I have finally able to post my JGRO pics. It looks a little confused but it works.
The three pieces of threaded rod running side to side under the bed pull the sides together therefore not allowing the gantry to sprawl the seamless rod outwards. The rods are also supported from underneath.
Cutting cedar curtain rod brackets for a friend's woodwork shop at 1016mm/min or 40" min with a 4.6mm plunge, 3/16" with 5 plunges to complete the cut. Cut time was 3 min for each one.
The little machine is a JCKleinbauer hawk mill. I will try circuit board routing with this one.
Here is a picture underneath the bed. The bed is supported by square steel tubing parallel with the Y-axis. The "T" nuts are used along with bolts to hold down work. Thin carpet tape also works well. I will post pictures of numbers I have cut out later.
Some numbers and brackets I have cut out. The small inside pieces to be cut out within "8" and "6" have actually been milled out instead being cutout as you would avoid a small piece of wood flying out and doing some damage.
These images should serve as encouragement to anyone considering undertaking such a project. The obviously only just good enough construction (and that is by no means meant as an insult) compared to the fineness of the output clearly show the practicality of Doing It Yourself.
Thanks for the encouragement guys.
The machine is also accurate enough to engrave animal tags with the finest lines of the letters matching up spot on, I did one for my dog Rocky in aluminium after he somehow lost his original and I decided I was not paying another $18.00 (US $9.00) for a repalcement. I will post a picture of it if he lets me take the picture without licking the camera.
I forgot to add in the first post that I was using 1/4-20 Stainless steel allthread on the X and Z axis but was unable to find any 6 foot lenghts for the Y axis and therefore using standard allthread there. The stainless steel allthread is very clean and operates more smoothly than the standard coated version.
Read the G-code first to ensure you do not get any surprises. Some of them are different to others as I have not adjusted all the parameters. If the file has "fullprogram" as part of its name it will the code will indicate 5 plunges of 4.6 each, finishing at 23.0. The files values represent mm/min feed rates as well as mm units so make the relevant adjustments if you use inches.
You can use NC-Plot to check the toolpath. The demo gives 60 days of testing time.
Jason
Last edited by Jason Marsha; 10-13-2005 at 05:45 PM.
Reason: error
Thanks.
I built it over a period of months. If you have access to a woodworking shop it can go a lot faster. I had access to a friends shop occasionally which helped a lot. To build it again I would say about 20 hours minus torsion box. Tapping MDF takes up quite some time. A drill press is critical as it almost impossible to dill a hole straight with a hand drill.
When I built my jgro I got frustrated with all the tapping. I finally chucked up the tap in my hand drill and made quick work of threading. Nice to see you've supported the rails.
Zoltan,
The supports are not perfect as I drew them with my rather poor CAD skills. The one with the 5 cutouts although not perfect are copies of each other.
Lionclaw,
Did you have a problem with stripping the threads while using the drill?
The files were cut with multiple passes for 22mm (7/8 inch) thick material. In some cases I used 5 passes at a depth of 4.6mm, leading to the following: 4.6mm, 9.2mm, 13.8mm, 18.4mm, 23mm. The ellipse for the open sign in the first post of this thread was cut out in 3 passes of 7.4mm each.
The maximum feed rate I have used is 1016 mm/min (40"/min ).
The bit used was a Whiteside (RU2100) 1/4" upcut spiral with 1" of cutting length, and 2" overall length.
When the bits are new you can have a fast feed rate, as these bits just run through the wood leaving clean surfaces and fine wood chips behind.
Upcut bits will leave a little fuzz on the top of the cut which can be easily cleaned off with a fine sandpaper, however the bottom of the cut is clean.
Jason
Last edited by Jason Marsha; 10-20-2005 at 12:13 PM.
Reason: incorrect values
Thank you again. I downloaded your files and run them on simulator and it run in one pass of 23 mm for all cuts, no multiple passes. This is the reason I have asked. I will look to the code itself. Thank you. Very nice cuts.
Zoltan
Use a perspective or isometric view (X,Y,Z) to see multiple passes you will not see it in the X,Y view. Only the files with "Fullprogram" contained in the name will show multiple pass in a plotter. To have 3 passes ,copy the original code and paste it twice underneath. You will now have three identical pieces of code, Edit the Z plunge in the second and third set of code to reflect your depth of plunge per pass, with the third pass having your finish depth.
eg
Z5.0 (First set of code)
Z10.0 (Second set of code)
Z15.0 (Third set of code)
Note that the program code only has the Z axis above the work at the beginning and at the end of the program. You will edit the code that was copied and pasted to ensure this occurs.
You will also see in my "Fullprogram" G-code the feed rates have been adjusted for the plunge and then readjusted to reflect the original feed rate. The reason for this is that the machine is not as rigid as a metal one, therefore if I plunge too fast the entry point will have a slight gouge in the material face.