February 12, 2016 at 4:40 pm #7467
Okay I just can’t come up with an easy way for me not to have to build 3 machines for the 3 size conduit that is out there. Adding hardware and cost seems like a bad trade off. So I ask for your help, the people that get to use there machines more than I do!
So here is my thinking.
1) The original was designed to be 3d printed, and possibly injection molded That is why I tried to keep the unique number of parts down, plan was to even use the middle joiners x’s 4 instead of the middle ends. I also thought, foolishly, that all conduit was the same size worldwide.
Find another rail, unique part count doesn’t matter if we are printing it anyway.
2) As it is now it is dirt cheap and works better than I had ever hoped. People can continue to use this without feeling like they are missing out if they don’t want to spend a bit more for a slightly better version. Conduit is nice but we can do better.
Second version can cost bit more without pissing anyone off.
3)More rigidity = faster and more accurate. We are carrying a much larger tool than I had originally designed this for and who doesn’t like faster?
Slightly larger diameter tube/rail will benefit the most here, as well as some design tweaks that I learned from the first.
So I am thinking Stainless Steel rails, and a redesign. Nothing drastic same bearings, steppers, controller, same basic layout. Good?
Stainless has better wear properties, more rigidity, but what size? At least 25mm, what size is available in your corner of the world? I can probably find a good source that can ship a kit worldwide and perhaps even ship custom kits. But I am not trying to spend hundreds on rail when the previous stuff works great for $7.
4) Name, Need a name for this. It is a cnc platform, a multi function machine. Stick with the original “Melvin”? Not fond of V2 or version 2 or beta.February 12, 2016 at 6:32 pm #7469
love the thought of another project, been having a lot of fun with the current version. Living in Australia had a issue with the 25mm tube and have been using aluminium 6061 grade tube with little ware but still yet to give it a hard time.
priced stainless and was way to high to even consider, how would aluminium extrusion running on 608 bearing and nylon rollers go?? i know price would be high but don’t know as high as stainless, might also be able to get larger spans on square/rectangle extrusion
looking forward to hearing more responsesFebruary 13, 2016 at 8:25 pm #7505
I was thinking alu extrusion too. maybe even a gear lock design to help stabilize. you still use the conduit for the middle z, roller setup. but run it all on 2040 or 2060 extrusion with a tooth gear roller instead of the belt driven setup. maybe smaller diameter pipe for the z drive to shrink down the z package and give a little more build volume.February 13, 2016 at 8:30 pm #7506
Tooth driven how? Sounds slow and likely to skip steps. I’d go leadscrew or ballscrew but they are both expensive and slower than belts. Most 3d printers use belts for the x and y axis for speed.February 14, 2016 at 6:02 am #7517
my thought is if he goes to aluminum extruded rails and wheels sliding along those rails then he is just like 105462 other make your own CNC machines out there, for the same price as they are.
i like the design of this. the locally sourced parts, the option to print my own.
I doubt i would have made this if i just bought a kit with parts, etc for $500 and built it.
maybe thats just me 🙂February 14, 2016 at 2:20 pm #7559
Square extrusion takes too much hardware, you need to touch all 4 sides instead of just 3 with tubes or hex.
Gears, leadscrews, ect all have backlash. Software compensation is crappy at best.
I really like belts. I think I will be belt driving the z as well with a multiple pulley system? Probably not but I would like too and will be trying it out.February 14, 2016 at 2:25 pm #7560
Yup, I agree. I just want to make some improvements and don’t want to maintain multiple versions. It really kills motivation when you know you will have to do some thing a few times.
What else can be easily sourced though? Just larger conduit? The stuff is nice for milling but I am seeing it’s weaknesses with the laser etching that is going on. Closet rod is cool and fairly inexpensive but the sizes are all over the place. That is how I keep ending up at stainless. The nice hand rail, boat rail, stuff is thin walled and in a few standard sizes.February 14, 2016 at 3:10 pm #7567
would it be worth asking all corners to find a common tube/pipe/rod size, us trouble makers down in Australia can get 25.4mm/1″ off the shelf.
agree with the other comments re staying unique re staying away from the common alu extrusion, after my original post re alu did some research and yeah defeats the purpose of this awesome project. maybe larger diameter tube for lager spanning options.
1 question what are we really trying to fix apart from maintaining 2 designsFebruary 14, 2016 at 3:35 pm #7569
A few things, I hope.
More rigid. I designed this for the small lightweight rotary flex shaft grinder, turns out the dewalt and a few others are a better choice. I have 2 new center assemblies not sure which one is better yet I printed the prototypes with light infill so they both flex. Need to reprint and give it another test.
Ideally multiple size tubes- but doesn’t really seem feasible with this geometry. So a single size is the tradeoff.
Less parts or faster printing? I hope
Smaller foot print, maybe, a little less dead space.
Less unique sized hardware, for easier assembly.
Better rails for us heavy users. The conduit from my first machine has been rotated several times. And I think the smoothness of the rails is having an effect on the laser etching resolution. And the stuff that comes off the rails as they wear is making a nasty mess of my machines.
I could keep going. Really this is just a better version for those that want it. I am leaving the other ones up I think they work great.
I guess I can just improve the machines we have, right? People can buy 25mm stainless and use that one and I could just redesign some of the parts I want to change…..
Maybe I should do that. Just make some changes. Cool! Nevermind, maybe I should take this thread down.February 15, 2016 at 9:20 am #7617
Have you found any good sources for the IE version Stainless steel sourced from the US?February 16, 2016 at 11:18 am #7680
Well I found some but I don’t want to pay that much. 25.4mm (1″) is a little less than a 1/3rd the price. Grrrrrr, I don’t want to make a 3rd version.February 16, 2016 at 1:14 pm #7685
what countries can not source 25.4mm cheaper than 25mmFebruary 16, 2016 at 3:33 pm #7695
Drop the 25, for a 25.4….?February 16, 2016 at 4:26 pm #7700
I’d vote for 25.4 – Australia
can you see any benefits in going larger again???February 16, 2016 at 5:22 pm #7704
I really have not seen many large builds. As far as I know I get asked once a week about it but no one has actually built one. I am pretty happy with a smaller build. I can see some people wanting to cut full sheets 4’x8′, but really that is only a few people not the majority. With some mid spans support I still think it could cut full sheets as is of wood.
So larger rails is just a larger footprint for us people who want smaller machines. Best of both worlds?February 16, 2016 at 6:12 pm #7712
I’d like to go all stainless as well. but I have to find a good cheap place to get it. You should be able to find 3/4emt in stainless at an electrical supply house, but I am unsure of the actual dimensions (larger, smaller?). maybe instead of making vast improvements just up the size of the motors…February 16, 2016 at 6:26 pm #7717
I’m happy with the power we have. If you need more you can just turn up your drivers. We are under powering our steppers!
I want more rigidity and precision. Steppers won’t change this.February 16, 2016 at 8:15 pm #7726
You’re not going to get much more precision and rigidity than where you’re at right now. unless you want to go to hardened stainless rods. At 25mm and 30in lengths, well i don’t know what that would cost. if you go any smaller on the emt size you will lose rigidity. if you go any larger than 1″ maybe 1 1/4″ You’ll start to lose rigitity as well. and cost effectiveness. keep focusing on making your awesome designs better.February 17, 2016 at 6:50 am #7748
In Italy 25.4 is hard to find…February 17, 2016 at 10:24 am #7754
What sizes do you have available? In conduit and stainless?February 17, 2016 at 11:21 am #7757
20 and 25 mm can be found quite easily. The “inches” fractions are very difficult to find in Italy and, I think, I the major part of the EU. Except of the uk, maybe…February 17, 2016 at 9:20 pm #7809
I think overall your most common stock/rail/rod/conduit size worldwide is going to be 1″ or 25.4mm. Second is going to be 25mm. You could easily account for this by using some sort of eccentric spacer with the bearing holes. This would also be good for adjusting pre-load/tension on the rails. I mean you can buy eccentric spacers, get them made in bulk for cheap from china, make them yourself…drill a piece of round stock off center or and even print them. If you aren’t sure what I am talking about, it’s basically a nut where the hole is off center, that way when you tight it it moves the bearing/wheel closer or further since it’s off center.February 18, 2016 at 8:34 am #7822
The problem is conduit is measured by its ID, so the OD varies that’s how we end up with 23.5mmOD. Unless we find something priced within reason with similar rigidity/mass measured as tube (OD) instead of pipe (ID). Or I pay for a large quantities of hex extrusion (deluxe version, but would need to ship rail) I think its best to stick with what we have. I do want to give 25mm Carbon fiber or fiberglass tubes a try, super common and cheaper than stainless, but I doubt it can handle the point loads without delaminating quickly.
Updates on the current parts, then maybe a deluxe version down the road seems to be the best coarse for now.
Eccentric is actually very expensive and they all vary in size and shape, it would be hard to spec a universally available one. I though about a printed insert to change one of the three bearings position but the problems happen on the combo axis with the Z. It can be done but I think we would be sacrificing rigidity to make it happen, If I just suck it up and make my models more friendly to edit I should be able to maintain 3 builds for the best rigidity. Then to handle the tool holder fragmentation issue, people only making holders for there sized conduit. We can just switch to a universal mount, which I have been worried about because another point of attachment reduces rigidity. The users here have already came up with a nice one so I will leave that alone until the design changes significantly.February 19, 2016 at 3:54 am #7858
+1 hereFebruary 19, 2016 at 1:07 pm #7887
I am near to the end of the build of my MPCNC.
I have finished the basic frame, legs and gantry.
But I still have to print one rigid joiner and that will complete the middle assembly.
Then I will measure & print the Z motor support and tool holders.
I decided, early on, to go for stainless steel tubes despite the additional expense.
(I couldn’t find the original galvanised steel conduit in the UK where I live and because I reckoned stainless would last longer and be resistant to rust).
But I chose 1 inch diam tubes, and only discovered, after delivery, that Ryan’s design would not accommodate that diameter!
So, having committed myself, I realised I would have to modify the design to cater for 1 inch tubes.
The main change (apart from inserting triads of bearings circling a one-inch diam tube section) was to split most of the monolithic parts, like middle_ends, into two parts.
The idea was to insert/remove shims as required to adjust the ‘grip’ of the bearings on the tubes.
My redesign removed a small amount of plastic from each mating face so as to allow for closing down the size if needed.
Also, I planned to make the single bearing bolt hole into a slight slot, so I could shim that bearing in/out.
So, I could move the two Y bearings closer or further away form the tube by flat shims between the new joints in my split middle_ends, and slide the single bearing in/out by putting curved shims around its bolt.
Did it work?
Yes. When first assembled it was a loose fit. I just shimmed until the fit was correct for each part of the assembly.
It took most of a day but was straightforward.
Thanks very much, Ryan, for a really original design and thanks for being so generous by making it open-source.
I am really enjoying the build and the technical hurdles I introduced and am looking forward to adding the electronics very soon.
Thanks for the cable kit and the drag knife; it all arrived safely. (postage was a killer though!)
I have attached a photo of my middle_assembly with dark lines in places to try to show the joins I have mentioned.
Attachments:February 19, 2016 at 7:27 pm #7902
Dang, that’s a lot of work, nice job!
Have you checked to make sure your z axis is 90 degrees from the bed in all directions? That center section is a beast to get right.February 20, 2016 at 7:16 am #7920
I admit I hadn’t thought about vertical alignment!
I will check that out when I have completed the last part (rigid joiner) of the mid assy and assembled it.
Have just bought a dial indicator which can show dimension differences down to 0.01mm. That should reveal the truth about alignment!
If necessary, I plan to ream out the appropriate bolt holes in the rigid supports and shim until the twin tubes are vertical in both planes.
I should find out in the next couple of days.February 21, 2016 at 4:52 am #8022
The fundamental issue:
Metric or Imperial
then we need to add tolerance
The original design (V1?) already mixed it up with the bolds and nuts, M8 and 5_16, M3 / M4 vs 6_32. Being from Europe in a big city i can source the Metric parts easily (just round the corner in some hardware stores) Imperial becomes a bit more expensive, and time consuming option (i need to order them).
Same as for the Conducts/Tubes/Pipes
Coper, Aluminium or Steel, Metric or Imperial, i can order impirial, but if i go to the local hardware store….. i’m stuck with 22mm (dirt cheap) and if i get 25mm it might be close to 25.4 depends on the store/the batch and tolerance.
Example : i printed the International version of the Roller, to find out that the rollers (bearings) and the 25 mm Aluminium pipes don’t fit properly, there is around <1mm space.
So for V2 i would say that there is a need for some tolerances in the (great) design.
ps: i’m still in printing phase…..February 21, 2016 at 7:08 am #8026
I was afraid of that. To build in some adjust-ability to the z axis I will probably have to add hardware to the z axis.
Still in the planning phases again. I will update the current design first. The next one will probably involve new rail material.February 21, 2016 at 9:50 am #8032
While waiting for the next part to finish…. i designed a kind of eccentric spacers (mentioned also by Atomist)
Not having a working CNC yet (still printing and building) the part seems to work to adjust the single bearing side of the roller
No clue if it would hold and can work with the forces of a CNC machine, i can’t try yet 🙂
It converts from M8 to M5 to bold the bearing, but you can tune it so the bearing nicely gets to the tube
(update:) tried to upload the .scad files… didn’t work…
// fix the MPCNC issue regarding the roller and the Pipe/Tube
// print 2 of those and the other part
translate([0,0,3.75])cylinder(h=3.5,r=3.9, center=true, $fn=360);
translate([0,0,1]) cylinder(h=2,r=6, center=true, $fn=360);
cylinder(h=20,r=2.6, center=true, $fn=360);
The M8 replacement
// MPCNC 608ZZ Bearing adjuster
// bearings specs : M8 , 7MM high
/// M8 to M5
translate([0,0,10.5]) cylinder(h=21,r=4.0, center=true, $fn=360);
translate([1.5,0,2.5]) cylinder(h=5,r=6.3, center=true, $fn=6);
translate([1.39,0,0])cylinder(h=50,r=2.6, center=true, $fn=360);
You must be logged in to reply to this topic.