Back
[02:09:28] <DJ9DJ> moin
[11:06:07] <Chemeleon> anyone know of a better option than round supported rails/linear bearings, that wouldnt cost significantly more?
[11:07:04] <archivist> depends what you are making
[11:07:30] * JT-Shop has everything put away and is going to take a nap now
[11:07:40] <Chemeleon> very rigid router, 12x12 moving table design, goal being to cut as absolutely fast as possible
[11:07:54] <Chemeleon> wood primarily, but I figure its overbuilt enough metal should certainly be doable
[11:08:09] <archivist> 12 mm,inches,feet,metres
[11:08:14] <Chemeleon> http://www.chemeleon.net/diycamCNC_wip2.jpg still working on the design, but here it is so far
[11:08:16] <Chemeleon> ah, sorry, inches :)
[11:08:29] <Chemeleon> I want something desktop sized and super fast for testing code as I write it
[11:08:58] <Chemeleon> and plan to document it and release the plans for free, so trying to stick to sources that are easialy duplicated, ie not just waiting for a great deal on rails on ebay
[11:09:11] <archivist> do you really care then
[11:09:20] <Chemeleon> ?
[11:10:12] <archivist> machine for testing code does not need to be as good as a machine that cuts solids like titanium
[11:10:21] <andypugh> Bigger supported rail?
[11:10:23] <syyl_ws> the cad rendering looks almost as elegant as an brick ;)
[11:10:41] <Chemeleon> syyl, yea, its a lot of very chunky 8020 :)
[11:11:04] <Chemeleon> going for completely over the top overbuilt, as currently a lot of my cnc work is 3d carvings, and I want to try and push the speed way up on them
[11:11:05] <archivist> anyway just use the simulator to test code
[11:11:20] <Chemeleon> test code as in code from the programs I'm writing
[11:11:29] <Chemeleon> not just gcode from commercial cam apps
[11:11:43] <Chemeleon> so want physical results to confirm theres nothing slipping past :)
[11:11:50] <ssi> Chemeleon: if you don't want to spend for profile rail, the cncrouterparts.com rail system is a reasonable low-cost option
[11:12:01] <andypugh> I have a feeling that a moving gantry might be better than a moving table.
[11:12:01] <Chemeleon> thanks ssi, I'll look into that
[11:12:10] <syyl_ws> i double that
[11:12:22] <Chemeleon> whys that andy? all logic I can come up with puts moving table as significantly stiffer
[11:12:29] <Chemeleon> and at this scale, the extra size isnt a big deal
[11:12:33] <syyl_ws> the gloryfied cheesegrader i work on has also a moving gantry
[11:12:52] <andypugh> Though it is only a 1D moving table, and a moving gantry does have the maximum level of stacked axes.
[11:12:54] <syyl_ws> but runs 30m/min with brutal acceleration
[11:12:57] <ssi> for that size, I'd stick to the moving table
[11:13:12] <ssi> I built a little desktop machine with a 12x12 table, and did moving gantry, and it wasn't nearly solid enough
[11:13:22] <Chemeleon> I have a mid sized moving gantry out in the shop I built with mdf to save money on my first cnc, works fine, but can only do 3d cuts around 50ipm before vibration from direction change starts to appear in the wood
[11:13:31] <ssi> you need such a wide stance for the gantry bearings for it to be stable that you use up at least as much space as the moving table anyway
[11:13:41] <Chemeleon> yea, my thinking exactly ssi
[11:13:42] <archivist> I have found small size has been a hindrance often
[11:14:14] <Chemeleon> I'm hoping with this being super stiff, I could potentially hit 300ish ipm for 3d carves with a 1/8" bit
[11:14:16] <andypugh> I think I changed my mind.
[11:14:18] <andypugh> :-)
[11:14:28] <syyl_ws> good highspeed 3d milling is more a matter of good gcode anyway
[11:14:47] <Chemeleon> which would be a huge time saver, even beyond not having to run from my office to the shop out back :)
[11:15:04] <syyl_ws> bad code lets even a well built hsc-machine look bad
[11:15:17] <andypugh> At 12x12 it might not be inconceivable to use iron castings and slidways.
[11:15:34] <syyl_ws> with crapy surfaces, overshots in the corners and so on..
[11:15:42] <syyl_ws> i learned that the hard way D
[11:15:45] <kwallace3> My thinking on the the rail connector plates is that in relation to stiffness the plates are like hinges.
[11:15:51] <Chemeleon> I thought I'd found profiled rails from a chinese seller cheap, but looking at his quote again, seems he just quoted individual prices for his "total" rather than the quantities I'd need
[11:16:17] <ssi> yeah profile rail isn't cheap, and never will be I don't think :/
[11:16:32] <ssi> my plasma table languished a year because I didn't want to pony up $500 for rails
[11:16:49] <ssi> for the gantry axis, I needed a matched pair of fairly big rails
[11:16:50] <Chemeleon> part of the limit, is I'm trying to keep this under 1k too :)
[11:16:59] <ssi> ended up with a pair of 60" 30mm rails
[11:17:06] <Chemeleon> if I do it with round rails, looks like thats feasible so far
[11:17:10] <ssi> ebay, from korea
[11:17:15] <kwallace3> There should be a plate on each side of the rail.
[11:17:21] <syyl_ws> 1k, rigid and fast in one machine makes me feel a bit fuzzy :D
[11:17:26] <andypugh> Round rails + air bearings might be fun.
[11:17:43] <Chemeleon> syyl, yea, I've not really seen plans that fit all those goals before :)
[11:17:53] <Chemeleon> though I'm sure its been done
[11:18:00] <ssi> shit, it costs $1k to build a 3d printer in that size
[11:18:03] <ssi> and they're the opposite of rigid
[11:18:08] <Chemeleon> heh, yea
[11:18:15] <andypugh> "flaccid" ?
[11:18:19] <ssi> "not rigid"
[11:18:20] <ssi> :)
[11:18:28] <Chemeleon> though there are 3d printers down around the 500 mark last I checked, but they also use significantly smaller electronics and whatnot to cut the price
[11:18:31] <archivist> who needs silly travel speeds when cutting speed is the real limit
[11:18:39] <syyl_ws> with good reason in my mind, Chemeleon
[11:18:42] <ssi> archivist: agreed
[11:18:44] <Chemeleon> archivist, yep
[11:18:51] <ssi> the only machine I care about speed on is my plasma table
[11:18:53] <syyl_ws> the real limits are physics
[11:18:54] <syyl_ws> in my mind
[11:18:59] <Chemeleon> my bigger machine jogs at 300 fine, but no way it comes even close to that cutting :(
[11:19:00] <ssi> cause travel speed == cutting speed :)
[11:19:07] <syyl_ws> you cant go with 6m/min in a sharp corner
[11:19:21] <syyl_ws> or around a corner
[11:19:26] <syyl_ws> it will look like crap
[11:19:27] <andypugh> It depends on what you are cutting I guess. 3D-profiling of foam is likely to be limited by axis speed rather than tool/spindle
[11:19:35] <Chemeleon> I figure with those sort of speeds, I'll probably also have to bolt this machine to my desk
[11:19:42] <Chemeleon> luckily my desk is rediculously heavy :)
[11:19:45] <Tom_itx> you want uber fast use linear slides
[11:19:58] <ssi> http://www.youtube.com/watch?v=wNo5JsQFgco
[11:19:59] <Tecan> (wNo5JsQFgco) "Cnc plasma axis test" by "imcmahon" is "Tech" - Length: 0:00:14
[11:20:06] <ssi> 3000ipm, 200i/s/s
[11:20:10] <Tom_itx> not uber cheap though
[11:20:21] <Chemeleon> very nice ssi :)
[11:20:35] <CaptHindsight> why all the connector plates since the t-slot has 4 holes that you can tap and use for bolts?
[11:20:39] <ssi> I need to rebuild that axis
[11:20:50] <ssi> it only has one rail, and it gets warbly at high speed
[11:20:56] <CaptHindsight> but that might be too difficult for DIY
[11:21:00] <Chemeleon> CaptHindsight, trying to keep tapping and drilling to a minimum for the design
[11:21:06] <ssi> why? I DIY'd it :P
[11:21:14] <ssi> that machine is 100% scratchbuilt
[11:21:15] <Chemeleon> I'll make my own plates, as a friend can get me aluminum plate really cheap (3.50/pound)
[11:21:19] <ssi> belt reductions and all
[11:21:24] <Chemeleon> but others could buy them premade that way
[11:21:28] <CaptHindsight> connector plates + gussets
[11:22:00] <Chemeleon> and I still need to redo the plates a little around the gantry, shifted the design slightly when I started adding rails, and didnt put them back on yet
[11:22:23] <andypugh> There are self-tapping screws that fit the T-slots. Then all you need is a tool-acess hole drilled through from the back.
[11:22:45] <CaptHindsight> 8020 offers drilling and tapping services, you just call out the the profile + length and holes size and location, they even tap them
[11:23:04] <Chemeleon> any idea what the pricing is like CaptHindsight?
[11:23:27] <Chemeleon> probably would be good to work up a price list if someone wanted to buy things ready to go like that
[11:23:39] <ssi> have you thought about using smaller rail
[11:23:40] <CaptHindsight> heh, yeah only $2 each for the self tapping screws, they make lots on all the fasteners
[11:23:44] <ssi> 8020 is huge for that little machine
[11:24:03] <ssi> I'd use 20mm rail in something like that
[11:24:09] <Chemeleon> ssi, yea, its completely overkill, but when I ran my early estimates for smaller 8020 sizes, the cost difference was minimal
[11:24:15] <Chemeleon> the rail in those pictures is 16mm
[11:24:26] <Chemeleon> would run about $200 shipped for all 6 rails and the carriages
[11:24:30] <andypugh> http://uk.rs-online.com/web/p/connecting-components/4667304/
[11:25:01] <ssi> no, not the rail sorry
[11:25:02] <ssi> extrusion
[11:25:02] <andypugh> £0.40 per screw, not quite $2
[11:25:07] <CaptHindsight> source from a Chinese fastener maker
[11:25:11] <ssi> http://www.mcmaster.com/#t-slotted-extrusions/=of3xos
[11:25:16] <ssi> I use a pile of that 20mm solid extrusion
[11:25:23] <ssi> it's much cheaper than 8020
[11:25:28] <CaptHindsight> andypugh: now try that same fastener from Bosch :)
[11:25:35] <ssi> get it from mcmaster or misumi
[11:25:43] <Chemeleon> I'll check into that ssi :)
[11:25:43] <ssi> misumi will cut and drill and tap to order
[11:25:50] <Chemeleon> I've been using 8020 ebay prices for my estimates mostly
[11:26:02] <Chemeleon> its prevalent enough on there for the same general prices for me to consider it reasonably repeatable for others
[11:26:03] <andypugh> CaptHindsight: That is from Bosch. Via RS components (never a route to bargain pricing in itself)
[11:26:35] <CaptHindsight> Chemeleon: just send 8020 the BOM, they will even redesign it for you and send you all the pricing
[11:26:49] <andypugh> I wonder if there would be an advantage in adding glue to the screwed joints?
[11:27:03] <Chemeleon> yea, will do so once done CaptHindsight :)
[11:27:36] <Chemeleon> looks like with the cncrouterparts bearing setups, I'd be looking at around $218 + the metal for it to ride on, which should fall within budget
[11:27:42] <Chemeleon> ya'll think that'd be superior to round rails?
[11:27:57] <ssi> so, I suggested that before I looked at your machine
[11:27:58] <ssi> I don't know that their stuff is SMALL enough for you!
[11:28:30] <Chemeleon> I'd have to shift dimensions slightly, but I dont think it'd expand it enough to be a negative
[11:28:34] <ssi> ok
[11:28:50] <ssi> here's the little machine I built
[11:28:50] <ssi> https://sphotos-b-mia.xx.fbcdn.net/hphotos-ash4/309920_10100133657681442_251706433_n.jpg
[11:28:54] <Chemeleon> atm, total footprint is around 22"x32"
[11:28:54] <ssi> it's all round rail
[11:29:03] <Chemeleon> so its a little big for desktop sized, but still small enough for mine at least :)
[11:29:22] <ssi> here's another machine I built:
[11:29:23] <ssi> https://sphotos-b-mia.xx.fbcdn.net/hphotos-ash4/253334_10100133670565622_1712486917_n.jpg
[11:29:36] <ssi> used prebuilt linear actuators I bought at a surplus store in sunnyvale ca
[11:30:05] <Chemeleon> I've wondered about that sort of design in the past - do those actuators not let it flex too much?
[11:30:30] <ssi> they're acme screw/profile rail slides
[11:30:33] <ssi> they're VERY rigid
[11:31:03] <ssi> I paid like $40 apiece for them
[11:31:07] <ssi> I wish I could buy a hundred at that price
[11:31:07] <ssi> heh
[11:31:13] <Chemeleon> cool, if I ever find some at an auction cheap, would make a nice addition to my beast of a drill press :)
[11:31:28] <ssi> they only have like 3" travel
[11:31:31] <ssi> maybe 4"
[11:31:36] <ssi> my plan was to make that an engraver
[11:31:40] <ssi> but I haven't finished it
[11:31:45] <ssi> I have a lot of half-finished projects :P
[11:31:51] <Chemeleon> dont we all? :)
[11:31:59] <ssi> no I don't think you understand hahahah
[11:31:59] <Chemeleon> though I'd say mine are more typically 90% finished ;)
[11:32:12] <andypugh> Chemeleon: I like the idea of making a machine by screwing/bonding linear guides to a section of granite countertop.
[11:32:19] <ssi> I have no fewer than five half-finished cnc machines
[11:32:20] <CaptHindsight> http://www.automationtechnologiesinc.com/products-page/cnc-router/kl4530-desktop-cnc-router-with-3-motors
[11:32:22] <ssi> and a half-finished airplane
[11:32:28] <ssi> and that's just scratching the surface :(
[11:32:32] <Chemeleon> even my drill press qualifies for that, there were a few spots I couldnt easily repaint while it was disassembled, so put it back together intending to paint afterwards, and...still havent done so
[11:32:42] <CaptHindsight> Chinese surface plates are low cost
[11:32:45] <Chemeleon> in my defense, the main part left to paint has a lot of moving pieces to avoid getting paint on :)
[11:32:51] <Chemeleon> www.chemeleon.net/drillPress_wip10.jpg
[11:32:57] <ssi> CaptHindsight: yea they're very inexpensive
[11:33:12] <ssi> that's a cool machine
[11:33:25] <Chemeleon> got it from an auction of an old machinists shop that closed up
[11:33:33] <Chemeleon> place was being torn down the next day
[11:33:43] <Chemeleon> I finally got it taken apart and loaded into the uhaul at 2am the night before
[11:33:54] <ssi> "drill press" doesn't really do it justice
[11:33:55] <ssi> heheh
[11:33:55] <Chemeleon> it didnt look quite so big in the auction picture ;)
[11:33:59] <Chemeleon> heh, yea
[11:34:23] <Chemeleon> 5hp motor (biggest physically I've ever seen for that sort of hp), gear driven with variable speed, power feed, reversing, and 11" stroke
[11:34:31] <ssi> anyway I need to go cut the grass and go shopping so I can come home and code
[11:34:32] <Chemeleon> rated for 1.5" hole in steel in one go
[11:34:35] <ssi> bbiab
[11:34:56] <Chemeleon> I suspect its close to as big as you can go before you're talking radial arm drills :)
[11:35:00] <Chemeleon> complete and total overkill
[11:35:04] <Chemeleon> but just cool :)
[11:35:52] <Tom_itx> andypugh, maybe loctite
[11:36:32] <andypugh> I am sure that there is a Loctite for the job.
[11:36:33] <CaptHindsight> http://www.shars.com/products/view/2340/Grade_B_12quot_x_18quot_Black_Granite_Surface_Plate $39.85
[11:36:47] <Chemeleon> I got a bit of an overkill surface plate at that same auction too :)
[11:36:57] <Chemeleon> 4'x6' 6.5" thick granite slab
[11:37:09] <Chemeleon> and a stand built from 3/8" thick tubing and angle iron for it
[11:37:17] <Chemeleon> the metal in the stand alone would have cost more than I paid for the whole thing
[11:37:18] <andypugh> CaptHindsight: Yeah, that would make a really nice base for a small CNC wouldn't it?
[11:37:44] <Chemeleon> huh, thats an interesting idea, dont think I've seen any built around a slab of granite before
[11:37:50] <Chemeleon> would definitely add a lot of mass to stabalize things
[11:37:53] <CaptHindsight> andypugh: it's what I use, I just bought 2 of those for small printers
[11:37:55] <Tom_itx> wonder if a slab from the undertaker would be cheaper
[11:38:19] <andypugh> Kitchen worktop is an option too.
[11:38:32] <jthornton> a fouled up marker stone maybe
[11:38:43] <andypugh> I am thinking about a CNC lathe of that construction. But then I accidentally bought a Rivett.
[11:39:39] <CaptHindsight> drilling granite might be a problem for DIY
[11:40:11] <andypugh> Should be no trouble with an SDS drill.
[11:40:48] <andypugh> I would reckon on oversize holes and resin-bonded anchors aligned by jigs.
[11:41:51] <andypugh> Might need a drilling jig to prevent the bit wandering.
[11:42:19] <andypugh> I have naver seen an impact drill in a bench press
[11:42:36] <CaptHindsight> bond solid aluminum in the holes , then drill and tap
[11:42:45] <Tom_itx> just a diamond tip drill should do the trick
[11:42:46] <andypugh> Plan B might be diamond core drills and a milling machine.
[11:43:01] <CaptHindsight> I have a mill dedicated to drilling granite
[11:43:07] <andypugh> CaptHindsight: Yes, that's a good plan.
[11:43:47] <andypugh> OK, ignore my suggestions. Go with CaptHindsight and bond-in oversize Alu plugs then drill and tap :-)
[11:44:26] <CaptHindsight> 4" slabs don't take too long to bore through with a diamond core bit, they just get pricey for longer lengths
[11:44:44] <Tom_itx> that's a fix i use when repairing pulled screws from wood. drill out quite a bit and insert a dowel plug
[11:46:31] <Chemeleon> I see two ways I could do those cncrouterparts carriages on the table for this router, not sure which would be best though
[11:46:52] <Chemeleon> I could do the metal plate oriented horizontally, short carriages on each corner pushing "out" to lock the table in place
[11:47:05] <Chemeleon> or I could do two long ones on each side, pinching a vertical plate between them
[11:47:33] <Chemeleon> the first seems easiest to build, but theres no adjustment built in for any side to side slop if things arent tight from the getgo
[11:48:11] <Chemeleon> it would also take up a bit less space
[11:48:46] <andypugh> I have no opinion. I like prismatic slides. I might accept Turcite at a push.
[11:50:51] <Chemeleon> guess I could design some sort of way to adjust tightness for the first, probably simpler than the mild gymnaistics for orienting the steel and carriages for the second style
[12:07:47] <CaptHindsight> automation technologies complete router $800 and ~1/2 that if you buy direct from China in bulk
[12:08:37] <CaptHindsight> if you offered an improved design using a granite surface plate the shipping might be an issue
[12:10:09] <Chemeleon> looks like my 1k goal budget may be a bit trickier than I'd thought, just reran the numbers, and I'm coming in around $1050 before shipping costs :\
[12:10:20] <Chemeleon> have to see where I can somehow trim that down without affecting performance
[12:10:28] <CaptHindsight> whats your ultimate goal? open plans for a DIY with a BOM <$1k? or is there something else?
[12:10:44] <Chemeleon> yea
[12:10:56] <Chemeleon> free plans, and a cost to build it under 1k that others can duplicate easily
[12:11:05] <Chemeleon> so trying to avoid my usual deals since they tend to be one off things :)
[12:11:20] <CaptHindsight> you do learn quite a bit from designing and building
[12:11:24] <Chemeleon> if it was just for me, and I was being patient, I bet I could do it for at least a few hundred less collecting parts as I find them cheap
[12:11:51] <Chemeleon> I am basing the price on things like drilling those plates manually rather than buying prebuilt, but if someone wanted to spend a bit more they could buy them from 8020 instead
[12:12:04] <CaptHindsight> but your kit is $200 more than a complete cnc router
[12:12:11] <Chemeleon> while I'm trying to keep assembly simple, drilling a grid of 1/4" holes in aluminum isnt too hard imo
[12:12:26] <Tom_itx> you still have an international audience so sourcing parts is still an issue
[12:18:53] <Chemeleon> also going for easy to build
[12:19:03] <Chemeleon> bolting things to a granite slab would up the complexity for the average person
[12:19:16] <Chemeleon> so far, I think this should be buildable in a weekend with just run of the mill tools
[12:19:47] <andypugh> I forgot that you wanted it to be buildable by boobs.
[12:19:52] <andypugh> noobs, I mean.
[12:19:55] <Chemeleon> heh
[12:20:06] <Chemeleon> depending on your local slang, either is appropriate :)
[12:20:14] <andypugh> That will necessarily reduce the available quality.
[12:20:45] <Chemeleon> I dont think end quality needs reduced to be easy to build, though its possible someone with no skill may not build things accurately enough
[12:20:58] <Chemeleon> you still have to be able to cut to length accuruately, drill holes where you want them, etc
[12:21:18] <Chemeleon> but thats easy stuff for most woodworkers at least
[12:22:08] <CaptHindsight> spend a few days lurking in #reprap, you might end up changing your mind about how well these things actually work after assembly
[12:24:06] <archivist> woodworker and closer than a chalk mark may be asking a bit much
[12:25:38] <Loetmichel> Chemeleon: to drill holes where you want them CAN be a challenging tast, especially in (real) wood.
[12:25:42] <Loetmichel> task
[12:26:31] <archivist> and you are asking for .0002 and using an extrusion that slumps after the extrusion die
[12:27:45] <Chemeleon> archivist, not asking for that, just saying deflection due to strength of the extrusion shouldnt be an issue
[12:28:49] <Chemeleon> I know accuracy wont be in the 10 thousandths range :)
[12:29:09] <Chemeleon> if accuracy matches my bigger one but at a much higher speed, I'll be more than happy with things
[12:33:18] <andypugh> Machine tool setup and fitting is a skilled and difficult job, and the final result will strongly depend on how that is done.
[12:33:36] <Chemeleon> yep
[12:33:56] <Chemeleon> I know I'm capable of it though, and I dont consider myself overly skilled in my shop, so its reasonable to assume others can too
[12:34:21] <andypugh> So your clueless noob might end up with a pile of junk and blame your design, not his incompetence.
[12:34:26] <CaptHindsight> if you have everything precision cut, drilled an pinned you might be able to remove much assembly error but then you add to the price of the BOM
[12:34:29] <Chemeleon> and if they really are that bad, they can always pay extra to have 8020 precut everything I guess
[12:34:53] <Chemeleon> I'll definitely have a disclaimer at the start, unfortunately thats required for nearly anything these days :(
[12:35:55] <andypugh> Now, if you press-fit solid carbide rods into all the circular holes in the extrusion, that would help the stiffness. Possibly not the price.
[12:36:24] <Chemeleon> heh, yea that sounds a little more expensive than I'm aiming for :)
[12:36:45] <IchGuckLive> hi all B)
[12:36:57] <andypugh> AHoy
[12:37:03] <Chemeleon> I could easily spend 5-10x as much for the same size chasing increased stiffness and accuracy, but its not needed for this build
[12:37:07] <Chemeleon> #1 goal is cutting speed
[12:37:39] <andypugh> Then you might not want steppers.
[12:37:46] <IchGuckLive> speed is money in some cases
[12:37:56] <andypugh> But that is all you will get at the price, I suspect.
[12:38:05] <Chemeleon> andy, yea, I suspect they may be a bit of a limiting factor
[12:38:06] <IchGuckLive> steppers are good up to 5m/min
[12:38:15] <Chemeleon> but the numbers I've run so far says they "should" be ok
[12:38:17] <IchGuckLive> 5000mm
[12:38:40] <CaptHindsight> 5m = 5000mm, got it :)
[12:38:49] <IchGuckLive> with the 5i25 7i76 up to 8500
[12:38:49] <Chemeleon> I know my bigger machine "can" cut at the speeds I'm talking, its just not stiff enough to do so with any accuracy
[12:39:06] <IchGuckLive> at 100kg mashine pull
[12:39:15] <Chemeleon> and I plan to use the same screw and electronics setup for this smaller one
[12:39:35] <IchGuckLive> i use this on my 3 plasmas
[12:40:57] <IchGuckLive> with 8Nm steppers
[12:41:27] <IchGuckLive> on a not needed 0,01mm accurat
[12:41:45] <IchGuckLive> the plan is to go for woodcarving as well
[12:41:56] <IchGuckLive> plasmas with 0.2mm is goog ro go
[12:42:17] <IchGuckLive> plasmas with 0.2mm is good to go
[12:42:21] <IchGuckLive> O.O
[12:43:06] <CaptHindsight> Chemeleon: IchGuckLive has built lots of low cost routers of your type
[12:43:16] <Chemeleon> cool :)
[12:45:28] <IchGuckLive> count has been endet at 120+ mashines no more counts
[12:45:44] <Chemeleon> why've you built so many, to sell them?
[12:45:59] <IchGuckLive> school units most
[12:46:00] <CaptHindsight> I need a large spatula for flipping 400lb surface plates 180deg, but I don't see one in the Harbor Freight catalog
[12:46:18] <Chemeleon> heh
[12:46:38] <Chemeleon> I still need to either figure how to tilt mine, or partially dismantal the doorway
[12:46:42] <CaptHindsight> time to make a jig
[12:46:47] <Chemeleon> have a small room sectioned off in the corner of my shop I want to put it in
[12:46:49] <Chemeleon> 48" door
[12:46:56] <Chemeleon> 49" wide surface plate at the widest points
[12:47:05] <Chemeleon> doh
[12:47:32] <IchGuckLive> Chemeleon: what is the goal the router needs to be on
[12:47:47] <IchGuckLive> mashine parts milling
[12:47:48] <IchGuckLive> wood
[12:47:51] <Chemeleon> 12"x12" cut area, doing 300ipm with a 1/8" bit in wood for 3d heightmap carvings
[12:47:54] <IchGuckLive> model scape part
[12:48:21] <IchGuckLive> Chemeleon: are you in europ
[12:48:22] <Chemeleon> www.chemeleon.net/diycamCNC_wip2.jpg is where the design currently stands, still have some work to do on it, and debating switching the rails for something different
[12:48:24] <Chemeleon> usa
[12:48:59] <IchGuckLive> hevy duty
[12:49:07] <Chemeleon> yea, significantly overbuilt
[12:49:09] <IchGuckLive> you do not need that kind of fixure
[12:49:36] <Chemeleon> but tired of dealing with flex on my current bigger router, so decided to just ensure this one will have no issues whatsoever with that
[12:49:39] <awallin> except the z-motor isn't attached at all :)
[12:49:51] <Chemeleon> awallin, yea, like I said, its not done yet :P
[12:50:12] <Chemeleon> still have to setup screws and nuts and their mounts and whatnot, and possibly tweak dimensions a little more
[12:53:06] <IchGuckLive> Chemeleon:
http://www.sammellothar.de/X_RAHMEN.mpg
[12:53:12] <IchGuckLive> as simple as it coudt be
[12:53:26] <IchGuckLive> this is 250USD
[12:53:44] <IchGuckLive> modern stable version is 400 euros
[12:54:06] <Chemeleon> how do you do it for 250usd? electronics alone are more than that over here
[12:54:54] <IchGuckLive> here is the plasma at 1500Euros
http://www.youtube.com/watch?v=V6kllGspehQ
[12:54:55] <Tecan> (V6kllGspehQ) "Plasma konstruktion auf Freecad (view)" by "magic33de" is "Education" - Length: 0:03:36
[12:55:53] <IchGuckLive> Chemeleon: electronics is around 50 USD including stepeprs driverer and power
[12:56:19] <Chemeleon> where do you get them that cheap? I'm looking at about $50 just for a single stepper
[12:56:50] <IchGuckLive> i use epspn laser printer stepeprs on 2USD etch
[12:56:58] <Chemeleon> ah
[12:57:06] <IchGuckLive> ok your size
[12:57:08] <Chemeleon> nice :)
[12:57:14] <Chemeleon> what sort of cut speeds do you get with that?
[12:57:21] <IchGuckLive> 1200mm/min
[12:57:38] <Chemeleon> quite good for that price range :)
[12:57:43] <IchGuckLive> 4x4x1inch
[12:58:02] <Chemeleon> I'm aiming for about 7600 mm/min though
[12:58:05] <IchGuckLive> no sorry 160x160x50mm
[12:58:25] <Chemeleon> found a calculator that said a 1/8" bit can handle just over that cutting, so trying to hit it :)
[12:58:45] <IchGuckLive> NP with the 5i25
[12:59:14] <IchGuckLive> pricise is good at 0.1mm
[12:59:37] <IchGuckLive> 0.004
[12:59:49] <IchGuckLive> sox calculate on that
[12:59:55] <Chemeleon> nice
[13:00:05] <IchGuckLive> for your need
[13:00:48] <IchGuckLive> my router part supply
http://stores.ebay.de/cnc-discount
[13:01:19] <IchGuckLive> electronic
http://www.pro-tos.de/shop/
[13:01:31] <IchGuckLive> for the real working mashines
[13:02:23] <Chemeleon> barring finding something comparable and cheaper, I'll probably just use this kit:
http://www.automationtechnologiesinc.com/products-page/stepper-nema23-3-axis-kits/cnc-stepper-motor-3-axis-kit-2
[13:02:33] <IchGuckLive> this steppers are the bunch of force
http://www.pro-tos.de/shop/Schrittmotoren/Schrittmotor-34H2120-42-8-8-Nm.html
[13:02:42] <Chemeleon> same as on my bigger machine (though I used 2 motors on the gantry movement there to prevent skewing) and works well there
[13:02:55] <Chemeleon> I did have one stepper motor sheer off its shaft at the start, but they replaced that without complaint
[13:03:20] <IchGuckLive> 48V best to go
[13:03:21] <CaptHindsight> http://img265.imageshack.us/img265/8929/cncrouteruj1.jpg how about DIY for the forms for something like this?
[13:03:33] <Chemeleon> that all granite?
[13:03:59] <CaptHindsight> epoxy granite
[13:04:18] <IchGuckLive> cool stone design
[13:04:19] <Chemeleon> interesting, never seen one like that before
[13:04:26] <Chemeleon> bet thats just a little on the heavy side :)
[13:04:29] <IchGuckLive> somewone recicled the kitchen
[13:04:36] <andypugh> Concrete might be just as good in that sort of nulk
[13:04:40] <andypugh> (bulk)
[13:04:57] <IchGuckLive> jeh GFK concrete
[13:04:59] <Chemeleon> any idea how epoxy granite costs compared to concrete? at some point, I intend to build a large wood lathe for bowl turning, figured I'd use concret for it
[13:05:34] <IchGuckLive> the problem is wher to get the shreddered strings
[13:05:46] <IchGuckLive> rest is standart foam concrede mix
[13:07:00] <CaptHindsight> the polyester resin part ~$45 a gallon + whatever gravel
[13:07:32] <Chemeleon> lot more expensive then, think I"Ll stick to concrete when its time for that project :)
[13:07:40] <CaptHindsight> figure 20% by volume is resin, 80% aggregate
[13:08:29] <IchGuckLive> Chemeleon: only wood i wood go at your opinion
[13:09:37] <andypugh> I am not sure that you need that sort of stiffness in a wood lathe. even the good ones are fairly flimsy
[13:09:57] <Chemeleon> for large bowls, its needed
[13:10:13] <Chemeleon> you need a lot of mass for things not to wobble when the stock isnt yet turned round and balanced
[13:10:22] <andypugh> This might take things a bit far:
https://www.aldi.co.uk/en/specialbuys/sunday-8th-september/
[13:10:27] <andypugh> (Eeeew!)
[13:10:34] <Chemeleon> heh
[13:10:56] <Chemeleon> even ignoring the awfulness of that design, I'm talking 20-30" diameter logs mounted to turn :)
[13:11:16] <Chemeleon> I have a 12" lathe currently, works fine up to its capacity
[13:11:25] <Chemeleon> but I want my next lathe to be capable enough that its never the limiting factor
[13:11:32] <ReadError> im not sure i would buy a lathe from a place that sells "Mini Potato Waffles"
[13:11:33] <ReadError> ;)
[13:11:35] <archivist> google the fella turning lampshades from wood at that diameter
[13:11:36] <Chemeleon> and to buy something commercially made like that, I'm looking at 4-10k
[13:11:45] <Chemeleon> seen it :)
[13:11:57] <Chemeleon> laughed at the people complaining that he was "wasting" wood in the comments
[13:12:18] <archivist> clue, they do not have it
[13:12:59] <andypugh> Well, he is wasting wood, but more justifiably than many, many other uses of wood.
[13:16:41] <IchGuckLive> ALDI Germany good god Things B)
[13:16:55] <IchGuckLive> O.O
[13:17:34] <Chemeleon> right, I should probably stop working on a design I wont start building for at least a month if not longer, and go get some stuff done that needs done now :)
[13:17:59] <Chemeleon> got a small sign/nameplate to run off for a friend, and then a carving to start that'll take around 14 hours on my current machine
[13:18:02] <IchGuckLive> winter is coming so plenty of time
[13:23:07] <IchGuckLive> ok by friebnds til tomorrow
[13:41:19] <CaptHindsight> there are several studies you can Google about cast iron vs epoxy granite or concrete
[13:43:07] <CaptHindsight> besides the differences in flexural modulus and compressive strength, you'll see the difference in dampening vibrations
[13:45:49] <CaptHindsight> but on a small router I bet the spindle and how it's moved and held was probably the problem
[13:48:07] <Loetmichel> on small hobby gantry machienes it can be cost effective to fill the hollow parts of the gantry/bed with buckshot or fine gravel
[13:48:18] <Loetmichel> to dampen vibrations
[13:48:21] <CaptHindsight> ReadError: gee thanks, now I'm craving mini potato waffles
[13:50:14] <CaptHindsight> aluminum tube vs t-slot with cavities filled with resin+aggregate
[13:50:49] <CaptHindsight> or steel tube filled
[13:51:43] <Loetmichel> capricorn_1: LOOSE fillings are way better do dampen vibrations!
[13:51:48] <Loetmichel> grr
[13:51:50] <Loetmichel> CaptHindsight
[13:56:06] <CaptHindsight> something very rigid that won't conduct a wave
[14:01:41] <Tom_itx> andypugh, not necessarily carbide rods, even hardened dowel pins used in punch dies would suffice
[14:01:50] <Tom_itx> and those are cheap
[14:03:06] <Tom_itx> or even drill rod
[14:07:38] <Tom_itx> CaptHindsight, what mold release agent is used in that epoxygranite stuff?
[14:09:16] <CaptHindsight> Tom_itx:
http://www.henkelna.com/frekote-mold-release-agents-18643.htm?utm_source=google&utm_medium=cpc&utm_term={keyword}&utm_campaign=Frekote%2Bmold%2Brelease
[14:09:27] <Tom_itx> valen made a column out of that stuff for his mill
[14:10:04] <Tom_itx> i wonder if regular fiberglass release agent would work
[14:10:11] <Tom_itx> i've got a gal of that already
[14:10:15] <CaptHindsight> yeah
[14:10:32] <CaptHindsight> fiberglass resin is usually polyester
[14:11:52] <CaptHindsight> polyester with some fillers (talc) to reduce the shrinking when it polymerizes
[14:12:25] <Tom_itx> yeah, i've got a small jar of glass beads leftover from previous projects
[14:20:19] <dosas> hey guys, i wondered if it is possible to run halcmd while linuxcnc is running to get the state of certain parameters
[14:20:38] <dosas> i know i can run halmeter but i want to do it via cmd line
[14:21:06] <awallin> good epoxies stick like crazy to the mould. I've damaged several moulds when trying to release a part, even when using frekote or similar good release agent
[14:22:34] <CaptHindsight> they are good adhesives :)
[14:25:10] <CaptHindsight> silicones might have worked better for those apps
[14:27:04] <skunkworks> dosas: Yes - open up a terminal and run halcmd -fk
[14:27:09] <CaptHindsight> hydrogels work even better but it's difficult to make for large molds
[14:28:14] <skunkworks> (think it is fk
[14:28:29] <dosas> thanks
[14:34:27] <CaptHindsight> PVA and wax seem to work for DIY, the PVA is a porous and flexible structure when dried
[14:47:22] * JT-Shop should get some work done
[15:09:50] <kengu> maybe i'll just go and look around in workshop to see if there is some suitable plastic around to use as wood chip guard in router
[15:10:13] <kengu> maybe no need for a brush strip as long as I have no idea of local supplier
[15:12:03] <andypugh> Brush strip is probably available from the local hardware shop as draught excluder.
[15:13:32] <kengu> can be but i would not bet on it
[15:19:39] <andypugh> <checks ip> Ah, yes, Finnish draughts probably need something a bit more industrial :-)
[15:20:31] <andypugh> B&Q have it:
http://www.diy.com/nav/build/insulation/draught-excluders/seals/B-and-Q-Draught-Excluder-Brush-Strip-12628580?skuId=13149286 But not Clas Ohlsen.
[15:21:10] <kengu> yeah.. i was about to comment on that that here we usually use something that is called "insulation" more and less just draught guard
[15:21:21] <andypugh> This is about the point that I order from eBay and put that project to the back for a few days.
[15:22:07] <kengu> yeah.. it is project.. not for next week but the week after
[15:23:18] <kengu> sure available here also somewhere..
[15:24:47] <kengu> the problem with those draught excluders is the height..
http://www.ebay.ie/itm/Stormguard-838mm-Bottom-of-The-Door-Brush-Strip-Draught-Excluder-Real-Wood-/261251472521 like this.. where the brush is only 25mm
[15:25:23] <kengu> i was looking at one of those yesterday but that fact then turned me down
[15:27:16] <kengu> i will still need to go with the jaseals or formseal or some other commercial way
[15:27:45] <kengu> but I think I will test or prototype it with just flexible plastic strips
[15:31:35] <dosas> hey guys, is it possible to define multiple postgui_halfiles in the .ini file ?
[15:33:48] <andypugh> No, sadly not
[15:34:16] <andypugh> In fact I think that the PostGUI HAL file is actually called by the GUI iitself.
[15:39:54] <dosas> too bad
[15:40:08] <dosas> and i have to use postgui if i want pyvcp, right?
[15:41:14] <andypugh> Yes. The pins from pyVCP don't exist until the GUI is loaded.
[15:41:52] <andypugh> Why is it a problem to only have one postgui hal file?
[15:45:11] <dosas> it is not really a problem
[15:45:17] <dosas> but i just like modularity
[15:45:23] <dosas> if that is a word
[15:45:48] <dosas> so if you have different types of machines and some need the sam modules some not
[15:45:53] <dosas> this would be convenient
[15:46:41] <andypugh> If you refer to a signal in the postgui hal file (rather than a pin) that does not exist then all that happens is that that signal is created
[15:47:38] <andypugh> So, the PyVCP file can refer to signals that might or might not exist in the rest of the config. As long as it doesn't refer to any pins that don't exist.
[15:48:04] <andypugh> I don't know if that helps.
[15:49:12] <dosas> maybe yes
[15:49:36] <andypugh> I have never tried it, but it is just possible that a HAL file can read the contents of another HAL file using "source"
[15:50:19] <andypugh> It works at the halcmd prompt. I can see reasons it might not work in an actual file.
[15:59:07] <dosas> no then i'm just gonna cat it all together
[16:05:53] <andypugh> The best solution would be to fix all the GUIs to accept multiple postgui files :-)
[16:08:57] <dosas> yes that would be great
[16:09:05] <dosas> and also accept multiple pyvcp files
[16:09:24] <andypugh> You can have multiple PyVCP files.
[16:09:29] <dosas> another question: is it possible to show parameters like pid gain in the pyvcp panel
[16:09:48] <andypugh> You can only have one docked on the left, but you can have floating ones and ones embedded in tabs too.
[16:10:07] <Tom_itx> 2 litre bottle might work as a chip guard
[16:10:21] <Tom_itx> comes in colors too
[16:10:21] <dosas> i have
[16:10:22] <dosas> setp pid.0.Pgain 0.3
[16:10:30] <dosas> net pgain pyvcp.pgain pid.0.Pgai
[16:10:37] <andypugh> You can't show parameters in a PyVCP panel
[16:10:42] <dosas> fuck
[16:10:45] <dosas> okay
[16:10:47] <andypugh> You can show pins. PID gains are pins
[16:11:06] <dosas> sure
[16:11:23] <dosas> because when i connect them to pyvcp they are always zero
[16:11:26] <dosas> if not it works
[16:11:32] <andypugh> PID gains used to be parameters but were changed to pins for just the reason you are talking about.
[16:11:47] <dosas> so the lines above should work
[16:11:58] <andypugh> But, they are IO pins. Which are a bit strange.
[16:13:13] <andypugh> http://www.linuxcnc.org/docs/html/man/man9/tristate_float.9.html is part of the answer, but that is for driving the IO pin, not for reading back the value
[16:13:35] <andypugh> What sort of widget is pyvcp.pgain?
[16:14:07] <dosas> its a number
[16:14:53] <andypugh> Can you be more specific?
[16:15:13] <Tom_itx> what flavor number?
[16:15:24] <Tom_itx> chocolate, rasberry, orange orange...
[16:15:31] <andypugh> What widget name did you use?
[16:15:40] <dosas> 1.3f
[16:15:43] <dosas> is the flavour
[16:15:52] <dosas> and the widget is <number>
[16:16:10] <dosas> <halpin>"igain"</halpin>
[16:16:10] <dosas> <font>("Helvetica",12)</font>
[16:16:10] <dosas> <format>"1.3f"</format>
[16:16:10] <dosas> </number>
[16:16:32] <dosas> for the xml file?
[16:17:00] <andypugh> OK, it's a display not a control (which is what I was wondering about) and so it ought to work.
[16:17:00] <dosas> sorry i don't get the question
[16:17:34] <dosas> without the tristate?
[16:17:46] <andypugh> tristate goes the wrong way, I think
[16:17:57] <JT-Shop> http://www.youtube.com/watch?v=NiCCnbDVt5c
[16:18:21] <DJ9DJ> gn8
[16:21:18] <andypugh> dosas: You might find some useful stuff in this PID tuning GUI
http://www.linuxcnc.org/hardy/index.php/russian/forum/10-advanced-configuration/6495-development-pid-tuning-gui-and-hal?limitstart=0#6495
[16:21:57] <andypugh> Presumably he has solved the problem you are struggling with.
[16:22:19] <dosas> thanks for the link
[16:23:15] <dosas> okay i will have a look at it tomorrow no it's time for bed
[17:26:58] <tjtr33> andypugh: when you come back, Apple has a remote controlled thermostat called the Nest , adverts sez "control from anywhere"
