Posted: 4/19/2010 5:19:06 PM EDT
|
I intended to keep updating my progress on my homemade lower, but I work too slow and it fell into the archives: http://www.ar15.com/archive/topic.html?b=8&f=25&t=384650
The Cliff's Notes version - I set about automating an old chicom 3-in-1 machine to learn about stepper motors and hopefully automate some roughing and decking work. The more I got into writing the control program and doing simple machine work, the more I realized it was capable of much more accuracy and complex cutting than I expected. I decided the ultimate challenge for me, the software, and the machine would be to write my own program (based on the Ray-Vin manual ) to turn a Spike's 0% lower forging into a working lower receiver. Today, I'm up to roughing out the mag well. I dreaded this part, but it's turning out OK. Ray-Vin starts with drilling 1/8" holes at the corners. 
Here's a snip of the code I wrote to drill these holes. This shows my coding philosophy for the project, complete with warts. I wrote several primitive modules like line, drill, comp, and a few others to cut circles, arcs, angled lines, etc. Then I can just add these into a Linux shell script. ( .BAT file for you Windoze folks ) Lines that start with # are comments where I show the math I used for the next move so I can remember why I did it later. Machine movements are program lines like ./line .409 wI used compass co-ordinates and directions rather than x/y/z co-ordinates because I wasn't consistent with use of x, y, z in all my modules and it was getting confusing, so now directional movement is expressed as map-reading directions. When I'm standing at the machine, North is the work moving away from me, West is the work moving to my left. Well... I said there were warts. ./line .409 w moves the work 0.409 inches to my left. ./comp sis backlash compensation. My software isn't smart enough to figure out when it needs to compensate. I have to keep track of where the machine is going and write it in myself. The movement modules don't have to be in a batch file. I can type them in one at a time. Linux allows me to save the history of what I've done lately, so I can perform a few operations on the fly, save them to a file and repeat the series of operations later. Instant program based on what I already typed once.
More warts! Using a jacobs chuck to hold a 6 inch end mill. One more lesson in why chicom tools suck. I bought 3 of these long endmills. The first one (In this photo) had an oversize shank and wouldn't fit in a 1/2 inch endmill holder. Later I chucked up another one that fit the holder just fine - but - it broke in half while I was sliding it in the holder... with my bare hand! The 3rd one has screwed up flutes on the end but I'm working on the sides so it was workable. I would have bought American cutters if the local shop had any. 
Finally, the mag well is mostly hogged out and ready to start finishing work. This is the 3rd cutter with the bad end flutes. I ain't never buying these again, I'll just plan ahead and buy American off the web. 
I've added a new goal along the way. There is plenty of info on the web on how to make your own lower from a 0% forging, Ray-Vin being a really nice manual. There are also plenty of cheapo 3-in-1 machines and small mills available, and a lot of people buy CNC kits to convert the machines the same way I did, with stepper motors. It would be simple to add a calibration module to my software so that somebody that has a mill with a stepper motor add-on could get the Ray-Vin manual, load up my softare from a live CD, re-calibrate my software, and recreate most of the milling and drilling operations. I looked around the web for anybody that had already done something like that and couldn't find anything. I was pretty surprised that I couldn't find anything already. It might sound like a pipe dream, but I wrote the software from scratch, and it's not limited to any particular G-code implementation... because it doesn't use G-code. Neither does it do everything. It is written for an unsophisticated import machine and requires intervention to change tools manually. The operator has to pick up an edge or manually locate home more often than a serious industrial CNC machine needs. What it does do is automate and semi-automate the more difficult machining operations of making an AR15 lower, enough that at least a few people who wouldn't tackle it now will be able to turn out their own homebuilt lower receiver. But I guess I'll see how my first and second lowers turn out before I announce what a great thing it is outside the OKHTF! |
|
So how long until it's done?
Will it cost less than a goverment toilet seat? What upper on it? Will it be better than 'forged'? Any failed previous attempts? I must admit.....while not finished...does seem to hold more promise than the average 'garage made' zip-gun. 
|
|
I'm liking the idea of a charity raffle, but I'm going to hold out for more than just an AR lower.
How long? Well, looking back at the dates on my first experiments adding motors, I've already been working on the CNC control part for a year, and started making the AR fixtures last May. I think after the mag well is done, it ought to go a lot faster and maybe I can be finishing up by the end of this summer. If you look at it as "It cost this much to make this one lower" it won't be cheap. If you look at as "It cost this much to convert a manual machine to CNC and make one lower" it starts being a pretty decent price. I have an 80% A1 upper, and that's what I would like to finish up and mate to this lower, but I'll put my Bushmaster A2 upper on it to start with. I started with a forged blank, so it ought to be as good as forged. :) ETA: It may look like its being machined from billet because of the flat sides, but the workpiece is being held between two clamping fixtures. I haven't failed so bad that I've had to throw one away, at least not yet, but I planned on it, and bought two forgings so this first one could be my learning experience. I've done a couple of ooglies to this one so far but nothing bad enough to turn it into a doorstop. I also partially finished an 80% lower after I got antsy that maybe I couldn't really make the tight tolerances needed for the hammer and trigger pin holes. I needed a little confidence that the FCG was going to work before I started on the mag well and THEN found out I had a doorstop after I had done all that work. I would have started with a zip gun if it was legal! At least I would be shooting it already. :) |