2012-05-15 00:08 (t500) well it's just good enough for me 2012-05-15 00:09 http://paste.debian.net/168773/ 2012-05-15 00:09 better old solid stuff than the newest hottest crap 2012-05-15 00:11 man that's some dose of tech pessimism here 2012-05-15 00:11 luckily I think most new tech is actually *far* better than the old one, so I'm not worried :-) 2012-05-15 00:13 you won't get a quality like T500 for 600 bucks new 2012-05-15 00:14 and actually it's terribly hard to find any decent anti-glare non-glossy display in a laptop nowadays 2012-05-15 00:14 computer prices are dropping and will continue to do so, driven by semiconductor economics as you know 2012-05-15 00:15 I would agree there is scary little innovation in notebooks, but we all know this for 10 years, no need complaining. 2012-05-15 00:15 let's do better 2012-05-15 00:15 hey btw, since we have the pessimists meeting here - will nokia go out of business? 2012-05-15 00:15 :-) 2012-05-15 00:15 who's complaining? 2012-05-15 00:16 I just stated I prefer used high quality 2 year old stuff to new stuff that's inferior quality for 4 times the price 2012-05-15 00:17 (taking diy here) ok :) 2012-05-15 00:17 wpwrak: but watch out we have the pessimists coming after us :-) 2012-05-15 00:17 for that - a sip of nice yunnan coffee... 2012-05-15 00:18 DocScrutinizer: I think I bought my last notebook, took it apart after buy and reassembled in new case, which was fun 2012-05-15 00:18 but then I switched to a screen with back-mounted thin-itx board and battery and yes, it works just fine 2012-05-15 00:18 wpwrak: wolfgang confusing sane pragmatism with pessimism 2012-05-15 00:18 :-) 2012-05-15 00:19 wpwrak: do you have any explanation why whitequark described his pcb-making in such dark words (months of fiddling, unstable, messy, eventually gave up), whereas what I saw at your place looked quite charming actually? 2012-05-15 00:20 what would help me if I would pick a realistic goal first - maybe atben clone? 2012-05-15 00:20 or you think that's too difficult 2012-05-15 00:23 DocScrutinizer: what's your take on the future of nokia? 2012-05-15 00:23 whitequark: i usually get my boards right one the first try :) they're small, though. i try keep things simple ;-) it only takes a lot of attempts if i try to rush things. haste makes waste, quite literally in this case ;-) 2012-05-15 00:23 DOOOOOOM 2012-05-15 00:23 I somehow think in the downfall there maybe a chance of a spinoff that could continue with a Linux phone 2012-05-15 00:23 (get right) in the sense of the PCB being what i had in the design. then i discover the flaws in the design .. ;-) 2012-05-15 00:23 but I'd say only 10% chance of that happening, more likely it will all be grinded down to nothing 2012-05-15 00:24 (spinoff) hard to see this perspective. I think they already fired all inhouse competence wrt linux/FOSS 2012-05-15 00:26 I expect Nokia becoming a M$-dependency 2012-05-15 00:26 whitequark: (scaling) in fact, it scales negatively. if you make a lot of boards, the error rate goes up :) 2012-05-15 00:26 of course, for scaling you send to a fab 2012-05-15 00:26 and for cost-down 2012-05-15 00:28 although I also think there are a lot of interesting pcbs with unique features a standardized fab cannot easily do, but that needs to be seen case by case 2012-05-15 00:28 hm? 2012-05-15 00:28 like what? 2012-05-15 00:28 I remember a long discussion with Andrey the Elphel founder about how he planned to make some special cuts into pcbs to be able to perfectly aline an image sensor with the lens 2012-05-15 00:29 (nokia) why the future tense ? ;-) 2012-05-15 00:29 and how he was unable to find any pcb maker willing to work with him on that, for months, until he eventually gave up 2012-05-15 00:29 wpwrak: :-P 2012-05-15 00:29 indeed 2012-05-15 00:29 wpwrak: please help! :-) which initial goal? atben clone? 2012-05-15 00:30 something else that is more interesting now? 2012-05-15 00:30 labsw? 2012-05-15 00:30 yeah, sounds good 2012-05-15 00:30 not too much too fine pitched stuff I assume 2012-05-15 00:31 wolfspraul: (atben clone) it's a little demanding on the ME side. that uSD shape ... depends a bit on how confident you are with a dremel 2012-05-15 00:31 I don't even have a dremel. need to get a number of tools. 2012-05-15 00:31 duh, so I mistaken sth 2012-05-15 00:31 my drilling machine broke yesterday because I abused it in cutting a line :-) 2012-05-15 00:32 well, in good chinese quality it didn't totally break, but at high speeds it sputters (stops for a few seconds). middle speeds still work :-) 2012-05-15 00:32 wolfspraul: i would pick something simpler. than atben/atusb. why not design your own circuit ? we have plenty of parts. i'd make something with an MCU, two LEDs, and a button. add USB if you want fancy. 2012-05-15 00:33 nah I can only motivate myself for something I actually need 2012-05-15 00:33 how about labsw? 2012-05-15 00:33 you dismissed anything with the xc6slx9 fpga as too difficult initially 2012-05-15 00:33 wolfspraul: RF has one nasty property: if you want it to work well, you need make sure grounding is good. now, good grounding makes the soldering harder, too, because all those large ground areas near your chips act as excellent heat traps 2012-05-15 00:34 an fpga/atrf board would be nice 2012-05-15 00:34 how about labsw? 2012-05-15 00:34 that could come in handy 2012-05-15 00:34 labsw is good but also a bit large. more like a 2nd stage project :) 2012-05-15 00:35 for the first try, the goal should be to make something where you can learn quickly. extra complexity only slows you down. 2012-05-15 00:35 :nod: 2012-05-15 00:35 atben clone 2012-05-15 00:35 once you're comfortable, you can move to more interesting things 2012-05-15 00:36 more interesting like doublesided ;-) 2012-05-15 00:36 atben = hairy ME + RF. that's not an easy project. also, the board is densely populated and your vias must be quite precise 2012-05-15 00:36 which is like one magnitude more demanding than singlesided 2012-05-15 00:37 oh, double-sided is good. particularly if the circuit is simple. so if you screw up, you can still fix the problems with patch wire. see if the rest works. 2012-05-15 00:38 the bottom side of atben is just gnd? 2012-05-15 00:38 honestly, start with a single or 1.5-sided PCB (one trace side plus GND backplane, so you can't mess up the via fitting) 2012-05-15 00:38 single-sided SMT is kinda boring. too easy ;-) 2012-05-15 00:38 I do see some traces there though... 2012-05-15 00:38 wolfspraul: it's as much ground as possible :) but yes, i needed a few traces, too. 2012-05-15 00:38 cladamw has joined #qi-hardware 2012-05-15 00:39 wow there is almost nothing on the board, right? 2012-05-15 00:39 the power only comes from ben? 2012-05-15 00:39 no converter, nothing? just from ben into the atrf chip? 2012-05-15 00:40 I think there's a mc as well? 2012-05-15 00:40 but a local crystal, I remember you tried without but the ben's clock was usable/stable enough? 2012-05-15 00:40 power comes from the ben. and yes, very few components. very densely packed. lots of ground all over the place. a lot of vias to drill. 2012-05-15 00:40 was not 2012-05-15 00:40 aah that's atusb I guess 2012-05-15 00:40 DocScrutinizer: atben and atusb have a crystal. remember the phase jitter issue ? 2012-05-15 00:40 yep 2012-05-15 00:41 (MCU) yes, that's atusb. atben is "dumb" 2012-05-15 00:41 I recall a shematics with two chips 2012-05-15 00:41 which I think was atusb 2012-05-15 00:42 atrf, 8 caps (0201), 1 resistor, 1 crystal, and then one other IC - what is that one? 2012-05-15 00:42 I should look at the schematics, open hardware here right? :-) 2012-05-15 00:43 wpwrak: btw what fixed the jitter issue finally? 2012-05-15 00:43 wolfspraul: the components are 0402. i didn't venture into 0201-land yet. most people find 0402 challenging enough ;-) 2012-05-15 00:43 LOL, indeed 2012-05-15 00:43 yeah, 0603 is already hard 2012-05-15 00:43 ah ok, yes 2012-05-15 00:43 I'm just too old for that shit 2012-05-15 00:44 0201 may actually not that bad. i mean, already 0402 is more by smell than by sight ... 2012-05-15 00:44 LOL smell 2012-05-15 00:45 I heard 0201 is oxidizing away from mere plain air, too terrible mass/surface ratio ;-P 2012-05-15 00:45 for experts: 0201, with a trace crossing under the component, and have a via in one of the pads. if you can pull that off with DIY, you rock ;-) 2012-05-15 00:45 DocScrutinizer: actually, I have some problems placing SMT components because my hands are shaking. not much, but enough to shift a 0603 horizontally by one position 2012-05-15 00:45 sure 2012-05-15 00:45 and I'm 19 and I never drank alcohol, nor do I commonly drink coffee 2012-05-15 00:46 and I'm not focusing anymore on things closer than 40cm 2012-05-15 00:46 so I guess that's implicit limitations of human beings 2012-05-15 00:46 a nogo for unarmed eye 2012-05-15 00:46 SMT soldering 2012-05-15 00:46 wpwrak: I still think atben clone is a good first target 2012-05-15 00:46 I need something I can use myself 2012-05-15 00:46 that or labsw 2012-05-15 00:47 uh, I'm somewhat lucky here because I don't focus on things further than 40cm :) 2012-05-15 00:47 anything fpga you outlawed, for now. ok. accepted. 2012-05-15 00:47 hihi 2012-05-15 00:47 with atben it's just a few parts and I focus on the pcb and see how I can get it done and documented, should be neat. 2012-05-15 00:47 wolfspraul: actually, just try making an atben. it's not like you are commiting to a longterm relationship with it or something like that 2012-05-15 00:47 whitequark: maybe you're not placing your arm/hand well. or maybe yuo have a small defect in your motoric system. try to rest hand and arm as much as possible. use small angled tweezers. if you notice your hand is shaking, do something else for a while. then 0603 should be no problem. 2012-05-15 00:48 wolfspraul: if it will work, you're two orders of magnitude better at making pcbs than me 2012-05-15 00:48 wolfspraul: if it won't, try something else 2012-05-15 00:48 there is no way it will not work 2012-05-15 00:48 wpwrak: I don't really have problems placing 0603's, it's perfectly manageable but requires some concentration, accuracy and preciseness 2012-05-15 00:49 wpwrak: whereas for 0805's or even 1206's it's mostly a no-brainer 2012-05-15 00:49 wpwrak: what tools do I need? I start with a 0.8mm fr4, dual 1oz copper 2012-05-15 00:49 whitequark: (0603) ah okay. that it does :) 2012-05-15 00:49 do you etch nowadays, or drill? 2012-05-15 00:49 I mean the copper 2012-05-15 00:50 for etching, the process is still what I saw with laser printer and toner transfer? 2012-05-15 00:50 1) i cnc-drill all the via holes 2012-05-15 00:50 2) i cnc-mill any larger holes and the board outline 2012-05-15 00:50 wpwrak: ahh, so that's how you managed to diy atbens. cnc explains everything. 2012-05-15 00:50 3) i clean the board (steel wool and alcohol) 2012-05-15 00:51 4) i toner transfer the layout. this is a bit tricky. if you can find that HP paper, that would be best. else, you need to search around. may taken 5-10 attempts before you find a paper that's good. 2012-05-15 00:52 HP will pay you to claim that they are still related to anything innovative :-) 2012-05-15 00:52 whitequark: yeah. atben is a bit tight for manual work. particularly the uSD card shape. of course, it can be done, but ... 2012-05-15 00:52 you have this Roland some-15 machine, right? 2012-05-15 00:52 wolfspraul: yeah, and even with such an old paper ;-) 2012-05-15 00:52 mdx-15, yes 2012-05-15 00:53 the engineer's 3D-printer 2012-05-15 00:53 any recommendations what is cheap and good nowadays, and easy to program/control? 2012-05-15 00:53 ;-D 2012-05-15 00:54 not sure if it makes sense to get such a critter, though. i mean the mdx-15. it's a nice package, much like an office printer, but it also has its drawbacks 2012-05-15 00:54 oh mdx-15. I WANT IT 2012-05-15 00:54 you'll not get a decent CNC for <1000 bucks 2012-05-15 00:54 drawbacks being the weird language, the weak spindle, and the small work area 2012-05-15 00:54 but $3.5k isn't really manageable for me 2012-05-15 00:55 DocScrutinizer: any advice on a good CNC in $1k-$1.5k range? 2012-05-15 00:55 plus, inconvenient mounting (you can't clamp/screw things down. the only option is double-sided tape) 2012-05-15 00:55 not at all 2012-05-15 00:55 ask roh - he seems more up to date with such stuff 2012-05-15 00:56 i don't know how whose low-cost chinese machines compare, though. maybe they have their own set of issues 2012-05-15 00:56 I will probably just buy something cheap here in beijing 2012-05-15 00:56 that's the great thing about all the chinese "crap", it's wonderful for experimenting and learning, and then knowing what is really worth spending money for 2012-05-15 00:56 hehe ;-) 2012-05-15 00:56 since they will challenge every screw - "not needed" 2012-05-15 00:56 just to find you got a reproducable precicion of >1mm ;-D 2012-05-15 00:57 that's certainly up to me to find out :-) 2012-05-15 00:57 both words "reproducible" and "precision" are not translatable into Chinese... 2012-05-15 00:57 :-) 2012-05-15 00:57 no need 2012-05-15 00:57 indeed 2012-05-15 00:57 they make their billions without them 2012-05-15 00:57 rofl 2012-05-15 00:57 so true 2012-05-15 00:58 yes once you get the hang it's so empowering really 2012-05-15 00:58 yeah, for PCB on CNC they simply produce 50 boards and select the 2 that incidentally were good 2012-05-15 00:58 you say pcb is hard 2012-05-15 00:58 ha ha 2012-05-15 00:58 I can do a brain surgery tomorrow! 2012-05-15 00:58 yeah. the mdx-15 is pretty accurate in X/Y. the table moves quite a lot in the Z direction, though. maybe 0.5 mm if you push hard. now, normally there shouldn't be much vertical pressure, but you can still sometimes run into problems. 2012-05-15 00:58 definitely 2012-05-15 00:58 chinese docs can do it too, so can I 2012-05-15 00:58 just give me some tools, and I will start 2012-05-15 00:58 first couple patients will probably die, but after some hundred it will get better 2012-05-15 00:59 a friend here has one of those hyped makerbots 2012-05-15 00:59 let me see what that is 2012-05-15 00:59 happy ? :) 2012-05-15 00:59 except for unpacking and showing people how cool they are they will have no use for it 2012-05-15 00:59 ;-) 2012-05-15 01:00 makerbots and the like are additive. so you need to apply the material you want in the end. alas, they don't print FR4, copper, and such 2012-05-15 01:00 ah ok, that probably won't help then 2012-05-15 01:00 so for pcbs, you want a mill 2012-05-15 01:00 I just want a simple drill/mill like you have 2012-05-15 01:00 ideally fully controllable from standard linux tools etc 2012-05-15 01:01 you'll need a more or less dedicated PC for the parallel port 2012-05-15 01:01 dremel? you mean a hand dremel? you need that for what? 2012-05-15 01:01 all the cheap ones leave the control work to the PC. and the PC is bit-banging the motor signals. as low-level as it gets :) 2012-05-15 01:02 (dremel) that would be the tool if you don't have CNC. or for the occasional retouching. 2012-05-15 01:02 toldya, CNC mill == engineer's makerbot / 3-D printer 2012-05-15 01:03 I fail to see the application for the additive 3d printers I've seen so far 2012-05-15 01:03 they are insanely slow, large tolerances, just a joke really 2012-05-15 01:03 maybe I could make doorstops or so 2012-05-15 01:03 or candle holders 2012-05-15 01:03 that would fit aesthetically 2012-05-15 01:04 3d printers are less messy than CNC mills. and some can also print overhangs. that's something the basic CNC mills can't make 2012-05-15 01:04 proper 3D printers start at nn kBucks 2012-05-15 01:04 and weigh tons 2012-05-15 01:04 yes sure, but we try to be very practical here 2012-05-15 01:04 wpwrak: yup 2012-05-15 01:05 wpwrak: it seems you are happy with your mdx-15 and will continue to use it... so that's a good guidance 2012-05-15 01:05 and you built a neat noise-cancelation box around it :-) 2012-05-15 01:05 what exactly is that HP paper? 2012-05-15 01:05 perhaps an in-house 3d printer can be good for very rough prototypes. so that you can hold the thing in your hands, turn it around, feel it. 2012-05-15 01:05 goldelico has printed GTA02 cases now, in translucent white 2012-05-15 01:06 the mdx-15 is all i have :) if i had a lot of money, i'd probably try to find a 4 axis machine. but we're quickly talking USD 10k+ 2012-05-15 01:06 i may need a bigger apartment as well ;-) 2012-05-15 01:07 still need more data 2012-05-15 01:07 what HP paper you recommend? 2012-05-15 01:07 how did that transfer work again? 2012-05-15 01:07 C6039A: HP Premium Photo Paper, glossy 2012-05-15 01:08 o.O 2012-05-15 01:08 first print on the laser, then use your stepper-motor modified whatever-it-was to transfer/burn the toner onto the copper? 2012-05-15 01:08 or the same kind of paper but in a different size. the product code changes with size and number of sheets in the package. 2012-05-15 01:08 how about just using an iron to transfer it? 2012-05-15 01:08 the laminator, yes 2012-05-15 01:08 you need heat & pressure? 2012-05-15 01:09 yup, just like inside laserprinter 2012-05-15 01:10 you can use a clothes iron instead. need a bit of figuring out the technique, though. ironically, i found that women seem to have a lot less difficulties with it than men ;-) (a few friends have built circuits using this process) 2012-05-15 01:10 heat and pressure, yes 2012-05-15 01:11 ok, sounds like iron first 2012-05-15 01:11 what chemicals do you need for the etching? 2012-05-15 01:11 umm, now it gets funny 2012-05-15 01:11 yeah. also modifying the laminator can be a bit messy. i made a very simple circuit that works but poorly 2012-05-15 01:11 anything to watch out for in the laser printer? 2012-05-15 01:11 sure I will try to skip that 2012-05-15 01:11 i use muriatic acid and peroxide 2012-05-15 01:12 yeah, you might mess it up 2012-05-15 01:12 I will go out and buy cheap chinese tools, whatever I can find, and then reverse from there. rather than following a perfect spec. 2012-05-15 01:12 it was a 2-step process, right? 2012-05-15 01:12 you have the etching bath, put it in there for xx seconds, then out and into a second cleaning bath? 2012-05-15 01:13 also I think tones can differ vastly 2012-05-15 01:13 toners* 2012-05-15 01:13 xwalk has quit [Ping timeout: 248 seconds] 2012-05-15 01:13 maybe but no point to over-plan and assume the worst 2012-05-15 01:13 make sure you're printing as 'fat' as possible 2012-05-15 01:13 (laser printer) the straighter the paper path, the better. some have a backdoor you can open and then the paper doesn't get turned around 2012-05-15 01:13 like I said, when the experiments are cheap and fast, they are fun (and valuable) no matter what the outcome 2012-05-15 01:13 straight - excellent, makes sense 2012-05-15 01:14 xiangfu has joined #qi-hardware 2012-05-15 01:14 but in general, it seems that any printer will do. you may have to vary the transfer paper, though 2012-05-15 01:14 can you prepare the copper in some way for better transfer? 2012-05-15 01:14 I would think there are many ways to transfer it over 2012-05-15 01:14 I think wpwrak already mentioned copper conditioning 2012-05-15 01:15 cleaning and polishing 2012-05-15 01:15 you must clean the copper before the transfer. so you scrub it with the steel wool until it's shiny. then you wash/rub off the steel wool with alcohol. 2012-05-15 01:15 sandpaper also ok? 2012-05-15 01:15 eeeek 2012-05-15 01:15 probably not 2012-05-15 01:16 then quickly, before it oxidizes again, do the transfer. if it fails, remove the toner with paint thinner or acetone, then repeat the cleaning process 2012-05-15 01:16 maybe if you can find very very fine sandpaper. but steel wool should be easier. any supermarket has that ;-) 2012-05-15 01:16 sandpaper vs steel wool? 2012-05-15 01:17 very fine, hmm 2012-05-15 01:17 corning 500 is no problem here 2012-05-15 01:17 they have sandpaper everywhere, from about 100 to 500 I think, have never tried to look for extremes though 2012-05-15 01:17 more like 1400 2012-05-15 01:17 and the steel wool is less intrusive? I would have never thought 2012-05-15 01:18 well, steel wool is less hard than sand paper 2012-05-15 01:18 I would think that the steel wool creates deep scratches, whereas say a corning 500 sandpaper feels quite smooth 2012-05-15 01:18 ok, thanks 2012-05-15 01:18 I will try 2012-05-15 01:18 and copper is harder than most other materials you usually apply steel wool to 2012-05-15 01:18 you want the fine steel wool. with almost hair-thin steel. not the coarse with "noodles" 2012-05-15 01:18 depends on applied pressure too, of course 2012-05-15 01:18 ok [hair thin] 2012-05-15 01:19 the etching... 2012-05-15 01:19 need more data 2012-05-15 01:19 2 step? 2012-05-15 01:19 you can get scratches, yes. you can also experiment a bit with what works best. 2012-05-15 01:19 you buy the chemicals or make yourself? which chemicals do you buy? 2012-05-15 01:19 I recently discovered a "home improvement" market nearby 2012-05-15 01:19 actually VERY NEARBY 2012-05-15 01:19 a bit scary 2012-05-15 01:20 something you go to prison when buying in USA ;-D 2012-05-15 01:20 they have little shops who are piling up chemicals in their little barracks until under the roof 2012-05-15 01:20 kids playing there, on the barrels etc. 2012-05-15 01:20 urgh 2012-05-15 01:20 and I am *sure* those people know *nothing* about chemistry 2012-05-15 01:20 nothing, zero 2012-05-15 01:20 yeah 2012-05-15 01:21 etching: mix the etchant, toss the pcb inside, move it around until it's finished, remove, rinse, check visually. if you find insufficiently etched bits, toss it back in. otherwise, wash carefully, then proceed with tinning. 2012-05-15 01:21 that's China how we love and know it 2012-05-15 01:21 I will keep my shopping there very focused 2012-05-15 01:21 in, want this, that and that, pay, out 2012-05-15 01:21 how nearby ? downstairs neighbour ? ;-) 2012-05-15 01:21 with all my wikipedia research about chemicals, I know 100 times more about those things than them who are storing 100 times more right where they sleep with their whole family... 2012-05-15 01:21 scratch that, 10,000 times more than me 2012-05-15 01:22 always keep the muriatic acid next to the caustic soda and nothing can go wrong ;-) 2012-05-15 01:22 if you can get ammoniumperclorate, I think it's a good etching 2012-05-15 01:22 maybe also keep a few canisters of 50% peroxide around, just in case :) 2012-05-15 01:22 ok one by one 2012-05-15 01:22 IIRC 2012-05-15 01:22 so it's only 1 bath, the rinse with normal water? 2012-05-15 01:22 wtf is muriatic acid? 2012-05-15 01:23 yes. it's only 1 ml or so that you need to get rid of 2012-05-15 01:23 tap water here probably has a few hundred unknown chemicals in it anyway 2012-05-15 01:23 DocScrutinizer: HCl 2012-05-15 01:23 aah 2012-05-15 01:23 wait, I need this in the right order 2012-05-15 01:23 1. what do you buy 2012-05-15 01:23 2. how to you mix/prepare 2012-05-15 01:23 wolfspraul: there are several chemical processes what people use for etching pcbs 2012-05-15 01:23 3. the bath you described and remember from my visit 2012-05-15 01:23 4. rinse with tap water 2012-05-15 01:23 5. how to dispose of the rest 2012-05-15 01:23 sure I can imagine, but let's just focus on yours 2012-05-15 01:24 and watch out joking with me, I may not get it 2012-05-15 01:24 :-) 2012-05-15 01:24 there if Fe(III)Cl (or some such), there is ammonium chloride, there is natrium chloride, and there's HCl+Peroxide. plus a few even more exotic ones. 2012-05-15 01:24 then I end up with some bomb, since the people I buy from certainly know even less than me 2012-05-15 01:24 yes but let's focus on your process 2012-05-15 01:24 ah, let's talk about the fertilizer then .. ;-) 2012-05-15 01:25 can you walk me through? 2012-05-15 01:25 just what you do 2012-05-15 01:25 what do you buy 2012-05-15 01:25 how do you prepare/mix 2012-05-15 01:26 FeCl = traditional, needs heating, considered pretty "safe" to handle. 2012-05-15 01:26 the application I think I'm clear (as described above), then disposal 2012-05-15 01:26 slow and messy 2012-05-15 01:26 much dirt 2012-05-15 01:27 ammonium chloride and natrium chloride (actually, it's not NaCl ... must be something slightly different) = cleaner than FeCl, still considered "safe", work at room temperature 2012-05-15 01:28 natrium chlorite maybe? 2012-05-15 01:28 FeCl drawbacks: it's a ugly brew, rost-brown to nearly black. stains everything it touches. loses transparency after a bit of use. will not evaporate. 2012-05-15 01:28 what is your known best and latest practice? 2012-05-15 01:29 riuns your cloths and skin 2012-05-15 01:29 I will focus on that 2012-05-15 01:29 ammonium chloride and such drawbacks = harder to obtain, slow 2012-05-15 01:29 there's also ammonium persulphate 2012-05-15 01:29 yeah 2012-05-15 01:29 it's pretty awesome: easy to obtain, clean 2012-05-15 01:30 not perchlorate but persulphate 2012-05-15 01:30 through it ruins clothes even faster than FeCl(III) 2012-05-15 01:30 I guess I mixed it up somehow 2012-05-15 01:30 now, HCl + H2O2 = clear solution, stays transparent, works at room temperature, you can choose how fast you want it to be, you can let it evaporate and then only have to dispose the solids 2012-05-15 01:30 that's your current process? 2012-05-15 01:30 HCl + H2O2: ingredients extremely cheap and easy to find (hardware store and pharmacy) 2012-05-15 01:31 yes, HCl+H2O2 is my current process 2012-05-15 01:31 you buy hcl and h202 separately? 2012-05-15 01:31 keep away any organic solvent from this though!!!! 2012-05-15 01:31 esp acetone 2012-05-15 01:31 hcl = hydrogen chloride 2012-05-15 01:31 drawbacks: need to control the chemistry a bit. releases Cl gas. 2012-05-15 01:32 hcl = muriatic acid ("salzsaeure") 2012-05-15 01:32 ah ok, hydrochloric acid, i.e. solution of hydrogen chloride in water 2012-05-15 01:32 and keep clear of the gas! 2012-05-15 01:32 h2o2 = peroxide. at high concentrations, peroxide is used as rocket fuel. 2012-05-15 01:33 at low concentrations, as wound desinfectant 2012-05-15 01:33 this one? http://en.wikipedia.org/wiki/Hydrochloric_acid 2012-05-15 01:33 fortunately, you'll only need the 5% variant 2012-05-15 01:33 easily found at barbers 2012-05-15 01:33 you may want to keep a bottle of ~30% around to recycle your acid. but that's already advanced use. 2012-05-15 01:33 wait wait 2012-05-15 01:34 I'm still with the hcl acid 2012-05-15 01:34 what concentration of that? 2012-05-15 01:34 any pharmacy has 5% peroxide. liters of it. probably hundreds ;-) 2012-05-15 01:34 the "standard" concentration. i think it's 35%. could be 30%, though. 2012-05-15 01:35 the one they have at hardware stores. it's used for plumbing work. not sure what exactly the plumbers do with it, though. 2012-05-15 01:35 cleaning copper ;-D 2012-05-15 01:36 peroxide is a liquid too? 2012-05-15 01:36 yup 2012-05-15 01:36 and I need a 5% concentration 2012-05-15 01:36 wasserstoffperoxid 2012-05-15 01:36 a bottle of 30% concentration for "recycling your acid"? 2012-05-15 01:36 don't understand 2012-05-15 01:36 better forget that 2012-05-15 01:36 then you mix the 35% hcl and 5% h2o2 ? 2012-05-15 01:37 in which way? 2012-05-15 01:37 30% H2O2 is actually dangerous 2012-05-15 01:37 i think it was 1/3 hcl, 2/3 h2o2 2012-05-15 01:37 my #1 concern with all this would be that what I think I buy is not what I actually buy 2012-05-15 01:37 but you may want to experiment a bit. you can also add a bit of either while etching. 2012-05-15 01:37 the sellers here operate with 0 knowledge, if anyone can imagine this 2012-05-15 01:37 wolfspraul: indeed 2012-05-15 01:37 I think you cannot, unless you lived in china for a while and get the feeling on how this works 2012-05-15 01:38 so I actually would only trust the concentration if I measured it myself, in some way 2012-05-15 01:38 and if the measurement equipment was imported :-) 2012-05-15 01:38 seriously 2012-05-15 01:38 but anyway 2012-05-15 01:38 let's keep on our over-knowledge path... 2012-05-15 01:38 anyway you'll easily check for general properties of your chemical. HCl is easily detecable by smell, H2O2 will act like sodawater on any wodden surface 2012-05-15 01:38 wooden* 2012-05-15 01:38 the 30% H2O2 would be for increasing the H2O2 concentration if the solution gets stale. or if you've let it sit for a while. and yes, that's something you don't touch without gloves :) 2012-05-15 01:39 1/3 hcl, 2/3 h2o2 ? 2012-05-15 01:39 how much do you make typically? 2012-05-15 01:40 wolfspraul: i doubt the pharmacy will have peroxide much stronger than 5%. maybe 10%, which would be used for hair bleaching. anything stronger and the customer feedback will be quick and unfriendly :) 2012-05-15 01:40 those are both liquids - any danger of them reacting with air, something else, evaporating? 2012-05-15 01:40 I go to the 'home improvement' market around the corner 2012-05-15 01:40 a basin large enough to hold your PCB, and 5..10mm high 2012-05-15 01:40 I wouldn't be surprised if they are selling used equipment from all sorts of nuclear facilities around beijing, without knowing what it is 2012-05-15 01:40 HCl can evaporate out of the bottle (unless the bottle is good enough). so better keep the bottle outdoors 2012-05-15 01:40 the problem is not what I can get, but whether I know for sure what I get 2012-05-15 01:40 but that's my problem 2012-05-15 01:40 wolfspraul: yes, toldya DO NOT GET THEM in contact with acetone 2012-05-15 01:41 I think this is actually a small miracle. the semiconductor industry uses large amounts of really dangerous chemicals 2012-05-15 01:41 like arsine 2012-05-15 01:41 gallons and gallons of it 2012-05-15 01:41 the 5% peroxide should be as harmless as it gets. i mean, i wouldn't drink it, but i don't think it makes a mess with anything else. 2012-05-15 01:41 but I think until today, there never has been 1 fatality 2012-05-15 01:42 but then, china is still behind in that industry, and if there were a fatality it would surely be covered up 2012-05-15 01:42 (in china) 2012-05-15 01:42 they probably just seal off the valley and wait until the screaming stops. no bad news ;-) 2012-05-15 01:43 do you know how many and what large amounts of *really* dangerous chemicals are used in semiconductor making? I had no idea... 2012-05-15 01:43 wolfspraul: pretty please read http://de.wikipedia.org/wiki/Acetonperoxid 2012-05-15 01:43 well, you need some pretty aggressive etchants .. 2012-05-15 01:43 alright, so... 2012-05-15 01:43