#qi-hardware IRC log for Monday, 2012-04-16

qi-bot[commit] Xiangfu: split common to c/r/led for follow one lib per symbol (master) http://qi-hw.com/p/kicad-libs/82f842901:12
xiangfucladamw, wpwrak c/r/led done01:13
xiangfucladamw, please update your local git repo. :)01:14
cladamwxiangfu tks. :)01:27
qi-bot[commit] Adam Wang: added Xiangfu's new libraries for symbols (master) http://qi-hw.com/p/kicad-libs/9f18a2901:40
qi-bot[commit] Adam Wang: added pwr.lib (master) http://qi-hw.com/p/kicad-libs/b62279901:41
cladamwwpwrak, one interesting question, I changed +3.3 symbol text size to 0.06" under pwr.lib, but everytime I import +3.3V symbol but its text size still 0.03", does KiCad always have its rules to link internally ?01:56
cladamwi have to manually change it into 0.06" size. phew ~ bad01:59
wpwrakmaybe you still have the other lib in your .pro02:01
cladamwno, i removed "power.lib" from .pro02:08
cladamwmaybe caused by "../../../../usr/share/kicad/library/ is still existed there under "User define search path" ?02:10
wpwrakmaybe you also have to delete *-cache.lib02:12
cladamwhmm...after I removed "../../../../usr/share/kicad/library/" there under "User define search path", it still being imported a 0.03" size. 02:13
cladamwno, i'm editing under /kicad-libs/components/ , so there's no *-cache.lib though :(02:15
wpwrakso where do you see it appear with size 0.030" ? in the component editor or in the schematics editor ?02:16
cladamwdon't know why caused it. I hope it's not Kicad internal detection system. but seems to be ?02:17
cladamwI compared both. :(02:17
wpwrakso it happens in both ? maybe you simply didn't save your changes then02:17
wpwraknote that you have to save the component into the library and then the library to disk02:18
cladamwin components editor I see 0.06" but in schematic editor it's 0.03". 02:18
wpwrakah, good02:18
wpwrakthen exit kicad. make sure there's no *-cache.lib, check that .pro is correct02:18
wpwrakthen start eeschema, delete the old component, and put the new one. that must work :)02:19
cladamwno i saved it in component editor. :(02:19
cladamwhmm... okay... check file itself. phew~02:19
cladamw1. closed KiCad, no *-cache.lib in there02:21
wpwrakif it doesn't work, you have to debug it ;-)02:21
cladamwwpwrak, nice, yes just fixed this bug by directly editing pwr.lib phew ;-)02:28
wpwrakwhich means that you didn't actually save your changes before :)02:32
cladamwhmm ? but i clicked "Save current library to disk" after i changed. :-)02:35
wpwrakdid you also save the component to the current library ?02:36
qi-bot[commit] Adam Wang: saved a 0.06" text size font (master) http://qi-hw.com/p/kicad-libs/0144a6502:36
wpwrakthere are two steps: component -> library and library -> disk/file02:36
cladamwyeah... :-)02:37
cladamwbtw, do you think that we should keep the same symbol 3V3 as m1r4's AD schematic symbol ? or use a +3.3V ?02:39
wpwraki kinda like 3V3. it's short :)02:40
wpwrakand negative voltages are fairly uncommon in modern circuits, so the "+" adds little information02:40
qi-bot[commit] Adam Wang: changed +3.3V to 3V3 text (master) http://qi-hw.com/p/kicad-libs/5cc635302:58
kristianpaulviric: perhaps, but then i will need something symbolic to give in exchange03:04
qi-bot[commit] Adam Wang: added 1V2, 1V8, 2V5, 5V, USBEXT5V symbols (master) http://qi-hw.com/p/kicad-libs/3f8046603:12
qi-bot[commit] Adam Wang: changed TP's reference text to size 0.06" (master) http://qi-hw.com/p/kicad-libs/8b1d18003:32
qi-bot[commit] Adam Wang: added chip name "C_P" with polarized symbol (master) http://qi-hw.com/p/kicad-libs/606579804:06
qi-bot[commit] Adam Wang: added con.lib (master) http://qi-hw.com/p/kicad-libs/5bed93e06:03
virickristianpaul: give an autograph06:55
qi-bot[commit] Adam Wang: added GND, AUDIO_AGND, VIDEOIN_AGND symbols (master) http://qi-hw.com/p/kicad-libs/6012de107:10
qi-bot[commit] Adam Wang: added filter.lib to include FB (master) http://qi-hw.com/p/kicad-libs/f0c590a07:35
qi-bot[commit] Adam Wang: added crystal.lib (master) http://qi-hw.com/p/kicad-libs/a0f1bf008:45
qi-bot[commit] Xiangfu: new components switch, take from default device (master) http://qi-hw.com/p/kicad-libs/ef6023909:31
cladamwxiangfu, there's a fsmra2jh.lib i created already. you can use it for SW2. :-)09:42
cladamwbut you just created a switch.lib, i think it's nice. so maybe just let it includes fsmra2jh ?09:44
xiangfuagree. 09:46
wpwrakswitches should have generic names anyway, reflecting their connectivity but not any specific parts09:46
wpwrakotherwise, good luck with the thousands of part numbers just for simple on/off switches with two contacts :)09:47
cladamwwpwrak, fsmra2jh.lib i created which is exactly the on/off switches with two contacts/ for soldering. so the name may just not good only. We can just rename it to be nice one. But so far the p/n always be used for m1. :)09:51
xiangfuthe fsmra2jh have a simple SW. 09:52
xiangfucladamw, I think we can just remove hte fsmr... to switch and remove the switch.lib I have added.09:53
cladamwxiangfu, yes, just on/off but with another two contact for shields' contacts.09:53
cladamwyeah...i agreed, just includes it into your switch.lib :-o09:54
cladamwwpwrak, btw, since the switch has many different types. like http://en.wikipedia.org/wiki/Switch09:56
wpwraki think switch configurations are too complex for a full taxonomy. so we may want to only have a systematic part of the general properties, such as on/off vs. pulse, number of poles, number of ground pins, etc., and then just number or A/B/C/... them09:57
qi-bot[commit] Xiangfu: merge fsmra2jh.lib and switch.lib (master) http://qi-hw.com/p/kicad-libs/7b8cb3109:58
wpwrakad yes, SPST etc. should be very good for this taxonomy :)09:58
cladamwwpwrak, so for example: SPST-2S which means Single pole, single throw, two shields somethings like that. :-)10:00
wpwrak's more for relays: http://en.wikipedia.org/wiki/Relay10:00
wpwrakin particular then closing vs. opening switch (NO/NC)10:00
cladamwyeah .. more one parameter (NO/NC)10:01
wpwrakyeah, that sounds pretty good. maybe add a SW_ prefix. or maybe PUSH_10:01
cladamwwell.. i think that we just one-by-one to create each catogry10:01
wpwrakyeah, we can make a first try, then see how it goes10:02
cladamwsince that switch belongs to "Tact Switch" even not for ST10:02
wpwrakwe also have the possibility of imposing another, more dynamic hierarchy on top of all this in the "catalog". http://downloads.qi-hardware.com/people/werner/tmp/out2.pdf10:03
DocScrutinizerif you're struggling with switches, thank God you're not using electron Valves ;-)10:05
wpwrakso the component name would give a hint of what it is and it would make sure there are no name conflicts. for actually finding a part, you could use the catalog or kicad's component browser. the catalog is more flexible (you can move things around easily) and has an arbitrarily deep hierarchy10:05
cladamwwpwrak, you means creating a like catalog hierarchy page to get into sub page for symbols ?10:05
wpwrakcladamw: not in kicad. outside of it. see the PDF above10:06
cladamwwpwrak, yes, i see it. how "dynamic" ?10:06
wpwrakcladamw: in kicad, it would be too inconvenient because you'd have to switch between projects10:06
wpwrak(dynamic) there's is a text file that describes the hierarchy, so it's very easy to reorganize things. and changes don't affect schematics that already use the component we're moving around10:07
wpwrakhere's the hierarchy file for the PDF above: http://downloads.qi-hardware.com/people/werner/tmp/EXPAND.txt10:08
cladamwwpwrak, sorry that. now i know your ideas on "dynamic" (i.e. is folder text ) :-)10:09
wpwrakyeah, it's much like folders10:09
DocScrutinizerwpwrak: what a nifty think are you planning now?10:23
DocScrutinizerand Hi Adam!10:24
cladamwDocScrutinizer, Hi Hi. :-)10:26
wpwrakDocScrutinizer: oh, it's just a catalog that also allows us to review pin types. pin types are normally not visible and are hard to get right10:28
wpwrakDocScrutinizer: nothing new really. we had something similar already for gta02-core. but there, it was all done in kicad and thus less flexible10:28
qi-bot[commit] Adam Wang: added bnc, switch, filter (master) http://qi-hw.com/p/kicad-libs/0c0c8ea13:24
Aylamth, larsc: on sound/soc/codecs/jz4740.c:112, the "DAC Switch" available from userspace seems to only change one bit of a register, am I right?14:57
LunaVoraxHello everyone!15:35
larscAyla: sort of. if there is no other active part, it will shutdown the whole analog output domain16:14
larscand also the dac itself16:14
larscDAPM uses a graph16:14
Aylawell, if I understood correctly, several steps are needed to power on/off the DAC without creating noises16:15
kristianpaulviric: lol16:17
DocScrutinizerAyla: good point :)16:19
DocScrutinizerI'm generally no friend of too much "AI" in kernel(-drivers)16:20
DocScrutinizeresp when there's no way to override or even just check what that AI is doing16:21
DocScrutinizerfirst and foremost a driver shall export hw functions completely and in a basically unfiltered way. Then on top of that it *may* implement some *optional* "intelligence"16:24
DocScrutinizeryou could argue if a driver should allow to even write invalid values to registers or rather limit acceptable values to the range that's defined by the chip's/circuit's specs16:28
DocScrutinizerfor some drivers like e.g. lp5523 you'd think they better had limited /sys/class/leds/lp5523:kb3/led_current to values that won't fry your LED16:30
DocScrutinizerfor other drivers (e.g. FMTX) you'd think it would be nice to test what results you get when writing values higher then the "allowed" 120d to tx_power16:31
DocScrutinizermy take on that usually is "get a module parameter 'sanity-check=disable' to pass to the driver on modprobing it"16:33
DocScrutinizerbut definitely I think a driver that doesn't even expose *any* way to fiddle with e.g. audio codec function block suspend, that's broken by design16:34
qi-botThe build was successful: http://fidelio.qi-hardware.com/~xiangfu/build-nanonote/openwrt-xburst.full_system-20120415-1203 16:48
viricwpwrak, kristianpaul: YPF, argentina! :)17:25
viricI can't tell much for the argentinian part, but repsol is not my list of friends.17:36
viricBut it may be all circus, as usual.17:36
wpwrakviric: you're congratulating argentina for being de facto blocked from equitable investment for at least the next three years, after setting this sort of precedent ? :)17:39
viricwpwrak: among other things, yes ;)17:44
viricbut things are not so simple.17:44
viricwpwrak: today one of those in the Spanish government said: "Ya no estamos en campaña; es momento de decir lo que de verdad pensamos"17:45
viric(unrelated to ypf :)17:46
wpwrakviric: well, if you google for moreno or boudou, you should be able to get a fairly quick impression of how politics are going here these days18:01
wpwrakviric: i appreciate your schadenfreude at your government receiving a blow below the belt, but i think this may just be a case of everyone ending up on the losing side18:03
viricit can be, it's hard to know in advance.18:03
viricIt's some kind of elite against some other kind of elite18:03
viric(or same kind, but different people at least :)18:03
wpwrak(know in advance) right. if you jump out of a 10th floor window, you can't be sure in advance what exactly will happen :)18:04
viricah well, you seem quite sure :)18:04
wpwrakwhen you see patterns repeat for the third of fourth time, you begin to have a hunch :)18:05
wpwrakonly a year ago, the same president was praising ypf for their great investments in argentina18:06
viricI've no idea on the argentinian government. As for the Repsol side... do you think it has been a trick, and the old owners have received benefit of this operation?18:06
wpwraknow, in the middle of a major crisis involving the vicepresident, all of a sudden they pop out that "patriotic" card. malvinas, anyone ? :)18:06
wpwrakthere will probably be endless lawsuits following this coup18:07
wpwrakmuch like after the "haircut" in 200218:07
wpwraksome of the 2002 lawsuits are still going on. and present a danger for argentine assets abroad. as soon as someone can plausibly show that they're under the control of the government, they can be sized. that has happened already.18:08
viricwpwrak: do I read right that the nationalisation has been realized by Dromi, who prepared the former privatization (of YPF)?18:09
wpwrakso the YPF coup may also have some unintended consequences18:09
viricIf someone is psichopatic, egoist, greedy, and corrupt, that has lots of chances to participate with great power in a country government ;)18:10
viricwpwrak: I hope the spaniards don't decide to go into some kind of 'war'... as in the Perejil case18:11
virichttps://es.wikipedia.org/wiki/Incidente_de_la_isla_de_Perejil I hope we can see "Perejil: the movie" some day18:14
wpwrakdromi ? haven't even heard that name. i think on the argentina side, kicillof had a lot to do with it. a rather dubious character.18:20
viricRoberto Dromi18:20
wpwrakargentina has almost no military force these days. which is good. keeps the government at least from that sort of folly18:20
viricHere they don't know whom to give more money: military or church18:21
wpwrakfor spain, it may be attractive, though. kill off a few angry young man. maybe they could put them on board of decrepit ships, sink them, and then blame argentine submarines :)18:22
wpwraka good investment in either case :)18:22
viricwpwrak: if you google from roberto dromi, you see him both in the privatization and in the nationalisation :)18:22
viricwpwrak: could be :)18:22
viricwpwrak: google for.18:22
wpwraki only see him mentioned on the side of privatizations18:25
viricwpwrak: when I said "google", in fact I was using ddg :)18:26
wpwrakin that deal, he was quite on the privatization side18:27
wpwraki don't see any mention on him pushing nationalization anywhere18:28
wpwrakof course, some things were nationalized. but that wasn't his doing18:28
kyaklittle spammer...18:53
kristianpaulviric: i'm living at argetina btw19:38
kristianpaulnot living19:39
virickristianpaul: ah :)20:39
virickristianpaul: colombia? I can't recall20:39
wpwrakyes, the country of the traitors who didn't put argentina's malvinas on the agenda :)20:40
wpwraki think argentina will now retaliate by excluding columbian students from some scholarship programs20:50
wpwrakat least that's what i gather may be behind an odd "interview" that popped up in some approval process a brazilian friend is presently going through20:52
wpwrakthe timing isn't entirely right, though. that thing came up a few days before the summit. maybe it's in response to police reports indicating that some 80-90% of burglaries are committed by columbian gangs that send "students" to act as beach heads (so they have all their papers in order an can rent an apartment. then the rest of the gang follows without having to bother with bureaucratic niceties)20:55
viricwhat a complot20:56
wpwrakyeha. maybe it's just the ritualized feeding of xenophobia :)20:56
viricDo you think this may work?21:04
viricsounds very cheap21:04
kristianpaulwpwrak: ;)21:04
viriclooks like wanting only usbhost21:04
kristianpaulnow i will not travel to argetina and visit wpwrak :D21:04
viricI don't know if there are limits on unidirectional transmission power.21:06
viric*legal* limits21:07
viricit's not that I want to play being Tesla21:07
wpwrakit doesn't really say what TX power it has, does it ?21:08
kristianpaulor RX gain limit? ;-)21:08
wpwrakkristianpaul: afraid or insulted ? :)21:08
viricI don't know that unit though.21:09
wpwraknaw, you can make your RX gain as high as you wish21:09
kristianpaulwpwrak: neither :)21:09
wpwrakviric: that's the amplification. but what's the input ?21:09
virichm no21:09
viricdbi is the relationship between directional power / omnidirectional power21:10
viric(just checked)21:10
kristianpaulwpwrak: may be afraid of ticket costs ... 21:10
viricit may be the usual 15 or 20dBm21:10
wpwrakviric: (dBi) provided they got that right :) 60 dB of anything seem suspicious, though21:12
wpwrakif it's really dBi, that would be a rather narrowly focused directional antenna21:13
wpwrakmaybe it's just dBM, with M = Marketing21:13
viricyes, it sounds very narrow.21:15
viricI've really have zero experience with unidirectional antennas21:17
wpwrakif i understand the concept right, the beam width would have to be around 22 arcseconds for that. still not as good as a decent laser, but close21:23
wpwrakwell, the beam divergence, to be precise21:23
viricI decided not to engage into such a calculation...21:24
viric22 arcseconds is quite narrow ;)21:24
viriclaser at radio freqüency; nice!21:24
wpwrakyou could mount a laser on the antenna to indicate where to place the opposite antenna ...21:24
viricfair :)21:26
whitequarkI heard that the main reason such laser-antennas aren't produced is because of their marginal usefulness21:31
whitequarkdue to them being very hard to position21:31
viricof lasers at visible spectrum you mean?21:34
viricI remember seeing them for high speed links between buildings21:34
whitequarknope, by "laser-antenna" I meant an RF antenna with a very narrow link21:34
viricas visible spectrum today I saw this http://www.interconnect-ns.co.uk/Laser-Links21:35
whitequarkhaha, nice. my classmate once did a simple version of that21:35
whitequarkwith a typical chinese laser pointer (it was before those EXTRA BRIGHT green lasers and stuff)21:36
viricwell, I claim it's for visible spectrum, but it does not say that.21:36
whitequarkwell it's possible21:36
whitequarkeven with a chinese pointer, that is21:36
wpwrakwhitequark: they may also be a bit inconvenient to carry around. e.g., the arecibo dish seems to have 60 dBi ...21:36
whitequarkof course he only got 120 baud or such, but hey, that was a one-weekend project21:36
viricgood enough21:36
whitequarkwpwrak: :-)21:36
viricwpwrak: for high enough frequencies, maybe it can fit your pocket.21:37
whitequarkI sometimes wonder about THz spectrum21:37
whitequarkthat is, the visible part of THz spectrum21:38
whitequarkit's still "radio" waves, right? it's electromagnetic radiation21:38
whitequarkso, can we use metal antennas to emit it?21:38
whitequarkjust imagine an antenna which glows when turned on21:38
viricyes I wondered about that too21:39
whitequarkit's like LEDs, but other way around21:42
whitequarkwe can generate light by shifting electrons back and forth between the orbitals in a LED21:42
whitequarkand we can generate radio waves by doing the same, for example, with hydrogen21:42
whitequarkthe freq would be something like 2MHz21:43
whitequarkor 20, I don't quite recall21:43
whitequarkthrough the antennae for THz waves should be quite small21:45
viricI can't tell if you understad all this better or worse than me21:46
whitequarkand after a certain wavelength (gamma? xrays?) you'll need to make them smaller than atoms21:47
whitequarkon the other hand, what an antenna is? a structure where a wave can propagate well because of lots of free electrons21:47
whitequarkthat is, when we talk about antennae and waves, we don't treat quantums or their relations as separate particles21:48
whitequarkso in a certain sense a single atom can be an antenna21:49
whitequarkno idea if I'm telling nonsense or not :)21:49
--- Tue Apr 17 201200:00

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