#qi-hardware IRC log for Friday, 2014-03-28

whitequarkwow, coolant is so awesome01:17
whitequarkI decided I'll just fucking flood the entire area, and that worked perfectly01:17
whitequarkI have a perfect mirror finish and only broken one endmill, and even that one because of stupidity01:18
DocScrutinizer51so you did my basin suggestion?01:27
whitequarknope, simply poured water all over that01:29
DocScrutinizer51btw it's not just cooling, it also acts as lubricant and transports away the dust01:29
whitequarkone downside, now everything is covered with sludge and/or water01:29
whitequarkand given that my table clamps and table are wooden...01:29
whitequarkwell, victim table01:30
DocScrutinizer51wax goes a long way01:30
whitequarkdon't have any01:30
DocScrutinizer51parafin oil01:31
whitequarkneither that01:31
whitequarkI think I have earwax but probably not enough to make a basin01:31
DocScrutinizer51or simply parafin (candle)01:31
whitequarkwhy the hell would I have any candles? it's 201401:31
DocScrutinizer51not for basin, to seal wood01:32
whitequarkyes, good idea01:32
whitequarkfuck. no. I did not figure it out.01:45
whitequarkanother broken endmill.01:45
whitequarkfucking shit.01:46
DocScrutinizer05it just occurred to me that the lateral force on the endmill bit is identical to the lateral force on the PCB01:58
DocScrutinizer05when your clamps are a bit loose, it might move away PCB before endmill breaks01:59
DocScrutinizer05if you even had sonsors there, dang nice stuff you could do01:59
DocScrutinizer05350 devices, 400 donors, 7355502:01
DocScrutinizer05aesthetic numbers02:01
DocScrutinizer05though I wonder how the HECK somebody managed to donate 5 EUR02:02
wpwraka diffusely symbolic token of appreciation :)02:12
wpwrakwhat was the largest donation you received ?02:13
wpwrak(endmill broken by stupidity) yeah, that's a common cause of death there ;-)02:14
wpwrak(sensors) i don't think you want your clamps loose enough for anything like that to happen. loose workpieces give you all sorts of exciting new problems02:19
wpwrakin any case, i'd try to work with lower feed rates. at those breakneck speeds, it's no surprise the endmills die like files02:20
whitequarksooo... it seems the main reason is that at first pass, water simply disperses very quickly from surface of acrylic02:32
whitequarkand later it tends to stay there02:32
whitequarkso I changed initial depth to .5mm and pour a LOT of water right on endmill at all times while milling the initial groove02:32
whitequarkthankfully there is nothing in my cnc that can rust!02:33
wpwrakif you want to go the water route, you may indeed want to think of using a basin02:33
wpwrakhah, final last words :)02:33
wpwrakthe wood will of course be useless in a basin, but you could just use acrylic as victim material02:34
whitequarkno, I just placed a lot of tissues under the machine02:37
wpwraklet's hope you're right about there being no parts that could rust :)02:38
whitequarkbtw, it's not like I *want* to use those breakneck feeds02:50
whitequarkI have to, kinda02:50
wpwrakobsessive-compulsive ? ;-)02:53
whitequarkno, I mean at lower feeds acrylic really tends to accumulate on the endmill02:54
wpwrakeven with all the water ?02:55
wpwrakwith my mill, most of it is reduced to powder. some strings may remain, but they just fall off after a while and don't get in the way02:56
whitequarkdo you use water?02:59
wpwraknaturally humidified air :)02:59
whitequarknaturally ?03:00
wpwrakthe air in buenos aires is usually fairly humid03:01
wpwrakmy powder is a bit more coarse. i don't think i have a picture from the middle of a run. have to take one the next time i mill something.03:14
whitequarkwpwrak: so I settled on 7000rpm 560mm/min for 1.2mm endmill03:55
whitequark(I broke all the smaller ones)03:55
whitequarkit doesn't stress the endmill too much and leaves a really nice finish03:55
DocScrutinizer05(diffusely symbolic token) the funny part in it: you can only select multiples of 10.-08:54
DocScrutinizer05(largest donation) 4k iirc08:55
DocScrutinizer05(want clamps loose enough) this wasn't a suggestion but a assumption to develop a suggestion on08:56
DocScrutinizer05(basin) you know those vacuum cleaners that have a combined vacuum and waterjet nozzle? does that give you some idea for alternatives to basin and lots of water?09:04
DocScrutinizer05what became of your whistler tool?09:05
DocScrutinizer05is that a vacuum that can cope with water?09:05
DocScrutinizer05sth like this: http://www.certeo.de/nasssauger-und-industriesauger/kaercher-nass-und-trockensauger-nt-27-1-me-adv-m12355.html  ?09:07
DocScrutinizer05of course you as well could use a huge bottle with two pipes for in and out of air, to collect the water _outside_ of the vacuum cleaner09:08
DocScrutinizer05then use sth like an medical infusion bag http://german.infusion-bags.com/china-comfortable_1000ml_medical_grade_pvc_infusion_bag_easy_hanging_dual_port_bag-924616.html09:10
DocScrutinizer05fill it with water09:10
DocScrutinizer05mount a 1mm syringe needle to the end of tube/hose and attach that to the spindle head/motor so it adds a constant tiny jet of water on the endmill, attach the vacuum hose next to it, so it will instantly suck up all the water from your acrylic surface - incl the acrylic dust09:12
DocScrutinizer05you probably even could cut the end of the vaccum hose at a 45° angle so the end can hoover 1mm above the acrylic surface and the hose rises at 45° form surface. Then cut an aperture into the hose exactly above center of the ellipsoid footprint, where you insert the endmill and syringe needle09:16
DocScrutinizer05the air flow inside that vacuum hose end will be perfect :-) some air from top along spindle axis, to blow away the water and slurry. And most air from circumference of hose footprint along acr. surface and going straigt up into the hose, taking all the water and slurry with it09:18
DocScrutinizer05also the downward air along spindle axis will avoid any water creeping up the tool bit / spindle to the motor09:20
DocScrutinizer05sounds like a mad useful extension for your CNC09:21
whitequarkDocScrutinizer05: yes, that was the idea11:26
whitequarkexcept for a first proto, I'll just make a small reservoir for collecting used coolant11:26
whitequarkwell... by the way11:26
whitequarkI bought a centrifugal pump, it has a DC motor11:26
whitequarkhow do you regulate power of such beast?11:26
wpwrakmilitary-grade resistors ?12:20
eintopfwpwrak: do you are a expert in workqueue? :(12:50
eintopfand I mean linux workqueues12:51
wpwrakhmm, what does an expert "in workqueues" do ? :) i have used them in the past. i don't use them often enough that i would get the API right from memory.13:42
whitequarkwpwrak: no, I said "regulate"15:45
whitequarkmeaning, what do I do? limit voltage? limit current?15:46
whitequarkcould also just plug it in reverse (to somewhat reduce pumping power)15:47
wpwrakboth ? :) but yes, i'd adjust voltage and let current follow15:48
whitequarkhmm. so, an LM31715:48
wpwrakreverse thrust also sounds interesting, though. please make sure you include the audio track when you make a video of that :)15:48
whitequarkaudio track? what's about it?15:49
whitequarkI've did reverse trust trick on this specific pump in the past, it's nothing special really15:49
whitequarkit just works worse, and in this case that's exactly what I want15:49
wpwrakoh, you mean constant reversal15:49
wpwrak(lm317) maybe you could also cook something with a PWM. not sure. that's more DocScrutinizer05's domain.15:50
whitequarkwell, I have LM317 on hand and don't care about power so...15:50
whitequarkthough, going from 24V to ~10V would dissipate some massive amount of heat15:50
whitequarkbtw, I bought sulfuric acid15:52
wpwraksteadily working your way up towards HCl :)15:53
whitequarkmeh no15:54
whitequarkI decided I will still etch with persulphate, but add a fishtank heater and aerator to it15:54
wpwraki'm already looking forward to the euphoric outburst when you finally get drunk enough to dare to try it ;-)15:54
whitequarkat 50°C persulphate should work very well15:54
wpwrakhow to put a nail into the wall: method #1: place pointy end on wall, take hammer, hit non-pointy end a few times. method #2: fix nail with structure that's independent of building, lift building with crane, repeatedly move building sideways to make it bump into nail, release nail from fixture, return building to original position.15:57
wpwrakwhy choose method #2 ? because you don't risk hitting your fingers with the hammer15:58
whitequarkremind me again how do I avoid corrosion while storing and using HCl in a 20sq.m flat16:00
wpwrakbtw, i think the next nobel prize for literature should go to gcc. i asked it to make a little gpio-polling loop. instead, it wrote an epic poem about redundancy, inefficiency, and bloat ...16:00
whitequarkwith no balcony or a ventilation slot to attach to16:00
wpwrakhow about using two plastic containers ? one opening down, the other opening up. ventilate on open window from time to time to avoid excessive cl buildup16:01
whitequarkconstant maintenance or acidic corrosion, awesome choice16:02
wpwrak(containers) the "airtight" type16:02
whitequarkalso do you know that HCl eats through plastic?16:02
wpwrakwell, you'll probably access it often enough that the maintenance will be quite implicit16:02
whitequarkwell, not technically "eats", it extracts the plastifier basically, and the base matrix becomes extremely brittle16:03
wpwrakwell, buy only a quantity you can reasonably expect to use within the next, say, 5 years16:03
whitequarklast time it ate through bottle in 116:04
wpwrakget a real bottle :)16:04
wpwrakmaybe ask a chemist how they do it. they're probably a simpler way than what i'm describing16:05
zrafaor watch breaking bad16:05
zrafaand get ideas from there :)16:06
wpwrakyeah, it may have "dual" uses when getting unexpected visitors16:07
DocScrutinizer05(lm317) DC-motors you operate like buck converter, with FET based PWM and clamp diode17:37
DocScrutinizer05HCL storage - baby simple: give it some victim material, probably weak NaOH solution enclosed in a sealed container together with the HCl bottle works just *great*17:41
wpwrakah yes. even easier. or maybe baking soda. not as aggressive as raw NaOH.17:44
DocScrutinizer05you'll want wet phase17:46
DocScrutinizer05hi dos117:46
DocScrutinizer05or stuff like Mg or Al17:47
wpwrakisn't CaOH about as wet as NaHCO3 ?17:47
DocScrutinizer05depends on how much water you add ;-)17:47
wpwrakyeah :)17:47
wpwrakbut do you actually need to add water ? the Cl is likely to spend a lot of time buzzing around17:48
DocScrutinizer05sure, but I tell you what I don't like about backing soda: the CO2 it releases. NaOH + HCl -> NaCl + H2O17:51
wpwrakah, i see17:51
DocScrutinizer05CaOH alias chalk is probably the best17:52
wpwrakyeah, if it really gets bad, the co2 could begin to act as a buffer. though i kind doubt whitequark would get near such quantities17:52
DocScrutinizer05well Ca(OH)217:52
wpwrakyeah, sounds tame enough, including the CaCl2 you get in the end17:57
wpwrakactually ... no, wrong reaction ...17:58
wpwrakyou get an explosive ;-) Ca(ClO3)217:59
wpwrakplus some CaCl217:59
wpwrak6 Ca(OH)2 + 6 Cl2 = 5 CaCl2 + Ca(ClO3)2 + 6 H2O18:00
wpwrakyou don't believe it ?18:05
DocScrutinizer05where the heck is there any Cl2 ?18:05
wpwrakit's that what escapes from the HCl ?18:06
DocScrutinizer05HCl != Cl218:06
wpwrakwell, whether HCl or H2+Cl2 escapes, you get some CaCl2 and maybe a bit of Ca(ClO3)218:10
wpwrakneither sounds overly worrying to me. it's not as if the quantities we're talking about were particularly large18:11
DocScrutinizer05would be weird to get a pretty exlosive substance from every single vomiting on chalk18:11
wpwrakand in case of a spill, it would be just CaCl2 (and H2O)18:12
DocScrutinizer05prlly the Cro Magnon had already discovered the gunpower then18:12
wpwrakhappens to you often ? :)18:12
DocScrutinizer05actually nope, I thinke last time I puked been last millenium18:14
wpwrakbut yes, it's probably HCl and not Cl2 that escapes18:14
DocScrutinizer05due to a rotten ham on some pizza18:14
wpwrakyum ;-)18:14
eintopfmhhh, yumyum noodles...18:25
DocScrutinizer05however the feel like puking is more constant by the month recently18:25
wpwrakah, those soldering fumes ...18:26
Action: wpwrak nogs knowingly18:26
DocScrutinizer05or, according to the old german proverb: "ich kann gar nicht so viel essen wie ich kotzen moechte"18:26
wpwrakfirst finish neo900, then you can food-bounce all you want ;-)18:27
DocScrutinizer05not cause of soldering, cause of other members of my species18:27
DocScrutinizer05from an evolutionary aspect, we need another huge meteorite or a pandemic disease18:29
wpwrakMalthusian catastrophe not happening fast enough ?18:30
DocScrutinizer05whatever it is, it's already overdue18:32
DocScrutinizer05random sidenote: funny name: "Steinunn Sigurðardóttir"18:46
wpwrakah yes, those icelanders18:46
DocScrutinizer05yeah, those sons and dottiers18:47
DocScrutinizer05greatly simplifies genealogy though18:48
DocScrutinizer05I can already tell the "first" name of her father ;-)18:48
DocScrutinizer05Also you never have to wonder if that's a male or female first name :-D18:51
DocScrutinizer05annoying however gets the name plate on your family door18:52
wpwrakuse apartment numbers ?18:52
DocScrutinizer05particularly when multiple generations live in one household18:53
wpwrake.g., in argentina, nobody puts a name on the door. it's all by address18:53
DocScrutinizer05heh, I bet that has advantages18:53
DocScrutinizer05add to that a lax registration office practice, and you're quite safe from unadvertised visits18:55
DocScrutinizer05afaik in USA it's similarly lax, that's maybe one of the reasons why parking tickets are not on the owner but on the car18:56
wpwrakwell, there are those who visit robin hood style, taking from the (suspected) rich, giving to the self-identified needy. for these, name or number don't really matter. at least not the simpler ones18:59
DocScrutinizer05sure. For those I had a few nice surprises installed when I'd live where you do19:03
DocScrutinizer05you know, along the line P99 (actually not P9)19:04
wpwrakcause of death: own booby trap :)19:04
DocScrutinizer05depends on your confidence in your own acting and in your projects21:30
wpwrakyeah. i try to err on the side of caution, especially when it comes to designing deadly traps :)22:23
wpwrakand if you know Dr. Strangelove, then you can also appreciate why MAD wouldn't be an effective deterrent22:26
DocScrutinizer05sorry, the last one doesn't make sense to me, prolly because of me not really knowing "Dr. Strangelove"22:54
DocScrutinizer05I tend to design such traps with at least triple ANDed trigger, with a unusual but loud and obvious warning on every single trigger getting pulled. A similarly unusual but fool proof big emergency stop button. And no instant deadly force, rather start with something that I'm even more immune against than most others: high voltage ;-)22:57
DocScrutinizer05a very good first defense line is an electronic dog, loud and nasty, with the speaker right behind the door22:58
wpwrak(Dr. Strangelove) ah, one of the best movies ever made. very nice pitch-black humor.23:01
wpwrak(no immediate lethal force) ah, dragging out the experience. well, why not :)23:02
DocScrutinizer05(loud) probably an evil growl and some snuffle like a grisly bear is more effective than loud barking ;-)23:07
wpwrakmake a little robot arm that scratches on the door, with tungsten claws :)23:10
DocScrutinizer05180dB 8kHz is also quite effective23:11
DocScrutinizer05(robot) lol - just thought about exactly same . At a height of at least 1.50m23:11
DocScrutinizer05better 1.80m23:12
DocScrutinizer053s after the 8kHz attack, gas kicks in, all still in front of the door, when somebody tries to open it23:13
DocScrutinizer05break it23:13
wpwrakthe arm would also be dual-purpose - if they get in, those tungsten claws could teach them the error of their ways23:14
wpwraktarantino style :)23:14
DocScrutinizer05well, when somebody already got in, you can do only two things: knock them out for at least 30min, or lock them in for at least same time23:16
DocScrutinizer05several kW of focused microwave energy might do23:19
DocScrutinizer05or tasers23:19
wpwraksharks. with lasers.23:22
--- Sat Mar 29 201400:00

Generated by irclog2html.py 2.9.2 by Marius Gedminas - find it at mg.pov.lt!