older motors. You should check your motors datasheet!

Quote:

---

Originally Posted by JerryBurks

... (regarding joining 2 motors via timing pulleys)...
But that worst case scenario the active motor (when moving slow) may still try to push one gantry leg forward with 400 pound force (assuming 50% nominal current, 800 in-oz effective torque with the 15mm pitch screws at 80% efficiency). I believe I am building a pretty strong machine but I am afraid it may be a bit too much.

---

You should be fine with your well thought out design and overkill (ie the big motors and safety timing belt).

It's not really likely for one motor so skip steps, as the minimum distance it can skip is a full pole cycle which is 4 full steps or 1/50th of a rotation. That should not be possible as the timing belt won't allow 1/50th of a rotation difference between the motors. They should share the load quite well and back each other up long before any one motor can fail. It's more likely the 2 big motors will rip your wooden machine apart before they fail! ;)

Regarding the torsion flex Ger21 brought up, if you used 2 thicknesses on the top and bottom gantry plates that will increase its torsional strength in a big way for very little cost in materials or weight, and without needing much re-design.

Overall your design looks good, the only thing I would change is to move the gantry leadscrew down as far as you can (as close to the bottom rail as possible). Where you have it in the centre of the 2 bearings is pretty much the worst place, compounded too by the fact your rails are very far apart. :)

Quote:

---

Originally Posted by stevespo

A.... A complete driver/motor failure is also pretty unlikely....

---

Agreed. But I mentioned already I am a bit paranoid.:rolleyes:

Quote:

---

Originally Posted by RomanLini

That sounded a little extravagant to me, I quickly checked one size34 stepper motor specs in a Pacific Scientific motors PDF;
"max axial load 135kg, max radial load 49kg (radial load measured half way along the output shaft)".

I'm sure I have seen motors specced for MUCH less axial load, I think some motors have a springy wave style washer removing axial play, that was more common on older motors. You should check your motors datasheet!

---

Well, don't know. This motor/driver kit came from China (Longs-Motors) and the mechanical load ratings are not specified on their web site. I should probably ask them. But I can see the bearings (15mm shaft lower and 17mm shaft upper deep groove bearing, type 6202Z and 6203Z from NSK) and found the attached load specs. I did not disassemble one and therefore don't know if they are spring loaded or how the bearing holder is designed. In any case, the shafts look quite impressive.

I did some more research and understand better the issue of the floating bearings with a spring washer. The rating of the bearings don't really matter then. There is actually one supplier who boasts a captured front bearing with high axial capability but I would be surprised if my steppers had that.

Oh well, I will then bite the bullet and do it right, i.e. design in another thrust bearing.

Quote:

---

I did not disassemble one and therefore..............

---

Do not disassemble them, or they will be worthless. They become demagnetized when taken apart.

Quote:

---

Originally Posted by ger21

Do not disassemble them, or they will be worthless. They become demagnetized when taken apart.

---

:eek: O.K. thanks, did not know that.

Yes! Don't disassemble the motor!

I had a read last night after logging off the forum, I found a larger Pacific Scientific tech PDF showing all the features of their hybrid motors. Their motors are all rated for high axial loads as the primary bearing (front bearing) has a retaining ring so it is axially locked in the motor front casing. So the axial load rating is then the rating of the front bearing, which is the larger of the two bearings. The rear bearing still has the wave spring washer so it preloads the bearings, not that those tight bearings need much preload.

They also had charts showing the de-rating of the axial load specs based on speed, basically it's 130kg at zero speed, dropping to about half that at typical stepper speeds and maybe 1/3 that if you want real high speeds.

I have some interest in this as my small router/mill uses size 23 steppers with no other thrust bearings, solid couplings onto the leadscrews, and it has performed superbly for a year or so including cutting aluminium.

I have a feeling that all (or majority?) of the new "hybrid" motor packages use the front bearing lock ring, as that's a feature of the front bearing plate and I haven't seen any of the 60mm "new hybrid" size23 motors that have any axial shaft movement at all. On some of the older size23 steppers, especially the non-hybrid round ones, it is common to be able to push the shaft back against the spring washer with a few kg force. Those older type motors are rated for 6kg axial load on size23 and 10kg axial load on size 34, I checked a datasheet.

At this point I have made a thread in the Keling forum section, asking if they have axial load specs for their Keling stepper motors as they have a decent range of size 23 and 34 motors in the new larger hybrid package.

I'm keen to find out, because if they are rated for higher loads I can crank the accels and speeds up on my own machine and I have always run it on "girly" setting! :)

That is great info. I had also seen this for the Portescap steppers. They look like my Chinese ones but who knows what is inside.

But like yours, the CNC Shark that I am using right now has also 287 in-oz NEMA23 steppers directly coupled to all the lead screws and I have yet to hear about motor/bearing failures from users.

Somewhat slow progress due to varnish drying time and some small stuff but I got the table substructure pretty much finished now. Next step will be to attach the 80/20 profiles and the x-rails.

So far I got the table flat to 15-20 mils over the 36x53" surface (judging from my steel ruler). Not bad for a lumber construction but will probably need shimming of the 80/20 profiles once I am done (or flat milling of the waste plates)

https://lh4.googleusercontent.com/-a...M/P1020756.JPG