"CNC project" - Technique for creating the guide rails
Beau Schwabe
Posts: 6,566
Just Getting Started:
I have 'finally' figured out a way that I am happy with in the construction of a CNC router project.
In this post I will focus on the gantry (Future posts as time permits will focus on other portions of the CNC) ... The hardest part in my opinion to any CNC is getting the gantry right. I have figured out a hardware solution to the gantry that I am very pleased with that I can expand to any axis. I call this ... 'The poor man's CNC at the expense of precision from others' ... the reason is because truly I have had to do very little drilling. (<-- something I have had difficulty with in terms of gantry alignment in the past.... this design is a dream for starters.) The parts can be found very easily at most hardware stores and on E-bay. For example I picked up 32 3/8" I.D. roller skate bearings for $5 on E-bay ... there are many other deals out there you just need to look.
Attached is a video file of the Z-Axis Gantry manually moved, as well as still images and exploded views of the Gantry and Slider mechanism.
Note:
- The I.D. of the Bearings and the square metal stock with the holes are 3/8" inch.
- Notice the bolts threaded at 45 Deg to the square metal stock... These are 1/4" inch and the square metal stock must be drilled in order to make this work. The purpose of these are to provide an easy mechanical standoff that is perpendicular to the axis motion to connect to other equip. (since the square stock rides at 45 deg to lock the bearings ... this step translates the 45 Deg back to something easily manageable.)
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe
IC Layout Engineer
Parallax, Inc.
Post Edited (Beau Schwabe (Parallax)) : 2/15/2010 3:52:20 AM GMT
I have 'finally' figured out a way that I am happy with in the construction of a CNC router project.
In this post I will focus on the gantry (Future posts as time permits will focus on other portions of the CNC) ... The hardest part in my opinion to any CNC is getting the gantry right. I have figured out a hardware solution to the gantry that I am very pleased with that I can expand to any axis. I call this ... 'The poor man's CNC at the expense of precision from others' ... the reason is because truly I have had to do very little drilling. (<-- something I have had difficulty with in terms of gantry alignment in the past.... this design is a dream for starters.) The parts can be found very easily at most hardware stores and on E-bay. For example I picked up 32 3/8" I.D. roller skate bearings for $5 on E-bay ... there are many other deals out there you just need to look.
Attached is a video file of the Z-Axis Gantry manually moved, as well as still images and exploded views of the Gantry and Slider mechanism.
Note:
- The I.D. of the Bearings and the square metal stock with the holes are 3/8" inch.
- Notice the bolts threaded at 45 Deg to the square metal stock... These are 1/4" inch and the square metal stock must be drilled in order to make this work. The purpose of these are to provide an easy mechanical standoff that is perpendicular to the axis motion to connect to other equip. (since the square stock rides at 45 deg to lock the bearings ... this step translates the 45 Deg back to something easily manageable.)
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe
IC Layout Engineer
Parallax, Inc.
Post Edited (Beau Schwabe (Parallax)) : 2/15/2010 3:52:20 AM GMT
1024 x 768 - 182K
1024 x 768 - 175K
1024 x 768 - 135K
1024 x 768 - 177K
Comments
Ok, so how do you connect something round like a pipe to a bolt with very minimal contact and leave plenty of room for rollers, while at the same time allow easy adjustments to be made?
I discovered this quite serendipitously when trying to work out another solution, but this method works rather well. It turns out that the 3/8" nuts that I have just fit perfect INSIDE the copper pipe.
So what I did to the copper pipe was to drill 3/8" holes properly spaced to the square stock. Next was to slide a nut inside the pipe (using a pencil, or longer bolt, i.e. all thread) and align it with the drilled hole in the pipe. Finding the nut through the hole and threading into it was effortless.
So that's basically the trick, Once you have a thread bolt secured perpendicular to the pipe, it becomes very easy to make fine tune adjustments.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe
IC Layout Engineer
Parallax, Inc.
Post Edited (Beau Schwabe (Parallax)) : 1/30/2010 8:40:16 AM GMT
Full speed ahead, I gotta see this!
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
·"If you build it, they will come."
Thanks!
I have several 3ft lengths of straight copper tubing, I thought I would use (I do believe that they are red-hardened) .... If I see that it's posing a problem, I can always move this concept to just about any pipe. Other than oxidizing, I think it will see much worse abuse. The rail/roller connection should remain clean just through use alone.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe
IC Layout Engineer
Parallax, Inc.
Post Edited (Beau Schwabe (Parallax)) : 1/23/2010 6:56:41 AM GMT
On the nut & bolt combo on the ends (no bearings)·of the rolling rails (for attachment of equipment?), did you round off the shoulders on the nuts to keep them from interfering with the square tubing when tightened? I can see where the shoulder on the bolt helps to retain the bolt and keep it from turning.
Ingenious setup, BTW.
Duffer
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Any technology, sufficiently developed, is indistinguishable from magic.· A.C. Clark(RIP)
No rounding of the 1/4" inch bolt heads, they will tighten and form a natural "lock" against the square tubing.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe
IC Layout Engineer
Parallax, Inc.
There was a slight change to the Z-Gantry so that part of the frame would double up as a roller support for the Y-Gantry.
The adjustment bolts for the Z-Gantry were removed and replaced with 3 inch all-thread making adjustments easier.
One rail of the Y-Gantry is fixed and flush with the square hole stock, while the other has a total of six 3 inch all-thread
5/16th inch bolts for fine adjustments.
Note: - I mentioned earlier that the hole diameter of the square hole stock and roller bearings was 3/8ths of an inch as well as the bolts. That's all correct except for the bolts. - The bolts are 5/16ths of an inch... if the bolts were 3/8ths, I would have to tap each hole in the square hole stock. (not gonna happen) - All 45 Deg bolts are 1/4th of an inch, and a 1/4th drill bit must be used on the square hole stock in the direction you plan on using the bolt in order for it to properly fit. (there is a ver small amount of metal removed in this step)Next weekend I'll try to work on the X-Gantry, but for now the weekend is over and my attention
needs to be back on the Propeller II.
Enjoy!
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe
IC Layout Engineer
Parallax, Inc.
Post Edited (Beau Schwabe (Parallax)) : 1/26/2010 3:17:48 AM GMT
Here is the X,Y, and Z Gantry frame ... since a good sturdy·frame is one of the·most important parts of a CNC, the·motion control can·take a back seat for·now.
I was asked·how I·worked out the roller bearing distance so that I was able to get the clearance I needed between the copper pipe and the square stock so I have included an image to help with that.
Not much if anything has changed for the Z and Y axis since last weeks post.· The main difference here that the X axis has that the others do not is that it uses 6 roller bearings·on a side rather than the other axis that only use 4 roller bearings on a side.· Also, the X axis roller bearings run true to the axis rather than at 45 deg.· This and the extra two rollers on a·side were added for stability since the X axis in this design·bears·most of the weight.
Attached are images of the XYZ-Gantry frame mostly complete.· keep in mind the entire gantry structure is sitting on it's side locked into a vise at the moment.
Here is a You-Tube video that I took early this morning ... BTW) we ended up getting·6 inches of snow
http://www.youtube.com/watch?v=JCh9aaikIyw
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe
IC Layout Engineer
Parallax, Inc.
Post Edited (Beau Schwabe (Parallax)) : 1/30/2010 8:41:26 AM GMT
How do you keep the rectangular frame square?
Thanks... I call it mad, only because it's one of my crazy unfinished ideas.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe
IC Layout Engineer
Parallax, Inc.
that is briliant! I can't wait to see the finish.
Jim-
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Signature space for rent!
Send $1 to CannibalRobotics.com.
This is an excerpt from a post on I put up on·linear drive screws and mantaining XY perpendicular without allot of precision machining.
This goes back to·eliminating precise machining and eliminating trouble mantaining parallel on your X-Y table, here is a rig I used that works incredibly well. It's a throwback to the old drafting board days and was used to keep parallel bars parallel. I'm surpized it's not seen more as it's cheap and very reliable and simple. It uses a cable that snakes around the perimeter and crosses in the middle of the bar. It's ingenious in that the once it's anchored down and taught there is no way the bar can slip out of parallel without movement of the anchors.·I used it on an X-Y table I built. Using nylon drawer pulleys (1/4" bolt hole inside of bearing) and steel 'brake' cable from a bicycle. I built the whole thing for less than $20 and it uses one stepper motor and has incredible accuracy. I put some photos of it on the·thread too.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Signature space for rent!
Send $1 to CannibalRobotics.com.
CannibalRobotics,
It's funny you mention the "old drafting board days" I have one of those boards and considered using that design. The one thing that I might see as a potential problem is that if there is any unexpected tension or binding. The tension spring will take the slack, and end up torquing the overall alignment so that it's no longer true.
Attached are pictures of my drafting board... sorry for the artwork, my 7 year old (about 3 at the time) decided it would make a nice coloring board.
mpark,
"How do you keep the rectangular frame square?"
Here is a video explanation: www.youtube.com/watch?v=HVQ6cnwR25w
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe
IC Layout Engineer
Parallax, Inc.
Post Edited (Beau Schwabe (Parallax)) : 2/7/2010 10:39:57 PM GMT
Jim-
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Signature space for rent!
Send $1 to CannibalRobotics.com.
I shopped around on Ebay and found 32 of them for $5 ... these are close cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=360228723482&ssPageName=ADME:B:FSEL:US:1123
Edit:
this is the item, only I won it on a bid, still $19.95 isn't bad...
cgi.ebay.com/Lot-of-32-Inline-Skate-Rollerblade-Roller-Blade-Bearing_W0QQitemZ130361936834QQcmdZViewItemQQptZLH_DefaultDomain_0?hash=item1e5a2d4bc2
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe
IC Layout Engineer
Parallax, Inc.
Post Edited (Beau Schwabe (Parallax)) : 1/31/2010 1:51:13 AM GMT
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Signature space for rent!
Send $1 to CannibalRobotics.com.
Is the weight of the carriage and router going to hold the gantry down? I see side and top bearings to ride on the copper pipe, but no bearings on bottom. I don't think you need them if the system is heavy enough.
And what about the plan for motors? What do you envision for mounting motors?
Carry on. . . in between P2 layout of course! This DIY Poor Man's router will provide a good relief from the other project.
Ken Gracey
Parallax Inc.
Thanks... "This DIY Poor Man's router will provide a good relief from the other project." - actually it's a good workout for the arms ... <-- I do it the old fashioned way with a hack-saw and metal, after my cuts, it's a wire brush attached to a drill (works better than sandpaper, and does wonders after cutting all-thread so you can easily re-thread the nut). ...I have confined this project mostly to just weekends and during my lunch break, so yeah absolutely P2 takes priority.
For the X axis (remember the video shows the CNC on it's side locked in a vise) there are a total of 12 roller bearings that lock the x-axis in place. Two on top and two on bottom that sandwich the copper tube... the other two that are hard to see ride in the same orientation as a spinning top with reference to the ground.
I have three stepper motors, that have all of the coils individually brought out... this means I can configure them to many different coil drive methods. I have a few ideas for mounting, however the Z-axis will be the toughest but the others should be straight forward.
The next step might be to build a table or frame to eliminate the awkwardness of the vise ... the weight when removing/securing to the vise is getting to the point that it is awkward with one hand on the CNC and the other hand operating the vise.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe
IC Layout Engineer
Parallax, Inc.
Post Edited (Beau Schwabe (Parallax)) : 1/31/2010 2:23:56 PM GMT
Some progress this weekend ... www.youtube.com/user/ICPolyman#p/a/u/0/FUFkplcmDE0
I decided to build a base frame for the CNC because it was starting to get a little hard to manage being suspended in a vise. ...and besides, I needed the vise for other things that were CNC project related.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe
IC Layout Engineer
Parallax, Inc.
Post Edited (Beau Schwabe (Parallax)) : 2/7/2010 11:10:52 PM GMT
I have some steppers that I ordered some time ago on E-bay that were made in Holland, not that that makes any difference, I really don’t know but what I like about them is that all of the coil terminals are brought out separately which means that I can configure them for just about any kind of stepper drive I want.
I did some 'holding torque' tests tonight, but I would like to see how my results compare to other steppers.
I created a cross beam similar to a method used to measure torque here. The cross beam extends opposite directions from the motor shaft so that the weight of the beam itself doesn’t have any influence on the measurement. At 1 foot out, I slowly added weight until the stepper motor could no longer hold it’s position.
Holding torque results:
10.0g/ft @ 3.7V motors drawing .3A
17.5g/ft @11.1V motors drawing .9A
Note: batteries used in test were Li-Poly single and triple pack.
I figure at the 17.5g/ft that's 210g/in and with a rod that has 20 threads per inch that's about 4000g of force (or 9 pounds of force) to the X, Y and Z gantries ... that should be plenty, and I'll probably go with a dual coil energizing sequence which effectively doubles the amount holding torque. (18.5 pounds) ... I don't think I'll have a problem, but I just wanted to be sure. If not I'll need to find some alternative stepper motors soon.
Consider also, that the torque on the threaded rod will be even greater than 210g/in since the diameter of the threaded rod is 1/4 inch. (more like 850g/in)
Thanks
@Scope - Thanks!
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe
IC Layout Engineer
Parallax, Inc.
Post Edited (Beau Schwabe (Parallax)) : 2/15/2010 3:57:00 AM GMT
9 pounds of force would be fine for something like a PCB mill but I think you will be underpowered for wood routing etc.· As you probably know, the faster steppers rotate, the less torque available.· Acceleration also seems to be an area where steppers don't behave so nicely and servos perform better.
What are you planning to use for a spindle on this router?· Knowing what spindle will help to guide what force is needed on each axis.· If you are planning on a 2 HP router motor to cut wood, you would be better off in the range of 30 to 50 pounds of force.· For something small like a dremel tool, 9 pounds of force would be okay.
On the CNC machine and control I am working on at the moment, I use belt drives on X & Y axis coupled to nema 23 steppers (around 300 Oz/in torque).· The Z axis is driven with a ball screw which is .2" pitch and driven with the same size motor.· It has been a long time since I calculated or measured the force on this machine but know that it is more than adequate to damage itself :-(
Also, the required force is something that is hard to estimate based on cutting forces alone (what you plan to cut).· A mechanical system has the potential to have "Sticky spots" throughout its motion of which are hard to estimate or predict so you want to make sure you have enough surplus power to overcome those.· "sticky spots' can be caused by dirt on the guideway, mechanical binding at certain speeds, racking of axis due to uneven load caused my cutting forces, and so on.
BTW, I am submitting a CNC related project for the propeller contest.· I would post some videos and information about it but I am not sure if the contest rules allow for posting information about the project prior to the contest results.· If it is okay to do so, I will post some videos and other information.
Chris
The 9 pounds was a rough calculation at a radius of 1 inch ... realistically that radius is 1/4 inch so 36 pounds or so would be more on target. I'll do some real weight tests before this next weekend and see where I actually stand.
I would like to cut wood, plastic, and some light metal ... speed really isn't an issue with this 'first generation" CNC so I might use a metric threaded rod with a finer pitch to increase my drive force at the cost of speed for the moment.
As far as your submission to the contest and showing video and pictures of what you have done, I am ok with it since the deadline for any 'new' entries has been closed. You might however double check with any one of the judges (Lauren Davis, Jeff Martin, David Carrier, or Andy Lindsay) just to make sure your not going to violate any of their rules.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe
IC Layout Engineer
Parallax, Inc.
Not much progress this long weekend, I've been fighting a cold. (probably induced from working in the garage on the CNC
...The next big step is going to require cleaning/re-arranging a portion of the garage where the CNC can live. I have a location scoped out, but this means that the actual CNC building progress may halt for the time being.
I haven't forgotten, but I will try to take some real torque measurements of the motors I have.
A few people have asked how I constructed the rails. Aside from the information in this forum thread here is a video that briefly explains "What I Did" <-- For some reason that was a popular phrase that I used in the video without even realizing it until I played it back to myself. I must have said it at least a dozen times ... Ohh well
www.youtube.com/watch?v=dy083h_oC7Y
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe
IC Layout Engineer
Parallax, Inc.
Post Edited (Beau Schwabe (Parallax)) : 2/15/2010 3:59:00 AM GMT
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Quit buying all those fixed voltage regulators, and·get an Adjustable Power Supply·for your projects!· Includes an LED testing terminal!
(no longer new) SD Card Adapter·Now available!· Add extra memory to your next Propeller project with ease!
Great idea, THANKS!
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
"puff"...... Smile, there went another one.
Dragon
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Dragon