#qi-hardware IRC log for Friday, 2014-06-13

whitequarkjust seen that. not all hope is lost, it seems00:38
wpwrak(small speaker) what if you leave both ends open ?00:46
whitequarkoooh, DocScrutinizer05 finally confirmed transfer of funds to neo900 UG01:20
whitequarkI started worrying already01:21
dos1that was me actually, on his behalf ;) and well, Nikolaus is still far from being done with all transfers01:22
dos1so you were lucky that yours was processed at the beginning ;)01:22
whitequarkI see. still good01:26
wpwrak(tesla) i couldn't have said it better myself :)01:26
wpwrakdos1: maybe a refund status update on tmo couldn't hurt, to spread the word01:26
wpwraki.e., that about half of the things have been processed at all, that it takes a while for things to actually make it through the system, and then some more time before the transfer is manually processed and an ack is sent01:27
wpwrakso that people don't get too nervous :)01:27
whitequarkwpwrak: honing my skills... http://i.imgur.com/dpPEhA3.jpg02:32
whitequarkor even http://i.imgur.com/cMCXdvS.jpg, with better contrast02:34
wpwraknot much of a challenge with such huge components ;-)02:35
whitequarknote how I've massively improved my etching02:35
whitequarkclean, straight edges02:35
whitequarknow I only need a laminator with better temperature control, because the bubbling of the mask is likely the result of that being off...02:35
wpwrakoh yes, the edges are very clean02:36
wpwrakdo you know what a steiner tree is ? that's roughly how you'd want to organize your routes. you have some that are a bit circuitous02:37
whitequarkugh, wikipedia is down02:41
whitequarkwpwrak: can you explain from your meat memory02:41
wpwrakdown, really ? en.wikipedia.org/wiki/Steiner_tree_problem02:41
wpwrakto join three points with lines, you pick a fourth point "in the middle", then draw from each point to that middle point. that minimizes the length of the lines.02:42
wpwrakgeneralize for N points :)02:42
whitequarkdown, yes02:43
whitequarkmaybe I hit a different path in the load balancer02:43
wpwrakof course, in a pcb you won't calculate those exact points but approximate the routing to what a steiner tree would look like02:43
whitequarkI didn't really route it to minimize the trace length02:43
wpwrakmaybe putin found an article on homosexuality :)02:43
whitequarkI routed it to satisfy my obsession, mainly02:44
wpwrakhehe :)02:44
whitequark(putin) no way, I put openvpn on my router like months ago02:44
whitequarkso everything goes through UK.02:44
whitequarkand yes, that happened (not literally that but similar accidents)02:44
whitequarkthey would've not affected me, though, because my ISP plain out ignores federal law02:44
whitequarkfun, eh?02:44
wpwrakwell, short traces are usually desirable: less can go wrong when etching, less unintentional emission, less undesired immission, less resistive loss, typically less in-circuit interference, etc.02:46
whitequarkI don't accept any etching that isn't perfect02:46
whitequarkotherwise, yeah, generally correct, but absolutely pointless for this circuit02:46
whitequarkwait, immission?02:46
whitequarkimmission (plural immissions)02:47
whitequarkThe act of immitting02:47
wpwrakyou should set up another vpn that sends it to .au and then to .us, in case the're changing tapes in the uk when your packets fly by :)02:47
wpwrak"n.1.The act of immitting, or of sending or thrusting in; injection; - the correlative of emission."02:48
whitequarkah, you mean, it wouldn't catch rf02:48
wpwrakor other noise02:48
wpwrakthere can also be very bad things at very low frequencies :)02:49
whitequarkhmmmm, for example?02:49
wpwrakpeople in the downstairs lab experimenting with large coils :)02:49
whitequarkwouldn't it require a really huge trace?02:51
whitequarklike, pcb antennas only really work for >=dozens of mhz02:51
wpwrakwe had that happen once at university. a sun kept on crashing. field services replaced every single bit of the machine, but even with just the case still in place, it still crashed. i suggested we ask them to send an exorcist instead of a technician.02:51
wpwrakwell, one fine day our sysadmin had lunch with the folks from the lab below us. they told him of their cool new experiment. that experiment happened to be precisely below that machine. 1 m to the side and you wouldn't have problems.02:52
wpwrakoh, anything can mess with your stuff. it's not "all or nothing". e.g., that sun probably was just being pushed the tiniest bit over the edge. a few percent less and it would probably have worked perfectly.02:53
Action: whitequark sighs03:02
whitequarkI disassembled this chinese UV led flashlight. a fine piece of #yoloengineering03:02
whitequarkno current regulation whatsoever03:02
wpwrakwhy would it need one ?03:03
wpwrakUV LED fail. you buys new UV LED. everyone happy.03:04
whitequarkI don't get it. somehow this stuff doesn't work.06:34
whitequarkI've connected a pair of LEDs to the outputs on the control board, and they light up just fine06:34
whitequarkbut when I connect my relay board, which I have tested independently and established that it works fine, too, extremely odd behavior happens06:35
whitequarkwhen I enable (i.e. set high) output 1, relay 1 triggers. when I enable output 2, relays 1 and 2 trigger.06:35
whitequarkI'm starting to think it would be easier to use an optocoupler there07:33
whitequark... which is the case for which it's clearly been designed07:33
astrDocScrutinizer05: is this a closed speaker?  http://www.aliexpress.com/item/20pcs-lot-Acoustic-Speaker-2030-3020-Oval-Tablet-Phone-MP3-Speaker-1W-8-Ohms-L30MMxW20MM/1680409694.html09:18
nicksydneywpwrak:  new, tiny and sexy http://www.acmesystems.it/arietta14:03
DocScrutinizer05wpwrak: what experiment was this? magnetic neuro stimulation?14:11
Action: DocScrutinizer05 shouldn't read stuff like https://en.wikipedia.org/wiki/Post-hoc_analysis before first coffee14:30
DocScrutinizer05astr: I can't answer your question since there are no datasheets. However it *looks* like both speakers have a chamber with aperture, which generally needs further considerations what "to do" with the air/sound coming out of such aperture14:32
DocScrutinizer05seems like that oval 20*30mm 5mm thick speakers are widely used, maybe in a popular tablet or laptop or whatever. You should try to find either the datasheet of the speaker or the device it's used in and then copy what those devices' designers did. They are supposed to (but not guaranteed to) have had consultations with the speaker manufacturer or at least got detailed datasheets for the speaker14:34
DocScrutinizer05odds are that those speakers are actually meant to get mounted to baffle, which in practice means you glue the speaker to wall of case behind an aperture that the sound escapes from case, and backside of speaker (with "chamber") is "open" to the inside of device and has no further acoustic elements attached14:38
DocScrutinizer05when you can get a sample of such speaker, there are test procedures to acquire its Q, F(0) etc parameters, and then calculate the performance in various mounting variants by using tools like the mathcad sheets from quarter-wave site14:41
DocScrutinizer05generally building a good speaker is really difficult and not only depends on buying a transducer of matching size and "good quality"14:43
astrDocScrutinizer05: I wonder if I can just buy an existing thing and hack out the speaker + chamber (+ amp?). I can't face spending days working out how to make one.16:07
DocScrutinizer05that's exactly what I'd suggest. Amp is dead simple. The speaker is a standard 8..32Ohm and any semi-matching amp can drive it16:15
DocScrutinizer05ask if you want to know about my notion regarding possible small tweaks to chamber geometry, you _can_ do some bending and folding to it without changing acoustic properties much16:17
astrDocScrutinizer05: I'd like to know18:02
DocScrutinizer05I meant you can ask me about particular changes you plan to apply to a copied design, and if I hink they can work18:04
astrre the amp, the troble is it will be batt charge or other bits I don't need that will make it bigger when I can buy a tiny module like http://www.aliexpress.com/item/Free-Shipping-5pcs-lot-DC-5V-2-Channels-3W-Digital-Power-D-Audio-Amplifier-Board-Amplifier/1758901278.html18:04
astrwhat about victimising something like this http://www.aliexpress.com/item/Portable-Mini-Ball-Speaker-mini-mobile-speaker-Tiny-portable-mobile-phone-speaker-sponge-ball-speaker/1482323150.html18:05
DocScrutinizer05sorry, I can't give advice regarding amp design to use18:05
DocScrutinizer05amp design depends on your particular needs18:06
astrre the amp, I guess I just find one that matches the watts of the speaker. but it seams a bit wastfull18:06
DocScrutinizer05and of course you as well can check how that sponge ball is done18:06
astrmy sister has something like this, it's quite loud :). http://www.aliexpress.com/item/Portable-notebook-computer-audio-player-Cute-black-laptop-audio-table-pc-Bomb-Speaker-mobile-phone-cellphone/1768145090.html18:08
astrthough it distorts at max vol. sounds like a issue with the casing18:09
astrsome sort of vibration18:09
astrperhaps hacking this is that answer http://www.aliexpress.com/item/Portable-notebook-computer-audio-player-Cute-black-laptop-audio-table-pc-Bomb-Speaker-mobile-phone-cellphone/1768145090.html18:16
astrhmm feedback complains of tablet being louder than it18:18
DocScrutinizer05sorry, no idea18:19
astrdidn't expect you to have one :)18:31
wpwraknicksydney: seen these yet ? https://www.gumstix.com (and they're ancient :)19:29
--- Sat Jun 14 201400:00

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