#qi-hardware IRC log for Monday, 2014-05-12

nicksydneywpwrak: http://www.drdobbs.com/embedded-systems/breadboarding-arm/24016814103:51
wpwraknicksydney: very retro :)08:33
nicksydneygot the inductor but the footprint i'm using is too small for me to solder it properly...so doing it manually inside PcbNew11:55
wpwrakdesign a bigger footprint ? or add a new size option to an existing one ?12:11
Action: wpwrak notices that it's very hard to get people to try fped. once they've used for a few minutes, they love it. but getting them to make the first step is nearly impossible.12:12
larscsame as with heroin12:21
viricif you look only at similarities, almost everything is the same12:28
nicksydneywpwrak: making the footprint larger inside PcbNew ..if it works out than will create a new module with different footprint12:31
wpwrakheroin, the universal similant. now there's a new thought :)12:31
wpwraknicksydney: are you using footprints from the kicad libs or from qi-hw'12:32
wpwraks kicad-libs project ?12:32
nicksydneywi0h12:32
nicksydneyqi-hw12:32
wpwrakthen you already have the original in fped12:32
wpwrakyou're using iunductor-2p-3mmx3mm ?12:33
nicksydneyyeah12:34
wpwraktry  cd kicad-libs/modules; fped inductor-2p.fpd12:34
wpwraksee the nice little table ? :) just has one row at the moment, but there could be lots and lots more :)12:35
wpwraksize{x,y} is the component size; p{x,y} is the pad size; gap is the space between the pads12:35
nicksydneyhttp://postimg.org/image/bep50rokz/38b232d4/12:35
wpwrakmove the divider to the right or increase to full-screen12:36
nicksydneyok i've increased the px and py12:36
nicksydneybut it did not chnage12:37
wpwrakdid the background turn blue ?12:37
nicksydneypx turns to magenta12:38
nicksydneyand at the bottom there is a green bar where i can enter the value12:38
nicksydneyhttp://postimg.org/image/n2p95zbz1/574ef59f/12:38
wpwrakyup. don't forget the mm12:38
nicksydneyahh ok12:39
nicksydneyit works12:39
wpwrakif you forget the unit, then it prints an error in the status line at the very bottom and the background of the drawing turns blue12:39
wpwrak(and it doesn't update the drawing)12:39
nicksydneyso once i save it ...and go back to kicad it will refresh ? or any other steps i  have to do ?12:40
wpwraknaw, don't change the existing entry. else you'll lose the old footprint.12:42
wpwrakquit, restart, and make a new row instead.12:42
wpwrakor is your inductor 3mm x 3mm, too, just differing in px/py ?12:42
wpwrakwhat's the part number anyway ? (or datasheet link)12:43
nicksydneyhang on let me get the datasheet12:43
nicksydneyhttps://github.com/nanikjava/RA8875/blob/master/documents/datasheets/Murata-LQH1C_LQH3C_LQH4C.pdf12:44
nicksydneyit's the LQH4C Type12:44
nicksydneyLQH4C6R8M0412:44
wpwrakah, you're on the wrong footprint with this12:45
wpwrakthese come in standard sizes. so this one is an 1812. lemme see if we have that already ...12:45
wpwrakyup. just use 1812 from stdpass.fpd12:46
nicksydney1812 or 1812P ?12:46
wpwrakit's not polarized, so just 181212:46
nicksydneyok let me print it out now12:47
nicksydneyok much better than the one i did manually :)12:50
nicksydneycool12:50
nicksydneylearn something new today about a chip called Super I/O ... only on x8612:52
nicksydneynifty little chip from this company http://www.fintek.com.tw/eng/products.asp?BID=1&SID=17&layer=012:53
nicksydneywonder if in ARM world they have something similar12:53
nicksydneyapparently can do HW monitoring too12:54
whitequarkSuper I/O is the remnant of the 8086 era12:55
whitequarkbasically a kitchen sink of various compatibility shit they have to still carry around12:55
whitequarkproper CPUs simply shed that nonsense when it becomes obsolete12:56
nicksydneywhat i find interesting is the HW monitoring part 12:59
whitequarklook up ARM PMICs13:00
nicksydneycool thanks whitequark13:01
nicksydneyahh Power Management Integrated Circuits (PMICs)13:01
nicksydneywpwrak: http://www.dialog-semiconductor.com/products/bluetooth-smart/smartbond-da1458013:06
nicksydneyspoke too soon that's not BLE13:08
wpwrakactually, it is. "BT Smart" is the new official name for "BTLE"13:18
whitequark*facepalm*13:19
wpwrakwhether the idea of replacing a nice and descriptive name with something vague and generic is really a smart idea, seems somewhat dubious13:20
nicksydneyohh....interesting they change the name..making people more confused13:20
whitequarkwpwrak: well it sure sounded well to marketing guys13:20
wpwrakin think the official acronym for BTLE was actually "BLE". but that also robs is of the distinctive "BT" prefix.13:21
whitequarkthat's basically their job, replacing nice and descriptive with vague and confusing13:21
wpwrakso i just use a more understandable and descriptive acronym ;-)13:21
wpwrakyeah, they exceeded themselves in this case13:21
wpwrakit's somewhat galling that those certainly very smart people who make FETs had to place GSD (Gate, Source, Drain) counter-clockwise for single FETs, but clockwise in dual packages14:51
wpwrakthis means that when going from single to dual, not only the dimensions but also the topology of the circuit changes14:52
wpwrakthat is, unless ... hack, hack ... yeah, compatibility achieved in one case14:54
wpwraklet's see if i can do the other one, too14:55
apeleteHi larsc20:31
larschi20:39
apeletelarsc: got a minute ? would to share some debug thoughts with you20:42
larscbut only one20:43
apelete:)20:44
apeleteseem like EIO is set by drivers/mmc/card/block.c20:44
apeletehttp://paste.debian.net/99145/20:44
apeletedon't know if that makes sense20:44
apeletewhen mmc_start_req(card->host, areq, (int *) &status); fails, it goes into MMC_BLK_RETRY states 5 times in a row before going into MMC_BLK_ABORT20:46
apeletethat's where it seems to set EIO with blk_end_request(req, -EIO, blk_rq_cur_bytes(req));20:47
apeletenot sure about that, so I also stepped through jz_mmc_irq() in mmc driver (that's one of the two codepath in the mmc driver that also set EIO)20:49
apeleteit didn't triggered the EIO codepath though20:49
apeleteso last thing to do would be to step through jz4740_mmc_transfer_check_state() which has another EIO codepath20:50
apeleteand see if ot actually triggers an IO error (and if it is linked to drivers/mmc/card/block.c at some point)20:51
larscdo you know why mmc_start_req fails?20:51
apeletenope20:52
apeleteshould I look into that ?20:52
larscyes20:55
apeleteok, will do so20:57
apeletelarsc: thanks for your time :)20:57
nicksydneythe RA8875 has been shipped from taobaoring.com by EMS so hopefully in 2 days time it will reach...interesting to see if i get what i paid for :)22:46
nicksydneyaccording to EMS tracker it's in Beijing right now22:47
nicksydneyVideo: Putin hits the ice to compete with legendary hockey players  -- https://www.youtube.com/watch?v=TUdYCzdpzdY&feature=player_embedded23:22
nicksydneybetter make sure you don't get in the way of Putin when he is playing hockey :)23:22
nicksydneywpwrak: found this company (website) that can print things out based on your design in NZ https://www.ponoko.com/23:39
nicksydneythey have leather, paper, plastic, wood, cardboard23:40
nicksydney"PETG is a clear thermoplastic polyester with a glossy finish." sounds sexy :)23:40
wpwraka local shapeways ?23:52
--- Tue May 13 201400:00

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