#qi-hardware IRC log for Wednesday, 2014-03-05

nicksydneywpwrak: BLTE everywhere https://www.kickstarter.com/projects/192833321/pebblebee-the-most-versatile-ios-android-bluetooth?ref=category 01:19
nicksydneyseems like using a QFN chip https://s3.amazonaws.com/ksr/assets/001/636/504/6d699f10c187b7c9e148d07ffe557394_large.JPG?139241982501:20
nicksydneythe video mentioned that the product is sealed with ultrasonic waves ?? how ?01:30
nicksydneythis is cool https://www.kickstarter.com/projects/1761670738/ring-shortcut-everything01:34
nicksydneyso the gesture recognition must be processing the finger movement using accelerometer01:35
nicksydneyi presume01:35
wpwrakor a combo with .. what's the name, gyrometer ?01:44
nicksydneythis could be a nice addition to create a PCB https://www.kickstarter.com/projects/1008225922/the-microslice-a-mini-arduino-laser-cutter-and-eng?ref=footer02:01
nicksydneyahh gyrometer ..right02:01
nicksydneyneed to build one of this https://www.kickstarter.com/projects/1554904482/the-fumefan ... haven't got fume fan02:08
DocScrutinizer05freaking vimeo videos freeze my konqueror02:52
sharekhi all10:00
whitequarkwpwrak: oooooh found a cool idea10:13
sharekanyone knows how to disable the screen power off ?10:13
whitequarkas an alternative to activation with thermolysis of copper hypophosphite, you can use cramolin's GRAPHITE aerosol or its analogue10:13
sharekI found /sys/class/lcd/ili8960-lcd/lcd_power10:14
sharekthe value changes 0 or 4 when on and off10:14
shareksetting with echo 0 > lcd_power don't enable it again10:14
whitequarkshould try that idea as well10:15
whitequarksharek: setterm -powersave off10:17
sharekthanks whitequark 10:18
shareksetterm: cannot (un)set powersave mode: Invalid argument10:18
sharekwtf?  -powersave <on|vsync|hsync|powerdown|off> 10:19
sharekis in the help 8-?10:19
whitequarktry -powerdown 010:20
whitequarkwpwrak: another guy posted this http://content.foto.mail.ru/list/mial/92/i-108.jpg as a PCB he made10:21
whitequarkpretty damn impressive10:21
sharekthis don't fail the command but not re-enable the screen10:21
whitequarksharek: hmmm10:21
whitequarkno clue10:21
sharekwhitequark: no problem, thanks also10:23
sharek*thanks anyway10:23
whitequarkwpwrak: http://vrtp.ru/index.php?showtopic=16042&view=findpost&p=412939 more10:26
whitequarkDocScrutinizer05: take a look also ^10:26
wpwrak(pcbs) neat !10:56
whitequarkbut from posts it seems that this required a fair amount of twiddling with chemicals10:56
wpwraki'm not surprised :)10:57
whitequarke.g. the concentration of NaCl in plating bath must be strictly between 0.05 and 0.08 NaCl equivalent mg/l10:57
wpwrakhmm ......10:58
whitequarkwould be a bit tricky to get that precisely. I would need to get something better than my kitchen scales with 1g resolution11:00
whitequarkdeionized water is rather expensive to throw it around mindlessly11:03
wpwrakor just make at least 100 l at a time :)11:07
whitequarkwell, no, you can dilute it several times. homeopathy of sort :p11:08
wpwrak... and improve the process by adding details about degradation in long-term storage ;)11:08
whitequarkoh, some of the guys report that solution is stable over 8 years11:08
wpwrakso you'd have to use 12 l per year. sounds feasible.11:17
DocScrutinizer05Graphite Aerosol - seems I mentioned that method a few days ago11:20
whitequarkDocScrutinizer05: I think you didn't mention the aerosol part11:20
whitequarkat least I imagined a process where one would manually put graphite on every hole. not pretty.11:21
DocScrutinizer05well, I was a tad unspecific, when I said "any conducting coating"11:21
wpwrakmix the graphite with some weak explosive. then detonate it in the hole and graphite will automatically coat the walls :011:36
wpwraklike the exploding wire, but with proven russian graphite technology :)11:37
DocScrutinizer05you're giving me ideas again11:48
DocScrutinizer05not what you think :-P11:48
DocScrutinizer05you got a plate with holes and two conductive planes, each side one. It's easy to fill the holes with a substance that "shorts" the two planes, so you can make electric current flow through the substance. Should be useful for _something_ maybe11:50
Action: DocScrutinizer05 idly thinks about alkaline batteries that - when leaking - allegedly can cover whole devices with some crystalline substance that grows like ... mildew11:53
whitequarkit totally does, I've had a "pleasure" to clean my usb keyboard from that shit recently11:53
whitequarkhad a few fun hours with sandpaper11:54
DocScrutinizer05is that stuff conductive maybe?11:54
whitequarkI don't think so11:55
whitequarkit's a crystallized salt11:55
whitequarkNaCl isn't conductive when dry11:55
DocScrutinizer05not even copper salts are conductive, and only few of then disassociate into gas and conducting copper at low temperature. yep, I know ;-)11:56
DocScrutinizer05anyway when you fill a hole in a PCB with some conductive goo or putty (or porridge) and then send current through it from the two copper planes, the current density is magnitudes higher at the edges of the hole inside the goo than anywhere else12:00
DocScrutinizer05basically theoretically you even could use something loosely reminding to electromigration in such setup12:01
DocScrutinizer05a problem is "race"12:01
DocScrutinizer05the process has a positive feedback and even stops as soon as one single sub-process finished, effectively defeating multiple sub-processes to complete12:02
DocScrutinizer05incandescent bulds have same problem. You're using metal-halogenid gas to turn the positive feedback (thinnest part of wire heats up most and thus evaporates fastest) into a negative one - well, almost12:05
whitequarkwell, that's why I don't fill holes with good12:06
whitequarkI rather cover the internal surface of hole with it12:06
whitequarkI've seen a process where you cover PCB with conductive ink and then use compressed air to blow it out12:06
whitequarkor a vacuum cleaner12:06
DocScrutinizer05btw that's the reason why you shouldn't dim halogen incandescent lamps too much. the halogen-metal will not dissociate anymore12:06
DocScrutinizer05sounds reasonable12:07
wpwrakhehe, just saw that putin is among the peace nobel prize nominees this year :) maybe that ukraine thing is just to boost his chances :)13:20
whitequark>Over time many individuals have become known as "Nobel Peace Prize Nominees", but this designation has no official standing.13:22
wpwrakwell yes, but one of them will get it, right ?13:22
larscwell hitler was nominated as well13:23
larsc(as a joke though)13:24
pcercueiHitler didn't do only bad things, he's the guy who killed Hitler after all13:27
pcercueithough he also killed his assassin13:27
wpwrakso he's a martyr as well13:27
DocScrutinizer05you were even able to top first post13:28
Action: wpwrak bows13:28
wpwraka while ago, one politician made headline news around here with the following statement: "¡Hitler era un tipo espectacular! ¡Era muy importante en el mundo!” ("Hitler was a spectacular guy ! He was very important in the world !")13:30
wpwraksomewhat surprisingly, he got away with it, though with some scorch marks13:30
wpwraktechnical challenge of the day: goal: place pogo pins exactly vertically. the problem:: tactile feedback doesn't really help you, i.e., you don't feel a difference between them merely touching the pcb or touching it with their entire base13:34
wpwrakthis is for soldering, so you can't grab them directly but go through tweezers, which may contribute to the weak feedback13:35
whitequarkwpwrak: use 2 drilled plates for guiding them13:35
whitequarkmaybe 1 ~5mm thick plate13:35
wpwrakthat could easily get in the way of soldering13:35
whitequarkoh? why?13:36
wpwrakso any guide would have to be localized, either small, or limited by angle13:36
whitequarkwhat exactly are you trying to do?13:36
wpwrakalso, you must be able to apply some pressure. so just something that grabs it on the side won't be enough13:36
wpwraksolder pogo pins13:36
whitequarkwell... you have throughholes right?13:37
wpwrakdoens't help :-(13:37
wpwraki thought it would, but no, it doesn't13:37
whitequarknono. see, you stick pogopins through drilled guide plate, then place pcb exactly on top and stick end of pogo pin through pcb13:38
whitequarkthen solder it13:38
whitequarkshould be perfect13:38
wpwrakah, they're too short for proper soldering at the bottom. have to be soldered from the top13:38
wpwraki didn't make them ...13:38
DocScrutinizer05you have a problem13:39
whitequarkphoto maybe?13:39
wpwraki was thinking of the following approach: take a piece of wood and make a hole roughly the diameter of the lower "pipe", so that it will fit snugly13:40
DocScrutinizer05only idea I could come up with: glue13:40
wpwrakthat would could be used for holding the pin while soldering. the problem: it would still "float"13:40
DocScrutinizer05err, didn't you just reject that approach?13:40
wpwrakit would be small and end some mm above the base, giving room to reach in with the iron13:41
DocScrutinizer05so what's the problem then?13:41
wpwrakkeeping it exactly vertical13:41
DocScrutinizer05now you lost me too13:42
wpwrakso one way to do that would be to put a pcb with an acceleration sensor on top of it, and use that to detect the orientation and to indicate when the pcb is level (and thus the wooden cylinder glues to its bottom would be vertical)13:42
wpwrakthe pogo pin needs to be placed vertical with respect to the pcb. let's assume i have a sufficiently level table that the pcb lies completely flat13:43
DocScrutinizer05get a wood 6mm*30mm*100mm, drill as many (deadend) holes as needed into the 6*100 surface13:43
DocScrutinizer05when really needed, add contraption to keep that wood level to PCB13:44
wpwrakah, you'd make the "cylinder" long, like a pencil. yes, who not13:45
wpwrakthat would also work with the sensor approach. and may even be accurate enough with just dead reckoning. thought that's oddly imprecise for this sort of task, probably because you can only see inclination in one plane at a time13:47
wpwrakthat's also the problem with drilling vertically without a drill stand. you think you're perfectly positioned but then you find you've been way off ...13:48
DocScrutinizer05or, very straight forward prgmatic: grip the pogopin like a mill in your CNC, make CNC insert it into PCB hole, solder13:48
wpwrakargh ;-)13:49
whitequarkactually a good idea13:49
wpwrakwith your BIG mill, perhaps13:50
whitequarkbig? mine is 30x20. BIG one is 70kg 60x40 model13:50
DocScrutinizer05for the wood approach: when you get a L-shaped wood with short end 100mm and long end 200mm and you drill the pogp holes to the open end of short part, you have a hard time holding that *not* parallel to the PCB13:50
wpwrakhmm, not sure about that. the basic problem remains that you only see one axis/plane13:52
whitequarkwpwrak: just drill it with CNC13:53
whitequarkit will be vertical the13:53
DocScrutinizer05are you able to keep the long end of a 200*400mm object in <20mm proximity to your table's surface?13:53
DocScrutinizer05if yes then you don't need to see *anything*13:53
DocScrutinizer05I leave it up to you to calculate the sin() and cosin() etc for 400mm vs (20mm * 10%)13:55
wpwrakyou could tilt the L also along the long axis13:55
DocScrutinizer05yes, when you ponder long enough how to act like an idiot to do that, you can13:55
DocScrutinizer05an object in a plane that touches amother plane in 3 point is defined in relation to that other plane13:56
DocScrutinizer05one point is your pogopins, you might actually be able to find a way to make 2 more contact points between L and your table13:57
wpwrakyes, you could make a "bridge". but then you need to match heights. not quite trivial, need space, and will change with tool wear13:59
DocScrutinizer05you know, 3-leged stool doesn't wiggle14:00
wpwrakexactly. but the distance between legs would be very small since the pcb isn't large14:00
DocScrutinizer05again: I leave it up to you to calculate the sin() and cosin() etc for 400mm vs (20mm * 10%)14:00
Action: DocScrutinizer05 headdesks14:01
wpwrakfor 400 mm, i would to add whatever the pcb is mounted on as well. this makes the system even more complex.14:01
DocScrutinizer05isn't the pcb flush on your pretty large table?14:01
wpwrakno, not necessarily14:01
DocScrutinizer05meh, I'm out14:01
wpwrakit's level with what support structure i use, though. and that one is flat on the table.14:02
DocScrutinizer05try segway technology!14:02
DocScrutinizer05btw the pogopin itself for sure isn't precise enough to try and make it sit in PCB exactly perpendicular, based on the rim at bottom alone14:06
wpwrakdunno. it would seem that this ought to work if feedback was better14:07
DocScrutinizer05never will14:07
wpwraki mean, you can usually feel if something makes proper contact or not. and the solder it liquid at that time, so the pad is very level14:08
DocScrutinizer05starts with pogopin rim not absolutely plane and perpendicular to pin's long axis14:09
DocScrutinizer05then continues with the hole you drilled into PCB and the gaps and edge rounding you have there14:10
DocScrutinizer05but sure, it's definitely way more exact plane to surface of PCB than PCB surface is plane to table it sits on¡14:11
DocScrutinizer05after all there culd be a grain of dust under one corner of the PCB and lift it from table14:12
DocScrutinizer05and then the wood, OMG! that might even deform with air humidity. Yu'll need to use massive platinium14:13
DocScrutinizer05do those pogopins adjust the laser targeting at the moon and there particularly to the reflector the Apollo astronauts placed there? Or why do they need that precision?14:14
wpwrak;-) btw, this is what they look like: http://www.digikey.com/product-search/en/connectors-interconnects/contacts-spring-loaded-and-pressure/1442508?k=0906-114:16
wpwrakso the base is fairly substantial14:16
wpwrakactually no, this is the one i have, a bit taller: http://www.digikey.com/product-search/en/connectors-interconnects/contacts-spring-loaded-and-pressure/1442508?k=0906-314:17
DocScrutinizer05is soldering end the thick or thin one?14:19
wpwrakmaybe what i should try first is just a piece of wood to get a firmer grip (compared to tweezers, which add at least one degree of freedom on their own). maybe that'll be enough already.14:19
wpwraksoldering is at the base where the thin through-hole pin emerges14:19
DocScrutinizer05and that fsckng through hole pin is too short to get soldered from other side of PCB? :-o14:20
wpwrakyeah, pcb is 1.6 mm, pin is nominally 1.75 mm. it's barely visible14:21
DocScrutinizer05honestly, glue that shit into place before soldering14:22
DocScrutinizer05or make sure your PCB hole fits snuggly14:23
DocScrutinizer05worst case a pair of pliers slightly flattening the thru hole pin can do wonders14:24
wpwrakyeah, that could be an option. they specify a fairly large (minimum) tolerance14:24
wpwraki think by the time it deforms, you've already torn it off :)14:25
DocScrutinizer05tear off a wire by applying force perpendicular to its axis, from both sides? how does that happen?14:26
DocScrutinizer05I've flattened wires from o to I that way, but never ever torn one off14:26
DocScrutinizer05unless I used wire cutter instead of flat pliers14:27
DocScrutinizer05when piers are too weak, I sometimes use a hammer14:29
DocScrutinizer05or if you're really worried, rather apply some leadfree solder to surface of wire. Just so much that it sits in your PCB hole tightly14:31
DocScrutinizer05though I consider the mechanical flattening of a 0.3mm in middle of those 1.75mm wire the better option14:33
DocScrutinizer05usually you have suqare-shaped nails instead of round wire for such purposes, on thru hole components14:34
DocScrutinizer05and the diagonale of the square is a few tens os a mm wider than the hole you drill to the PCB14:34
sharekwhitequark: remember my question of yesterday ? (disable poweroff screen) ? I found the answer... is easy! setter -blank 0 > /dev/tty014:35
sharekthanks for the tip, you are in the right way for solve them!14:35
DocScrutinizer05s/os a /of a14:36
DocScrutinizer05makes for a nice press fit14:36
DocScrutinizer05square shape isn't mandatory though. Just make sure that your wire has a different shape than your hole, with one dimension slightly larger than hole diameter14:38
DocScrutinizer05heck, you even could cut a thread on the wire ;-P  -  then screw it into the PCB14:39
DocScrutinizer05when your PCB is through hole, you could as well consider spot-welding, with high current. I'd rather test that on a few scrap samples though ;-)14:45
DocScrutinizer05apropos wire cutter - you know those isolation stripper pliers with the two V-shaped blade claws and the screw to adjust how far they close. Ideal to make some dents into the wire without cutting it (or tearing it off). You just need careful adjustment of the screw14:49
DocScrutinizer05when adjusted correctly (and actually not too short wire to pass the front blade thickness) this thing will make almost perfect 4 teeth to your wire, that hold the wire perfectly in a hole of matching diameter14:54
wpwrakhmm, i'd rather refrain from abusing these little pins. they're not that fat after all.17:31
wpwrakbtw, for wire stripping, i'd hotly recommend these: http://toolboom.com/en/Pro-sKit-1PK-3001E-Precision-Wire-Stripper-With-Conductive-Handle.php17:34
wpwrakthere's also a 3002 for AWG 20-1017:34
wpwrakthey're almost perversely good. better than anything else i tried17:35
wpwrakalready at pro'skit there are a number of variants of the theme. PK-3001, 1PK-3001E, 8PK-3001D, maybe more. they look a little different but all do the same17:37
DocScrutinizer05magic for all down to 0.2mm^217:39
DocScrutinizer05and up to maybe 16mm^217:39
DocScrutinizer05no adjusting, no picking right tool or tool-section17:40
DocScrutinizer05lightning fast, one handed action17:40
DocScrutinizer05another recommendation of me: http://www.youtube.com/watch?v=LOiCEeJQ24g17:43
wpwraknaw. i have that kind, too. but it's no comparison17:44
wpwrakmine has a flat blade, so it's very convenient for UBB cables (small ribbon) ;-) but for individual wires, i just use the simple ones17:46
DocScrutinizer05everybody has his favorites17:51
DocScrutinizer05had that flat cable straight blade thing too, and fiund it not to the point since you don't need stripped flat cable where the single wire attaches to other wires so distance between the wire strands is raster of the flat cable17:53
DocScrutinizer05also had that tool you favor and found it quite inconvenient to use17:54
DocScrutinizer05when I actually want to solder a flat ribbon cable to a number of solder points in raster distance, I use the effect of plastic retracting 0.5mm on heating it17:57
DocScrutinizer05also you can poush the isolation a 1mm or so *before* you cut the flat cable. After cutting the isolation will retract to original position slowly, exposing a 0.5mm to 1mm of blank copper wires17:58
wpwrak(flat cable) hmm, for UBB (or any case where the PCB pattern matches the cable) the raster is just fine ... obviously, because it's designed for it :)17:58
wpwrake.g., here: http://downloads.qi-hardware.com/people/werner/anelok/tmp/ybpgm-0-parts.jpg17:59
wpwrak(push) you mean pull/stretch. yeah, that sounds nice. i wouldn't count on plastic retracting with heat, though. it will, but it can get messy, too, with molten/burnign plastic where you don't want it18:00
DocScrutinizer05ok, when you often have that usecase then the stright blade tool is obviously useful18:00
wpwraki keep it around just for that ;-)18:01
--- Thu Mar 6 201400:00

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