CNC daydreaming

[rabler]

But I'm another vote for starting with a commercial built CNC machine. All of the expensive and hard mechanical work is already done and all you have to do is hang the new controller/motors of your choice on the machine and press go to make parts.

Personally if I only had one mill in the shop it would be a BP or clone knee mill with both manual and CNC capability. These machines pop up on eBay and Craigslist occasionally, normally around mid-80's vintage. Nothing wrong with getting a 35 year old machine if it is in reasonably good mechanical shape. For more serious production work, you need a VMC with auto tool changer. These can be found, with dead controllers, for scrap price or less. I saw one for free the other day. Normally if the old machines are lubricated properly through their life, they just don't wear out. My 35 year old Hardinge CNC lathe runs every day and will hold +/- 0.0002'' all day long.

Good input from everyone! Glad I posted this, kind of felt like it was a aimless thread but I'm getting some great ideas. Thanks you everyone!!

I've certainly seen VMCs go for fairly cheap recently at auction. Chicago and much of Ohio are within range and I have a 10 ton trailer.
I'm strictly a hobbyist, and have no desire toward production. But I already have a couple machines that are in the 4-5 ton size range, so big is not out of the question. I don't foresee getting rid of my manual equipment, so I'd lean toward a dedicated CNC machine rather than a mixed machine.

How much different is it to take a commercial VMC and rebuild the electronic than building around a knee mill? Any things to particularly look for or look out for on either one? Or particular models to look for or avoid? All the pictures of VMC's look like cabinets, so I really have no idea how similar or different they are from manual machines.

 

[Karl_T]

The big question here is do you want a tool changer? For a first timer, making this work can be quite the challenge. And its not going to happen with the inexpensive hobby controls unless you are a wizard with interfacing a PLC.

The cabinet is just an enclosure for flood coolant and chips to not get out in the room. Nice if you got the space.

My two cents you won't go wrong with a good heavy duty knee mill with dead control. Look for box ways and taper 30 or taper 40 tooling - this is WAY MORE RIGID than an R8 dovetail way machine. These should come with an air or hydraulic tool release, makes hand tool changes real quick.

I wore out my excello with these features and then rebuild a Vectrax. there is a thread on this website for the rebuild.

 

[JimDawson]

How much different is it to take a commercial VMC and rebuild the electronic than building around a knee mill? Any things to particularly look for or look out for on either one? Or particular models to look for or avoid? All the pictures of VMC's look like cabinets, so I really have no idea how similar or different they are from manual machines.

There is really very little difference from the controls end. Just a bit more wiring for the additional hardware. They normally come equipped with an automatic tool changer (ATC). 16''x20'' and 20''x40'' are common sizes in the smaller machines (<10K lbs.) Normal spindle power for the smaller machines is 7.5 to 10 HP, with maximum spindle speeds in the 5K to 10K RPM range.

EDIT: I should note here that an ATC does not mean that you have to use it, you can also do manual tool changes with a button push to release and install the tool in the spindle.

There is a huge difference in the VMC machine frame compared to a knee mill. The VMCs are bed mills, a much more robust arrangement. All of the Z axis movement on the column. These machines are engineered and built for heavy production use.

 

[Reddinr]

Does Solidworks "Maker / 3dexperience" version include a way to post process for making g-code? Their marketing videos are frustratingly slim on information about what you get and what you don't

Thanks, R

 

[rabler]

My two cents you won't go wrong with a good heavy duty knee mill with dead control. Look for box ways and taper 30 or taper 40 tooling - this is WAY MORE RIGID than an R8 dovetail way machine.

I don't know what "dead control" means? I have a K&T 3K, that for it's age is in pretty good condition in terms of table/bed wear.

There is a huge difference in the VMC machine frame compared to a knee mill. The VMCs are bed mills, a much more robust arrangement. All of the Z axis movement on the column. These machines are engineered and built for heavy production use.

I think rebuilding the electronics in a VMC is going to be the plan. Now I just need to find which models are of an appropriate size. Some of them are ridiculously large. The 7.5-10 HP range fits my power/weight budget. I'm pretty sure I'll have room in the new shop, even with thoughts of a few other toys. I think the ATC is something I could start with not using and then eventually incorporate into the electronics.

 

[coherent]

I'm daydreaming of building a small CNC mill some day. And probably a plasma cutter too, although I could see myself buying one of those, possibly even used. But the CNC mill is likely first.

This would be at least a year down the road before actually doing. I've got a new shop that's schedule to get framed up in January, and it'll keep me busy for quite a few months (and keep my budget slim) as I'll need to finish it out with insulation, electric, etc. So in part I'm just killing time daydreaming until they start building the shop.

I'm inclined toward starting with something like a PM-940V (that's my current thought), hardened ways, get the Arizona Video ball screws and servo motors, then build it up myself from there. I've read a few threads discussing that. I'm fairly comfortable with messing with electronics. I will probably go with linuxcnc. I'm an old Unix/linux hack going back to the early 80's. I have looked at Centroid Acorn too. I'm still kicking around options so none of this is set in stone. I've looked at Tormach, and while the DIY aproach isn't all that much cheaper, I'm just more inclined to do it myself.

I don't currently use any CAD environment for my hobby work. There was a time in my professional career many years ago where I was heavy into AutoCad, hell I've even done a bit of VLSI layout. But the pendulum swing on retirement was to get away from working in front of a monitor.

Right now one of my biggest challenges is internet bandwidth. I'm using HughesNet satellite service. This means YouTube, while a good resource, is pretty much out of the questions (except if I have insomnia, 2am - 8am doesn't count against the usual quota). It also hampers using any internet CAD tools, like Fusion360. I have a deposit in with Starlink, so hopefully that situation improves some in the very near future.

So what are some good learning resources for doing a CNC mill conversion? What advice do those of you with hands on experience doing this wish you would have heard before starting?

I'm not sure what your internet speed has to with wanting to get into CNC but maybe I missed something. If you haven't yet, check into Starlink internet and see if it's available in your area yet. I'm also in a limited options rural location and the best I could get was 15mb or less from Suddenlink (which is the worst ISP and customer service imaginable) . Signed up for Starlink and now average 150-200mb down, 15-25mb up and no monthly limits for 99 a month. Best move I ever made.

It sounds like you are on the right track decision wise as far and getting a CNC mill.
I started in CNC nearly 20 years ago. Things have come a long long way for the budget and "do it yourselfer". Read, read and read some more. Lots of online information. The more you read, the more it will all come together for you. Once you actually get into a build and then use the machine a bit, most of the mystique will go away and it's all pretty simple. If you want a CNC mill, first find a model that lots of others have converted (sounds like you already have) and and buy a new or used mill in the size and price range you want. You'll learn much much more than buying a new cnc mill or a machine that someone else has converted. I think the CAD CAM side of stuff has the biggest learning curve, but there is no reason you can't start on that now and be ahead of the learning curve before you actually have a machine. I've build a number of machines including CNC Laser cutters, CNC mills, CNC routers and 3 or 4 CNC Plasma cutters and other than some minor hardware and cut file programming differences CNC is CNC no matter the machine. I have a couple of 3d printers and they do differ in software and CAD modeling to some extent but most general techinques and CAD/CAM skills are the same.
If you can find a fellow forum member or a neighbor who has a CNC router, mill or plasma cutter a little time watching or having them explain and use the machine will go a long way toward simplifying things and helping you decide on the route you want to take and machine you want for your purposes. Tons of videos out there on specific machines and conversions.
Good luck!

 

[rabler]

I'm not sure what your internet speed has to with wanting to get into CNC but maybe I missed something. If you haven't yet, check into Starlink internet and see if it's available in your area yet. I'm also in a limited options rural location and the best I could get was 15mb or less from Suddenlink (which is the worst ISP and customer service imaginable) . Signed up for Starlink and now average 150-200mb down, 15-25mb up and no monthly limits for 99 a month. Best move I ever made.

Glad to hear that you're getting good performance from Starlink. As I mentioned earlier I'm on the waiting list. HughesNet limits you to some minimal data per month at a fairly modest bandwdith, it makes streaming any internet video prohibitive. The hobbyist version of Fusion360 uses cloud storage, so slow internet makes it somewhat clumsy.

 

[JimDawson]

I don't know what "dead control" means?

I think Karl means a CNC knee mill with a broken computer. A rather common problem with old machines. The mechanics of the machine far outlives the old computers (or even the new computers). I expect my machines will still be mechanically functional 100 years from now, but the computers will be long dead.

 

[rabler]

I think Karl means a CNC knee mill with a broken computer. A rather common problem with old machines. The mechanics of the machine far outlives the old computers (or even the new computers). I expect my machines will still be mechanically functional 100 years from now, but the computers will be long dead.

That makes sense. Karl, pardon me being a bit slow.

 

[Karl_T]

Say Jim, what control would you suggest for the OP? My Camsoft and your custom controls are a bit over the top for a first timer. Mach used to be suggested, sounds like that one is not so much any more. linux CNC?? Centroid?? I really have not kept up here.

I do like those DMM servos you turned me on to. They come with a step driver for the low end controls. After ALL YOUR HELP, getting +/- 10 volt analog servo mode going, they are wonderful. But I do not know a control to suggest that can use this. My .02 putting low end stepper control on a VMC is just wrong.

Rabler - you should know stepper controls do NOT take feedback from the servos. They just issue commands and expect the machine did it right. I put this concept on rebuilds in the 90s. After ruining only the high dollar parts, I went out and found a servo control with position feedback. My opinion - stepper control is for low end hobby use only. maybe I am a snob. But its a well earned opinion.