#qi-hardware IRC log for Thursday, 2012-07-19

wpwrakcladamw: heya ! have you seen my comments on the qi-hw list about the footprints ?01:33
wpwrakcladamw: as you can see, i'm slowly catching up with you ;-)01:33
cladamwwpwrak, yes, i saw them. Also thanks for fixes directly. :-)01:35
wpwraki hope i didn't break anything :)01:36
wpwrakregarding the through-hole-pads, do you agree that there should never be solder paste on them ?01:36
qi-bot[commit] Werner Almesberger: gui_status.c (show_pad_type): share pad type to name translation (master) http://qi-hw.com/p/fped/57dbeb101:43
qi-bot[commit] Werner Almesberger: postscript.c: move drawing part of ps_pad_name and ps_pad to separate functions (master) http://qi-hw.com/p/fped/238830c01:43
qi-bot[commit] Werner Almesberger: postscript.c (hatch): change argument from layer set to pad type (master) http://qi-hw.com/p/fped/26b74e801:43
qi-bot[commit] Werner Almesberger: postscript.c: show key for pad types at bottom of fullpage drawings (-P -K) (master) http://qi-hw.com/p/fped/d250ed201:43
qi-bot[commit] Werner Almesberger: postscript.c (prologue): set line join style to "round" in showoutlined (master) http://qi-hw.com/p/fped/c28f70101:43
qi-bot[commit] Werner Almesberger: postscript.c: cleaned up geometry calculation for pad type key (master) http://qi-hw.com/p/fped/ed69b7b01:43
qi-bot[commit] Werner Almesberger: postscript.c: only show the key for pad types actually used in the package (master) http://qi-hw.com/p/fped/b4d61a801:43
qi-bot[commit] Werner Almesberger: Merge branch 'master' of projects.qi-hardware.com:fped (master) http://qi-hw.com/p/fped/52c288e01:43
wpwrakah, xiangfu is in hot pursuit :)01:43
cladamwwpwrak, not exactly that should never for the future, it depends on h/w applications.01:44
qi-bot[commit] Werner Almesberger: genkicat/fped.c (fped_ps_entry): use fped -P -K (key) and adjust position (master) http://qi-hw.com/p/eda-tools/d67e8ec01:44
wpwrakcladamw: under what circumstances would you want solder paste on through-hole pads ?01:44
xiangfuwpwrak, I understand now how, .sch --> .net --(fped/footprints)--> .brd.01:50
xiangfubut do those jobs from 0 is not easy. why the C/R is here. why there is a ferrite_chip etc.01:52
cladamwwpwrak, the solder paste layer on PTH hole can be kept there once. Since stencil maker they identify/determine automatically which PTH with solder paste or be noticed by our demands. Considerations or comments from you are good though. 01:54
cladamwwpwrak, in eastern stencil maker, when they see gerber, the PTHs will directly be disable by s/w. We don't need to take care of them. It's more like a common condition to speed up process. But good comments from you. so no feedback to you on thread. :-)01:58
wpwrakxiangfu: yeah, the whole workflow is a little tricky ;-)02:14
wpwrakcladamw: hmm. i don't like the idea of relying on the fab to fix our gerbers for us :)02:14
cladamwwpwrak, sure. :-) as better work flow in Qi is good on own rules. :-)02:15
wpwrakcladamw: i wonder in what cases you'd want to have solder paste on TH pads. it seems to me that the default should be no solder paste, agreed ?02:16
wpwrakcladamw: in pads.fpd, the round ones are "bare" while the square and rectangular ones are with solder paste. shall i make them all bare ?02:20
cladamwyou could do so. I've been requested twice when did 19" chassis(frame) stack up to hook main boards with PTH(3.5mm diameters) with extra 8 small PTH surrounded 3.5mm PTH, then used solder paste to cover both, to solve system ground issues. the solder paste there for purposes are for adding strength on contacts of current flow and strengthen mechanical tolerances.02:22
cladamwwpwrak, yes, you can make them all bare. :)02:22
xiangfuwpwrak, where is the information indicate if there are solder or bare?02:23
wpwrakxiangfu: in fped, it's in the color of the pad. and when you select the pad, there's a clickable field in the input area that says "normal", "bare", etc.02:23
wpwrakxiangfu: if you make postscript, different pad types are drawn in different styles. and the very latest version of the catalog even tells you what is what :-) http://downloads.qi-hardware.com/people/werner/tmp/kicad-libs-modules.pdf02:25
wpwrakxiangfu: (upload just finished :)02:25
wpwrakcladamw: (chassis) okay, that sounds exotic enough that you may even want to make a special footprint for that case :)02:26
cladamwwpwrak, that time to fix EMC. :-)02:27
xiangfuwpwrak, got it. thanks.02:27
wpwrakcladamw: so i guess it's safe to make all the TH pads bare02:27
wpwrakcladamw: (EMC) i'm not sure that process is actually reliable :)02:28
wpwrakxiangfu: most of the footprints are quite boring. things get more interesting when you go to QFN parts, with solder paste islands02:31
wpwrakBGAs get solder paste, right ? or are the balls already enough and the pads are bare ?02:32
xiangfuwpwrak, QFN, no 'foot' but only 'footprint' :-)02:34
wpwrakcladamw: what do we do with parts that have TH and SMT ? e.g., usb_a_plug_smt.fpd02:51
wpwrakcladamw: solder paste everywhere ? solder paste only on the SMT pads ?02:51
cladamwwpwrak, pad[1::4](SMT) with solder paste, the other two pad[5::6](TH) are not necessary. :-)02:55
wpwrakatusb has that critter ... it actually looks as if they soldered the whole connector manually :)02:57
cladamwbut like i said, if you make all 6 pads with solder paste, it won't be trouble when making stencil. :-) but if we want to define well. then we define rule for pad[5::6] :-)02:57
cladamwyes, i saw atusb samples here are mostly like human soldering not to use a through hole reflow machine. :-)02:59
wpwrakyeah, that connector is a tricky beast. it gave us a bit of headache. if you include it in the SMT process, you may as well have solder paste on all pads, including the TH ones.02:59
wpwrakhowever, if you do it manually, you wouldn't want solder paste in the holes ...02:59
wpwrakfor atusb, that made a lot of sense. it's just one silly little component anyway. nothing like the TH orgy that M1 is ;-)03:00
cladamwwpwrak, sure, one won't make/face a hole (aperture) as solder paste. but for speeding up process in manufacturer, stencil maker works for it. :-)03:01
wpwrakvery well. transformed :)03:07
qi-bot[commit] Werner Almesberger: modules/usb-a-pcb.fpd: change pad type from "normal" to "bare" (master) http://qi-hw.com/p/kicad-libs/faedf5603:12
qi-bot[commit] Werner Almesberger: modules/: make pads of all fully through-hole components "bare" (master) http://qi-hw.com/p/kicad-libs/00711fd03:12
qi-bot[commit] Werner Almesberger: modules/pads.fpd: make also square and rectangular pads "bare" (master) http://qi-hw.com/p/kicad-libs/9be533b03:12
qi-bot[commit] Werner Almesberger: modules/usb_a_plug_smt.fpd: mixed TH/SMT; change the two TH pads to "bare" (master) http://qi-hw.com/p/kicad-libs/c0e33e503:12
qi-bot[commit] Werner Almesberger: modules/meander-2450MHz.fpd: make all pads "trace" (regression) (master) http://qi-hw.com/p/kicad-libs/cbfba8303:12
wpwraki just try to imaging how much fun doing all this would have been with kicad's module editor ...03:13
cladamwthis case may not be easily proposed out in kicad list as targets may be not the same. 03:16
wpwrakyeah, i think kicad's module editor implicitly makes pads with a hole "bare". so you couldn't run into this specific problem.03:18
viricwpwrak: yes, m1 looks quite flexible :)07:12
whitequarkinternet is censored in russia now12:14
wpwrakso that's good news ? now you're allowed to say openly that there's censorship ?12:31
whitequarkoh, not exactly12:33
whitequarkprior to the law which has passed, sites sometimes were closed, but that was rare, exceptional, and had nothing to do with ISPs/registrars12:34
wpwrakah, i see. you're still not allowed to admit that you weren't allowed to talk about such things in the past :)12:34
whitequarknow this is common, and ISPs/registrars must oblige and have a requirement for filtering hw.12:34
Action: whitequark shrugs12:34
wpwrak(rare, exceptional, ...) like a visit from the secret police ? :)12:34
whitequarkwolfspraul: remind me, which VPN service do you use?12:34
whitequarkwpwrak: more like a visit from copyright holders12:35
whitequarkit was the single most common reason12:35
whitequarkthere were only few cases of politics-related stuff removal, and, well, with the Streisand effect that wasn't effective anyway.12:36
whitequarke.g. there were one or two blogs removed from livejournal.com, and frankly nobody actually knew about them before12:37
whitequarkbut yeah, stopping delegation of russian thepiratebay has caused some resonance12:37
whitequarkin a ~day it migrated to .org domain.12:37
whitequarkwell, what could you expect from a country where a nationwide social network Vkontakte basically hosts the biggest archive of pirated material.12:37
whitequarkin case you are curious, the law has the number 89417-612:40
whitequarkoh, and of course the alleged reason is child pornography. in the last two weeks there were more discussions about that than in previous ten years combined.12:42
whitequarke.g. someone said that Wikipedia and Yandex [biggest russian search engine] are driven by, quote, "child pornography lobby".12:42
wolfspraulwhitequark: I use my own with openvpn, tunneling to a 7.90 EUR / month hetzner virtual machine12:44
whitequarkyeah, sounds sane. I'll make the same setup, I think12:45
wpwrakwhitequark: i sometimes wonder if child pornography wasn't invented specifically for the purpose of censorship :)13:02
qi-bot[commit] Werner Almesberger: dsv/dsv: add support for PDF embedded in ZIP (master) http://qi-hw.com/p/eda-tools/e5e67dd14:31

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