Considering CNC Mill... what should I look for?

[ArmyDoc]

I am planning on building my shop this fall, and have been looking at manual mills and lathes. I had it in the back of my mind that I might want to convert the mill to CNC at some time down the line. A friend said if that was my plan, I should look at getting a CNC mill in the first place. I was surprised to find that the CNC weren't all that much more expensive. Besides the things I would look for in any used mill, like wear on the ways etc, What things do I need to be aware of and watch out for when evaluating a CNC mill?

 

[BGHansen]

I've picked up two used CNC mills. Not an authority by any stretch, but here's what I looked at and questioned. First thing I asked is "why are you selling the mill?"

First one was a 1981 Bridgeport Series 1 2HP mill with an Anilam Crusader 2 controller (circa 1985). Answer to my question was he'd used the mill in a side business shop (was a tool and die maker). He was retiring and selling off his equipment.

The table runs on ball screws, so should have little backlash, but ball screws can wear out. I put a 10th indicator on the spindle and checked backlash relative to the table X/Y hand wheel micrometer dials at various points in the table travel.

Also checked for table flatness by running the table back/forth with the indicator on the table. Checked the spindle/quill for looseness at the bottom of the 5" travel. Ran the mil through its speed range for wobble/noise. Verified that boring and the kick-out worked.

I knew nothing about the CNC, so let the owner plug in a few routines to run it through its paces.


Second one was a Tormach 1100 Series 3. Answer to my question was he used the Tormach in a side business (www.yourlittleCNCshop.com) but instead of running his parts on the Tormach, had worked out a deal with his boss (he is a machinist) to run his stuff on a Haas CNC at work. He had jobs that took 20 minutes on the Tormach, but only 3 on the Haas.

Did the same sort of checks though the Tormach is a square column mill, so no quill to check. We checked for "lost motion" (Tormach's term) with an indicator against a 1-2-3 block. Also cut a round pocket in a chunk of aluminum and measured the diameter across the X and Y (were different by just under 0.001").


You might want to look at age of the controller. My BP Anilam has a problem with the RS-232 port. I cannot load programs to/from the controller. There is a shop in North Carolina that works on them at a shop rate of $250 / hour. I'd likely pull the Anilam and go with a retrofitted Centroid Acorn controller before trying to get the Anilam fixed. All other functions on the Anilam work fine.

My BP has a 2-axis controller. The Z-axis is manual. After learning the programming in 2-axis, I had a bit of buyer's remorse and wished I'd had a 3-axis machine. I was planning on adding the 3rd axis on the quill or knee (Jim Dawson has a thread here on his quill CNC addition, CooterBrown has one on a BP knee addition). I ended up buying the Tormach instead, kept the BP also.


Bruce

 

[Boswell]

Many conversions use Mach3 (so do I currently). Mach3 has been around for a while and was originally setup to a parallel port to connect to the Breakout and motion control boards. Microsoft support for the parallel port ended with Microsoft XP. It is an easy upgrade to use an Ethernet Smooth Stepper to replace the parallel port connection but you should be aware that if you look at a used system that relies on a parallel connection and and old MS XP based computer, you can expect to upgrade all of that sooner or later and probably sooner. I guess the underlying message here is that you need to evaluate the CNC control software chain as there are many out there and some are better than others. Three key parts to evaluate. CNC software (Mach3), Breakout board and motion controller. These can be all separate or combined is various combinations based on the system. Some are open source (LinuxCNC) other are proprietary. Both have pros and cons. Some target a hobby user and others will be full on commercial systems. As you start to find some specific system, share the details here and I am sure you will get a lot of expert feedback.

 

[Mini Cooper S]

I was also looking at converting a manual machine into a CNC also, as I was watching for something to come on the market, I found a Series 1 Bridgeport CNC circa 1978. The hardware was in very good shape but as you can imagine, the controller was less than reliable. I wound up stripping the guts out, and one of the two electrical cabinets on the machine. I then fitted a spare PC that I had, along with a PMDX- 454 Smartbob, PMDX-407 spindle control, Gecko G214 Microstep Drives and Mach4 software. It took a while along with a lot of internet research to learn how to make it all work, but it turned out to be a descent machine when I was done.

I say "when I was done", that really didn't happen as I am always adding some new feature to my project. All totaled I have $5900 in the machine.

Richard

​

​

 

[samstu]

I'm no expert but will give you my opinion.

Maybe first decision is the mill a project or is the mill for projects? If the mill is a project, old mills with anilam, boss and other controllers make great conversion candidates as they can be found in low hour condition for little green. They are cnc ready with ball screws and sometimes servos can be reused. So doing homework before purchase is key. Converting a manual mill to a cnc is usually not worthwhile due to needing to change to ballscrews, etc. Mach 3 has a good following while Centroid has diy conversions to real industrial controls. Masso is another interesting controller but isn't fully baked in my opinion.

If the mill is for doing projects, get one ready to go. Meaning modern controllers and conversational controls have real value, especially if time is of value. You can make older anilam etc work but parts, support, etc are spotty and can be pricey.

I almost purchased a bridgeport with 12 year old centroid control. It worked great and was accurate but when I researched this machine, I found that it had been for sale for nearly a year (it's still for sale) confirming a limit resale potential. I then purchased a cnc mill from HGR, but when I got it home, I learned how badly worn it was. So, I returned it for full refund.

In the end, I purchased a Tormach 1100 series 3 with a full enclosure and power drawbar. These mills sell quickly, have good resale, have "factory" support, and parts are cheap. Spindle bearings from tormach are $126 a pair for lower and same price for uppers. Pathpilot software updates are free. Sure there are limitations and yes, I'd rather have a Haas TM or even a nice bridgeport conversion but it's a hobby and the tormach is forgiving, repairable, and not nearly as messy as an open bridgeport. It's more saleable than a DIY conversion as these vary in success and usability.

 

[JimDawson]

I would try to find a CNC machine that is in good mechanical shape. I wouldn't worry about the controls at all, that's the cheap and easy part to fix. A CNC with dead controls should sell at near scrap price. My preference is a manual/CNC machine. There are a few of these around.

 

[kb58]

Another thing to keep in mind if this is your first mill - will you want to be able to use it as a manual machine? I did, and giving up the handles for CNC didn't make sense in my case, since I do odd one-off things. Some CNC mills retain the cranks for manual operation and some don't, so choose wisely.

 

[ArmyDoc]

Maybe first decision is the mill a project or is the mill for projects? If the mill is a project, old mills with anilam, boss and other controllers make great conversion candidates as they can be found in low hour condition for little green. They are cnc ready with ball screws and sometimes servos can be reused. So doing homework before purchase is key. Converting a manual mill to a cnc is usually not worthwhile due to needing to change to ballscrews, etc. Mach 3 has a good following while Centroid has diy conversions to real industrial controls. Masso is another interesting controller but isn't fully baked in my opinion.

If the mill is for doing projects, get one ready to go. Meaning modern controllers and conversational controls have real value, especially if time is of value. You can make older anilam etc work but parts, support, etc are spotty and can be pricey.

I almost purchased a bridgeport with 12 year old centroid control. It worked great and was accurate but when I researched this machine, I found that it had been for sale for nearly a year (it's still for sale) confirming a limit resale potential. I then purchased a cnc mill from HGR, but when I got it home, I learned how badly worn it was. So, I returned it for full refund.

In the end, I purchased a Tormach 1100 series 3 with a full enclosure and power drawbar. These mills sell quickly, have good resale, have "factory" support, and parts are cheap. Spindle bearings from tormach are $126 a pair for lower and same price for uppers. Pathpilot software updates are free. Sure there are limitations and yes, I'd rather have a Haas TM or even a nice bridgeport conversion but it's a hobby and the tormach is forgiving, repairable, and not nearly as messy as an open bridgeport. It's more saleable than a DIY conversion as these vary in success and usability.

Thank you. There were a lot of good points in there. The part about having ball screws to start vs the cost of replacing is why I had heard it would be better to start with a CNC machine.

I guess that I lean more towards the mill being for projects than the other way round. I suspect that tearing apart a mill is a larger project than I am ready for, and the idea of replacing electronics is terrifying, rather than exciting. I have no problem replacing components - I can pop out an IO card, hard drive etc and plug a new one in it's place, but that's about the extent of it. I've never tried to build a computer from various components.

I read that Haas, Fadal, Fanuc and Wells-Index are all worth looking at, because they are still in business and supporting their machines. I will add Tormach to that list. Had heard anillam and Boss were to be avoided due to age. Real problem is it seems that only Bridgeports are available around here.

I would try to find a CNC machine that is in good mechanical shape. I wouldn't worry about the controls at all, that's the cheap and easy part to fix. A CNC with dead controls should sell at near scrap price. My preference is a manual/CNC machine. There are a few of these around.

How easy and cheap? If it's a matter of pull out the old system and plug in an new one, I can probably handle that. But I don't know about wiring at a pin level. I guess if you're replacing everything, it's pretty much all together and all you are doing is connecting it to the servos. That wouldn't be too hard. I guess what I'm saying is plug and play with components is about my speed. If it's wiring soldering etc, not so much.

Another thing to keep in mind if this is your first mill - will you want to be able to use it as a manual machine? I did, and giving up the handles for CNC didn't make sense in my case, since I do odd one-off things. Some CNC mills retain the cranks for manual operation and some don't, so choose wisely.

I have been going back and forth with this. Part of me says yes, I want to use it as a manual machine. But that may be just because I have been thinking about manual machines and making things in those terms. That's also why I lean towards a Bridgeport VMC type machine, rather than a fully enclosed unit. Part of me says "Plugging in a piece of metal, and pressing start isn't really making it yourself." But the part of me is focused more on the part , and doesn't want to be doing all of the repetitive steps needed to make it says, "just let the machine do that." Also, some of the things I want to make would require 4th axis control with simultaneous movement, which I couldn't do manually anyhow.

I've read that a pendant can provide full manual control that is even better than hand dials. What do you think about that?

 

[kb58]

...Part of me says "Plugging in a piece of metal, and pressing start isn't really making it yourself." But the part of me is focused more on the part , and doesn't want to be doing all of the repetitive steps needed to make it says, "just let the machine do that." Also, some of the things I want to make would require 4th axis control with simultaneous movement, which I couldn't do manually anyhow.

I've read that a pendant can provide full manual control that is even better than hand dials. What do you think about that?

I meant more about the best use of your time. If you just want to drill some holes, or face a surface, sitting down at the PC and setting that up takes way longer than just doing it yourself, manually on the mill. That said, yes, if you have a pendant/dongle/control box, then I guess handles have gone the way of the dodo.

 

[ArmyDoc]

I meant more about the best use of your time. If you just want to drill some holes, or face a surface, sitting down at the PC and setting that up takes way longer than just doing it yourself, manually on the mill. That said, yes, if you have a pendant/dongle/control box, then I guess handles have gone the way of the dodo.

Ah. I see what you mean. Yes, for simple operations, CNC programming would take a lot of time.
I would definitely want some form of manual control.