C-Beam Machine feedback

Hi Crew:

Just though I would drop you guys a line to point out some, well, flaws shall we say in the C-Beam machine design.

Don't get me wrong, its a great machine, but I have found some issues that were mostly due to mistakes on my part, but certainly are avoidable with a bit of engineering.

Firstly, the Z-Axis on my machine started acting funny, so I took a close look at the insides. I discovered that one of the mini v-wheels was cracked. Thinking this was odd, I looked a bit closer and found that all of the internal wheels had marks around the perimeter just inside the v-portion of the wheel on the flat surface, but only on the inside (side closest to the lead screw) of the wheel. After scratching my head as to how they got there, I realized that the wheels were hitting the lock collar at the extremes of travel, and the sharp corner of the lock collar was marking up the perimeter of the wheel. Thinking a bit more, I realized that this was probably the cause of the cracked wheel as well.

Now, before I go any further, yes, you are right, this was my own fault for running the axis right to it's end stops, but in my own defense this would hardly be an uncommon occurrence with a new owner testing out his new machine and software.

Also, on the X axis, I noted the same problem with the outside v-wheels having marks where they hit the triangular brace that attaches the c-beam to the uprights. This one is a bit more serious in that it marks the angled side of the wheel so would make it roll with a bump every rotation. Again, totally my fault for running the axis to the end stops, but still should be avoidable with a bump stop of some sort on the ends of the axis.

Yes, I know I should have limit switches but I just haven't got around to installing them yet. Anyway, a belt and suspenders approach would be prudent, limit switches and soft bump stops for all axes.

So, my suggestion would simply be to put some rubber or neoprene collars on the lead screw inside the lock collar. This would give a soft bump stop for the wheels, so as to avoid damaging them in the likely event of an axis crash. This only really applies to c-beam with inside mini v wheels. Perhaps some kind of bump stop for the outside wheels as well. I was thinking a small bracket that goes into the holes on the top (open side) of the C-Beam end caps. Maybe even just a screw with a large grommet on it that the gantry plate hits first. I know I will be doing this when I repair the machine, and for all future C-Beam builds.

One other issue I found. My z-axis seemed to have a lot of slop in one end of the lead screws at the bearing. The inner diameter of the bearing calipered out to 8.00 mm, as it should. However, the diameter of the lead screw was nominally 7.85 mm and even 7.80 in a couple of places. This is a lot of clearance, almost .008", and may result in the lead screw just turning inside the bearing race and the bearing doing nothing. The lead screw certainly seemed to have a lot of side to side play in it. Not sure this would affect the accuracy very much, but it may make the system noisy and certainly will not do it much good.

In the assembly video, Mark notes that the lead screw may be a bit large for the bearing, and advocates chucking it in a drill and using some sandpaper to take down the diameter for the last inch or so to get it to fit in the bearing. This is the opposite problem, and does not bode well for the consistency or quality of the lead screw supplied.

And, lastly, I have some kind of weird sound on the Y axis at it's rear most position. There is a loud clicking sound for about the last 30mm or so of the axis travel. I have looked inside the axis as best I can and can see no reason for this sound. It does not do it on the other extreme of table travel. This is not due to the motor slipping, as far as I can see. I am going to have to take the axis apart and see if I can figure out what is causing this sound.

Anyway, just my .25 worth, I call 'em as I see 'em. Off to the parts store to buy some spares.

MetalGuru

 

As far as the lead screw goes, I checked the diameter with a digital caliper along the whole length of the screw and it was 7.80 to 7.85mm diameter for the whole length, not just the ends. Definitely not wear because my machine is brand new, has only been operating for a week or two, and just small test cuts in wood. I am going to check diameters of the lead screws in the other two actuators and let you know what I find.

As far as trueness, when I assembled the actuators, I always leave the end cap mounting screws a bit loose, and run the carriage right to the end of its travel. Then I tighten the end cap screws on that end. This should ensure that the lead screw is positioned on the same center as the lead screw nut, and any play in the end cap position is reasonably centered when the screws are tightened. I then run the carriage to its other extreme and tighten the screws on that end.

Locktite is not going to fill a gap that large, especially with the threads on the lead screw. I am considering putting locktite on the outside of the bearing to lock it into the end cap of the C-Beam axis, however. I am also looking into using the KFL08 pillow blocks on the lead screw. They would seem to be quite a bit more robust, and have the locking collar built in with 2 setscrews. See picture below:


They might be just a hair too wide at 48mm to fit inside the c-beam rails, but it could either be mounted at a slight angle, or a recess machined into the end block to fit the diamond shaped flange. These also have a spherical bearing which self aligns to any minor shaft misalignment. And, they are only $2-3 each.

 

Gray:

Yeah, I think half inch would be much more suitable for the 1000mm actuators. I've been thinking about a device that would slide along the inside track and support the lead screw halfway between the carriage and the end plate. Perhaps like a nylon sleeve with some kind of flange that would slide inside the extrusion rail. There would have to be one on each side, and they would have to move with the carriage, but not interfere with the movement. Not an easy problem, still thinking about it.

My solution for bump stops ended up being easy and cheap. I just took a 10mm screw and tee nut, and put a small 5 mm id x 10mm od grommet on the screw. I then installed one of these at each end of the track. Then I slipped a larger 10mm ID x 15mm od grommet over the top of that. Works like a charm, soft bump stop for $0.30. see photos below. If it saves one v-wheel that's $5 and worth it.

 

Rick:

Yeah, those ideas would work, but I was thinking of something much smaller and riding along the inside of the c-beam, against the back of the track.

I'll see if I can come up with a drawing. I like the solid joiner piece in your second drawing, anything like the string idea might get jammed up in the wheel on an inside-the-track system.

 

Another bump stop idea.

 

Hi guys:

So, some more feedback on C-Beam construction. Some notes I jotted down as I repaired my broken wheel on my C-Beam Machine.

1. Motor Mounts - As suggested by Ronald van Arkel | OpenBuilds, I replaced the motor standoffs with shorter, heavier ones. (Mcmaster Carr 94669A195) 5mm bore x 13mm diam. x 30mm long unthreaded aluminum spacers. These have much stronger support for the motor but do necessitate either cutting off the leadscrew, or as I did, just sliding a bit out the other end of the C-Beam.



Also, Ronald pointed out a potential problem, the wide head screws used to mount the motor actually put a great deal of stress on the motor mounting hole pockets because they are larger than the pocket. If you put one in the motor and push down, you will notice that the screw does not come out straight, but at an angle. Over tightening these screws could break the motor mount tab off the motor. Simple solution is to grind down the diameter of the head a bit, and put a slight radius on the bottom of the head to better fit the motor pocket. Grind enough off until the screw sits straight in the hole when pushed all the way down. Or use standard socket head or round head screws instead of the low profile socket heads.

Another note on the flex coupling - Always tighten the collar screws fully BEFORE you tighten the setscrews. If you do it the other way around, the setscrews will prevent the collar from fully tightening, and you may end up with slippage. Again, insert a playing card between the flex coupling and the c-beam end plate to make sure you have clearance before tightening any screws.

2. Locking Collars - When I removed the gantry to change the wheel, I noticed it was difficult to get off the last bit of the leadscrew. After I had removed the carriage, I looked at the leadscrew and discovered it had been quite damaged by the setscrew on the locking collar. This is going to be an issue with all C-Beam builds, if you ever have to take one apart, which you inevitably will. Make sure to check and repair the threads a bit with a jewellers file before removing the leadscrew, or at least loosen the setscrews on the flexible coupling and remove the leadscrew with the carriage as one piece. Otherwise, you risk damaging the threads in the nut block, which I am now sure that I did with mine. I would suggest future kits come with a one or two screw split locking collar like these:


These are made by Ruland (www.ruland.com), and are p/n MSP-8-F and MCL-8-F respectively. I'm sure tons of other places make them as well. McMaster had some but they were too big on the OD.

3. Preload on the leadscrew - In the assembly video, Marks says he leaves the end cap on the c-beam "a bit loose" when he installs the locking collar to give the leadscrew a bit of preload when he tightens it up. "A bit loose" is not very specific, and if you leave too much slack on the end plate screws you can bend the leadscrew when tightening it up. I found that inserting a playing card between the end cap and the c-beam, and tightening the top two end caps screws before tightening the locking collar setscrew will give just the right amount of preload. Tighten the locking collar while taking up any slack in the leadscrew by pushing back on the carriage and forward on the locking collar as you tighten the screw. Then, loosen the two screws on the end cap. remove the playing card, and reinstall and tighten all 4 screws. You will find that there is no play in the leadscrew, and everything is tight.

I actually cut slots in the playing card where the two screws and the leadscrew go into the end cap, about halfway down the card. It then slips down a bit and holds the end cap a fixed distance from the c-beam and keeps it parallel to the c-beam end. I always keep an old deck of cards laying around the shop. They make excellent, consistent shim stock.

4. End cap lineup - When I tighten the end caps for the final time, I make sure that the carriage is moved all the way to that end of the c-beam. This ensures that the shaft is aligned as much as possible with the center of the nut block. Then, run the carriage to the other end of the c-beam, and loosen and re-tighten the end cap screws to align that end.

That's all I can think of for now. I'm sure I will find more things to ***** about ಠ⌣ಠ as I learn.

 

Teflon:

I'd arrived at about .25mm by playing around with various spacer thicknesses. Great minds think alike. Your way would work just as well. My cards are exactly 0.3mm, YMMV. I tried 1mm first by removing the spacer washer between the bearing and the locking collar. That was way too much, the leadscrew was bowed several mm. Just went down from there. I just like to have a repeatable, measurable way to do things, "a bit loose" was way too nebulous for me.

I'm a bit worried that the bearings may wear prematurely because they are not really designed for thrust loading. Some of the guys (Ronald van Arkel | OpenBuilds) were talking about using needle thrust bearings. In the interests of KISS, I think you could just use an angular contact bearing which handles both radial and thrust loading, something like this:



Not sure if you can get a bearing that small in an angular contact version. But. for a $1 apiece or whatever those bearings cost, it's cheep to replace them if they wear out.

Anyone know the part number of the bearings used in the C-Beam?

 

Holy Crap! $80 each? I can replace a whole buttload of the 688's for that price. The F8's look kind of cool, but in this application I would think they would get filled with little chips of wood and aluminum pretty quickly...