Converting an SX2 Mini Mill to CNC - build thread

Talk about anything and everything here within reason.
User avatar
oatnet
100 MW
100 MW
Posts: 2929
Joined: Apr 26, 2007 3:03 pm
Location: SoCal, USA

Converting an SX2 Mini Mill to CNC - build thread

Post by oatnet » Jan 12, 2013 6:30 pm

I'm converting a mini-mill to CNC, and I wanted to create a build log.  I thought about putting it on CNCZONE as yet another X2 conversion, but since it is intended to make ebike parts anyhow, I decided to put it here on the sphere instead - in the spirit of nechaus' "makerbot" thread. There are a lot of folks here with skills and tooling that far exceed mine and will regard this project as laughably naive, but hopefully there are others who are interested in seeing what is involved.  Usually I post here ONLY based on my direct experience, and that will be the case for the rest of the thread as the build progresses, but this is new ground for me so the first posts are OPINION based on the reading and research I used to design the project.  I am sure I misunderstood some things along the way, so take it as the ravings of a CNC noob trying to learn.

------

Since I took a Machine Tools class a few years ago (my sole experience to date), I have wanted a vertical mill at home, so I could make my mistakes in private, and the tools I break are my own instead of depriving students trying to learn a trade.  I also wanted CNC control, if possible.

During the class I almost bought one of the cheap chinese router/engraver tables that had just hit the market at the time, which today are available on eBay from dozens of vendors.  I'd start out with one of those $800 engravers in my cart, upgrade to the VFD spindle, throw in the bonus 4th axis table, add kings ransom in shipping, and suddenly, you are way past $2k...  I held off long enough to read feedback from others, finally concluding that the mills are better suited for engraving, and were not powerful, rigid, repeatable, or capable enough for the type of milling operations I had in mind.

I realized for the same $2k+ the Chinese engraver cost, I could get a decent large vertical mill - so what if it didn't have CNC.  I started looking at the Sieg X3 clones, then I rationalized in upgrade after upgrade - consider getting 220v service installed at the house, to handle the larger Sieg X4 clones, then the RF-45 clones...  Then I rationalized down in size a few models as a trade off for adding in CNC.  Then I'd work my way back up the size range again until I was at the RF-45 with a CNC kit.  Then I'd figure well since I was spending all this money, I should get a professionally converted mill...

About this tiime, I realized I was planning on spending 10k on a mill, as well as getting 220v service installed and giving up my spot in the garage permanantly, to install a mill to build the occasional ebike part.  I'd get frustrated with the futility of implementing the ultimate mill, and all the time I'd wasted thinking about and researching it, so I'd forget all about it for a few months.  Then some project I'd read about, some part I'd need, would trigger the craving for a mill again, and I'd go through the whole cycle again, from Chinese engraver to 10k CNC.  This repeated for a few years.

--------

Recently, I came across an ad for the Harbor Freight varient of the Sieg X2 mill for $450.  I liked that the owner had done the nasty job of cleaning off the shipping grease, had lapped in the ways already, and was only 10 minutes away. Still, I dithered about it for a few days, starting the usual cycle of rationalizing bigger and better mills, until I had a moment of clarity.  This HF X2 mill was more than up to the most immediate job on my project work list - milling a brake adaptor to mount a 9" disk brake on a Boxxer fork.  Sure, I might get lost in the acres of backlash, and have to go find the edge a few times, but eventually I can locate a spot with precision.  Yes my operations will not be precise to .0005 - but I don't need that level of precision - so I bought it.

I figured I'd see how it fared in my workshop, whether it was a tool I only wanted to have, or a tool I actually wanted to use and grow proficient on.  If I used it enough, I thought I'd do the CNC conversion, which is a strong driving reason for me getting a mill.  I had a taste of G-code in class, I wanted more but I realized because of my learning style it is something I could only pick up on my own, figuring out what I needed for each project until I built up a good body of working knowledge.  

As it turns out, while the HF Sieg X2 mill is a good fit for the home mill I needed, the tilting column makes it a challenge for CNC, as the torque of a cut can move it out of adjustment.   For light-duty manual milling it is not a big deal to tram the head, even between each cut if needed, adjusting with a rubber mallet - but CNC needs to do cut after cut without interaction, and if it gets a little out of whack each time, eventually you get way out of whack and crash the head.  Some people build supports that brace the column to prevent is, and I'd consider that if I had to.

However, there is a single X2 variant, one of the "SX2L" models sold by littlemachineshop.com (LMS), which has a solid column that eliminates the whole movement problem, and therefore is more capable of CNC duty.  It is harder to tram in the head because you have to shim the solid column to the base instead of rotating the adjustable column, but without the ability to rotate it will stay square much longer.  The LMS X2 also has a bigger table, with so much longer travel in all 3 axis that the total work envelope is DOUBLE the size of other X2 mills.  It also comes with belt drive instead of the noisy/weak plastic gears on other X2 clones, and a 500w BLDC spindle motor with a tachometer output, instead of a 350w brushed motor.
SolidColumnLMS-HFAdj2.jpg
HF on the Left, LMS on the right. Note the hollow behind the dovetails on the HF mill. The walls of the LMS mill are visibly thicker, and the area behind the dovetails is not hollowed out, presumably more rigid. Between the backs of trhe two mills, notice the large nut at the base of the HF mill. It secures the column to the base, and the column pivots around this point for angled cuts. This nut does not supply enough tension to maintain alignment between the base and column, but it does make it easy to tram the mill. The LMS column bolts directly to the base, but requires shimming to achieve square.
SolidColumnLMS-HFAdj2.jpg (125.37 KiB) Viewed 31181 times
As you can see from this borrowed pic, the walls of the solid column are thicker than the ones on the tilting column, and not hollow inside the dovetail section, adding rigidity.  LMS sells the column separately for $175, and it can be used to upgrade other X2 mills because their smaller tables' saddles will fit on it. The Solid column does not lose space to the tilting mechanism, allowing for a little extra Y axis travel.  The base of the solid column juts out further, so it's dovetails are lengthened further along the Y axis, also adding travel.

I could overcome the two biggest weaknesses of the HF mill with the solid column upgrade, and replacing the plastic gears with LMS's $145 belt drive upgrade kit.  However, the total cost of parts/cost/shipping/HF mill approaches that of the LMS mill, and I still wouldn't have the bigger work envelope or BLDC spindle.  I decided I would be better off selling the HF mill to someone who needs just a mill, and buy the LMS mill for a CNC conversion.

--------

Once again I started thinking about the 389lb x3, or the 879lb RF-45, and roughly priced doing conversions on them too.  The X3 only cost 40% more after conversion, and had a much bigger work envelope, and is way more rigid and has a knee and can be ordered with a rotating head...  Then I looked at the HF mill I have, the space its flooded enclosure will need, and realized that was already more space than I wanted to give up.  I also realized that I'm getting older and the 139lb mill was just on the edge of what I feel comfortable moving around.  I don't want to have to go through a major 400lb rigging ordeal every time I want to do something to the mill.  Skilled people are tuning X2 mills to get .001 cuts out of them, and I could invest time someday to approach that, but very little of what I expect to make will need more than .010 accuracy.  It can't handle the occasional large piece, but 99% of what I've thought about making will fit into that work envelope.  

That perspective convinced me that a solid-column LMS conversion was the right solution, and despite my earlier plans to wait, I ordered the mill, the conversion kit, and the electronics/software.  I read a lot of posts complaining about how high the liftgate shipping was, but because LMS is pretty close to me, Residential Liftgate Delivery was cheaper than picking it up at a freight company.  After shipping and tax, the LMS 3960 solid column mill was $972.75.
http://littlemachineshop.com/products/p ... 1387807683

My research found many X2 CNC conversions that used the CNCFUSION.com kit, who also offers a special version for the larger table on my LMS mill, so I went with their kit.  The Z axis mount will require me to drill/tap two holes in the column - maybe I can use my HF mill to do that work.  Their design's Z axis has some issues - such as blocking the gib screws, and having the screw mounted to the side causes an alignment issue.  However, it will suffice to get me started, and I can always upgrade it later.   Each axis requires a $24 precision coupling - par for the part - and I'll take them up on $33 to have each ballnut pre-loaded and matched to its screw.  That adds $47 to each of the 3 axis, $141 overall.  Shipped, the precision hardware was $821.00 .
http://www.cncfusion.com/

For the electronics, I settled on the 50v Gecko G540 driver, after reading a lot of documented experiences, it is the sweet spot of price/features/reliability.  I researched and considered Geckos 80v drivers and a BOB (break out board), but the power of the g540 is really sized correctly to the X2's rigidity and the simplified installation a better fit to my skill level. 
http://www.automationtechnologiesinc.co ... xis-driver

I had trouble settling on stepper motors, until I recognized that the 381oz/in Kelig stepper motors many people were praising because their volt/amp requirements neatly fit the G540's ratings, also had super-low inductance and resistance so they run cooler.  Then it clicked - the motor scaled to the gecko driver and both scaled to the mill, and I can use one of my Vicor PSU's to power everything.
http://www.automationtechnologiesinc.co ... -381-oz-in
http://www.kelinginc.net/NEMA23Motor.html

Software is another issue.  CAD lets you design the part.  CAM converts it to G-Gode.  CNC software drives the mill, and executes G-code.  I haven't figured out what I want for cad/cam, but for now I'll just be happy to have the mill under digitial control with CNC software.  I found the $175 Mach3 CNC software at automationtechnologies for $149, and since they had the steppers at the same $49 price as Kelig, and the G540 driver at an typical $279. They were the cheapest place to single-source all the electronics/software on my list, and while I was at it I bought a $10 emergency-stop switch for the controller, and a $16 single motor cable as a model for my own cables.   $640.50 delivered for all the electronics and software.  
http://www.automationtechnologiesinc.co ... cense-file

So, $2,285.25 to buy the LMS solid-column mill and CNC conversion parts, about the same price as a Makerbot 2 but with the ability to work in a wide range of materials in addition to plastic, at 10x the resolution. Another $50 for the PC/keyboard/mouse and $149 for Mach 3 CNC software, so you can build a complete, functional, basic CNC mill 2,484.25 total. Whups, I had a spare monitor and parallel cable, so those costs are not included.

Upgrade costs will include an enclosure with drains, coolant nozzles and pump, home switches.  Then there is the cost of a stand, tooling, CAD/CAM software, more tooling, a 4th axis, more tooling...
-----

I got the crate with the mill on Tuesday, my wife received it for me because Hollywood freight missed their delivery window, and there was a dent in the box.  I mailed pictures of the dent to LMS, who was very supportive, and said if anything was damaged they would replace the part (they stock all the parts). I freed up time to open the crate last night, and the dent appears to stop short of the table, and I have not detected any problems with the mill yet - yay!   I put it up onto the workbench and took comparison pictures, so people could see the difference between the HF X2 and the LMS SX2.  On top of all the benefits I mentioned earlier! in person the LMS SX2 presents better quality.  The column looks bigger and better, and the edges of the dovetails are sharper and consistent, like they were cut better.  The connectors for the wiring are better.  It's all the same parts, just better made in many instances.  I noticed the SX2 table also has a t-slot.cut into the front.

Anyhow, here are pics:
00IMG_9083.JPG
As delivered
00IMG_9083.JPG (97.9 KiB) Viewed 31181 times
01IMG_9091.JPG
The dent
01IMG_9091.JPG (72.9 KiB) Viewed 31181 times
02IMG_9094.JPG
The dent on the inside
02IMG_9094.JPG (46.11 KiB) Viewed 31181 times
03IMG_9095.JPG
With the crate removed
03IMG_9095.JPG (84.41 KiB) Viewed 31181 times
04IMG_9107.JPG
HF (Harbor Freight) X2 on the left in red, LMS (Little Machine Shop) mill on the right in blue.
04IMG_9107.JPG (235.65 KiB) Viewed 31181 times
05IMG_9108.JPG
The LMS table is much larger.
05IMG_9108.JPG (212.96 KiB) Viewed 31181 times
06IMG_9110.JPG
HF has a round black 350w DC spindle motor on top, LMS has a square silver 500w BLDC spindle motor.
06IMG_9110.JPG (211.7 KiB) Viewed 31181 times
07IMG_9111.JPG
The tan box on the back of the LMS houses the motor controller, the corresponding red box on the HF is smaller
07IMG_9111.JPG (207.06 KiB) Viewed 31181 times
08IMG_9112.JPG
The LMS has an additional t-slot in the front of the table
08IMG_9112.JPG (207.3 KiB) Viewed 31181 times
09IMG_9113.JPG
Size Matters
09IMG_9113.JPG (176.51 KiB) Viewed 31181 times
10IMG_9115 NO DIAL.JPG
The base of the HF mill has a mechanism+dial for tilting, this is open space on the LMS mill allowing for longer Y axis travel.
10IMG_9115 NO DIAL.JPG (190.75 KiB) Viewed 31181 times
11IMG_9116.JPG
The spindle speed control layout is different, as are the emergency stop buttons. At first glance, the LMS buttons are more durable, but the HF button is more slappable.
11IMG_9116.JPG (224.9 KiB) Viewed 31181 times
12IMG_9117.JPG
On the LMS, the blue case under the spindle motor houses the pully system that drives the spindle. The HF mill has a corresponding black case, which is thinner and houses an assembly of plastic gears to drive its spindle.
12IMG_9117.JPG (235.15 KiB) Viewed 31181 times
13IMG_9118 COMPARE BASES.JPG
The base of the LMS mill is a lot longer, note the increased distance between the tabs that are used to bolt the mills to the bench.
13IMG_9118 COMPARE BASES.JPG (197.73 KiB) Viewed 31181 times
14IMG_9120 lowered.JPG
The heads lowered.
14IMG_9120 lowered.JPG (212.19 KiB) Viewed 31181 times
Last edited by oatnet on Feb 02, 2013 2:38 pm, edited 1 time in total.
Member 117 of 26,000+

153v DUNE BUGGY © 164v Vectrix E-Moto © 72v Norco A-line/x5403 © 60v Specialized Enduro Comp/BMC © 72v x5tracycle © 48v TF IO/BMC © 36v Kepler FrictionDrive © Spot-Welding a123 © Pelican Front Packs © Vectrix Battery Replacement © MORE !!


Image大鸡巴 Motors
Knowledge is acquired through experience, everything else is just information - Albert Einstein
"Cheater?" I'm not competing, I'm Commuting!

User avatar
oatnet
100 MW
100 MW
Posts: 2929
Joined: Apr 26, 2007 3:03 pm
Location: SoCal, USA

Re: Converting an SX2 Mini Mill to CNC - build thread

Post by oatnet » Jan 13, 2013 4:36 am

.
Although they specified 2 weeks before shipping, CNCFusion has already sent me a tracking number for the precision hardware. I'm not looking forward to the next part, but if I prep the mill now the parts can be installed as soon as they arrive. My tasks will be remove the head/table/saddle to strip the mill down to a bare column, clean the waxy shipping oils from every surface of with low-VOC mineral spirits, then coat the ways with white lithium grease and everything else with Mobil 1 synthetic non-detergent motor oil (oils with detergents will cause wear).

As Michael from CNCFusion joked, these are "Snowflake Mills - no two alike" so I anticipate some challenges fitting the precision hardware. If I am lucky, the better build quality I see on the LMS mill will result in more accurate tolerances that reduce the number and magnitude of those challenges. I'll use this detailed photo-rich article about installing the CNCfusion kit to guide me through the teardown, as well as removal of the many parts made obsolete by CNC Conversion. That guide did the best job of getting me comfortable with what I will be doing with the CNCFusion kit, but there are additional articles from CNCFusion here, here, and here.

Once the mill is reassembled with precision ballscrews and motor mounts, I can bolt on the steppers, wire them to the g540 driver, install/configure MACH 3 on the CNC computer, and start to dial everything in.



[attachment=0]KL23H2100-35-4B.pdf[/attachment]



IMG_9146.jpg
Above is a picture of the Electronics order.

Upper left - (3) 381oz/in double-shaft stepper motors, two still wrapped. The 2nd shafts can be used for handwheels, although nobody really uses them after CNC conversion - digital control is much easier, plus the ballscrews have such low friction that they won't hold position so I'd really have to crank down the gib locking screw with varying degrees of accuracy. Handles also increase the square footage requirement of my enclosure. Maybe I'll machine the shafts off, or maybe I'll leave them for some future vague and redundantly usless closed-loop encoder project.

Lower left - Gecko g540 drive with integrated BOB (break out board) and enclosure. The parallel port connects to your computer to receive the control signals sent by the MACH3 software. Each of the (4) db-9 connectors on the front provides power and control for a stepper motor/axis. The first three connectors control the X-Y-Z axis, the 4th can be used for a rotary table, or a dual-screw Z axis. The green terminal strip along the bottom of the case supplies connectons for the emergency stop switch, spindle speed control, to drive relays for limit/home switches etc, and a couple of other things I forget. I believe this is the current "Rev 8" model because the trim adjustments are on the front next to the DB9 connectors; apparently Rev's 6 and 7 were not marketed, and the earlier versions had the trim adjustments somewhere difficult to access. I'll probably need to design a cooling system for the drive too.

Upper right - The gecko DB9 4-wire stepper motor cable I got as a visual aid for making the others, on top of the vacuum-packed bag of (4) db9 connector kits that come with the Gecko g540. I'll buy some 4-conducter stranded cable and make my own cable for the Y/Z axis. These cables carry power and control to the steppers.

Lower right - Emergency stop button- used to instruct the g540 to power down all steppers and the spindle.

I bought a MACH3 license on this order too, and got the license key via email. I haven't installed it yet because I haven't built the PC that will be doing CNC control. It requires a PC old enough to have a parallel port and not much horsepower. I have an old PC that should be suitable, but it needs a new hard drive and OS. I am considering cutting a hole in the side of the PC case to mount the g540 drive right in it.
.
IMG_9146.jpg (240.53 KiB) Viewed 34594 times

IMG_9148.jpg
Above is a picture of tooling procured from Shars and eBay.

Upper left - 3-piece T-handle tap wrench set for tapping threads, and a pair of 1-2-3 blocks for securing work.

Lower left - 6" x 3/16" precision parallel set (varying heights) for holding work squarely in a vice.

Center, top to bottom - HSS 3/8" center drill/countersink, HSS 1/4" drill bit, 1/2" R8 collet, 3/8" R8 collet, 1/4" R8 collet, tap guide (spring loaded center, fits in niche on top of T-handled tap wrenches to provide precision alignment square to the drilled hole), (2) 1/2" HSS 2-flute square corner end mills I'll use on aluminum, 1/2" center/edge finder (one of my favorite tools)

Upper right - M2.5 - M12 metric tap set, these things are SHARP. I need a matching counterbore set so I can recess socket-head screws.

Lower right - Genuine Kurt D30 3" precision vice from eBay - 3" is the right proportional size for an X2 mill, and you can't beat a Kurt vice, although their 3" version is rare and hard to find. I'll use the vice for manual (albeit CNC controlled/pendant?) milling, I'll use fixtures/tooling plates for CNC operations.
.
IMG_9148.jpg (229.54 KiB) Viewed 34594 times

IMG_9150.jpg
The supervisor inspecting my post photo-shoot cleanup.
.
IMG_9150.jpg (118.13 KiB) Viewed 34594 times
Attachments
KL23H2100-35-4B.pdf
Above is a link to the Spec sheet/wiring diagram for the stepper I'm using, and below are some additional details:

NEMA 23 Stepper Motor: KL23H2100-35-4B (1/4” Dual shaft with a flat)
381 oz In. Hybrid Motor
1.8° /200 Steps Per Rev.
3.5 Amps Current Per Phase
4-wire Bi-polar
.
(56.56 KiB) Downloaded 2039 times
Member 117 of 26,000+

153v DUNE BUGGY © 164v Vectrix E-Moto © 72v Norco A-line/x5403 © 60v Specialized Enduro Comp/BMC © 72v x5tracycle © 48v TF IO/BMC © 36v Kepler FrictionDrive © Spot-Welding a123 © Pelican Front Packs © Vectrix Battery Replacement © MORE !!


Image大鸡巴 Motors
Knowledge is acquired through experience, everything else is just information - Albert Einstein
"Cheater?" I'm not competing, I'm Commuting!

User avatar
nicobie
10 MW
10 MW
Posts: 2425
Joined: Aug 07, 2008 1:48 pm
Location: Central Coast CA,USA

Re: Converting an SX2 Mini Mill to CNC - build thread

Post by nicobie » Jan 13, 2013 2:29 pm

Looks like a fun project. Is the mill big enough to work steel, or is aluminum / plastic all you will be milling?
Image

May your tote always stay tight and your edge eversharp :wink:

my eTownie build thread: https://endless-sphere.com/forums/viewto ... =6&t=23701

User avatar
oatnet
100 MW
100 MW
Posts: 2929
Joined: Apr 26, 2007 3:03 pm
Location: SoCal, USA

Re: Converting an SX2 Mini Mill to CNC - build thread

Post by oatnet » Jan 13, 2013 5:27 pm

nicobie wrote:Looks like a fun project. Is the mill big enough to work steel, or is aluminum / plastic all you will be milling?
Thanks!

I have seen youtube vids demonstrating that even the tiliting column X2 will mill steel, it is just a matter of how fast it can do it. The steppers I chose have abundant torque, so they can press the tool into the workpiece with great force resulting in great speed. However when the mill pushes too hard, it makes the column flex until the tool skips along the workpiece, risking damage and leaving a rough surface. Since the LMS's column is thicker, and better braced by bolts at (4) dispersed points instead of the tilting column's single point, I hope the added rigidity will let me mill steel even faster than I saw in those vids. Of course, locating/drilling/tapping holes in steel is a slam dunk, like a precision drill press.

After the mill is assembled, I'll be able to assess how deep I can cut and how fast I can feed steel before it induces chatter. If I want more speed, I can increase feed rate by changing the pulley on the belt drive to drive a carbide bit at higher RPM, or upgrade to a faster/more powerful spindle motor. I also have been percolating on an upgrade that will add rigidity by bracing the top of the column (actually the side benefit of an upgrade to increase usable z-axis travel).

That said, most of what I think about milling is aluminum, with plastic test runs. However, like you, I still found it important to make a machine that would CNC steel fast enough so that programs don't take days to complete. I think I've met that design criteria, stay tuned and when the mill is built I'll try to prove that out.

-JD
Member 117 of 26,000+

153v DUNE BUGGY © 164v Vectrix E-Moto © 72v Norco A-line/x5403 © 60v Specialized Enduro Comp/BMC © 72v x5tracycle © 48v TF IO/BMC © 36v Kepler FrictionDrive © Spot-Welding a123 © Pelican Front Packs © Vectrix Battery Replacement © MORE !!


Image大鸡巴 Motors
Knowledge is acquired through experience, everything else is just information - Albert Einstein
"Cheater?" I'm not competing, I'm Commuting!

User avatar
oatnet
100 MW
100 MW
Posts: 2929
Joined: Apr 26, 2007 3:03 pm
Location: SoCal, USA

Re: Converting an SX2 Mini Mill to CNC - build thread

Post by oatnet » Feb 02, 2013 1:50 pm

Progress over the past few weeks. First step its to tear the mill down and clean off the thick shipping grease. Then lube it up, mount the precision hardware, setup the electronics, etc etc.
15 IMG_9189.JPG
The Precision hardware from CNCfusion, as packaged for shipping. Note the upper left corner, where my assistant is pointing out the X-Axis motor mount.
15 IMG_9189.JPG (118.06 KiB) Viewed 30912 times
16 IMG_9192.JPG
Top to Bottom:
Z axis assembled unit - includes motor mount, zero-backlash coupling, Ballscrew, and ballnut block that bolts to the spindle head.
Y Axis ballscrew, ballnut, Motor Mount and coupling. One of the stepper motors is included for scale.
Cap Screws/bits supplied installation, X axis Motor mount, X axis ballnut spacer
X axis Ballscrew with ballnut and coupling
16 IMG_9192.JPG (108.11 KiB) Viewed 30912 times
17 IMG_9244.JPG
The first step is to disassemble the mill and clean off the shipping grease it is saturated with. I started by removing the spindle controller + box. With the back of the box removed, I had ready access to (2) bottom screws that secure it to the column, but the other (2) are hidden behind the BLDC controller's circuit board.
17 IMG_9244.JPG (109.57 KiB) Viewed 30912 times
18 IMG_9245.JPG
With the PBM loose, you can see the two remaining screws at the top.
18 IMG_9245.JPG (117.86 KiB) Viewed 30912 times
19 IMG_9246.JPG
Back of the column with the Controller Box removed
19 IMG_9246.JPG (95.85 KiB) Viewed 30912 times
20 IMG_9247.JPG
After unscrewing the stop at the top of the column, I could lift the head right off. This pops off the plastic plate, which comes off anyhow.
20 IMG_9247.JPG (82.32 KiB) Viewed 30912 times
21 IMG_9250.JPG
Both handwheels are off, as is the Y axis handle block and the table's left plate.
21 IMG_9250.JPG (109.12 KiB) Viewed 30912 times
22 IMG_9252.JPG
After removing the cap screws on the right side bearing mount the table slides right off
22 IMG_9252.JPG (77.55 KiB) Viewed 30912 times
23 IMG_9255.JPG
The right side bearings. The instructions said to wiggle it off, not so easy, and I spent more time removing this bearing than I did the rest of the disassembly. I could get the bearing block almost off, then it wouldn't budge, and finally resorted to a gear puller. Save yourself time, just use a gear puller. Note the shipping grease coating the bearing block.
23 IMG_9255.JPG (65.06 KiB) Viewed 30912 times
24 IMG_9257.JPG
Mill, disassembled. It took me a little over an hour to complete the teardown, It would have been a half hour if I had used a gear puller in the last step. The rusty-looing spots are actually gobs of grease.
24 IMG_9257.JPG (123.35 KiB) Viewed 30912 times
25 IMG_9265.JPG
Here, all of the parts have been carefully cleaned with mineral spirits to remove the grease. When I carried each component outside to be cleaned, I was glad I got the X2 mill instead of the X3, because I would have needed help carrying anything heavier. Unprotected surfaces can rust quickly, so it is important to lube them up as soon as possible.
25 IMG_9265.JPG (107.21 KiB) Viewed 30912 times
255 IMG_9323.JPG
These are the parts that are replaced by precision CNC hardware. Too bad I couldn't buy a LMS model without this stuff and save the extra expense, no idea what I will do with it now.

From the top left:
X axis leadscrew Assembly
Z axis ruler (see note below)
Y axis leadscrew assembly
Z axis Pinion/ fine adustment Assembly and head stop
Torsion Arm, Z axis rack, (3) gib stops, column cover


Note: Although the spindle head has its own gib stop screw, a second gib stop slides up and down the Z axis, independant of the spindle head. It is useful when removing/working on the spindle head so I wanted to keep it, but its locking screw is on the right side so it interferes with the Z-ballscrew. With the spindle head off, I slid the gib stop off, flipped it, and put it back on with the locking screw facing the left instead. Normally this would interfere with the Torsion arm, but the Torsion assembly was removed for CNC, so that was not a problem. I found it also interfered with the Z-Axis ruler, so I removed that too. I might drill out the mounting holes on the ruler so it will sit further back and clears the gib stop, I'll have to see if I care.

Seems that most people leave the Z-Axis rack in place, but I thought it was one more thing for swarf to get caught on, so I removed it.
255 IMG_9323.JPG (92.68 KiB) Viewed 30912 times
Member 117 of 26,000+

153v DUNE BUGGY © 164v Vectrix E-Moto © 72v Norco A-line/x5403 © 60v Specialized Enduro Comp/BMC © 72v x5tracycle © 48v TF IO/BMC © 36v Kepler FrictionDrive © Spot-Welding a123 © Pelican Front Packs © Vectrix Battery Replacement © MORE !!


Image大鸡巴 Motors
Knowledge is acquired through experience, everything else is just information - Albert Einstein
"Cheater?" I'm not competing, I'm Commuting!

User avatar
oatnet
100 MW
100 MW
Posts: 2929
Joined: Apr 26, 2007 3:03 pm
Location: SoCal, USA

Re: Converting an SX2 Mini Mill to CNC - build thread

Post by oatnet » Feb 02, 2013 1:59 pm

26 IMG_9269.JPG
The base and the saddle lubed up with white lithium grease. What I've read makes me think I need White Lithium grease to lube the ways annually (Philistine, note the last word has a "nu" in it, didn't mean to get you excited :D ), and non-detergent Mobil 1 on everything else daily. I coated all of the hidden surfaces with grease to protect them.
26 IMG_9269.JPG (96.33 KiB) Viewed 30909 times
27 IMG_9301.JPG
I slid the saddle on and tilted the base has been tilted back on the column, so I could install and the Y axis ballscrew/ballnut that is sitting in front. You can see the round R8 spindle through the hole in the base.
27 IMG_9301.JPG (118.16 KiB) Viewed 30909 times
28 IMG_9302.JPG
I slid the saddle on and tilted the base has been tilted back on the column, so I could install and the Y axis ballscrew/ballnut that is sitting in front. You can see the round R8 spindle through the hole in the base.
28 IMG_9302.JPG (87.61 KiB) Viewed 30909 times
29 IMG_9303.JPG
Y axis motor mount/enclosure has been installed at the top of the picture.
29 IMG_9303.JPG (76.93 KiB) Viewed 30909 times
30 IMG_9304.JPG
One of the zero-backlash connectors supplied by CNCfusion
30 IMG_9304.JPG (56.91 KiB) Viewed 30909 times
31 IMG_9305.JPG
Motor mounted, ballnut block adjusted, Y-axis assembly roughed in.
31 IMG_9305.JPG (53.87 KiB) Viewed 30909 times
32 IMG_9309.JPG
Top view of the completed Y-axis assembly. OMG the steppers are enormous. I put the Z-Axis fine-feed adjustment dial on the 2nd shaft to move the Y axis for testing. I like the X/Y/Z axis diagram on the motor mount enclosure.
32 IMG_9309.JPG (63.37 KiB) Viewed 30909 times
33 IMG_9310.JPG
Y axis installed. The right-side bearings that were a bitch to remove, easily slid right onto the ballscrew. The extension on the right side allows clearance for the larger ballnut. Note on the left side, the motor does not seat firmly against the motor mount. The holes for the motor mount screws were drilled plenty deep, but only tapped halfway down.
33 IMG_9310.JPG (100.18 KiB) Viewed 30909 times
34 IMG_9319.JPG
Motor flange on top, compared to the cap screws in as far as the holes are threaded, shows the problem. Luckily, I had bought a set of metric drills/taps to make the brake adaptor, and these seemed to be M4 screws. My M4 tap was really tight though, as it turns out it was probably just an issue of my brand-new tap vs CNCfusion's tap that has been worn down by usage. I sent CNCfusion an email with a few questions, one of them pointing out the problem and asking them to verify that it was an M4 tap; they answered one of the questions but nothing about the M4.
34 IMG_9319.JPG (32.77 KiB) Viewed 30909 times
35 IMG_9338.JPG
I decided to try the M4 tap, and it worked. Here I am tapping the last hole. I ordered a set of (3) t-handled tap wrenches from Shars, but they were really odd, and didn't fit any of the taps. I was able to thread the tap in by hand, and used a mini-crescent wrench to tap it the rest of the way.
35 IMG_9338.JPG (77.21 KiB) Viewed 30909 times
36 IMG_9340.JPG
The Y-axis motor then bolted right on. Looks like I didn't take many pictures during the Z-axis installation. I used the z-axis gib stop to support the head as I lowered it onto the column, and bolted the Z-axis ballnut assembly to it. One of the existing holes in the head that is used for this was really shallow, so it had to be drilled out and tapped to accept the mounting screw. It did not drill easily, so maybe the manufacturer broke a tap off in it. I only have (1) screw securing the Z-axis motor mount at the top, I still have to drill holes in the column for the other (2), but I wanted to have everything moving before I commit to the location of those holes.
36 IMG_9340.JPG (77.66 KiB) Viewed 30909 times
37 IMG_9344.JPG
From the other side. You can see the z-axis gib stop just above the bellows, with the locking screw now pointing left. Between tooling and workpiece height, it is hard to see how it would ever interfere.
37 IMG_9344.JPG (82.24 KiB) Viewed 30909 times
38 IMG_9352.JPG
The converted LMS mill really dwarfs the HF mill now
38 IMG_9352.JPG (102.51 KiB) Viewed 30909 times
39 IMG_9354.JPG
The column at full height, and the Kurt precision vice mounted on the table, all the gibs adjusted.
39 IMG_9354.JPG (80.38 KiB) Viewed 30909 times
Currently, the Z-Axis ballnut seems wierd. Ballnuts have a channel that takes a ball bearing out of one end, and recirculates it to the other end. When it is ready to recirculate one of these ball bearings, I get high resistance until I hear/feel one of the bearings snap into the channel, then it moves smoothly until the next one is ready to pop. I can move the head up or down smoothly over a range of an inch or two without causing this circulation, but as soon as one of the bearings needs to pop, I hit resistance. CNCfusion told me where to look to find instructions to try to resolve this, but I haven't gotten to that yet. The X-axis feels a little catchy too, so maybe this is normal behaviour, but the Y-axis is smooth as silk so I have concerns.

To run the Mach 3 CNC software, I picked up a used Pentium D computer at a local swap meet for $50, 250gb drive, 2gb ram, and a Microsoft XP license tag on the side. I had to install XP from scratch; haven't done it for a few years, seems that Microsoft has tried to obsolete XP-SP2 by keeping Windows Update from installing correctly. I had to manually install Windows Update, then it was fine. None of HP's drivers worked, even though they were specified for the exact model I have - I had to go to Intel's website and install their generic drivers. I've had this problem with HP before - I believe more attempts to obsolete hardware by confusing end-users. Anyhow, after hours and hours of drivers and M$oft patches and virus scanner, Mach 3 installed smoothly and picked up the license file I had stored on the root of C:

I ran a straight-through parallel cable to the Gecko G540 drive. I configured a 48v PSU from one of the Vicor Megapacks I've been using to charge batteries, and powered up the G540 to the anticipated fault light. Then I installed the emergency stop switch, and the G540 powered up to the green ready light.

I haven't hooked up the motors yet. The Gecko manual uses nomenclature I am not used to, it refers to the (4) wire connections as "A", "A/", "B", and "B/". I assumed that A=A+ and A/=A-, but that completely reverses the wire colors on the "sample cable" I bought, and looking at other people wiring the same motors are referring to "A/" as "A+". If anyone has insight on the "/" nomenclature, I would sure appreciate the tip.

I guess polarity doesn't matter as long as I am connecting both "A" leads to the same coil, if I do it backwards on both coils, then the motor should just run backwards... Still, not fully understanding, I find it hard to proceed. Soon I'll just go with it anyhow.
Member 117 of 26,000+

153v DUNE BUGGY © 164v Vectrix E-Moto © 72v Norco A-line/x5403 © 60v Specialized Enduro Comp/BMC © 72v x5tracycle © 48v TF IO/BMC © 36v Kepler FrictionDrive © Spot-Welding a123 © Pelican Front Packs © Vectrix Battery Replacement © MORE !!


Image大鸡巴 Motors
Knowledge is acquired through experience, everything else is just information - Albert Einstein
"Cheater?" I'm not competing, I'm Commuting!

User avatar
oatnet
100 MW
100 MW
Posts: 2929
Joined: Apr 26, 2007 3:03 pm
Location: SoCal, USA

Re: Converting an SX2 Mini Mill to CNC - build thread

Post by oatnet » Feb 02, 2013 2:04 pm

Oh, and since Nicobie asked about the X2 milling steel - here are a couple of random youtubes of SOMEONE ELSE'S X2 milling steel... I have a way to go yet:



Member 117 of 26,000+

153v DUNE BUGGY © 164v Vectrix E-Moto © 72v Norco A-line/x5403 © 60v Specialized Enduro Comp/BMC © 72v x5tracycle © 48v TF IO/BMC © 36v Kepler FrictionDrive © Spot-Welding a123 © Pelican Front Packs © Vectrix Battery Replacement © MORE !!


Image大鸡巴 Motors
Knowledge is acquired through experience, everything else is just information - Albert Einstein
"Cheater?" I'm not competing, I'm Commuting!

User avatar
Farfle
10 MW
10 MW
Posts: 2090
Joined: Jun 27, 2010 10:42 am
Location: Moscow, ID
Contact:

Re: Converting an SX2 Mini Mill to CNC - build thread

Post by Farfle » Feb 02, 2013 2:35 pm

Beautiful, that machine is going to be awesome. Not so much important with the cnc, but I couldn't imagine milling things without a rotary table.
The race bike:
24s5p 50c nano tech
Badass custom mofo outrunner 205x177mm 26kv
Peak power in : 420A at 87v
Peak power out: 34 Hp and 105ft/lb

Build to last...
"It will be assaulted by the elements, It will be ravaged by time and it will be destroyed by the user. BUT, if you Stick to the little rules: K.I.S.S., Overbuild everything, and test, review, revise, repeat. It will last"


Team Farfle blog!! http://www.farfleselectrics.blogspot.com

User avatar
nicobie
10 MW
10 MW
Posts: 2425
Joined: Aug 07, 2008 1:48 pm
Location: Central Coast CA,USA

Re: Converting an SX2 Mini Mill to CNC - build thread

Post by nicobie » Feb 03, 2013 2:01 pm

I wouldn't have thought that such a little machine would be capable of doing steel. Looks like you picked a winner!
Image

May your tote always stay tight and your edge eversharp :wink:

my eTownie build thread: https://endless-sphere.com/forums/viewto ... =6&t=23701

User avatar
oatnet
100 MW
100 MW
Posts: 2929
Joined: Apr 26, 2007 3:03 pm
Location: SoCal, USA

Re: Converting an SX2 Mini Mill to CNC - build thread

Post by oatnet » Feb 03, 2013 11:39 pm

Got the X and Y axis powered up and moving today, yay! I used the HF mill to drill/tap holes in the x-axis motor mount, so I I could mount the wire-track for the X-axis. The Y axis and Z axis motors stay in fixed location, so they don't need track. I freehand-drilled/tapped the column for other side of the wire track. Then I sorted out the motor wiring and Mach 3 configuration/operation basics enough to get X/Y moving, as well as a feel for the limits of travel.

I still have to wire the Z axis, and I haven't tuned the motors yet, or learned how to adjust the speed in mach3. Here are two long vids of it simultaneously moving X + Y, just reaallly slooowwwlly.

-JD



Member 117 of 26,000+

153v DUNE BUGGY © 164v Vectrix E-Moto © 72v Norco A-line/x5403 © 60v Specialized Enduro Comp/BMC © 72v x5tracycle © 48v TF IO/BMC © 36v Kepler FrictionDrive © Spot-Welding a123 © Pelican Front Packs © Vectrix Battery Replacement © MORE !!


Image大鸡巴 Motors
Knowledge is acquired through experience, everything else is just information - Albert Einstein
"Cheater?" I'm not competing, I'm Commuting!

scriewy
100 W
100 W
Posts: 280
Joined: Dec 02, 2009 12:37 am
Location: tel aviv, israel
Contact:

Re: Converting an SX2 Mini Mill to CNC - build thread

Post by scriewy » Feb 04, 2013 6:42 pm

You saying the scriewballs problem went away by it self ?

nice toy oat (:

is there a difference between a rotary table and a dividing head ?

i can mill steel as well with my getho drill prss lol
it's a pitty i wont stay to see the world crumble, u jedi Scum

clyte 405,
went through battle, bloody resurected Dewalts 4x28v 2s2p usable 2.7ah, 52v drops to 44v at 54kmh
78v 2ah Vdrop to 58v at max 66kmh.
72kg rider + 23kg bike.
keywin 20a 30-70v mod to 6 4110 as methods advised for noobs, shunted 38a
clyte 24-72v 40a mod 4310 to 4110, shunted to 77a
17.5.11 - 5000km
at 78v 4ah 72.7kmh Vdrop to 68v
10.11 - 82.1kmh
30.8.12-17000km

a place for lasers as ES is for e-vehicles - laserpointerforums.com

User avatar
Farfle
10 MW
10 MW
Posts: 2090
Joined: Jun 27, 2010 10:42 am
Location: Moscow, ID
Contact:

Re: Converting an SX2 Mini Mill to CNC - build thread

Post by Farfle » Feb 04, 2013 9:23 pm

A dividing head typically moves in two axis, and more often than not has a chuck. A rotary table is always just one axis, and has slots like a standard mill table
The race bike:
24s5p 50c nano tech
Badass custom mofo outrunner 205x177mm 26kv
Peak power in : 420A at 87v
Peak power out: 34 Hp and 105ft/lb

Build to last...
"It will be assaulted by the elements, It will be ravaged by time and it will be destroyed by the user. BUT, if you Stick to the little rules: K.I.S.S., Overbuild everything, and test, review, revise, repeat. It will last"


Team Farfle blog!! http://www.farfleselectrics.blogspot.com

User avatar
rkosiorek
1 MW
1 MW
Posts: 1778
Joined: Jan 18, 2007 2:55 pm
Location: Belleville, Ontario Canada

Re: Converting an SX2 Mini Mill to CNC - build thread

Post by rkosiorek » Feb 07, 2013 1:10 pm

nice job.

be interesting to see if there are any issues driving the "Z" axis. gravity helps the head going down and opposes the head going up. that means that the acceleration during "Z" axis contouring will be different depending on whether you are plunging or withdrawing the tool. that could affect the accuracy of your vertical contours that you may not be able to resolve by simple tuning.

rick
DISCLAIMER:
We will not be responsible for damage to equipment, your ego, blown parts, county wide power outages, spontaneously generated mini (or larger) black holes, planetary disruptions, or personal injury that may result from the use of this material.
- Samuel M. Goldwasser -

User avatar
Farfle
10 MW
10 MW
Posts: 2090
Joined: Jun 27, 2010 10:42 am
Location: Moscow, ID
Contact:

Re: Converting an SX2 Mini Mill to CNC - build thread

Post by Farfle » Feb 07, 2013 1:18 pm

rkosiorek wrote:nice job.

be interesting to see if there are any issues driving the "Z" axis. gravity helps the head going down and opposes the head going up. that means that the acceleration during "Z" axis contouring will be different depending on whether you are plunging or withdrawing the tool. that could affect the accuracy of your vertical contours that you may not be able to resolve by simple tuning.

rick
It shouldnt, good cnc servos will be able to slam the head around like it weighs nothing, and for some reason it can't, then it will slow down the other axis to match it. On the cnc at the school if you are loading a servo enough for it to bog like that, the mill throws an error amd shuts down. Normally only happens if you leave the axis locked.
The race bike:
24s5p 50c nano tech
Badass custom mofo outrunner 205x177mm 26kv
Peak power in : 420A at 87v
Peak power out: 34 Hp and 105ft/lb

Build to last...
"It will be assaulted by the elements, It will be ravaged by time and it will be destroyed by the user. BUT, if you Stick to the little rules: K.I.S.S., Overbuild everything, and test, review, revise, repeat. It will last"


Team Farfle blog!! http://www.farfleselectrics.blogspot.com

User avatar
rkosiorek
1 MW
1 MW
Posts: 1778
Joined: Jan 18, 2007 2:55 pm
Location: Belleville, Ontario Canada

Re: Converting an SX2 Mini Mill to CNC - build thread

Post by rkosiorek » Feb 07, 2013 2:56 pm

weight means something but mass means much more.

these are open loop stepper motors. there is no feedback mechanism or feedback loop to continuously adjust or correct any deviation from expected behavior or motion. the stepper motor is commanded to move but there is no guarantee that the table or head will follow.

your mills at school are likely commercial mills and those are servo systems. they have a feed back mechanism that uses either glass scales on the axis or rotary encoders on the ball screws to measure if the table moves when and only by as much as it is supposed to. the feedback corrects position and speed errors, things that simple stepper controllers are just not capable of. feedback loops are also essential for overload detection. the simple stepper systems do not have this luxury.

counterbalance is important. many mills have either counterbalance weights and pulleys or spring/hydraulic mechanisms hidden inside the vertical column. it all depends on the design. but the truly massive and rigid mills with a moving head or gantry, that you could take a big beefy cut with were likely to have something to compensate for the weight of the head. in our shop we had some mills by Cincinnati Millacron and Mori Seiki that did, but others by Bridgeport and Moog that did not. On larger rail or gantry mills i can't remember one that did not. our largest mill was a 5 axis Cinci 2 head gantry mill with 32x8x4feet (10,000X2,500x1,250mm) travel. and it definitely had a pair of massive weights balancing the weight of the gantry.

rick
DISCLAIMER:
We will not be responsible for damage to equipment, your ego, blown parts, county wide power outages, spontaneously generated mini (or larger) black holes, planetary disruptions, or personal injury that may result from the use of this material.
- Samuel M. Goldwasser -

User avatar
etard
10 MW
10 MW
Posts: 2074
Joined: Aug 03, 2008 3:28 pm
Location: Redlands, CA

Re: Converting an SX2 Mini Mill to CNC - build thread

Post by etard » Feb 09, 2013 12:51 am

Looks good so far man! Here is another mill conversion, worth signing up to see the pics. He also has a nice thread on the CNC lathe conversion of a 9 x 20.

http://www.hobby-machinist.com/showthre ... conversion
Four wheels moves the body, Two wheels moves the soul
Thanks to Justin @ http://www.ebike.ca He brings the soul to ES

User avatar
hillzofvalp
10 MW
10 MW
Posts: 2181
Joined: Dec 26, 2010 2:09 am
Location: Los Angeles, CA

Re: Converting an SX2 Mini Mill to CNC - build thread

Post by hillzofvalp » Feb 17, 2013 11:35 pm

I used to want to convert a small mill.. but seems like a more rigid, massive machine is the best option if you want to get good quality work (aka machine product is what your model is). Obviously for most bike accessories it's not necessary.. but any time you don't want to spend Saturday afternoon and evening to get a press fit .5" deep bearing bore in a hunk of steel...I'd probably get the massive guy.

I'm not at any point to buying one, but someday I'd like to save up for an entry level Tormach.

http://www.tormach.com/product_pcnc_770_main.html

User avatar
oatnet
100 MW
100 MW
Posts: 2929
Joined: Apr 26, 2007 3:03 pm
Location: SoCal, USA

Re: Converting an SX2 Mini Mill to CNC - build thread

Post by oatnet » Mar 01, 2013 1:16 pm

hillzofvalp wrote:I used to want to convert a small mill.. but seems like a more rigid, massive machine is the best option if you want to get good quality work (aka machine product is what your model is)
I think you are confusing speed with quality. A larger mill gives a bigger work envelope, and the extra ridigity allows it to take deeper cuts that translates to speed, but I have pictures below of mirror-smooth cuts that are accurate to .001, how do you get better "quality"?

That said, I can't imagine it taking more than 45 minutes or so mill out a slot for a 1/2" deep bearing an inch or two across. Even if it took all 'Saturday and Evening" as you assert, I'd rather spend that time than 400% more $$$, the expense of installing 220v, give up my parking spot, and rig up a lift every time I want to take the head off the column. If I did, I would do an RF-45 or SX-4 conversion.

It's all about tradeoffs - sure, the Ferrari is nicer than the volkswagen, but the car payment/insurance are much higher and you can't park it on the street. If you need groceries more than track time, the VW is probably a better choice.
etard wrote:Looks good so far man!
Thanks Etard! :mrgreen:
rkosiorek wrote:nice job. be interesting to see if there are any issues driving the "Z" axis. gravity helps the head going down and opposes the head going up. that means that the acceleration during "Z" axis
Thanks Rick! Well, I am a ways from doing conturing :oops: but I'll be on the lookout for that. I read a few reports of people taking their counterweight/air lift mods off their X2 mill after CNC conversion. My research didn't turn up problems with the Z-axis, even on conversions using 220oz-in steppers, which are probably more than a mill this size needs. I hope that by using 380oz-in steppers I'll won't have an issue - the spindle head is nothing like the size of the monsters you are used to working on. :lol:

-JD
Member 117 of 26,000+

153v DUNE BUGGY © 164v Vectrix E-Moto © 72v Norco A-line/x5403 © 60v Specialized Enduro Comp/BMC © 72v x5tracycle © 48v TF IO/BMC © 36v Kepler FrictionDrive © Spot-Welding a123 © Pelican Front Packs © Vectrix Battery Replacement © MORE !!


Image大鸡巴 Motors
Knowledge is acquired through experience, everything else is just information - Albert Einstein
"Cheater?" I'm not competing, I'm Commuting!

User avatar
oatnet
100 MW
100 MW
Posts: 2929
Joined: Apr 26, 2007 3:03 pm
Location: SoCal, USA

Re: Converting an SX2 Mini Mill to CNC - build thread

Post by oatnet » Mar 01, 2013 1:52 pm

I've made a lot of progress on the mill over the past few weeks, just slow on posting it. Here is a video of it moving in all 3 axis:

IMG_9528.JPG
I got some use out of the HF mill, drilling/tapping holes on the back of the X-Axis motor mount to mount the flexible cable track. Here is the first hole after being center-drilled for accuracy, then drilled to size.
IMG_9528.JPG (63.06 KiB) Viewed 30562 times
IMG_9530.JPG
The hole tapped and a screw test-fit.
IMG_9530.JPG (38.56 KiB) Viewed 30562 times
IMG_9532.JPG
Both holes tapped
IMG_9532.JPG (47.86 KiB) Viewed 30562 times
IMG_9533.JPG
Track-end mount installed
IMG_9533.JPG (58.09 KiB) Viewed 30562 times
IMG_9534.JPG
Holes drilled/tapped into the back of the column to accept the other track end. I did these freehand.
IMG_9534.JPG (63.24 KiB) Viewed 30562 times
IMG_9535.JPG
Track installed, sorry for the dark/blurry shot
IMG_9535.JPG (60.12 KiB) Viewed 30562 times

*******************************************************


IMG_9850.JPG
I used sheets of aluminum foil to shim the base into alignemt with the table. Shimming was a challange - loosen all four column bolts, insert a piece of foil, torque the bolts down and tram the head in again. It took a lot of trial and error, but I was able to align the table to .002" from one side to the other. I may spend more time on this in future, chasing perfection. Thanks to the solid-column, the table has maintained that accuracy for weeks, despite all the work I've done to the mill since then.
IMG_9850.JPG (56.7 KiB) Viewed 30562 times
IMG_9848.JPG
I also had to shim the bearing block on the Z axis, to get the motor mount on the other side of the ballscrew to line up cleanly with its bolt-hole.
IMG_9848.JPG (72.15 KiB) Viewed 30562 times
IMG_9841.JPG
I waited to drill the column for the Z-axis motor mount until I had the rest of the conversion sorted out. The center screw (left side in the picture) uses the existing hole intended for the column stop, but I have to drill/tap the other holes. I originally planned to strip the column off and drill them on the HF mill, but I really needed to have the Z axis assembled to accurately locate the holes. I used a drill, freehand, so the holes were not as precise as I would have liked, but they will do.
IMG_9841.JPG (74.25 KiB) Viewed 30562 times
IMG_9844.JPG
Tapping the first hole for the Z-axis motor mount. You can see how the second hole laps over the edge of the dovetail, which causes problems later.
IMG_9844.JPG (85.01 KiB) Viewed 30562 times
IMG_9845.JPG
Drilling the second hole for the Z-axis motor mount.
IMG_9845.JPG (108.97 KiB) Viewed 30562 times
IMG_9847.JPG
Here is the problem. The left side hole has a nice margin to the end of the dovetail, but the right side hole blew out the side of the dovetail. As a result, I couldn't get the tap to bite, it kept walking itself out of the hole. I drilled out the top of the hole, and then the tap worked. I could have just drilled straight holes secured the bolts with nuts inside of the column, but I tapped them for the best possible grip - then I put the supplied nuts on the ends of the screws.

This is exactly where the holes need to be to line up with the existing hole, and minimize binding on the Z-axis. I think the misalignment is a characteristic of these "snowflake mills" - no two the same.
IMG_9847.JPG (90.57 KiB) Viewed 30562 times
===================

I had to tune the motor accurately. The motors are 1.8 degrees/step, aka 200 steps per revolution. The Ballscrew is .200 pitch, which means it needs to turn (5) times to travel an inch. The Gecko g540 does 10 microsteps per step.

200 steps x 5 revolutions x 10 microsteps = 10,000 steps per inch. This is handy, because each microstep represents .0001" . The drives seem pretty happy jogging at 100 inches per second.

I used the dial-indicator mount to verify I had the correct steps per inch, as well as repeatability - that the machine keeps returning to the same point every time. Good thing, because I discovered I had not properly stored the SPI for the Z/Y axes - they were still set to the 8000 SPI I used on my first try.
IMG_9671.JPG
I bent some .25" steel rod and threaded the ends so I could mount a dial indicator to them. The L-shaped piece is a sweep bar used to tram the table, and the the slightly bent piece is used to square the vice
IMG_9671.JPG (55.81 KiB) Viewed 30562 times

Verifying that the mill is square. There is a scratch in the replaceble jaw plate, you can see the dial skip -.001" as it goes over it. Note that otherwise, the needle is solid on "0"


After programming in backlash, verifying that the Z axis returns to the same spot every single time.


I was driving the mill with the keyboard, I don't have home/limit switches installed, so I was worried about crashing the mill and didn't take it to the ends of each axis - so this does not demonstrate the full range of travel. With the table square, movement accurate and repeatable, I was ready to start milling. I used the keyboard (arrow keys, and pageup/down for z axis) to move parts under the spindle. I couldn't get the keys working on my early experiments, because I had been using g0 commands to move the table around, and g0 mode locks out the keyboard control. I learned to use the TAB key to bring up the jog dial, and used it to slow the motors to the correct feed rate and give smoother cuts. That is the wrong way to set feed rate, but at this point I was more focused on making chips than advancing through the MACH3 learning curve.

User avatar
oatnet
100 MW
100 MW
Posts: 2929
Joined: Apr 26, 2007 3:03 pm
Location: SoCal, USA

Re: Converting an SX2 Mini Mill to CNC - build thread

Post by oatnet » Mar 01, 2013 2:07 pm

IMG_0006.JPG
So the first project on the mill was not an eBike part. While building on the mill, the mount for my work lamp broke, causing me frequent angst. So I put a piece of 6061-T6 in the vice, drilled a hole, cut it off, and made a replacement.
IMG_0006.JPG (54.97 KiB) Viewed 30561 times
IMG_0009.JPG
The metal rectangle replaces the plastic piece that snapped.
IMG_0009.JPG (45.17 KiB) Viewed 30561 times
IMG_0011.JPG
Next I started on the brake adaptor. I started by cutting out a chunk of 6061-T6 aluminum, and milling the 4 sides square.

The downside to Boxxer World Cup forks is that they don't place their brake caliper holder in an industry standard location, so I have to buy adapters to mount brakes. Adaptors are readily available for standard brakes, like the AVid Elixer 8" brake I installed on my Norco Aline.

When I discovered the A-Line needed more brake, I immediately thought of the 4-pot 9" gatorbrakes I bought from Methods a few years back, which were made for the 2010 zero eMoto. Nobody ever used the 9" on a Boxxer, so if I wanted to use that brake, I needed to make an adapter.
IMG_0011.JPG (60.76 KiB) Viewed 30561 times
IMG_0012.JPG
The 6061 coupon, milled square and in the vise.
IMG_0012.JPG (54.77 KiB) Viewed 30561 times
IMG_0020.JPG
Holes located, drilled, tapped. The threaded holes are 51mm apart, as are the unthreaded holes.
IMG_0020.JPG (54.81 KiB) Viewed 30561 times
IMG_0021.JPG
I have the factory adaptor bolted to my adaptor-in-progress. I made the plastic template above it (months ago) by bolting it to the forks, clamping the caliper onto the disk, and cutting holes in the plastic where the caliper lined up. I then was able to bolt the caliper to the plastic and verified everything was aligned correctly. It was hard to judge depth, but it seemed like the part needed to be about .25" thick and I figured I could mill/shim it from there.

I drilled the threaded holes first. I drilled the first hole in a rough location, then moved the x axis 51mm for the second hole.

With those holes in place, I bolted my plastic template to the part, and used a center-finder on it to locate where I needed to put the other two holes. The holes in the template are definitely not round, so this only gave me a rough idea of where they are located. I located and drilled the first hole as accurately as I could, but it was essential that the second hole be exactly 51mm away from it. I bolted the factory adaptor to the first hole, moved the spindle to where the 2nd hole _should_ be, rotated the adaptor so its second hole was roughly under the spindle, and used a centerfinder on that second hole to precisely locate the spindle. This wasn't really CNC milling, but the part bolted up perfectly as a result.
IMG_0021.JPG (67.23 KiB) Viewed 30561 times
IMG_0043.JPG
I needed to mill off the excess material, so I drew a rough outline of the final adaptor so I wouldn't take too much off.
IMG_0043.JPG (60.16 KiB) Viewed 30561 times
IMG_0048.JPG
At a proper feed rate, the mill produced a mirror finish on the edges. The top edge shown here was spectacular.
IMG_0048.JPG (32.6 KiB) Viewed 30561 times
IMG_0053.JPG
Test fit of the caliper on a boxxer fork so I can see how much material I need to mill out to clear the caliper.
IMG_0053.JPG (60.83 KiB) Viewed 30561 times
IMG_0056.JPG
Moved over to the vice. My machine tools instructor would have a fit if he saw me mount a part like this, but it was just a hair too long to fit into the vice flat.
IMG_0056.JPG (66.88 KiB) Viewed 30561 times

IMG_0058.JPG
First rough cut
IMG_0058.JPG (67.43 KiB) Viewed 30561 times
IMG_0059.JPG
All rough cuts completed.
IMG_0059.JPG (52.26 KiB) Viewed 30561 times
IMG_0062.JPG
First pass at filing onto shape
IMG_0062.JPG (42.99 KiB) Viewed 30561 times
IMG_0063.JPG
On the first test-mount, I was stunned to discover everything lined up perfectly, with a whisper of disk scrape. I only needed 2 of the super-thin Gatorbrake shims to align the brake into silence.
IMG_0063.JPG (73.54 KiB) Viewed 30561 times
IMG_0065.JPG
Notice that the edges of the pads line up perfectly with the braking surface of the disks.
IMG_0065.JPG (74.6 KiB) Viewed 30561 times
I have a few pics of the part mounted, but they seem more appropriate for myNorco A-line build thread (CLICK HERE)

-JD
Member 117 of 26,000+

153v DUNE BUGGY © 164v Vectrix E-Moto © 72v Norco A-line/x5403 © 60v Specialized Enduro Comp/BMC © 72v x5tracycle © 48v TF IO/BMC © 36v Kepler FrictionDrive © Spot-Welding a123 © Pelican Front Packs © Vectrix Battery Replacement © MORE !!


Image大鸡巴 Motors
Knowledge is acquired through experience, everything else is just information - Albert Einstein
"Cheater?" I'm not competing, I'm Commuting!

scriewy
100 W
100 W
Posts: 280
Joined: Dec 02, 2009 12:37 am
Location: tel aviv, israel
Contact:

Re: Converting an SX2 Mini Mill to CNC - build thread

Post by scriewy » Jul 12, 2013 6:26 am

oatnet i didn't understand how the stepper motor can do 2000 microsteps per rev if the motor itself is limited to 200 steps per rev ?
i get the board is capable of sending microsteps, but i thought the steps a stepper is designed for that's what it can do, unless there are some gears on output of the stepper. (i aint expert on steppers, saw few vids of how they click every step).

next in relation to non hardened steel, as SAE 4140/4340
what company/type/coat mill bits you use ?

what's your feed rate with a 5mm end mill bit or 7mm at a depth of ?

did you find optimal rpm before too much vibration/chatter if at all ?

what max depth you feel comfortable milling befor you feel any flex if at all ?

if you havent tried steel, then regarding aluminum.
it's a pitty i wont stay to see the world crumble, u jedi Scum

clyte 405,
went through battle, bloody resurected Dewalts 4x28v 2s2p usable 2.7ah, 52v drops to 44v at 54kmh
78v 2ah Vdrop to 58v at max 66kmh.
72kg rider + 23kg bike.
keywin 20a 30-70v mod to 6 4110 as methods advised for noobs, shunted 38a
clyte 24-72v 40a mod 4310 to 4110, shunted to 77a
17.5.11 - 5000km
at 78v 4ah 72.7kmh Vdrop to 68v
10.11 - 82.1kmh
30.8.12-17000km

a place for lasers as ES is for e-vehicles - laserpointerforums.com

kfong
10 MW
10 MW
Posts: 2111
Joined: Sep 02, 2008 11:34 am
Location: SE Michigan, USA
Contact:

Re: Converting an SX2 Mini Mill to CNC - build thread

Post by kfong » Jul 13, 2013 8:07 am

Nice word JD, you will find the CNC to be a whole new hobby by itself. we have been busy too, just finish this. Waiting on a 2hp water cooled spindle. It should be able to hog out aluminum at a good rate. Working also to get the X3 online, x an y axis are done.

http://m.youtube.com/watch?v=q9FHNgrY64U

User avatar
oatnet
100 MW
100 MW
Posts: 2929
Joined: Apr 26, 2007 3:03 pm
Location: SoCal, USA

Re: Converting an SX2 Mini Mill to CNC - build thread

Post by oatnet » Jul 13, 2013 8:42 am

kfong wrote:Nice word JD, you will find the CNC to be a whole new hobby by itself. we have been busy too, just finish this. Waiting on a 2hp water cooled spindle. It should be able to hog out aluminum at a good rate. Working also to get the X3 online, x an y axis are done.

http://m.youtube.com/watch?v=q9FHNgrY64U
Nice DIY build kfong, and you also went for the x3, I'd love to see a build thread on that. I thought long and hard about the x3, I wish my build had its work envelope, but the X2 does fit the space I had better.

I still need to install home switches, and I got hung up struggling with Autodesk Inventor, but I recently used Sketchup to design the battery box for my VW dune buggy conversion (putting the 9" motor that was in the 3500lb bus into a 1,000lb street-legal dune buggy, CLICK HERE), I liked it so I think that will clear my CAD obstacle.



Sorry screiwy, I don't have the practical experience with this machine yet to answer your question. This project enabled me to build the brake adaptor to complete the Norco Aline 5403 build, and then after that I got sucked into the mid-drive "Ktrak" snow/sand bike, and finally the Dune Buggy conversion I mentioned above. I will likely cycle back to the mill next, now that I have CAD sorted.

-JD
Member 117 of 26,000+

153v DUNE BUGGY © 164v Vectrix E-Moto © 72v Norco A-line/x5403 © 60v Specialized Enduro Comp/BMC © 72v x5tracycle © 48v TF IO/BMC © 36v Kepler FrictionDrive © Spot-Welding a123 © Pelican Front Packs © Vectrix Battery Replacement © MORE !!


Image大鸡巴 Motors
Knowledge is acquired through experience, everything else is just information - Albert Einstein
"Cheater?" I'm not competing, I'm Commuting!

kfong
10 MW
10 MW
Posts: 2111
Joined: Sep 02, 2008 11:34 am
Location: SE Michigan, USA
Contact:

Re: Converting an SX2 Mini Mill to CNC - build thread

Post by kfong » Jul 28, 2013 11:05 am

We do keep a running build on our website http://www.embeddedtronics.com/x3.html, our gantry project even made it to hack a day http://hackaday.com/2013/07/26/cnc-rout ... ore-100262, and the gantry on our site is
http://www.embeddedtronics.com/cnc_gantry.html. 3D printers will be our next are of interest, part of the reason for the gantry build. I need (4) 12 inch custom aluminum disks for the structure.

We started on the X1 and wished we went with the X2 at the time due to the column flex. Still it has earned it's keep over the years, it even milled all the aluminum parts and the stainless steel torque arms for my Motobecane build. I think you will be quite happy with the X2. It has enough iron and mass to mill out steel with reasonable cuts. We are hoping the current gantry mill will be stiff enough for steel, won't really know for sure until we do actual cuts. Very confident it will cut aluminum like butter using the 2.5hp VFD drive. The work area is 2'x2.3', quite a bit more than the X3. The problem you encounter with going larger is the flexing, aluminum isn't the best choice for a stiff mill of this size, but it's easy to work with. We have backup plans to stiffen it with plate steel if needed.

Cad took much longer to accomplish. I had to teach myself solidworks, mastercam and mach3. Learning 3D, took the longest to do without a class. YouTube is your best resource in learning this, but the payoff is worth it.

oatnet wrote:
kfong wrote:Nice word JD, you will find the CNC to be a whole new hobby by itself. we have been busy too, just finish this. Waiting on a 2hp water cooled spindle. It should be able to hog out aluminum at a good rate. Working also to get the X3 online, x an y axis are done.

http://m.youtube.com/watch?v=q9FHNgrY64U
Nice DIY build kfong, and you also went for the x3, I'd love to see a build thread on that. I thought long and hard about the x3, I wish my build had its work envelope, but the X2 does fit the space I had better.

I still need to install home switches, and I got hung up struggling with Autodesk Inventor, but I recently used Sketchup to design the battery box for my VW dune buggy conversion (putting the 9" motor that was in the 3500lb bus into a 1,000lb street-legal dune buggy, CLICK HERE), I liked it so I think that will clear my CAD obstacle.



Sorry screiwy, I don't have the practical experience with this machine yet to answer your question. This project enabled me to build the brake adaptor to complete the Norco Aline 5403 build, and then after that I got sucked into the mid-drive "Ktrak" snow/sand bike, and finally the Dune Buggy conversion I mentioned above. I will likely cycle back to the mill next, now that I have CAD sorted.

-JD

User avatar
oatnet
100 MW
100 MW
Posts: 2929
Joined: Apr 26, 2007 3:03 pm
Location: SoCal, USA

Re: Converting an SX2 Mini Mill to CNC - build thread

Post by oatnet » Dec 19, 2013 1:24 pm

I haven't had time to do any CNC, but I have used it for several operations on my Dune Buggy conversion (link in signature). That said, I wanted to issue Customer Support Kudos to LittleMachineShop.com.

The spindle on my mill suddenly stopped working, it looked like a bad phase, spinning very slowly, pausing, and spinning slowly again.   My mill is out of the warrantee period, but only has an hour or so of use, so Chris of LMS generously offered to replace the failed part.

He sent me a link to the controller test regimen, which showed voltage on all three phases, so we then assumed it was the motor.  I was surprised, expecting a blown FET, but I could see how maybe I lost a hall sensor or a winding failed...

I got the replacement motor fast, but when I swapped it in, same problem.  Chris immediately sent me a new controller, and told me to hold onto the motor until the problem was fixed.  I got the replacement controller fast, but when I swapped it in, same problem.   If it wasn't the motor or controller, that left only the throttle...  I popped the throttle box, put the DMM on the pot, but didn't see resistance to the wiper change as I rotated it up and down...  So I powered it up, jumpered the wiper as shown in the picture below, and the spindle immediately spun up to speed.  Chris sent me a new potentiometer assembly, and that resolved the problem.  Anyhow, the support and excellent customer service made me really glad I bought from LMS, they have earned a loyal customer.

Thanks LittleMachineShop! :D

-JD
Member 117 of 26,000+

153v DUNE BUGGY © 164v Vectrix E-Moto © 72v Norco A-line/x5403 © 60v Specialized Enduro Comp/BMC © 72v x5tracycle © 48v TF IO/BMC © 36v Kepler FrictionDrive © Spot-Welding a123 © Pelican Front Packs © Vectrix Battery Replacement © MORE !!


Image大鸡巴 Motors
Knowledge is acquired through experience, everything else is just information - Albert Einstein
"Cheater?" I'm not competing, I'm Commuting!

Post Reply