Zach's Homebrew CNC Mill

[jsheerin]

I've just about finished his book, so yeah, I do reference it a lot. :>

I'm sure stress relieving would send your epoxy up in smoke and I'd be afraid even local heating would damage the epoxy. From what I've found mild steel needs 1100-1200F. My current plan is basically what you said - weld, stress relieve, machine as necessary, maybe stress relieve again, then grout / replicate joints for assembling the various parts of my frame and attaching components to it.

[RotarySMP]

Nice.

0.018" over 12" is not much over the mating surfaces. Why down you scrap then into alignment. A simple scrapper ground into the tip of a file will be fine for this job.

[jsheerin]

You can also buy a HSS scraper at MSC for about $20. That's what I used to scrape some aluminum - works fine. You just have to grind a radius on the tip and then sharpen it on a stone.

[louieatienza]

That mill came out awesome! Great job...

My suggestion for coating the steel would be Line-X or similar bed liner material... durale, impact resistant, and might even have some dampening effect as well....

[mwood3]

Nice build, but im curious about your Y axis setup. hard to see from your pics posted but did you use 2 bearing blocks for the Y or 4? I have a set of linear rails I was wanting to use for a similar application but was worried that 1 block per rail wouldnt give enough support. also how long are the bearing blocks you used?

Thanks

[tskguy]

Wow great job, it looks great! But what a teez, I want to see it cut something!

E

[Zach_G]

Sorry ya'll haven't posted a status update recently. I've been rather unmotivated this past month to do a whole lot, plenty of other summer activities to enjoy! Got a ways to go before it starts cutting:

To answer your question mwood3, yes there are only two bearings on the Y and Z axis. That said, they're the long (~5.25") block version of 35mm bearings, and also use 4 races, each of which has 4 contact points in the groove and are heavily preloaded to be very rigid. I don't think I'd rely on just 2 regular blocks otherwise. For reference my table is 8.25" wide and there will be about 10" of Y travel.

I did manage to weld up a stand with casters and leveling feet that should make things easier to move around and work on. The mill is still torn down while I plan for squaring the column, adding waycovers and installing limit switches. It has gotten a few coats of rustoleum though, seemed like a logical choice so we'll see how it holds up. Right now I'm getting set up to fill the base with the epoxy/granite/sand/iron oxide mix. doing some experimentation I decided on a 11/58/21/10 % mix by weight respectively which is somewhat similar to other mixes I've seen posted on the zone. May need more epoxy since my experimentation was with water which is much less viscous.

Im also working on a drip tray. Probably from 0.06" steel. Cant decide if its worthwhile to have a sheet metal shop bend it up for me or just cut it out myself and weld up all edges. Need to get an estimate locally if I can find anyone even willing to take on such a small job.

I also need to rework the fixed bearing mounts for the ballscrews. All three are jacked up in some way. They're from China, linearmotionbearings2008 on ebay. One the bearing hole was bored crooked to the mounting surface so it takes a good bit of forcing the screw to align it. Another the rubber seal is so tight it actually started smoking! The last bearing hole was bored too deap so the flange that holds the bearings in has close to 0.06" gap to the bearing race. Under large forces that translates to .06" of backlash... All these problems seem to be a common from what I've read. Oh yeah and they screwed up the lengths so I ended up having to redo the end machining anyway, at least it was longer rather than shorter. I emailed Chai about these issues and go figure, didn't hear anything back. Guess that's the gamble we play when going cheap.

Phew, plenty more to do, alot being that annoying detail work.

[TarHeelTom]

Im also working on a drip tray. Probably from 0.06" steel. Cant decide if its worthwhile to have a sheet metal shop bend it up for me or just cut it out myself and weld up all edges. Need to get an estimate locally if I can find anyone even willing to take on such a small job.

0.060 steel seems like a lot of steel for just a drip tray.

But you might try your local HVAC shop. They could whoop you up a drip tray in just a few minutes.

Tom

[Zach_G]

Well I've made some progress in the past month. After weeks of waiting I finally took delivery on the drip tray. Probably overpaid too, at least it came out looking nice, that is until I unleashed my welding talents (or lack thereof). I went with 0.06" steel sheet since it also needs to support a full enclosure which is still tbd. Might just throw together some PVC pipes and shower curtain, I really want to get this thing running soon! Rustoleum primer and enamel was used on the frame and mill. Seemed to scratch off pretty easy with just a fingernail, so I tried 1 part appliance epoxy on the drip tray. A little better but still can scratch it with a fingernail. Should've gone with a 2 part epoxy. Oh yeah, the epoxy granite filling went well for the base of the mill. That sucker is solid now, around 350lbs. I'll have to take pictures and explain the process when I fill the Z column later. Figured its a good idea to wait until after placement on the base while I can still lift it myself. Getting the filled Y column in the drip tray was no simple task. As you can see, its way more fun to rig up levers and blocks to lift a few hundred pounds rather than buy an engine hoist. That's how they built the pyramids. Or maybe it was aliens. Didn't want to wait on them though so I went with blocks and levers.



You may also now notice there are 4 blocks on the Y rails rather than the original 2. I was mulling mwood3's question on 1 blocks per rail and messing around with a 6" vise on the table. The whole thing just looked wonky and after a quick calculation with machining forces and comparing to bearing ratings, it seemed certain conditions would get dangerously close to the bearings max moment. Fortunately I was perusing ebay and serendipitously discovered another set of NSK bearings! Their mounting height is a little shorter and will require some standoff plates but this allows me to take the Z bearings and put them on the Y. In order to not lose the10" Y goal I bastardized 2 bearings by removing the end caps and sandwiched them together so that their balls now recirculate through both bearings. This saves me nearly 2" off ball return tubes and thus giga-bearing is born! Probably destroyed the P4 rating since the balls are individually sized for each track but wont be noticeable due to the limitations of the ballscrews. They still slide smoothly which is what concerned me most. I also machined a new saddle plate to bolt the X and Y axis together. This will connect to the old saddle plate still attached to the X axis, making life so much easier when it comes time to combine the X and Y since now I dont have to turn bolts in confined spaces. The only drawbacks I can see is a loss of 0.75" of Z travel due to the extra saddle plate, and the stiction of 4 heavily preloaded bearings is up to 80lbs. May have to turn on backlash compensation in Mach3 since I doubt the cheap chinese ballscrews have the rigidity to push 80lbs without deforming several thousandths. Speaking of, I've also begun overhauling the ballscrew bearing ends.



While the ballscrews and nuts themselves are satisfactory, the cheap pieces of Chinese garbage bearing ends need some serious rework. The rubber seals are so tight on the shafts that they actually start smoking when turning quickly. Off with their seals! Next, all 3 needed nearly 0.025" of shims between the outer bearing race and housing endcap. The one housing that wasnt a press fit would let the bearings float that amount, which shows up as backlash. That's alot of backlash! Lastly, since these AC bearings are not duplex pairs, I decided to shim between the inner races to take out any additional backlash. I managed to destroy 2 sets of AC bearings pressing them off the shaft and out of the housing. Probably more my fault for not having a proper press, but at $6 a replacement from VXB.com its no huge loss. To shim the one good set left I placed arbor spacers (complements of McMaster) between the inner races until the outer races just barely grazed eachother under finger pressure, and then added another 0.001" shim for good measure. This set required 0.006" of shim, though I didnt notice that amount of backlash before due to the press fits probably helping to offset the races some. Could use a little more but I'm hesitant the ABEC1 tolerances would survive heavy preload for long. You can see the super high tech test rig. With the shimming it's getting 0.0015" movement from ~80lbs thrust. Tis all for now, waiting on some boxes with parts to continue. Then I can attempt squaring Y and Z columns once axis are reassembled.

[louieatienza]

Well I've made some progress in the past month. After weeks of waiting I finally took delivery on the drip tray. Probably overpaid too, at least it came out looking nice, that is until I unleashed my welding talents (or lack thereof). I went with 0.06" steel sheet since it also needs to support a full enclosure which is still tbd. Might just throw together some PVC pipes and shower curtain, I really want to get this thing running soon! Rustoleum primer and enamel was used on the frame and mill. Seemed to scratch off pretty easy with just a fingernail, so I tried 1 part appliance epoxy on the drip tray. A little better but still can scratch it with a fingernail. Should've gone with a 2 part epoxy. Oh yeah, the epoxy granite filling went well for the base of the mill. That sucker is solid now, around 350lbs. I'll have to take pictures and explain the process when I fill the Z column later. Figured its a good idea to wait until after placement on the base while I can still lift it myself. Getting the filled Y column in the drip tray was no simple task. As you can see, its way more fun to rig up levers and blocks to lift a few hundred pounds rather than buy an engine hoist. That's how they built the pyramids. Or maybe it was aliens. Didn't want to wait on them though so I went with blocks and levers.



You may also now notice there are 4 blocks on the Y rails rather than the original 2. I was mulling mwood3's question on 1 blocks per rail and messing around with a 6" vise on the table. The whole thing just looked wonky and after a quick calculation with machining forces and comparing to bearing ratings, it seemed certain conditions would get dangerously close to the bearings max moment. Fortunately I was perusing ebay and serendipitously discovered another set of NSK bearings! Their mounting height is a little shorter and will require some standoff plates but this allows me to take the Z bearings and put them on the Y. In order to not lose the10" Y goal I bastardized 2 bearings by removing the end caps and sandwiched them together so that their balls now recirculate through both bearings. This saves me nearly 2" off ball return tubes and thus giga-bearing is born! Probably destroyed the P4 rating since the balls are individually sized for each track but wont be noticeable due to the limitations of the ballscrews. They still slide smoothly which is what concerned me most. I also machined a new saddle plate to bolt the X and Y axis together. This will connect to the old saddle plate still attached to the X axis, making life so much easier when it comes time to combine the X and Y since now I dont have to turn bolts in confined spaces. The only drawbacks I can see is a loss of 0.75" of Z travel due to the extra saddle plate, and the stiction of 4 heavily preloaded bearings is up to 80lbs. May have to turn on backlash compensation in Mach3 since I doubt the cheap chinese ballscrews have the rigidity to push 80lbs without deforming several thousandths. Speaking of, I've also begun overhauling the ballscrew bearing ends.



While the ballscrews and nuts themselves are satisfactory, the cheap pieces of Chinese garbage bearing ends need some serious rework. The rubber seals are so tight on the shafts that they actually start smoking when turning quickly. Off with their seals! Next, all 3 needed nearly 0.025" of shims between the outer bearing race and housing endcap. The one housing that wasnt a press fit would let the bearings float that amount, which shows up as backlash. That's alot of backlash! Lastly, since these AC bearings are not duplex pairs, I decided to shim between the inner races to take out any additional backlash. I managed to destroy 2 sets of AC bearings pressing them off the shaft and out of the housing. Probably more my fault for not having a proper press, but at $6 a replacement from VXB.com its no huge loss. To shim the one good set left I placed arbor spacers (complements of McMaster) between the inner races until the outer races just barely grazed eachother under finger pressure, and then added another 0.001" shim for good measure. This set required 0.006" of shim, though I didnt notice that amount of backlash before due to the press fits probably helping to offset the races some. Could use a little more but I'm hesitant the ABEC1 tolerances would survive heavy preload for long. You can see the super high tech test rig. With the shimming it's getting 0.0015" movement from ~80lbs thrust. Tis all for now, waiting on some boxes with parts to continue. Then I can attempt squaring Y and Z columns once axis are reassembled.

Nice work so far! Like the color sccheme as well... I have the same bearing blocks and will be dismantling them in a similar fashion soon...

[jsheerin]

Looking nice! I have those bearings too, so thanks for posting the information.

[Zach_G]

Good luck disassembling and not destroying bearings. A word of advice, remove the rubber seal from the housing first to ensure you can press on the outer race of the bearings. I didn't and the majority of pressure acted on the inner races, which popped it off the outer race.