Chapter 2 - Tools
CNC - JGRO Router
This category will contain articles relating to the JGRO CNC Router project and upgrades.
The JGRO CNC Router - The Past, The Present and The Future ...
I started the JGRO project quite some time ago looking for a way to make better wood patterns for metal casting. In hindsight, it is rather funny, A guy makes a wood CNC machine to do his wood working because he is such a lousy wood worker. Seems like a catch 22 to me! Well, I can only work wood to a certain point and beyond that I just stay away from the big dogs and stay on the porch. When I tried to make patterns like the cross slide piece for the gingery lathe I went through four iterations before I made a pattern that I was happy with and then was not satisfied fully. Sand casting only requires a couple of degrees in taper to pull from the sand and more than that makes the part look funny to me. That was the main reason for the CNC machine. I could cut tapers, fillets, sprues and riser plates all at the same time and I could go on my merry way with the casting and other fun stuff.
I completed the initial build in 2006 if I remember correct and then never really used it. Life got in the way and I had no time to cast and other things. I was busy remodeling and building a shop and a myriad of other stuff. The machine was stuffed into the basement of my out building where it got water damaged. It sat there till last year when I dug it out wanting to make a couple of plaques for two of our pets we buried in the back yard. To my wife the cat and dog were like family. So I thought it would be a nice thing to do.
After replacing the upper and lower bed and the gantry bottom on the machine to get it running again, I started doing some cutting. My problems were rigidity or rather the lack of it. First it was the X-Axis pipes, they were fairly rigid, but lacked and real room for adjustment. I never could get the bearings to ride on them the way that I wanted them too. That led to another change using a design from buildyourcnc.com that utilized box slides rather than pipes. That was a world of improvement, but still I am plagued by lack of rigidity in the Y and Z axis. They flex if I run too fast and as a result I have to take slower cuts and more passes. Now this would not have been a problem if I were only occasionally cutting something for me. However, since folks have learned what the machine can produce, I get more and more requests for things made from it.
That brings me to now. I have decided that I need to build a new machine for two reasons. First, I want a machine that is rigid enough that I can take some light passes on aluminum and have the ability to cut wood at a decent speed. Second, I want a larger cut area. I effectively have about 14" x 24" to cut in. At the time I thought that would be just fine, but I would like to cut some doll houses and route designs into our kitchen cabinet doors and well as a few other larger projects. This time I thought I would look around to see what was out there as far as plans before I pulled the trigger. I knew that I wanted a machine that I could build myself, and I didn't want to have to design one. The two that I had in mind were the mechmate and the Kronos Robotics KRMx01 or KRMx02. I decided that I will build the Kronos Robotics KRMx01 with some modifications. I would like hardened steel rails for the slides and use the KRMx02 Z axis for further router support and rigidity. I have to thank Michael Simpson of Kronos Robotics for helping me to come to that solution. I have to give credit where credit is due. Michaels KRMx02 machine is a ROCK SOLID RIGID machine, but I want the added accuracy of a lead screw over the rack and pinion. Having said all that, one never really knows where a CNC build will take them. Watch the KRMx01 CNC pages for what I actually do and how I get there.
Now, in closing let me say some good things about the JGRO router. Then entire project was a pleasure to build. It made me scratch my head too many times to count trying to figure things out. I learned everything I know about Vector Graphics, Motor control, G-code, tool offsets and tool changing, CAD software, CAM software, EMC2 (now LinuxCNC) and more from this one project. The learning opportunities that are presented in a project like this will fill many hours of time and when comprehended and put together are a very fulfilling exercise. Would I do it again??? YOU BETCHA!!! If you just want to learn the basics of these systems and your willing to take the time and do it. I would say go for it. Joe, the guy who designed the JGRO Router and gave it to the rest of the world is forever in my debt. He has caused me to cuss, bleed, scratch my head, get out of my comfort zone more than anyone ever has and I never met or talked to the guy. Thanks Joe!
That pretty much wraps up what I have to say about this project. I will continue to posts any changes to the machine that I make and any projects that I cut with it if I can remember to take a picture. So you might want to occasionally look and see if there has been an update. When I finally put the machine to bed for the last time, i will be sure to let you all know. Till then, Happy Cutting, Casting, Mold Making and Solder Melting!!!
Projects made with the JGRO CNC Router
The JGRO CNC Router project has been entertaining and fun and a HUGE learning experience. But the fruit of the labor comes with what you get to make with it. This page shows some of the things that I have made with the router. Of course it isn't everything but a small group of things that I actually remembered to take pictures of. I hope you enjoy. Also, look forward to a final instalment of the JGRO Router project where I sum up what I have done, what its strengths and weaknesses are and where I plan to go from here.
A Christmas Plaque.
A coworker and friend of mine, Sarah, has a fine drawing hand. As a matter of fact, I am trying to get her to do some vector graphics to sell. She is just that good. Well Sarah thought it would be nice to take a quick pencil drawing and route it into wood as a gift to one of her family members. I told her I would do what I could. This project taught me a few things. First, if you plan to scan a drawing be sure to ink over it if it is done in pencil. Pencil lead just done not scan worth a crap for me at least. The scan was then converted to vectors but then the vectors had to be cleaned up. There were WAY too many control points and several of them had to be removed and the curves massaged into position. Then it was scaled quite large and the vectors offset to create a second pair so that I could route between them. The other option would have been to route them on the vector. But, you know me, I like to over complicate things. After the offset was made, the resulting vector was scaled to fit the 11-1/4" x 16" piece of clear pine, leaving and 1-12" border. Once the design was routed the only thing else done was the edge rounded over with a rounding bit. Now before we get to the photos, I have to make a disclaimer. Her drawn image was much nicer than my final vectors. Hindsight being 20/20 I would have had her ink them. Better would have been to have her draw the drawing in a program like Inkscape and just use the vectors from that. So hopefully, I can convince her to learn to draw using that tool or a similar one. Remember, you can click on an Image to see a larger version.
A Bi-Wing Airplane
After I completed the change over from the pipe rails to the box style of slide from BuildYourCNC.com, I thought I would take a break from doing upgrades to the machine and cut something to see how it was doing. So, over to cnczone.com I went looking through the forums at the different vectors folks had made available of 3D puzzles. Well I downloaded a couple of simple ones and decided I would cut out a Bi-Wing Airplane. The first one I cut was done in tempered hardboard and even though I had to clean a lot of fuzzies from the cut pieces and manipulate the vectors to get a better fitting puzzle, it turned out pretty good. A picture of that one can be seen on the router clamp upgrade section of the JGRO Updates section. Anyway, I took the puzzle to work to show it around and the next thing I knew, I had folks wanting one for themselves to give as Christmas presents. Well, I rescaled the vectors for 4.75 mm Lauan plywood and started cutting airplanes. Here are a couple views for you to look at.
A Majestic Horse
Once some of my coworkers found out I could make 3D puzzles, they started asking what else could I make. Well seeing an opportunity for the machine to start paying for itself, I set out to find vectors that have already been tested and were tried and true so to speak. That led me to a website called makecnc.com. This website sells a lot of CNC related project material, including 3D puzzles and an assortment of other goodies. So I bought the Majestic Horse vectors and set off to make some. Below are a couple of pictures of the first one I made.
A Howling Wolf
The Next puzzle to be done was the Mournful Wolf puzzle. I bought these vectors from makecnc.com. By this time I am starting to see some of the inherent weaknesses of the wood CNC Machine. There is a tiny amount of flex at the router and the Y and Z axis. Needless to say, I am starting to think even more about the future of this machine and CNC in general. Below are a couple of pictures of the first wolf puzzle that I made.
The Spinosaurus Dinosaur
The Spinosaurus Dinosaur was a fun project to do. This puzzle when cut from 4.75 mm Lauan, was about 30 inches long and 18 inches tall. I made it for a co-worker but I believe I will have to make one for myself as well. Below are a couple of pictures of the puzzle.
The Kangaroo was a special request from another co-worker. I made some changes to the machining on this puzzle compared to the previous ones. I noticed on the earlier puzzles that the outside of the board I was cutting looked better than the puzzle side of the board. There were fewer fuzzies to clean off and looked like a better cut. So this puzzle was cut using a conventional profile path instead of a climb path. Also, I adjusted the machine somewhat to make it a bit more rigid. The results were better than before so I knew it was a step in the right direction. Also, I went over the sheet with a palm sander before cutting the parts loose from the board. This save a LOT of hand sanding!!! :-) Below are a couple of pictures of the Kangaroo 3D Puzzle for you to look at.
The Shark Helicopter
Another one of my co-workers has a relative in the military that likes Helicopters. This puzzle was selected by her for that person. This one was purchased from makecnc.com as well. This puzzle is a little different from the ones I have made earlier in that the slots are a little longer than actually required to fit the piece into. So if the slot is a little loose there is a potential that you will have loose pieces that have a hard time staying put. For example the side tail pieces are only held in by the short tab on them. This puzzle has the most pieces to date, coming in at over 80.
The Train Engine
The train is a slotted puzzle that was requested by another co-worker for her grandson. This was a pleasant puzzle to make and put together although all the little cutouts was a pain to sand. Here are a couple of pictures for you to gander at.
The angelfish so far has been the most challenging project. So far it has the most pieces of any puzzle I have cut, but that isn't what made it so challenging. I found the vectors for this project on one of the forums at cnczone.com. If you want to know which one, email me and I will post it. The exact name of the thread escapes me. Something about 3D dinosaurs DXF files. Again, if it is important to you email me and I will post it. The problem I encountered with the vectors was twofold. First, some slots appear to be in the wrong location and second, there are missing slots. Particularly on the tail section where the upper and lower fins attach. If anyone wanted to do this project, I would suggest just buying the vectors from MakeCNC.com and be done with it, rather than fiddling with the vectors to make them right. Well, unless you like doing that sort of thing. I just don't have the time. Anyway, below are a couple images of the Angelfish that I made.
Letters for a Co-Worker
A co-worker wanted some 7 inch tall letters cut for her daughters room. I thnk they turned out pretty good! Anyway, you can click on the images below for a bigger version.
Below is the Pegasus. This project wasn't requested by anyone, but I think it would sell nonetheless. At least I hope so, or it will end up adorning my office or something or be a gift to someone unsuspecting. Anyway, I hope you like it.
The Pirate Ship
Redesigning the X-Axis slides
After I installed the new motors and was seeing how much performance compared to the Genicom stepper motors I salvaged from some old printers, I noticed that I had some jittering and jumping from the gantry at a speed of 29 i.p.m. The problem after a closer inspection was that all four bearings were not making contact on one side. At a higher speed, the gantry would rock and cause it to jitter at the top.
The problem was I didn't have enough room to adjust the pipe. It would run up against the bore of the front and back plate before I could get it adjusted into position. I could open these bores up but then would have to bore out the pipe adjusters as well. Then I started wondering if there was deflection in the pipe in the middle. I don't know how much the gantry weighs but it is pretty heavy. Well, I fought with it until I was thoroughly frustrated and decided I would look at something different. So it was off to the Internet and see if I could find something different.
I spent a ton of time on BuildYourCNC.com watching Patrick put together his machine, and watch all the videos of him putting together his v1.3 machine. Well his site offers the nc files free, so I downloaded them to take a look. They were a little problematic for me for a couple of reason. If my machine was running correct, it wasn't big enough to cut some of the parts, and if it were big enough, EMC2 seemed to choke on the g-code in the files. I din't really want to build his machine, I just wanted to use the I-Beam slide that he was using.
I set out to reverse engineer the nc files back to a CAD drawing so I could get some measurements. I used a G-code to Cad converter and LibreCAD to create the drawings. It was a good learning exercise for LibreCAD anyway. Armed with the knowledge I gleaned from his work I set out to build some slides for my now Frankensteined JGRO machine. If you are interested in the drawings for what I done, let me know and I will either post them to the site, or email them to you. They were drawn up using LibreCAD. LibreCAD is free CAD software that will run on Windows and Linux. (Not sure about Mac).
Here is a side view of the finished X-Axis slide. The I-Beam construction resists horizontal and vertical deflection. I did lose an inch and a half of travel on the x-axis because of the plates that bolt to the front and back plates, but I am no longer using the bottom support on the machine either.
This is a shot of the X-Axis slide from another angle. Here you can see the box that wraps around the I-beam and bolts to the gantry. The nice thing about this is that you have support for the weight of the gantry and the thrust from the cutting tool plunging into the material. Not to mention that the slide is much smoother than the black pipe I was using before on the original design.
Some final thoughts.
With the machine back together I set back to see how fast it would run before losing steps. I can rapid up to 32.5 inches per minute with no problems. Not as good as I had hoped but still better than what I used to get. My lead screws are 3/8" 16TPI rod. I suspect I would do a little better with 1/2" 13TPI rod or 1/2" 10 Acme rod. But that is an upgrade for another day. I want to get back to cutting, after all, I still have those plaques to make. My poor wife probably thinks I have forgotten about them.
If I do more to the machine, it will probably be build a new gantry and Z-Axis assembly using the I-Beam method of Patrick's. I will move the gantry sides to the outside of the box slides rather than the inside to get a few more inches of Y-Axis travel.
Thanks for visiting. I appreciate it. If you have questions or thought or even constructive criticism, please feel free to email me. :-) Happy cutting!
Adding a New Router Clamp
One of the downfalls of a wooden router clamp is that through constant use it slowly loosens up, wears down and starts to flex. Well, I had enough of that and needed to do something about it. Fortunately, I have a friend named John who is an excellent metal working sort of fellow. Well, at least excellent in my opinion, because if you ask him you may get a different answer. He is a rather modest individual!
John took a look at the problem and suggested that he could use some strap and angle and whip up a stout and usable router clamp. The images below show the clamp installed and I have to admit, it went a long way towards stiffening the router assembly up. Thanks John for making this most excellent clamp for me. I owe you one (more like 10 or so) for this one! You can click on the image for a larger view.
In order to make the clamp work with the JGRO machine I had to make a change or two. First, I drilled out the threads on the Z-Axis plate. They were starting to loosen up and I was afraid that I would strip them out. I guess wooden threads and super glue only gives you so much mileage. With these holes drilled through I inserted a 1-1/2", 1/4x20 bolt and used a nut and lock washer on the outside. But then I noticed that the bolt head would not clear the heads of the adjusting bolts for the Z-Axis pipes. The solution was to grind the heads of the bolts holding the clamp down to about half of their original height. This gave me enough clearance to run the Z-Axis up and down without interfering with the Z-Axis pipe adjuster bolts.
After adjusting the clamp and Z-Axis pipes so that the router was square and purpendicular to the bed and the rest of the machine, I thought I would try cutting something. I have been looking on CNC Zones forums and found several DXF files for 3D puzzles. Well, I took one of the simpler ones and created toolpaths for it. It wasn't too bad but the vectors need a little adjusting. The images below show the finished Bi-Wing airplane. I cut this from 1/8" Tempered Hardboard, also called masonite. I used an 1/8" upcut flat endmill at a feedrate of 20 i.p.m. This will be the first and last thing I hope to cut from this stuff. The cutting took about thirty minutes to finish, but the cleanup from that nasty fuzz it left on the edges took about an hour with the help from my two oldest boys.
If you want to try this little project out either email me and I will send you the DXF file or you can search the CNCZone forums for it.
New Motors and Electronics
The stepper motors I salvaged from the Genicom printers work OK, but I can only get about 25 inches per minute from them.· If I push them any harder I lose steps, which is a bad thing.· I would like to cut at faster feeds so I don't have to spend too long baby sitting the machine. So, I bit the bullet and bought new stepper motors and electronics from BuildYourCNC.com. Patrick was quick and reliable with the shipping.
Here is the stuff I ordered from Patrick at BuildYourCNC.com. Three new 425 oz/in stepper motors, three controllers, a power supply, breakout board and a USB cable. I don't have a clue why they sent the USB cable.
Here are a few pictures of the new motors installed. Here is the Z-Axis motor installed. The wiring is managed with the plastic cable manager stuff you can buy at the auto parts house and I used little brass hooks, sold at Walmart as coffee mug hangers to organize the wiring on the machine.
Here is another shot of the cable management along the X-Axis. I am still not real happy with it, but at least the wires are all bundled together and out of the way. If you look closely, you can just see the X-Axis motor peeking out in the back with the mess of wire and electronics.
Some concluding thoughts and other stuff.
I was expecting a pretty good speed gain from the new motors. Well, I started at the 25 i.p.m. that I was able to get from my old motors and that seemed to do alright, so I slowly bumped up the speed to see how fast I could run it before I started losing steps. Well, when I hit 29 i.p.m. I noticed that the gantry would jitter or jump back and forth at the top. That wasn't good. Looking at it closer, I noticed that all four bearings on one side of the machine wasn't making contact allowing the gantry to bounce a bit at the higher speed. I tried adjusting the pipe but had very little adjustment available due to the restriction of the pipe in the hole bored in the end plates and the pipe adjusters. After some thought and painful deliberation, I decided that I had to either replace the end pieces and bearing adjuster with larger bores or do something different. I ended up deciding that I would just do something different for the X-Axis. But that is the next installment.