#qi-hardware IRC log for Saturday, 2013-12-07

paul_boddiewpwrak: Yes, that's where I saw it. Wonder which exploit they're trying for!01:23
larsc2none09:00
larscthey are trying to post a random comment09:01
nicksydneyhello all10:28
nicksydneywpwrak: what's happening over there :)10:28
wpwrakthe evil daystar just woke me10:36
wpwraklarsc: at least they're bold. trying that right in the midst of the lion's den.10:37
nicksydneyhahahah10:41
wpwraknicksydney: how's the lab setup going ? getting ready to roll ? :)10:44
nicksydneyyeah....got the printer and the paper....jaycar closed when i reach there so didn't get the PCB kit10:44
nicksydneyi'm going to try to print using the photo paper i bought10:45
wpwrakthat pcb kit you mentioned wouldn't be much use anyway since it's for photochemical10:46
wpwrakif you have printer and paper, all you need are some copper-clad pcbs, an acid, and some containers that don't melt from the acid and that have a flat bottom10:47
nicksydneywhat kind of acid ?10:48
nicksydneyi will get the PCB from that same store that sell the PCB kit10:48
nicksydneythey sell copper pcb10:48
nicksydneythis is what i got today http://ctrlv.in/26728510:48
wpwrakfor the container, i'd say tupperware but if possible, i'd get something made of glass instead of plastic. more resistant :)10:50
wpwrakvery good :)10:50
wpwrakacid: if you can get HCl (hardware store, pool supplies, etc.) and Peroxide (pharmacy), then you're good10:50
nicksydneymade of glass...hmmm.....maybe some kind of microwaveable dish that normally used for lasagna :)10:50
nicksydneyHCl -- what's common name if i go to hardware store ?10:51
wpwrakthere are many other acids people use but that is about the easiest one10:51
nicksydneyhttp://www.bunnings.com.au/search/products?q=Hydrochloric%20acid10:52
wpwrakHCl = muriatic acid10:52
nicksydneyis that the one ?10:52
wpwrakyes10:52
wpwrakwhitequark: see, if you need some, you can just fly over to australia :-)10:52
nicksydneyok will get this one .. http://www.bunnings.com.au/bondall-1l-hydrochloric-acid_p096023210:53
nicksydneyonce i get it how should i use it ?10:53
nicksydneythis is interesting "Bondall Hydrochloric Acid is the solution for acid etching concrete, brick and pavers to adequately prepare the surface prior to sealing or painting. It is also great for cleaning and polishing brick or metal and adjusting the pH level in pools." .... now they should add also for etching PCB :)10:53
wpwrakyeah, don't get more than 1 l. even half a liter would last for a very long time. do you have a place outdoors where you can store it ? garden, shed, balcony, ... ?10:54
wpwraka place where kids and pets won't get to it10:54
nicksydneyyeah...i can keep it outside in the garden 10:54
wpwrakperfect. the bottle will give off a little bit of chloride gas. enough to corrode metals if it's locked up together with them. that's why it's best to keep it outside, so the gas gets just vented off10:55
wpwraknext, you'll need H2O2, peroxide. you should get a small bottle of ~5% at the pharmacy. if you can in addition find something with a higher concentration (up to 30%), even better10:57
wpwrakthe weak one is enough to make a suitable solution. but it gets better: you can reuse the acid almost indefinitely, but you have to add small amounts of HCl and peroxide from time to time. especially the peroxide escapes relatively quickly.11:00
nicksydneyhttp://www.chemistwarehouse.com.au/product.asp?id=543911:01
wpwrakfor "topping off" the etchant, peroxide with a higher concentration works best, since it adds less water and thus the brew doesn't get diluted so much11:01
wpwrakyeah, that's good11:02
wpwrakmaybe get some of this, too: http://www.chemistwarehouse.com.au/product.asp?id=69464&pname=Hilift+Peroxide+40+VOL+12%25+200ml11:02
wpwrakmixing is about two parts 3-5% peroxide and one part HCl11:03
wpwrakyou need just enough of both that both side the of the PCB will be covered. so a layer of about 2-3 mm at the bottom of the "tupperware"11:04
wpwrakregarding the "tupperware", make it the side of your typical PCB, which will generally be small. something like 6 x 6 cm may be enough. no need to go bigger than 10 x 10 cm. the bottom should be perfectly flat. and it should have a lit for airtight sealing (since you'll want to keep that acid for reuse)11:06
wpwraknote that the HCl+H2O2 combo doesn't need heating. some other acids do, especially FeCl. so if you find online articles explaining how to balance all the mess on your oven, just ignore them :)11:07
wpwrakah, and while you're at it, get discardable gloves. latex or similar. they usually come in boxes of 100 units or so.11:08
wpwrakregarding the mixing: mix the solution before putting the PCB. then stir it, e.g., with a wooden stick (skewer or such)11:09
wpwrakif you need to add anything with the PCB already in the bath, then do that as far as possible from the PCB and stir immediately. else, you may get spots where the etchant is a lot more potent than elsewhere, which can produce highly uneven etching11:10
wpwraklots of details but it's not really all that complicated :)11:11
nicksydneyok let me write a summary for myself 11:11
wpwrakah, when done etching, just grab the PCB with your gloves fingers, wipe the etchant off the PCB with the stick, then wash off the remaining acid with tap water11:12
nicksydneyif i want to throw the HCL+H2O2 combo can i just throw it in the garden ?11:13
wpwrakthe quantity of acid that's left is very low, so it won't attack anything. first rinse, then take off the gloves and rob the board with your fingers to get rid of the last bits. you want to make sure the board is completely acid-free when done. (else, it would continue etching away, eventually destroying your board)11:14
wpwrakonly if you hate your garden and what lives in it ;)11:14
wpwraka better approach is to let the open container sit in the sun for a while, so that all the water can evaporate. then you're left with blue-green crystals. you can scrape them out, wrap them watertight, and discard them as solid garbage.11:15
nicksydneyhahaaha...ok got it11:15
nicksydneyok understand11:15
nicksydneyexplain to me bit more about " 3-5% peroxide and one part HCl"11:16
nicksydneylet's take the measurement of the cap of the boottle11:16
nicksydneybottle11:16
wpwrakbut as i said, that etchant can be reused. so disposal is usually a very rare event.11:16
wpwrakyes, any small container will do. measure the quantity you want, then drop it into the "tupperware". then add the other11:17
wpwrakmany sources recommend doing this with the perixode first, then the acid. to avoid dropping things into a concentrated acid. not sure if it matters.11:18
wpwrakin any case, don't spill anything (and if you do, wash it off with a lot of water, then apply some soapy detergent - usually alkaline - over the affected area to neutralize the acid)11:19
wpwraklemme snap a picture of my setup ...11:19
nicksydneycool11:19
nicksydneypicture worth thousands word :)11:19
wpwrakhmm, in half an hour. camera commands me to change thebattery pack :-(11:21
nicksydneyok no worries......do I need to add water into the container when I combine HCL and H2O2 together ?11:23
wpwrakthey already come with the water. HCL is only 35% and the peroxide even much lower. all the rest is good old water.11:25
wpwrakyou'd only need to worry about adding water if you used a very strong concentration of peroxide. in theory, peroxide could go up to 100% (or almost). but at some point it'll just spontaneously explode :)11:26
wpwrakperoxide is an interesting chemical. starts as desinfectant. then you can use it to bleach hair. then you can use it to burn off layers of skin. then you can use it as rocket fuel. and then you get an unstable explosive :)11:27
nicksydneycrazy11:27
nicksydneyok must be very careful in handling it11:27
wpwrakat 3-5% you're in the desinfectant range. at 10% you'd be in the bleaching range. 30% and it's the burn off of skins (at that point, gloves are no longer just a precaution :)11:28
wpwrakand unless you have good friends at the next missile base, you may not find it easy to go beyond 30% :)11:28
nicksydneythis one http://www.chemistwarehouse.com.au/product.asp?id=69464&pname=Hilift+Peroxide+40+VOL+12%25+200ml is only 12% so it should be ok to handle11:29
wpwrakalready 30% may be difficult. also because it's used to make drugs (all, what isn't ?)11:29
nicksydneyi just opened the paper i bought...i think i make a mistake....it's a photo paper 11:30
nicksydneyit's white..not transparent11:30
nicksydneythought it is called photo paper as back in mind thought it will be glossy and transparent..but looks like it's not11:34
wpwrak(12%) yes, it's hair bleach. so wear gloves and don't splash it around. but it won't kill you.11:34
wpwrakthe paper is good for toner transfer11:35
wpwrakremember the two different processes ? toner transfer and photochemical ? toner transfer is laser printer, white paper, ironing11:35
nicksydneyok i thought that normally you use the transparent paper http://www.ebay.com/bhp/overhead-transparencies11:37
wpwrakthat would be for photochemical. photocemical is (any printer) transparencies, photo-sensitive coating, UV light11:37
wpwrakbut we already discussed this at length yesterday or so, the irc logs are your friends :)11:38
nicksydneyok i take a look at the log .... very bad with words...i learn faster with pictures :)11:39
nicksydneyok will do printer testing with the new paper and see if this is going to work11:41
wpwrakah, here's a nice collection of peroxide for hairdressers: http://articulo.mercadolibre.com.ar/MLA-484877319-oxigenada-alfa-x-1-litro-_JM11:43
wpwrakalso shows the "volume" notation vs. percent. so 10 Vol is 3%, etc.11:43
wpwrakthat scale sensibly cuts off at 100 Vol = 30% :) (30% comes in a pitch-black bottle, with warning sings)11:45
wpwrak(for peroxide doesn't like sunlight)11:45
nicksydneywohooo...successfully printed on the paper...uploading the pics hang on11:56
nicksydneywhen you print it ...what is the scale ? 11:59
wpwraki think you can guess that ;-)12:00
nicksydneythis is what i have --> http://snag.gy/jvEf2.jpg 12:02
nicksydneythis is the output on the photo paper --> http://ctrlv.in/267307 12:04
wpwrakalmost perfect12:05
wpwrakyou also need to click "mirror"12:05
wpwrakthat is, for the front layer. the back layer will be mirrored by default12:06
wpwrakyou can tell whether it's mirrored by looking at text. if you can read if normally, it's wrong. (that why i always put some text on both sides)12:06
nicksydneyok...the measurement of the board is 3cm wide is that right ?12:07
nicksydneydo i need to change anything on the 'Approx. Scale' ?12:07
wpwrakjust keep accurate scale. no idea what "approx scale" is supposed to mean. sounds scary.12:08
nicksydneylooks like something is offf...it's not printing the round wheel 12:12
wpwrakah the arc12:13
wpwrakah yes, that's because you used "print"12:14
wpwrak"print" doesn't work. you have to use "plot" ,make a postscript, then print that postscript12:14
nicksydneyok need to find out way to print the postscript 12:16
nicksydneywpwrak: how do you normally do  ?12:22
wpwrak"lpr filename" :)12:22
nicksydneyusing plot got few files with extension .gbl, .gbo, .gbr,etc12:23
nicksydneybut no .ps file12:23
wpwrakthat's because you didn't tell it to generate postscript12:24
wpwrakdefault is Gerber, hence the *.g*12:24
nicksydneymanaged to generate .pdf f12:28
nicksydneyis it correct that the size of anelok is like 3cm ?12:29
wpwrakthe pcb is 25 x 80 mm12:29
nicksydneybloody hell.....even my pinky size is double the size of the freescale processor12:30
nicksydneyhow in the hell i'm going to solder this :)12:31
nicksydneyhahahhahaha12:31
wpwrak;-)12:31
wpwrakwith a steady hand and a lot of flux :)12:31
nicksydneybasically you dip the pcb into the flux :)12:33
wpwrakwell, that may actually work :)12:35
nicksydneythe processor is it LQFP or QFN ?12:36
wpwrak48-QFN, 0.5 mm pitch12:37
nicksydneygoing to have a hard time soldering those if i manage to get a hand on that12:42
wpwrakthe actual soldering isn't the diffcult part. what's tricky is the positioning.12:44
wpwrakit's kinda difficult to align the pads with the traces. and since it's a large package, it's also possible to get confused and have it off by one position.12:45
wpwrakso it's not the errors in the 0.1 mm range (general imprecision) that should worry you but the ones in the 0.25-0.5 mm range (caused by proper alignment, but with the wrong trace or with the space between traces)12:47
nicksydneymaybe should try to rework the pcb and make it one sided only to make it easier for me to try it out12:50
nicksydneyi'll try to do that for my exercise12:54
nicksydneywhen you say you do 'lpr filename' what is the file that you print...is that the .ps file ?12:54
wpwrakhmm, i'd recommend starting with a new design. just oled, mcu, and touch sensor12:54
wpwrakyes, the .ps12:55
wpwrakoh, assuming you have a postscript interpreter in the path12:55
wpwraknot sure what happens with non-ps printers. i'd expect cups to take care of it.12:55
wpwrak(i only have ps printers. never worry about printer drivers :)12:56
nicksydneyit generate few .ps files12:56
nicksydneywhich means i need to combine those files together ?12:56
wpwrakjust print  anelok-F_Cu.ps and anelok-B_Cu.ps12:57
wpwrakthe others are for other layers. layers you don't really need.12:58
nicksydneywpwrak:  i have the board in this picture http://downloads.qi-hardware.com/people/werner/anelok/tmp/brd0-d1-top.jpg .. KL25Z13:06
wpwrakperfect13:06
nicksydneywpwrak: but it doesn't work in Linux .. need to flash a new bootloader as the current bootloader does not work with Linux..did you have to do the same with yours ?13:07
wpwrakthen all you need is oled, the few caps and resistors for the oled, and the touch sensor13:07
wpwrakah yes, i did that. wasn't painful.13:07
wpwraksee anelok/pgm/README13:08
wpwraksection "Updating/installing the P&E firmware"13:08
nicksydneyyes i'm reading that13:15
nicksydneythe painful part is finding Windows box13:15
nicksydneywpwrak: if you look at this picture ... the board with anelok printout (http://ctrlv.in/267322) it shows that anelok is very small but if you look at your picture  ( http://downloads.qi-hardware.com/people/werner/anelok/tmp/brd0-d1-top.jpg ) the size is almost the same...which makes me wonder whether the file i'm printing is correct in terms of size or i'm missing something13:18
rjeffriesxiangfu do you know what happened to remaining stock of Ben Nanonote? Thank you.14:15
nicksydneywpwrak: ok i'm bit lost here ..according to this product brief (http://www.farnell.com/datasheets/1691542.pdf) the MCU size is 7mm x 7mm while I measure thr printout the size for the MCU is aaround 3mm ... so there must be something wrong in the print out14:29
rjeffriesnickydney if you munge wpwrak 's practical, step-by-step advice re: making PCBs into a guide, you (and he) could publish and help many people, including me. 14:38
nicksydneyrjeffries: wpwrak has published a tutorial about making PCB 14:40
rjeffriesyet you need to ask many questions, so it woudl seem the tutorial needs work. LOL just sayin'14:43
nicksydney:)...nah sometimes there are things that does not work according to documentation. .. so need to confirm 14:46
rohyep. your printout looks wrong size14:49
rohmaybe a metric/"ยง$%& problem.. maybe just factor 214:49
nicksydneytried printing using File --> Print and selected 'Scale 3' from 'Approx. Scale' and I can get measurement for the MCU at 7mm x 7mm15:07
rohmeasure the longest side of the pcb15:10
rohbiggest structure. the mcu is too small for a proper scale check15:10
nicksydneythanks roh...will try again tomorrow....15:21
nicksydneyheading off to la-la-land :)15:21
wpwraknicksydney: it does look a little small. but ... where did you get the MCU size ? the square pad isn't 7x7 mm. it's only about 5x5 mm16:24
wpwraknicksydney: in any case, in "plot", remove the check mark from "Exclude PCB edge layer from other layers". that way, you get the pcb border (you'll want this for positioning later on, too)16:25
wpwrakthen you can measure the board dimensions. should be 25 x 80 mm16:25
wpwrak(i kinda wonder whether you really used "Plot" ... i don't even see PDF there, so it seems a bit surprising that you ended up with PDF insead of PS. well, maybe it's a difference between our versions. but still, strange)16:27
wpwraknicksydney: here's finally my etching setup: http://downloads.qi-hardware.com/people/werner/tmp/acid-box.jpg17:59
wpwrakall is is a plastic box. so i can move it out of harm's way. the small container (white lid) is what i use for most things. it's 10 x 10 cm. it sits in a larger glass container that catches any spills. not that spills would be very common - and when they happen, i just let them dry.18:01
whitequarkyeah distortions are the worst :/18:07
whitequarkI've once had my printer (driver?) add a nonlinear distortion somehow18:07
whitequarkthat was painful to debug18:08
wpwraksounds suckish, yes18:24
wpwrakkewl. case snaps closed and wheel actually can be spun. first time ever :)19:15
wpwraknow i just wish the OLED wasn't dead ...19:15
nicksydneywpwrak: this is where I get the MCU size dimension ... http://www.farnell.com/datasheets/1691542.pdf .. page 822:06
nicksydneywpwrak: this is what I'm seeing when I'm using File --> Plot .. http://ctrlv.in/267508 (top left saying PDF)22:08
nicksydneywpwrak: this is my another attemp using 'Exlude PCB...' ... you see the measurement is seems very off .. http://ctrlv.in/26751022:14
nicksydneywpwrak: could it be that the driver is the issue ?22:14
nicksydneyok be back later going out first22:14
wpwrak(pdf) ah, interesting. so they added PDF recently. that's nice.22:26
wpwrak(size) 7x7 mm is correct. but that's the size of the chip, not the size of the center pad. so unless you calculate the size of the chip from what you see in the printed layout, you'll get something incorrect.22:28
wpwrakah yes, that's rather small :)22:28
wpwrakthat's about 37.5 mm22:29
wpwrakoff by a factor 2.1322:29
wpwrakcan you generate PS directly ? maybe something goes wrong when going PDF -> PS22:30
wpwrakcould of course also be that the printer driver does weird things22:30
wpwrakat least your printer is good enough to handle even such tiny sizes. so you can keep up if i use smaller chips ;-)22:30
--- Sun Dec 8 201300:00

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