A lot of discussion! I love this list haha. Well, I will try to answer this and focus on the main subjects we've been covering.
As you pointed out Les, an hydraulic rod is what I'm planning to use. Hard chromed and also lubed to avoid premature wear. I plan to add some sort of plate to collect the oil so it's not dripping all around the machine. They are costly, but they are made here in Argentina, as opose to the linear guides that we've been discussing. And since our currency suffered a severe devaluation in the last months, there are some components that are a lot cheaper to buy in here. About not using bronze, is this only because of the wear? Or do you think it will behave worse than HDPE or UHMW? The reduction I thought of originally was two step down timing pulleys and then the pinion connected to rack to give me the 10 to 1 ratio. I would like also to add an encoder directly coupled to the pinions for position feedback and may be compensate the backlash by software. I've been reading about using two pinions one fixed and the other spring loaded to cancel out the backlash but I think it's easier to cancel it in LinuxCNC. I've already done that on a grinder and works pretty well. Off course the rack on that grinder is far more precise than the one I'm going to use here, but this is for wood and melamine and I can allow some inaccuracies. Spindle, yes I thought about using 3 to 5 kw. I'm mostly going to use 1/4" to 1/2" diameter mills to cut through melamine. I'm still thinking about adding an ATC but I think that part will come later, because to start, there's not much work that needs a lot of tools. The ATC will have the tools resting on a side and that's one of the reasons the machine has 3.8 meters long in the work surface. Since the longer sheet we may use is about 3.6 meters long. The extra space is for the tool fixture. About the work holding subject. Yes, the idea is to use vacuum. I didn't calculate yet what amount of power we might need but It's good to have your expertise in here. The idea is to use what you mentioned, an MDF table wich is where the air is going to be suck from and take a skim cut to level the table. Here's a link of something similar from what I want to do: https://www.youtube.com/watch?v=fOrkvPF0pro It's a little difficult to answer one by one since I don't have quoting anymore but I thank all of you for your inputs and expertise! El sáb., 6 oct. 2018 a las 9:47, Les Newell (<les.new...@fastmail.co.uk>) escribió: > > No, I plan to support 50 mm bars every 600 mm more or less. I'm attaching > > some pictures of the design I'm working on > > If you really want to go with this I would second Chris' suggestion of > using HDPE or UHMW. You are still gonna wear those shafts out pretty > quick unless you use chromed shafts. Hydraulic rod is hard chromed and > accurate diameter. I don't know if you can get 50mm hydraulic rod in > those lengths but if you can it would still be pretty costly. The bushes > will be service items so make plenty of spares! > > > I also thought about > > reducing the 3000 max RPM with the worm and gear to 100 RPM on the shaft > > Don't use a worm or planetary gearbox. They have backlash (especially > worm boxes) and don't like repeated drive reversal. Harmonic drives work > well but don't have enough advantages to offset the cost in this > application. > You are looking for an overall 10:1 ratio so that can be done by two > belt reductions. Say 3.2:1 from the motor to the shaft then another > 3.2:1 from the shaft to the pinions. Cheap, simple and reliable. > > > I uploaded the pictures because the list doesn't allow me to attach them. > > Here's the link: https://imgur.com/a/7kLUWsq > > I'd like to see some triangulation in that frame. Try to make sure the > screw/rack/whatever is well below the top of the table. You want to be > able to easily slide sheets on and off the table without fear of > damaging anything. > > In my opinion screws are out unless you use high pitch screws, say at > least 25mm pitch, 30+ mm diameter. Even with rotating nuts there are > limits to how fast you can spin the nut before the balls start jamming > in their guides. > I have used endless belts with a fixed motor, as Roland suggested, on a > couple of machines. One was a 2.5m x 1.5m table machine and I had > problems with the long belts flapping around, especially when changing > direction. This was a plasma cutter so there wasn't much load. The other > machine was a 3.5m long feeder which only needed to push accurately in > one direction before retracting. I didn't have any problems with the > belt flapping but stretch under load was a significant issue. On the > plus side, that machine has been running every day for 10+ years with no > noticeable wear in the drive system. > Using a belt that is fixed at both ends and looped over the pulley is > more than twice as rigid as an endless belt and it won't suffer from the > flapping issues. For a machine your size I'd use two 30mm wide belts on > X with lots of tension. I really like the servo belt idea and am > thinking of using it on a machine I am planning on building. Keeping > dirt out could be an issue. > I have serviced a lot of commercial CNC routers and generally they > either use high pitch screws or rack. > > Todd suggested a minimum of 5KW for the spindle. It does depend a lot on > what you are doing. You can do a lot with a 1/2" cutter which will work > fine on a 3KW spindle. If you want to go bigger you will need more > power. Don't forget about extraction. You'll be generating a LOT of > dust. Making a good extraction hood isn't nearly as easy as you would > first think. If you need multiple tools for each job you might want to > allow for a tool changer. If you can't have a tool changer at least have > a tool setter. Tool setters are easy enough to make and save a lot of > time. Just make sure it is well protected so you can't damage it when > loading/unloading sheets. > > > If indeed its that critical, one would need a dynamic distance detection > > method of some sort riding the work pretty close to the tool... > > Whoa Gene, before you get too esoteric the solution is actually really > cheap and simple. Make the top of the table out of something that is > easy to machine, say MDF or plastic. After building and leveling the > machine skim the whole table using the router. The distance from Z to > the table will be pretty darn accurate over the whole area. As long as > the machine is built reasonably accurately the table will be pretty > close to flat. Remember this is a wood router, probably cutting large > sheets. If the machine has a slight bow the sheets will bend enough to > sit flat on the table, especially if it is a vacuum table. > > This brings up another point. How do you intend to hold the work down? > The simplest solution is to have a MDF bed and screw your work down but > that can get old fast if you are working with big sheets, especially if > you are cutting lots of different parts. You only need one miss placed > screw to wreck a cutter and probably the job as well. > Vacuum works well. There are two options. The first is a pod system. You > have a number of vacuum pods like these <http://vacuumpods.com/>. They > use a high vacuum and hold well but are best suited to production work > on pre-cut blanks where you can set up and run hundreds or thousands of > parts. A matrix bed pulls a vacuum through the whole bed. They rely on > high flow and a lower vacuum. They are best for sheet work, especially > if you run lots of different parts. You need to make sure unused areas > of the bed are sealed off. > > For that size machine you are looking at a vac pump rated at 10kw at the > very minimum for a matrix bed, maybe 5-7kw for pods. My 8'x4' Rye router > with a matrix bed has 2x 4kw vane pumps and they are barely enough. > Another 8'x4' machine I use occasionally has 2x 3kw side channel blowers > and they are also only just enough. I have been investigating using > multiple vacuum cleaner motors (say 5 or 6 of them). They are relatively > efficient and should generate enough vacuum for a matrix bed. They are > much cheaper than proper vacuum pumps though in this application they > would need to be replaced at fairly regular intervals. > > Les > > > _______________________________________________ > Emc-users mailing list > Emc-users@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/emc-users > _______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
