Need help - Lathe Build, trying to understand requirements/limitations

Results 1 to 4 of 4

Thread: Need help - Lathe Build, trying to understand requirements/limitations

  1. #1

    Default Need help - Lathe Build, trying to understand requirements/limitations

    Hi Everyone,

    I'm pretty new to CNC so I do apologize for the newbie questions! I just want to ensure I'm accounting for everything I need and not missing a key piece of hardware or barking up the wrong tree.

    Some background on me first - I make custom fountain pens, generally out of wood, resin, and metal accents - sometimes delrin or ebonite as well. I may or may not start making metal pens, but if I did it would likely be brass/aluminum. I started last year, but my small business has taken off and I'd like to automate / open up some of the possibilities of what i can do. A lot of my tedious work involves drilling, threading (int + ext), rounded square blanks, and shaping my pens once they are done - I have a number of pens in different sizes which all have different curves/profiles. I'll also be making accent rings out of round stock.

    My original plan was to buy a 10x22 metal lathe and outfit it with a DRO - my current process is that I do everything by hand on a wood lathe, and it gets quite tedious. Not only sizing tenons, threading, etc, but getting my pens down to the size for the model I'm making is a pain in the butt, and then i have to go to town with a file, and then sand the crap out of it to get rid of the file marks. Being able to dial in the diameter and kick on the auto-feed would be great, but we all know that full CNC control is the bees knees and if the DRO is gonna cost half of the CNC cost, I might as well go all the way!! My backup plan is a template system for profiling, but i sense that it's only marginally more work to do a CNC retrofit.

    Since I "just make pens" I don't have a 3D printer or a mill - so my first thought is that I will use the lathe as it is, and drive everything with belts. My hope is that this will let me learn the ropes, get my feet wet, and get used to the CNC workflow. I'm a little shy about replacing the leadscrews with ballscrews right away because of the mods that would be required (milling out the cross slide which I can't do). I don't know the first thing about encoders but I get the sense that I need to encode the spindle somehow for multi-start threading. I'm in canada so there aren't any pre-made kits for the lathes that I can purchase, and pre-retrofit lathes from CNC Conversions Plus are out of my price range. My plan for now is to use Centroid Acorn because I should be able to hit the ground running.

    So with all of that out of the way, what I'm wondering is -

    1) With backlash compensation, will threading go OK, or should I assume that until I do a ballscrew replacement I'll need to keep using my taps and dies? if I can use CNC for tenons, shaping, and some drilling, that will be a huge help and I can always put off threading.
    2) What kind of spindle encoder should I be looking for? I can't connect anything to a shaft, because the lathe spindle is a gigantic hole. Is there a simple approach and a readily available optical/magnet sensor that I can employ?
    3) assuming I can create a reliable zero where i can start and end the CNC, is it reasonable to assume that I could do manual operations by removing the belts? or should I anticipate that i'll need to either (a) get good at Intercon or (b) do turning operations using the MPG? I've been told I "can't do manual turning using the MPG", but some folks are just biased. but maybe without using ball screws that's true?
    4) As a worst-case scenario, is it safe to assume that if nothing else I could do profiling with a belt+leadscrew setup? That alone would be worth the effort for me. It takes me two hours to shape a pen, sometimes more.

    I think I've got a list of all the hardware I need, so I might make a separate thread for that - I just want to make sure i'm on the right track. If I really need to be looking at a full retrofit I'll have to take a step or two back and probably wait a year.

    Similar Threads:


  2. #2
    Community Moderator Jim Dawson's Avatar
    Join Date
    Dec 2013
    Posts
    5717
    Downloads
    0
    Uploads
    0

    Default Re: Need help - Lathe Build, trying to understand requirements/limitations

    Quote Originally Posted by guitarchitect View Post
    -

    1) With backlash compensation, will threading go OK, or should I assume that until I do a ballscrew replacement I'll need to keep using my taps and dies? if I can use CNC for tenons, shaping, and some drilling, that will be a huge help and I can always put off threading.
    It is possible to do any lathe operation without ball screws. But you do need mechanical backlash compensation to pull that off. I won't go into the details here, but it's possible to do it at relatively low cost.

    2) What kind of spindle encoder should I be looking for? I can't connect anything to a shaft, because the lathe spindle is a gigantic hole. Is there a simple approach and a readily available optical/magnet sensor that I can employ?
    The least cost and easiest method is to use a timing belt drive to the spindle encoder. You just bore a timing belt pulley to the OD of the spindle and attach it. Then use the same size pulley on the encoder. There are a few more ways to do this also.


    3) assuming I can create a reliable zero where i can start and end the CNC, is it reasonable to assume that I could do manual operations by removing the belts? or should I anticipate that i'll need to either (a) get good at Intercon or (b) do turning operations using the MPG? I've been told I "can't do manual turning using the MPG", but some folks are just biased. but maybe without using ball screws that's true?
    With any CNC you can set the work zero to anywhere you want.

    I don't know what ''intercon'' is.

    It is possible to do some simple turning using a MPG. I do that occasionally on my CNC lathe. But for profiling shapes or anything that requires more than a 1 axis move is pretty much impossible.


    4) As a worst-case scenario, is it safe to assume that if nothing else I could do profiling with a belt+leadscrew setup? That alone would be worth the effort for me. It takes me two hours to shape a pen, sometimes more.
    Yes, it's possible. But rather than using the lead screw, I would use the rack & pinion that most lathes have on the Z axis. Again you would have to provide mechanical backlash compensation.


    Starting with a manual 10x22 and then doing a full CNC conversion would not be my first choice. If you want a CNC lathe, then just simply buying one would be my first choice. Most would require a controls retrofit, but that is normally the cheap part. At least you would be starting with a machine that has all of the expensive bits installed and is a much more substantial machine to start with. A Hardinge CHNC would be a perfect size for what you are doing. These are available on eBay from about $1000 and up. There are many other options available also. If you can find an EMCO CNC lathe that be a good option also for your required work envelope, very nice machines. A quick search of eBay shows them from about $800 and up.

    Jim Dawson
    Sandy, Oregon, USA


  3. #3

    Default Re: Need help - Lathe Build, trying to understand requirements/limitations

    Quote Originally Posted by Jim Dawson View Post
    It is possible to do any lathe operation without ball screws. But you do need mechanical backlash compensation to pull that off. I won't go into the details here, but it's possible to do it at relatively low cost.
    I'd love to hear more about this! Is there a reason that software-based wouldn't work? It looks like Centroid's CNC12 software does it reasonably well.



    Quote Originally Posted by Jim Dawson View Post
    The least cost and easiest method is to use a timing belt drive to the spindle encoder. You just bore a timing belt pulley to the OD of the spindle and attach it. Then use the same size pulley on the encoder. There are a few more ways to do this also.
    Makes sense, thank you!



    Quote Originally Posted by Jim Dawson View Post
    With any CNC you can set the work zero to anywhere you want.

    I don't know what ''intercon'' is.

    It is possible to do some simple turning using a MPG. I do that occasionally on my CNC lathe. But for profiling shapes or anything that requires more than a 1 axis move is pretty much impossible.
    Sorry, intercon is the CNC12 approach to conversational CNC - effectively I could build a macro at the terminal/lathe to say, face off the part, give it a tenon, thread it, and drill it out (with me managing tool changes of course)... I can imagine that after getting into the groove of using it I could easily build a small library of stuff i do all the time and assign M-codes (macro numbers) to them, long as i can easily set my Z!


    Quote Originally Posted by Jim Dawson View Post
    Yes, it's possible. But rather than using the lead screw, I would use the rack & pinion that most lathes have on the Z axis. Again you would have to provide mechanical backlash compensation.
    Fair enough - I think the only challenge with that is if I release the halfnut that drives the carriage on the lead screw, the controller now has no idea where the carriage actually is. I can imagine it being a pain to re-set that every time but with a simple DRO maybe it's easy enough to "home" it as i alluded to in my first post.

    Quote Originally Posted by Jim Dawson View Post
    Starting with a manual 10x22 and then doing a full CNC conversion would not be my first choice. If you want a CNC lathe, then just simply buying one would be my first choice. Most would require a controls retrofit, but that is normally the cheap part. At least you would be starting with a machine that has all of the expensive bits installed and is a much more substantial machine to start with. A Hardinge CHNC would be a perfect size for what you are doing. These are available on eBay from about $1000 and up. There are many other options available also. If you can find an EMCO CNC lathe that be a good option also for your required work envelope, very nice machines. A quick search of eBay shows them from about $800 and up.
    Yeah, I get that... unfortunately being in Canada border issues and shipping can really cause some suffering. All the Emco machines i've found were just way too small for my use (I really need about 16" length on the bed, 7" for swing is fine). I considered an 8x16 but I can't get one well-made enough here to be useful, and it's a little constrained for DRO let alone CNC. My choices of new lathes with decent parts availability/support are a 7x12 or a 10x22. I strongly considered a taig (no tailstock, not enough Z) and sherline (underpowered, workholding too limited) but didn't go that route, and retrofitting the CNC onto a new lathe will get me going for only $5,000CAD - much less than the taig and quite a bit less than leasing a tormach 8L. I've been keeping an eye on the local used market but just haven't gotten lucky yet! and i'm assuming that I will probably need to do a bunch of operations manually, primarily drilling with the tailstock because of how finnicky my resins and acetates can be... if they heat up and you don't retract quickly they melt onto the drill bit. could be fixed with flood cooling but that's out of my scope for now



  4. #4
    Community Moderator Jim Dawson's Avatar
    Join Date
    Dec 2013
    Posts
    5717
    Downloads
    0
    Uploads
    0

    Default Re: Need help - Lathe Build, trying to understand requirements/limitations

    Quote Originally Posted by guitarchitect View Post
    I'd love to hear more about this! Is there a reason that software-based wouldn't work? It looks like Centroid's CNC12 software does it reasonably well.
    Software based backlash compensation is marginal at best, and doesn't work well at all on any cut with an arc. You can bias the mechanics with an air spring to always keep the moving load tight against the driving device, be it a leadscrew or rack & pinion. This works really well, I use this method on the quill of my mill and I can count on it to hold +/- 0.0001'' An air spring is just an air cylinder connected to a regulator and a relatively large air reservoir (portable air tank, available at Princess Auto for cheap). This method will work both in the Z and X axis and will give you 100% backlash compensation.

    Sorry, intercon is the CNC12 approach to conversational CNC - effectively I could build a macro at the terminal/lathe to say, face off the part, give it a tenon, thread it, and drill it out (with me managing tool changes of course)... I can imagine that after getting into the groove of using it I could easily build a small library of stuff i do all the time and assign M-codes (macro numbers) to them, long as i can easily set my Z!
    You are thinking about the workflow from a manual operation perspective, for CNC work you have to look at it a bit differently. You would want to create a library of jobs that you want to produce. Then post the G code for each job, then when you want to run that job again, setup your tools and load the program. This is why they make CAD/CAM software. After about 50 years of manual lathe operation, it took me a while to adjust my thinking for CNC work. I run about 20 different parts on my CNC lathe, the G code for each job is in a folder on the hard drive. Each job has multiple operations. I only have to set up the tools, select the program,.load the material, and press Go. Very simple.

    Use a tool to set 0. Put the stock in the chuck, move a tool to Z 0, move the stock to the tool, tighten the chuck. Work Z 0 is now set. You will be setting up tool offsets in the CNC software.

    Fair enough - I think the only challenge with that is if I release the halfnut that drives the carriage on the lead screw, the controller now has no idea where the carriage actually is. I can imagine it being a pain to re-set that every time but with a simple DRO maybe it's easy enough to "home" it as i alluded to in my first post.
    You are still thinking about driving the carriage with the lead screw, use the rack & pinion instead. The hand wheel drives from a rack & pinion on almost all lathes. Connect the servo motor to the hand wheel, not the lead screw. Use the air spring to bias the carriage to the to the right since most work is done with the carriage moving right to left.. You will never lose position this way. Even better is to use linear scales on the carriage and cross slide and close the loop that way. But this precludes the use of Acorn which is an open loop system only.

    ..... and i'm assuming that I will probably need to do a bunch of operations manually, primarily drilling with the tailstock because of how finnicky my resins and acetates can be... if they heat up and you don't retract quickly they melt onto the drill bit. could be fixed with flood cooling but that's out of my scope for now
    For CNC use, you will want to set up gang tooling (or a turret). You drill bit will be in one of the tool stations, then peck drill. Google ''gang tooling''

    Jim Dawson
    Sandy, Oregon, USA


Tags for this Thread

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  


About CNCzone.com

    We are the largest and most active discussion forum for manufacturing industry. The site is 100% free to join and use, so join today!

Follow us on


Our Brands

Need help - Lathe Build, trying to understand requirements/limitations

Need help - Lathe Build, trying to understand requirements/limitations