#qi-hardware IRC log for Tuesday, 2014-03-25

DocScrutinizer05cut/mill plastic (acrylic)? *REALLY* dead-slow RPM or you run into trouble03:18
DocScrutinizer05even better, slow RPM and constant water cooling. I *seem* to recall a dash of detergent in cooling water is revommended, might be a delusion03:20
DocScrutinizer05wpwrak: yeah, living molten red screaming acrylic03:27
DocScrutinizer05btw who spiced the drinks?03:27
DocScrutinizer05((dash of detergent in cooling water is recommended)) might help a lot to stop any screaming03:29
DocScrutinizer05also iirc it's highly deprecated to use mineral oil based (or containing) cooling liquids for plastic cutting/milling, unlike for metal03:30
DocScrutinizer05I guess aunt google is your friend03:31
DocScrutinizer05OT: http://talk.maemo.org/showthread.php?p=1418385#post141838504:00
test1 is anyone in here?05:43
test1so librewrt = openwrt + linux-libre?05:44
test1is this correct?05:44
Action: DocScrutinizer05 wonders if erdogan and Putin are meeting every weekend to smoke crack and drink the strong Absinth, then discuss their plans how to conquer the world05:55
DocScrutinizer05maybe both could use some councilors from China, who have more expertise about building Great Firewall05:58
kyakthey should also ask USA and GB for a bit of help06:12
kyakcurrently they are leaders in establishing of total control of citizens of their states and their puppet states06:14
kyakcool, Obama makes impression that he is going to call off NSA ;)06:16
kyakDocScrutinizer05: btw, you should be more bothered with intermediate results of elections in France06:17
DocScrutinizer05Front Nationale successfully pimped up their image06:18
wpwrakDocScrutinizer05: I doubt he his mill has a liquid coolant system. and he shouldn't really need one. not sure what he's doing to his acrylic ...06:31
wpwrakkyak: the general rule seems to be that all countries try to behave roughly equally well. then one gets bold and commits some atrocity, which is promptly condemned by all the others. after an appropriate amount of time has passed, at least a couple of weeks, the rest make that atrocity their standard procedure as well. and it repeats.06:35
DocScrutinizer05(no liquid cooling system) nothing a bit of putty or even a simple plastic tub couldn't fix06:40
DocScrutinizer05a 3mm high level of water above the surface of the acrylic to mill will go a long way06:41
DocScrutinizer05you don't necessarily need a circulating coolant system06:42
wpwrakmessy :) naw, this shouldn't need any coolant. 06:42
DocScrutinizer05at 100°C water will start to boil and not allow acrylic or toolbit to heat up any further06:42
test1Does the Ben Nanonote have a GPU?06:43
DocScrutinizer05I heard for other materials they use alcohol or even liquid nitrogen as coolant bath06:43
test1Or is everything on the screen rendered by the CPU?06:43
wpwraktest1: yup, all safely done by the CPU06:44
wpwrakDocScrutinizer05: you better figure out your process well before cooling with alcohol :)06:44
DocScrutinizer05I never owned or even seen a NN, so sorry no info here :-)06:44
DocScrutinizer05wpwrak: yep, sure :-D06:45
test1wpwrak: have you heard about GPU-based paravirtualization06:45
DocScrutinizer05and keep a fire extinguisher close06:45
wpwrakliquid nitrogen ... nice but may make some things brittle. also, better make sure you have good ventilation ...06:45
test1So, I supposed the nanonote will note be susceptible to this kind of attack?06:45
DocScrutinizer05forbes, omg06:46
wpwrak(GPU-based paravirtualization) naw06:47
wpwrak(BIOS flashing attack) well, there is a ROM in the CPU, for the USB boot loader. we don't know for sure whether it is truly a ROM or perhaps Flash. but that's about the only place i could think of06:48
kyakwpwrak: i can't agree more07:01
DocScrutinizer05oh yeah, those nasty BIOS attack vectors. Particularly when abusing audio ultrasonic as hiden channel for stealth networking07:05
test1I would be cool to set up your own audio network though07:09
test1for time when you can't use the wifi card on your laptop without using its propriatary firmware xD07:10
DocScrutinizer05hmm, April 12th, 2013, 12:56 PM is date of first post on http://www.wilderssecurity.com/showthread.php?t=345273, makes me wonder...07:15
DocScrutinizer05sure, rather use the proprietary firmware in audiocard ;-P07:16
DocScrutinizer05btw fsck forbes, they refuse top show any proper content without you allowing them to place cookies on your box. Sure, the first thing I wanna do when trying to read about possible security threats is to allow cookies07:33
DocScrutinizer05s/top /to /07:33
DocScrutinizer05LOL http://forum.sysinternals.com/gpu-based-paravirtualization-rootkit-all-os-vulne_topic26706_post138481.html#13848107:36
test1DocScrutinizer05: I use Self-Destructing Cookies07:37
DocScrutinizer05self destruction cookies? nice! :-P07:37
test1DocScrutinizer05: Yup that's actually the first one I read07:37
DocScrutinizer05http://forum.sysinternals.com/gpu-based-paravirtualization-rootkit-all-os-vulne_topic26706_page1.html is already giving it away07:41
DocScrutinizer05test1: sorry, I'm not visiting sites that would require me to do silly stuff to meet the crap they want to forcefeed on me07:43
DocScrutinizer05and forbes, c'mon *really* now?07:43
test1oops :P07:44
DocScrutinizer05classical layer9 virus07:57
wpwrakone of those mysterious "an error occurred" clips08:22
wpwrakbtw, roh would also know some things about milling acrylic. he's got a big mean mill.08:23
DocScrutinizer05wpwrak: one of the worst thing with acrylic is when you use old worn out tool bits that aren't sharp anymore. And acrylic is known to cause tools to lose their sharpness quickly08:24
DocScrutinizer05seems it's pretty hard08:25
wpwrakdunno. my trusty old 3.175" mill still works fine with it. so it can't be too bad.08:28
DocScrutinizer05probably the "one size (err type) fits all (materials)" approach doesn't pan out too nicely for headmill bits and acrylic08:47
DocScrutinizer05http://www.plexiglas.de/sites/dc/Downloadcenter/Evonik/Product/PLEXIGLAS-Sheet/Verarbeitungsrichtlinien/311-5-tipps-zur-verarbeitung-von-plexiglas-de.pdf  (german, good)09:02
whitequarkwpwrak: DocScrutinizer05 ok, I think I kinda solved it09:07
whitequark1.0mm endmill, layers of 0.5mm, S18000, F1000 (yes, one thousand mm/sec)09:07
DocScrutinizer05by removing the "chips"?09:07
whitequarkI can't quite get the nice finish I want though09:07
DocScrutinizer05use water09:08
DocScrutinizer05you'll be surprised09:08
DocScrutinizer05try to get plastic drills09:10
whitequarkI should make a full-sized cover for my CNC09:10
whitequarkthen I could even maybe use flood coolant09:10
whitequarkor at least mist09:10
DocScrutinizer05meh, simply plase a few drops of water on the surface where you plan to drill a hole09:11
DocScrutinizer05a) you should use low RPM anyway. b) you're likely to use small diameter drills which won't cause the water drop to get blown away09:12
DocScrutinizer05ooh, when you mill the hole with smaller tool, make sure you always work against the spin 09:14
whitequarkthis worked really well09:14
DocScrutinizer05don't let the mill add to the forward force09:14
whitequarkfirst, chips don't accumulate on the endmill anymore, they turn into a kind of sludge and don't get in the way09:14
whitequarkand definitely don't melt09:14
whitequarksecond, the finish is much much nicer09:15
whitequarklet's try going slower09:15
DocScrutinizer05(against the spin) I'm fuzzy. When you mill out a hole with a bit that spins clockwise, then your toolpath also needs to be clockwise09:16
whitequarkyeah, I know about conventional/climb milling09:17
DocScrutinizer05the surface where the feed adds to the circumference spped of the mill will look vleaner then the surface where the both subtract09:19
DocScrutinizer05typo galore09:19
DocScrutinizer05the surface where the feed adds to the circumference speed of the mill will look cleaner than the surface where the both subtract09:20
DocScrutinizer05btw WTF? your CNC does 1m/s?09:25
DocScrutinizer05is this feed or circumference speed?09:25
DocScrutinizer05when it really does 1m/s, you better mount it to concrete of your floor with 20*300 bolts, and leave the room before starting it ;-)09:26
whitequarkmy CNC does 1900mm/min09:27
DocScrutinizer05per minute, yeah maybe09:28
whitequark31mm/s it is.09:28
DocScrutinizer05[2014-03-25 10:07:34] <whitequark> 1.0mm endmill, layers of 0.5mm, S18000, F1000 (yes, one thousand mm/sec)09:28
whitequarkright, typoed09:28
whitequarkit's just in contrast to wpwrak, which goes so painfully slow I don't think I'd be able to even watch it09:28
whitequark2mm/min is absurd09:28
DocScrutinizer05for acrylic slow feed is toxic09:29
DocScrutinizer05slow RPM, high feed09:29
whitequarkoh, /s09:29
whitequarkso he went 180mm/min09:29
DocScrutinizer05it's a bit counter intuitive, until you think about melting09:30
DocScrutinizer05slow feed causes too much heating and finally melting of the acrylic09:31
whitequarkyeah yeah, I got it already, after two dozen cuts09:31
whitequarkI'm not *that* dumb09:31
DocScrutinizer05never said you were09:31
DocScrutinizer05but water is the killer, eh?09:31
whitequarkit totally is09:31
whitequarkS8000 F1200 is really good09:32
whitequarklemme try even lower rpm09:32
DocScrutinizer05eventually your mill breaks09:32
DocScrutinizer05tricky to find the right RPM vs feed09:32
DocScrutinizer05when the ratio goes too low, your feed cracks the bit09:33
DocScrutinizer05ratio too high makes acrylic melt09:33
whitequarkthe hard part is making water stay there09:33
whitequarkand not go away09:33
whitequarkand I just broke my second to last endmill09:33
DocScrutinizer05i'm typing too slow09:34
whitequarknah, not while cutting09:34
whitequarkwhile removing melted acrylic from it09:34
DocScrutinizer05you might be able to remove molten acrylic by milling through wood09:35
DocScrutinizer05with very high RPM09:35
DocScrutinizer05and moderate, initially dead slow feed09:36
whitequarknah, I can usually just spin it off with pliers and while spinning the axle manually09:37
whitequarkS8000 F1500 was an error09:37
whitequarkI'm out of endmills09:37
Action: whitequark sighs09:37
DocScrutinizer05now you got plenty of time to think about my various suggestions to build a foce detector into your CNC09:38
DocScrutinizer05for carbide tools it won't work with the one suggested solution, since those break before any noticeable bending09:39
whitequarkbest finish I got so far (at S8000 F1000): http://i.imgur.com/kJheXqz.jpg09:51
whitequarkit's even moderately transparent09:52
wpwrakwhitequark: (painfully slow) my mm/s are mm/s ;-)10:05
wpwraklooks pretty good. don't expect too much from the mill's finishing powers. if you want it "neat", then it's sanding (many times, with decreasing grain sizes) and polishing10:07
Action: whitequark glares at http://i149.photobucket.com/albums/s57/jollytrapshooter/2012-03-12140819.jpg10:10
wpwrak(polishing) e.g., with dremel's polishing compound. 10:10
wpwrakthis is a not so nice example: http://downloads.qi-hardware.com/people/werner/anelok/tmp/case-top-polished-outside.jpg10:10
whitequarklet me try to put some paper on top of that acrylic. that should make it more hydrophilic10:11
wpwrakthe window is polished, the area around it isn't10:11
whitequarkoh and I still have my 3.175mm endmill10:11
whitequarkcould also add water in process to make sure it's in abundance10:11
wpwrakthis is a before and after comparison, again with polishing: before http://downloads.qi-hardware.com/people/werner/anelok/tmp/case-top-diffuse.jpg10:12
wpwrakand after: http://downloads.qi-hardware.com/people/werner/anelok/tmp/case-top-polished.jpg10:12
wpwrakif you have larger steps or, much worse, case halves, then you need to sand10:13
wpwrakthere are also alternatives to abrasion, e.g., fire or solvents10:13
wpwrakfor solvents, the "best" seems to be the vapour of a somewhat nasty solvent, forgot which10:14
wpwrakfor fire (i knew you'd like that :), you want a very hot flame. so traditional russian technology should serve you well10:15
wpwrakthe red stuff in the pictures is that dremel polishing goo. alas, very hard to get rid of10:16
whitequarkooooh, I just had an excellent idea for how to keep water near work area10:19
whitequarkI noticed the results are best when water turns into that kind of sludge before it's dispersed by the endmill10:19
whitequarkso, what about pouring a gel there?10:19
DocScrutinizer05that's a cause/result mesup thinking10:22
DocScrutinizer05water turns into that sludge because it gets micro-circulated by mill, and thus manages to move away the dust from the tool active surface10:23
DocScrutinizer05when you use gel, no micro circulation, dust stays where it comes into existence, and result will be way inferior10:24
whitequarkhm, good point10:26
DocScrutinizer05even to the point of gel and dust building up a nice sticky gue on toolbit surface that eventually dries and then melts 10:26
wpwraki guess the water could still get squeezed out of the gel. so i wouldn't dismiss the idea. i guess it depends a bit on how watery the gel is10:28
DocScrutinizer05when your water "rolls" away on the hydrophobic acrylic surface, then use some drops of detergent to add to the water to cure this 10:28
DocScrutinizer05and, as already suggested, nothing a bit of putty couldn't fix10:30
DocScrutinizer05make a nice ring forming a mini basin around your milling area10:31
DocScrutinizer05fill in 1 or 2 mm high of water10:31
DocScrutinizer05make sure it won't flow out of your basin too fast when you have milled in your first "sink" into the acrylic10:32
wpwrakthen convince it not to leave through the holes you mill ;-)10:32
DocScrutinizer05faster ;-P10:32
Action: wpwrak needs coffee10:33
DocScrutinizer05since your acrylic is supposed to be tightly and flush attached to a victim material below, there shouldn't be any noticeable sink path10:33
wpwraki went through the liquids ordeal when milling aluminium. there, you need lubrication. messy business.10:34
DocScrutinizer05yeah, for that you need oil or similar, a mess compared to plain water10:34
wpwrakindeed. also because the oil acts as solvent for the adhesive tape ...10:35
DocScrutinizer05usually an emulsion of oil in water gets used for that10:35
DocScrutinizer05the oily component can be rather low, like maybe 10% or lower10:36
DocScrutinizer05in german it's called bohrwasser10:37
wpwraki just used plain oil. i need it only for lubrication, not for cooling10:40
DocScrutinizer05plain oil is way more messy, and particularly for alu I guess the water even gets active in some way10:48
whitequarkyep it is... if you strip the oxide layer, aluminium is an active metal10:49
DocScrutinizer05I know. I just think it kinda "kills" or rather dissolves the alu dust10:50
whitequarkno, not that fast10:50
DocScrutinizer05at least passivates it, so you can remove it easier. Maybe10:50
DocScrutinizer05for sure plain minaral oil is way more sticky, and with the alu dust mixed in, it's a nasty polish10:51
wpwraki've heard that brass and bronce should be self-lubricating. might be better suited for playing with metals if it doesn't have to be Al10:51
DocScrutinizer05brass particularly self lubricating on steel10:52
whitequarkDocScrutinizer05: air passivates al just as well10:53
DocScrutinizer05machine builders love that combo, for bearings10:53
Action: whitequark headdesks11:18
whitequarkso I just figured out the hard way that there are exactly two kinds of shower hose thread11:18
whitequarkone with d=22.5mm, and another with d=21.5mm11:18
whitequarkapparently one is "imported" and other is "russian"11:29
whitequarkthey even looked almost same, which is why I bought wrong one in the first place. neither russian nor western packaging considers necessary to write the diameter on the packaging11:29
whitequarkand I couldn't use my caliper to measure whatever's in plastic11:29
DocScrutinizer05err, that last one doesn't parse11:31
whitequarkwell, the hose was in factory packaging.11:32
DocScrutinizer05never heard of different diameters either11:35
whitequarkwell I guess you aren't in russia!11:35
wpwrakthey're probably nominally the same :)11:35
DocScrutinizer05smells like 3.5mm plugs11:39
wpwrakjust saw this: https://www.kickstarter.com/projects/1516846343/microview-chip-sized-arduino-with-built-in-oled-di12:35
wpwrakit's depressing. a new design, based on AVR. in 2014. yet 10 times over-funded12:36
wpwrakto make anelok kickstarterable, i think i should say somewhere "Based on 32-bit ARM core, like <B><BLINK><DANCE>Arduino</DANCE></BLINK></B> Due."12:36
larscit has a OLED!!! and built-in, imagine the things you could do with that12:36
larscI mean the blink pattern possibilities are endless12:36
whitequarkthat's depressing, yes.12:38
whitequarkcircle! rectangle! wow    so 1st grade geometry     such compexity12:39
wpwrakit's like having 3072 individual LEDs. you can play with a different one ever hours and it'll still last a whole year !12:39
wpwraki like their comment on "The huge difference between our 3D Printed and CNC Milled Prototypes." though12:42
wpwrakwhitequark: note what they have to say about polishing :) "Each prototype enclosure took 8 hours to polish by hand! You really need a beer after that!"12:44
whitequark"handcrafted fresh green electrons delivered by you from local sustainable hackerspaces"12:47
whitequarksomething along these lines12:47
viricare you laughing of people who know how to make money? :)12:50
whitequarkviric: I am. I never considered money to be the pinnacle of personal achievement12:51
whitequarkI mean, by this measure Putin & co are best of the humanity.12:51
viricIt requires a mixture of ignorance and malice to make money12:52
viricbut then there is the paradox of knowing how to make money, just to outperform evil people, and then invest the money on things you'd like to grow12:53
whitequarkI won't agree here either, you absolutely can make money on a good cause12:53
wpwrakkeep things simple: 1) define "a good cause" as "me making money". 2) start making money13:20
DocScrutinizer05thanks for the advice. will do it this way with my next project. After I recovered from this one14:06
zrafawpwrak: swd tests.. we have not done more tests, just last Friday. I have really bad news from tests, so I think that I should write an email to tell you the current story, so others benefit as well if that problem appear for others of course.16:46
zrafawpwrak: But, in short: it has not worked yet :(16:46
zrafawpwrak: worse, the last two boards we did are okey, and I see some 3.3V from ben16:47
zrafa+ubb pins so I guess that jlime is not the issue ( I disabled jz mmc driver of course)16:47
zrafaThen, I think board with mkl26 is okey, and ben+ubb with jlime is working with your swd software as well. But, still nothing from MCU.16:48
zrafaEvery test, every reset, swd werner software and we always get ACK 7 as answer from MCU16:48
zrafaI should write an email = I should write an email to qi hw mailing list I meant16:49
whitequarkokay... so I jot jigsaw blades16:54
whitequarklemme try to (not) saw off a finger16:54
wpwrakwhitequark: pro tip: drink a bit of vodka first. that way, if you miss, it will hurt less.16:55
whitequarksure, brb16:55
whitequarkwpwrak: in general, how does one select the blade?16:55
wpwrakpro tip #2: don't drink the entire bottle, so you'll have some desinfectant ready16:55
whitequarkhere I have one for 2-15mm, another for 3-30mm16:55
whitequarkwhat one do I use for 3mm plywood?16:55
wpwrakguess both should work. i don't really know the magic but i'd pick the one with the smallest teeth. 16:56
wpwrakzrafa: sounds suckish :-( did you try with the scope ?16:57
whitequarkwpwrak: guessed as well16:57
zrafawpwrak: no, because I do not have skills for that. And Rodolfo is very busy everyday to give me help. Maybe the electronic engineering student who is working with us could try16:59
wpwrakzrafa: grab him ! :)17:00
zrafawpwrak: btw, we just power the MCU with pin 3 (vdd) and pin4 gnd. Does it need more vdd pins connected? 17:00
zrafa(for example, pin 30 and 48 are vdd as well)17:00
wpwrakyou should always connect all the power pins17:00
wpwraknot sure what happens in this case if you don't. what package do you have ?17:01
zrafamkl26z256vlh4 LQFP/QFN (64 pins)17:02
zrafawpwrak: and you suggest to connect all the vss as well?17:03
wpwrakof course, yes. and don't forget the decoupling caps17:04
wpwrakwhile you're at it, you'll also want to add a USB connector17:05
whitequarkthis was surprigingly painless17:10
whitequarkI mean, my fingers are even mostly intact17:12
wpwrakoh, then you didn't pass yet17:12
zrafawpwrak: decoupling caps.. : reading wikipedia now... :)17:13
wpwrakif you omit decoupling caps, some of the internal rails could collapse on sudden load changes. not sure how picky the chip is, though.17:13
wpwrakyou basically have to put 1 uF between each VDD/VDD pair. they're next to each other for that reason17:15
wpwrakand place the caps close to the pins, not centimeters way.17:16
wpwraksame treatment for VDDA/VSSA and VREFH/VREFL17:16
zrafaargh.. so we need as well vdda/vssa connected as well??! I wanted to avoid those to save work :)17:20
whitequarkconnecting all supply pins is pretty important17:20
whitequarkyou'd get really magical behavior otherwise17:20
wpwrakyeah, connect them as well. they may not really be needed but it's better to be safe than sorry17:20
wpwrakalso, equip the regulator. that way, you can power the board from USB.17:21
wpwraksee anelok for all the caps and connections17:21
zrafawpwrak: ok. We were trying using vdd from ben+ubb17:22
zrafato power17:22
wpwrakyour 64 pin package has one VDD/VSS pair more, so you'll need an extra 1 uF cap. but the rest should be the same as with anelok's MKL25Z128VFT417:23
zrafawpwrak: yes, I see those. Thanks a lot for all the tips, we will try to follow all of them for the next try ;)17:26
wpwrakhmm, my circuit is also a little dodgy. the way i mix USB and battery power isn't entirely clean17:31
whitequarkwpwrak: I think there are dedicated switches?17:32
wpwrakwhitequark: you mean chips ? or switches in the MCU ?17:34
wpwrakwell, i'd call them "FET" :)17:35
wpwraki actually have one. just not doing that sort of thing.17:35
whitequarkwell... it's like, it already has convenient outputs for detecting power switches and connects right to the typical battery charger IC17:35
whitequarknot strictly necessary but can be convenient and/or takes less place than discrete stuff17:36
wpwrakah, but i use primary cells17:36
wpwrakno fancy charger neenen17:36
whitequarkno liion? :/17:36
wpwrakCR2032. all the Li you need :)17:37
whitequarkhm, I think you can recharge CR**** cells17:38
whitequarkie, they're not actually primary. they're more like rechargeable fuel cells17:38
whitequarkwait, no, that doesn't make sense17:38
whitequarkit does, they're not *fuel* cells. methanol there is just as ion transfer medium17:39
whitequarkI think17:39
whitequarkwpwrak: don't PC motherboards actually recharge CR2032 RTC cells?17:40
whitequarkpretty sure they doi17:40
wpwraki wouldn't expect them to. they may draw no current while the system has other power, though.17:42
DocScrutinizer05they definitely don't recharge LiIon primary bupbat17:43
DocScrutinizer05that's maximum forbidden 17:44
wpwrakzrafa: anelok's VSYS should be the same as VBAT. i.e., connect VOUT33 to all the VDD/VDDA/VREFH pins and also connect the ben 3.3 V to that net17:44
whitequarkDocScrutinizer05: seems like true17:44
whitequarkinteresting. so this is a myth. seems quite prevalent in RU though17:45
wpwrakmaybe un RU they do ... to regenerate the Pu, perhaps ? :)17:45
whitequarkwpwrak: btw, metal blades cut acrylic just fine17:45
wpwrakno Li-ion, just Pu-ion17:45
wpwrakexcellent !17:46
whitequarkthe finish is even moderately nice17:46
DocScrutinizer05pretty simple: power up you rPC, remove CR2032, prbe between + and - of bat holder with ampere meter (ramge 10..100mA)17:46
whitequarknooope, I only have this PC I bought used. I think it wasn't cleaned in maybe 5 years17:46
whitequarkI'm really just afraid to open it17:46
whitequarkit's better to not know sometimes17:46
DocScrutinizer05whitequark: re sawing acrylic, see the links i posted this morning17:48
wpwrakzrafa: maybe put your design on qu-hw or github. that way, we can check it17:51
whitequarkso I wrote the local endmill supplier asking them to sell me some endmills for acrylic18:07
whitequarkhopefully they have a better idea of proper feeds&speeds than me18:07
whitequarkthey better do, for the amounts they charge18:08
Action: whitequark looks at eighteen acrylic circles he now posesses18:18
whitequarkwonder if I can do something with them18:18
whitequarkapart from "burn"18:18
wpwrakmaybe you can use them as feet/spacers some day18:24
whitequarkhm, burned one18:28
whitequarkfunny, it doesn't really emit any soot, only slightly stinks, and eventually burns down to nothing18:28
wpwrakthe perfect crime18:29
whitequarkalso, apparently pmma combustion products aren't even toxic18:30
whitequarkthis is so boring18:30
DocScrutinizer05not at all. acrylic afaik is a component for industrial scale rocket solid fuel18:47
DocScrutinizer05oh yes, there been that youtube video, somebody used a acrylic cylinder with a hole along the axis, through which he feeds some gas, prolly oxygen18:49
DocScrutinizer05>>every home should just have one of those. It's essential!<< LOL18:53
DocScrutinizer05the latter one shows how to drill acrylic the right way18:55
DocScrutinizer05it creates one (or 2) long curl of acrylic 18:56
DocScrutinizer05however I think the design like shown in this videos isn't useful yet. You at least need to fix two things: enclose the acrylic into a proper containment that is able to handle the pressure in burning chamber even when the acrylic got hot, and a "constant current" regulation for the the oxygen. The constant pressure regulation is useless19:07
DocScrutinizer05you easily can change the ordinary constant pressure regulator into a constant "current" one, by feeding back the pressure after the valve you place after that regulator (to adjust the volume of gas passing trough) to the "environment" pressure inlet of the regulator. Simple way to do that: put the whole regulator+valve into a chamber to which the valve's outlet feeds the gas. Then from that chamber you feed the Oxygen to the actual 19:12
whitequarkyep, very cool design19:15
whitequark(drilling) well, it's really a very different setup19:16
DocScrutinizer05sure, but it shows how proper drilling looks like, regarding the material taken away19:16
whitequarkI know :)19:19
whitequarkDocScrutinizer05: note though that he uses extruded acrylic and not cast19:21
whitequarkthat may have some effect19:21
--- Wed Mar 26 201400:00

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