realistic CNC purchase?

[bestrada11]

Is it realistic to think/hope to get into the 3 axis CNC milling operation for under 5k? I would not be running production runs just onesy twosy prototype pieces. the biggest thing i would want to run would be something like a reproduction of like valve cover, intake and exhaust manifold gaskets. so basically valve cover spacers and intake and exhaust flanges so i can build my own manifolds/headers. the biggest piece would end up being like 22X8 overall that is only a rectangular ring that is open in the center cut out of .375 6061. accuracy can be fairly foregiving but to be within .002-.003 would be the max variance i would like to have? i want to become aquinted with the process before i go and spend like 25k on a nicer unit that i cant even use all of its capacity. I own a cnc plasma table so im fairly aware of the concepts and geometry application. but then again this is all 2 axis work but the addition of the z axis into the equation is not mind boggling by any means. let me know if i should start out with this mentality or not.

[jcc3inc]

Sir,

A friend purchased a used Bridgeport Boss I machine, removed the old controls but kept the stepper motors. I am quite certain that he has less than 5K in it, and he seems very satisfied with his end results. It is a nice piece of iron and his CNC update made it very practical.

For more information, see www.outbackmachineshop.com
He has pictures and comments on his site.

Regards,
Jack C.

[bestrada11]

wow that is truly amazing.. i dont know if i have enough knowlage of cnc technology to perform a retrofit... i was hoping that someone knew of a hobbyist plug and play that i could learn from and then move onto bigger and better things. so if i had a machine with power feeds on all three axis than i could retrofit the controller to these motors???

[jcc3inc]

Sir,

Note that Jim Glass has an email; you could correspond with him and see all about the transformation.

Jim is a very capable and innovative person. He has made several machines that are impressive.

Are you in Ariz? Maybe someone can do some of the parts for you, if you are reluctant to dip into certain areas.

Regards,
Jack C.

[bestrada11]

im in new mexico... the only thing i wouldnt be clear on is how the controller will monitor position and movement?? i do know a little bit about the controlling aspect, because i have a cnc plasma cutter that if done major servicing to and that unit uses optical encoders mounted on the motor shaft to keep track of itself... but i will contact this gentleman and see if he will donate some of his time to talk with me about my options... because i have found a mill that i really like with power feeds on all axis that is around 4500 and i wouldnt mind spending another 1-2k on converting this unit if it is doable... thanks for the input and the replies on this!!!

[Mark Hockett]

Here is a CNC mill on ebay that looks like it wouldn't take much to get going. http://cgi.ebay.com/CNC-VERTICAL-MILL-M ... .m20.l1116

It looks like the aluminum casting for X axis servo is broken. Milltronics still makes and supports this machine the getting the parts should not be a problem. The machine sells for about $27K new. I have this same machine and it is a very accurate CNC mill. Here is a video of mine in operation,
http://www.youtube.com/watch?v=NpbN1UDwtQM

It uses linear guides for the Z axis so the quill is still functional like a normal Bridgeport mill.

If you are not an electronics wizard buying a machine that is still supported is the way to go. If you have problems the factory is very helpful. I am going to upgrade mine in the next month or so to the latest control option and the entire machine control upgrade with a new single board mother board is only $1600, that is fairly cheap for an industrial CNC control.


The machine has a very powerful and easy to use conversational control so you don't have to know how to program in G code its no problem, you just answer questions. Trying to retrofit and older machine would probably add up to a whole lot more than this one would cost. You could email the company selling it and offer them the $4950 and they pay for shipping, times are rough for machinery dealers right now. I wouldn't sell mine for less than $10K.

You can download a control simulator from Milltronics web site and learn how to use it before buying,
http://www.milltronics.net/CNCControl.aspx
click on view product demo. If you need help getting started feel free to call me at 360-914-6026

[Michael_Moore]

I paid $5K for my running Tree 325 with Dynapath Delta 20 control The iron seems in pretty decent shape though I did put fresh balls in the X and Y screws. Replacing all the belts also seemed like a good thing to do. Keep in mind that once you buy the machine that still leaves you easily spending another $3K+ on vises, CAT40 toolholders and other tooling.

The control was functional but I did do a retrofit. However, that was my decision and not something that was forced on me.

There's a factory brochure on the Tree CNC line in this folder:

http://www.eurospares.com/graphics/meta ... eebrochure



cheers,
Michae

[bestrada11]

Thanks for all the great input!! it looks like i got alot of research to do !!

[JimGlass]

I bought an old Bridgeport Series I CNC with a boss control for $1500. A week later and another $1000 and a few pointers from Jcc3Inc I had the retrofit completed. I have been running the Bridgeport with a Compaq 7770 laptop for over three years now.

A few months after completing the retrifit I learned how to convert CAD drawings to G-code. Now customers email CAD drawings to me and I convert them to G-code in minutes instead of days of keying in every move. The software does not make mistakes either.

I run something in that old Bridgeport just about every week. Sometimes the machine runs all week. Not a bit of trouble with it.

I remember people saying a laptop cannot run a CNC Bridgeport, but it does. Others said the laptop will fail because of the machine vibration but it has not even though I should retire the old WIN 98 laptop. Other people said the Gecko 202 driver boards will fail and they have not.

The old Bridgeport is not real fast but for $2500 and having the CNC capability in my garage I have learned to accept the slower speeds.
I love that old machine.

The day it came home;


Jim

[bestrada11]

Are these machines, once dialed in and fixtures made able to be a "set it and forget it" type of process my cnc plasma tables over the years have progressably gotten closer to this but, one can never control unattached metal.... especially the thin stuff... if this was capable of initial set up and then walk away and let it run for a couple hours i dont mind that one bit for 2500 dollers....

[JimGlass]

Are these machines, once dialed in and fixtures made able to be a "set it and forget it"

I would say "Yes", that is part of the idea of CNC machining. Most CNC operators are able to run more than one machine at a time. Sometimes I will have production work where I make a batch of six pieces in a single run that will take twenty minutes. During the run I'll sweep the the floor, cleanup, or if late in the day have "a" beer during the run.

However, as an old manual machinist I'm reluctant to leave any machine running unattended. I have tried it and stuff happens when I'm away. Most of my CNC work is expensive tooling and most of the time making only one piece. Most of the time I'm right there watching the CNC machine run making sure the tool has coolant and nothing moves.

Below is a special cam for a machine:


Jim

[toastydeath]

I'd like to second the "don't set it and forget it." Sit around and read a magazine, or get on your laptop (my personal favorite), but be by the machine. There are shops that can, and do, set it and forget it. But they spend a lot more than 2500 bucks to get that capability.

There are heisenbugs and bohrbugs that pop up all the time in CNC work. This past friday, two occurred at once where I work in the same day. It's okay to go to the bathroom and whatnot, but long term unattended operation isn't the easiest thing for a shop to accomplish, and has a steep (and expensive) learning curve.

In terms of accuracy, all our (three) bridgeport CNCs are repeatable to .0005". That means they can usually be made to cut to .0005" employing some finagling by sneaking up on dimensions and juggling program numbers. They are "vendor retrofits" from whoever we bought them from. Ballscrews on X and Y, and scales providing feedback to the control. We do not use them for anything but short runs, because changing the tooling is a pain. The controls are conversational, which is bad in some ways and good in others. Given the choice, I'd rather have a real vertical machining center for almost every job I have to do on those knee mills.

I also recommend not using mastercam or whatever CAM package you prefer for as long as possible if you pick a machine up. I feel that some folks are relying on the pre-programmed expertise of the CAM system rather than having a comfortable and full understanding of what the code is doing. There are speed advantages in many cases, absolutely. But I look at my boss, who has mastercam on his desktop, and he rarely uses it. Not for lack of understanding of mastercam; he's pretty darn good with it. It's just he has a better grasp of g-code than most people have mastery over English, and one of the most humbling and reoccurring experiences I have is him walking up to a control and hammering out code faster than I can even begin to think about what I'm going to do first. At least personally, I'd rather struggle with hand coding and get to a fraction of what he can do and to gain a fraction of that insight before switching over to CAM. Perhaps that's just my bias, and CAM is the right solution for you right off the bat, but this seems to be a recurring theme I've heard others express.

The following is a list of experiences that taught me that it is Not A Great Idea to walk away from machines for a long time. There's an even larger set, all caused by one operator we had who liked to hang out on the other end of the building.

Bottom line, throw M01's before each toolchange. This way, if you're feeling sketchy about something and have to use the rest facilities or grab another cup of coffee, you can do it immediately and without worry for what might happen.

• A machine with a random access tool magazine decided to put a large facemill away right next to a large drill - the facemill lost, and fell out of the tool arm. It was a soft fall, considering the bed of chips. If the machinist hadn't been sitting there, nobody would have noticed the soft "thunk" of the tool dropping, and the next part to change into the machine would have had a very nasty crash.
• Two, a vertical mill I was running had the toolchanger break during a tool change. The claw that holds the tool, for some reason, sheared in half. The 30 lb facemill (it will always be the most expensive tool in the machine that drops) the machine was trying to swap dropped right onto the part. And the machine went to go on with it's business, believing the toolchange to be finished. That would have been another, even more expensive crash as either the spindle jammed on the tool, or the machine went to go and mill this tool out of the way with a 2.5" x 10" insert drill.
• Another machine had a drill break during a pattern (silently), then the tap going into that pattern broke when it found the first hole that wasn't there. The operator was away, doing who knows what, but if someone hadn't noticed the sound of the tap grinding itself up, the next tool would have ran and blown up on tap shards. And then the next one, and so on.
• A mill had an endmill shear in half, while it was doing a fairly deep pocket. Silently. So when the next tool went to rapid to the bottom of the pocket, there was a little bit of interference. It was a small tool, and it too broke off silently. The machinist noticed the "crunch" of the small tool breaking, and managed to stop it before any large tools came in to work on that pocket.
• While standing around, I heard my machine go to change tools, followed by no cutting sounds on what should have been a very heavy roughing operation. The toolchanger had become lost, shifting every pocket down one. It went to the right tool, somehow thought it'd gone too far, and backed up one. It did the same thing later that day, which is how I got to see the actual "backing up" part. Freaky.
• On a lathe, turning copper. 5" deep bore, 4.5" diameter. Very stringy chips, such that they'd fill the whole bore and there was a M00 (program stop) right after the first roughing pass so you could clean the chips up. On the 23rd of 24 parts, the chips packed just right and ripped the chunk of copper right out of the chuck.