Self-serve ads are available again! Check out our new advertising page here.
Thread punching edition.Thread hymn.https://www.youtube.com/watch?v=MebuKbDuVew [Embed]Last thread: >>2015219>Haas automation videos.https://www.youtube.com/user/haasautomation/playlists>Titans of CNChttps://www.youtube.com/user/titanamericanbuilt/playlists>Edge precisionhttps://www.youtube.com/watch?v=E-kgF0sJFno [Embed]N.e.w.>Setting up a grindall spindex on a surface grinderhttps://www.youtube.com/watch?v=J9rcyIRTfz0 [Embed]>Precision ground toolroom stoneshttps://www.youtube.com/watch?v=DVLXsq7pi9Y [Embed]>narrow pallet jacks, internal barcode work order system and morehttps://www.youtube.com/watch?v=ooKIIwbkw1E [Embed]
>>2033061So the hole gets the thread done but also a second high pitch helix cut in it?Isn't this undesirable?
>>2033153yes, it has a helix broached into it, two of them actually.the near-straight helix of the flutes punch a line down each side as it is inserted while slowly rotating to match the helix angle, then it reverses direction and cuts the entire thread in the first rotation or so, then speeds out of the hole quickly.For most things it probably won't matter if it has a helix punched into it.
>>2033157that can't be good for the spindle can it? I feel like this method is just asking for a broken spindle crash... you are effectively plunging into the workpiece to full depth, whereas atleast with using a tap cycle you might be able to stop it in time...If you are doing lots of tapped holes, why not use Tapmatic, or other rigid tapping heads?https://www.youtube.com/watch?v=cm2gZh5k4F0they've had this stuff for a while it looks like, and there's no second pitch helix cut, or extra strain on the spindle.
>>2033173it's not going to rapid to the bottom of the hole like it does with a threadmill because it has to synchronize the first rotation with the helix angle of the flutes
>>2033173Because those aren't used for production stuff like automotive machining apparently, there must be a reason they don't use them.I guess they would have to buy one for every tap in the machine and a large machining center could have like 300 tools.
>>2032718what is it even doing in position 3?
>>2033317threading or some shit with the threader removed for video
I think the best tip I learned today is that you can measure a radius by setting an indicator to a ring gauge diameter, then run the indicator along the radius to get a comparative measurementpretty awesome from Gotteswinterhttps://www.youtube.com/watch?v=dMKLZEgh_SU
if you have a bolt that bottoms out at the same time that the head hits the surface which way do the threads stretch?
How exactly would you indicate in a drill chuck/ER holder on a QCTP?Not my pic, but same idea. >AXA QCTP -> MT2 Boring Bar holder -> MT2 ER collet chuckI need to use it to bore 4" long 3/16" diameter holes in brass, I cannot seem to get it on center. This is just in the garage on a small CNC, learning as I go. In what operation would you indicate the X travel or the rotation of the tool post or the tool height, and what surfaces would you indicate?The actual drill bit is fucking flimsy
>>2034460>ow exactly would you indicate in a drill chuck/ER holder on a QCTP?Stick a noga on the lathe chuck face and sweep whatever you have in the drill chuck, like the picture shows.
>>2034460>>2034471actually a travel indicator on a magbase works the best for that, there is just less droop due to gravity.a nogo probably isn't strong enough to use like that, it will deflect and fuck up your readings.Sucks but it's true.
>>2034475>nogo probably isn't strong enough to use like that, it will deflect and fuck up your readings.I used to do this daily on a CNC lathe indicating in drills, If you have a decent large noga it's not a problem indicating in tenths.
>>2034486which one? I have a couple and I'm sure one of them deflects too much.
>>2034510It's one of their heavy duty ones.https://www.amazon.com/NOGA-Heavy-Duty-Holder-Magnetic/dp/B001VY1DYQ/ref=sr_1_30?dchild=1&keywords=noga+tool&qid=1613663047&sr=8-30
>>2034486>>2034475Well I have a real Noga im not sure which model it is.Ive done this and the rotation of the QCTP is still off.Now that im thinking about it, I should be able to sweep the side of the tool to get the rotation square, and then indicating around the tool will give me the height and the X axis?
>>2034536I actually have one of those, but I don't think that's the one I used for sweeping around.I'll try that one later today. Shop is heating up and I'm ready for work.
>>2034460https://www.youtube.com/watch?v=ztTQC1bkuzAThis is the basics, I'm not a fan of indicating the flutes like this guy does though, either have the drill sticking out far enough where you can indicate on the shank or use a pin gauge in the drill chuck. Hopefully the video will get the point across.
4/200and snow sucks
So it's about time to automate this job a bit more, apparently this is a list of variables corresponding to the work coordinate variables on Fanuc stuff.#5201 COMMON X#5202 COMMON Y#5203 COMMON Z#5204 COMMON W#5205 COMMON B#5221 G54 X#5222 G54 Y#5223 G54 Z#5224 G54 W#5225 G54 B#5241 G55 X#5242 G55 Y#5243 G55 Z#5244 G55 W#5245 G55 B#5261 G56 X#5262 G56 Y#5263 G56 Z#5264 G56 W#5265 G56 B#5281 G57 X#5282 G57 Y#5283 G57 Z#5284 G57 W#5285 G57 B#5301 G58 X#5302 G58 Y#5303 G58 Z#5304 G58 W#5305 G58 B#5321 G59 X#5322 G59 Y#5323 G59 Z#5324 G59 W#5325 G59 B
>>2035091ahhh here we gohttps://cnc.uno/training/en/macro.php
Hot opinions on a pocketNC for home machining and small parts manufacture?
>>2035244It's a lot of money for very little capability
>>2035244>>2035332yes very true, there's just not a silver bullet for manufacturing things cheaply unless you have an army of chinese people with files and drills.The only thing I would say fits the mold is something like a markforged 3d printer except reasonably priced. Carbon fiber reinforced parts are actually usable sometimes.
https://tv.gab.com/channel/subarctic_50/view/every-workshop-needs-this-tool-tap-602718e87827503d6fc105fchmm large heavy duty tap guide for those manual machine boys
why am I so slow? it took me like 2 hours to make 4 dead simple parts with only 7 ops and 10 tools.
>>2035930that's normal though
This is pretty brilliant.https://www.instagram.com/p/CLfsBpzppJO/
Can milling be done with a countersink bit?
>>2036379countersinks usually have a reduced shank diameter and therefore are not a very rigid toolit will probably bend on you if you try it, unless it's a carbide countersink or a center drill or something.
Well shit, I need to investigate updating the G54 Z zero offset on Fanuc 16TB more, first attempt didn't work out.
woah check out this millturn, it's a 5-axis but the table rotates and it has a tailstock at the 90 degree rotation positionpretty damn coolhttps://www.youtube.com/watch?v=LlK5CZJ2oiQ
Post sick looking cuttershttps://www.youtube.com/watch?v=05AYsCzKz8U
Is there any reason not to use G52 for changing Z zero in looping the program?https://www.mmsonline.com/columns/shifting-program-zero-on-machining-centers
>>2037388I think G52 is going to work.
>>2037430Reeee it's not incremental so I have to calculate the value to stepover each time.
>>2037388>>2037479Why wouldn't you subprogram, incremental, and loop call the sub?
is it possible use autoCAD without having Internet Explorer 11? I need to use AutoCAD on my home PC because of wu flu but it won't let me use it until I download IE 11. Internet Explorer is a truly cancerous program so I would prefer to avoid having it even installed on my PC
>>2037551I don't know how to program gcode. Why would I do all that when I can just compute the stepover and set G52?
>>2037551Also since I have someone that can program Gcode, what is the problem with the syntax::1005 (BACKPLATE TURRET1 PROG)G0 G20 G40 G80 G96 G99#1 = 0.0 (loop count)#2 = 3.0 (Max iterations)#3 = 0.262 (stepover CONST)WHILE[#2 GT #1] DO1#1 = #1 + 1...G52 Z - #3*#1 (Z zero update) <--- likely problem syntaxEND1G52 Z0.0M5M9G0 G28 U0.G0 G28 W0.M46 (C AXIS JOINT RELEASE)G54M30
>>2037565there are concessions you must make to have business software, mostly your soul.
>>2037593>>2037595Hard to give any advice when I don't know what you're doing, what you want to do, what CAM program you/your programmer are using, what controller it is, and why your programmer can't come up with something that works.
>>2037669I am the setup guy, operator, programmer and business owner.It's a Fanuc 16TB lathe, Fusion 360 and a custom post for an old ass machine.I am inserting manual-NC into the program so I can just post it every time and have the custom code added in.It's a 5" diameter bar that sits stationary, no bar pulling. I want to machine the first part off the end of the bar, then the next part, etc.I figure I can safely run 5 parts at a time (I am using the sub-spindle to catch the parts as a I part them off on a 1/2" steel rod). 5 parts will buy me around half an hour of un-attended machining.
>>2037674>>2037676and sometime later, more than half an hour of unattended machining.basically totally automated for a few hours per bar.
>>2037674To be honest if you want to try G52, you have an easy way of doing it. Run machine with no parts in it at lowest speed lowest rapid and see what happens. I've used G52 once just to offset a particular toolpath without having to use subs or incremental, but I would disable it immediately with G51 (IIRC, it's been a while) because offsetting your shit and forgetting about it gets crunchy fast.
>>2037678Yeah but the code is wrong, control complains about this line:G52 Z - #3*#1 (Z zero update) <--- likely problem syntax
>>2037682I really haven't done anything like that, I don't use fusion at all. From what I can see of your requirements of just parting-off and moving over for another part-off, I could probably just type the whole fucking thing up by hand faster and more easily than messing with G52 offset stuff that uses syntax. No idea how that stuff works.It would be tool change, tool offset call, work offset call, spindle rpm, coolant if necessary, move to first location, sub program call. Incremental, type out the part-off movements (literally G1 xyzf), back to absolute, exit sub.move in z, call sub.move in z, call sub.home, end program.
>>2037692Yeah I could do that in CAM, but the program would be larger.I do the milling/drilling on the part in the lathe as well, so the parting is interrupted and I need a good surface finish on it. It's a thin ass parting blade sticking out about 2.5" so it takes forever to part off. Don't want to use a thicker parting blade because brass is expensive.
>>2037693Look into a Y-axis parting tool?The thing about an incremental sub is that it's very small, just the movements required for the parting, and all you do is move over and call it again. Example.(main)Z-1.M98P003Z-1.M98P003Z-1.M98P003(sub)003G91[parting movement for one parting operation]G90M99If you want it to take 5 slices, you move and call 5 times. Pretty sure the old controller has enough memory for like five more lines of code, right?
>>2037699each side of the control has 500kb, but what I would have to do (those holes are countersunk for flat head screws) is line up 3 part models with the right offset from each other in CAM and do the toolpaths for each part just ignoring the simulation thinking it's running into the part it just parted off.It would work, but it's not really that great compared to just making the fucking loop work with G52.
>>2037705(prog size would be 3x or 5x what it is now)
>>2037705>it's not really that great compared to just making the fucking loop work with G52.Are you even sure the controller can do it? Sorry I can't help you more.
>>2037713Well yeah I mean I think it can.It should be able to do a simple multiplication of two varaibles afaik. It worked with G52 just fine for the first two pieces, but I was just setting G52's Z to -0.262 every time so it won't go beyond the second piece without a larger offset.Yes I know I could pre-program the offsets and just do that but that's fairly silly when I could have it do a simple multiplication and make the code more re-usable and easier to modify.For instance, if I want the stepover to now be 0.272 I would only have to modify the stepover variable at the top instead of all the pre-computed stepovers for each part.
>>2037724at some point you gotta ask yourself if you're saving time by spending time trying to get it to work versus just doing it a bit quicker and a bit dirtier. You're talking about parting what, 5-ish parts at most? Is changing 5 variables that much trouble?More trouble than all this discourse?
>>2037728Well it is also important that I learn the proper syntax for doing what I want in the program.I have a post up on practicalmachinist about it, hopefully someone knows some gcode stuff there.But why settle for quick and dirty? I actually have plenty of time on this job to get it done right, and I will see this job again and again, I may end up wanting to part off 10 of them at a time and all those variables would just be annoying.
>>2037674You got any posts for a OT-F machine?
>>2037810nah youd have to talk to a post provider like nexgencam
PM came through maybeYour current following BlockG52 Z - #3*#1needs to be written as followsG52 Z-[#3*#1]However, its better practice to express the DOC ((step-over CONST) with direction (+ or -). Accordingly, had:#3 = 0.262been expressed as#3 = -0.262then the G52 Block would be written as follows:G52 Z[#3*#1]Regards,Bill
>>2037872so fucking awesome, now I can dick off all day even harder while making money
96/200 time to do other stuff for a while
How do I setup a dial indicator properly for checking pinion backlash or zeroing out a differential? What other measuring tools would I need?
>>2038538backlash? i don't know, on the end of the bar?is it a test indicator or the kind with a plunger on it?
well shit, need some hydrualic fluid before I can continue
is there a decent online toolstore for us home shops? mscdirect's online catalog is dogshit, and the book isn't much better. can't find a good lathe 45* tool anywhere and i've bought most of my mitutoyo stuff off ebay
>>2038937travers, micro100, one of harvey/helical's distributors for their tools, carbidedepot
looks like american machine and gear is making metric change gear sets for old ass south bend lathespretty cool if you have an old south bend
>>2038937>bought most of my mitutoyo stuff off ebayBe careful, there is a lot of fake Mitutoyo stuff on ebay.
>THAT FEEL WHEN YOU FUCK UP A $400 PIECE OF BRASSsheeeit maybe it's time to finally make a trip to the scrapyard and make some of it back
>>2038913I talked to a guy that hand engraved $6000 shotguns. I told him I engraved something with cnc one time and he basically called me a pleb.>>2038937what's wrong with the online catalog?
>>2039472>what's wrong with the online catalogpoorly designed with too many disjointed subcategories. inserted endmills are not in the milling>endmill catagories, oh no you have to go to a specific indexable section amd click through vaguely defined subcategories to find the real one. Just a bunch of little shit online retailers have ironed out over the years but would be a monumental undertaking trying to reign in that beast.
I have heard you can model dxfs and then convert to program for conversational controls. Can anyone explain this a bit better for me? Pic is my controller. I have programmed dxfs before, and I have longhand programmed this controller.
damn I spent all afternoon trying to reverse engineer a screw. I knew it was finer pitch than 40tpi but I didn't have a gauge that went past that and no optical comparator to measure it. I tried holding it up to a scale and counting threads over a set distance but it didn't work. then I just started cutting test threads to see if it would match starting at 56 and going down to 48 before I realized it was metric. fuck metric.
>>2040174yeah it can be a bitch sometimes, I should probably order some new screws for my tool blocks but fuck if I remember what it needs
see this on lathewat do
>>2040760Chew her out for having long hair not tied up, and no safety glasses
>get engineering degree>china coof, no one is hiring>search for any job>figure machine operator will be good to strengthen my knowledge>hired onDAY 1>orientation BS for 5 hours>get put on a machine, ask supervisor questions to learn fast>can tell i'm asking too much, back off for a bitDAY 2>get in, run machine, run 800pc/hr>machine starts throwing parts out everywhere>wire spool slams into machine>producion manager comes over, points out that kill wire was still cut, maintenance didn't do a job properly>put on new machine, NOGOs out of the gate>make boxes while they fixing everything>hope it isn't always like thisDAY 3>machine 1 up again, run all day>2,600pc/hr average for 8 hours>production manager asks how I'm doing>"i'm alright, i cranked out about 20k parts today, should be better tomorrow>he stares at me blankly, tells me no one has ever produced nearly that many in a day>break open a few boxes, all specs are good because i check everything>he sends me home stunnedDAY 4>13k parts in morning half>lunch>come back, parts ejecting everywhere, get supervisor>ejector cam pushing part before work holder releases>supervisor and i observe and adjust for 2 hours>prod. mngr comes over and is pissed we're not producing>he tries to fix problem with cardboard and leaves>not the problem, still shitty>end of shift, still broken, no solutionManufacturing feels so bipolar like absolutely no one knows what they are doing on the line. How do I climb up and out?
Do tolerances scale up and down with size? What are the tolerances for a crankshaft this size?
>>2040786Are you in a high volume production shop? These are dead end jobs for mexican immigrants, get away ASAP
>>2036379Yes definitely. I have done it with some success. But i think the cone type relief limits the feed (probably a good thing to avoid chatter while counter sinking) so i went slow and with a lot of oil. Came out okay but you're probably way better of with a dbit or a real chamfer mill.
made like 24 parts today, need 76 more. Pretty run down from these damn antibiotics.
>>2040786>>get engineering degree>>china coof, no one is hiringThis is how it was before the coof too. Get comfy because that degree is worth the paper it was printed on. Trust me, I know.
>>2040794a high volume machining shop with just machine tools should be using the latest and greatest flavor of the month from sandvik etc. seeking that extra 5 seconds shaved off the cycle time
>>2040786>2,600pc/hr average for 8 hours>because i check everything2600 parts per hour is one part every 1.38 seconds. It takes longer than that to reach into the bin and grab one, let alone to "check it" for anything other than its existence. It doesn't allow for a single bathroom break, you to skip one song on your mp3 player of choice, or even to itch your nose.And by the very next day, you had that whittled down to a part every 1.107 seconds. A solid 20% improvement, even with all the down time.
did joe pieczynski take down the videos where he made an ar15?
>>2040982i bet you can find it on bitchute or something, try google
>>2040982The guy has skill, but holy shit is he a faggot who is grating to listen to
>>2041078I've been watching his videos while I pretend to be doing stuff at work. he reminds me of the tool and die boomers who taught me machining tricks at school.
https://youtu.be/yy0c8uqj9M8?t=47>only $130k>can probably snag one used for around $100k
>>2041133wtf a DMU-50 is only $100k?https://mmi-direct.com/machine/used-5-axis-machining-center/dmg-mori-dmu-50-ecol-1