The CNC challenge, a different sort of mod
Posted: Sat Jul 21, 2007 6:41 pm
Noticed that Color0 is picking up a CNC mill so I thought I’d start a thread, have been using a small Taig
while I work on getting one that will hopefully do what I really want.
Santa Please: Haas OM-1A
Office Mill; 8" x 8" x 12" (xyz), 5 HP, 30,000 rpm ISO 20 taper spindle with 20 tool carousel, includes Visual Quick Code, 1MB, Beacon Light, coolant pump, high-speed machining, etc.
At a buck under $45K it’s sadly just a little out of my price range so it’s time to mod something
The challenge is to make a home desktop mill that will let you do the same type of precision micro work without robbing a bank first. Enter the Xmod mill otherwise known as the Seig X2 Mini-mill sold by Harbor Freight, Grizzly, and others.
At around $500 the Seig has a great price tag but as far as precision marching goes its pure slop, and even worse it’s manual. However as a solid piece of iron for the foundation of a DIY CNC retrofit project it’s darn near perfect.
I actually started my CNC adventure last year but soon discovered that like modifying micro/mini RC cars sometimes what you order is exactly what you get but not always what you really needed.
Definition: Backlash - play or lost motion between loosely fitting machine parts.
That’s the real enemy, a Seig X2 mill can have up to 0.1” of backlash in the X & Y and I don’t’ even want to think about the Z axis slop. The little Taig CNC desktop mill, approx. $2.3K as a turn key CNC mill, I’m currently using has around 0.02” which although an improvement is far from the ideal since a critical eye can detect discrepancies down to 0.001” and even that can be too much for perfect fitting micro car parts. So basically I want a mill with effectively “0” measurable backlash.
The backlash fix starts with ball screws but like modding our cars success depends on choosing the right parts and how you install them. I needed/wanted that zero backlash ideal but it turns out that most of the ball screw Seig X2 turn key packages and kits although nice and a great price are in what I call the Xmod class, great for making small display panels, 2 to 3 inch parts, and other home CNC projects where a high degree of precision isn’t exactly necessary but not something that will let you compete on the tack with a Mini-Z in other words not precise enough for the micro mini car parts I wanted to make. Most use inexpensive rolled ball screws with lead errors of up to .004 inches per foot along with single ball nuts so that they can still have as much as .02” of backlash making them no more precise than a Taig or Sherline at best.
The simple solution would be precision machine ground zero backlash ball screws with high quality zero backlash ball nuts, but again the price!!! Parts like those are why that Haas costs $45K. Never fear there is usually always a solution you just need to know what it is and why you need it, simple really go for high end rolled ball screws that have a minimum of 0.001” of lead accuracy some are now available at affordable prices that actually exceed the quality of low end machined ball screws, then use two apposing ball nuts so that you can tighten them against each other sandwiching the ball screw threads between them and eliminating the backlash. Takes more room and requires modification of the mill frame itself, along with doubling the cost which is why it’s not popular as a turn key kit or package but in the end can achieve the coveted zero backlash for the X and Y axis. Darn that leaves the Z (up and down), the problem being that there is no easy way to use two apposing ball screws for the Z.
Time to compromise since the Z rarely gets used for precise circles and such instead mostly you plunge down thus only going in one direction and drastically lessening the consequences of not having absolute “0” backlash. For the Z I I went with the high end rolled ball with a “preloaded” ball nut, that single preloaded nut cost more than the two nuts for the other axis but aging it’s forking out more money since on preloaded nut cost more than two standard ball nuts but in this case it’s again worth it even though not a zero backlash arrangement it’s still much better than a single standard ball nut.
Additionally the quality and setup of the mounting hardware is another consideration. Yea I planed on, well hang-down-blocking is the only appropriate term that comes to mind, my precision ball screws by supporting the X & Y on only one side with a single low cost low quality ball bearing instead of on both ends with precision mounting hardware.
Setting up the Taig mill, funny they don’t really come “turn-key” there is still quite a bit of work and learning involved before you’re machining anything, I soon discovered that not all controller electronics are equal. Shouldn’t have been a surprise, after all an Xmod CB is not truly proportional, and the old Iwaver CB although an improvement is in no way equal to the Mini-Z CB, which in turn is not a KoPorpo AD Band Mini-Z CB, and like building a race car to compete at the track a bigger motor is not always a good choice nor is that AD band board necessary. After the type of ball screw mod the motor choice was the second major change to the now lamented blissful ignorance of my throw a lipo in an Xmod and its as “good” as a Mini-Z CNC retrofit learning level. After slogging through books, videos, and browsing CNC forums and the like I started to notice quite a few qualifies in statements about stepper motors, like you can do “almost” as much with a steeper, or “nearly” anything you “might” want to do, those red flags led me to a disturbing discovery almost all of the what I term “Pro” CNC retrofit gurus after starting with stepper motors in their projects later upgraded to servo motors for their own personal home mills, there goes the budget again. BTW just like RC forums CNC forums have good, lopsided, and bad advice, it’s a matter off sorting out the members into their actual skill, experience, and even honesty levels then relating what they have to say to what you want to do always taking anything anyone says with a healthy grain of salt.
Long story but true small desktop mill size servo motors are hard to come by, most end up being high torque DC motors with an encoder thrown on, which although fast and again suitable for larger less precise work can have cogging problems, and for high resolution work I needed a motor made to be a servo motor with a good optical 500 CPR encoder otherwise there would be no point in moving up form a stepper motor.
Ok by this time that $45large price tang on the Haas was starting to look like a deal, not really it turns out that the old “you get what you pay for” saying needs to be balanced with “only if you know what you need” and a few other qualifies. Turns out that yes after selling off most of the parts I originally ordered for my X2 CNC mill retrofit, the new parts did not cost all that much more, although admittedly there is a lot more work involved not only in tracking them down but also in the installation.
After the modified CNC mill is 100% complete I will list each individual part and it’s source, but until I have concrete proof that each part is doing the job I wanted it to I don’t want to recommend any specific parts or choices.
These are the major changes I made to my original X2 mod plans:
1. Not just ball screws but “high precision” ball screws with zero or near zero backlash ball nuts. Along with “precision” ball bearing mounting hardware.
2. Actual servo motors with high quality encoders. Usually recommended for faster return speeds but in this case they are for high resolution accuracy.
3. Reliable servo drivers for glitch free operation, I’m going with individual Gecko G-320’s instead of a ready-built 4-axis all in one controller/ps/driver unit.
4. 54v 15amp power supply, with additional regulated 12 and 5 vdc outputs, for encoders, relays, and other controller hardware.
5. Planned addition of a Gecko G-100 “G-Rex” USB/Ethernet controller module, not cheep at $400 and requires updated controller software, most PC controller programs like Mach-3 are still working on compatible versions so I decided to wait until all of the bugs have been worked out before taking the plunge, but in the end a good controller module will be a must have.
6. An enclosed flood coolant system that not only lubricates and cools the piece while cutting, but washes away the debris cleaning a clean path for the cutting tool to aid in extreme cutting accuracy on extra small “micro-sized” parts as well as a clean cut finished surface without excessive polishing.
7. Vibration dampening, again for precision and finish.
8. Little bells and whistles that are often overlooked but can make all the difference in the world, like; limit switched, computer controlled spindle speed, computer controlled operation of extras like the coolant system, so you don’t have to be present 100% of the time during slow cutting processes.
9. At this time a high quality tool changer is not in the budget but a tool change high switch and program macro will make the whole tool change process much less labor intensive.
What prompted me to make most of the above changes to my original X2 modification plans was visiting some new buds (yea their machinists which apparently causes them to drink Budweiser I hope it’s not catching) and seeing a couple of retrofitted X2 mills in operation, they where great but all except one similar to what I’m describing above where not suitable to the type of work I want to do.
while I work on getting one that will hopefully do what I really want.
Santa Please: Haas OM-1A
Office Mill; 8" x 8" x 12" (xyz), 5 HP, 30,000 rpm ISO 20 taper spindle with 20 tool carousel, includes Visual Quick Code, 1MB, Beacon Light, coolant pump, high-speed machining, etc.
At a buck under $45K it’s sadly just a little out of my price range so it’s time to mod something
The challenge is to make a home desktop mill that will let you do the same type of precision micro work without robbing a bank first. Enter the Xmod mill otherwise known as the Seig X2 Mini-mill sold by Harbor Freight, Grizzly, and others.
At around $500 the Seig has a great price tag but as far as precision marching goes its pure slop, and even worse it’s manual. However as a solid piece of iron for the foundation of a DIY CNC retrofit project it’s darn near perfect.
I actually started my CNC adventure last year but soon discovered that like modifying micro/mini RC cars sometimes what you order is exactly what you get but not always what you really needed.
Definition: Backlash - play or lost motion between loosely fitting machine parts.
That’s the real enemy, a Seig X2 mill can have up to 0.1” of backlash in the X & Y and I don’t’ even want to think about the Z axis slop. The little Taig CNC desktop mill, approx. $2.3K as a turn key CNC mill, I’m currently using has around 0.02” which although an improvement is far from the ideal since a critical eye can detect discrepancies down to 0.001” and even that can be too much for perfect fitting micro car parts. So basically I want a mill with effectively “0” measurable backlash.
The backlash fix starts with ball screws but like modding our cars success depends on choosing the right parts and how you install them. I needed/wanted that zero backlash ideal but it turns out that most of the ball screw Seig X2 turn key packages and kits although nice and a great price are in what I call the Xmod class, great for making small display panels, 2 to 3 inch parts, and other home CNC projects where a high degree of precision isn’t exactly necessary but not something that will let you compete on the tack with a Mini-Z in other words not precise enough for the micro mini car parts I wanted to make. Most use inexpensive rolled ball screws with lead errors of up to .004 inches per foot along with single ball nuts so that they can still have as much as .02” of backlash making them no more precise than a Taig or Sherline at best.
The simple solution would be precision machine ground zero backlash ball screws with high quality zero backlash ball nuts, but again the price!!! Parts like those are why that Haas costs $45K. Never fear there is usually always a solution you just need to know what it is and why you need it, simple really go for high end rolled ball screws that have a minimum of 0.001” of lead accuracy some are now available at affordable prices that actually exceed the quality of low end machined ball screws, then use two apposing ball nuts so that you can tighten them against each other sandwiching the ball screw threads between them and eliminating the backlash. Takes more room and requires modification of the mill frame itself, along with doubling the cost which is why it’s not popular as a turn key kit or package but in the end can achieve the coveted zero backlash for the X and Y axis. Darn that leaves the Z (up and down), the problem being that there is no easy way to use two apposing ball screws for the Z.
Time to compromise since the Z rarely gets used for precise circles and such instead mostly you plunge down thus only going in one direction and drastically lessening the consequences of not having absolute “0” backlash. For the Z I I went with the high end rolled ball with a “preloaded” ball nut, that single preloaded nut cost more than the two nuts for the other axis but aging it’s forking out more money since on preloaded nut cost more than two standard ball nuts but in this case it’s again worth it even though not a zero backlash arrangement it’s still much better than a single standard ball nut.
Additionally the quality and setup of the mounting hardware is another consideration. Yea I planed on, well hang-down-blocking is the only appropriate term that comes to mind, my precision ball screws by supporting the X & Y on only one side with a single low cost low quality ball bearing instead of on both ends with precision mounting hardware.
Setting up the Taig mill, funny they don’t really come “turn-key” there is still quite a bit of work and learning involved before you’re machining anything, I soon discovered that not all controller electronics are equal. Shouldn’t have been a surprise, after all an Xmod CB is not truly proportional, and the old Iwaver CB although an improvement is in no way equal to the Mini-Z CB, which in turn is not a KoPorpo AD Band Mini-Z CB, and like building a race car to compete at the track a bigger motor is not always a good choice nor is that AD band board necessary. After the type of ball screw mod the motor choice was the second major change to the now lamented blissful ignorance of my throw a lipo in an Xmod and its as “good” as a Mini-Z CNC retrofit learning level. After slogging through books, videos, and browsing CNC forums and the like I started to notice quite a few qualifies in statements about stepper motors, like you can do “almost” as much with a steeper, or “nearly” anything you “might” want to do, those red flags led me to a disturbing discovery almost all of the what I term “Pro” CNC retrofit gurus after starting with stepper motors in their projects later upgraded to servo motors for their own personal home mills, there goes the budget again. BTW just like RC forums CNC forums have good, lopsided, and bad advice, it’s a matter off sorting out the members into their actual skill, experience, and even honesty levels then relating what they have to say to what you want to do always taking anything anyone says with a healthy grain of salt.
Long story but true small desktop mill size servo motors are hard to come by, most end up being high torque DC motors with an encoder thrown on, which although fast and again suitable for larger less precise work can have cogging problems, and for high resolution work I needed a motor made to be a servo motor with a good optical 500 CPR encoder otherwise there would be no point in moving up form a stepper motor.
Ok by this time that $45large price tang on the Haas was starting to look like a deal, not really it turns out that the old “you get what you pay for” saying needs to be balanced with “only if you know what you need” and a few other qualifies. Turns out that yes after selling off most of the parts I originally ordered for my X2 CNC mill retrofit, the new parts did not cost all that much more, although admittedly there is a lot more work involved not only in tracking them down but also in the installation.
After the modified CNC mill is 100% complete I will list each individual part and it’s source, but until I have concrete proof that each part is doing the job I wanted it to I don’t want to recommend any specific parts or choices.
These are the major changes I made to my original X2 mod plans:
1. Not just ball screws but “high precision” ball screws with zero or near zero backlash ball nuts. Along with “precision” ball bearing mounting hardware.
2. Actual servo motors with high quality encoders. Usually recommended for faster return speeds but in this case they are for high resolution accuracy.
3. Reliable servo drivers for glitch free operation, I’m going with individual Gecko G-320’s instead of a ready-built 4-axis all in one controller/ps/driver unit.
4. 54v 15amp power supply, with additional regulated 12 and 5 vdc outputs, for encoders, relays, and other controller hardware.
5. Planned addition of a Gecko G-100 “G-Rex” USB/Ethernet controller module, not cheep at $400 and requires updated controller software, most PC controller programs like Mach-3 are still working on compatible versions so I decided to wait until all of the bugs have been worked out before taking the plunge, but in the end a good controller module will be a must have.
6. An enclosed flood coolant system that not only lubricates and cools the piece while cutting, but washes away the debris cleaning a clean path for the cutting tool to aid in extreme cutting accuracy on extra small “micro-sized” parts as well as a clean cut finished surface without excessive polishing.
7. Vibration dampening, again for precision and finish.
8. Little bells and whistles that are often overlooked but can make all the difference in the world, like; limit switched, computer controlled spindle speed, computer controlled operation of extras like the coolant system, so you don’t have to be present 100% of the time during slow cutting processes.
9. At this time a high quality tool changer is not in the budget but a tool change high switch and program macro will make the whole tool change process much less labor intensive.
What prompted me to make most of the above changes to my original X2 modification plans was visiting some new buds (yea their machinists which apparently causes them to drink Budweiser I hope it’s not catching) and seeing a couple of retrofitted X2 mills in operation, they where great but all except one similar to what I’m describing above where not suitable to the type of work I want to do.
I'll be sure to use that one a lot. (The Taig vise I found in the package works quite well for fixturing.)
Thanks LBRC!
. Also because I’m using larger screws and therefore larger ball-nuts than the typical X2 retrofit I needed to cut out some of the table bottom and slide block for the X axis too but fortunately I could do most of that on the Taig.
.