[CarveOne]

Yep, told ya, probably something simple. Going to a steel frame will be fun to do and will make it a much stiffer machine.

I probably have one more steel frame machine left in me but haven't made any decisions as to what it will be so far.

CarveOne

[Zygoat]

i do a lot of Ethernet wiring, i used to strip the wires as well. had nothing but problems, until i realized the same thing. DON'T STRIP THE WIRES! i have been there man, glad you got it sorted out. :banana:

[nickswimsfast]

Don't have a lot of time, start work early. Quick post.

I labeled the pictures with descriptions, so hopefully that should be enough. This was work I did over the weekend. Most importantly was strengthening the frame with cross members. Then my friends came over and helped. They took care of the z-axis riser blocks, and started on wiring up limit switches.

I am temporarily using a dremmel until a proper spindle arrives in the mail. I have purchased a TAIG headstock assembly from Nick at cartertools.com who sells TAIG equipment. It cost me about $300 shipped. He helped me select all the components I needed because the individual part #'s can get confusing. Need more nice and helpful fella's like him!

Anyways, I need to attach the z-axis assembly to the ballscrew still. Haven't completely decided how i'm going to do that, risk drilling into the plywood (and splitting it) or using some sort of adhesive/tape combination in tension.

I can just feel it coming together. I still need to address a grinding noise coming from my X-axis ballscrew at the limits. I suspect a binding issue due to inaccurate parallelism. Probably going to leave that off for the weekend.

[nickswimsfast]

The video goes in great detail of what was done. @ 5:00 minutes is where I show my most recent issue. Any ideas?

[ame=http://www.youtube.com/watch?v=b1nVsmIUJZs]‪CNC Build Video #4‬‏ - YouTube[/ame]

Let me know if you have any comments on the picture or video! Also ideas concerning the grinding noise is definitely welcome!

[jsheerin]

Your y coupler looks really deformed. If it's not just an effect of the video, can you adjust the motor mount to line up your motor any better with the screw? The couplers are supposed to compensate for slight axial and angular misalignment of the two shafts, but you want to get them as close as possible.

On the noise, it sounds like it occurs once per revolution of the screw? It's tough to tell - you can probably get a better idea standing next to it than we can through the video. If it's twice per revolution, it could be misalignment of the screw, either due to a bend or the ball nut pulling the shaft out of alignment due to its attachment to the carriage. If it's once per rev I would probably look for crap in the bearings first. Did you disassemble the bearings and clean them out? If not you might want to, and relube them when you put them back together. You might also try removing the motor and turning the shaft by hand to see if you can feel anything when you hear the noise.

[nickswimsfast]

Your y coupler looks really deformed. If it's not just an effect of the video, can you adjust the motor mount to line up your motor any better with the screw? The couplers are supposed to compensate for slight axial and angular misalignment of the two shafts, but you want to get them as close as possible.

On the noise, it sounds like it occurs once per revolution of the screw? It's tough to tell - you can probably get a better idea standing next to it than we can through the video. If it's twice per revolution, it could be misalignment of the screw, either due to a bend or the ball nut pulling the shaft out of alignment due to its attachment to the carriage. If it's once per rev I would probably look for crap in the bearings first. Did you disassemble the bearings and clean them out? If not you might want to, and relube them when you put them back together. You might also try removing the motor and turning the shaft by hand to see if you can feel anything when you hear the noise.

You were spot on with this. I loosened the motor mount and the noise disappeared. I disconnected a motor and turned both shafts, noise was gone in both. It's not a carriage/nut alignment issue, Yay! It is because of the motor alignment.

After retightening the motor, I put too much stress on the coupler and it broke. Now i have an excuse to buy some lovejoys. I did try to attach some flex tubing i had (for Mcguyver fix :P), but while forcing it on the ballscrew shaft i banged my hand pretty well.

All in all, I'm happy to be convinced the issue is motor alignment, because i can fix it.

[nickswimsfast]

Purchased the following via amazon:

Lovejoy Size L050 Solid Type Rubber Jaw Coupling Elastomer
Lovejoy Size L050 Jaw Coupling Hub With 5/16" Bore
Lovejoy Size L050 Jaw Coupling Hub With 1/4" Bore

It was a bit unclear what to get at first, so far anyone who uses linearmotionbearing ballscrews from ebay the ends are 8mm or 5/16", I think they were RM1605 ballscrews. Hopefully everyone will learn from my mistake or correct part selection. TBA when the parts arrive.

[nickswimsfast]

Couplers
The lovejoy couplers arrived. (Picture attached) The 0.3125 couplers don't fit on the ballscrew shaft. Measured with the calipers I believe something is "egg-shaped" or close to within a few thousandths because I measured both with my calipers. I believe the proper course of action is to manually increase the ID of the coupler since they are 10$ parts with some sandpaper on some sort of rotary device. Unfortunately i don't have a lathe. This plan is opposed to grinding the OD of the ballscrews ($$$). I'll have to get creative increasing the ID without a lathe.

Ballscrew grinding shennanigans
I also tested the motor as seen in the video to get a feel for whats going on inside the ballscrew bearings:

[ame=http://www.youtube.com/watch?v=50CI_aac2U0]No grinding noises - YouTube[/ame]

I am slightly concerned that I may have not have properly installed the bearings to the bearing mounts. If I recall correctly, there is two bearings (angular contact, radial) in the big mounting block; and one bearing (radial) in the smaller mounting block. I just don't recall any documentation or indication which bearing was which. I did not crack the case on the sealed bearing to lube them... I would have hoped i don't need to do that? I mean they are sealed after all...

I don't know that i'll even be able to move the bearings much at all anyways, as they are interference fit and I had to make do without a press. I was reasonably careful when i installed them as i don't have a hydraulic bearing press...

Thanks!
I want to thank jsheerin for his often insightful and very useful responses, and everyone has commented. Hope you all are finding some sort of interesting information in this thread. Let me know if you found something useful or entertaining, because it just fuels my fire to work on this even more!

[jsheerin]

Glad it helps. A bunch of people here helped me when I was building my first machine, so I'm happy to repay the favor.

My bearing blocks came with the bearings in them. Did you have to install yours yourself? I haven't used mine yet, so I can't provide much more input than that, but if there are bearing numbers on the bearings, you might be able to look them up and see which is a plain radial bearing and which are the thrust bearings (I'm guessing they're just deep groove radials).

[louieatienza]

Couplers
The lovejoy couplers arrived. (Picture attached) The 0.3125 couplers don't fit on the ballscrew shaft. Measured with the calipers I believe something is "egg-shaped" or close to within a few thousandths because I measured both with my calipers. I believe the proper course of action is to manually increase the ID of the coupler since they are 10$ parts with some sandpaper on some sort of rotary device. Unfortunately i don't have a lathe. This plan is opposed to grinding the OD of the ballscrews ($$$). I'll have to get creative increasing the ID without a lathe.

Ballscrew grinding shennanigans
I also tested the motor as seen in the video to get a feel for whats going on inside the ballscrew bearings:

No grinding noises - YouTube

I am slightly concerned that I may have not have properly installed the bearings to the bearing mounts. If I recall correctly, there is two bearings (angular contact, radial) in the big mounting block; and one bearing (radial) in the smaller mounting block. I just don't recall any documentation or indication which bearing was which. I did not crack the case on the sealed bearing to lube them... I would have hoped i don't need to do that? I mean they are sealed after all...

I don't know that i'll even be able to move the bearings much at all anyways, as they are interference fit and I had to make do without a press. I was reasonably careful when i installed them as i don't have a hydraulic bearing press...

Thanks!
I want to thank jsheerin for his often insightful and very useful responses, and everyone has commented. Hope you all are finding some sort of interesting information in this thread. Let me know if you found something useful or entertaining, because it just fuels my fire to work on this even more!

It's only a little more tha .002" too small. You could probably order an 8mm drill at Enco and it should self-center itself fairly well. Use a clamp and drill press...

I also have fixed and free end bearing blocks from Chai. Although they're decent, I found that my screws turn tightly when attached; I don't know if it's the rubber seal. The locknut rubs against the housings, even though I have the spacer installed (I'm using BK12/BF12.) I bought two sets from Marchant Dice and the differences are night and day. The screws turn smoothly with almost no friction, and the fit and finish was eons better. Of course, the Marchant Dice ones were over twice the cost of Chai's. (they're both made in China) With both, I did not have problems getting the journal to slide in; they were both a high-tolerance fit.

edit: is it possible there's a problem with the bearings in the stepper motor? Have you tried running the stepper sans screw?

[nickswimsfast]

Well after fumbling around I returned the 3-way switch and bought a single pole switch and wired up my spindle motor to work with the switch. Next time i'll buy the housing that allows for a screw to go from the side so I can permanently affix it to the frame.

Since metric drill bits were no where to be found locally, I purchased a number drill set (29 different drills black oxide craftsman brand from Sears.) In order to bore out the 0.3125 lovejoy connector to 8mm I used a 21/64 which is slightly oversized.

I clamped it as shown and drill it by hand. The bit would not fit in the work range of my TAIG mill. This also provided the opportunity for the drill bit to self center as I bored it out. It went quick.

I then took a piece of scrap wood cut it to provide a simple shim, I imagine a more rigid shim would be best longterm, but I wanted to test it. It worked. No grinding noises, alignment of the motor was better as I loosely attached it then rotated it a few turns to allow the motor to self-align a bit, and then tightened it down.

I still hear the grinding noise on the other two axis, and I need to go through and install lovejoys on them. This will be a bit involved because I screwed the motors to the frame, I'll need to bore them out to utilize bolts for adjustability. This was unwise, but a harmless lesson.


-----
jsheerin, I did have to install the bearings. They came separate. It was a bit nerve racking to install them as it is an interference fit. I'm wondering if the black locknut even serves any purpose. *shrug* hasn't effected performance yet, and the bearings seem to be locked in place, there is no play in the screw along the axial direction, much less any direction.

louieatienza, link for marchant dice? Sounds like you have it figured out.

[louieatienza]

Well after fumbling around I returned the 3-way switch and bought a single pole switch and wired up my spindle motor to work with the switch. Next time i'll buy the housing that allows for a screw to go from the side so I can permanently affix it to the frame.

Since metric drill bits were no where to be found locally, I purchased a number drill set (29 different drills black oxide craftsman brand from Sears.) In order to bore out the 0.3125 lovejoy connector to 8mm I used a 21/64 which is slightly oversized.

I clamped it as shown and drill it by hand. The bit would not fit in the work range of my TAIG mill. This also provided the opportunity for the drill bit to self center as I bored it out. It went quick.

I then took a piece of scrap wood cut it to provide a simple shim, I imagine a more rigid shim would be best longterm, but I wanted to test it. It worked. No grinding noises, alignment of the motor was better as I loosely attached it then rotated it a few turns to allow the motor to self-align a bit, and then tightened it down.

I still hear the grinding noise on the other two axis, and I need to go through and install lovejoys on them. This will be a bit involved because I screwed the motors to the frame, I'll need to bore them out to utilize bolts for adjustability. This was unwise, but a harmless lesson.


-----
jsheerin, I did have to install the bearings. They came separate. It was a bit nerve racking to install them as it is an interference fit. I'm wondering if the black locknut even serves any purpose. *shrug* hasn't effected performance yet, and the bearings seem to be locked in place, there is no play in the screw along the axial direction, much less any direction.

louieatienza, link for marchant dice? Sounds like you have it figured out.

World of CNC - Marchant Dice Ltd - CNC Shop CNC Machine Leasing CNC World CNC Parts Bearings Pillow Block Kress Motor, Spindles Ballscrew Hiwin Trapezoidal, Nut Round Rails Coupling Desktop Packages

[jsheerin]

The square locknut should hold the screw rigidly against the bearings, afaik. There are different ways the bearings can be arranged and I'm not an expert on that, but for one example, the bearings are a matched set and when you tighten the bearings together by tightening the locknut, there is a predetermined preload set up in the bearings. The locknut allows the bearings to resist thrust loads in one direction (the other direction would be resisted by the shoulder on the screw). So if it's not tightened, the only thing preventing movement is friction between the screw shaft and the inner bearing races. If you tighten it and things bind, assuming you're not tightening it excessively, something is probably wrong (which is definitely possible with these blocks).

[louieatienza]

The square locknut should hold the screw rigidly against the bearings, afaik. There are different ways the bearings can be arranged and I'm not an expert on that, but for one example, the bearings are a matched set and when you tighten the bearings together by tightening the locknut, there is a predetermined preload set up in the bearings. The locknut allows the bearings to resist thrust loads in one direction (the other direction would be resisted by the shoulder on the screw). So if it's not tightened, the only thing preventing movement is friction between the screw shaft and the inner bearing races. If you tighten it and things bind, assuming you're not tightening it excessively, something is probably wrong (which is definitely possible with these blocks).

There's two ways, either the angled bearing races face each other, or they're back-to-back. Where they face each other, they have a spacer between the inner rings and the outer rings are "pressed" together for the preload. The pressure angle of the race determines the thickness of the spacer and therefore the overall "strength" of the system. This is the stiffer of the two assemblies. Most commercial fixed blocks are made this way (I believe even Chai's though I haven't opened it up yet.) In the other scenario, the spacer is put on the outer rings, and the locknut of the screw determines the preload. This is the weaker of the two assemblies, since it cannot be preloaded as much, and the "bearing" surfaces of the bearings are closer together.

Looking at Chai's blocks again, I think it's the loose spacer or bushing provided that goes under the locknut and past the seal (don't know its proper name) that is not long enough (at least on mine) causing the locknut to rub against the seal and bearing retaining plate. In fact when I mount my ballscrew to the Marchant Dice blocks (Kuroda GG 15-5 C5 ground ballscrew) I can hold the the bearing block, and spin the screw easily by simply pushing and pulling the ballnut.

I'm pretty sure when you insert the bearings, they should slide in like a tight piston, without using much force. It is possible the inner rings got deformed when inserted, and that might cause some "roughness." Also I'd check to make sure the bearings are both pointed in the right direction with the spacer in the middle (if there is one.) Finally, I'd make sure thar the bushing under the locknut is the right length so that it does not interfere with the housing or seal... You can adjust the bearing journal carefully bu adhering a piece of fine automotive sandpaper to a stiff piece of cardboard, wrap it around the journal and "roll" the screw carefully, checking each time, until the bearings just slip on.

[JohnZ]

Since metric drill bits were no where to be found locally, I purchased a number drill set (29 different drills black oxide craftsman brand from Sears.) In order to bore out the 0.3125 lovejoy connector to 8mm I used a 21/64 which is slightly oversized.

I purchased my 8mm lovejoy couplers from Brent Hubbard on eBay. They have always given me stellar service, and I typically have the parts I ordered within 48 hours from payment to having them in my hand. The 8mm end is about 8 bucks shipped.

If you find that you really need to get an 8mm coupling, I highly recommend them. (I have no affiliation with them at all, just a happy customer for many years) - Their eBay store - Power Supply Wall Transformer, Lovejoy Drive Couplings items in HUBBARD CNC INC store on eBay!

[nickswimsfast]

About linearmotionbearing ballscrews
So while browsing 5bears website again, i stumbled on this page: 5 Bears - Y axis ballscrew

This cleared up quite a bit about the ballscrew assembly. I took off the flange bearing housing lid. This exposed the bearing, which was securely in place, I sprayed a bit of general purpose white lithium grease in there and closed it back up as it looked ok. The bearing serial identification can be seen in the pictures.

It has been nearly two years but I believe the ballscrews came attached to the fixed bearing block with both bearings already installed. The free end bearing block was not installed.

I examined the locking nut for the ballscrew carefully and realized that I had installed them in the incorrect orientation (not that there were instructions :P) The raised surface needs to face the lid. This preloads the bearings and removes the play in the screw. I fixed this on all three of my ballscrews.

I am a little confused how the nut set screws work... You can see the discoloration on the threads where it looks like the set screw goes. I wonder how it applies pressure to lock the nut. Is it just applying a compressive force through the threads in the discolored region? So wierd.


no grinding noises
While figuring out the ballscrews, installed the lovejoy couplers on the remaining 2 axis. So they are now installed on all three. In order to install them i bored the 0.3125 couplers to 21/64, since i needed to fit on an 8mm shaft (where i did not have an 8mm drill or a lathe to clean them up on. I had to widen the mounting holes for the motor mounts in order to obtain more adjustability to get the motors aligned. After installing the lovejoy couplers, all the grinding noises disappeared. That was a relief.

Attaching the table
I had to shim my table up with 2x4's in order to allow my endmills to touch off on the table. The lowest point for the spindle is roughly 1.5" above the table. This leaves me some flexibility in fixturing, and drilling holes through. I then used the machine to drill through holes to bolt the table to the saddle.

First g-code program
I wanted to do something other than manual jogging. I wrote a simple 4 line g-code routine to mill countersink holes for the bolts that hold the table down. (picture) Time permitting I'll finish this, as I was playing with the speeds and feeds.

Stupidly, this was my first cut:
-1100 RPM (Tachd)
-29 IPM
-0.375 diameter center cut end-mill
-0.040? DOC

This did not work well. I discovered why having a wood frame is bad. The "unrigidity" of the machine allowed it to chatter badly. I quickly shut the machine off, and re-evaluated my arrogance.

Raised the RPM to get it out of resonance, and slowed the feed. Better Cutting parameters were:
-4360 RPM (Tach'd)
-3 IPM
-0.375 diameter center cut end-mill
-0.150 DOC

Seems I have a bit to learn about machining wood, but it shall be good fun!

I'd post the machine cutting a circle, but it's not that exciting. I'll post something else later. I still need to attach limit switches and the E-stop :P

[louieatienza]

The "discoloration" is actually a copper or brass plug that pushes against the threads when the setscrew is tightened, without deforming the threads...

Usually when I cut wood, it's at about 12,000-18,000 rpm, and about 75-150 ipm. I find lower rpms temd to make the bit "catch" the wood grain.

edit: Check this build out; here some "adjustments" were made to get the bearing blocks to run smoother...
(last post)
http://www.cnczone.com/forums/vertic...tml#post982544

[nickswimsfast]

So I've finished attaching the table to the saddle plate. I've countersunk, all the bolts with .25" of clearance to the top of the table. I measured with the spindle, and I need to level the table by milling 0.090 off the top. It appears the right side is raised, in effect the whole surface is lifted on the right. I could shim and machine. Or just machine.

I tested using double sided tape to hold down this block, and did a series of tests. @6370 rpm .120 doc, 0.375 diameter cut 30 ipm was the most aggressive cut I did. It's interesting to note that @ .085 doc resulted in resonance.

I desire more rigidity. but who doesn't? I still intend to make a metal frame, but I'll refine and run this for awhile first.

[nickswimsfast]

Here is a status update.

[ame=http://www.youtube.com/watch?v=7WYZMfLxzHc]CNC Build Video #5 - YouTube[/ame]

I'm still contemplating the way cover/bellows to protect the ballscrews and rails from chips. I've started folding that plastic, but it doesn't like to hold a crease. I may try again with an iron to help hold the edge, or a different material.

[nickswimsfast]

It's been awhile since I've worked on this machine, AND documented it. I have made some progress on it since I last posted. I came up with a temporary dust containment box for the table, and actually ran the machine to make at least one "real" part seen below (2.5 hour run).









While disassembling the machine and moving to my new job in Washington, a power line wire came loose in my control box, and fried my old hobbycnc control board. 70$ visibly fried dead. I'm not even going to try and fix it, as now I have a valid excuse to upgrade to an opto-isolated gecko g540 system. I also have ordered some DB9 motor cables from CNC Router Parts with the built in resistor. I am happy to stop running inappropriately sized and unshielded wire. While upgrading the system I will explore rebuilding various aspects of the machine to improve its overall quality.

To get the machine at a usable standing height, I built a simple base. (See attached pictures for your enjoyment.) I quickly modeled it up on solidworks and went to Home Depot which is now a block away... DANGEROUS to my pocket, but so convenient. The attached pictures show the build of the stand, and re-assembly of the CNC machine on it. It should be noted I put a 1/2" plywood plate on top of the table before re-assembling the machine.

Lots more to do...


-------
Between graduating for my second time and working my second internship, I haven't had time to keep this updated. I recently moved to Washington to begin working at a major northwest aerospace company. Now financial burdens are lifted from this project, and my goals for the machine have changed.

Apologies for the "necro" post, I believe it is warranted as I intend to pick things up again.