#qi-hardware IRC log for Thursday, 2011-01-27

wolfspraulrjeffries: I apologize if I offended you. I am seriously personally interested in this clarify of language thing... maybe suffering from being misunderstood as a non-native speaker.00:53
wolfspraulso I often pause when I use certain words, and wonder 'is this clear to someone else', etc. I actually enjoy that - cut down bloat :-)00:54
wolfsprauls/clarify of language/clarity of language/00:54
rjeffrieswolf no problem. let's agree that you find "proof of conept" to lack meaning. cool, and in fact atben IS going to be maore. it will become a product and be sold. No argument. so that was my mistake00:55
wolfspraullong way to get there :-) (don't tell me what you think it is right now :-))00:55
wolfspraulit starts with p...00:55
rjeffriesMy point was that IMO until 6LoWpan is integrated into a ??future?? device it will have limited appeal. that is all00:56
wolfspraulactually I'm more worried about ipv600:56
wolfspraulipv6 only. well I'm worried and excited at the same time.00:56
rjeffrieson a differnet topic I can understand your frustration with Ingenic, and desire to control your SC future00:56
wolfspraulipv6 is probably a headache worth going through.00:57
rjeffriesvery ambitious, but eventually diable.00:57
wolfspraulno frustration, they have a clear business model which they are open about and are not pulling any dirty tricks.00:57
wolfsprauluntil today the 4740 is their main cash cow00:57
rjeffriesI watched the EXCELLNT 6LoWpan video and am convince that a small reaLITIVE LOW COST DEVICE WITH 6lOePAN CAN BE A REAL SELLER00:58
wolfspraulI haven't even watched it yet.00:58
wolfspraulwhich one was it?00:58
rjeffriesSORRY about caps lock00:58
rjeffriesit is damned good. they guy knows his stuff. there is far more going on than I dreamed00:58
rjeffriess/SC future/SOC future00:59
rjeffriesand on MM without memory management, one User app can easily crash other apps or the system01:01
rjeffriesis memory management required? No. Do we want CPU w/omemory managment in 2012? prolly not01:02
wolfspraulmmu is a very important missing feature, agreed01:16
wolfspraulbut let's have some faith in sebastien's prioritization01:16
wolfspraulthe first one is to make the product work as a product, that's why I teamed up with him and in the last year we executed brilliantly along those lines01:16
wolfspraulthere is soooo much work left to do there though...01:16
wolfspraulif we continue well, the mmu is just a missing feature. does anyone have a patch? :-)01:17
DocScrutinizerhmm, PoC is usually demonstrating working principle, though bells and whistles missing, like a decent GUI etc01:18
DocScrutinizermoo btw01:18
DocScrutinizerweird, IRC feels like full moon the last few hours, and yet it's not even Friday01:22
DocScrutinizerwell, the user I +q'd an hour ago said thanks to me in private query01:23
DocScrutinizer#openmoko had a weird spammer01:24
DocScrutinizerand here the mood seems quite agressive01:24
wolfspraulin #qi-hardware ? ooops :-)01:25
wolfspraulmaybe you are right and we should be more peaceful and understanding around here...01:25
DocScrutinizerhmm, I think *you* are as peaceful and understanding as it gets01:26
DocScrutinizerI have to agree with rjeffries though - I also have no f'ng clue why he spends time here :-P01:28
wolfspraulI can't figure out which dc jack m1 is actually using :-)02:13
wolfspraulhere's a table http://en.wikipedia.org/wiki/Coaxial_power_connector02:14
wolfsprauland this is the link to the datasheet from the m1 bom http://www.scpre.com.tw/D1-D21/d8.htm02:14
wolfspraulhere's an alternative ds http://downloads.qi-hardware.com/hardware/milkymist_one/datasheet/Supply/DCJ-R-20-LF.pdf02:15
wolfspraulthe bom says "DC POWER, 2.1mm"02:16
DocScrutinizerwolfspraul: *PIN* diameter 2.002:51
DocScrutinizerso it seems it's a standard 6mm barrel connector with a 2mm pin(-hole). There are connectors with same outer diameter but different pin02:53
DocScrutinizerwolfspraul: seems my universal PSU here comes with 4 6mm-outer barrel plugs, with 4 different inner hole diameters02:54
DocScrutinizerthen there's also a so called 2mm barrel connector, that has *outer* diameter of 2mm02:56
wolfspraulwhat does that correspond to on the wikipedia page?03:02
wolfspraulwikipedia talks about 5.5mm od and 2.1mm pin diameter03:05
wpwrakGRRR. blackout from 22-5 h :-(03:08
DocScrutinizerwolfspraul: well, your 5.5mm are my 6mm03:13
DocScrutinizer2.1mm == 2.0mm03:13
DocScrutinizerI.E probably the hole in plug is 2.1mm while the pin is 2.0mm03:14
qi-bot[commit] Wolfgang Spraul: first product label draft http://qi-hw.com/p/m1/5cff85a03:15
wolfspraulwpwrak: the qi servers suffered a similar fate - down for 8h!03:15
wolfspraullooks like a kvm bug/crash, oh well03:15
wpwrakwolfspraul: oh, fun :)03:16
DocScrutinizerwolfspraul: DIN 45323 power connectors:  6.00 mm OD, 1.98 mm ID [1]  (wikipedia)03:19
wpwrakwolfspraul: (label) looks quite messy ?03:21
wpwrakCE touching the "thankyoiu", TIP+ printed over RoHS. no spacial separation between 2A and TIP+, FCC also very close to the "thankyou", ...03:22
wpwrakif there's not enough space, maybe just remove the "thankyou" line ?03:23
wpwrakthe "made on earth" also looks a bit silly. also, doesn't the place of manufacture have some significance for customs ? (as meaningless this may be in practical terms)03:24
DocScrutinizerdamn, those barrel connectors are a nightmare of a zoo03:26
tuxbrain_awaywolfspraul: yes is a legal matter , the device must especify his origing....03:29
tuxbrainand yes I agree with wpwrak is a pure mess....03:32
tuxbraina pity we are now overhelmed by other matters , but if it not urgent, If you pass the requeriments, and space I can ask victor to take a look at, btw where is suposed to this label will appear?03:33
wpwraklooks like the label that goes on the box. hence little need for the "thankyou" advertisement link. i like the +/- 5%, but i wonder if it is true ;-)03:35
DocScrutinizerI wonder what's a "2A TIP"03:36
DocScrutinizer(+/-5%) unusual for PSU03:37
DocScrutinizeris this direct VCC?03:38
wpwrakDocScrutinizer: either TIP+ is some certification/compliance, like CE, FCC, or RoHS (after all it appears right next to them), or it is an unusual way of repeating the information already present in the drawing below03:39
DocScrutinizeror just input to some sorta voltage regulator / converter?03:39
wpwrakDocScrutinizer: in any case, I wouldn't trust it ;-)03:39
wpwrakDocScrutinizer: with that accuracy, it ought to go straight to TTL Vcc ;-)03:40
DocScrutinizeroh, that's TIP+. Missed the +03:40
wolfspraulwpwrak: are you looking at the svg?03:41
wolfspraulwithout the font it will be wrong, look at the png instead03:41
wpwrakwolfspraul: looking at the svg, yes. let's see about PNG ...03:42
DocScrutinizerUsual writing would be DC 5V 2A, plus the pictogram below. Or - if it's really +/-5% - you should replace that with a "warning!! use only specified PSU"03:42
DocScrutinizerI don't see a png link03:43
wpwrakyeah, no png. (substituted svg with png)03:44
wolfspraultuxbrain: where does it say that it 'must specify its origin' and what is the origin?03:44
DocScrutinizerwell, some countries seem to require that03:46
tuxbraincountry of origin, I think CE countries have permision tu use CE instead of the name of the country, let me check the exact rules but I'm pretty sure about it.03:46
wolfspraulI'm listing origin as Earth until I find something better.03:46
DocScrutinizerwhere's schematics of m1?03:46
tuxbrainwhat something about made in china in partnership with all world03:47
wolfspraulnothing made in China03:47
tuxbrainor taiwan03:47
wolfspraulif you ask Beijing then it's made in China03:47
tuxbrainor whatever is manufactured03:47
wolfspraulwe know better than to randomly pick a country03:49
wolfspraulics from the us, south korea, japan03:50
wolfspraulpcb from taiwan, maybe china in the future03:50
wolfspraulcase from germany03:50
wolfspraulsmt and assembly in taiwan03:50
wpwraki think it's something like the point of assembly. like that, you also have TVs "made in argentina"03:50
wolfspraullogistics center hopefully hk soon03:50
wolfsprauloh I'm sure, so what? I'd rather stick with the truth, and see how far I can go.03:51
wpwrakwhy make things hard for customers ?03:51
wpwrakit's them who have to wrestle with their customs03:51
wolfspraulI doubt there will be a single problem, this is just a label.03:52
wolfspraul'country of origin' is mostly just a meaningless field in customs paperwork, you can fill anything in.03:53
DocScrutinizerwell, for all I can read from that confusing page 8, I suggest to change the writing to not include any +/-x% rating03:53
wpwrakpng looks better than svg, but still a bit disorderly. i'd drop the "TIP+". i understand what it means because i saw it being discussed here, but otherwise i would find it confusing.03:53
wolfsprauland nothing woudl stop you, or us, or anyone, from still filling whatever fits in there.03:53
DocScrutinizerand there's A FUSE MISSING!!03:53
wolfspraul'made on earth' or 'made by the people' is still accurate, if some customs bureaucrat wants to see a member of the UN listed in his form, we'll give him that name :-)03:54
wpwrakDocScrutinizer: is there a fuse ? in the device, i mean ?03:54
DocScrutinizernope, it's missing03:54
wolfspraulDocScrutinizer: the power circuit is 'weak' by my standards, is that what you mean?03:54
DocScrutinizeryou can't labek such device CE03:54
tuxbrainIt has pass CE?03:55
wolfspraulCE is a self certification03:55
DocScrutinizerCE is self-cerified03:55
wpwrakwolfspraul: i can so see it. the bored customs official needs to put some country. the program doesn't have "earth" or "people". he's having a bad day, so stamps it as "incomplete customs paperwork". and it either goes back or the customer gets a letter asking them to handle it personally.03:55
wolfspraulyes, it passed including the power adapter03:55
wolfspraulwpwrak: who says that anybody would fill in 'earth' in that field?03:56
wolfspraulwe know they have no humor, or reason03:56
wolfspraulDocScrutinizer: tell me more about the fuse03:56
tuxbrainWolf, I understand you point but.... don't angry customs please!!! I'm in the other end!!!, and they are reaaaaaallly picky lately03:56
wolfspraulwhere is it missing?03:56
wpwrakwolfspraul: so, why not just make it "Taiwan" and be done with it ? is this a fight you really need to pick ? :)03:56
DocScrutinizer2A slowblow, in J11:1 -> C14203:57
DocScrutinizeror a littelfuse03:57
tuxbrainwolfspraul: listen wpwrak please03:57
tuxbrainplease please please03:57
wolfspraultuxbrain: did you have any issues with the latest nanonote or m1 packages?03:58
DocScrutinizer"MAde in" usually is location of final assembly03:58
wpwrakfor the voltage rating, it would be very useful to determine the real range (with whatever safety margin you need to keep this simple)03:58
DocScrutinizerwolfspraul: this fuse issue is no joke03:58
wolfspraulDocScrutinizer: I'm too electrically challenged, don't understand you. what would you change/add regarding the fuse? And why do you think without that change it cannot be self-certified as CE compliant?03:59
wolfspraulDocScrutinizer: oh I think the power circuit on m1 can become much stronger, no doubt.03:59
tuxbrainluckily not but I have a bureocratic nightmare with some Pakages form USA.03:59
wolfspraulwe simply didn't have that time/feedback yet, now thanks to you maybe we do...03:59
tuxbrainjust because they don't like the invoice format...04:00
DocScrutinizerAIUI a CE device must have a safe failure mode. Shorting external PSU on arbitrary regulator failure is not safe, it will make your PCB go up in flames04:00
wolfspraulthe guy checking for CE compliance had no issue at all with the current design and performance (there were a lot of tests that included the power supply)04:00
DocScrutinizerwolfspraul: my suggestion: change order: cut connection from J11,pin1 to C142. Insert a 2A fuse slowblow04:01
wolfspraulso if we think we know CE better than him, we need to pinpoint him to the exact regulation04:01
DocScrutinizero.O on some guy doing CE. It's usually self-issued04:02
wolfspraulyes, but how can I (a non-electrical guy) self-issue? only after having done CE tests with a professional, in this case reichl-emv.de04:03
wpwrakwolfspraul: maybe ask him about the fuse issue ?04:03
wolfspraulirrespective of that, at some point we will probably want to make the power circuit more robust04:03
DocScrutinizerwolfspraul: I don't really know about regulations. I know about proper design though, and about common practice in industry (which maybe allows conclude about regulations)04:03
wpwrakwolfspraul: maybe he thought you were only concerned with EMC ?04:03
wpwrakDocScrutinizer: old friend of yours ? ;-)04:04
DocScrutinizerif that's what I think it is, then no chance they find such issues :-P04:04
wpwrakDocScrutinizer: http://www.reichl-emv.de/NEU_index.html04:05
wolfspraulor maybe the 'issue' doesn't matter because there are tons of products with this level on the market, and it actually doesn't conflict with CE regulations at all04:05
DocScrutinizerELV is some strange company with a monthly magazine that frequently has poor schematics like that04:05
wpwrakDocScrutinizer: does look like EMC-centric specialists, no ? :)04:05
DocScrutinizerOhh, nope. Completely off topic. Sorry04:06
wolfspraulDocScrutinizer: for the m1 power supply, I'm mostly worried about the lack of wrong polarity protection, and maybe some over-voltage protection too.04:07
DocScrutinizeryes, that's an EMC specialist, and he probably doesn't know much about DC04:07
wolfspraulguess what - I take his word and self-certify as CE compliant :-)04:07
DocScrutinizerOVP is maybe bearable. Fuse and inverse polarity prot isn't04:07
DocScrutinizerand it goes hand in hand04:07
wolfspraullike I said - I know the m1 power circuit is not robust enough, in my personal opinion04:08
wolfspraulbut among the many priorities, we simply haven't gotten to making it stronger yet04:08
wolfspraulwith your help maybe we can :-)04:08
wolfspraulwe may even try to fold that change into rc3, if it's not too intrusive04:08
DocScrutinizerwolfspraul: my suggestion: change order: cut connection from J11,pin1 to C142. Insert a 2A fuse slowblow. add a 2A diode parallel to C142, so it shorts VCC when polarity is inverted04:09
wolfspraulwill there be very large components / big relayout / big electrical risks?04:09
wpwrakDocScrutinizer: is adding a (chip) fuse something that could be done with rework ? or is it too unreliable ?04:09
DocScrutinizerrework like field change order?04:10
wolfspraulonce we're at it we can also look at the ovp. right now it's weak that's why we have this +/- 5% thing04:10
DocScrutinizerdepends on PCB layout04:10
wolfspraullet's focus on rc304:10
wolfspraulwe have layout changes anyway for rc304:10
wpwrakwolfspraul: are you really sure about that 5V +/- 5% ? sounds very very unusual04:10
wolfspraulthis is the perfect time to suggest more things for rc304:10
DocScrutinizerchop fuse is fine. I'd suggest a littelfuse self healing anyway04:10
wolfspraulwe had a chart/graph somewhere that listed this04:11
wolfspraulI think up to 5.6 or so may still be ok04:11
wolfspraulbut over 6.2 will definitely be above what some ics are speced for04:11
wpwrakwolfspraul: still pretty narrow. (6.2 V) wow.04:11
wolfsprauleven less I think04:11
wolfspraulthere was a chart/graph somewhere, but I doubt I can find it now04:11
DocScrutinizermake that diode suggested above a 6.3V Zener, that acts like a normal diode in reverse voltage04:12
wpwrakthen i'd agree with DocScrutinizer: this should be a warning label04:12
wolfspraulwe have to wait for Adam, who is currently at Xilinx office to track down the power-up bug (#1 on the wiki page)04:12
DocScrutinizerwolfspraul: o.O04:12
DocScrutinizerthe regulators will blow chunks on 6.2V input?? how that?04:12
wolfsprauldon't take my word, we need to find the exact chart04:13
wolfspraulor wait for Adam04:13
wolfspraulwho is offline right now because he's with Xilinx :-)04:13
wpwrakvirtually anything that works on 5 V also tolerates 6 V +/- 10%, usually much more. so there's a real risk that the device could be connected to a supply that goes too high.04:13
wolfspraulI would love to make the power supply stronger, that's really very helpful and great timing too, if we can fit it in for rc3.04:13
wpwrakwill tht sticker go on rc2 or only >= rc3 ?04:13
DocScrutinizerif this schematics wasn't so messy, It'd be simple to track down which chips are connected to %V VCC04:14
DocScrutinizerbut I can't swear there's no other weird 5V inverted ground symbol anywhere in these 12 pages04:14
DocScrutinizerwpwrak: absolutely - yes04:15
DocScrutinizerwpwrak: but I really doubt there's any chip that will blow on 6.2V04:15
DocScrutinizerI mean, it seems that's all regulators. as long as R182 NC04:16
wpwrakDocScrutinizer: well, some chips that connect to power have very low "absolute maximum values". usually usb-ish things, though. for common regulators, it seems very picky.04:16
wpwraki've seen some freaky ultra-low-power regulators that have a very low maximum vin, but i suspect that's more for thermal considerations than anything else. can't get all that much heat out of a 2x2 mm BGA ;-)04:18
DocScrutinizeralas the pdf isn't searchable. So no way to make semi-sure there's no other 5v offsheet symbol somewhere on p5604:19
wpwrakand of course, the supply voltage should be specified for normal operation, not the maximum non-destructive voltage. so if your regs go into thermal shutdown all the time but don04:19
wpwrakt take any real damage, that's still not good enough04:20
wpwraknote to self: should write a net to components lister for kicad04:20
DocScrutinizeralso I wonder what for are those L7 L8 L9 L10 L12 L2004:21
DocScrutinizerlooks contraproductive to me04:21
wpwrakbeads ?04:21
wolfspraulthere's something, but not what I was looking for http://en.qi-hardware.com/wiki/Milkymist_One_Power_Tree04:22
DocScrutinizerusually you don't want to worsen the regulator's phantom impedance04:22
DocScrutinizerlooks like an invention of an EMC specialist :-P04:23
wolfspraulI definitely want to limit changes in rc3 to low-risk stuff04:25
DocScrutinizerme too04:25
DocScrutinizerjust kicking out high risk cruft :-)04:26
wolfspraulbecause not only is rc2 fully functioning, but it also passed the tests I need to happily self-certify CE04:26
DocScrutinizerwell, the above mentioned Lnn don't help. Your tests were not passed because but rather despite them being there04:27
wolfspraulI will find that overvoltage analysis chart later, when Adam is bac.04:27
DocScrutinizerrisk of such a design is regulator starting wild oscillating04:27
DocScrutinizerlook, the regulators are all linear type, and they don't create any noise or RF that needs to get filtered from the rest of circuit04:29
DocScrutinizerand they aren't sensible to noise feed back from rest of circuit either04:29
DocScrutinizerbut they might start oscillating when a mostly inductive load is seen on output04:29
DocScrutinizerso Lnn are dangerous, if anything04:30
wolfsprauljust gimme the bottom line :-)04:30
DocScrutinizerthe good part of the story - you can replace them by 0R04:30
wolfspraulI am not aware of any problem there. How would one see it?04:30
wolfspraulwhat does need improvement is the power supply circuit, wrong polarity04:31
DocScrutinizeryou'd see it on more noise in the rest of circuit VDDs making the thing instable, and occasionally you might find one board or another actually oscillating on one of the regulator outputs04:31
wolfspraulok we need to see what Adam or lekernel think of it, for me no need to change something I don't understand and afaik causes no problem.04:32
DocScrutinizerit increases chances you will see problems eventually, occasionally04:32
wolfspraulwe've had two runs now, first one 6 boards, second one 40 boards.04:33
wolfspraulnext one will be 80-90 boards.04:33
DocScrutinizeryou also won't usually see problems without fuse04:33
DocScrutinizerit's all about risk management: design so the potential problem sources get minimized04:34
DocScrutinizerthere's no use in cargo cult designing. And I don't see any other reason for those C236, L20, C237 and siblings04:35
DocScrutinizersuch design makes sense for maybe electron tube radios04:35
DocScrutinizernot though for linear regulator IC04:36
DocScrutinizerfor OVP you got my above suggestion - BOM add: 1 fuse, one Zener04:37
DocScrutinizerI can scrutinize the whole schematics later on, and pull the datasheets for the regulators04:38
DocScrutinizerto verify this suggestion is sufficient for the purpose04:38
DocScrutinizerI'd prefer a Zener + Thyristor design though, as otherwise on small OV (like 6.5V) the Zener might break before the fuse trips04:42
DocScrutinizeranyway I guess all the regulators are rated sth like max in = 30V04:43
DocScrutinizerand probably even have OverTemp shutdown04:43
DocScrutinizerso no probelm at all with OVP04:44
wolfspraulDocScrutinizer: alright, let's write your suggestions up on the wiki a bit04:49
wolfspraulwhat should I write exactly?04:49
wolfspraulwhat should we write about L7/8/9/10/12/20, C236/237?04:49
wolfspraulthis is our work basis for rc3: http://en.qi-hardware.com/wiki/Milkymist_One_RC2_Known_Issues04:50
DocScrutinizerinsert fuse between powerjack and first capacitor (see above). Add a diode parallel to first capacitor, to short inverse polarity and make fuse trip04:50
DocScrutinizermake all the Lnn as mentioned above 0R  (those L that are in regulator outputs together with the 100nF Cs)04:51
DocScrutinizerthe C's nectby are cruft then, one of two, as well04:52
DocScrutinizerbut don't hurt, so you might consider removing one of two, and make the other 200nF, for BOM reduction04:52
wolfspraulalright, I just copy/paste it for now http://en.qi-hardware.com/wiki/Milkymist_One_RC2_Known_Issues04:54
DocScrutinizerscratch R182 pads, looks dangerous04:54
wolfspraulI'm pretty sure there are good reasons for all of this, but we can go through...04:56
DocScrutinizerDAMNIT! midi tx U7, p604:56
DocScrutinizerok, add a <tbd>R in 5v VCC of U7, add a 5.2 Zener to gnd04:59
wolfspraulyes, what?04:59
DocScrutinizerwolfspraul: that's your chip that emits magic blue smoke on 6.5V04:59
DocScrutinizerneed to check max I it takes from Vcc and min voltage it can work on, then specify above mentioned R accordingly. to be done later05:00
DocScrutinizerI need to check the whole schematics, if there's some other similar bugger05:01
DocScrutinizerif somebody could toss me over the midi bus specs, that'd help05:02
wpwrakwolfspraul: btw, will adam work through the CNY week ? i'm going over my prototype schedule. with the delays and infrastructure (power) problems, i probably won't have them ready before middle next week. when i ship them right then, they'll arrive in the middle of CNY.05:03
wpwrakDocScrutinizer: your google is probably as good as everyone else's :-)05:03
DocScrutinizerI bet it is05:03
DocScrutinizerI'm just incredibly lazy :-)05:04
wpwrakDocScrutinizer: i know ;-))05:04
DocScrutinizerprobably even no google, just wikipedia05:04
DocScrutinizeranyway, have to care about my own processor's VDD now, bbl05:05
wpwrakwolfspraul: an alternative schedule would be to complete the atben boards, then try the avr-based atusb design. if that works well (and quickly), i'd scrap the current atusb pcbs (still unpopulated) and make avr-based ones instead. this would add about one more week, if things go well.05:06
wpwrakwolfspraul: if things don't go well, i'd fall back to the current atusb design and save the avr for later. i have parts for either choice.05:08
DocScrutinizerbtw it seems to me like midi-rx is actually a tx circuit, and midi-tx is a rx circuit then? Then I wonder why midi-tx needs 5v Vcc at all05:09
DocScrutinizerwhile midi-rx is sending with only 3.3V current loop05:09
DocScrutinizeraah, scratch that05:10
DocScrutinizerfriggin missing offsheet symbols05:10
lekernelMM runs only _one_ app. do I need a MMU for that? no. so stop nerding around...05:18
wolfspraulnothing better than good focus :-)05:18
wolfspraullekernel: can you quickly glance over some of DocScrutinizer's ideas and see which make sense to you?05:18
wolfspraulwpwrak: you are talking about changing the usb firmware chip to avr?05:19
wpwrakjust imagine how much more focus you'd get with 0 apps ;-)05:19
wolfsprauladam definitely will not work over cny, and I don't think our schedule is that precise/fast anyway05:20
lekernel+/- 5% comes from the USB standard (the DC input powers directly the USB ports)05:20
wpwrakwolfspraul: yes. didn't you see that one ? i mentioned it on the list a while ago05:20
wolfspraulyes I saw it, I just wasn't clear on the practical consequences yet05:20
lekernelbesides, there are current limitations in every regulator, on the USB ports and in the wall plug... do we really need a FUSE? I've never had any power supply problem05:21
wpwrakwolfspraul: (adam) so he'll be afk or even away from his mailbox in the week feb 3-7 ?05:21
wpwraklekernel: i think the role of the fuse it to blow when that one-in-a-million problem happens that then causes your board to overheat, burn, and torch your apartment. doesn't happen often, but when it does, it's no fun.05:22
wolfspraulwpwrak: he may not even be in Taipei, yes05:24
wpwrakwolfspraul: the practical consequence is a redesign of the board (the easy MCU side, not the messy RF), hopefully without changing the size, evaluation of the noise filter on the clock signal (that's something new it needs, due to the AVR needing an external clock source),05:24
wpwrakwolfspraul: and then new firmware for the AVR. there's a USB stack I may be able to use, which should help speed things along. the transceiver-specific side of the firmware is very simple.05:25
DocScrutinizerlekernel: there's no power limitation in the PSU, simply because you allow to plug in random ones. The current limits in regulators are fine as long as no shit happens. Fuse is for the case when a regulator breaks and shorts VDD to GND05:26
lekernelremoval of ferrite beads: maybe, I put them there in case we had to fix EMI problems or to be able to disconnect a power supply rail in case of issues05:26
lekernelbut I wouldn't remove them since:05:26
lekernel1. they are dirt cheap05:26
lekernel2. it works with05:26
lekernel3. it would be wise to re-do some EMC tests after removing them, which takes time05:26
wpwrakwolfspraul: in addition, i'll also need a new programming cable (one more signal) and maybe just write a little avr programmer (probably quicker than adapting one of the existing monstrosities)05:26
lekernel4. there are super-more important issues like software05:26
lekernelwhen we make 1000 units a month, we'll talk again about that05:27
DocScrutinizerbeads in power rails are generally a bad idea05:27
lekernelI don't care, it works05:27
DocScrutinizerand will increase EMI problems, not lower them05:27
DocScrutinizeroh well, ok05:28
lekerneldepends, sometimes the device has to accept incoming power perturbations05:28
lekernelin this case the ferrite beads and the caps help05:28
DocScrutinizerand ciao05:28
wpwrakwolfspraul: (adam away) okay, seems that the avr will get its try :) (ah, and i think it's a few cents cheaper than the silabs chip as well :)05:29
wpwraklekernel: they're on the primary side then ?05:29
lekernelwhat? the beads? primary side of what?05:30
lekernelthey're in series with each rail05:31
lekerneland right at the DC jack input05:31
wolfspraulok if we cannot discuss and go through joerg's feedback I will remove it from the wiki page to not confuse adam. I'll leave it there another day or two to see what happens, then it moves to the page history :-)05:31
wpwraklekernel: primary side of the regulators, i mean. if the perturbations come from the DC supply, you'd want to catch them there. if they make it past the LDO, i doubt any filter on the secondary side will accomplish much.05:38
lekernelmaybe they're useless05:40
lekernelbut finding out is extra work that's not worth it imo05:40
wpwraklekernel: well, in case they confuse the LDO (not sure how likely that is to happen). you can always 0R them. that's an easy change :)05:42
lekerneland I didn't mean to offend Joerg, but I had to say clearly that we shouldn't spend time on that until we run into actual power supply problems...05:43
lekernelyeah, exactly, it's an easy fix05:43
wolfspraullekernel: the discussion started when Joerg thought a design without a fuse is bad.05:43
wpwraklekernel: properly fusing/protecting the whole thing is probably more work05:43
lekernelthere are already current limits everywhere... isn't a fuse overkill?05:44
wolfspraulif we make changes in rc3, I would like to get consensus between lekernel and adam that they are good and will not introduce regressions05:44
lekernelalso, a fuse has some parasitic resistance, too05:44
wolfspraulif we cannot get such consensus -> no chnage05:44
wpwraklekernel: a fuse is basically your safety belt. of course, everybody drives carefully, so accidents should never happen ...05:45
wolfspraullekernel: I trust Joerg's judgment, although there may be misunderstandings somewhere since he just started to look at the schematics, and has stopped for now, I'd think :-)05:46
wpwraklekernel: (offend joerg) i think the styles of both of you are very fertile for conflict ;-)05:46
wolfspraulfor sure05:46
wolfspraulwpwrak: that was an energetic and quick 'dialogue'05:50
wolfsprauland it all started with a little label at the bottom of the case...05:54
kristianpaulyou can swich made with assemble05:55
kristianpaulthere are lots oh mobile phone like that05:55
wejpwolfspraul: thanks for your suggestions about gmu05:57
kristianpaul(5V +/- 5% ? sounds very very unusual) indeed05:59
wpwrakkristianpaul: "you can swich made with assemble" qu'est-ce que c'est ? ;-)05:59
wpwrakkristianpaul: the highly precise 5 V looks like trouble. particularly if you need to strip the original supply for import reasons. such an accurate monster will be hard to source reliably.06:01
wpwrakalso the absence of OVP is worrying. it's not unheard of that such adapters leak a bit of AC.06:02
kristianpaul(do we really need a FUSE? I've never had any power supply problem) use the FUSE and you never will :-)06:06
wpwrakkristianpaul: well, how often has a seat belt in a car saved your neck so far ? :)06:08
kristianpauli meant, sorry i just wake up, in the M1 Label you can _replace _ Made on with Assemble in06:11
kristianpaulwpwrak: (highly precise) sure, you'll never get that06:12
wpwrakkristianpaul: aah ! yes, you see that from time to time, true. not sure what exactly is required there. the "made in ..." seems to have some semi-legal meaning.06:12
kristianpaulhe i just discover my M1 power adapter, have a 5V laberl overwritten a 12V one :p06:12
kristianpaulah legal..06:12
kristianpaulwpwrak: i'm afraid there is not detection/protection in the a case wrong polarity power adapter is pluged..06:14
kristianpaulI need check sch06:14
wpwrakkristianpaul: (never get highly precise) the first challenge would be to find a power supply where the tolerance is even specified06:14
kristianpaullater, i gotta run work now !06:14
wpwrakkristianpaul: most of the switching supplies are actually surprisingly accurate. but then, you never know under what conditions...06:15
wpwrak(12 V overwritten) ouch !06:15
kristianpaulbut i tested before !!06:15
kristianpaulthe label was TOOO obvious06:15
kristianpaulmay adam or wolfspraul can explain that ;-)06:16
tuxbrain_awayany one can spend with me some time to help me in update the MM to Flickernoise 0.2 and SoC-R2?06:18
tuxbrain_away0.1 seems to not support softw reflashing...06:19
tuxbrain_awayI just realize on one thing!06:30
tuxbrain_awayops maybe I put this in mm chanel06:30
wolfspraulwejp: oh very welcome, I'm happy to use your software.06:31
wolfspraulI'm slower/later than others, but I really enjoy it now :-)06:31
wolfspraultuxbrain_away: you should have the jtag-serial boards with your m1, no?06:33
wolfspraulif 'made on earth' is not as enthousiastically welcomed as I hoped, maybe I just remove it altogether06:33
tuxbrain_awayyes I have but I have no Idea on how to usit and MM wiki instructions are a little bit messed for my level :(06:33
wolfspraulalthough any implications with actual customs are far far exaggerated imho. that's not how the world works.06:34
tuxbrain_awayis a good marketing motto but bad idea in the adminstrative side on things06:34
wolfspraulno way.06:34
wolfspraulcustoms works with forms, it's a bureaucracy06:34
wolfspraulfill out something wrong, and there is a problem06:35
wolfspraulbut whether something is blatantly wrongly declared or not won't matter in 99.9% of cases06:35
wolfspraulI can say 'made in togo' and build a multi-billion USD business although it's made in iran, of all places.06:35
wolfsprauluntil finally someone says 'hey, there are these sanctions, no?"06:36
tuxbrain_awayMade on earth /Assembled in Taiwan doesn't fit?06:36
wolfspraulI don't like 'assembled in', it reduces the copyleft value for no good reason.06:36
wolfspraulI make a second label for the chickens.06:37
wolfspraulthat will be glued all over, top, bottom, sides, box.06:37
wolfspraulhopefully that will not conflict with some regulation somewhere, but hey, you asked for it...06:37
wolfspraulwhat I care about is the thankyou url, to give credits to the various people and companies involved06:39
wolfsprauland also the cc-by-sa logo, as well as ce/fcc logos because these are good quality standards that I actually care about06:39
wolfspraulsame for rohs06:39
wolfspraulor other environmental standards, carbon footprint, etc.06:40
tuxbrain_awayman, we are brave enough to build, pay and market hardware against any previouse way done in hardware industry.... don't make all this faith steps fail for fight against custosms, onces we move full cargo ships of our material, we will have a position where we can argue against06:41
wpwrakwolfspraul: (customs) blatantly wrong but plausible (within the system) would be better than implausible (in the context of the system) :)06:41
wolfspraulwpwrak: give you an example about your beloved power supplies with the 'right' symbol on them06:42
wolfspraulif the symbol is not visible somewhere, we are in trouble06:42
wolfspraulI'm 100% with you06:42
wolfspraulbut if the stupid symbol is printed on a label somewhere, which odds do you think there are that any _actual_ technical compliance with that standard will be checked/verified?06:42
wolfspraulso the issue is just on the other side to know about the symbol, and go through the pain of modifying the label and printing a new one06:43
wolfspraulthat's it06:43
wpwrakwolfspraul: unlikely. but with "made on earth", it sticks out. so if there's some obscure requirement to declare the "origin" (in whatever sense), then you made it clear that you're refusing to comply.06:43
wpwrakwolfspraul: that's different from a white lie06:44
wolfspraulthose are _declaration_ requirements06:44
wolfspraulin a form06:44
wolfspraulnot on a label somewhere06:44
wolfspraulget the label right, form wrong: not pass06:44
wolfspraulget the form right, label wrong: pass06:44
tuxbrain_awaynot, agree, get the form right, label wrong and you maybe pass06:44
tuxbrain_awayif a visual inspection is done (and I suffer this ones) you will not pass06:45
zrafaI got an freerunner with its charger wihtout problems, without any label.06:45
wpwrakwolfspraul: if you're sure about that, okay. one question is also whether customs know precisely what is required. sometimes they don't know the fine print of the law either. for them, it may be paperwork and label agree -> pass. else, be more suspicious.06:45
wolfspraultuxbrain_away: do you have more precise info? I always like to learn, just need to have hard data.06:46
tuxbrain_awayyes and time to time I get custom passed without paying taxes but that's not the point06:46
wpwrakzrafa: i had three shipments held up because they contained chargers.06:46
wolfspraulyou had a problem with 'country of origin' declaration somewhere?06:46
wolfspraulwhat was it exactly?06:46
wolfspraulyou attach way too much significance to a label06:47
zrafawpwrak: yep, I know.. which kind of stuff had those three shipments?06:47
wolfspraulthere is no regulation that says a product with a faulty label cannot be imported06:47
wolfspraullegal is funny.06:47
wolfsprauladamw_: hi good evening!06:47
wpwrakzrafa: eventually i grew bored of driving to ezeiza twice per shipment and spending half a day on each trip, plus a considerable amount of money to get those things unblocked. so then i started to request that power supplies be removed before shipping. never had an issue since.06:48
tuxbrain_awayI have found references a lot of references  but are in spanish and relative vage, not clear  legal article yet06:48
wolfspraulsure they make a big fuss about 'country of origin'06:48
tuxbrain_awaylegal is hell06:48
wolfspraulmostly for their flawed statistics06:48
wpwrakzrafa: an OQO 01 with DHL, and OQO 01+ with FedEx, my first openmoko "kit" (phone, jtag, and unfortunately some external power circuit) via DHL06:49
zrafawpwrak: yeah.. and ezeiza has some mafia about shipments surely.. they ask money and you better be quiet06:49
tuxbrain_awayit also has to vio of extra taxes depending on country of origin06:49
wolfspraulbut what about traders who buy, relabel, resell?06:50
wolfspraulyou think those things don't exist?06:50
wpwrakwell, there was one more where i had troubles: another openmoko phone with DHL. there, they did an extra examination because they weren't sure what this thing did. that was in a single trip, though, without bribes06:50
wolfspraulit's not even illegal!06:50
wolfspraulthose are very legitimate and honoroable businesses :-)06:50
wolfspraulthey deal with real problems in our real world06:50
tuxbrain_awayyes !! and incompte firmwares or partial asembled hardware and alot of tricks06:50
wolfspraulship from a to b, relabel, ship from b to c06:50
wolfspraulif someone thinks this should be valuable, so be it06:50
wolfspraulthat's what we have capitalism for06:51
adamw_good evening!06:51
wolfsprauladamw_: before you talk about Xilinx, I had another question06:51
wpwrakwolfspraul: (attach significance to a label) it's not me per se. it's those fine people working at customs. they look at those labels. they believe it's harder to falsify a label than the documents.06:51
tuxbrain_awayI thinkg I going thoug TFTP approach to flashing MM06:51
wolfspraulwe were talking about the overvoltage protection in m106:51
wolfsprauland I remember one time you published a little graph/chart about voltage ranges, but I couldn't find it anymore06:52
wolfsprauldo you remember where that was?06:52
wolfspraulwpwrak: ok, we keep our label easily removable anyway, that's a good quality to have.06:52
wpwrakwolfspraul: ;-))06:52
wolfsprauland I remove the 'made on earth' upon popular request.06:52
wpwrakwolfspraul: excellent !06:52
wolfspraulwpwrak: how about 'made by the people'?06:54
wpwrakwolfspraul: please ...06:54
wolfspraulok ok06:54
wolfspraulyou guys are holding me down, what can I do06:54
wpwrakwolfspraul: you could also put the A-in-circle amindst the certification logos (-:C06:55
wpwrakamidst even06:55
tuxbrain_awayput f&#ing made in taiwan , damn it06:55
tuxbrain_awayhahaha wpwrak06:56
tuxbrain_awayalso a beautiful logo with an axe and a snake twisted on it will sure be very wellcome here in spain06:58
wpwraktuxbrain_away: there's a place called Fuxing in taipei. "MADE IN FUXING TAIWAN" ? ;-)06:59
wpwraktuxbrain_away: what's the axe-with-snake ?06:59
tuxbrain_awaysome svasticas in the corners and a little of arab "Ala agbar"06:59
tuxbrain_awayETA's logo07:00
adamw_wolfspraul, why do you want to know this? I bet very few people noticed this carefully before! :-)07:00
wpwraktuxbrain_away: i was just about to suggest the swastika ;-))07:00
zrafa"made in antartic, SC dark side industries"07:00
wpwraktuxbrain_away: and let's not forget a few unflattering pictures of mohammed07:00
wpwraktuxbrain_away: (eta) nice ! :) actually, to piss of argentina customs, "MADE IN FALKLANDS" ought to work miracles ;-)07:02
tuxbrain_awayoh yeah!07:02
wpwrakadamw_: eventually, all your dark secrets will be uncovered ;-) wikileaks will be followed by qileaks (wolfgang: too bad it's too late for mokoleaks - that would have been fun :)07:03
tuxbrain_awaywhy not "we qi leaks" :P07:04
wpwrakhehe :)07:04
adamw_wpwrak, I even really didn't know if I had have put this doc in the wiki already...:-) I also didn't find in Milkymist One wiki page myself.07:06
wpwrakwolfspraul: for the "thankyou", maybe put it elsewhere ? also, the URL sounds like the content would be "valued customer, thank you for choosing our excellent product", which doesn't seem to be what you have in mind ?07:06
adamw_so I think I really must "forgot" to put this link connected to wiki somewhere, damn it. :-)07:07
adamw_oah! guys! here: http://en.qi-hardware.com/wiki/Operating_Voltages_of_Milkymist_One%27s_Key_Parts07:08
adamw_let me put it to connect to home page now. :-)07:09
wolfsprauladamw_: we discussed this earlier, with Joerg07:12
wolfspraulwpwrak: do you think the 'A' symbol for Argentine would also matter on m1, or only the power adapter?07:12
adamw_wolfspraul, hmm..good, any news?07:13
wpwrakwolfspraul: only on the power adapter07:13
wolfspraulnot really, Joerg made some suggestions but we didn't find the right way to discuss them yet07:13
wolfspraulhow did Xilinx go? did you go there today?07:13
wolfspraulwpwrak: tuxbrain_away on the issue of labeling and customs, note that I shipped nanos to 30+ countries in the last year, and in every case, I took the full risk, meaning that I would have refunded the customer if there would have been any unsolvable problem.07:14
wolfspraulyet I am still the biggest risk taker in this, I wouldn't hesitate for 1 second to write whatever on such a label07:14
wolfspraulmy experience with those 30+ countries has been positive overall, I mean customs is still a pita, but in the end it's all just paperwork07:15
adamw_wolfspraul, i went to there of course.07:16
wpwrakwolfspraul: but you kept the labeling sane. CE, FCC, "made in China". and no power supply.07:16
wolfspraulnow the next level :-)07:17
adamw_got new ideas, so I need to modify one board and try to check.07:17
wolfsprauladamw_: that .odp is not what I meant.07:17
wolfspraulI meant a chart I remember once where you looked at which IC would fail at which OVERvoltage07:18
wolfspraulthe result was that up to 6.2 or so was still ok?07:18
wolfspraulor 5.7?07:18
wolfspraulI forgot07:18
wolfspraulanyway it was a little table that looked at the consequences of overvoltage to the various ics07:18
wolfspraulthe .odp you have here (which is very nice too) looks at operating voltages07:19
adamw_ha..right, you should see them carefully. :-)07:19
wolfspraulwpwrak: I'm still learning in all this, the big fish one day will undoubtedly be the phone.07:19
wpwrakis the plan to just fix small flaws for rc3 or could also be more substantial changes be considered ? e.g., i wonder if the whole power circuit shouldn't just begin with a big DC-DC converter that has a wide input range and that gives clean power to the rest.07:19
wolfspraulwpwrak: it depends whether we get this feedback in a form quick enough and high quality enough to make use of it.07:20
wolfspraulif it's just "the power circuit sucks", then rc3 == rc207:20
adamw_at the rightest brown color for LM4550 CODEC.07:20
wpwrakhmm, send a pack of valium and a sixpack of beer joerg's way ? :)07:20
wolfspraulin my urgency scale this is not at the very top, although if someone like Joerg would come and think it through a little and work on it together with Adam, we probably could do it.07:21
adamw_moment, let me pointed the page of datasheet. :-)07:21
wolfspraulactually ben-wpan will also allow us to get to 'the next level', because it is an 'intentional radiator'07:21
wolfspraullots of FUD associated with them, so let's clear that up...07:22
wolfspraulcountry of origin is not something that makes me loose a lot of sleep07:22
wpwraki think a less fussy power supply would also be beneficial for the intended usage. imagine the vj shows up and finds that he forgot his adapter, that the adapter has the wrong plug (when travelling), etc. then the temptation of just plugging in the next adapter that's at hand would be overwhelmingly difficult to resist.07:22
wolfspraulI agree with you, now what?07:22
wolfspraulwho moves to action07:23
wolfspraul(i'm not saying you should, I'm just saying that is exactly where things stand)07:23
wolfspraulas you know I was very happy to see Joerg giving us feedback, well I didn't get too far with it...07:23
wolfspraulwithout someone taking action, rc3 == rc2 in that area07:23
wolfspraulthe rc2 known issues wiki page lists all we plan to do for rc307:24
wpwrakwolfspraul: hmm, maybe first find out how strongly sebastien feels about the perfectness of the current power circuit design ?07:24
wolfspraulI can speak for him on that, because we discussed it many times.07:24
wolfspraulhe would not argue very long that the current design is super robust, or 'consumer standard'07:24
wolfspraulon the other hand you need to understand the only reason why he was able to get to where m1 is today is laser sharp focus07:24
wpwrakwolfspraul: then ping joerg and see if he's willing to help. connect him with adam maybe. or sebastien on a relaxed day (he seems stressed out lately)07:25
wolfspraulhe single-handedly built the whole thing from ic design to compiler bugs to kernel to vj app to presenting at conferences07:25
wolfspraulthat's a miracle that he made it so far :-)07:25
wolfsprauland along the way he would have had 1000 distractions that would have made his path one of 500 years, not 2 years07:25
wolfspraulin that context you need to see the power circuit design07:25
wolfspraulit works, it passes certification, done07:26
wpwraksure. it's an amazing achievement.07:26
wolfspraulsebastien is not stressed out over this.07:26
wolfspraulhe is focused.07:26
wpwrakno, i mean in general07:26
wolfsprauldidn't notice that, he is traveling a bit at the moment07:26
wolfspraultravel is terribly time consuming...07:27
adamw_wolfspraul, AVDD min. is described on p.3 of http://www.national.com/profile/snip.cgi/openDS=LM4550B07:27
wolfsprauladamw_: I added a little section to http://en.qi-hardware.com/wiki/Milkymist_One_RC2_Known_Issues07:27
wolfspraulat the beginning "joerg's feedback"07:28
wolfspraulthis is just copy/paste from this channel07:28
wolfspraulnot necessarily action items07:28
wolfspraulmaybe you can think about it a bit07:28
wpwrakthat may add pressure. i would also think that the launches may make him nervous. first with the question whether it will succeed, and then also with the things he probably thinks really ought to be done before people use the box.07:28
wolfsprauladamw_: for rc3, we definitely only want to introduce changes that are low risk, where you feel it's safe.07:29
wolfspraulalso by now we have passed a whole test of ce/fcc compliance tests, that adds more pressure to making changes07:30
wolfspraulbut if there are changes in what Joerg suggested that are safe, we can think about it07:30
wpwrakwolfspraul: is redoing the compliance tests an option ?07:31
wolfspraulof course, no problem07:31
wolfspraulI love testing, quality, etc.07:31
wolfspraulif we think our design got better, why should we be afraid of that?07:31
wpwraki was more thinking in terms of money, logistics, and all that07:31
wolfspraulno it's just a little additional cost and delays, sure, like many other things.07:32
adamw_h/w all inputs or/ suggestions are needed to introduce on Milkymist list not just put wiki or irc i think.07:32
wolfspraulcan be organized, no problem.07:32
wolfsprauladamw_: fair enough. Don't let this distract you.07:32
wolfspraulit's just fyi07:32
wolfspraulI just put it there so it's remembered somewhere.07:32
wolfspraulit's under 'joerg's feedback' so you know we have no consensus that we want this stuff.07:32
wolfspraulalso it is not clear to me whether/how sebastien and joerg can agree upon the significance, risk, correctness of those changes.07:33
wolfspraulif they cannot agree, we don't change anything.07:33
wpwraki think fuse and polarity inversion diode should be no-brainers07:33
wolfsprauladamw_: how long was the xilinx meeting?07:34
wolfspraulhow open minded were they to helping you?07:34
adamw_oah..surely put first. then send to list. and 'discuss' on list. and proposed by whom needs to put more clear reasons or explanation. :)07:35
adamw_around 2.5 hr...haha07:35
wolfsprauladamw_: 100% agreed. you know how this works nowadays...07:35
wpwrakthe junk at the output of the regulators, i don't know. according to joerg, they can only make things worse, and can be hidden traps (e.g., when the frequency of power consumption changes ... changes)07:35
wolfspraulwpwrak: no foss chaos can easily confuse adam anymore!07:35
wpwrakwolfspraul: hehe ;-))07:35
wolfsprauladamw_: were they helpful? can we go there for more support?07:36
adamw_i really understood some pins on fpga side: like DONE/INIT_B/PROGRAM_B. :- )07:37
wpwrakPLAN_B ? :)07:38
adamw_and I drawed the MM1 power up / down whole timing power seqeunces on supplies, flash and fpga pins..then we found a very potential status will cause that 'bug'. :)07:38
adamw_yeah....before like PLAN_B or so, I needs to make clear timing sequences, otherwise I think my poor english will let people known what I said. :)07:40
wolfsprauladamw_: well that sounds great!07:40
wolfspraulmore good documentation coming from you, nice :-)07:40
wpwrakadamw_: (likely cause) that sounds very promising !07:40
adamw_tomorrow I'll work on this first...then find a another part to add it..07:40
wolfsprauladamw_: still my question - how friendly were they?07:40
wolfsprauldo you have access to their whole lab?07:41
wolfspraulthey give you an fae to help, or two?07:41
wolfspraulcan we go there again, or one-time only?07:41
adamw_friendly? I entered their office not lab..07:41
wolfspraulno tech work, just discussions?07:42
adamw_surely , I need to go to there again...the guy who is very busy, he has have pm3:30 for other customer, but I made a pointment first...well we are still small...ha:-)07:43
adamw_but surely they are friendly to me.07:43
adamw_no..tech work, discussion enough...ha I brought my works to him..and described to him...then secrets sounds explored.07:44
wolfspraulwpwrak: you mentioned simulation once in conjunction with ben-wpan07:44
wolfspraulthat made me think - how much can simulation help us discover problems of an entire design early?07:45
adamw_the xilinx reference boards surely have the same status.07:45
wolfspraulyeah, that's in line with what Sebastien thinks07:45
adamw_i can explain later...ha. :) I just need to do drawing. and search part to do experiment..07:45
wolfspraulgreat, good luck!07:45
wolfsprauladamw_: wpwrak was asking how many days you take off for cny, and whether you are in taipei at that time07:46
adamw_this bug is very much 'un-usual'...as we already knew.07:46
adamw_what' is 'cny'?07:46
wolfspraulchinese new year07:46
adamw_oah..probably 2/1(or 2) ~ 2/7, send gift to me? :-)07:47
adamw_atusb etc?07:47
wolfspraulwpwrak: [simulation] can we simulate an entire m1? can we play through proposed changes such as the ones from joerg, maybe even including emi effects etc?07:48
wolfspraulif not today, is this realistic at some time in the future?07:48
wolfsprauladamw_: yeah! gifts! :-)07:48
wolfspraulso we just assume you are not in taipei 2/1 to 2/8 or so, should be no problem07:48
adamw_the bug is not related Xilinx itself , it's about whole h/w design. we guessed :-) just need take actions first then see.07:49
adamw_good...send it out maybe on after 2/5 or so. then I must receive it.07:50
wpwrakwolfspraul: (simulation) i use it for small subsystems. more to help the design process, e.g., the dimensioning of analog components, check for parameter stability, and such07:52
wpwrakwolfspraul: (entire mm1) you'd have to simplify and abstract it a lot. i'm not sure how detailed you can get. not the whole fpga for sure, but maybe some aperiodic power consumption patterns might work. (you'd have to measure/model them, though)07:54
wpwrakadamw_: perfect. gives more some more time to tinker :) i'll try the silabs->avr conversion for atusb. but this will take a few days.07:54
wpwraks/gives more/gives me/07:55
wolfspraulroh: morning07:56
adamw_wpwrak, tinker? can't follow you..well, i know you are working for that avr conversion. :)07:57
wolfsprauladamw_: tinker means he wants to work some more08:02
rohmorning wolfspraul08:02
adamw_wolfspraul, ha..08:03
wpwrakadamw_: (tinker) something between work and play ;-)08:05
adamw_wpwrak, hahaha...good hint. sometimes not easy, only by whom possesses enthusiasm ;-)08:08
ArtlavHi all.12:40
ArtlavAnyone familiar with NanoNote problems?12:40
ArtlavSpecifically, how can i fix the cd following symlinks?12:40
Artlav a symlinked to /b/c12:42
ArtlavRegular linux: cd /a -> /a. cd .. -> /. symlink /a/b to .. -> /12:43
ArtlavOpenwrt: cd /a ->/b/c. cd .. -> /b. symlink /a/b to .. ->/b12:43
ArtlavOr does this at least have a googleable name?12:44
ArtlavSo, noone here?12:55
ArtlavCorrection: It's Midnight commander problem, and cd .. is normal behavious, my mistake.13:04
ArtlavSo, "Cd follows links" option in mc is not working.13:11
kyakis it working on your linux host?13:11
ArtlavYes. And everywhere i ever tried before.13:12
kyakdoes it bother you?13:12
ArtlavYes. I have symlink going into the depth of the sd card, and it exits in there instead of the root. Annoying.13:13
ArtlavWell, i guess there is no probblem for now - mc is very defective in default image, and i'm on the way to rebuilding it anyway.13:14
kyakin what way is it defective?13:14
ArtlavNo syntax highlighting, no extfs - archive viewers. Maybe some more.13:15
kyakah, so it's very limited13:16
ArtlavYes, basic or stripped model.13:16
ArtlavI tried adding the missing files with all the definitions, but it won't notice them. I guess things are simply not compiled in.13:17
kyaki'm not using mc myself, but i guess you can enable features of full version and enjoy it13:17
ArtlavSure, i hope.13:18
kyakyeah, it's Makefile is full with "--without" :)13:18
OrnotermesSparkfun have a MicroSD Sniffer that can be used as a break out board: http://www.sparkfun.com/products/941914:03
OrnotermesBut for one that have the stuff and know how needed to etch own boards it's cheaper to make one... or a bunch14:05
rjeffriesin USA I think this price is low. <Ornotermes> Sparkfun have a MicroSD Sniffer that can be used as a break out board: http://www.sparkfun.com/products/941914:20
rjeffriesHaving said that I still plan to build something for 8:10 just as (this is for The Wolf) PROOF OF CONCEPT14:21
rjeffrieslet's hope Wolfgang has a sense of humor. But since I think he may be German, "sense of humor" may not be the first thing that comes to mind.14:22
rjeffriesYes, I have German friends and have visted that greta country several times. Not sure I  ever saw a German smile. ;)14:23
rjeffriesSTEREOTYPE  so don't band me yet pls. well, if I deserve to be banned, that is ok. must accept my medicine14:24
wpwrakphew. today's daytime blackout: 5 hours.14:26
tuxbrain_awayrjeffries: Me and wolfie have put MONEY in this copyleft thing.... I can asure both have TONS of sense of humor :P14:27
wpwrakrjeffries: (never seen a german smile) i guess you don't drink beer then ;-)14:28
rjeffrieswprak maybe that is my problem. I am totally BSing you but there is SCIENTIFIC evidence that those of Germanic heritage (including AUSTRIAN ;) kaugh less smile less and in general are NOT AMERCANs. That's the real probelm. They are not Yanks.14:32
rjeffriess/kaugh/laugh (silly typo)14:34
rjeffrieswpwrak maybe we need to send our Secretary of state to Argintina to look into the disaster of no electricity. This is affecting your productivity and slowing down mission critical engineering of atBen and atUSB.14:36
wpwrakrjeffries: (humourless germanics) an interesting statement considering the immigration pattern of the US, let alone more distant common ancestors ;-)14:37
wpwrakrjeffries: i guess a secretary of state to EDESUR (the company - occasionally - supplying electricity to my area) might be useful14:37
wpwrakrjeffries: of course, they're privatized, so ...14:38
rjeffrieswpwrak I am so totally joking14:47
rjeffriesheh my wife comes from german  Russian stock  she does smile,14:47
rjeffriesI assume it is those Russian genes??14:47
rjeffriesnew topic14:47
rjeffrieshow can we get Carlos back in the club?14:48
rjeffriesI miss him, and really like his SIE (nee SAKC) device a lot14:48
rohhm. good question. i havent really understood his problem15:01
rohs/problem/the logic behind him leaving15:01
rjeffries<roh> human nature. he disliked the fuss Wolf made about those crocheted (knitted?) Nanonote cases15:04
rjeffriesI can sympathise with both sides. but that difference of opinion is (I think) root cause.15:04
wpwrakroh: maybe he realized that he wasn't doing enough to move the SIE forward, had somewhat lost interest anyway, and was looking for an excuse for dropping it ?15:04
rjeffrieswe can and must get past that because Carlos and his students are a great resource15:04
rjeffrieswprak I don't think so. that sounds a little harsh to me.15:05
rohrjeffries: i think one needs to differenciate marketing stuff from the technical side15:05
rjeffries<roh> agree! anyway the other question is how closely affiliated iwth Carlos is the othe project with a camera I forget the meaningless (to me) acronymn15:06
wpwrakrjeffries: (too harsh) maybe. it looked like this pattern of picking a fight only to find an excuse for breaking up to me, though.15:08
wpwrakrjeffries: (xue) i think only very loosely right now. as in "he's helped at some point in time and he knows andres" (last man standing, it seems, although virtually disappeared for about a month, too)15:09
rjeffriesyou may be right but I do not think so. Carlos was really upset about the focus on something outside of hardware/softwar eengineer.15:09
wpwrakrjeffries: (upset) oh, i agree with his sentiment. i also see plenty of things i don't think are important and where i believe energy could be channeled into more productive directions.15:10
rjeffriesCarlos was a BIG help early on. He got Ben schematic into Kicad for example. <wprak? have you ever booted up your 64MB Ben with USB Hots made from Kidac design?15:11
wpwrakrjeffries: alas, no. never got to play with it so far.15:12
rjeffrieswell this is wolf's sandbox. but sometimes people forget that Wolfgang is Sharisim a commercial company15:12
zrafarjeffries: nah.. the crochet thing was the last excuse like anyone would do if he wants to leave for previous reasons.. I think that wpwrak thougts could be the reasons. If you read Carlos last mails15:12
rjeffriesand qi-hardware is supposed to be an open community, one where wolf is a member and perhaps first among equals15:12
zrafarjeffries: he said he was tired to send mails people ignored completely, and that15:13
wpwrakrjeffries: btw, i don't know if that design was really done with kicad. from what i've heard about the role of other projects, it would seem not.15:13
rjeffriesyes that had happend I think15:13
zrafarjeffries: he needed some help on kernel side of software side and nobody replied15:13
rjeffriesso there you have it15:13
rjeffriestaht would frustrate me as well15:13
zrafarjeffries: crochete was something he maybe did not like (because a lot of answers to that crochete mail).. but that is not a real reason to leave15:14
zrafarjeffries: previous facts are the real15:14
rjeffrieswpwrak ar eyou saying that the 64MB qty 10 experimental Ben was NOT done using Kicad15:14
rjeffrieszafra I AGREE you are correct his questions on email fell on deaf ears15:14
rjeffriesanyway I have consumed my allotment of electrons on that topic15:15
rjeffriesI will try Carlos again off list15:15
wpwrakrjeffries: i'm saying that i don't think it was done (schematics and layout) with kicad. but i'm not an authoritative source on this.15:16
zrafarjeffries: on the other side I tried to reply every mail carlos sent asking something on software side (I replied from jlime point of view), but he never gave feedback about if my answers were useful to him. I think that he wanted other people to reply his questions. Now I am angry because I replied him all his software questions and he did not say thanks..15:16
rjeffrieswprak who can tell me (off list if need be)15:16
zrafaI LEAVE!15:16
zrafanah.. I will stay, party just starts15:17
rjeffries<zrafa> I apologize for not knowing who you are (yet) are you Mr JLime? if so, cool beans15:17
rjeffriesor limes I guess15:17
zrafarjeffries: no problem, Mr JLime does not exist. Jlime exists ;) And if you want to know the real answer : Kristoffer Ericsson is the real Mr Jlime15:18
zrafarjeffries: but well, on nanonote side I did the userland yes15:18
wpwrakzrafa: or he posted weird questions that don't have a useful answer. like "recommend an ADC". it's like "what means of transportation should i use ?" (without knowing where you are, where you want to go, and so on)15:18
wpwrakzrafa: no reponding to such things is actually a matter of respect. give him time to correct things himself :)15:19
zrafawpwrak: ;-))))15:19
tuxbrain_awayIMMO I thing just Carlos don't know how to team play, and that anyones here does his best with his limitations15:25
tuxbrain_awaythe SIE batch done was a production nighmare, with a lot of things to fix for the next batch...15:26
tuxbrain_awaybut Carlos don't feel that need, the board was ok for his teaching needs...15:27
tuxbrain_awayso nor wolfgang and I where confortable on producing another bach of v2, I was really interested on finance an V3 batch but this V3 never comes ...15:28
rjeffries<zrafa> is Jlime distro more focused on providing PIM funcyionality15:28
tuxbrain_awayat some point he was mad on me, and wants me out of the project15:29
rjeffries<tuxbrain> comuter h/w and s/w is fairly simple stuff. PEOPLE are hard!15:29
zrafarjeffries: jlime distro is mainly for HP Jornadas and HP Palmtops. For nn we wanted to use our experience on previous years making jlime for hp jornadas (similar ancient devices)15:29
tuxbrain_awayand wolfgang just tell him the real price of a producion of another v2 batch15:29
zrafafor nn15:30
zrafarjeffries: for nn. We did jlime for jornadas with X and gtk1, sdl, etc applications15:30
rjeffriesunderstood tucbrain. sounds reasnable15:30
zrafarjeffries: always lightweight applications (that is why we use ancient gtk1 software and no gtk2 which are slow on embedded)15:30
rjeffriesI have a perfectly working UNUSED for a long time Palm Tungsten C. what a sweet machine, way ahead of its time. now a relic15:31
zrafarjeffries: and we use OpenEmbedded. So the jlime distro for nn is very similar to our distros for HP Jornadas15:31
tuxbrain_awaythis plus the lack of tolerance on the increasing non tech topics in the list (that I find really enriching the project as a whole) has made him decide he is not part of qi15:31
rjeffrieszrafa so Jlime is open embedded rather than OpenWRT. I;ll bet you are another OpenMoko refugee?15:32
zrafarjeffries: if IIRC jlime is older than openmoko15:32
tuxbrain_awaya looooot older :)15:33
rjeffriesOpenMoko also used Openembedded iirr15:33
zrafarjeffries: so no.. jlime was living at the same time than openmoko.. and at that time we are not related with openmoko. And as tuxbrain_away says .. it is a loooot older15:33
rjeffrieshwt was that cool HP palmto that was wide\, clamsheel, faitrky good keybaord? talk about ahead of its time15:34
zrafarjeffries: yes, many use OE :)15:34
rjeffriesI though OE has a bad reputation and was avoided by W+Qi Hardware for the reason of complexity15:35
tuxbrain_awayThanks to jlime and my HP jorndada I was able to finish some university practice on the train, editing and compiling C programs :)15:35
rjeffries(reducing complexity)15:35
tuxbrain_awayOE has his ... let's call it temperament15:36
tuxbrain_awaybut OE and OpenWrt are mostly same philosophy15:36
zrafatuxbrain_away: yeah.. you are right. just one list, many limitations from all of us.. no all the people will be happy like it is15:36
rjeffrieswprak hypothetical question: starting with exiting Ben design, and ignoring case issues completely (in this context) how practiacl would it be to add a USB hub chip?15:37
tuxbrain_awaybut due it has been used for very diferent porpouses it has a loooot more package ready than openwrt mainly focused for routers untin NN comes to play15:38
zrafarjeffries: I have my thoughts about.. OE is widely used because you can build a lot of software like distributions for PC, but for devices with few resources. Openwrt is a distribution for routers, and a lot of effort is being put on make openwrt useful for few specific devices, like nn.  OE is more general. On the other side15:40
zrafarjeffries: and I already said it several times ...15:40
zrafarjeffries: I do not like how all the projects use OE and Openwrt to build rootfs again and again and again.. millons of times15:41
zrafarjeffries: I like both things to build repositories of software.. YOu build it once. If you like to improve you can add packages or improve the development system to support/improve new/current packages15:42
zrafarjeffries: but many projects use Openwrt and OE to build the rootfs (the thing you install for the first time) all the time, improving that all the time, building that all the time. I guess that they have a lot of idle cpu and want to use it a bit :P15:43
zrafarjeffries: I am not an expert neither openwrt nor OE, that is why I do not understand why projects do that15:43
tuxbrain_awayzrafa: totally agree, but you should have a really good release plans , and for example I thing opkg is not so good package manager to be able to manage a "distro" upgrade without breaking things (or opkg it self)15:45
rjeffrieswhat is the IRC command to state an action. I thought it was /action15:45
Jay7/me ?15:45
xMfftuxbrain_away: it can actually work well if kernel related packages and uclibc stuff is set to hold15:45
xMffand if enough ram is available15:46
tuxbrain_awaythat a lot of human energy that I thing is translated to CPU energy, but I agree that no need of reflash on every update should be the way to go15:46
zrafatuxbrain_away: yes, that is true. But upgrade and good release plans could be managed different I think. Developing a distribution building its rootfs all the time does not sound to me useful for upgrades.. I would improve the repository on specific and planned releases and, of course, giving users some way to upgrade properly15:47
Action: rjeffries nods15:48
tuxbrain_awayRAM  It's just me or is quite demanded lately on qi-community15:48
rjeffriesthe time for more RAM has arrived yes it is obvious15:48
zrafatuxbrain_away: no you, it is quite demanded.. people want openoffice and firefox15:48
Jay7we have 64Mb on Zauruses15:49
Jay7it is not enough :)15:49
rjeffriesWolf need not worry he will still sell a boatload of Ben15:49
wpwrakrjeffries: (usb hub) possible, but what for ?15:49
rjeffriesI agree 64MB is the wrong answer15:49
Jay7xorg-server + udev + dbus ... etc15:49
Jay7128Mb is usable minimum now :(15:50
tuxbrain_away+compiz+KDE&Gnome together+wifi+320Gb HD+1080p tv output :P15:51
Jay7but personally I would prefer 512Mb :)15:51
rjeffrieswprak I'll let Orotermes defend that idea15:51
Jay7tuxbrain_away: no, just any modern web-browser + mail client :)15:52
tuxbrainJay7: just joking :P15:52
zrafaI have asked for examples about which applications users would like to run (with more ram).. No many answers15:52
rjeffriesso here's thedeal with RAM the current SOC only talks to static RAM as werner has said it costs ,mor ethan DRAM15:52
zrafaif the screen is 320x240.. which applications we would run with 512MB of ram and 320x240?15:52
Jay7for 320x240 64Mb is enough :)15:53
rjeffriesxrafa I would think more programs running at same tiem, e.g.. conoile and itehr apps15:53
Jay7for 640x480 128Mb is minimum15:53
tuxbrainzrafa obious answer: we need higher resolution for that 512 RAM :P15:53
wpwrak128 MB may already be hard. with SDR SDRAM, the best price point may be around 64 MB. with DDR SDRAM, 128 MB would be better. but the 4720/4740 doesn't support DDR.15:53
rjeffriescan we get VGA resolution on a reasobly priced LCD15:54
Ornotermeswpwrak: one usb port is a good thing, several ports would accept even more hardware extensions15:54
zrafaJay7: yeah.. 64MB and 320x240 is nice.. but still we need use cases15:54
wpwrakrjeffries: no. it's not SRAM vs. DRAM. it's SDR SDRAM vs. DDR SDRAM ;-)15:54
zrafaJay7: tell me some software you can not run on nn now15:54
Jay7zrafa: opie + browser?15:54
Jay7or fbreader15:54
wpwrakrjeffries: as time goes by, the number of qualifiers increases ;-)15:54
zrafacan not you run that?15:54
Jay7with big book opened15:54
Jay7or may be some pdf15:54
Action: rjeffries nods and realizes he was WRONG15:54
wpwrakzrafa: don't worry ... as soon as we have networking, someone will want firefox :)15:55
Jay7zrafa: I see your point now :)15:55
Action: tuxbrain will want firefox15:55
wpwrakOrnotermes: err yes, but where do the connectors go ? :)15:55
rjeffrieswhen we get 6LoWpam this baby can go many news ways, some can generate high unit volumes15:55
Ornotermesmaybe it would be practical to let some onboard hardware use USB15:55
wpwrakOrnotermes: urgh. USB is power-hungry.15:56
rjeffrieswpwrak my question was NOT assuming a board that fits inNN case15:56
tuxbrainwith java and flash and vrml rendering plugin :P15:56
Ornotermeswpwrak: the bus it self or the devices that use usb?15:56
rjeffrieswho has watched that 6LoWpan video? That technology will be VERY widely used15:57
wpwrakOrnotermes: well, kinda both. when the device is active, you have a lot of polling.15:57
tuxbrainwhat video?15:57
zrafatuxbrain: btw, Blizzard was able to build the Java ME environment for nn.. and he is using that on latest jlime.com images.. No idea if you still have future buyers who want that15:57
wpwraktuxbrain: oh, i have no doubts people will ask for plugins just minutes after you give them firefox ;-)15:57
Ornotermesan internal device that use say 3.3V could use the main supply instead of its own 5V -> 3.3V regulator15:57
rjeffriesa small peratble device, van be larger than NN that supports 6LoWpan at attractive price (does NOt have to be $100 USD, maybe $200 USD) will sell like gangbusters15:58
Ornotermesthat would be a little better15:58
rjeffriesWolf would need to fine dbiiger suppliers and factores15:58
wpwrakOrnotermes: that depends ... some usb-powered devices don't have that option. but which internal device do you think should use USB ?15:58
Ornotermeskeyboard maybe, would free up a lot of IO15:59
wpwraktuxbrain: (video) http://portal.sensinode.com/downloads/6LoWPAN-seminar.mp415:59
rohhm. you know that something like a thinkpad x120e is also only 400$? and lots of 10" atom class computers for 270E15:59
tuxbrainrjeffries: I can become as big as market demands :)16:00
wpwraktuxbrain: (roh found it)16:00
wpwrakOrnotermes: (kbd) naw, then you need another controller for that. you might as well use SPI then.16:00
Action: rjeffries thinks Tux must be so skiny now, but soon, a world clss Sumo wrestler16:00
wpwrakOrnotermes: USB makes sense for a bit larger devices. like netbooks. there you have more wiggle room16:01
rjeffries<wpwrak> I ythink formfactor larger than BEN but smaller than netbook that is copyleft and has right feature set could be ahit16:01
wpwrakanyway .. i need to get raw material for some ice cream research ...16:02
rjeffries<wprak> have you conssidered designing in a keyboard controller to free precious GOIOs16:02
rjeffriesI though the ectricity was out werner16:03
wpwrakrjeffries: phew. yet another product line. besides, it has been tried by psion. not a great success. i think the tablets will eat you for breakfast if you try to compete there.16:03
rjeffrieswith no AC your ice cream will melt in a heartbeat16:03
rjeffrieseither there is a market for coplylefty very open of not16:03
wpwrakrjeffries: (kbd) no, i want a cpu that makes more i/os available. problem solved ;-) we're not in real-life communism, where you design your country to run on scarcity :)16:04
rjeffriesyou ar eright marketplace is so fumm of cool new shit16:04
Action: tuxbrain is searching for another so not high class meals this year to survive but will still invest in caviar16:04
zrafawpwrak: (video guy) do you know where is he from ?16:04
wpwrakrjeffries: so far, the fridge still maintained < 0 C through all those messes. that isolation lasts a good while. you'd be surprised.16:04
rjeffriesSweden I think. he wrote the book on 6LoWpan16:05
wpwrakzrafa: no. maybe UK, but i'm not so good at detecting dialects.16:05
Action: rjeffries don't let chatting on irc come between you and your ice cream. you deserve it16:05
wpwrakindeed :) afk for a bit16:06
rjeffriesthank god he is not from Argintina. Becaus if we put him and wpwrak in the same town, watch out16:06
zrafarjeffries: he has a very clear english for me. Sometimes I have to do a big effort to understand.. no with this guy16:07
Traceigif i have ATX PSU16:08
Traceigwith 20+ 4 pin16:08
Traceigdo i plug that 4 pin for CPU16:08
rjeffrieswrong channel Traceig sorry16:10
Traceigbye then :)16:10
rjeffries<zrafa> Agree, he is also a very good presenter IMO16:10
Jay7they are coming on -hardware part seems :)16:12
Action: rjeffries nods16:13
wpwrakokay. i;ll tempt murphy again and switch on the soldering iron ...16:37
tuxbrainI have seen the video.... wpwrak move you ass and finish that atben thing damn it! :P16:46
Action: rjeffries back from lunch17:05
rjeffriesnow w esee tux getting all excited abur 6LoWpan. He is one smart cookie17:06
rjeffriess /w esee/we see/17:06
tuxbrainI see this intersting link about there is also a 6LowPan kernel driver http://www.nanork.org/wiki/6LoWPAN-installation17:28
tuxbrainbut I dont find any further information on this sixlowpan.ko module?17:28
wpwraktuxbrain: (finish) working towards it :) takes about half an hour to solder the chips on atben. 20-30' to place the vias. haven't timed soldering the vias, testing, etc.17:37
wpwraktuxbrain: for atusb, add another 30' for the larger quantity of components. there are also more vias, which probably means a total of 30' more, too.17:37
wpwraktuxbrain: and of course, making the boards takes time, too. the goal is to have about 5-6 of each. 1-2 sets for me, 2 sets for adam, one set for roh, and one for richard.17:39
Action: tuxbrain will wipe wolfgangs as to hurry up to subtitute you with adams toys once this jewel works17:41
wpwrak(nano-rk) looks interesting. one thing to do is to find out where the most "official" ieee 802.15.4 and 6lowpan code really lives. the linux-zigbee project is eerily silent.17:41
wpwrak"wipe wolfgang's ass" ? hmm, you may want to look that up :)17:42
tuxbrainyep, it only mentioned 6lowpan in one page but I don't see any other extra info17:42
wpwrakroh: oh, and as i was strolling through the local diy store, killing time while i had no power anyway, i found non-corrosive silicone. seems that solvent in the usual adhesive/sealant attacks metals.17:44
wpwrakroh: (well, a little. can't be too bad, since they recommend it for metals, too)17:44
wpwrakroh: but probably enough that you don't want it on your pcb :)17:44
wpwraksoldering the vias, 15', cutting them, cleaning, and non-functional testing (connectivity, idle current, clock out), another 15'18:20
wpwrakso that's ... 1.5 hours per atben starting with the pcb ready and when i don't make any mistakes. (and have power ;-)18:21
rohwpwrak: hrhr18:26
rohyeah.. its impressive how much time simple things need18:26
zrafaroh: so those things are not simple at all then :)18:29
wpwrakroh: the large ground zones and all the (RF) vias are a bit of a pain. otherwise, a board of that size should take maybe half an hour to solder and check.18:30
zrafa1.5 hours plus make time to do pcb .. without any mistakes.. that is a lot if it is a simple thing18:30
rohzrafa not neccessary. simple doesnt mean fast18:31
rohon the contrary.. often things get much simpler as soon as one takes 'as fast as possible' out of focus18:31
wpwrakroh: oh yes .. i have an example for *that* ... lemme take a picture :)18:32
zrafaroh: yeah, simple does not mean fast.. but it depends against something to measure if it is simple or not... we could say .. if it takes time then it is no so simple.. or something like that, as reference.18:36
zrafaOf course, maybe time to make it is not the best thing to measure if it is simple18:36
wpwrakroh: http://downloads.qi-hardware.com/people/werner/tmp/rush-job.jpg18:47
wpwrak(and no, i won't try to find out if it works anyway ;-)18:49
steve|m.oO(you need a phillips screwdriver to insert batteries in a kids toy?!) (girltech im-me)19:04
wpwrakthat's the latest (hopefully final) version19:04
wpwraksteve|m: these girls today ...19:04
steve|mlike to.. screw?19:04
rohsteve|m: how much money should i wire you? 20E?19:05
steve|mroh: damn, in which channel did I post the price I calculated? but anyway, it was around 20¬.. is fin19:06
wpwraknow .. let's try a bit of creative ice-cream abuse for a change19:07
rohsteve|m: hm.. osmocom?19:07
rohmeh. egypt just killed the internet for endusers.19:09
roh~15minutes ago.19:09
steve|mroh: recalculated it.. 18,26¬ ;)19:10
rohincluding you shipping it?19:11
rohok. will send you 1919:11
steve|m(32,62 + 3,90) / 219:11
Jay7wpwrak: is atben sdio-wifi module?19:13
Action: Jay7 tries to understand19:13
tuxbrainnice pics dude19:13
tuxbrainwpwrak: atben 8:10 6lowpan module :)19:14
tuxbrainups that was intended to Jay7 , wpwrak already knows :P19:14
Jay7m? :)19:15
Jay7well.. now I should read about 6lowpan :)19:15
tuxbrainI recomend you to watch this video http://portal.sensinode.com/downloads/6LoWPAN-seminar.mp419:16
tuxbrainor I think we have to start calling THE video19:16
steve|mroh: the USB stick that comes with it has a CY7 PSoC (which everybode is desoldering).. let's see if they protected their firmware better than reinerSCT ;)19:16
Jay7I've read wikipedia :)19:17
tuxbrainwell dudes, was a pleasure be online again , but is time to sleep here19:18
tuxbrainbig hugs to you all, yes ron for you too :P19:19
Jay7good idea19:23
Action: Jay7 gone sleep too19:23
kristianpaulwpwrak: saldly i saw the last project from Carlos is a basic-stamtp like board based on a freescale procesor.. even more sad, i saw libre word in that, but how come?? he know milkymist exits.. anyway... :(19:47
xMffoh, I just seen the plplot stuff on the list, thats nice19:59
wpwrakkristianpaul: yea, he really seems to go in a different direction20:06
rohkristianpaul: there are many definitions for libre22:51
rohand fpga are not cheap. you can get basically all soc i know of cheaper than a fpga which could run even a good part of the same22:52
rohas long as the datasheets are free and complete and or one has complete foss drivers for something i think we can also speak of open hw.22:53
rohi dont want to suggest the free soc idea is bad. on the contrary. but one needs to see whats the goal of a specific board/project and what makes one get there in feasible time and money22:56
rohnanonote also has a 'done' soc.23:00
rohmoko had too.. sometimes i wonder if oem-ing something in the price and power range of this: would make any sense. http://www.pollin.de/shop/dt/OTU4ODkyOTk-/Computer_und_Zubehoer/Hardware/Notebooks_Netbooks_Tablets/Tablet_PC_mit_Android_Betriebssystem_ARCHOS_7.html23:01
--- Fri Jan 28 201100:00

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