[djbird3]

For all those interested in the LMS HiTorque Mini-Mill I've decided to start this thread to discuss this particular mill as well as the CNC conversion I am planning for it. One of the reasons I'm doing this is that there are no other threads that discuss this particular X2. Another reason is this is my first Mill CNC conversion so it will be a learning experience for me and a oppurtunity to get feedback from others. My intention is to use the experience gained converting this small mill for a larger X3 class mill I want to do eventually. In addition to the CNC conversion I intend to also design a power drawbar attachment as well as a ATC unit so that this mill will be fully automated or allmost fully. Anyway, here is how I intend to proceed:

1) Disassemble machine and evaluate.
2) Start deciding what hardware to use.
3) Make CAD drawings and 3D models of all pertinent machine(the mill itself) parts.
4) Workout design for necessary mounts, brackets, etc.
5) Lap dovetails on all axises to eliminate any tight spots.
6) Install hardware starting with the table then eventually the Z axis.
7) Figure out what motor driver combination meets my needs.
8) Install limit switches, opto-couplers, etc. and debug elctronics.
9) Design eclosure for electronics and assemble everything.
10)Use resulting machine to make parts for Power Drawbar and ATC:)

I've already got a jump on the first three tasks and will discuss what's going on so far, but for now here are some pictures of the machine. All questions and comments are welcome so feel free to do so.

~Don~

[djbird3]

OK, to start with my reasons for selecting this particular mill for CNC conversion are:
1) larger table with 30% more travel
2) belt driven spindle
3) brushless motor
4) overall better quality

My initial impression of this machine when I got it was that it was much cleaner and coated with light machine oil instead of thel nasty red goop that these Chinese machines usually come covered with. Another thing was that the paint job was much nicer but of course these are only cosmetic things. The spindle runs very quietly with only a little belt noise but there seems to be a problem with the motor controller because the speed is unsteady. Another problem is that this controller has no reverse which I don't like, but for now I will leave the motor controller issues for another post and deal only with the mechanical aspects. I've already disassembled the machine so that I could inspect it as well as make measurements for my CAD drawings and found everything to be fine with only a couple problems. First they didn't bother to remove the paint from the dovetail of the spindle when they painted it and secondly the pulley on the motor was in excess of .010" off center. Apparently the Chinese are still having some problems drilling holes in the center of things. Anyway, it only took about an hour to sleeve and rebore pulley and a minute to wirebrush the dovetail. Here are some pictures of the disassembled machine so everyone can see what I'm talking about. I've just about finished the CAD drawings and have already ordered the ballscrews and bearings I want to use (should be getting them this week). I'll be discussing in my next post the what, how's and why's of the hardware I intend on using for this conversion as well as the design parameters, so stay tuned, much more to come!

~Don~

[Fastest1]

Great pics and descriptions. I am sure it will help quite a few people.

[djbird3]

OK, today my ballscrews arrived and they look pretty good. I spent about 2 weeks researching ballscrews because I've never dealt with them before and I finally decided on REX Industries. I chose the REX screws and nuts because they were the best bang for the buck. There are only 4 primary brands to choose from that I'm aware of excluding the chinese stuff, which are; REX, Nook, Thomson and Roton. I checked prices and specs on all of these and REX was by far the best choice. The REX screw is class C7 guaranteed to not exceed more than .0018" error per foot, Nook XPR series screw is rated at no more than .001" error per foot but cost about $1400 for 4 feet of screw and 3 preloaded nuts as apposed to $400 for the equivalent in REX hardware. Roton has fairly descent product for the buck but they don't offer a preloaded nut and there ballscrews are only rated at .003 to .006 error per inch which isn't very good but you can get 3 nuts and 4 feet of screw for around $150. You can also get the equivalent in Thomson with their screw accuracy rated at no more than .003" error per inch for around $250. BTW these screws I'm comparing here are all 5/8" - 5 rolled screws except for the REX screw which is metric.

For the sake of those who aren't familiar with ballscrews and ball nuts let me explain a few things I've learned about them in the forums. First, the accuracy of a screw is not as big of a deal as you may think, screw inaccuracy can be compensated for by using screw mapping in Mach 3. A better quality screw of course should be prefered because they run smoother and are less prone to binding when using a preloaded nut but the bottom line is you don't need top of the line to get decent accuracy. Secondly, when it comes to ball nuts there are 3 basic ways of eliminating backlash which are:

1) using a double nut
2) using a preloaded nut
3) using oversized balls in a standard nut

The first approach is to use 2 nuts back to back adjusted to take up the backlash. This is a good approach, especially if your using Roton nuts @ $25 a piece, but not an option for a small machine like the X2. The second way is to use a preloaded nut which uses offset tracks to take up the slack. The pictures below are of the ball screw and nuts the I just got. These nuts have 4 ball circuits in each of them. The small red plugs are what returns the cirulating balls from the end of their respective circuits back to their beginings. These nuts turn very smoothly and have no backlash and are the best solution for a small machine like the X2. The last approach is what I call "The Cheap and Dirty" approach. This technique utilizes oversized balls to remove the excess play but the problem with this approach is that the undersized balls are used for a reason and that reason is to prevent binding. When you stuff bigger balls into the nut there is less wiggle-room for the balls to circulate so they tend to jam up under load. This isn't usually a big problem under light or medium loads but heavy loads can cause gualling of the screw. Regardless of the situation the ballscrew will wear out much quicker than if you used a preloaded or double nut. Anyway, if anyone has anything else helpful to add to this please do. In my next post I'll be dicussing design specs and such for this conversion, so untill next time...

~Don~

[hoss2006]

The lead error for Roton and Thomson should be per foot, not per inch as you've stated.
.004/foot is a typical rolled ballscrew lead error spec.
precision rolled come in under .002/foot.
Preloading a nut with oversized balls is an affordable solution to getting reduced backlash, not zero backlash.
Proper application typically results in .001 or less without putting undue strain on the ballscrew.
Lots of CNCfusion preloaded nuts doing the job out there.
I would like to see Roton come out with a nut similar to the offset track type but with the same threaded end.
Hoss

[mcphill]

I bought one of these with the intent to do the exact thing you are doing, but life got in the way... I was planning to work with CNC Fusion to put a kit together... There are only a few part that need to be change from their existing kits to support this model.

[djbird3]

The lead error for Roton and Thomson should be per foot, not per inch as you've stated.
.004/foot is a typical rolled ballscrew lead error spec.
precision rolled come in under .002/foot.
Preloading a nut with oversized balls is an affordable solution to getting reduced backlash, not zero backlash.
Proper application typically results in .001 or less without putting undue strain on the ballscrew.
Lots of CNCfusion preloaded nuts doing the job out there.
I would like to see Roton come out with a nut similar to the offset track type but with the same threaded end.
Hoss

Hey Hoss,
Yeah, it was late last night when I typed this and apparently I SCREWED-UP on the error per foot in regard to Thomson and Roton. BTW I wasn't suggesting that it's wrong or anything to use oversized balls but that the other methods are better as far as durability goes. Anyway, I appreciate your correction as well as your contributions in these forums that I've learned a good bit from over the years. Keep up the good work.
~Don~

[hoss2006]

Not to nitpick but a double nut system is the worst as far as durability goes.
It relies on constant spring pressure between the nuts, 100-200 lbs worth typically,
to counteract backlash. That puts the balls and screw under a lot of strain 24/7.
The springs will compensate for the accelerated wear but only up to a point.
Have fun with your build, should be a nice one.
Hoss

[djbird3]

Not to nitpick but a double nut system is the worst as far as durability goes.
It relies on constant spring pressure between the nuts, 100-200 lbs worth typically,
to counteract backlash. That puts the balls and screw under a lot of strain 24/7.
The springs will compensate for the accelerated wear but only up to a point.
Have fun with your build, should be a nice one.
Hoss

Once again I guess I should have been more clear about what I was talking about in regard to double nut arrangements. My idea is to use 2 standard nuts back to back adjusted apart from one another just enough to take up the slack, no need for houndreds of pounds of preload just a few pounds at the most. Essentially you would be accomplishing the same thing as a preloaded nut but by offsetting the nuts instead of offsetting the tracks in the nut. Another way to look at it is that one nut takes the load in one direction while the other in the opposite. But, as I stated before this is much more bulky arrangement then a single preloaded nut. You might want to try this on your G0704, there should be enough room for you to fit two of those Roton nuts under the table. You'll just have to workout a design to mount the nuts in an arrangement that will allow you to adjust the preload between them and lock them in position, if you follow what I'm saying. Anyway, this is something worth expeimenting with and definetly a cheaper alternative to preloaded nuts.
~Don~

[hoss2006]

We're on the same page then, I discussed this same idea a couple months ago on my thread.
http://www.cnczone.com/forums/836074-post276.html
Hoss

[LatheMaster]

Hey all , I'm faily new here but have been doing a lot of reading of threads . I am getting the LMS sx2 this month and the Fusion conversion kit . I just want to mention on the fusion kit , if you click on sx2 , there is a spot in the lower right corner of the screen to add modifications to the X2 kit for the extra travel of the LMS SX2 . Its $45 for modified lead screws . I plan on getting that kit with preloaded balls on every axis .

My question is I have been looking at boards and drivers and not sure what to get . I had one guy tell me he has had overheating problems with his Gecko g540 and he had to put cooling fins on it . Said he has had problems since day one with it . I would rather have separate components . I want something that gives options for external switches and buttons . I have been looking at the new digital drives by Keling which will be out in January .

For a breakout board I have looked at several that have all the bells and whistles . Most of the top end ones require you get a board called smooth stepper . What is this and why do I need it ? There is a breakout board that is only $28 too . Should I get a better board right away ? I am thinking the $28 board is one to just get you going and any add ons for other stuff would be additional boards that would have to be purchased .

For steppers what do I need ? I want 100ipm rapids and would like a feed rate at least half of that on all axis . I dont want a counterweight on the Z either . I see LMS has gas struts you can get . Not sure how well they work . I dont know all the formulas for figuring this stuff out . I do plan on adding a 4th axis and maybe a 5th in the distant future so I need to keep power supply needs in mind too . Thanks for any help you can offer me .

Keep in mind a lot of the educational stuff I will be doing will be on foam and machinable wax so I want fast feeds . I was also wondering why the new SX2 has only a 2500rpm speed ? It has a brushless motor and a belt conversion from the get go . Can different pulleys be added to increase this speed to 4k or more ?

Hoss2006 , just wanted to mention I was doing reading on preloading bearings and double nutting balls . What a debate . I pretty much left that forum mindboggled .(chair)

[rewster]

I finished my x2 conversion a few weeks ago. CNCfusion kit #4 w/ preloaded ballnuts, gecko g540 and 381 motors from keling. I have zero complains so far. Backlash is extremely minimal, less than expected. The 381 oz-in nema 23's were an obvious choice because of their compatibility with the g540. They have plenty of power, no counterweight/spring needed on the Z. I've been very pleased with the performance of the g540. Easy to set up and runs beautifully, plus has all of the bells and whistles built in. From my understanding, it is possible to replace single drives if something does fry, rather than replacing the entire unit. I have two 12v fans blowing on the back of it, and have not experienced a single symptom of the problems you speak of with this setup.

I highly recommend this setup. I'm glad I landed on this decision after trolling for months while I anxiously gathered build funds. Keling offers a g540 kit which is pretty affordable. I did a lot of wax cutting while getting the software tweaked and got great results with 30ipm feeds. I've been running 6061 at 15ipm, .075, but it could probably handle much more.

[cornbinder23]

FWIW,
The G540 does have separate drives, it is 4 G251's i believe
and can be serviced to replace one if need be, I have about 9 mo use from my G540 with the 381's and a 906oz on an SX3 with NO issues whatsoever so far, I do have a small fan blowing across the back of the case and it never gets hot at all, even after running 4hr plus programs :cheers:

JTCUSTOMS

[djbird3]

Hey all , I'm faily new here but have been doing a lot of reading of threads . I am getting the LMS sx2 this month and the Fusion conversion kit . I just want to mention on the fusion kit , if you click on sx2 , there is a spot in the lower right corner of the screen to add modifications to the X2 kit for the extra travel of the LMS SX2 . Its $45 for modified lead screws . I plan on getting that kit with preloaded balls on every axis .

My question is I have been looking at boards and drivers and not sure what to get . I had one guy tell me he has had overheating problems with his Gecko g540 and he had to put cooling fins on it . Said he has had problems since day one with it . I would rather have separate components . I want something that gives options for external switches and buttons . I have been looking at the new digital drives by Keling which will be out in January .

For a breakout board I have looked at several that have all the bells and whistles . Most of the top end ones require you get a board called smooth stepper . What is this and why do I need it ? There is a breakout board that is only $28 too . Should I get a better board right away ? I am thinking the $28 board is one to just get you going and any add ons for other stuff would be additional boards that would have to be purchased .

For steppers what do I need ? I want 100ipm rapids and would like a feed rate at least half of that on all axis . I dont want a counterweight on the Z either . I see LMS has gas struts you can get . Not sure how well they work . I dont know all the formulas for figuring this stuff out . I do plan on adding a 4th axis and maybe a 5th in the distant future so I need to keep power supply needs in mind too . Thanks for any help you can offer me .

Keep in mind a lot of the educational stuff I will be doing will be on foam and machinable wax so I want fast feeds . I was also wondering why the new SX2 has only a 2500rpm speed ? It has a brushless motor and a belt conversion from the get go . Can different pulleys be added to increase this speed to 4k or more ?

Hoss2006 , just wanted to mention I was doing reading on preloading bearings and double nutting balls . What a debate . I pretty much left that forum mindboggled .(chair)

Hi,

I'm in the same boat as you at the moment in regard to motor and driver combinations and primarily focusing on the mechanical aspects of my conversion but I can help you with some of your questions. To begin with the "Smooth Stepper" is a usb to dual parallel port adapter meant primarily for those who are using laptops that don't have a parallel port, which is the case for all new laptops. You don't need it unless you are using a laptop. If your PC lacks a parallel port you can purchase a expansion card for your PC that will solve that problem for about $20.

Next, the "breakout board" is the means by which you interconnect your drivers and other devices to their respective pins on the parallel port. Most use opto couplers to electrically isolate the port from the connections directly to your devices. The reason for this is to protect your parallel port and PC in the event of a failure of one your drivers or other devices. This is a common practice used to isolate equipment running from different power sources. The board I like to use is made by Probotics, it features RF isolaters rather than the the older optical type. The advantage of the RF(Radio Frequency) type is higher signal bandwith and better drive level, but this in most cases is not an issue. This board also features screw terminals as well as pin headers for connecting your hardware. I'll put some links at the end of this post so you can checkout these things for yourself.

As for motors and drivers, it's a matter of what your trying to achieve. In my humble opinion 100 IPM is modest, 200-250 IPM is what I'm shooting for. I intend on trying some 280 In/Oz steppers first because I already have them. As for drivers I'll first try my Probotics unipolar drivers that I use to operate my CNC router just as a test, but I will probably end up getting one of Kelings bipolar drivers. When it comes to advice regarding all things X2 I would checkout Hoss's website and his threads, no one has bolted more stuff on the X2 then him and generally I trust his opinions about hardware. Hoss seems to like Keling so I'll look into that first but at this point I'm not sure, I'll cross that bridge when I get to it.

The spindle on the SX2 is driven by a 500W brushless motor as you allready know and the controller is a digital 3 phase driver. I'm presently not happy with mine because it has a glitch that causes it run unsteadily, an issue I'm currently working on. Also there is no reverse, allthough it is possible to swap some wires around to get it to run backwards it doesn't run as well backwards because of the phasing of the feedback sensors in the motor. Worst come to worst I'll just get a different driver if I can't get the results I want. For you as far as spindle RPMs go, a simple solution is to do what a lot of other guys do and that is to mount a small trim router via a custom made mount to the head of your machine. Trim routers typically turn-up around 10000-15000 RPMs and have good bearings on there spindles that can handle the speed. If your looking to push a small endmill at high feeds its the simplest, least expensive way to go.

Regarding the Z axis I'm getting rid of the silly spring arm that counteracts the head weight. It exerts nonlinear pressure on the head which isn't what you want. If I can get the IPM that I want without using an airspring I will because it will be one less thing to wear out and get in the way. I'm figuring with a direct drive setup that I want to use on my Z axis it will probably take around 400-500 In/Oz of torque to get the speeds I want but I really don't know right now. At the moment I'm focusing on working out the mount designs and stuff but when I get that out the way in a week or two I'll concentrate on the motor/driver questions.

When I'm finished with this project I am going to release all of my mechanical drawings, electronic mods and specs to the public so that all interested can build their own if they wish to do so. So, anyone who is not in hurry, sit tight and probably within a couple of months at the most I should be done. I'll be posting in this thread on a regular basis about my progress and what I'm doing. Hopefully we'll all learn something usefull in the process.
~Don~

PC Parts Collection (parallel expansion card) MOSCHIP MCS9815CV Dual Parallel PCI Card - PcPartsCollection.com

Probotics (breakout board) PBX-RF RF Isolated CNC Breakout Board -

HossMachine (Hoss's Website) Hossmachine Homepage

[rwestbr]

" If your PC lacks a parallel port you can purchase a expansion card for your PC that will solve that problem for about $20."

Only you will not be able to get a accurate spindle RPM reading with MACH3 because they are not true parallel ports (must be on motherboard), Trust me I know from experience...

ROB

[djbird3]

" If your PC lacks a parallel port you can purchase a expansion card for your PC that will solve that problem for about $20."

Only you will not be able to get a accurate spindle RPM reading with MACH3 because they are not true parallel ports (must be on motherboard), Trust me I know from experience...

ROB

Rob,

I'm using the low profile version of this card in my HP Slimline and have never had an issue with it. I've been using spindle indexing under mach3 to do threading and it works perfectly. And as far as the parallel port having to be on the mother board is concerned it's just not true, there ain't any difference between it being soldered to the MB or it being on a expansion card, they both are accessed through the same bus. The only difference that could possibly cause a problem is which chip they choose to use and how they implement it, which in this case happens to be a MOSPEC 9815 chip which is a true parallel port take it from someone who has over 20 years of experience in the electronics field both repairing and designing. Remember it's not a perfect world sometimes things just don't work right for a lot of different reasons.
~Don~

[rwestbr]

Hey Don,
It was my experience that the fact that my motherboard had no BIOS setting for switching on EPP printer mode and the way windows reads this information for hardware,and this is what I was told from a parallel card manufacturer to be causing my problem. I never was able to resolve it until I changed to a motherboard with built in parallel port. Just my experience, I am glad your setup came out good, but to be honest mine was HELL!!!! FYI: I had a AMD X2 64 socket 939 compaq motherboard and tried no less than 4 different cards,changed every setting in MACH to no avail. The RPM's were not even close on my mill or on my lathe (CNC4PC C3 cards). Maybe my hardware or driver or maybe me? At any rate I hope this helps someone with this type of problem.

PS , I respect your incite and experience , wish I had it when I was in that mess.

Respectfully,
ROB

[luv2ride]

What are all of you guys doing with your factory lead screws? I am looking to extend my y-axis and am in need of a longer screw, 2-4 inches longer than the factory x2. If someone wants to get rid of one, please let me know. Later on in life I will refer back to these threads and install CNC.

[LatheMaster]

I finished my x2 conversion a few weeks ago. CNCfusion kit #4 w/ preloaded ballnuts, gecko g540 and 381 motors from keling. I have zero complains so far. Backlash is extremely minimal, less than expected. The 381 oz-in nema 23's were an obvious choice because of their compatibility with the g540. .......

I did a lot of wax cutting while getting the software tweaked and got great results with 30ipm feeds. I've been running 6061 at 15ipm, .075, but it could probably handle much more.

rewster , was your lowest feed rate on the z axis ? How much does the x y z differ from one an other in feed rates . Those were the steppers I was thinking of getting . I would almost think I might want to go bigger on the z cause it has a constant load and carries a lot of weight . I think Hoss did that .

Thanks for all the great info . Especialy djbird3 for the detailed explanation .

[rewster]

rewster , was your lowest feed rate on the z axis ? How much does the x y z differ from one an other in feed rates . Those were the steppers I was thinking of getting . I would almost think I might want to go bigger on the z cause it has a constant load and carries a lot of weight . I think Hoss did that .

Thanks for all the great info . Especialy djbird3 for the detailed explanation .

Yeah, I run a low feed on the Z. Like I said, all axes could probably handle a much higher feed rate. I've been using a couple old dull 2flute endmills, and I wasn't in a rush so I set to baby feed. I first had my rapids set to 150, but turned them down because I can't really justify needing the table to whip around like that in a pinch. The 381 has held its own on the z axis with no counterballast whatsoever....I haven't noticed any measurable amount of head drop while cutting. I first planned to convert then rig up some sort of airspring mod, but I don't think it will be needed. If you go with the g540, you're going to get the best performance with the amp rating of the 381s. Larger steppers won't necessarily produce any more power. Hope this helps.