Pm25-mv Conversion Planning

sbx

sbx

Registered
Registered
Joined
Dec 16, 2014
Messages
85
Hello,

So, I am starting to document and get my plans together to convert my PM25-MV mill to CNC. I'll start a separate build thread in the Build log section, but wanted to post in the general section to help get some ideas together and solidified. Hopefully you all can help clear up some of the remaining confusion.

Grand Vision: 4 Axis, Spindle control, flood coolant, full or mostly full enclosure, power drawbar and utilize TTS for toolholding (ER20 - Already in use).

Software: LinuxCNC (already familiar with Linux, hate windows) and want to incorporate a handheld game pad style pendant.

Now, here are the couple main questions I have confusion on, that I hope for help.

1. What tradeoffs are there for stepper motors to the ballscrews. I've seen builds using timing belts and others with direct connecting couplers and standoffs. But, I've never been able to pin down the benefits of one way over the other. Just not sure how I should start planning or where to look for more info.

2. I've been looking into the Mesa 5i25/6i25 FPGA I/O cards as opposed to dealing with parallel ports and old PCs. Thinking of building up an mITX Atom (or equivalent) based PC with the Mesa card as the primary motion control interface. If I do this I am somewhat confused about which stepper diver(s) that would be optimal. I think if I understand what I've read, this would also eliminate or equal the performance of having a ethernet smoothstepper involved. Thoughts?

3. Should I build the full enclosure from the start? Or build the mill, then finalize the dimensions and go forward building a full stand enlosure etc. The latter seems the most sensible, but also could be a pain disassembling a working set up.

Or, just chime in with " I should have done this first or better."

Thanks for any input you all can provide,

sbx
 
I enjoy my PM25, manual, spent 10 years punching in G Codes for CNC machining, my day job. Having said that, I would start with a larger, heavier machine. I must assume you're planning on powering the head movement. If so, build in an automatic column clamp. My PM25 moves .008 when I clamp it. Just my opinion.

Tom
 
Tom - When you say Automatic Column clamp, are you meaning locking the milling head to the column? Based on the research I've done the stepper/ballscrew can hold its position well, hopefully eliminating the need to have a clamp. None of the BF20 type conversions I've read about incorporated anything like this that I can tell. I do agree that when I engage the column lock on the head that it can move and is critical to do whenever doing any motion in X and Y and wanting to maintain any sort of static Z position.

Regarding a more rigid machine. If I had more space and/or my own house, I would have gone with an RF45 type. But I am currently renting a shop and have no idea what size space I will have in the future. So, the PM25 was chosen for being compact, 110v, and still somewhat movable.

Thanks for the input though. Will look into the z-clamp/ rigidity issue.
 
1. My understanding is the advantage to using a motor and belt is the increase in speed that you can move the axis at. I would thing that a direct connection would decrease the backlash in the syste, but you are limited to how fast you can move the axis. In a hobby situation accuracy would seem to be more important than speed.

2. Not familiar with those options

3. I would wait and keep your options open to modify your build as you go then build the enclosure.

Check out Hoss Machine he has plans for conversion of a G0704. You can mill all of the motor mounts, ball nut mounts etc. with your manual machine than make your conversion. I ordered balls crews from Linear motion on eBay. The prices were very good and the parts appear to be of nice quality. I am working on a CNC router project that has stalled a bit due to work and family. I also ordered Wantai stepper motors, power supply, controllers and control board.

If all goes well with the router I may take a stab at doing my PM-25 as well. Good luck!
 
Sounds like a solid plan and good advice. For a controler, have you thought about running linux on a beaglebone black with a Probotix PBX-BB-BeagleBone-Breakout-Board? I am considering this for my upcoming lathe conversion using a g540 driver.

I like direct drive steppers for the simplicity and rigidity.
 
Thanks for the suggestions so far!

Brooks - I am well aware of the Hoss site and plans. Plan to purchase them as additional reference. I am still unsure on the differences if any between the G0704 and the newer PM25-MV, but think I can figure it out.

Rescue35 - I haven't looked into beaglebone too much, but did see some interesting arduino stuff related to the TinyG board. But, I think there are quite a few limitations to them with either LinuxCNC or it just being too new. The Mesa cards look slick and have a bunch of features all ready designed to integrate with LinxCNC like separate Spindle control, and no need for extra breakoutboards. So, that is why I went there.

The more I think about the coupling concept, I think I'll just start with direct couple Oldham style connectors. That part can always be redone later without disassembling the entire mill.
 
1. Timing belt drives allow for mounting the motors in a more compact profile is space is a concern. You also have the ability to gear the motor up or down. Belts and pulleys need to be fully protected from chips, coolant etc.

Direct drive with the proper zero backlash couplers are easier to mount (less parts to make) and are more tolerant to chips and debris. Always protect from coolant. Direct drive will require your enclosure to be larger to accommodate the extra length of the motor/coupler/shields. Direct drive with dual shaft motors can be setup with hand wheels for manual use.

2. Linux/Mesa is a great combo if you are familiar with Linux and don't mind doing command line entry or a little coding. We are running a Linux/Mesa combo on the CNC router we built for the High School. Ours is a servo system but it is very robust and much more configurable than a Mach3/windows system. I'm a Linux hater so I don't operate the machine.

For a Windows based system the PMDX BOB's and Ethernet smooth stepper is an excellent combo with excellent support. PMDX also has some new offerings that may be even better. I'm running Chinese drivers and power supplies on my PM932 conversion and Gecko drives with a toriodal power supply on my CNC router. Of the two I prefer the Gecko drives. More robust and excellent support. I have had to replace two Chinese drives but I also push the limits of the machine.

3. Not necessary to build at the same time but it may dictate layout of other components such as one-shot oil systems, coolant pumps/supply lines/drain/holding tank, control box, etc. Plan ahead.

Some people like full enclosures. I kept mine simple to allow for running parts wider/longer than the enclosure will allow.

Good luck with your project!

Jay
 
sbx, do you have the newer PM25-MV that comes with the brushless motor and belt drive? I'm considering getting this mill for CNC conversion but I have read in other forums that it's not a Weiss machine and may have differences from the standard G0704/BF20. Are you planning to make the mounts etc. yourself or buy them from a kit? I'll be interested to know what you go with.

Cheers!
 
charlton said:
sbx, do you have the newer PM25-MV that comes with the brushless motor and belt drive? I'm considering getting this mill for CNC conversion but I have read in other forums that it's not a Weiss machine and may have differences from the standard G0704/BF20. Are you planning to make the mounts etc. yourself or buy them from a kit? I'll be interested to know what you go with.

Cheers!
Click to expand...

I have the PM-25 MV and it is in fact different in many small ways. The most critical would be the end plates that secure the lead screws. They are different than the G0-704. If you plan to make your own parts it would not be a problem. If you buy a kit they would probably not work. The other consideration is the spindle bearings are not designed for the higher RPM's used in CNC and would need to be replaced.
 
I would prefer to buy a kit since I have no experience with doing any CNC conversions so I worry about messing things up colossally. I understand that the bearings are probably not the tapered roller bearings but I wasn't sure why CNC would require higher RPMs from conventional milling. Are the spindle and feed speeds not typically dictated by the material and cutters?

Anyway, I don't want to hijack SBX's thread. I'll be watching attentively.

Cheers,
Charlton
 
You must log in or register to reply here.
Back
Top