#qi-hardware IRC log for Monday, 2012-08-20

wolfspra1loh that's nice, thanks larsc00:26
wpwrakso it takes about one month from the moment they begin work until they emit the final document ? and yes, it's those emissions in the low MHz range that i'd be most concerned about. i should have kept the old radios of my parents. they would cover that range :)00:54
wolfspra1lthe testing itself takes a few hours only, at most01:00
wolfspra1lI tried to put things in perspective, I think for your little thing you already worry too much about "fcc"01:01
wolfspra1lyour problem will be to find a lab that cares about a one-time little nothing like testing your board01:01
wolfspra1lwhere 99% of people who would make such type of device wouldn't bother with anything fcc :-)01:01
wolfspra1lthe idea of the fcc is to avoid something like "if I turn this on, will my cellphone stop working" etc.01:02
wolfspra1lyou can just say 'no' with your professional authority/design 'excellency', and done01:02
wolfspra1lat least if more people would think like you, testing labs wouldn't worry about their growth opportunities :-)01:04
wolfspra1lfor example when importing into the US itself, home territory of the FCC, you would check this on the import form, even for large quantities01:05
wolfspra1l"An FCC grant of equipment authorization and an FCC ID are not required, but the equipment complies with FCC technical requirements."01:05
wolfspra1lif you 'certify' this "complies with tech req" with your own professional authority, including some realization you had in your dream that there cannot be an issue, nothing will ever happen01:06
wolfspra1land everybody can focus on making money :-)01:06
wolfspra1lI doubt you find a lab to take care of this little business, that is exactly why the labs in europe just don't even pickup the phone unless you wire 25k usd upfront01:06
wolfspra1lI can understand them :-)01:06
wolfspra1lin China you may be lucky off-season that someone takes your thing, and I think it can easily all be finished in 1 day, testing, paperwork, stamp, done01:07
wolfspra1lthere will *not* be an FCC ID in any case01:07
wolfspra1lbut if there ever were a liability issue (very very unlikely), you could drag in that poor lab and point fingeres at them, I guess01:07
wolfspra1lso they make you pay a little01:07
wolfspra1lrisk = profit. if you outsource all risk, you cannot make a profit, by definition.01:08
wolfspra1lI suggest you just make that statement of compliance yourself, and it can be verified by an FCC approved lab at a later point shall that become necessary01:09
wpwrakwell, this sort of thing may come up when someone wants to import/resell01:13
wolfspra1lthat's why you make that statement01:13
wolfspra1lyou are the designer, you know the board, someone must STEP UP01:13
wolfspra1lyou want to point fingers at a lab and make them step up, fine. then you find such lab and pay.01:13
wolfspra1lyour device is an unintentional radiator in any case01:13
wolfspra1leven you admit that :-)01:14
wolfspra1lso just say so, just say "yes, it does comply with the fcc technical requirements for unintentional radiators"01:14
wolfspra1lcheck the box on the import form, and done01:14
wpwrakthe difference between me and the lab is that they have the equipment to actually see what's going on :) the paperwork itself is much less of a concern01:14
wolfspra1lof course if you keep saying "but I am not sure", then fine, pay for it01:14
wolfspra1lI can just tell you that business doesn't work with that kind of hesitation level01:16
wolfspra1lthat little thing won't radiate01:16
wolfspra1land even if it does, it can be addressed later01:16
wolfspra1ljust someone needs to step up, make a decision, take responsibility, and move forward01:16
wolfspra1lmake good design choices, and you are safe01:16
wolfspra1lreally01:16
wpwrak(good design choices) tricky without any feedback :)01:17
wolfspra1lthat's already much more professional love than 90% of electronic china toys will ever receive01:17
wolfspra1lyou worry about the wrong things01:17
wolfspra1lthat's why you make no money :-)01:17
wolfspra1lchinese little electronic plastic toys for kids are flooding the world, right?01:18
wolfspra1lincluding poisonous colors and what not01:18
wolfspra1land including the crudest possible electronics you can imagine01:18
wolfspra1lradiating?01:18
wolfspra1lha ha01:18
wolfspra1lWHO KNOWS?01:18
wolfspra1lfcc?01:18
wolfspra1lOH01:18
wolfspra1lMY01:18
wolfspra1lGOD01:18
wolfspra1lyou must be living in another world...01:18
wolfspra1lyou personally have 100 times the skill level of the entire hw engineering team of any chinese electronic toy maker01:19
wolfspra1lthey wouldn't even know what fcc is, actually01:19
wolfspra1lor that they are making 'unintentional radiators'01:19
wolfspra1lanyway, that's my perspective...01:19
wolfspra1lmy advice:01:20
wolfspra1l1) step up and make the professional statement that your device meets the requirements for an fcc unindentional radiator, by your own authority and design excellency01:20
wpwrakthe chinese have the advantage of anonymity ;-)01:21
wolfspra1l2) be open-minded to having this verified by an fcc approved lab at any time, shall there be doubts about your authority01:21
wolfspra1lno, not really01:21
wolfspra1lit's just a non-issue, you don't realize that you yourself may very well possess even more than the 'typical' professional whatever to make this kind of statement01:21
wolfspra1lyou think the lab has some geniuses?01:21
wpwrakno. but they have equipment :)01:21
wolfspra1land on a production line, you are dealing with component variations01:21
wolfspra1land even though you know all this, you can't just cut through01:22
wolfspra1lyour device will not radiate01:22
wolfspra1lnot at the level that the fcc would worry about01:22
wpwraklemme do some calculations ...01:22
wolfspra1lso it's an 'unintentional radiator'01:22
wolfspra1llike for example millions of cheap electronic plastic toys01:22
wolfspra1lout of which many may actually very well radiate like hell01:22
wolfspra1lbecause the 'engineers' that fumbled together those crude electronics make the absolute most ridiculous beginner mistakes you could imagine01:23
wolfspra1llike the worst worst worst combination, like in a textbook to demonstrate how *not* to do it01:23
wolfspra1lyet the devices are being made, shipped, sold, taxed, etc. everybody seems happy01:23
wolfspra1lexcept werner :-)01:23
wpwrakhmm. if i consider the high-current paths as RC filters, i get a cut-off frequency of 80 MHz. pretty high.01:24
wpwrakbut i take the message that FCC compliance checking may be something to do in parallel with other things, i.e., not something you actually busy-wait for01:26
wolfspra1lstep up01:27
wolfspra1lyour device is an unintentional radiator, at best01:27
wolfspra1lor are you hiding something?01:27
wolfspra1l:-)01:27
wolfspra1lif you follow the commonly understood principles to minimize radiation, you WILL BE FINE01:28
wolfspra1lbecause then you already do much more than all the chinese electronic toys that are flooding the world01:28
wolfspra1lthe only thing that radiates is your uncertainty about your own design :-)01:29
wpwrakhmm. the "commonly understood principles" would be multilayer and extra load caps to drive down the frequency. designing defensively has its price.01:30
wolfspra1lmake the extra effort and buy a 2 USD chinese toy outside somewhere, if Argentina can still afford to import them01:31
wolfspra1ltake it apart01:31
wolfspra1ltake a deep breadth01:31
wolfspra1land arrive in the real world01:31
wpwrakhmm. i also have a relatively high switching frequency in high-current paths. those toys probably don't.01:33
wpwrakthe test reports for the Ben are interesting. very specific design recommendations.01:40
DocScrutinizer05yay07:49
DocScrutinizer05wpwrak: voltage, length of *free* wires acting for antenna. all you care about07:50
DocScrutinizer05get a 100nF next to each chip's VDD-GND07:51
DocScrutinizer05done07:51
DocScrutinizer05unless you're dealing with switched mode PSU07:51
DocScrutinizer05which by design are nasty little radiators - I notice that every time my DCF77 clock refuses to sync when N810 or N900 (incl USB wallwart charger) is closer than 150cm07:52
DocScrutinizer05get a multiband radio/receiver (150kHz..50MHz), see if you can find some 'carrier' that vanishes when you power down DUT. place receiver 1m from DUT. all you need for "equipment"07:54
pangDocScrutinizer05: that's useful, thanks =)07:56
pangdo you have suggestions for receiver hw and sw?07:56
DocScrutinizer05err, Sony has nice stuff07:56
pangwill a basic RTL-SDR and gnu radio setup work?07:56
DocScrutinizer05way too complicated, introduces too much unknown variables07:57
DocScrutinizer05Sony ICF-SW7600GR,  Sony ICF-SW35,  Sony Weltempfänger Icf-sw11/s07:58
pangSony ICF-SW7600GR looks nice, with the scan mode08:03
DocScrutinizer05sony-icfsw7600gr is actually well suited for the task, continuous band from 150kHz to 30MHz08:04
DocScrutinizer05the smaller ones you'll hope for 1st harmonic of some spike out of band will be in band for your receiver08:04
DocScrutinizer05those "radiators" never are perfect transmitters not having any harmonics and sidebands08:05
DocScrutinizer05but honestly, up to maybe 3.3V supply I'd not worry for anything without a step-up converter08:07
DocScrutinizer05and the general FCC "paper" for consumer devices says: "you may experience interference from this device to any nearby radios, TV, etc. Increase distance between the two devices in that case"08:09
DocScrutinizer05fair enough, huh? ;-)08:10
DocScrutinizer05German regulations are way stricter here than FCC08:11
pangi'm based in Norway, and we just follow EEC regs08:14
DocScrutinizer05the whole stuff got massively unified worldwide during last ~30 years08:15
DocScrutinizer05otherwise trade was impossible08:16
DocScrutinizer0530 years ago plugging in your own 600R landline phone to the wall outlet was a crime in Germany ;-)08:17
DocScrutinizer05then came modems08:17
DocScrutinizer05eventually modems >2400baud08:17
DocScrutinizer05and lots and lots of people didn't get it why those mustn't be used in Germany while they work just fine in all the rest of the world08:18
DocScrutinizer05basically back when everybody became a criminal08:18
DocScrutinizer05;-)08:18
DocScrutinizer05then they relaxed stuff08:18
DocScrutinizer05FAX machine not rented from german telecom and configured properly by their service dude - WOOOOOOOOOOH EVIL! ;-P08:20
whitequark600R?08:20
DocScrutinizer05landline08:20
DocScrutinizer05a-b08:20
whitequarkno clue08:20
DocScrutinizer05analog08:20
whitequarkah.08:20
whitequarkbut why is it called 600R?08:20
DocScrutinizer05because that's the nominal line impedance08:21
whitequarkthat simple08:21
DocScrutinizer05yay08:21
DocScrutinizer05freenode \o/08:21
LunaVoraxHello!10:57
wolfspra1lhi11:39
LunaVoraxhi wolfspra1l :)11:40
paul_boddieThe Wiki has a NanoNote category and a Ben NanoNote category, the latter with significantly more pages. I guess we should at least put the former category's pages in the latter category unless they're too general for the model-specific category.11:59
wolfspra1ldefinitely, merge together into the bigger one sounds like a good idea12:05
wolfspra1lthanks for the initiative!12:05
paul_boddieThere aren't as many pages as I thought, thankfully. Most of the items are images which I may also try and recategorise, but I was mostly thinking of the convenience of seeing all useful resources in a single category.12:10
paul_boddieIt also helps me to find stuff out that I didn't already know. ;-)12:11
paul_boddieI got numpy working at the weekend, by the way. I think people asked for that once upon a time. The catch is that you need to build it with eglibc, and there are some issues with eglibc that I need to look into.12:14
paul_boddieI suppose that the relevant functionality could be ported to uClibc from eglibc, but that's somewhat outside my experience and expertise, currently.12:17
whitequarknews from the russian sonar-manufacturing company13:10
whitequarkthe main employee has ordered 4-layer boards with 0.1mm wide traces and p&p machine mounting13:11
whitequarkfor 0m5 RUR (around 15k USD)13:11
whitequarkand he _accidentally_ swapped top and bottom layers13:12
pang=)13:15
whitequarkpang: you probably have missed the immense stupidity of entire thing13:18
whitequarkgrep on http://irclog.whitequark.org by keyword "sonar"13:18
rz2kwhitequark: that is the one where grindars works?13:26
whitequarkyep13:35
paul_boddiewolfspra1l: The EU shop on the NanoNote page doesn't sell hardware any more as far as I am aware. Should we link to other resellers instead or is this going to offend someone? :-)13:36
rz2klol, I remember that sonar with hole in case for crystal.13:37
xiangfupaul_boddie, I agree. link to other resellers. :)13:55
paul_boddiexiangfu: Done already! People can always revert my changes. ;-)13:55
xiangfupaul_boddie, great.13:56
paul_boddiexiangfu: Did you ever look at eglibc, by the way? No-one seems to want to talk about it. :-)13:56
xiangfupaul_boddie, eglibc, no.13:57
xiangfueglibc, since the openwrt using uClibc by default. so I followed openwrt. using uClibc. I have no idea which one is good or bad.13:57
Action: xiangfu search eglibc vs uclibc13:58
paul_boddieI had thought about trying the latest "buildroot" to see whether I could fix other stuff as well, although my initial aim was to have as much working stuff on the NanoNote as possible, so I went for the 2012-04 build.13:58
xiangfu"The EGLIBC maintainers hope that EGLIBC will be binary compatible with GLIBC"14:00
paul_boddieIt probably should be as it's derived from glibc.14:00
lindi-afaik they are merging the changes back now finally14:14
qwebirc35176Hi for all14:17
qwebirc35176The link sources of the page http://projects.qi-hardware.com/index.php/p/qi-kernel/ is broken for a lot of days...14:17
qwebirc35176can you take a look at this?14:17
xiangfuqwebirc35176, Hi14:17
xiangfuthat is a known bug.14:18
qwebirc35176xiangfu: Tanks for the answer14:18
xiangfuthe web is broken.14:18
qwebirc35176xiangfu: There is a way to I see the code in browser?14:18
xiangfuqwebirc35176, you can still download all qi-kernel code by git clone: 'git clone git://projects.qi-hardware.com/qi-kernel.git'14:19
qwebirc35176But I'm in a Windows machine, my boring job obligates me to use this ¬¬14:20
qwebirc35176and I'm doing a radio player for OD, and I have some ideas14:20
qwebirc35176and I need to take a look at the code...14:20
qwebirc35176=/14:20
xiangfuyou can download a windows version git of cause. :)14:21
xiangfuqwebirc35176, I just update the http://projects.qi-hardware.com/index.php/p/qi-kernel/ page. notice people download code by using git clone command14:24
qwebirc35176Hum, that the only way I trhink =D14:27
qwebirc35176xiangfu: So, what do you think to remove the source entry of the web page?14:28
qwebirc35176xiangfu: Because I think that more people will try to see the code at web =D14:29
xiangfuqwebirc35176, thanks. done14:30
qwebirc35176xiangfu: Sorry, your note told the people that web is broken, thamks a lot xiangfu !14:30
xiangfunotice that after I review the page. all much better14:34
paul_boddieInteresting: https://dev.openwrt.org/ticket/9675 - I hope to get eglibc 2.16.1 to build, though, as something went wrong with 2.15, I think.14:34
paul_boddieLunaVorax: Which of these pages is the best: http://en.qi-hardware.com/wiki/How_to_reflash/fr or http://en.qi-hardware.com/wiki/Comment_reflasher ? My French isn't good enough to make a judgement, but you seem to have written one of them.15:17
LunaVoraxI'm reading them15:19
LunaVoraxpaul_boddie, http://en.qi-hardware.com/wiki/How_to_reflash/fr needs a little fixing but it's definitely better15:20
LunaVoraxIt's good, I fixed http://en.qi-hardware.com/wiki/How_to_reflash/fr15:22
paul_boddieLunaVorax: Things like choosing between the imperative and infinitive is always best done by a native speaker, I think.15:22
LunaVoraxMaybe15:22
LunaVoraxI have the advantage of having a French teacher as a mother15:22
LunaVorax:D15:22
LunaVoraxSo in any cases I can ask her15:22
LunaVoraxPlus, she isn't familiar with computer science so I can use her as noob-proof test15:23
LunaVoraxThat reminds me I should find time to help you translate the wiki15:23
LunaVoraxI was translating some things some time ago and College got in the way15:23
paul_boddieOK, thanks for the quick edit! I used to occasionally write simple French documentation a long time ago, and probably just followed the house style. I'll redirect Comment_reflasher to How to reflash/fr.15:23
LunaVoraxpaul_boddie, sounds like a very good idead15:24
LunaVoraxlead15:24
LunaVoraxdammit idead15:24
LunaVorax....15:24
LunaVoraxIt's too hot to type on the keyboard15:24
viricWhat do you think of the TI CC11XX radios?15:27
LunaVoraxBtw paul_boddie how did you know I spoke french?15:27
paul_boddieI saw your username in the history for the How to reflash/fr page. :-)15:28
LunaVoraxOh right15:30
LunaVoraxThe translation sounded familiar...15:30
paul_boddieYou made the right choice preferring your own page, then. ;-)15:34
LunaVoraxHahaha15:43
LunaVoraxIt would have been funny if I did not15:43
--- Tue Aug 21 201200:00

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