Ep02: Plasma CNC build | Z axis "fail" :(

[@WillemvanLonden]

Another epic journey. I'll be glad to tag along.

[@routercnc9517]

Hopefully not too many episodes to get this thing built!

[@yngndrw.]

In my opinion, all CNC machine designs should start with the tool. That way, you know the weight that the next axis needs to support/move, the forces, and the sizes. I have some suggestions: With the amount of work you've put into this, you should consider a machine torch instead of a handheld one - That would resolve the leverage issue and make it far more compact. From your response to other comments it sounds like your new design would support this as the middle spring has gone. You could print the ballscrew bearing mount in two parts without support. (Slice between the two bearings.) If you print on smooth glass, you can then superglue the two halves together and the axial load between the bearings will keep them together. Always try to design without the use of supports if at all possible, it makes life much easier. You could possibly use the breakaway sensors as the touch-off sensors and do away with a whole section. The ohmic sensing can then be used for thinner materials with the touch-off system acting as a backup if the ohmic sensing fails. The Nema 23 stepper motor is far larger than you need, I'd personally use a much smaller motor - Nema 17 or maybe even Nema 14. You can possibly get away with a single (wide) linear rail, that's what I'm planning to use in my plasma build.

[@routercnc9517]

Hi. I actually started with cutting area, then the maximum overall outer dimensions because it needs to fit under the worktop in the one remaining space in my shop. But then I did start with the cutting torch and head to get that about the right distance off the bed to maybe do tube cutting in the future, but still have enough travel down to the bed. Then I did the rest of the machine design from there. I did look at the machine torches rather than hand but I can't fit one in due to the height under the worktop. Unless it was quickly demountable since I would pull the machine out to use it . . . Good idea on printing in 2 parts and gluing. For now I used the PLA support material for the bearing mount and I'm OK with it, but one to think about as I don't like supports! I think the breakaway is quite stiff so most materials would deflect a bit. I'm only planning on cutting thin sheets to fold into other shapes. The latest design splits all the functions so the up and down is on the rails, the touch off is very light on another mechanism and the breakaway it quite stiff to hold the torch. Let's see how it works out, it will be in the next video. Agree on the stepper size for Z. I did look at the Langmuir steppers and these were Nema23 and about 2Nm so I went with a set of those in the end. I think smaller on Z would also work but I got a good deal on a set of 3 so went for all the same size as the X axis motor is probably needed at Nema23 2Nm as it moves everything else. I did also start with a single wide rail on Z but I already has 15mm rails from an old project so went with 2 of those. Lots of options there I think. Good luck with your build. If you have a channel I'd be interested in a link to see how you build yours. If not don't worry. Thanks for the comments, very helpful.

[@yngndrw.]

@@routercnc9517 I'm not sure if you can get a similar torch for the Bestarc, but HyperTherm does a shorter range of robotic torches. The internals are quite small so I wonder if you could cut down the internals of a full-sized machine torch? I suppose you do need some space above for the cable either way, to take advantage of a machine torch. My design was going to look very much like the JD plasma cutter, but I then completely changed it after watching Travis Mitchell's fibre laser (V2) videos. It's now a fully enclosed design, not especially space efficient in terms of volume to the work area, but it should hopefully keep my workshop a lot cleaner. I don't have any videos, but I'll try to include an OnShape link to the design after this comment - I'm not sure if KZbin will block it though. The OnShape model is really just a sketch, as I wasn't happy with the torch mount and I'll probably re-do it all to suit the final design. It does however show the general idea which is a welded box-frame with 20-40 extrusion attached for mounting the rails to reduce the threading requirements. The design uses 12mm supported round rails as I think they will be more robust with the plasma dust and they will bolt straight to the extrusion. I've also designed around a belt drive for the X and Y axes. The water tray is an off-the-shelf Amazon 30’"x28"x2.8" stainless tray as I can't easily make one myself and I'll sit it on top of some HDPE sheet to electrically isolate it from the frame. The drag chain and motors will all be internal, allowing for the outside to be easily sealed up. That's the plan anyway, the current torch mount is designed around an off-the-shelf Chinese linear actuator but doesn't have breakaway support. (Just noticed that the drain pan is no-longer available, I should have bought one in advance - I'll need to find another and re-create the design around that now!)

[@routercnc9517]

@@yngndrw. OK. I think I'll carry on with the manual torch and get it made, then re-assess. Do you need to electrically isolate the try from the machine? What is the purpose of that? I was going to custom make an extractor system under the machine from sheet metal, with fan for the smoke and bucket for the sparks. Also, I have been following Travis' laser build, impressive but not cheap!

[@yngndrw.]

@@routercnc9517 The idea of the isolation was to keep the torch current as localised as possible. A lot of people seem to have issues with interference, so my plan was to just bypass the issue by minimising the impacted area. I don't know if it will be needed, but it wasn't too much work to design for a water tray as I just need a sheet under it and some plastic clamps to stop the tray from sliding. It would be a lot more work with an extractor system, but it is still doable if you plan ahead. I'd be very interested to see how well your extractor system works, I've been concerned about getting enough volume but if it works then it's a lot more convenient than having to maintain water.

[@routercnc9517]

@@yngndrw. Ok thanks, makes sense. My machine will be right next to the workshop door so hopefully the fan under the extractor funnel will blow most of it out. Let's see how it goes. Like you I'm trying not to use water, also my use will be intermittent.

[@MickGalt]

nice work. have a look at how a touch probe in a machine works. you could have the Z touch detection and hit detection in one. it would also remove the need for the 4 proxies as it relies on breaking a single circuit when the mechanism rocks off the contacts.

[@routercnc9517]

I've looked at quite a few hit detection probe designs (there are some commercial ones available too) but as far as I can tell there is quite a bit of force required to detect the hit because otherwise the weight or movement of the torch could trigger it. I think this would then deflect thin materials, so drives a very soft mechanism, but then that could get false triggers. It does that on my initial design as the torch is hand held and the weight is off to one side. Maybe the machine torches (vertical) are better for that. I've now started again and designed and made what is hopefully a much better design and uses 2 proxies, one for the collision and one for the touch off. Let's see how it works when I test it!

[@sonsrc1326]

Nice shame u don’t like it but all trial and error looking forward to next

[@routercnc9517]

Yes I want to get this bit right so I don't have to come back to it, then the rest of the build should be fairly straightforward.

[@TheRainHarvester]

Can you counterbalance the torch's leverage? With weight on the other side? Or a spring?

[@routercnc9517]

Counterweight could help, but I still think I might get false triggering as the spring is in the middle so does not offer much control. Adding more springs could work, and in fact the redesign now has 6 springs radially rather than one in the middle.

[@kazolar]

You could get away without supports rotating some of the parts.

[@routercnc9517]

Yes, thinking about it that stepper mount could have been printed 90 degrees to that. Well the new design improves it further and avoids that problem all together . . .