#qi-hardware IRC log for Friday, 2011-04-01

wpwrakroh: btw, will you test the sugru ? note that it may change rf performance, so you'll have to check that too.00:18
rohuuuuh. nope.00:50
rohwould need to find some first00:51
wpwrakyes, or devise a theoretical absorption model and simulate it. that may be a few years more work, but i'm sure the world of RF designers would be eternally grateful ;-)00:53
lars_i have some sugru here, only 3 hours and roh could have it ;)01:10
lars_probably faster than running the simulations01:10
wpwrakdarn. scientific breakthrough thwarted by convenience !01:11
lars_you can have everything01:13
panda|x201it's a Unicon!01:19
wpwrakaw_: by the way, do you have any comments on the atben/atusb boards ? anything you'd suggest to change ? there may still be time :)01:25
aw_wpwrak, if a panel shown up with 10 pieces, it would be great. i ususally confirm the final panel gerber to let pcb maker to produce after i confirm it's okay.01:28
wpwrakaw_: okay, that's a good thing to consider. maybe also send it to the smt fab to check.01:29
aw_wpwrak, so you'd better to check if final panel regression compared to your current (working) design.01:31
aw_especially focus on edges on combined each piece of panel.01:32
wpwrakaw_: you mean a shift in the layers ?01:33
aw_yes, only after you confirmed the final panel gerber then they start to make pcb, then send that final panel gerber to let smt maker to produce stencil.01:34
aw_wpwrak, no; just check each edge of small pcb in panel.01:35
wpwrakaw_: i think pcb and smt fab know each other quite well in this case. so we won't need to act as an intermediary. but checking never hurts :)01:35
wpwrakaw_: (edge) okay. what kind of problems should i look for ?01:35
aw_yes, pcb maker always utilize materials as possible to place 10 piece into panel. so those 10 pieces maybe are not on the same row/column or direction.01:36
aw_make sure that the most usb connector & antena area.01:37
wpwrakaw_: what problems should i look for in the antenna area ?01:38
aw_10 pieces with few supporters linked together, you need to confirm if the strengthen is quite stronge enough when smt machine mounting.01:38
aw_just check if each pieces of antenna area have supporters links. if yes, you will pay more attentions on this.01:39
aw_second, let me show you.01:39
aw_wpwrak, http://en.qi-hardware.com/wiki/File:M1_jtagserialpod_run1_panel_front_pcba.png01:40
wpwrakaw_: ah, i see what you mean, thanks !01:43
aw_see this. imagine that 10 pieces are linked together. so you need make sure a "V-Cut" or "Stamps" links each piece and don't let "Stamps" linked your antenna's pcb edge. because those "Stamps" will be manually or cut by machine. so those edges around anttena is the most concern area.01:44
aw_only confirmed the panelized gerber then inform to pcb maker to produce it. :-)01:45
wpwrakaw_: i don't think "stamps"/tabs at the antenna would be a problem. in the case of atusb, there's of course the issue that there's a hole on the other side (for the usb connector)01:45
wpwrakaw_: but let's see how they solve this :)01:46
wpwrakaw_: (tab removal) i think it's tuxbrain + sand paper ;-)01:46
aw_wpwrak, also pay more attentions on usb connector's direction. like that panel link i sent you, if usb connector are placed in all the same direction, you will get trouble after smt mount. worker needs to "V-cut" your panel.01:48
aw_so sometimes a bi-directional arrangement is good: like one row 5 piece, they are on the same direction, the others are another direction. :-)01:49
wpwrakaw_: there will be a hole in the panel for the usb connector. the usb connector extends below the pcb: http://downloads.qi-hardware.com/people/werner/wpan/tmp/usb-off-board.jpg01:50
wpwrakaw_: so they'll just make a big hole for it. problem solved :) (maybe it's not terribly efficient, but i guess that's fine for now)01:51
aw_yes, i know it; let's how pcb maker arrange/merge your 10 pieces first. :-)01:51
aw_i didin't talk about that hole, make sure let david to send panelized gerber from pcb maker, so we can view it. don't directly inform them to make pcb. :-)01:53
wpwrakaw_: yup. will do01:54
kristianpaulwow sugru looks awesome, who need 3d printers after that ;)03:22
wpwrakkristianpaul: those who want to make > 1 units that look the same ? :)03:23
kristianpaulhttp://sugru.com/gallery/show/comfier-tap :D03:24
kristianpaulthis is the top http://sugru.com/gallery/show/padded-hook03:25
kristianpaulanyway .. :-)03:25
wpwrakkristianpaul: (padded hook) it'll be even more impressive when the sugru replaces the metal :)03:37
wpwrakthe tap looks pretty cool03:37
kristianpaul(metal) :-)03:39
tuxbrainlars roh, that sugru test will be awesome ;) I also have buyed a 12 units bag two days ago05:16
tuxbrainaw_ wpwrak:(panel gerbers) as wpwrak said the pcb and smt vendors usually work together and already talked about the requeriments, but I will send a mail to pcb to send me the gerber panel before it starts anyway, regarding stencil smt vendor was clear with this and will not do the stencils until they have phisically the pcb panel in their hands.05:20
aw_tuxbrain, yup...great. :-)05:22
wpwraktuxbrain: (smt guys being careful) good :)05:24
tuxbrainaw_  : has wpwrak comment you to sendme the atben/atusb material to me?05:27
Action: tuxbrain with greedy voice giveme giveme giveme05:27
aw_tuxbrain, no, not yet heard them from wpwrak05:28
aw_tuxbrain, wpwrak what will material be??05:29
tuxbrainaw_:, 2 atben, 2 atusb, and 2 atusb-pgm , I will sendyou back production units later05:34
aw_tuxbrain, okay.05:36
wpwrakaw_: that would be the samples i sent you, before it was decided that tuxbrain instead of sharism would do the production05:37
wpwrakaw_: sorry that this only materialized too late to avoid the detour05:38
aw_wpwrak, no problem.05:39
wpwrakaw_: thanks !05:44
wpwraktuxbrain: seems that we'll then have about one month to figure out a test process05:44
aw_tuxbrain, does the delivery address is the same as before?05:45
tuxbrainyeah also a good source of rf experience and equipment :) the guifi.net guys05:46
tuxbrainaw_: yes it is05:46
aw_tuxbrain, okay.05:47
tuxbrainwpwrak: what would be the best link to give them to descrive atben/atusb?05:47
wpwraktuxbrain: (guifi) got anything more cryptic ? ;-)05:48
wpwrakmaybe this http://downloads.qi-hardware.com/people/werner/wpan/tmp/2boards-20110305.jpg05:49
wpwrakand this http://downloads.qi-hardware.com/people/werner/wpan/tmp/atben-20110123-front.jpg05:49
wpwrakmention that it's an Atmel AT86RF231, doing IEEE 802.15.4 in the 2.4 GHz ISM band05:49
tuxbrainand some text descriving functionalyty and goals?05:49
wpwrakhmm, goals ... "poor man's wifi" ? :)05:50
tuxbrainguifi.net guys are implementing a WIFI a profesional high band and quality altenative network , they also obtain goverment telecomunication provider status, if this guys come on board we will have real RF pro ,06:06
tuxbrainhere is the guifi about in english http://www.guifi.net/en/what_is_guifinet06:07
wpwraktuxbrain: ah yes, they may have attracted some people with real RF knowledge and tools06:11
tuxbrainhere is the map of the guifi net nodes , lucky they are maily active in Catalunya :) http://www.guifi.net/en/node/3671/view/map06:15
wpwrakwith an outpost in africa. interesting :)06:18
tuxbrainI have contact them a couple of times and coincide in some meetings, but our lack of wifi and our low interest on implement it has make no real colaboration happens, I hope atben/atusb will change this :)06:18
wpwrakyeah, some folks there may like IEEE 802.15.406:19
tuxbrainI bet his first complain will be... but just 10 meters????? bah.... but I will try to redirect this to help us to make a high gain design :), we count with the advantage that they have the permision to operate at high power antenas on 2.4 band :)06:21
wpwrakyeah, there may be a lot of optimizations one can do06:22
tuxbrainok one insider is moving strings to arrange a meeting , I will try to have equipment access when aw samples arrive to perform some tests training for the production test :)06:24
wpwrakexcellent !06:30
tuxbrainwpwrak: aw_: confirmed, the pcb vendor will give us the panelization gerbers before production.07:58
aw_tuxbrain wpwrak great, let's check a bit. :-)07:58
wolfspraultuxbrain: remember the jtag-serial replacement?08:04
wolfspraulAdam will include 3 fixed jtag-serial boards into the shipment08:05
wolfspraulplease put the 3 old ones aside and return to me at some later opportunity08:05
wolfspraulI will definitely fix and reuse them.08:05
tuxbrainwolfspraul: yes waiting instutions,08:05
wolfspraulsince we got no replies from H:D or kd85, I assume their number is 008:05
tuxbrainok instructions recieved :)08:05
wolfspraulI just resent a mail to them08:05
wolfspraulif I have no number in the next 2 hours, Adam will send 3, it's that simple :-)08:05
wolfspraultuxbrain: you still have all 3 boards with you?08:06
wolfspraulthat would be best then you can replace all jtag-serial boards with properly fixed ones...08:06
tuxbrainyes, I was waiting for te replacement to send to pulster08:06
wolfspraulI didn't know that08:06
wolfspraulman we need to improve communication :-)08:06
wolfspraultomorrow Adam's package goes out, including 3*jtag-serial08:07
tuxbraindon't worry all undercontrol every body is happy here08:07
rohwhow you guys are moving fast08:07
tuxbrainroh who? sharims? tuxbrain? wpwrak? adam? all?08:08
wolfspraulroh: I got no answer to a simple question in 5 weeks08:08
wolfspraulfrom paying customers, distributors actually08:08
wolfspraulamazing :-)08:08
rohhehe.. just reading _long_mails08:09
wolfspraulbut yes, I assume they are so happy with the product, they don't even care about free fixes08:09
wolfspraulmust be that08:09
rohbtw.. better dont use my openmoko mail.. its working, but i dont check it that often08:09
tuxbrainI must admit that to be a the poorest international hardware group we are doing pretty well in terms of I+D and production, a bit o punch on marketing and we maybe left the status of the poorest ones :)08:11
tuxbrainroh what mail do you preffer08:11
rohtuxbrain: the smt quotes thread08:11
tuxbrain:) yeah, nothing like a fool in charge to accelerate things :)08:13
tuxbrainno knowledge, no equipment, no money, but leading a production of high frequency wireless device to use with a protocol not fully developed (6loWPAN) , put some metal as bkground and yeah that's Tuxbrain in pure state08:15
rohmy problem is that i havent yet found a nice 'concept' for casing both nicely and cheap08:15
rohi guess i will set up some kind of sheet-material built jig as a 'potting form' which helps aligning the pcbs and filling in something like sugru to a nice formed case08:16
tuxbrainat least you have found one nicely or cheap08:16
rohnah. neither. my bet is that sugru isnt really one of the expensive options. sinc one doesnt need that much in total08:17
rohthe usb thin could find a pre-made case, but sourcing that reliable can be hard08:17
tuxbrainthe sugru aproach doen't  look bad , both on atben and atusb,08:19
rohmaybe we should just try (rf performance)08:19
rohsadly i have no clue how to get my ben working08:19
rohwas some really early hw. i think ram or flash was smaller08:20
tuxbrainroh: lars_ can make you arrive sugru,08:20
rohah. ok08:20
tuxbrainmmmmm roh: send me your address08:20
rohwell.. i should check rf performance first08:21
tuxbrainyes a previous test unsugrued is a must to compare :)08:21
wolfspraulaw_: now they are moving, we are at 5 already :-)08:25
aw_wolfspraul, so 5 serial/JTAG pods?08:26
aw_wolfspraul, maybe more later. :-)08:26
wpwrakroh: (potting form) hmm, sugru may be far too viscous for that. i imagine it more like chewing gum than, say, silicone sealant used for construction12:44
wpwrakroh: and even the sealant isn't something i'd trust with anything resembling a cast at such small sizes12:46
wpwrakroh: but if things work, shipping with just a bag or two of sugru may be an attractive option. it would also justify calling the whole thing a "kit", which may help staying under the radar of FCC and the like.12:47
wpwrakroh: so what we would need is: 1) confirmation that it does indeed work, 2) an estimate of how much material is needed (i.e., is one bag enough ?), 3) report of any pitfalls, and 4) some pictures that show how cool it looks :)13:09
rohi will find out13:10
wpwrakroh: ah, you may be able to stamp it. nice way to add a personal touch.13:10
rohnice idea13:12
tuxbraindo you believe we should joint the oshw train ? at least for marketing :)13:16
lekerneltuxbrain: what's "the oshw train"?13:20
kristianpaulwell, it suposed copyleft hardware is just that, oshw is there13:20
kristianpauli mean not copyleft hardware is not to be confused with oshw13:21
kristianpaulof cours both share some things13:21
lekernelwhat's the difference?13:22
tuxbrainlekernel: http://freedomdefined.org/OSHW13:25
kristianpaul_for_ _me_ copyleft includes already oshw plus copyleft plans, (manufacturing, marketing.. )13:25
kristianpaulfree software13:26
kristianpaulfreedom to share a like and profit13:26
kristianpaulcopyleft hardware is not just one thing13:26
tuxbrainmust leave I will read you later13:27
lekernelah, yeah, those little people13:27
lekernelmy patience is far exceeded for all this crap13:27
lekerneljust put copyleft hardware into large retail stores. screw that OSHW thing. it doesn't work anyway.13:29
kristianpaulyou mean copyleft hardware like the mm1? :-)13:29
lekerneladditionally they don't want us (at least milkymist). believe it or not, it seems to me that they are more pesky than the largest french retail chain for technology products13:30
lekernelso I have the choice between selling 10 MM1s through pesky makezine people, or potentially 10K or more to this retail chain with relatively cool people. now pick one...13:31
wolfspraulcorrect we need to target professional retail and in general professional companies with strong marketing and strong existing customer base13:32
wolfsprauland we are creating the prerequisites for that right now13:32
kristianpaulCopyleft hardware as care for software freedom, must care for turnkey solutions i think13:32
wolfspraulfor us copyleft is a natural thing, I mean it's 2011, right? it's just the culture of the Internet, and we can focus on the product now13:32
wpwrakoshw take on the noble task for explaining to people who never chose to get out of bed the intricate details of the process of walking. meanwhile, we run around in the yard, playing :)13:38
rohwpwrak: i think we all agree that we implicate that besides other concepts13:38
rohthat meaning 'the concept of copyleft open hardware13:39
rohand i see the diversity and spattertness of the community as true feature13:41
rohe.g. the freakduino/chiboduino stuff.. also nicely documented and free code13:41
rohtoo bad these boards are in japan. 33$ are not even 25E anymore.13:42
wpwrakroh: (implicate) btw, that's not a good word to use. it it somewhat synonymous with "include", "involve", or "imply", but often has a negative sense.17:09
polobrazyhello, do you know if the driver of AMD enables the use of 3 monitors (full resolution) with the Radeon HD 5750 ?17:34
kristianpaulpolobrazy: /j #hardware17:35
polobrazythank you17:35
kristianpaulis is qi-hardware  Copyleft hardware  is out topic17:35
polobrazythank you, nice to learn a new concept!17:36
polobrazycould someone please invite me in the #hardware channel?17:37
qi-bot[commit] Werner Almesberger: atben: corrected type in RFP net name http://qi-hw.com/p/ben-wpan/84f2d2722:36
wpwrakgrr. s/type/typo/22:36
lekernel"Home > Products > Chemicals > Explosive"23:53
wpwraklekernel: bah, for that price, you can fly to tokyo where you get it for free !23:55
lekernelamazing xD23:55
lekernelis that for real?23:55
lekerneleverything seems to come from ukraine... scavenged from chernobyl?23:56
--- Sat Apr 2 201100:00

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