[Benonymous]

Well Zone, the time has come and I have begun building my machine. As the title suggests, the frame will not be welded. Instead, all joints will be bolted via tapped holes in the steel. The basic material I'm using is 75 X 50 X 4 mm RHS (rectangular hollow section). My build log will be a "warts and all" account and so I already don't apologise for being verbose and possibly getting sidetracked during the process of documenting the build.

Here is a basic render of the machine done in SolidWorks



The motors are Yasgawa brushless servos obtained from "FA Parts" on eBay. I have all THK SR20 trucks and rails. All drives will be via ballscrew with a single central screw on the X axis.

[Benonymous]

Okay, let's start...

I have been planning to build a machine for over a year and I've been lurking on the forum for much longer than that. The motivation to get planning started when a friend of mine who is a very skilled machinist and devout scrounger, gave me an old blueprint scaling machine or part thereof which yeilded enough parts for driving an x,y and z axis.



All good eh? Now what was missing was some form of linear bearings so I did a lot of searching on eBay and other places and came up with a vendor called "FA Parts" who had free shipping to Australia. From them I bought two sets of THK rails with four-row trucks which can be oriented in any way and still work.



:rainfro:

Armed with all this good gear, I began designing. Now I'd like to offer a little advice at this point and that is that you should thoroughly investigate all scrounged components to establish whether or not they will do the job you want. You have to be heartless about this and discard anything that hasn't got the tolerance, size or quality you need for your machine. if you design the machine around dud components, you've got a long and bumpy road ahead of you.

[Benonymous]

I was planning on purchasing a set of brushed servos and controllers from a local company but as I have FA Parts as a saved seller, I came across a super deal. A set of Yasgawa (Japanese) 400Watt brushless servos with 1200 line encoders, controllers and a Mach 3 interface board! Way too good to refuse for almost $600 less than the local guys with their Chinese servos and electronics.





No more excuses, I had all the components and a design based on them.

[Benonymous]

If you're going to build a machine out of 4mm wall RHS, you're going to need a chop saw. I don't own one but my friend the machinist, lets call him "Stig" from here on had one that he'd scrounged. The downside to this was the busted trigger switch. Stig made a deal with me that I could use the machine if I put in a new switch, the upside to this was that he had a new switch to put in the saw. SO I agreed. I got the thing on the bench, disassembled the handle and removed the trigger. It was completely different to the one he gave me as a spare

Two hours work with a Dremel and a spacer or two..



Success!

Now I had the means to cut up the RHS I'd had delivered the previous day.

When you're working with continuously welded tube you probably don't care terribly if its exceptionally flat and square on its faces. You're probably using it in a structural assembly that's got fairly broad tolerances. However, when you're making a 3 axis CNC router, these properties become much more significant.

So how square and straight is it?



Pretty good actually!

I used a lngth of thin steel wire to check for straightness by stretching it the length of the piece (4m long) and looking for gaps.

All in all I was impressed with how good it was. I made a note of the straightest/flattest end of the steel to cut the main X rails from.

[Benonymous]

Due to the pre-payment of a case of the Stigs favourite beer, I had him on-line to make some parts for me. The first bunch of bits were the right-angle brackets to bolt the bed of the machine together with. I had already bought an 8m lenth of 50 x 50 x 5mm angle iron to cut them from. When I suggested that this was adequate for the job, his nose wrinkled and he expressed a lack of confidence in my choice. "I've got some 75 x 75 x 10 angle at work" he said and I'll grind the faces square for you. An offer too good to refuse.



Here they are, all 12 of them.

Stig ended up facing only one side and cleaning up the edges.

From here I had a thrilling task ahead, drill 96 holes!

I marked all of the brackets out and then used a centre drill to make pilot holes.



Centre drills are good for doing pilots because they have practically no runout, are very stiff and if you bore far enough, you get a tapered lead in to the hole.

Drilling drilling drilling drilling drilling drilling drilling drilling drilling drilling drilling



( You'll notice I'm standing on a milk crate to operate my bench drill. I built the bench high on purpose over ten years ago and it's the only one I have. When i got the bench drill, it was designed to go on a lower bench...obviously!)


Here's the result.



Hoorayyy! :banana:

[hemsworthlad]

Ha ha you either built the bench very tall or like me you got the short arse decease. .Lol

Keep up the good work, Will be watching.

[Benonymous]

Ha ha, yeah I built the bench when I was making musical instruments it's about 1000mm tall. I built a work table to do this build but if I'd bolted the bench drill to that I wouldn't have been able to move it around.

[Benonymous]

I'll keep going with the posts until I've got the progress up to date for today.

So I got all 12 brackets drilled, then moved on to the RHS for the machine base. I started marking out the RHS and almost had it finished when, just for laughs, I thought I'd line up all the brackets over the marking out to see if the holes lined up OK.

They didn't....


by quite a few millimeters....


Not happy.


So I got two brackets, lined them up on the marking out and measured the discrepancy, almost 5mm. I went back over all my measurements and checked them in triplicate. I presumed I'd made an error of 5mm somewhere so, like a dolt, I re-did all the marking out with another 2.5mm incorporated. Then I grabbed two more brackets randomly and tried them out. DAMN!!!! they were out by 4mm!!

It was the brackets that were out wasn't it...

I checked all of them and after a bit I figured out that Stig had cleaned up the free end of all the brackets on a grinder. He'd just cleaned up the edge of the rolled steel for appearance, he hadn't done it square. I'd used my lovely Mitutoyo square on these cleaned up edges and marked them up off that edge.

GAAAAAAAHHHHHHH!

I was about to chuck all of the brackets out and ask Stig to do me a new set but the thought of that was just too humiliating. I looked them over and chose instead to re do the marking and use the edge of the actual angle (which I should have done in the fist place)

Fortunately, there's lots of meat on these so the end result is no less strong or structurally sound.



Believe me, I really took my time doing this and gave myself an imaginary kick up the backside for rushing into the job in the first place!

After drilling all of the brackets a second time, I finalised the marking out on the RHS and ...

drilling drilling drilling drilling drilling drilling drilling drilling....



BTW I used an old keyboard stand I had to make a feed-in bench for the drill.

After drilling I moved on to.......

tapping tapping tapping tapping tapping tapping tapping tapping



Annnnnnd finally, a bit of test assembly.

Rails on..



Bracket test and a test fit of the RHS




Both X rails.


And that's it, up to knock off time today!

[chimpera]

Hi, I like your build but it has week point in its design. The distance between the front and back gantry bearing blocks is insufficient. I know you will lose some travel but I think it would be worth the gained stiffness to widen the stance of the gantry. Also, it would be a shame to use unsupported rails for the z axis given the quality of all of the other components your using. Please keep going I think your onto a winner.

[Benonymous]

I agree chimpera, I'll probably do some re design on the gantry. I don't want to lose too much travel as this is already a modest machine from travel perspective. I hear what you're saying though and it would really beef the machine up and reduce any possibility of racking with the single X screw.

I've already dropped the unsupported rails and ordered some nice SSR20 ones from the new FA shop on eBay which is called "FA System" The SolidWorks render I put at the top of the page is really the MK1 version but the MK2 model is only a couple of components so far, brackets and rails

Thanks for the constructive comments, I hope to hear more as I progress :cheers:

[Benonymous]

Now for a salutary tale about using scrounged components.

I mentioned in a previous post that the Stig had gifted me a plans machine which I extracted a couple of ballscrews from and then began happily designing my machine around these drive components. I bought a set of THK SR20 rails for my Y axis to match up with the longer of the two screws I got from the plans machine.

Here's our subject..



So far so good. Anyway, now that I come to actually building the machine, I find that I need to get a bit of machining done on the screw. I need to get the nut off the screw so I use a method where I feed in a bit of brass tube as I unscrew the nut to keep all the ball bearings in place. This worked well with the similar ballnut on the Z axis. This time however, Murphy was watching and some of the balls escaped:tired: All things considered, Murphy actually did me a favour this time!

The escaped balls turned out to be plastic

Now plastic is just not going to stand up to being thrashed on a CNC router so I had no choice but to replace the screw. While I was at it, I considered that having two 5mm pitch screws and one 1/4 inch pitch on the Z wasn't really what I wanted, so I ordered two metric screws from "linearmotionbearings2008" off eBay. These are cut (I think, not rolled) screws and have decent looking ballnuts on them, 16mm diameter and 5mm pitch. All my screws are the same now except for the X screw which is 25mm dia I will have to wait another couple of days yet to get them as they're in customs at present.

Just as an aside. I did try to save the ballnut by putting steel ball bearings in it. I could only source 3mm balls here and they were fairly close to the plastic ones but the result with the slightly smaller balls was an unacceptable level of backlash. I was dreading having to get 200 odd balls in the nut but it wasn't that difficult. Here's the procedure..

Step 1. Take off recirculation tubes.

Step 2. Feed ballnut on to shaft leaving it at one end with plenty of shaft to feed through.

Step 3. Start dropping ball bearings in and turning shaft to feed them in to the thread.

Step 4. Keep feeding in balls until they show at the opposite end of the nut.

Step 5. To make sure the balls pack in get a piece of wire and push the emerging balls back until they won't go back any more. Don't fill the nut up too much the balls should be in the thread not coming up the recirculation tube holes

Step 6. Fill the recirculation tube with balls and use a blob of ordinary grease to block both ends of the tube so the balls don't all run out when you're trying to re fit it.

Step 7. Carefully re fit the recirculation tube without distorting it. Don't force it and take out a ball or two if it won't go fully home.

Step 8. Repeat for other side.

I won't say it's not a fiddly, time consuming job but it can be done!

[RomanLini]

Great build thread Benonymous.

Did you ever get that thought when you were drilling and tapping squillions of holes, that you could have just brought out the welder and gone bzzzt bzzt?

I've been thinking of doing a steel tube machine and I think I would choose welding, based on all the annoying drilling and tapping I had to do on my original machine.

[Benonymous]

Hi Roman.

I'll admit the idea of a BZZT here and a BZZT there and the job is done is very tempting but my idea with the drilling/bolting/tapping was that I'd be able to make small adjustments to the structure.

Welding is a much faster way to build, until you have a dimensional or alignment problem. With my welding skillz, these tend to add up and I'd hate to be within sight of a finished machine and find that all the accumulated errors had combined into a major problem.

I've seen lots of welded steel machines made on the forum and I'd be proud to own any of them. I wanted to make a machine for myself that would suit a person with no welding skill or facilities (I don't own a welder) even though I know I'll be thoroughly sick of drilling and tapping by the time I'm finished

[Benonymous]

Hoorayyy!

My new screws have turned up! Thanks Chai :-)





I was wrong, hey are rolled screws, not cut/ground. No matter though, the Warners were rolled too.

First impressions, very nice surface finish, ballnuts show no backlash at all (with the strong arm test). As you see, I have couplers and bearing blocks I didn't have these for the old screws so it just made sense to get them all in one hit, machined to fit.

Oh well, back to the SolidWorks drawing board for a bit of a re-design...

[hemsworthlad]

Hi Ben,

Chai's screws are great for the cost and work perfect if fitted correctly, I make and sell small machines and bought loads from Chai now never had a problem. I have had some on my machine coming up 2yrs and worked them hard in prity ****y conditions cutting ali plate 99% of the time and still no signs of wear or backlash.

Tip with the flexi couplers, put a smal indent on the shaft where the grub screws touch. If not they will slip ever so slightly and over time this will lead to positional error, basicly the tiny screws cant grip enough to hold.
To be honest I dont like them and have now started using the split type that clamp, they grip much better and seem to dampen vibration better.

Another thing to watch for is the fixed end bearing, I often have to press the seal further into the housing else it binds on the nut, but be carefull not to push it in too far else it rubs on the bearings. You will tell if it needs doing when you tighten the nut because it will be stiff if you try to spin it.

Cheers
Dean.

[Benonymous]

Hey Dean

Thanks for the tips with the couplings, I might take the Stig up on his offer to use some of the massive collection of industrial couplers he has in his bottom drawer at work. When I looked at the couplers today I thought they were a bit lightweight, especially when they'll be hooked up to my 400W servos.

Glad to hear there are no quality issues with Chai's screws. I know they're not top -line ground screws but they are easily a match for the Warners that I had from the plans machine.

Come to think of it, I'm no longer using anyof the hardware off the plans machine bar the X screw! So much for designing around scrounged parts!

I might also be able to salvage that Warner nut with some imperial ball bearings, the Stig strikes again! Maybe I can donate the ballscrews to someone else's project if I get it sorted out.

Stay tuned for the redesigned gantry.....

[hemsworthlad]

Maybe I can donate the ballscrews to someone else's project if I get it sorted out.

Get ready for it.!! . . Will be just like the seagul's in finding Nemo. . . ME ME ME ME ME ME ME ME ME ME ME ME ME ME ME ME. . LMAO.

Funny how all the best laid plans go Fubar when the tools come out.? Budget being the classic fairy tale. .:drowning: Ah well it will be worth it in the end.! so i'm told. . . . Just they never tell where the bloody end is.!! :wee:

[Benonymous]

[Benonymous]

Hi Zone.

There has been a delay in the build. Remember I needed to replace that Y axis screw? Well I ordered a replacement that was longer than the original mainly due to availability. So I get the screw and see almost 100mm extra travel on it and then look at the rails I had and I couldn't help thinking I was doing myself out of a bit more swept area and I'd kick myself later. An extra 100mm on the Y won't increase the dimensions of the machine greatly but imagine the recriminations if I want to do a job that's just a bit wider than my machine

So I went back to my old mates at FA System on eBay and ponied up for these ...



To add to the excitement, the trucks are SSR20's so they'll stay nice and clean with the scrapers clearing away the dust/swarf. The rails are 880mm to go with 730mm of travel on the screw. Should be enough room to space the trucks out on the Y axis to add rigidity.



So, on with the build! I have already designed them in to my SolidWorks model as I can always count on FA to supply accurate dimensions. The only problem now is that I think I may have less steel than I need due to the expanded dimensions.

Easy fixed.

[Benonymous]

Based on my new dimensions, I have cut the cross members for the bed of my router.



WooHoo! It's actually looking like a sort of structure!