[yantra3d]

I've been playing around with some CAD/CAM trying to get a few things designed for the mill that I can cut. I designed some small enclosures for the micro switches that I'm using for the limits (see first image). The switches are less than 3/4" in size and the mounting holes are too small to work with so I am going to try to cut these covers out of some stock Al which will also add some protection for the exposed wiring and terminals.

The other parts are for mounting the AC spindle motor to the top of the head. The idea is to have the top motor mount PL slide inside the two side brackets using an adjustment screw located at the rear for belt tension and 4 top-mounted fastening screws to lock it down.

I hooked the original spindle motor back up to the mill and wired it back together so I can cut these mounting brackets and plates needed to mount the AC motor that will be driving the spindle. I plan to use a 1.67:1 timing ratio which will give me about 5000+ RPM using a 20/12 pulley arrangement. I'll be using a 41 tooth 3/8" pitch 1/2" wide belt to drive them. I believe that 5k will be the absolute max that I would ever run this spindle (given the quality of the bearings, etc...) so that should be plenty. For anything faster than that I have another spindle solution that will take up to 1/4" bits.

I am going to sleep on this design for a couple of days I think before committing to cutting up my precious stock. It has a rather large footprint which may cause a problem later on when figuring out a powered drawbar solution. I wanted to allow enough range in the movement of the top plate so that I have some room to try different pulley ratios and belt sizes without having to scrap this design for another one should I decide against the 1.67:1 ratio at any point.

[Teyber12]

very interesting. I would love to re-mount my motor and do somethign like you plan to do (minus the servo motor). How hard do you think it would be to replace the gear thats currently attatched to the spindle?

Your work is ridiculously good mate.

cheers

[yantra3d]

Thanks Teyber, I appreciate that. Fortunately I have a lathe that I can use to bore out the pulley so it shouldn't be too difficult. The trick is finding the right pulley combo where the one attached to the spindle has a large enough hub diameter to accomodate the nearly 1" bore required without compromising your desired timing ratio. It looks like a 14 tooth pulley will be the smallest one I can get by with for the spindle. I'm about to order the aluminum pulleys so I'll post my progress here And let you know how it goes.

[MaX-MoD]

Hi there,

Just a post so I get email notification about your advancement

Keep on the good work

Max

[yantra3d]

Yes! I figured out how to control the ServoStar CD servo drive directly from the Kflop controller using step/dir signals. I found an appendix online that explains how to set the drive to function in gear mode so that it acts like a slave to a master drive through electronic gearing in the software.

Turns out that you don't even need the +/-10V analog signal at all using this mode since it is designed to accept digital signals in the form of an external encoder pulse (from master motor) through a d9 connector at up to 3 MHZ. In this mode you can also send it signals from a handwheel or in my case, a step controller pulse where the A channel connects to the step/pulse and the B channel connects to the DIR signal - changing direction depending on hi/low state. Essentially 3 wires is all it took (grnd, step, dir) from kflop BOB to a serial d9 connector on servostar.

I set up the gearing ratio in the motionlink software to account for a 1024 line input from step controller (GEARI = 1024) and the 2048 ppr +quad (GEARO = 8192) and then shortened the pulse width from the Kflop to about 60ns. Then I had to reset the velocity limit that was set by the previous owner that was limiting the RPM's to about 1900. Took me a while to figure out why it was topping out 1100 RPM's shy of the max!

Now I need to get the electronics enclosure finished up so I can get it mounted on the stand along with the mill so I can cut the mounting plates for the spindle motor to get this baby mounted. But first I need to cut and mount the limit switch mounting covers.

I have an idea for mounting the z axis limit (or trigger) using a magnetic base so that I can have the ability to adjust the lower Z limit's range in order to account for any given setup (vise, mounting hardware, etc...) that I choose to use at the time.

[yantra3d]

I finally finished up the electronics enclosure and mounted it back onto the stand along with the X3. I should be able to cut something this week. If all goes well then I'll try to get the limit switch mounting covers done and then move onto the spindle motor mount.

I created some CAD designs for the PC controller enclosure I am going to attempt to build. It's going to house a small ITX PC computer and 15" touch-screen panel and an attached keyboard and switch panel for misc buttons (PWR, E-stop, limit switch override, etc...). I will also be adding a removable hand-held jog pendant very soon.

Initially I planned on adding tons of buttons and switches for any and everything thing but after thinking it over I've decided to use as little mechanical switches as possible for this and rely on the touch-screen feature for the main control. I have a few reasons for this; for one, it's cheaper! It's also a HUGE time-saver and unless you enjoy pulling your hair out over the cobweb of wiring that's involved, it should save a headache or two. There are some mechanical switches that I would really like to have but all I really HAVE to have is an E-stop button switch and a limit switch override. I'll see how it goes after running it this way for some time and if I want to add more mechanical switches then I will.

I have a 24x24" flat plate of Al that I'll be using to cut these parts out with. The main enclosure that houses the touch-screen and pc will be made from a scrap steel plate that I will attempt to cut and bend into a simple box.

I need to start on a design for the swing-arm mount that will enable me to access it for both the lathe and mill. Not sure if I should mount it to a stand or a wall.

[Quazar]

Hey, great work on the whole conversion! Thanks for posting details of your progress.

I'm just starting an SX3 conversion, also using the CNCFusion kit. I've got the electronics sorted, but what I'm puzzling about is the placement of the limit switches. I'm using these, so I won't need to machine cases for them, but some of the placement may be awkward. I'm curious what your placement plan is - any advice would be most welcome.

My current plan is to install both + and - limit switches on all three axis (total of 6 switches). I can mount -Y flat on the machine's base just behind the saddle, and -Z can be mounted on the column. I was thinking of mounting both +X and -X on the back of the saddle itself, and bolting some stops onto the back edge of the table to engage the limit switches. I think there is room near the top of the column for a +Z limit switch, but I haven't looked into that yet. The +Y switch is the one that seem least obvious to me.

My thought was to double-stick-foam-tape all the limit switches in place before I drill/tap any holes, just to make sure everything clears and there is a clear wiring path, etc.

Regarding wiring: I'll wire each of the - limits to their own inputs, and these will double as home switches. The three + limit switches will all be wired to a single shared input. I figure that under normal operation, none of the + limits should be reached; if any + limit is reached I want all motion to stop. Does this seem like a reasonable plan?

Oh, and I took your advice regarding the connectors - I'm going with Tyco CPC connectors for everything that isn't shielded inside an enclosure.

Cheers,
- Dean

[yantra3d]

Hey thanks Dean! Congrats on your new SX3. It's always inspiring to see others' builds... can't wait to see how yours turns out.

Here are some pics of my CAD model that show how I plan to setup my limit switches for the X and Y axis. I am using a 1" thick 12x24" mic6 tooling plate that will be mounted to the top of the table as you can see in the pics below. This gives ample coverage of the X axis limit switch and triggers that I plan to mount in the front of the machine. The main reason for choosing the front of the table vs the rear is so I can take advantage of the slot on the front-side which will enable me to move the triggers around as I see fit. This should come in handy when fine-tuning their initial placement and give me freedom to change the limit position should I decide to mount a 4th axis to the tabletop at some point in the future.

For the Y axis I plan to just mount the triggers permanently to the base of the machine as I shouldn't have a need to adjust them for any reason.

I actually plan on using permanent magnets for the Z axis triggers. The reason for this is so that I can temporarily remove the gas struts should I ever need a few extra inches to the Z height at the top of the column and this would allow me to quickly move the limit to account for the change in height. (I lost a few inches of travel at the top by using the gas struts.)

Having the ability to adjust the lower Z axis trigger will definitely come in handy whenever I set up different work-holding fixtures, etc..., which will vary in height.

I am using separate switches (proximity sensors) for homing. I picked up some decent ones off ebay that were priced right. I'm not certain but I would guess that they will be much more accurate than the mechanical limit switches. I am using them on my cnc lathe (installed by manufacturer) and they work great. I wrote a routine (with help from Tom at Dynomotion) so it rapids to the sensor and then reverses travel very slowly until it backs off the switch and then it zero's out. This has a very repeatable accuracy that hasn't failed me yet. If you want I can post a link or pics to the home sensors I am using for this build. One thing I should mention is that I had to purchase a separate 24V>5V ttl conversion board (8 total inputs) to convert the home signal to digital ttl for interfacing with the kflop controller board. The sensors are rated between 6-36V however they only worked with a 5V pwr supply and not a 12, 18, or 24V supply! The conversion board has a switch for each input so that I can switch it from 24VIN to 5VIN, so that worked out okay in the end. Cost of board was about $25.

The AMP connectors are the way to go I think. I really like them. You can twist them on and off in a split-second and they are very rugged. I would stay clear of the s-video mini-din connectors, however I used some 3-conductor mini-dins for the home switches and they are holding up nicely.

I plan to build the small chip tray which will mount to the 12x24 tooling plate and have a front door that pivots downward that will cover the Y axis motor whenever I scrape chips off into a bucket. The rear side of the chip tray is open and so I will be making a 24" tall deflection shield that will mount to the back of the table to help keep chips on the table and from flying out behind the mill. I have a fogbuster mist-type coolant setup so I don't really need the full enclosure required for flood coolant.

The chip tray and shield will cost less than $50 in materials (18g Galvanized Steel) which isn't too bad. I have the 2D CAD files for the chip tray so if anyone wants to check them out or see the dimensions then I can load them up here. It's 9 inches tall and 26" deep (extends out 3" in front of tooling plate to give room for 4" vise). The width is about 29" with the sides extending out 10" beyond the tooling plate.

Hopefully I can get those limit switch mounts and triggers cut out this week. Had a lot of problems last week that I am sorting out now. Will post progress soon...

[Quazar]

This is tons of great info - particularly the drawings!

I like the idea of using the slot on the front of the table to attach the X axis trigger blocks. Though I won't need the adjustability since the switches I'm using have about 30mm of adjustability because of the arm.

I completely agree about the Y axis switches, and those will probably be the first ones I mount.

When I mounted my gas struts, I chose to loose Z travel at the low end, since I generally run with a vise anyway (or possibly a tooling plate when I switch to CNC). I like the idea of being able to change the limit on the fly like that, though.

I haven't really thought much about a chip tray. I want to be able to run metal, plastic, and MDF, so I may also have a dust collector attachment to think about...

Cheers,
- Dean

[yantra3d]

The tray is a must! I just started making my first cuts last week and only a couple of minutes into the first cut I decided to make the chip tray and shield a top priority. Chips were flying all over and landing on my keyboard, etc...

I will also need some form of way covers too. A layer of fine chips will also build up on the face of the column and the Y axis ways are totally exposed to them as well. The way covers and chip tray will be a top priority at this point. You are probably right about the dust collector if you cut MDF... that would be a good idea for sure.

I tried to keep the lower travel on the Z - at least so I could cut directly on the tooling plate with the smallest sized cutter. What size strut did you use? Were you able to keep the full height on the Z?

Also, did you go with one or two struts? The reason I ask is that I was doing some fly-cutting yesterday (trying to anyway) and noticed that there was an obvious offset whenever I removed one of the struts (so that I could temporarily mount the stock spindle motor back on the head). The one strut on the side was forcing the head to pitch over slightly and it was noticeably effecting the cut. Where did you end up mounting yours?

Have you gone over your gib strips to check for flatness? If not I would highly recommend doing so. My X axis gib strip was so warped that I had to eventually toss it. I wasted many hours trying to straighten it out with no luck. I then tried to make my own from some expensive ground stock only to destroy it worse than the original! Then I talked with Hoss and he turned me onto the replacement strips at LMS for less than $3!!! lol

[Quazar]

The tray is a must! I just started making my first cuts last week and only a couple of minutes into the first cut I decided to make the chip tray and shield a top priority. Chips were flying all over and landing on my keyboard, etc...

With my touchscreen and a dedicated control board, I plan to not keep a keyboard or mouse on the computer. I'll just plug one in for system maintenance. Of course, that doesn't mean I don't have to think about chip management...

I will also need some form of way covers too. A layer of fine chips will also build up on the face of the column and the Y axis ways are totally exposed to them as well. The way covers and chip tray will be a top priority at this point. You are probably right about the dust collector if you cut MDF... that would be a good idea for sure.

Yep - I'll spend some time thinking about this. It may also influence my switch placement and wire routing.

I tried to keep the lower travel on the Z - at least so I could cut directly on the tooling plate with the smallest sized cutter. What size strut did you use? Were you able to keep the full height on the Z?

Also, did you go with one or two struts? The reason I ask is that I was doing some fly-cutting yesterday (trying to anyway) and noticed that there was an obvious offset whenever I removed one of the struts (so that I could temporarily mount the stock spindle motor back on the head). The one strut on the side was forcing the head to pitch over slightly and it was noticeably effecting the cut. Where did you end up mounting yours?

I just installed the gas strut package from CNCFusion (here), which is a single strut mounted on the left side. It attaches to the machine base and runs vertically up to near the top of the head. The strut stud hole is drilled as close to the Z axis dovetails as I could reasonably make it, to minimize any extra torque generated by the strut. My mounting placement is very similar to the photos on this page.

Have you gone over your gib strips to check for flatness? If not I would highly recommend doing so. My X axis gib strip was so warped that I had to eventually toss it. I wasted many hours trying to straighten it out with no luck. I then tried to make my own from some expensive ground stock only to destroy it worse than the original! Then I talked with Hoss and he turned me onto the replacement strips at LMS for less than $3!!! lol

Good to know. I haven't looked at them yet (no point in tuning everything up just before I take everything apart), but I'll keep the $3 replacement in mind. Did you have to flatten the replacements a bit too, or did they arrive reasonably flat as-is?

Cheers,
- Dean

[Quazar]

I just installed the gas strut package from CNCFusion (here), which is a single strut mounted on the left side. It attaches to the machine base and runs vertically up to near the top of the head. The strut stud hole is drilled as close to the Z axis dovetails as I could reasonably make it, to minimize any extra torque generated by the strut. My mounting placement is very similar to the photos on this page.

Do folks typically remove the table locks and column lock when they convert to CNC? I can't thing of a good reason to leave them on, and if I remove the Z lock, I can fit another air strut such that it is symmetrical to the one I've already installed.

Cheers,
- Dean

[yantra3d]

Hey Dean! I can't speak for everyone else but I planned on going without the locks, however I have already used both the Y and Z locks so far!

I locked down the Y only cause my gib strip is toast and am awaiting a replacement so until then I have it locked down and am only using the X axis. (uggh)

I locked the Z because I was doing a really long slow (heavy) cut in some steel and wanted the extra rigidity of having it locked. This was fine since the Z travel wasn't necessary.

This sounds mad but you could probably make your plan work by just combining the two. You could just attach the strut to a modified bolt that could be adjusted to lock down the gib if required. One could thread a rod and then turn the end into a sphere using a lathe. May be a custom part you could post an RFQ for if you don't have a lathe, but I think it would work.

I'm almost finished fabricating my chip tray for the x3. It needs paint but it's below zero (snowing) and I have no interior space to spray it in! lol I'll post some pics of it in a day or so you can check it out...

[HackMax]

I left my locks in and tighten them just tight enough to allow the free movement of the axes. When I'm facing a piece of metal, which I always do by jogging manually, I lock any axis that I won't be using (especially the z.) I have my head trammed with the z locked so I want to have the z locked when I'm using a fly cutter or such to make sure the surface is as smooth and straight as I could possibly get it. Without the head locked I tend to get a .001-.003 edge where the tool takes the next swipe.

Rick

[yantra3d]

Thanks for the tip HackMax!

Just a quick update on some progress I've been making on the chip tray I've been making. I have been cutting a few things finally and the chips are taking over my little shop really quickly:



Here is my solution:



I used pop-rivets to assemble the panels that make up the sides of the tray and then I cut out some rubber strips to mount between with the tooling plate and the chip tray parts to help reduce shock/vibration and also to act as a seal for any coolant that may accumulate (not very likely with fogbuster spray unit!). I fastened the chip tray to the tooling plate with machine screws and vinyl washers (again to help with vibration).

I am finishing up a lexan door/shield and a rear shield that will swivel on a flexible hinge to accommodate the gas struts intruding into the rear wall of the chip tray region. I'll post more on that when it's finished and working.

You can see the process I used for making this chip tray in the pics below. I'll describe it here:

pic#1: I used a sawzall to rough cut the individual panels from 14g steel plate.
pic#2: Individually cut panels ready for the bandsaw for more shaping.
pic#3: The cheap HF bandsaw in vertical mode comes to the rescue.
pic#4: Panels all cut and ready for milling a 45' groove slot for bending.
pic#5: cutting a 45' groove using a countersink bit.
pic#6: more groove cutting for bend.
pic#7: close up of finished cut fresh off the mill. Took forever but looks good!
pic#8: using mill to drill holes for pop-rivet assembly. (jogging w/ keyboard)
pic#9: finished panel ready for bending.
pic#10: all panels finished slotting and drilling - ready to bend.
pic#11: my overpriced brake I used to bend the panels with.
pic#12: bending in process. I ended up using a hot air gun and it bent like putty!
pic#13: finished bend... not bad! A few seconds at 1000'F made a huge difference.
pic#14-15: more pics of finished bends on panel.
pic#16: test fit before painting.
pic#17: make-shift paint booth in my 10x12 shop (below zero temp outside!)
pic#18: I used a $7 spray can of hammered black metal even though I have a HF compressor and spray gun! lol
pic#19: panels are dried and ready for assembly.
pic#20: I used 1/8 pop rivets to fasten the sides of the tray panels together.
pic#21: Sides of tray are pop-riveted together and ready to mount to tooling plate w/ machine screws and some rubber seal.
pic#22: underside shot to show assembly via machine screws to mic6 plate.
pic#23: close-up of finish. Not too bad for a can of spray paint.
pic#24: close-up of rubber seal around tooling plate perimeter.

Next, I will be posting some progress on the limit switch covers and triggers I made and also some of the spindle conversion parts that I am now making. Too busy to keep up with any updates right now.

[yantra3d]

Here are some quick pics of the limit switch covers I made the other day:



The covers are very small - about .75" x 1.25". I used a 1/4 inch 2flute endmill for roughing and then a 3/16" 2f for finishing.

In the first pic you can see the mill taking a full depth cut with a 1/4" 2 flute endmill (cleveland brand from MSC). I'm satisfied with the finish overall but I would like to make some improvements.

The micro switches fit perfectly in their pockets (see pic#3) and the fit is so snug that I can't pop it out with any significant force. I think these will work well. I was very happy to see it float across the granite surface plate when I set it down on it. It glided like an air-hockey puck for about 5 inches!! There is a suction whenever you try to pick it up off the plate.

I also made some small adjustable triggers that mount to the front table slot. I will post those next and the spindle mod parts I have made so far...

[yantra3d]

I've been really busy lately and haven't had time to post my progress on the mill. I finally made some way covers for the Z and Y axis using some 10 oz denim. I used the typical spar varnish method and it was easy enough to make my own template without having to source one online. I made a test cover from some printer paper to test it and then came up with the measurements for both covers based on that prototype and it worked out well. As it turns out both of the covers are the same size in every way except for length.



I applied two coats of the varnish (10 hrs between coats) and waited 4 days for it to dry! I then coated it with a thin layer of Rubberized Undercoating which I picked up at the local Autozone. It's a 3M product and comes in spraycan form for less than $10 for 16oz can. I sprayed both covers and used the entire can. It also took days to dry!

The dimensions are 2.5" tall x 8" wide with 5/8" folds. The Z axis cover has 24 peaks or 48 folds and the Y axis has 15 peaks and 30 folds.

I'm not gonna lie - it was a time-consuming process. The spar varnish takes days to dry and has a hella potent odor. Making the folds takes a long time and it will wear the tips of your fingers raw like sandpaper. However, having way covers makes it worth the sacrifice imo. These work really well and there are no chips at all on the ways or ballscrew. The Z axis cover has very little flex and no apparent bulging at any speed after 2 weeks of use.

pic1: hanging varnish-soaked denim to dry (2 days+ per coat)
pic2: templates are drawn and Y axis cover is folded
pic3: close-up of Y axis cover folds
pic4: test fit of both covers
pic5: A quick and easy bracket to hold the rear Y axis cvr and btm of Z axis cvr. The Y axis cvr slips right under the brkt lip and can be easily removed to oil the ways between ops. The Z axis cover slide between the two side brkts and hold the btm of the cvr in place. Also very easy to remove for oiling.
pic6: cut out a quick mounting plate for the top of the Z axis cvr
pic7: mounted plate to top end of cvr using pop rivets
pic8: final cvr installed on the Y axis
pic9: small brkt I made to hold the Z axis cvr mounting plate via a tension screw (can be tightened with fingers)
pic10: photo of completed covers assembled and mounted

I'll be posting some pics of my spindle mod conversion...

[Quazar]

I have to say that your work is simply stunning!
Great job on the way covers (and the limit switches, and the chip tray, etc).
I'm taking my time with my conversion trying to do it right, but I can only hope my rig will look as professional as yours when I'm done.

Since I sometimes run MDF, I'm going to need something slightly different. Any dust that escapes collection floats in the air and will slowly find its way around most any mechanical barrier - over time the buildup can be significant. I was thinking of fitting some flexible air duct around the whole column (rectangular cutout for the head) and blowing in some filtered air to maintain positive pressure (and cool the electronics). It won't look as clean as yours, but I think it will work well enough for Z. Still noodling over protecting the X & Y axes...

Well done!
- Dean

[yantra3d]

Thanks a ton Dean! I like the vac idea a lot. I have been thinking about adding something like that to help clean up chips. I don't like blowing them around everywhere with a high-pressure spraygun. It seems like it would be really easy to just vac them up. The fogbuster produces very little extra coolant so I just use a brush right now. I actually like using a brush to clean up the chips. It helps to get me in the zone before I start making more chips. There's something very relaxing about it. lol

So, here's my spindle mod conversion I just finished up for the mill. I'm using a Goldline XT Ac servo motor w/ servostar drive that I picked up 2-3 yrs ago on ebay. I absolutely love the thing! It's rated at 1.81HP with a max speed of 3000RPM.



I'm using a 24 tooth pulley on the motor shaft and a 14 tooth pulley on the spindle shaft. This gives me a total speed of 5,145 RPM at a 1.714 ratio. I sourced the pulleys from SDP-SI (cnczone vendor) and bored the shafts out to fit both the motor and the spindle.

This is definitely one of the best mods that you can do to a mill this size imho. I can't tell you how liberating it is to get rid of the noisy gear-driven weak motor setup that comes stock with the base X3. This thing is so much more quiet and powerfull. I passed on the super X3 cause I knew that I could find a better motor for cheaper on the bay and swap it over to a pulley arrangement like this one and end up better off.

I am now working on the rigid tapping for it in the kmotion software. Right now it's sending back a differential encoder signal to the snapamp and I have closed the loop for spindle control in the kmotionCNC software. It runs off of step/dir signals directly from the kflop and is working really well so far.

Starting with the pulleys:

pic1: small pulley being drilled for custom-fit pulley shaft, then rough-bored with a 3/4" endmill.
pic2: using a 3-jaw chuck (haha) and a small 3/8" boring bar on the lathe to acheive final boring diameter for spindle shaft.
pic3: test fit to spindle shaft proves to be a perfect fit (beginner's luck)
pic4: large pulley was too big for lathe chuck so I got a chance to use my criterion boring bar (used ebay item) on the mill for this bore. It produced a nice finish even though I guesstimated the speed/feed rate for it.

Next up I'll post the pics for the side bracket parts that support the main motor mount plate.

[yantra3d]

Here are some photos from making the side supports for the spindle mod parts. There's two of these that support the top motor mount plate on either side of the head. They are also designed to act like ways for the motor plate to slide between when the tension screw in the back of the assembly is tightened for belt tension.

pic1: stock is rough-cut on the bandsaw, then squared up with a 3/4" endmill before being mounted flat for inner pocket and side profiles.
pic2: clamped in vise and mounted sideways on table for drilling and threading for tension beam mounting.
pic3: Using a Grizzly 2.5" face mill to finish up all outer edges.
pic4: finished part.
pic5: close-up of finish.

I'll post some pics from the main motor plate and assembly process next...