|
Yeah the laser fire hazard is the big one. Over time it builds up this dust inside the machine and that poo poo will go off like a Roman candle. My last job used to collect it in a cardboard box until the day it lit up, and they changed their method. I've never heard of anyone having eye issues from looking at the laser. I don't think that green plastic poo poo they use as a safety barrier is super expensive so I doubt cheaper machines have to compromise too much on safety in that regard.
|
#
?
Jul 30, 2019 04:19
|
|
|
|
| # ? May 29, 2024 23:35 |
|
|
I guess another part of it is that the laser shoots out and can get loose of the machine easily
|
|
#
?
Jul 30, 2019 04:20
|
|
|
shovelbum posted:I guess another part of it is that the laser shoots out and can get loose of the machine easily ?? this is not a realistic thing you have to worry about unless you are attempting to engrave a mirror and also you made the entire chassis interior out of mirrors instead of something normal. (i'm joking, this is not something that Just Happens). even if you did engrave mirrors inside a hall of mirrors machine, the beam would not have any means to leave the machine "easily" except thru a pane of filter glass that renders it harmless and quite safe to look at with no eye protection. errant reflection mishaps seem like a legit potential hazard w powerful laser pointers, which i'd guess is where your assumptions stem from- but a cutting laser head is a very different beast in most regards from a pointer. for one, a pointer is designed to dissipate as little as possible over hundreds of feet, a cutting head is highly-focused on a point just a few inches away and becomes ineffective a few inches past that. people wave laser pointers around willy-nilly in uncontrolled environments full of mirror-quality reflectors, maybe even shine them at people on purpose, while the laser head points down and only down in a tightly-controlled workspace built with safety in mind. nothing in its cutting path is reflective or consistent enough to just bounce the beam perfectly, still coherent and focused as the day is long. i cant imagine what incredible stupidity + negligence you'd have to do to somehow have the laser cleanly reflected right out of the machine through just air and into the shop in a state it could hurt someone, but it'd prolly have to be deliberate and literally involve mirrors/lenses or entirely compromising the chassis and filter glass and all of this involves a probable overabundance of caution, b/c actual ocular injuries from lasers in general are much rarer than you'd assume: wikipedia posted:Studies have found that even low-power laser beams of not more than 5 mW can cause permanent retinal damage if gazed at for several seconds; however, the eye's blink reflex makes this highly unlikely. Such laser pointers have reportedly caused afterimages, flash blindness and glare,[1] but not permanent damage, and are generally safe when used as intended. Volkerball posted:I've never heard of anyone having eye issues from looking at the laser. i have heard of one (1) second-hand possibly-anecdotal instance of someone catching a Permanently-Injurious eyeful of unfiltered laser, and (tellingly) it was a hobbyist eager beaver who jumped the gun on running his half-built machine to knock out some test parts. extrusion frame + motors and drivetrain + laser components set up and working, no enclosure or ventilation at all, and no laser goggle/shades PPE to compensate. iirc some finger-tight fasteners on the laser head came loose and the head swung down, pivoting on the last remaining fastener, having the bad luck to sweep one of his eyes as he hovered inches away staring at the goddamn thing. idk how things turned out long-term for him but his vision in that eye was "compromised for probably a long time". so yeah, that's the sort of thing you'd have to do to actually genuinely risk getting your corneas crisped. and even then he probably would ahve been fine if any single variable had changed- if he was doing literally anything other than staring at the laser working at point blank range when the head fell he prolly wouldnt have been injured; never mind if he'd just tightened the fasteners, waited for an enclosure, wore some goddamn goggles if he's gonna gently caress around in the open like that, etc etc
|
|
#
?
Jul 30, 2019 05:01
|
|
|
That guy would've been fine with a fuckin plywood box, jfc You're definitely right about the fear coming from pointers I think. I've seen a table saw, lathe, drill press, etc throw stuff all over the room anyway
|
|
#
?
Jul 30, 2019 05:32
|
|
|
I think overcaution is great for hobbyists playing with machines, and agree about the laser risks being mischaracterized to some extent. My hacked up K40 hasn't killed me in the 4 years i've had it. Fire risk is definitely the most likely thing to worry about, all the other risks are really in the realm of willful misconduct like running it with the enclosure open or trying to stick a screw driver in the power supply. Keep a fire extinguisher next to the machine, clean out the cardboard/wood debris once in a while, and start low on power and it's no problem. Something new-to-CNC hobbyists don't commonly realize is that your wood-cutting, high RPM CNC router will also start fires very handily, and the longer run times of milling/routing programs will lead to beginner laziness like "I better go check on something in the house for 45 minutes while this program runs".
|
|
#
?
Jul 31, 2019 01:44
|
|
|
Ambihelical Hexnut posted:Something new-to-CNC hobbyists don't commonly realize is that your wood-cutting, high RPM CNC router will also start fires very handily, and the longer run times of milling/routing programs will lead to beginner laziness like "I better go check on something in the house for 45 minutes while this program runs". This is actually not just a problem for hobbyists or wood routers. Even professional-level metal cutting machines can burn down under the wrong circumstances, either when they're using cutting oil as coolant rather than a water-soluble emulsion, or if they're cutting the easier-to-ignite metals like magnesium or titanium. I've heard stories about people trying to use ceramic inserted cutters on titanium and burning their machines to the ground. One shop I did some work for had a guy die a few years back because he welded over a barrel of magnesium chips. Metal fires are nasty. This is not relevant to the home-shop machines which are the subject of this thread (I hope!), I just think it's interesting (and terrifying.)
|
|
#
?
Jul 31, 2019 03:48
|
|
|
Oh definitely, I just assumed that most metal cutting in hobby CNC is low power aluminum work. Let us enjoy: https://www.youtube.com/watch?v=PsFNeiAu04M
|
|
#
?
Jul 31, 2019 04:03
|
|
|
Nowadays they've got these cool fire suppression systems you can put in the machine if you run a lot of magnesium and whatnot. Covers the whole table in foam when it senses a fire. I saw in the news a while back that a system like that saved a machine because the operator was in the bathroom.
|
|
#
?
Jul 31, 2019 04:07
|
|
|
Yeah, I think we require the fire suppresion systems whenever we sell a machine for magnesium or running cutting oil or something. Those crash compilations always disappoint me. Where are the videos of 2000ipm crashes straight into the top of a tomestone? If the pallet hasn't been ripped off its cones, it's not a real crash. Maybe ten years ago, one of our engineers was actually told to crash a machine. The designers had some predictions about how much energy was released in a spindle crash, but they wanted to verify it with emperical testing. So they sent him a machine and told him to start rapiding the spindle straight into the tombstone. So just remember: if you crash a machine, it's called stress testing.
|
|
#
?
Jul 31, 2019 04:20
|
|
|
Karia posted:So just remember: if you crash a machine, it's called stress testing. Spindle bearing thrust load limit study. Or a Dynamic table realignment test.
|
|
#
?
Aug 1, 2019 17:53
|
|
|
Karia posted:Maybe ten years ago, one of our engineers was actually told to crash a machine. The designers had some predictions about how much energy was released in a spindle crash, but they wanted to verify it with emperical testing. So they sent him a machine and told him to start rapiding the spindle straight into the tombstone. im trying to imagine this day at work and its impossible to do so without it turning into an absurd wish-fulfilment fever dream where everyone is patting me on the back and cheering in slow motion as i smile beatifically and set the machine steps to 1.0in/pulse, Z axis, and then just fuckin spin that pendant encoder like its a roulette wheel
|
|
#
?
Aug 1, 2019 20:11
|
|
|
I have done cycle testing on assemblies worth six figures and it's really stressful. Did not enjoy it at all
|
|
#
?
Aug 1, 2019 21:59
|
|
|
Karia posted:Yeah, I think we require the fire suppresion systems whenever we sell a machine for magnesium or running cutting oil or something. Yeah it always makes me wish we'd had a camera running on the machines at some of my old jobs. Wow, you hosed up a Z offset and buried your end mill in plastic, how intense. My favorite one so far in my career was in this model of integrex.  Guy ran a part on setup that ended up being .005" long. Proved everything out without any tools crashing though, so he figured he was good. He incremented the Z on the sub (right side) spindle .005, then hit go and walked away to run another machine. The problem was that the decimal point didn't take when he pushed the button, so it actually incremented positive 5 inches. There's a parameter in mazatrol that will let you set a confirmation screen whenever you make an adjustment over a specified amount, but it was not set on that machine at the time, so the machine said fuckin a bud we're gonna send it. After the cut off and transfer, the lower turret came up to face off the part in the sub spindle and smashed into the jaws at 100% rapid. The collision immediately snapped the locking pins on the lower turret, so the sub and lower turret became an involuntary bevel gear of chaos and destruction, breaking the jaws, the lower turret, and every tool and tool holder in it. I don't know what the final tally was on cost but it was into the tens of thousands. It's a tragedy we don't have video. Volkerball fucked around with this message at 22:42 on Aug 1, 2019 |
|
#
?
Aug 1, 2019 22:38
|
|
|
Volkerball posted:The collision immediately snapped the locking pins on the lower turret, so the sub and lower turret became an involuntary bevel gear of chaos and destruction, breaking the jaws, the lower turret, and every tool and tool holder in it. oh my god lol
|
|
#
?
Aug 1, 2019 23:12
|
|
|
I'm sorry to bother you guys with this, but I couldn't find anything on google. I picked up a genmitsu 3018 to mess around with. One of the things I was looking into doing was putting a stop button in case something gets hosed up. So my very simple question is what would be the best way to wire in a stop button? I was thinking of just splicing it into the power cord, but is there a better alternative?
|
|
#
?
Aug 6, 2019 12:54
|
|
|
AFewBricksShy posted:I'm sorry to bother you guys with this, but I couldn't find anything on google. You could also make a sort of generic e-stop setup by splicing a normal extension cord and putting in a little project box with strain reliefs. Don't know about the controller board on that, but it could also have its own stop input which you might be able to wire into, but I wouldn't consider that as an "absolute" emergency stop, since the depending on the circuit design/firmware it could potentially get ignored somehow.
|
|
#
?
Aug 6, 2019 22:01
|
|
|
It's been ages since I checked this thread and can't remember if this has been discussed to death or not -- has anyone tried converting a Harbor Freight mini-mill to CNC?
|
|
#
?
Aug 7, 2019 11:07
|
|
|
I have a CNC'd X2 (HF # 44991) with a ballscrew 3-axis kit, Gecko G540 . I would not recommend it unless you are only going to be doing 2.5D operations on a single side of a part and can put up with a lot of flat/square/parallel issues. The small work envelope is a pain in the rear end too.
|
|
#
?
Aug 7, 2019 23:23
|
|
|
I have the same kind of CNC setup on a Hitorque 3990, which is similar to the HF mill but solid column, gas assist, belt drive, and 10.6x11.8x5.1 vs 8.5x9x4 movements. I do not have complaints about accuracy or squareness, but agree that the small work envelope is a challenge particularly with doing multiple setups with a tiny vice. There are work arounds, but convenience costs money. I also hate the G540 and have found it to be finicky. On the other side of the garage, my HF dust collector is ready to suck down some wood chips: 
|
|
#
?
Aug 7, 2019 23:51
|
|
|
I don't have a converted mill like that, but from the two people I know who have, I haven't heard much good, also largely for accuracy/squareness reasons. I guess it'll depend on your specific needs; if for some reason you only want to rough out parts w CNC and aren't making anything to tolerances i suppose it would work without improvement, but that wouldn't be a great approach for most CNC milling tasks. If you've already torn one down and rebuilt it with the shimming and part refinement and etc necessary to bring one up to ~+/- 0.001" snuff already, you're probably in a good place, otherwise I'd want to take a step back and think about how to meet your needs most effectively and economically even if it means starting with a different platform.
|
|
#
?
Aug 8, 2019 00:01
|
|
|
Just found this thread, scanned the first few pages and the last few plus ran a search but didn't see any discussion of this. Any thoughts on "Maslow" style builds? My interest in having a CNC setup at home is mostly in making large parts with relatively simple cuts like arcade cabinet components and other furniture-sized items so this seems like a perfect fit for what I want, at least at a glance. It seems like it'd be weak on speed and maybe very tight precision, but as long as it can cut two cabinet sides close enough that I can easily sand down the difference I'm happy.
|
|
#
?
Aug 8, 2019 20:19
|
|
|
Wandering Orange posted:I have a CNC'd X2 (HF # 44991) with a ballscrew 3-axis kit, Gecko G540 . I would not recommend it unless you are only going to be doing 2.5D operations on a single side of a part and can put up with a lot of flat/square/parallel issues. The small work envelope is a pain in the rear end too.
|
|
#
?
Aug 9, 2019 12:18
|
|
|
Of general interest, a resource for single-path/single-stroke fonts. ideal for engraving/marking AND they actually look good enough to be used beyond jsut engineering prints http://imajeenyus.com/computer/20150110_single_line_fonts/index.shtml
|
|
#
?
Aug 21, 2019 04:30
|
|
|
hello thread, back in january I started working in assembly in a sign shop unaware they were mainly interested in my history with the CAD and design softwares i used in school and so after about a month they started training me on a shopsabre 4896. it's pretty cool and i like cutting parts and things for personal use out of stuff i save from the scrap pile. i mostly do acrylic with the occasional aluminum and very rarely brass. mostly i just wanted to say thanks for the information in this thread because work is pretty much only going to teach me as much as they have to in order to cover their purposes, but if I can learn more beyond that i think i might like doing this for a very long time.
|
|
#
?
Sep 1, 2019 05:04
|
|
|
I may be getting a big-boy CNC machine. Our CNC vendor has decided to be a bunch of asses, so my boss told me to find a CNC table. I'm cutting stuff in (max) 1/8" aluminum. Could this be a DIY thing, or do we need to go find an actual CNC router? What kind of spindle power do we need to make full-depth cuts on 1/8" aluminum?
|
|
#
?
Oct 10, 2019 00:21
|
|
|
babyeatingpsychopath posted:I may be getting a big-boy CNC machine. Our CNC vendor has decided to be a bunch of asses, so my boss told me to find a CNC table. Check out the shaper origin. Most routers can cut 1/8 aluminum if its adhered to a sturdy base well. What size do you need to cut?
|
|
#
?
Oct 10, 2019 01:45
|
|
|
Call it 24"x24", +-.003. Smallest typical feature .063" holes. If i could spot those at least, I'll drill them with appropriate drill bits.
|
|
#
?
Oct 10, 2019 04:56
|
|
|
babyeatingpsychopath posted:I may be getting a big-boy CNC machine. Our CNC vendor has decided to be a bunch of asses, so my boss told me to find a CNC table. Any full size router will be able to do that, but you have to ask yourself how much if your time worth, given the fact that the kits for real CNC routers aren't exactly what I'd call easy to assemble, or simple to troubleshoot. An off the shelf unit with a tool changer is a godsend if you can afford it. Being able to spot+drill holes, cut the aluminum, and chamfer it all without doing anything but dicking around on your phone is magical. Having to manually loosen the router collet, change bits, re height them on the tool setter (you did get a tool setter, right?) and hit start again gets real old after about tool change #50. You know, on day 2 or 3.
|
|
#
?
Oct 10, 2019 17:05
|
|
|
Your cutting requirements are very modest, but also the sort of thing you could reasonably get in machines that cost from "less than a thousand USD" to $25k+; as mentioned irt tool-changers, much of what you pay for in commercial CNC machine tools isn't raw cutting ability but additional automation of tedious non-cutting machining tasks, build quality and especially critical component quality (hobby machines subjected to full-time use will fall apart very quickly), turnkey-ness of the whole package, and product support. You pay for productivity and up-time, is a good way to think of it from a business perspective. Given that, you should assess how much you'll actually be running this thing (8 hours a day? one hour a day? once a week? only occasionally for specific projects?) and if your shop will rely on it exclusively for tasks you can't handle otherwise and let that help determine your budget. e: ^^^^^^^^^^^ already said a little more concisely, whoops
|
|
#
?
Oct 11, 2019 01:46
|
|
|
Ambrose Burnside posted:Your cutting requirements are very modest, but also the sort of thing you could reasonably get in machines that cost from "less than a thousand USD" to $25k+; as mentioned irt tool-changers, much of what you pay for in commercial CNC machine tools isn't raw cutting ability but additional automation of tedious non-cutting machining tasks, build quality and especially critical component quality (hobby machines subjected to full-time use will fall apart very quickly), turnkey-ness of the whole package, and product support. You pay for productivity and up-time, is a good way to think of it from a business perspective. General rule of thumb for machines like that is that you should be able to pay it off in a year of dedicated use. If you run it 8 hours a day, 5 days a week, and you're actually making $20/hour on it, then your budget should be about $40,000. If you use it twice a month for 10 hours, you might be able to get away with a much cheaper, less robust machine. A business machine should have business features, and business support. You'll desperately want a tool changer after about a week of using this for production parts. You'll want dust collection. You'll want a controller that doesn't poo poo itself and start losing steps because there is a microwave two units down making a burrito. Hobby tier machines are for tinkers or for people who are able and willing to argue with it for a few hours every so often because it's doing something stupid. A little Shapeoko is totally fine for youtube video making hobby tier work, and can actually do nice work if you baby it sufficiently. I gave up in like a month and bought a Tormach.
|
|
#
?
Oct 11, 2019 04:00
|
|
|
Since the laser cutting thread is archived: any setup tips for getting good results from 2-ply engraver's plastic? I ordered thru trolase, got some 1/16" metallic foil/black core sheets, plus this 20-mil thin n flexible sheet i'm already playing with. I know to unfocus the laser a touch for raster engraving and about masking to preserve the surface finish, but I've never worked with this stuff before and have never really had to care about finish quality for my previous acrylic work.
|
|
#
?
Oct 19, 2019 02:22
|
|
|
I want to know if something exists but I can't even combine the correct words to google it correctly. I was hoping there was a way to combine a bunch of step/direction connections into one (i2c maybe?) connection with an arduino, and then 3 wires go to a 2nd arduino that breaks out that i2c connection back to all the step/direction outputs, so I can have a more easy to manage cable harness running to my robot Is this a thing that can be done? googling i2c related things just gives me results of daisy chaining multiple i2c devices, not combining multiple analog signals, transferring them as a single digital signal, and splitting back into analog signals by copy+pasting something from github
|
|
#
?
Nov 24, 2019 17:29
|
|
|
bring back old gbs posted:I want to know if something exists but I can't even combine the correct words to google it correctly. I was hoping there was a way to combine a bunch of step/direction connections into one (i2c maybe?) connection with an arduino, and then 3 wires go to a 2nd arduino that breaks out that i2c connection back to all the step/direction outputs, so I can have a more easy to manage cable harness running to my robot e: Also i2c is typically not rated for much distance if you want decent speed, depending on your machine layout. SPI or some other serial protocol might be better, pretty sure you can find io expander for all kinds of signals though. peepsalot fucked around with this message at 18:21 on Nov 24, 2019 |
|
#
?
Nov 24, 2019 18:06
|
|
|
peepsalot posted:step/dir are already digital signals, so don't know why your talking about converting back and forth with analog. You basically just need an IO expander chip on the receiving end, not a whole other arduino. Ohh that's good I thought I was plugging into the analog arduino pins but I guess it's the digital side I don't know what an IO expander chip is but that's another term I can research, thanks!
|
|
#
?
Nov 24, 2019 18:10
|
|
|
Also, at the lowest level, you basically only need a serial in, parallel out (SIPO) shift register, since you only need output and not input. But you probably want something supported by an internal peripheral device that doesn't require arduino to bang out individual bits at a time, cause that might be too slow. Not sure if UART for example can go directly to a shift register. Maybe electronics thread folks could tell you more.
|
|
#
?
Nov 24, 2019 18:26
|
|
|
I'm not sure I completely understand what you're trying to do, but it sounds like on the physical level you are talking about a data bus and on the communication level you're maybe talking about packetization or multiplexing. Maybe Google those words?
|
|
#
?
Nov 24, 2019 19:01
|
|
|
Honestly I don't really know what I'm asking for either, I'm currently building a 6 axis arm that slides on a 2060 rail, and it's a very basic design, no encoders, just stepper motors and pulleys. But I'm still in over my head. Plan A: Just to have long (6ft or so) stepper motor wires run back to my stepper drivers and have those run through a cable sheath to clean it up. Are these wire lengths acceptable? the wire I have is the same gauge that are coming out of the motors. Plan B: Then I was thinking I could locate the entire stepper driver board to the slider gantry and just have the step/dir wires and motor voltage running to the gantry But then I was thinking it would be great if I could consolidate all the step/dir wires into a single data line and break them out with another board beside my controller. The thing is, my test bed is an Arduino just to make sure my motors spin and the thing moves, but the actual control box I'll be using is a proprietary stepper controller that just has inputs for endstops and outputs for step/dir so I won't be able to modify that thing to talk to i2c the way I can with an Arduino. That's why I wanted that 2nd board to break everything out into step/dir pins again, so as far as the controller is concerned I'm just running many long wires like normal. Maybe I'm overcomplicating this and I can just run long the stepper motor wires bring back old gbs fucked around with this message at 20:11 on Nov 24, 2019 |
|
#
?
Nov 24, 2019 20:07
|
|
|
Longer stepper motor wires are what most systems do. Make sure they're shielded and the shield is grounded.
|
|
#
?
Nov 24, 2019 20:35
|
|
|
ante posted:Longer stepper motor wires are what most systems do. Make sure they're shielded and the shield is grounded. Also only ground the one end near the controller side, otherwise you can end up with some interesting (and infuriating) ground loops.
|
|
#
?
Nov 25, 2019 00:56
|
|
|
|
| # ? May 29, 2024 23:35 |
|
|
Hey thanks to everybody who chimed in to help with my signal extension issue, I was barking up the wrong tree because an i2c connection can't really be extended very far anyways, 20cm max according to the arduino forum. For now I'm mounting my stepper drivers in a case under the arm, and extending the step/dir/gnd a few feet to the controller box. I'm going to try the wire I have on hand at first and if that doesn't work I'm going to get some RJ45 breakout boards and use Cat5 cable since it just comes ready to go with twisted +- pairs.
|
|
#
?
Nov 26, 2019 14:56
|
|