Ok, I have been so intrigued by Brainchild's DIY CNC posts that I have decided that I have to have one. However, I cannot convince myself, or more importantly, the wife, to spend $300-$400 on the micRo when I have no idea if I would even use it once built.

To that end, I have decided to try to build my own as cheaply as possible, while still making something that will function. I hope to learn quite a bit from this endeavor and see if a large scale CNC is in my future.

My goals for this project are to make a functional unit that has an X and Y travel of at least 6 inches, and a Z travel of 1-3 inches. I also want to be able to complete the whole thing in under $100. In order to achieve this financial target, I am going to use whatever I can find lying around the house; thus the FrankenCNC title.

To start, I did a lot of research on the internet over the past month to see about the actual controller that will interface the motors to my PC. I have come to the following decisions based on that research:

1. I will be using unipolar stepper motors (no bipolar steppers or servo motors).
2. I will design the driver to control the motors in full step mode only.
3. I will design the driver to work with specific motors, and not make a driver that can handle a variety of motors
4. The motors will be driven by applying the correct voltage to them (no over-voltage with PWM driving)

(I am going to use a 200W PC power supply that I have lying around to power the driver and motors.)

My first step to designing my controller was to pick some motors. I picked the 4018L-01 motors from Lin Engineering. They are NEMA Size 17 motors that have 30 oz-in of torque each. They run on 6V and use 0.8A of current per motor winding. They step 1.8 degrees per step (200 step/revolution). I will be driving them at only 5V (0.666A of current), so I will lose some speed with them. I have already received my 3 motors:
Click to view attachment

I have most of the logic figured out for my controller; I just have to design the parallel port connections; I am still trying to decide if I want limit switch support and maybe an enable signal for each axis. I will post the schematics once I have them finished.

Without in any way criticizing your idea; Good luck!

I guess the big issue comes down to: When you're done, and become addicted to CNC robotics, you'll have to throw all this junk away and start over with good stuff to get the performance you MUST HAVE...hehe...(ask me how I know!!! )

Thanks for the good luck wishes!

I realized long ago that this is a "throw-away" project. I had considered buying a 36V power supply, the HobbyCNC controller, etc., so that I could reuse some of those parts if/when I went to a bigger machine, but that made the overall investment cost soar (over $100 just for those 2 items). Since I am going to Frankenstein this one together, it may hint to the fact that I am a pack-rat that likes to reuse parts. So, when I get addicted to the CNC hobby, I can disassemble this project and reuse the parts for other hobby projects. And I am creating this from junk as more of a proof of concept than a life-long keeper tool. I am hoping that it will have enough accuracy to etch circuit boards, but I am not counting on it being able to cut aluminium or build anything too complex/3-D.

Besides, starting small and low cost is the best way to prove to the wife how cool CNC machining is. Building book shelves, cabinets, and a toy box are many things that she has asked for, but she doesn't yet buy into the $1500 initial cost of a CNC machine to help build them.

Sure. I'd just suggest buying bipolar motors rather than saving (~just a few bucks) on some surplus unipoles; the power is nearly doubled. But if you already have the unipolar motors, sure! To spend fresh money on them is another story....

Anyway, I'm not trying to "sell ya" or discourage you...technically it is better for my business if you "fail"; but I don't want that either! The bipolar motors will remain useful to you for some time, and if you can't make it work, they are still salable at least. (who'd buy used unipoles after all? ).

~Cheers

Power is nearly double, huh? Oh well. I at least am going to run the unipolars in full stepping mode, which means that I will have 2 of the 4 phases powered all the time. I think that is the way to go to get the most power out of a unipolar motor.

I also chose unipolar motors because of the cost of the controllers. I plan to make my own control circuitry to save money, and the bipolars require an H-Bridge chip that I could only find for about $9 each. That's $27 just for the bridges, not to mention all the other parts. The chips for a unipolar motor can be very basic and a lot cheaper.

Speaking of making my own controller, I have the design for a single motor controller pretty much complete. This design will use a computer power supply to power the board and motors. The +12V rail will be fed to a 7805 voltage regulator that outputs +5V to the circuitry on board. I plan on using the +5V directly from the computer power supply to power the motors because the +5V rails can drive more current. Plus, the circuitry on board is sensitive to voltage fluctuations and I don't want the motor phases turning on and off to impact the power to the circuitry. This means that my 6V motors will be a little underpowered, but I can live with that. Anyway, here is the schematic for one of the motors. Just multiply by 3 to get the circuitry for all 3 motors.
Click to view attachment
The X_DIR and X_STEP signals will come from the computer's parallel port.

One last thing to mention. I can't actually design a single circuit to drive all 3 motors due to the limitations of the software that I am using. Since I want to make this as cheaply as possible, I am using software that has a free, but limited use mode. Based on the limitations, I can build one circuit board to drive 2 motors, and then a second circuit board that has the parallel port interface and the driver circuitry for 1 motor. Then, I plan to use a single circuit board and etch it twice. The first etching, on the top half of the board, will be for the 2-motor circuit. Then I will perform the second etching on the bottom half of the board for the last motor circuit and the parallel port connection. I will then use a few jumper wires to connect the parallel port control signals to the top half of the board and I will end up with a single circuit board for my controller.

Hey, great work! It's great to see the design effort you are making.

FYI, "easy driver" is open source, and available for about $15. This little-chip-that-could can handle 30v and 750mA. It is also bipolar and self regulating requiring just one V source.

Now for the software; why aren't you using EMC2 again?

I think the limitations he's talking about are for his circuit design / PCB layout software. I assume he's using a free version of a commercial product that's limited to the number of components or size of the PCB.

Aren't there a wide variety of products offered in this area? Switching to a non-limited open source package would seem worth the effort when compared to having to create multiple boards. I've never used any of the open source PCB layout packages.

Can anyone recommend the best OS PCB layout packages for Pirin to look at?

Thanks. I have done a lot of research and hope this will be a solid approach.


Yeah, I saw those during my research (and recognized them when you posted the micRo motor kit). I guess it boils down to I knew absolutely nothing about stepper motors when I first started, and the unipolars seemed easier to create a controller for (I don't think I can solder surface mount chips, so I was looking for an all through-hole solution). I figured that I could make a controller cheaper than the $45 for 3 "easy drivers." Most likely, in my quest to make a really inexpensive solution, I became penny wise and pound foolish . Alas, I have already started down that road and am just going to push through to the finish.


Oh, I am planning on using EMC2 for the actual cutting, but I thought I had to design the schematic and layout in one tool, generate the g-code in another, and then import that g-code into EMC2 to actually route it? The software limitations I have run into are in the schematic/layout phase. In one program, they limit the layout to 4"x3". In the other program, they limit the schematic to 250 pins. Either program allows you to purchase a license to remove all limitations, but that would blow my $100 budget.

Yep, that's the issue. I tried to find an open source version, but my main requirement in the PCB layout arena is that the tool needs to have an auto-route feature. Routing by hand may be required, but I would love to have a tool do most of the grunt work. And, if I could find a single solution that could do the schematic, layout, and g-code generation that would be ideal. I am currently using the limited versions of Cadsoft Eagle Light Edition and Target 3001!.

If anyone has any suggestions, I am all ears. I don't mind "starting over" with a new tool, as the design is almost complete. Re-doing it in another tool doesn't require any redesign, just placing the existing design in the new tool.

Well, I have redesigned my controller many times, but have not revisited the software limitations in a while. Eagle's 4"x3" PCB layout limitation is still a problem, but my latest design may fit within the 250 pin limit of Target 3001! I am currently at 249 pins, but am missing the HOME signal (I have LIMIT signals, though). So, I may be able to fit this all on one board after all!

I think you meant www.linuxcnc.org

Indeed, thanks!

You may try:

MUCS PCB - University of Manchester PCB Software
A suite of PCB layout tools, from schematic capture, to auto-routing, to layout.
Authors: Doug Edwards, Alistair MacIntosh, Zahir Moosa, Fred Hoyle (WOW! I wonder if it's THE Fred Hoyle?)
Web Page: http://www.cs.man.ac.uk/amulet/pcb/index.html
Description: The main feature of this tool-set is it's stand-alone auto-router, which seems quite advanced and implements several algorithms. Documentation seems very good. File formats are well documented. This router is just asking to be integrated into PCB or gEDA...
Other MUCS tools include: Ncap, a textual schematic capture interface; Place, a graphical device placement tool; and various other utilities for manipulating and editing the design and generating GERBER or HPGL output.
Part Library: Yes

I'll be jumping around a bit in this construction, as I am trying to source many parts as cheaply as possible...

A while ago, I built some heavy-duty Dance Dance Revolution Pads for the PS2. In constructing those pads, I had to buy some radio-controlled airplane fuel line. I saved the left-overs and am using that now for this project.

I have my motors, which have a shaft outside diameter of 3/16". I then bought some 1/4" threaded rod (1/4 x 20). In order to match them up, I decided to use the radio-controlled airplane fuel line; which, coincidentally, has an outside diameter of 1/4". However, the inside diameter of the radio-controlled airplane fuel line is smaller than 3/16". This actually works out best because it will have to be stretched/worked onto the motor shaft, and once on, it doesn't move at all.

Finally, I bought some automotive fuel line that has an inside diameter of 1/4". The idea is to place the motor/airplane fuel line inside one side of the automotive fuel line, and then place the threaded shaft into the other end of the automotive fuel line. If need be, I can then zip tie the automotive fuel line to compress it to hold, but I don't think I will need that. I realize that this is far from ideal; I am curious to see what kind of backlash I actually get with this approach (automotive fuel line was < $2 and the threaded rod for all 3 axes was < $5).

Ok, here are both fuel lines. The automotive is black.
Click to view attachment
Here is the cross section of both fuel lines.
Click to view attachment
This shot shows that the model airplane fuel line's inside diameter is smaller than the 3/16" motor shaft diameter.
Click to view attachment
Finally, here is the model airplane fuel line piece stretched over the motor shaft
Click to view attachment

I have a Hobbycnc kit with a blown driver board (one of the driver chips exploded when I was testing it after I soldered it). Transformer, capacitor, rectifier, and 3 300 oz-in motors. I'd be willing to let it go pretty cheap. Let me know if you're interested.

Cool. Thanks for the offer. I'm going to wait and see how this turns out, but I will keep your offer in mind.

FYI, I checked their site yesterday and they sell replacement driver chips for like $15. I'd let everything go for ~$150

Continuing on with disassembling old stuff for this project... I needed some ball-bearings for my CNC machine. I looked at the local hardware store, and they want $15 a bearing!!!! So, I went rummaging and came up with some classic roller skates. Yes, that's right, roller skates.
Click to view attachment
I took off each wheel, and there were 2 bearings per wheel.
Click to view attachment
The bearings are a little rusted, but they still spin ok at low revolutions. I can always oil them later if need be.

I plan on using the bearings to guide my ways, so that my 2 ways don't need to be parallel. This saves a lot of time in the design. It is best to explain how this works by just building it and showing you. So here is how I have started.

I needed a way to mount the bearings to my platform, so I started by finding screws with a head large enough to fit over the shaft of the bearings. Unfortunately, the shaft is at least 1/4" inside diameter. I had some Truss head screws left over from my Dance Dance Revolution pad project that cover this well, but the threaded part of the screw was too thin. So, I got out the extra hobby airplane fuel line and added that to the screw shaft. It still wasn't thick enough, so I wrapped it in electrical tape. Then it fit inside the bearing perfectly.
Click to view attachment

Then, I took some left over aluminum square tubing from my projector's light box. I cut out each side to get aluminum strips.
Click to view attachment
Each strip had a hole drilled into it, and then each Truss screw with bearing was screwed into the strip. Finally, I took out my Dremel and ground off the extra length of the Truss screw so that it was flush with the back of the aluminum strip.
Click to view attachment
Unfortunately, this is as far as I have gotten. I hope to continue this more sooner than later. Getting free time for this project is proving quite difficult.

A commendable effort surely! Don't give up.

I had a steel lid left over from an electronics project box that I used (the box is plastic and comes with a plastic and steel lid). I also had a few 1" nuts left over from an arcade cabinet remake I did a while ago.

Using those items, I made a mounting mechanism that will "attach" the nuts to my moving CNC table surface. For sheer simplicity, my CNC machine will not us a gantry; instead, the rotary tool will only travel along the Z-Axis, while the table surface that the piece-to-cut is on will move in the X and Y-Axes.

I cut off tiny strips of the steel project lid and J-B Welded a 1/4x20, 1" nut to each strip. I can now mount the steel strips to my table surface and the nut will, subsequently, also be attached.

Click to view attachment
Click to view attachment

To make the base of my CNC machine, I used the top piece of a cabinet I bought. It has damage underneath, and the company shipped a replacement piece without taking this damaged piece back. So I added 2 aluminum channel rails, 16" long each, propped up on some hardwood strips. I had to prop them up in order to lift the X-Axis platform high enough for the motor.
Click to view attachment
I then took an 8"x10" piece of plexiglass that I bought at home depot, and mounted a single 8" rail to it. Then I mounted my aluminum strip/ball bearing unit to the plexiglass. Finally, I mounted the steel strip/1" bolt unit to the center of the plexiglass.
Click to view attachment

I then placed the plexiglass unit on top of the base (turned over from the photo above).
Click to view attachment
Hopefully you can see the following:

• The plexiglass just rests on top of the 2 16" rails, and slide quite well.
• The 8" rail, and the ball bearing strip sandwhich one of the 16" rails between them; this sets up the guide to keep the plexiglass from turning as it moves. This also allows me to not worry if my 2 16" rails are parallel, as the guide only follows one rail.
• I put the threaded rod through the 1" nut that is mounted in the middle of the plexiglass and then attached it to the stepper motor through the automotive fuel line.

The motor is not mounted to the base yet, and I have not made the circuit to control the motor, but the placement should give you an idea of how it will work. As the motor turns, the threaded rod rotates, causing the 1" nut to move, which carries the plexiglass platform. This will complete the X-Axis (?) portion of my movable table. I will mount the Y-Axis table and motor on top of this plexiglass platform.

With the 16" rails, and the 8" guide attached to the plexiglass, I will have 8 solid inches of travel for my X-Axis.

It looks like you are following in the footsteps of Tom McWire. His build is here.

http://www.instructables.com/id/Easy-to-Bu...Milling-Machine

It was actually this instructable that got me interested in going back and revisiting CNCzone after not being there for two years and eventually to this site. I love how low tech he is and yet still has a great proving ground test machine for very little money. I have thought many time of making this type machine to get my feet wet as well but only if I had printer motors lying around etc. and since I sold all of that I am just going to have to bit the billet (pun intended) and step up. Can't wait to see your take on the machine and how it turns out. He did some cool PCB boards and glass etching with his.

Salter

Yep, Tom's instructable is one inspiration for this build (I also liked this instructable). I am building my own motor driver board, though. And, of course, I am building it from as much scrap as possible. I am looking forward to being able to make circuit boards with this machine.

As for using printer motors, I ran into a problem. I bought a bunch of used laser printers, as a deal for ink jet printers fell through. Unfortunately, the laser printers I bought only had simple DC motors (no steppers). But, I was able to resell them for ~$15 profit, so not too bad. I did find a single ink jet printer to take apart, but the only stepper motor in their was VERY tiny and would not have worked for this.

The next step was to figure out a way to mount the motor to the wooden base. I had some steel corner brackets left over from my projector build, so I decided to drill mounting holes in the bracket and screw the motor to that.
Click to view attachment

I then screwed the corner bracket down to the wooden base to anchor the motor in place.
Click to view attachment

In order to finish the X-Axis, I had 2 more steps:

1) Mount the 16" guide rails that the Y-Axis will move along
2) Mount the Y-Axis motor

Mounting of the guide rails was pretty straight forward. The only issue here is that the X-Axis plate is only 8" long; so the 16" guide rails extend well passed the edge of the X-Axis plate. I also needed to start the Y-Axis plate out passed the gas-line tube coupler on the Y-Axis motor. Since this would put the Y-Axis plate starting at least 3-4" off the X-Axis, I ruled it out (for stability). To compensate, I used the remainder of the steel project plate (used for the 1" nuts earlier) to make a platform for the Y-Axis motor so that it could be mounted off of the X-Axis plate. Maybe a picture will help.
Click to view attachment

The plate on the right side will hold the Y-Axis motor. The Y-Axis plate will sit on top of the guide rails in the same manner that the X-Axis does (sandwhiching a guide rail between another rail and a ball bearing). My biggest concern at the moment is whether or not the X-Axis will tip when the Y-Axis is extended all the way to the left. I am going to see if the weight of the Y-Axis motor can compensate for it. If not, I will either need to add more weight to the metal plate, or figure out some sliding anchor for that side of the X-Axis plate.

The X-Axis is now finished. I ended up doing 2 more things before I called it done:
1) Mounted the Y-Axis motor to the metal plate.
2) I Moved the X-Axis motor closer to the 16" rail that the X-Axis is guided along. Since the X-Axis stage is only guided along one side, it acts as the pivot spot for the stage. The further away from the pivot point that I am, the more the stage wants to pivot. By moving the motor/threaded rod as close to the guide rail side as I can (moved up to the ball bearing), I reduce the desire of the stage to pivot.

Click to view attachment

In making the Y-Axis stage, I followed the same process as the X-Axis stage. Basically, I took an 8"x10" piece of plexiglass and mounted the following to it:
1) an 8" aluminum rail,
2) a ball bearing
3) a 1" 1/4"x20 nut

Again, I mounted the nut as close to the rail as the ball bearing would allow. Once placed on top of the 16" Y-Axis rails, I put the threaded rod through the nut and connected it to the motor.

Click to view attachment

And now the X-Y Stage is complete. I still have to mount the dremel tool above the Stage, in a manner that will allow the stage complete freedom of movement, but at the same time, allow the dremel tool to move up and down along the Z-Axis. In thinking this through, I decided that I am going to switch what I am calling my X-Axis and Y-Axis. The Y-Axis is now the bottom motor/stage, and the X-Axis is now the top one. The picture below shows why I did so. Note that the red circle in the lower left corner is where the Dremel will hover. In order to move the cutting tool in the Positive X direction, the top Stage will actually move to the picture's left. In order to move the cutting tool in the Positive Y direction, the top Stage will move to the picture's bottom.

Click to view attachment

EDIT: I just noticed that when I resized this picture to post it, the red circle got lost. The red circle is supposed to be where the 2 red lines intersect in the lower left corner of the top stage.

In my hand testing of each stage sliding along its guide rail, not only did I notice the pivot problem I mentioned earlier, but I noticed that my fix of tightening the "squeeze" between the 8" rail and ball bearing upon the guide rail caused it catch periodically. I assumed the periodic catching was the friction between the 2 aluminum channels. To address this, I sprayed both pieces of aluminum with silicone spray. The result is amazing. Everything slides very well now.

Click to view attachment
Note that I did not use grease. I thought about it, assuming that it would lubricate quite well, but I am afraid of saw dust getting stuck to the grease and gumming it up.

After "finishing" my X and Y stages, I decided that my gasoline tube coupler that connects the motor to the threaded rod wasn't a strong enough hold for my liking. So, I decided to anchor it a bit stronger. I started out by drilling a hole through the gasoline tube and the threaded rod:
Click to view attachment

Once that was done, I took a coat hanger and cut a strip from it. I had to file it down to fit snugly into to hole I drilled:
Click to view attachment

Once the hanger fit through the hole, I bent it around the gasoline tube so that it couldn't come out:
Click to view attachment

Now I felt more comfortable about the threaded rod turning when the motor shaft did. I am sure that I would have lost some steps if the motor encountered any form of resistance. For the motor shaft side of the gasoline tube coupler, I plan on using Zip-ties to squeeze that. Remember that I used airplane fuel line over the motor shaft first, before inserting it inside the gasoline tube. It is a really tight squeeze as is, but I figure the Zip-ties can't hurt.

For my Z-Axis stage, I decided to use 1/2" plywood, as I ran out of plexiglass. I originally cut the side of an old PC case, but the metal was too flexible.

The Z-Axis stage was going to be made the same way as the X and Y, but I actually ran out of aluminum 'U' channel. At the time I made the Z-Axis stage, I only had 16" of aluminum channel left. Since I needed two 8" strips for the stage to slide along, I replaced the normal 8" aluminum channel (that mounts to the stage) with 2 left-over ball-bearings. Here is what the back side of the Z-Axis stage looks like:
Click to view attachment

In this picture, the 8" aluminum channel guide will run left-to-right, sandwiched between the 2 bearings on the bottom and the top bearing. You can also see the 1" nut mounted to the stage for the threaded rod. Lastly, you see the 2 bolt heads that are used to hold the dremel tool mount.

Speaking of which, here is the dremel tool mount. I just used a section of 2x4, and drilled a hole in it. Once the hole was drilled, I cut a strip of wood out of the middle so that I could sandwich the dremel in what's left.
Click to view attachment

Seeing as I don't have a drill press, and the hole saw isn't deep enough to go all the way through in one pass, I had to drill the dremel support hole from one side, and then flip the 2x4 over and drill from the other side. Needless to say, I can't drill straight. This is easily seen in the picture below, where the Dremel is secure in it's mount, but it is not perpendicular to the mount. To compensate, I mounted the dremel support piece on such an angle as to make the Dremel square with the threaded rod, and parallel to the path of travel.
Click to view attachment

The Z-Axis stage is sitting on top of the board that I will cut out and use as the Z-Axis backing, but I have not mounted it nor cut it to size (though you can see my pencil marks for my cuts).

I was able to get the Z-Axis backing cut and the guide rails mounted:
Click to view attachment

Here is what it looks like with the Z-Axis stage placed on top of the Z-Axis backing:
Click to view attachment

It slides pretty well.

Of course, this is all fine and dandy lying on the floor. But what happens when I lift the whole thing up so that it can slide up and down?

Well, besides the whole Z-Axis stage sliding towards the floor (will be fixed when the motor is mounted), I run the risk of the Z-Axis stage tipping away from the base; kind of a head-over-heals tumble. To prevent this, I need a way to hold the Z-Axis stage to the base. My solution is to sandwich the stage between 2 bearings and the base. Here is a shot of the 2 bearing arms that I made:
Click to view attachment

These are made the same way as the bearing mounts for the X and Y-Stage; I mounted the bearings on the screw/airplane fuel line/electrical tape shaft, screwed it into a steel arm, and then JB welded the screw to the steel arm so that it cannot come off. I haven't yet mounted these arms, but the base of the arms will have a hole drilled into it and then it will be screwed into the Z-Axis base. These arms are long enough to place the bearing above the Z-Axis stage so that it can be sandwiched in between.

I was able to mount the Bearing clamps to hold the Z-Axis stage to its base.

Side View:
Click to view attachment

Top View:
Click to view attachment