[ger21]

Do you have any idea what kind of step - over you are using to make that wooden ball screw ?

If he's only got 28 lines of g-code, I'm guessing it's a single spiral toolpath.
It's not done using a raster type toolpath, so there really is no stepover setting.

[TrickyCNC]

That is a really fun project, with obviously very nice results, and testing for all kinds of future work.

This is why I like it here in the "ZONE"

Anything goes.
New ideas are discussed, not dismissed. Unlike some other forums I have tried

[JerryBurks]

If he's only got 28 lines of g-code, I'm guessing it's a single spiral toolpath.
It's not done using a raster type toolpath, so there really is no stepover setting.

This is correct. For a single spiral screw path only one "G1" command is needed. This video shows the motions:

[ame=http://www.youtube.com/watch?v=5U9lJAgU1oE]Fun Project: Wooden Ballscrew and Ballnut - YouTube[/ame]

You can buy ceramic bearing balls. That would be classy, and probably work well even lube-free.

Stainless steel balls are also easy to buy but may still give some rust on the wood, and steel parts in a wooden ballscrew assembly just seems like cheating to me.

Yes that would be nice and steel is cheating but I got to be practical here. The wood balls are 3 cents each, steel balls 9 cents, nylon balls 22 cents, brass balls 28 cents but alumina ceramic balls are a whopping 11 Dollar each. Since I will need a total of about 120 balls for the entire project (2 ball nuts and 2 angular bearings) I guess my wife would frown on the ceramic parts.

It's funny, if anyone else posted a thread and said they were making a wooden ballscrew and ballnut everyone would tell them they are crazy and it will never work, it's a dumb idea and don't bother etc etc, but based on your obvious successes with this wooden router so far I don't think anyone doubts you can do it even for a second!

Roman, I believe you are in a way responsible for this project. Last year when I built the machine and posted about my wooden timing belt pulleys, you joked back what would be next like wooden linear bearings or ballscrews. I just can not pass on such a challenge

...........Anything goes......New ideas are discussed, not dismissed. Unlike some other forums I have tried

I agree, on most forums (not only CNC) people seem to be just loud and spew nonsense only to be blasted back. In comparison it is very reasonable here. Being a technical forum there is a tendency for many new participants to reinvent the wheel over and over and the feed back is a great help to avoid unnecessary mistakes. If I had found this earlier (and assuming I had heeded the advice) it could have saved me a bit of rework for choosing the wrong steppers, drives and power supplies. Oh well, all part of the learning experience. On the other hand I have seen other forums where the conservative and well known old timers suppressed or ridiculed any new idea. That is also a problem because once a while a new and valuable idea comes up.
That said, nobody seems to believe that air cooling my water cooled spindle is a good idea

any chance of a peek at the program code ?

This is the g-code for roughing, finishing, widening and chamfering the spiral ball channel (or do you mean the generator code)?
I still need to tweak the program so that the bigger corebox bit does not whack into the un-machined ends.
%
G90
G21
F3000
M3
G0 Z 55.0000
G0 X 0.0000 Y 0.0000 A 0.00
F 7000
G1 X 0.000 Y 0.000 Z 18.860 A 0.00
G1 X 593.650 Y 0.000 Z 19.305 A11218.58
G1 X 593.650 Y 0.000 Z 15.305 A11218.58
G1 X 0.000 Y 0.000 Z 14.860 A 0.00
G1 X 0.000 Y 0.000 Z 13.010 A 0.00
G1 X 593.650 Y 0.000 Z 13.455 A11218.58
G1 X 593.650 Y 0.000 Z 12.955 A11218.58
G1 X 0.000 Y 0.000 Z 12.510 A 0.00
G1 X 0.100 Y 0.000 Z 12.510 A 0.00
G1 X 593.750 Y 0.000 Z 12.955 A11218.58
G1 X 593.750 Y 0.000 Z 16.395 A11218.58
G1 X 596.111 Y 0.000 Z 16.396 A11218.58
G1 X 2.461 Y 0.000 Z 15.951 A 0.00
G1 X -2.361 Y 0.000 Z 15.948 A 0.00
G1 X 591.289 Y 0.000 Z 16.393 A11218.58
G0 Z 55.0000
G0 X 0.0000 Y-18.0000 A 0.00
M5
M2
%

[TrickyCNC]

Roman, I believe you are in a way responsible for this project. Last year when I built the machine and posted about my wooden timing belt pulleys, you joked back what would be next like wooden linear bearings or ballscrews. I just can not pass on such a challenge

HaHa ... I remember that !


I agree, on most forums (not only CNC) people seem to be just loud and spew nonsense only to be blasted back. In comparison it is very reasonable here. Being a technical forum there is a tendency for many new participants to reinvent the wheel over and over and the feed back is a great help to avoid unnecessary mistakes.


Yep, I like to reinvent the wheel , if only to see why some things don't work !


If I had found this earlier (and assuming I had heeded the advice) it could have saved me a bit of rework for choosing the wrong steppers, drives and power supplies. Oh well, all part of the learning experience. On the other hand I have seen other forums where the conservative and well known old timers suppressed or ridiculed any new idea. That is also a problem because once a while a new and valuable idea comes up.


That said, nobody seems to believe that air cooling my water cooled spindle is a good idea

I have a few old VW's ... so Aircooled is a good thing for me


This is the g-code for roughing, finishing, widening and chamfering the spiral ball channel (or do you mean the generator code)?

Yes, I meant the BASIC program. It's been about 15 yrs since I wrote any code, and still keep intending to write some for CNC use.

%

Rich

[JerryBurks]

I made another one today, this time from a birch rod and to the full length of my a-axis, that is 33 inches. The birch machines so much better than the softwood and it came out about as good as I need it.


Wooden lead screws are actually not that exotic anyway. People used them for centuries

[RomanLini]

Wow that's a big set of thumbscrews!

Wood and leather these days mean something different to their meanings centuries ago. Often they were impregnated with some really specialised resins and things, like shellacs and baked plant resisns (tree sap) and there was something of a lost art. Things like ancient Asian "leather armour" where the leather raw hide was impregnated and sets to something like hard semi-clear plastic, that had the strength and light weight of carbon fibre or kevlar composite (because the leather fibres are very strong and semi elastic). These days we think leather is that baby-soft chemically tanned thin crap they make girl's jackets from.

We always think things like kevlar composite armours are modern magic and in the old days people were stupid, but some of those wood/paper/leather/fabric technologies were fine tuned over hundreds of years to be very "high tech".

Anyway sorry for the waffling but my point is that you could impregnate that wooden screw with some good shellac etc and some finishing and have something smooth and strong, something impressive like an "old school" carbon fibre ballscrew.

[fatal-exception]

I think I would have used buttress threads on that press...

Paul

[BanduraMaker]

Very fun project!

Roman, I believe you are in a way responsible for this project. Last year when I built the machine and posted about my wooden timing belt pulleys, you joked back what would be next like wooden linear bearings or ballscrews. I just can not pass on such a challenge

This being the case, you kind of have to make the linear bearings and whole machine at this point

Are marbles sized well enough to use?

[JerryBurks]

This being the case, you kind of have to make the linear bearings and whole machine at this point

Nah, you are not going to make me admit to that. This is just a fun and demo project. :violin:

Are marbles sized well enough to use?

Unfortunately not, they are usually 14mm and some are 5/8" and mostly sold in mixed sizes. Actually they sell pretty cheap agate marbles (which would be nice) in boxes from 50 years back but I have no idea how accurate they are. I am pretty much bound to 1/2" exactly because of the available router bits.


I hit a minor stumbling block today. I was trying to cut the blanks for the ball nut components from 3/4" plywood, which must be very precise because they are to be flipped over for 2-side machining.
They were supposed to be 5"x5" but after cutting I found them to be exact in x-direction (+/-2/1000") and on average about 20/1000" over-sized in Y-direction. Bummer. The only explanation is that I lost the preload on my Y-axis lead screw bearings. Maybe I step back and do the planned upgrade to proper angular contact bearings before I proceed with the wooden ballscrew project. The tapered roller bearings that I use now on my machine were cheap but I guess you get what you pay for.

[harryn]

Any chance it could be that small deflection in the lower Y rail shown in your simulation ? I would think that would show up before the bearing preload.

[BanduraMaker]

Nah, you are not going to make me admit to that. This is just a fun and demo project. :violin:

Yea, I understand. It's probably too challenging for you. :stickpoke

[JerryBurks]

Any chance it could be that small deflection in the lower Y rail shown in your simulation ? I would think that would show up before the bearing preload.

Not really. I did similar cuts before and they were precise. If it was the case then it would mainly show in x-direction. Also, the simulation was vastly scaled and there is no way I could bend the 20mm steel rod and the thick plywood plate that much with an 1/8" shaft bit.

Another giveaway is that I cut the square first in 0.1" passes in climb direction and then did a full depth final pass in conventional direction. Doing the final pass I could not hear the bit biting on 3 sides but quite a bit on the fourth side.

That is why I suppose the bearing is a somewhat loose or maybe even the ballnut bolts. The tapered roller bearings (cheap Russian made trailer bearings) are very difficult to set properly. There is little tolerance between loose and binding. Fortunately I bought nice angular contact bearings a while ago, just did not get around yet to install.

[JerryBurks]

Since I had to check the y-axis anyway, I replaced the tapered roller bearings with angular contact ball bearings and replaced the big 1600 in-oz stepper (22mH) with a much smaller 465 in-oz stepper from Keling (1.6mH parallel). Also changed the current for the 203V driver from 3.5A to 6A to match the motor specs. What a difference!

With the old motor I had the axis stall out regularly at 400 ipm and could use it reliably up to maybe 200 ipm. With the new motor and some very preliminary tests I can get up to 1200 ipm when jogging, maybe even more without stalling. Not that I want to machine at this speed but this is sure proof that the low inductance motors perform much better on typical cnc routers. Next I will replace the steppers and bearings for the x-axis as well. Maybe I leave the z-axis alone because it would be a bear to change.



To change the bearings and motor I opened the dust curtain gizmo and found that after 8 months of use there was indeed no serious trace of dust inside. Looks like it works very well.

[Mauser]

It strikes me that that wooden ball screw would make a magnificent lifting mechanism for a "Rolling Ball Sculpture" (AKA Marble Coaster, Do Nothing Machine, etc.)

[RomanLini]

Wow! Even the parts of my machine that I actually CLEAN look dustier than the inside of your box.

[Miata2k]

This being the case, you kind of have to make the linear bearings and whole machine at this point

Not to derail this thread, but this comment has me thinking. I realize that you probably can't build a real wooden linear bearing, but could you build a wooden slide of some sort that could be used for a basic CNC machine?

I was wondering around youtube like I sometimes do, and found this video about a router lift.
[ame=http://www.youtube.com/watch?v=QfjeS5jolBs]Home-made router lift - YouTube[/ame]

It uses a 3/4 threaded rod to raise and lower a router into a table. Think upside down Z axis. Anyway, they are not using any type a metal for the slides. It just got me thinking that maybe using a harder wood, a good lubricant and a little better design, maybe a all wood cnc is possible.

I figure a Bridgeport mill only use a metal on metal junction for its slide, and they have been well used for.... well forever

Just thinking aloud here.

Keep up the great work on the machine, I'm still envious of your dust covers.

-C

p.s. You may want to start a new thread for your all wood (including ball screws and slides) router project.

[JerryBurks]

It strikes me that that wooden ball screw would make a magnificent lifting mechanism for a "Rolling Ball Sculpture" (AKA Marble Coaster, Do Nothing Machine, etc.)

While I like to watch the marble machines and find some of them amazing feats of ingenuity and craftsmanship, I am more a person with a utilitarian interest. Well, kind of....not everything I do or try makes sense

[JerryBurks]

Not to derail this thread, but this comment has me thinking. I realize that you probably can't build a real wooden linear bearing, but could you build a wooden slide of some sort that could be used for a basic CNC machine.....It uses a 3/4 threaded rod to raise and lower a router into a table. Think upside down Z axis. Anyway, they are not using any type a metal for the slides. It just got me thinking that maybe using a harder wood, a good lubricant and a little better design, maybe a all wood cnc is possible...........

P.s. You may want to start a new thread for your all wood (including ball screws and slides) router project.

Interesting idea. I see a problem that the sliders need some preload to compensate for the natural surface inaccuracies of the machined wood. However, unlike scraped or ground (and oiled!) metal machine ways this would create a lot of friction.

But, like the ballscrew there is nothing wrong with trying to use large metal or plastic balls recirculating in a routed groove. But I am getting ahead of me here. I bet the ballscrew will take a while and a few experiments to get to work, if at all. It may also work but not be good enough for a practical purpose.

If I get some progress I will start a new thread but for now it is a bit premature.

[JerryBurks]

Not much progress on the machine modifications. These week long business trips take a toll on the hobby stuff

But I played and tweaked the settings a bit and took a video of a random zig-zag tool path at 1000ipm max and 30in/sec^2 acceleration. Since the x-axis still has the old steppers most of the motion is for y-axis. To compensate for the shorter motor I had to dump my pretty aluminum couplers and replaced with a piece of 3/4" thick walled aluminum tube and cheap shaft collars. And....works just as well.

The machine still has a lot of oomph at that speed. It is impossible to stop the y-axis by hand. Really amazing what that little stepper can do running 1700 rpm, considering the y/z assembly with spindle is about 75 pounds. The motor does get warmer than the big stepper when running for longer time but not too hot to touch.
[ame=http://www.youtube.com/watch?v=hpZR1Pq2Yw4]NewStepper.avi - YouTube[/ame]

[Walky]

Mindblowing! This image made my day, can't wait to see the full assembly!