Topic for #qi-hardware is now Copyleft hardware - http://qi-hardware.com | hardware hackers join here to discuss Ben NanoNote, atben / atusb 802.15.4 wireless, and other community driven hw projects | public logging at http://en.qi-hardware.com/irclogs
rz2k has quit []
wej has quit [Ping timeout: 260 seconds]
wej has joined #qi-hardware
wej has quit [Ping timeout: 260 seconds]
wej has joined #qi-hardware
LunaVorax has quit [Ping timeout: 260 seconds]
MistahDarcy has quit [Ping timeout: 255 seconds]
MistahDarcy has joined #qi-hardware
urandom__ has quit [Quit: Konversation terminated!]
<qi-bot>
[commit] Werner Almesberger: experimental directive "allow holes" to allow multiple holes per pad (master) http://qi-hw.com/p/fped/bb055fa
<qi-bot>
[commit] Werner Almesberger: Merge branch 'master' of projects.qi-hardware.com:fped (master) http://qi-hw.com/p/fped/1bc94d0
<wpwrak>
qi-hardware, meet the dark side of user friendliness ;-)
<kristianpaul>
actually the www subdomain seems to be the problem
<kristianpaul>
:-)
<wpwrak>
yeah, without it also qi-hardware.com works
<kristianpaul>
who uses www those days! ;-)
<wpwrak>
right, all the cool kids use gopher :)
<kristianpaul>
lol
<qi-bot>
[commit] Werner Almesberger: genkicat: new option "-t title.ps" to include a title page file (master) http://qi-hw.com/p/eda-tools/560bf9d
MistahDarcy has joined #qi-hardware
<qi-bot>
[commit] Werner Almesberger: add title pages to the catalogs, with useful information like the creation date (master) http://qi-hw.com/p/kicad-libs/e324adc
<DocScrutinizer05>
I'd probably place a lightning catcher next to the upward beam though, so that maybe the mirror could survive strikes
<xiangfu>
DocScrutinizer05: then. it's danger and looks like a big weapon. :-)
<DocScrutinizer05>
see 5:40
<DocScrutinizer05>
pulsed with peaks of 5kW
<DocScrutinizer05>
says the german text
<DocScrutinizer05>
I'm much in love with this monster
<DocScrutinizer05>
xiangfu: big weapon for pickup truck mount, yes ;-D
<DocScrutinizer05>
alas: 20min pre-heat
<DocScrutinizer05>
great weapon for mosquito killing with stereo camera & low-power red laser wall positioning and then taking those parasites down with a single 5kW pulse
<DocScrutinizer05>
also a plausible threat to keep *everybody* out my flat that I don't want to come in. "DANGER! LASER RADIATION! Wear eye protection. DO NOT TOUCH THE BEAM!"
<DocScrutinizer05>
then place 3 pairs of 0.99$ cheap protective goggles on nails under that big scary sign, and be sure *nobody* will ever dare to use them and enter the room without prior allowance
<wpwrak>
DocScrutinizer05: "Fasereinkoppelung vorhanden und ok." ... so that's what they use on star trek :)
<DocScrutinizer05>
hehe
<DocScrutinizer05>
I guess the whole device been originally used in medical skin treatment
<DocScrutinizer05>
to remove tattoos etc
<DocScrutinizer05>
the LCD display shown in that youtube video suggests that
<viric>
usually, 10mW max, and we are very careful on what we do with them
<GNUtoo-desktop>
ok
<viric>
we focus them either with protective glasses, or seen through a camera/screen
<viric>
we know people which work with the same lasers with less care, that have gone through doctors for problems in the eyes
<viric>
I'd be very very careful with a 200mW laser.
<DocScrutinizer05>
>>The power of this laser is no joke, at 1W the Spyder III Pro Arctic is a Class 4 Laser. Do not allow unprotected eye exposure to diffusely scattered light from the beam terminating on a wall, the floor or other flat surface from this laser closer than 3 feet.<<
Mistah_Darcy has joined #qi-hardware
<DocScrutinizer05>
plain speech: a blue laser of 2W is highly dangerous not only when staring into the beam but also when watching the projection spot on the wall
<DocScrutinizer05>
I'm just looking at the spot of my 5mW blue on the wall, ~5ft away
<DocScrutinizer05>
maybe 4
<DocScrutinizer05>
already creates scary optical artifacts in the eye
MistahDarcy has quit [Ping timeout: 250 seconds]
<DocScrutinizer05>
the spot looks much larger than it actually is
<DocScrutinizer05>
I don't think I like blue lasers that much
LunaVorax has joined #qi-hardware
<DocScrutinizer05>
the green allegedly 150mW is a killer in comparison, but feels less nasty nevertheless
<DocScrutinizer05>
I'd not dare to look at the spot if this was a blue one
GNUtoo-desktop has quit [Quit: [INFO] fsogsmd : received signal -11, exiting.]
<kristianpaul>
xiangfu: blinking with a counter !
<kristianpaul>
nice ;)
<xiangfu>
kristianpaul: Hi
<kristianpaul>
hola :)
<xiangfu>
that is for test clock is works fine.
<kristianpaul>
nice
<xiangfu>
kristianpaul: reballing: I tried on reballing.
<xiangfu>
kristianpaul: so far I think there are one problem is the soldering paste HAVE TO be good.
<xiangfu>
kristianpaul: so there is no good reballing chip so far
<xiangfu>
I already bought 2 branch soldering paste, one from Hongkong. another from Taiwan.
<xiangfu>
no time try reballing again recently. I will update the wiki once I try reballing again. now I have quiet some chips for reballing.
<xiangfu>
the best should be metal mask excatly same size as chip.
<kristianpaul>
how do you align the chip for the balls not to be unaligned when in the oven?
<kristianpaul>
i saw a video and the chip begun to "fall" in one side then the other
<xiangfu>
align before put then to oven.
<xiangfu>
when there are tin on your pcb bga footprint(all the factory pcb have tin but home-made-pcb)
<xiangfu>
things get eaier. even your chip a little bit shift. when reflowing the chip will automatic fit to footprint.
<xiangfu>
home-made-pcb have a lot of trouble.
<kristianpaul>
ah shift is fixed with more reflow good to know
<kristianpaul>
trouble, can imagine :-)
<xiangfu>
1. the oxide layer on home-made-pcb
<xiangfu>
2. align the chip. there is no outline of chip. but you can do some trick when pcb etching. add some cutout zone fit the chip outline.
<xiangfu>
s/there is no outline of chip/there is no slik layer on home-made-pcb
<qi-bot>
xiangfu meant: "2. align the chip. there is no slik layer on home-made-pcb. but you can do some trick when pcb etching. add some cutout zone fit the chip outline."
<xiangfu>
(shift is fixed) very tiny shift. not like shift one big bga ball. :)
<kristianpaul>
had you tried kapton tape?
<kristianpaul>
just asking not sure if hold the chip with a tape will actyally help..
<wpwrak>
xiangfu: you can tin your DIY PCBs, too
<xiangfu>
wpwrak: yes. I tried that. but I didn't get a good result.
<xiangfu>
wpwrak: I use the SMT solder. then hot air gun.
<wpwrak>
xiangfu: just heat up your iron to the maximum, apply flux to the PCB, then distribute a few drops of tin and spread them out
<wpwrak>
xiangfu: naw, just regular solder and the iron. no hot air or solder paste needed.
<xiangfu>
wpwrak: thanks for that. I will try that in next 2 days.
<wpwrak>
you can still get relatively thick (maybe 0.1 mm) solder deposits on isolated pads but it's usually very thin.
<xiangfu>
wpwrak: after tin my pcb. the align get more difficult. I guess my hand-tin not that smooth. :)
<wpwrak>
if you get too much solder on, say, an IC pad, you can "paint" it on an adjacent trace (if there's one)
<wpwrak>
with isolated pads, it's messier. you can use desoldering wick to remove excess tin, but you have to be very careful not to damage the pads.
<xiangfu>
("paint" it on an adjacent trace) most of them don't have trace. :(
<wpwrak>
of course, if you damage an isolated pad, it often doesn't matter ;-)
<wpwrak>
yes, BGA is a bit more demanding :)
<wpwrak>
factory-made pcbs also have a solder mask, which should help with BGAs.
<xiangfu>
(damage the pads) yes. when I put the iron to the isolate pad > ~3~5 secs. the pad will just get off from pcb.
<xiangfu>
(solder mask) yes.
<wpwrak>
but in general, i think they're a bit too much trouble for manual soldering. particularly since you can't reach (examine or fix) the balls in the middle.
<kristianpaul>
xray :-)
<kristianpaul>
;-)
<wpwrak>
(damaged pads) exactly :) in some cases it helps if you decrease the temperature. in other cases, it helps to increase it :)
<wpwrak>
decrease: if your iron is way too hot.
<wpwrak>
increase: if the iron is near the melting point of the tin. then it takes a lot of time for the tin to melt and the overall amount of thermal energy is actually larger than what you'd need with a hotter iron.
<xiangfu>
wpwrak: for bga: 1. soldering paste(flux). 2. align bga chip 3. oven 4. I have some probe-pad for test inner connection 5. luck. :)
<wpwrak>
the temperature also depends on how much heat is conducted away, e.g., to connectors, ground areas, or other large structures
<xiangfu>
wpwrak: are you using 'syringe' when your pase other pads? like 0402/0603/0805 etc.
<wpwrak>
so if you have a particularly nasty spot with a lot of ground fill nearby, it sometimes helps to heat up the iron a little
<xiangfu>
wpwrak: yes.
<wpwrak>
i normally solder things manually, without paste
<wpwrak>
if i have to use paste, then yes, i have a syringe (the paste already comes on one)
wolfspraul has quit [Quit: Lost terminal]
<xiangfu>
wpwrak: ok. I cannot even buy such thing(syringe + tin inside) in Beijing. at least I cannot find them on taobao or other web-shop.
<wpwrak>
(have to use paste) for things like center pads in QFN (if i even bother to connect them) or for SMT crystals. crystals are a problem for "traditional" soldering because the tin, applied from the side, is likely to stick to the metal case, so you end up shorting things.
<xiangfu>
good thing is I don't have much components. no qfp(0.5mm pitch)
<wpwrak>
"no qfp" ? i thought what you had is 0.5 mm qfp ?
<xiangfu>
(crystals) once I put too much paste. it also get short.
<xiangfu>
only bga (xc6slx9 ftg256) now.
<xiangfu>
so far I never try qfp + home-made-pcb. since I think the 0.5mm even harder compare to bga 0.8pitch.
<xiangfu>
my printer(600 DPI) give the qfp pins footprint very very close.
<xiangfu>
I think when I do soldering: qfp-short problems > bga-short problems
<wpwrak>
qfp sound be much easier to fix. i don't think i've had 0.5 mm yet, but 0.65 mm SSOP is very easy
<wpwrak>
if you have a short, just apply flux and pass over the pins with the iron again. either "paint" the tin to adjacent traces or just have it stick to the iron.
<wpwrak>
ah, and i solder them with the iron and regular tin, too. no solder paste or reflow oven.
<xiangfu>
how many pins?
<wpwrak>
for reflow, i think you need to make a stencil. otherwise, you'll have problems with the amount of solder paste.
<wpwrak>
(pins) small ones, up to 28 pins (ssop). but the number of pins doesn't really matter.
<xiangfu>
(stencil) yes. agree.
<wpwrak>
and QFP is just the same as SSOP.
<xiangfu>
yes. for hand-soldering. better is the QFP.
<xiangfu>
wpwrak: I will take a picture of my qfp144 etching pcb tomorrow. those pins are really close here.
<xiangfu>
btw. I think we have some problem on our bga.fpg.
<wpwrak>
ah, the dual FET in the ledtoy has a 0.5 mm pitch. the package is a bit of a hybrid between SSOP-like pins and QFN (i.e., very short pins)
<wpwrak>
and soldering that was no trouble at all
<wpwrak>
(bga.fpd problem) yes ?
<DocScrutinizer05>
((tinning homemade PCB)) do galvanic!
<xiangfu>
the pad size is 0.6mm. we edit the pad size base on the ball size.
<xiangfu>
but the pad size should be < 0.6mm.
<wpwrak>
DocScrutinizer05: how ? :)
<DocScrutinizer05>
there are galvanic baths commercially available to tin cover PCB
<xiangfu>
if I don't remember wrong there is a documennt on cx6slx9 chip bga footprint.
<xiangfu>
the document have the correct pad size.
<wpwrak>
DocScrutinizer05: and how do you connect all the traces ?
<DocScrutinizer05>
you don't do electrogalvanic
<wpwrak>
DocScrutinizer05: ah, i see
<wpwrak>
DocScrutinizer05: i think the chemical tinning approach is somewhat nasty. unpleasant and difficult to find witchbrews.
<DocScrutinizer05>
results are maybe sub-optimal, but worth testing
<DocScrutinizer05>
maybe silver coating even does better than tin
<xiangfu>
wpwrak: and the chip outline also wrong I think.
LunaVorax has quit [Read error: Connection reset by peer]
<wpwrak>
xiangfu: lemme check the references ...
rz2k has joined #qi-hardware
<wpwrak>
hmm, no footprints in the altera and xilinx references
<wpwrak>
xiangfu: using NXP's rules, the pad size should correspond to the nominal diameter
<DocScrutinizer05>
basically what those Siemens guys told us at OM was: the footprint libs are one of the core assets of a company usually. Each company has its own, and usually you don't want to use those that come from chip manufs. But ours are great and usable of course
<DocScrutinizer05>
iirc
<xiangfu>
wpwrak: I think we better use NOM
<xiangfu>
(the pad size should correspond to the nominal diameter) yes. agree. I will update the bga.fpd.
<roh>
hm. wasnt the size of the hole in the solder mask the defining variable to how much solder paste gets onto a pad?
<wpwrak>
DocScrutinizer05: (ours is great but all others suck) what a refreshingly original concept ;-)
<xiangfu>
DocScrutinizer05: (Each company has its own) I belive that.
<roh>
also it seems that these 'sizes' are all just recommendations.. not 'things that should work' ... from what ive heard one tunes these more often than what manufs try to admit
<roh>
btw... if one does make telco equipment, one can use lead based solder :) .. no crappy solder problems
erikkugel1 has joined #qi-hardware
<xiangfu>
wpwrak: what about the outline? should we make it a little bigger?
<wpwrak>
let's see what NXP say ... package is 4.5 +/- 0.1 mm
<xiangfu>
wpwrak: I used the wrong value on ftg256 and csg324. the 484 is right.
<wpwrak>
"occupied area" is 4.9 mm (squared, of course). so if outline == occupied area, yes, it should be a little larger. if outline == ideal location of package edges, the nominal or the max value would probably be best.
<qi-bot>
[commit] Werner Almesberger: title page: always generate title.ps when generating the catalog (master) http://qi-hw.com/p/kicad-libs/21513c1
<qi-bot>
[commit] Werner Almesberger: include common/Makefile after "all", so that it won't override the default (master) http://qi-hw.com/p/kicad-libs/2e07473