#qi-hardware IRC log for Saturday, 2013-12-21

wpwrakwhee, now we can have pets :)00:16
nicksydneyhello00:23
wpwrakheya ! show the pictures ! the pictures of the perfect board you made :)00:44
nicksydneynot yet my friend...was away interstate 2 days...today going to pick up the blue paper from the makerspace00:54
wpwrakdrum roll :)02:14
whitequarkoooo, there's a DE remailer I can trust02:55
whitequarkso goldelico won't even need to send me neo900 via russian post02:56
whitequarkalso works with CN, hmm...02:58
wpwrakhow do you know you can trust them ? because you know where their kids go to school ? :)03:01
whitequarkbecause a) they're huge and there's not a crowd of people spewing hate around b) I've used them in past and they consistently delivered03:03
whitequarkthe fees aren't something to scoff at (it starts at $50 and grows with weight), but if I combine several international shipments, it pays off03:04
wpwrakokay, that sounds fairly convincing03:04
whitequarkI believe it becomes profitable at three combined parcels03:05
whitequarkespecially so when intra-country shipment is free, e.g. amazon03:05
whitequarkplus, recently they started offering a service which doesn't involve russian post ever touching the parcel03:06
whitequarkas I understand, they themselves: rent airplane cargo spaces on US->RU flights, plan intra-RU logistics (say Moscow airports are really overloaded near end of the year), deal with customs and then send a courier from their warehouse03:07
whitequarkfun fact: it's cheaper than any other option: first class mail, EMS, etc, etc03:07
wpwraksounds nice.03:09
whitequarkindeed03:09
whitequarkthey even say when specifically they have shipments scheduled03:09
whitequarksay my parcel will likely be composed at 23th, then shipped the same day and arrive in about three more03:10
whitequarkwhich is even more impressive, considering the sheer number they likely have to process right before the new year03:11
wpwrakah, no "sometime between 8 am and 8 pm. just be at home, ready to respond within 10 seconds when the bell rings."03:11
whitequarkno, that's a completely different thing, I'm talking about US->RU shipment03:12
whitequarksince they stop them at 24th and stop customs processing at 27th03:12
whitequarkbut that's also taken care of. the courier will call you the day before and agree on a timeframe03:13
wpwrakah, i see. well, here this is so much easier: all government services are shut down from the 23rd to the 2nd, with some hastily issued decrees. apparently in response to widespread blackouts.03:14
whitequarkno government [services], no problem :p03:17
wpwrakindeed :)03:18
whitequarkI recall a saying: in US, public services are for those who can't afford private. in the rest of the world, they're for public [everyone]03:19
whitequarkbut, I find this more of a spectrum than a definite distinction by country03:19
orthogonal_Hi, I'm thinking of getting an atusb. I realise that the range is only 4-7m. If I were use a Wokfi ( http://www.usbwifi.orconhosting.net.nz/ ), then what would be an estimate for expected range?03:35
orthogonal_It is sort of like a metal dish put around the wifi usb to increase range03:36
whitequarkso a makeshift directional antenna. they claim 12-15 dB gain, so this means 15-30 times the range, in theory03:45
whitequarkbut, do keep in mind that you'd need to orient the antenna rather precisely at the source03:46
whitequarkand if you want bidirectional communication, you'd need two of them pointing at each other. it may quickly become nontrivial03:47
orthogonal_I see. So, 15-30 times the range would be about 1-2km. Ah, would I be able to orient it by just sight?03:57
whitequarkthat's about 60-210 meters, not 1-2 km03:59
orthogonal_Ah, I really should have calculated better :)04:00
whitequark(sight) at ~100m that may be possible. you could try one of those fancy laser pointers.04:00
orthogonal_Okay. Do you know if it is possible to go further? The XBee project seem to have 15 miles: http://www.digi.com/news/pressrelease?prid=45504:01
orthogonal_Not that I'd need that much, though04:02
orthogonal_They seem to use an rf04:02
whitequarkwell, in other applications (I've seen nRF24L01+ dongles) this is fixed by adding a power amplifier in front of the chip04:03
whitequarkI imagine it would be possible to make a custom version of atusb, but as-is it's likely not04:04
orthogonal_Hm... a power amplifier. I'll see if I can look into that04:09
nicksydneywpwrak : gone ?11:02
nicksydneywpwrak...this is the result using the blue paper...11:02
nicksydneyresult ... http://ctrlv.in/273127 ......http://ctrlv.in/27312811:02
nicksydneythe blue paper *after* ironing it to the board.....this is the side that face to the copper board ... http://ctrlv.in/273129 ...while this picture is the side to the iron ... http://ctrlv.in/27313011:03
xiangfunicksydney: Hi11:10
nicksydneyxiangfu: hi11:11
xiangfunicksydney: have you check out this: http://en.qi-hardware.com/wiki/Mini-slx911:11
nicksydneyuuhh thanks xiangfu haven't see that before11:12
nicksydneyxiangfu: did you do all those experiment ?11:14
xiangfunicksydney: yes.11:14
nicksydneyxiangfu: so seems like from your experiment the 'Lucky' gloss paper gave you the best result ?11:15
xiangfunicksydney: you can think the top5 are all same11:16
nicksydneyxiangfu:  noticed that the link to the shop is to taobao...which is hard to buy from :)....do you have any idea whether there is something similar like that on ebay ?11:17
xiangfuyou can try to buy the same 'type'11:19
xiangfunicksydney: where are you? maybe I can just send you some. I still have them in my home.11:21
nicksydneyxiangfu: thanks for the office but it's allright i will try to get it from ebay as the shipping cost will be too much :)11:23
nicksydneyxiangfu: will search on ebay and will get an idea from you to see what you think about the paper11:23
nicksydneyi'm going to try again to use the blue press-n-peel paper again tomorrow and this time will do it bit longer to make sure i apply enough heat...looks like i made mistake of not applying enough heat yah >?11:24
wpwrakaah, power failure while i slept. love 'em.11:25
nicksydney[22:03:01] <nicksydney> wpwrak...this is the result using the blue paper...11:41
nicksydney[22:03:04] <nicksydney> result ... http://ctrlv.in/273127 ......http://ctrlv.in/27312811:41
nicksydney[22:03:07] <nicksydney> the blue paper *after* ironing it to the board.....this is the side that face to the copper board ... http://ctrlv.in/273129 ...while this picture is the side to the iron ... http://ctrlv.in/27313011:41
wpwrakinteresting. it does seem to help a lot.11:50
wpwrakyour main problem now seems to be insufficient adhesion of toner11:50
wpwrakmaybe try more heat (if you can) and more pressure11:51
nicksydneywpwrak: yeah my gut feeling is also telling me i think i have not use the iron long enough perhaps and also perhaps not too much pressure11:53
wpwrakafter heating for a while, you should also pass with the iron's edge over the board. that creates extra pressure11:58
nicksydneywpwrak: true...initially i was bit worried that i put too much heat into it so i took it easy ... another try again tomorrow :)11:59
nicksydneywell at least it looks promising :)12:00
wpwrakdefinitely, yes. and once you figure out how to use the blue paper, you can see if you can get it to work with something that's easier to find as well12:00
nicksydneywpwrak: xiangfu told me about this page  http://en.qi-hardware.com/wiki/Mini-slx9 .... which is very useful12:02
wpwrakah, bgas, the next level :)12:05
wpwraknice page indeed. lots of details.12:06
nicksydneywpwrak: the ps plot from the back side is flipped so when i transferred it to the board it looks the other side ... ... while i can see in kicad the back side not flipped...how do i view it 'flipped' ?12:08
wpwraknot sure if you can12:10
wpwrakmaybe use a mirror ? ;-)12:10
wpwrakyou could use this, though: http://downloads.qi-hardware.com/people/werner/anelok/tmp/anelok-20131005.png12:11
nicksydneywhere did you get this from http://downloads.qi-hardware.com/people/werner/anelok/tmp/anelok-20131005.png  ?12:27
wpwrakgenerate the gerbers and .pos in pcbnew, then run "make overview"12:41
wpwrakah no, not .pos. the drill file. gerbers and drill file.12:42
nicksydneywpwrak: i have a question13:22
Action: wpwrak dims the lights, turns on the spotlights on the candidates. who will answer nick's final question ?13:24
nicksydneywpwrak: saw from schematics that there is a connection VSYS which is coming from AAT1217 and it is connected to pin 5 of MKL25 chip ..saw from datasheet that VOUT33 (labelled in Pin 5) is  "Regulator output voltage" does this mean that this is used to power the USB ? 13:25
Action: nicksydney put his thinking cap on and start getting his pencil and paper to write the answer down :)13:26
wpwrakthat's a little "6" there. the VSYS pin is VIN. VSYS is (more or less) max(battery, min(vusb, 3.3V))13:28
nicksydneyso is VIN the same as Pin1 (which is connected to battery) ?13:32
wpwraki'm talking about the boost converter. and no, VIN is fed by VSYS, which is an output of the MCU13:33
wpwrakhere's a diagram (a bit dated): http://downloads.qi-hardware.com/people/werner/pwsafe/block.pdf13:34
wpwrakUSB power goes to the MCU where it is "mixed" with battery power. the result (Vsys) goes to RF and to the boost converter. the boost converter then supplies OLED and memory card13:36
DocScrutinizer05I wonder if there's clear instruction somewhere in PCB_the_tonertransfer_way about the printout needs to get mirrored so it's correct orientation on the PCB after transfer13:40
nicksydneythe USB power that goes to MCU is USB_VBUS ?13:40
DocScrutinizer05btw same procedure recommended for photopositive method, to get 'mask' aka toner in direct contact to the photoresistive to avoid parallax and diffusion errors reducing width of traces13:42
DocScrutinizer05nicksydney: yes, this sounds like a generally correct assumption, on all designs13:42
DocScrutinizer05USB_VBUS is the commonly used name vor USB power input to device13:43
DocScrutinizer05for*13:43
DocScrutinizer05sometimes also for the 5V supplied by device to the USB, in hostmode13:44
DocScrutinizer05though in the latter case the term USB_VBOOST is aslo common13:44
nicksydneyDocScrutinizer05: cool...understood13:47
nicksydneyso that means USB_VBUS + VBAT = VSYS --> goes to boost converted to power up the OLED ... so that means without USB no OLED will be shown ..correct ?13:47
nicksydneyinteresting pin 6 on the MCU is called VREGIN and this is the one that is connected to USB_VBUS 13:51
nicksydneydatashee - VREGIN == USB regulator input13:51
nicksydneyahhh pg 40 found something13:52
nicksydneyUSB VREG electrical specifications13:52
nicksydneyVREGIN Input supply voltage 2.7  5.5 V13:52
nicksydneyam i reading all this correctly ?13:52
nicksydneywpwrak: question...why is there a mention in the schematic about VSYS connected to 1uF (C11) near Pin 22/23 ?14:14
nicksydneywpwrak: is that what is called a decoupling cap ? ... still learning here :)14:15
DocScrutinizer05rather a buffer I guess14:32
nicksydneyDocScrutinizer05: what is the use of buffer ?14:44
wpwraknicksydney: without USB, the OLED still works. the boost converter just uses battery power then. in fact, if we always had USB power, VSYS would be high enough to always have 3.3 V and we wouldn't need the boost converter14:46
nicksydneywpwrak: was looking through the AT86RF232 datasheet(http://www.atmel.com/Images/doc8321.pdf)  and I'm looking at page 10 where it shows the pinout config for the chip...i notice that for pin 13 & 14 is it pull to ground via cap CB3 and on page 11 it says CB3 is 100nf while in anelok schematic it is using 1uF is this ok ?14:47
nicksydneywpwrak: ahhh i see...so the reason why we have  boost converter because we are using battery and there is not enough 'omph' to power OLED14:50
nicksydneywpwrak: same with pin 29 of AT86RF232 it uses 1uf while datasheet using 100nf and anelok is using 1uF14:54
wpwrakah, interesting. they changes that from at86rf230/231 to 232. didn't even notice it. probably doesn't matter ;)14:55
DocScrutinizer05for sue doesn't matter15:02
DocScrutinizer05sure even15:03
nicksydneywpwrak: can you educate me why using 1uf and 100nf will not make any difference ?15:05
DocScrutinizer05nicksydney: it's basically like a tiny battery helping the regulator to provide smooth power15:17
wpwrakthey're buffers. making a buffer a bit too large doesn't hurt. it's like having a slightly larger tank in your car.15:17
DocScrutinizer05making that tiny battery a tad more beefy doesn't usually cause any harm15:17
DocScrutinizer05wpwrak: however check X5R vs X7R15:18
DocScrutinizer05ESR et al15:18
DocScrutinizer05Zind15:18
nicksydneyDocScrutinizer05/wpwrak: understood now for the cap15:19
nicksydneyDocScrutinizer05: what is X5R and X7R ?15:19
DocScrutinizer051uF actually might have properties not allowed for the CB1 CB315:19
DocScrutinizer05nicksydney: some special type of capacitors15:19
DocScrutinizer05regarding material etc15:20
DocScrutinizer05nicksydney: on page 101 (iirc) you got a pretty "schematic" of the regulator15:21
nicksydneyDocScrutinizer05: you mean page 108 ?15:23
DocScrutinizer05maybe, I already closed the pdf15:24
DocScrutinizer059.1 iirc15:24
nicksydney9.1 is receiver section15:25
nicksydney9.4 says voltage regulators15:25
wpwrakin any case, the transceiver will soon become a different chip anyway. so all that changes.15:28
nicksydneywpwrak: so anelok will not be using aT86 chip ?15:32
wpwraknope, i want to switch to the TI CC2543, which should be BT/BTLE-compatible (at the lowest layers - it'll need some tweaking for the MAC)15:34
wpwrakmore useful than 802.15.415:34
nicksydneywpwrak: ok...cool.....15:38
nicksydneytime to hit the bed15:55
DocScrutinizer05BTLE is a subset of BT, right?16:04
DocScrutinizer05or rather, an extension16:04
DocScrutinizer05extension of specs, regarding what's allowable etc16:05
DocScrutinizer05a subset regarding some of the complexity that is BT, maybe16:05
DocScrutinizer05BLE btw16:08
wpwrakthe relation is a little weird. as i understand it: the PHY is basically equivalent (though not identical as in word-by-word the same). not sure about MAC differences. there must be some in the data and also the hopping simpler. there may be some. the higher layers are apparently radically simplified.16:08
wpwrakofficially it's "BT Smart" :)16:08
DocScrutinizer05wibree BLP BLE BLS OMG!16:16
wpwrak;-)16:19
apeletelarsc_: hi there16:43
larschi16:46
apeletelarsc_: been a few days since last time, how are you doing ?16:50
larscperfectly well16:51
apeletenice16:54
apeletewas thinking about your trial with dma on jz4740, how did it end ?16:54
larscnot well16:56
larscat least not for the ethernet gadget16:56
larscthe problem is that all the ethernet packet are unaligned, while the dma only works if they are aligned16:56
larscso dma is bypassed all the time16:57
apeleteis that an issue we can solve or does it render the dma useless altogether ?16:58
larscat least for the ethernet gadget I would say dma is useless17:01
larscyou can still use it for maybe mass storage17:02
apeletelarsc_: okay, so no need for me to have a look at it then, I guess17:05
apeletebut maybe you can still push your patches ? just curious to read the code to know how it works :)17:05
larscI don't have them anymore17:06
apeleteah shoot :-(17:06
larscbut it was basically just allowing the same irq to be requested for both dma and the usb controller by adding IRQF_SHARED17:06
larscand adding the resource for the dma irq17:06
pcercueiif you're in need of implementing DMA, there's still the MMC driver :p17:07
apeletelarsc_: ok, maybe I'll try to do that just for fun, and ask for your help when I fail :)17:09
apeletepcercuei: MMC driver on the jz4740 ?17:09
pcercueiyes17:09
apeletepcercuei: so we need dma on the mmc driver too ?17:10
pcercueiit'd be nice to have, yes17:11
apeletepcercuei: ok then, I'm on holidays for a week or so, will take a look at dma both in usb and mmc driver17:14
apeletecan't order gcw-zero (not in stock yet) so I need to get my hands on something for the time being17:15
larscadding dma support to the mmc driver shouldn't be to hard with the dmaengine driver in plance17:34
larscplace17:34
larscbut before you start on this give me a ping, there are 1-2 bugs in the dmaengine driver that don't show up when being used for audio, but will show up when used for mmc17:35
apeletelarsc_: ok, will let you know before I start (not tonight anyway, I'm packing. will be traveling to Paris tomorrow)17:43
larscI'll be on the train tomorrow too17:44
apeletelarsc_: going on holidays too ?17:45
larscvisiting my family, not sure if you can call that holidays ;)17:46
apeleteha ha :)17:49
apeletesame here, visit my family and traveling by train. I hope to get some free time to spend on hacking though17:50
larscin Germany on the ICE they have poweroutlets on each seat these days. I just hope the train is not too overcrowded and I get a seat17:51
larscBut I'm getting in on the first stop (and out on the last stop) so I hope I should be able to find something17:52
pcercueiHolydays on ICE? :p17:52
larscyep17:53
apeletenot sure about poweroutlets on the TGV in France, but will be using my newly bought thinkpad X220, perfect to test battery time17:59
apeletegot it a month ago, but still working on my aging thinkpad X60, didn't have time to swith properly despite having debian installed on the X22018:00
wpwrakif you really want to test your battery, teleport to argentina, then join the exodus to the coast. even inside the city traffic has exploded today. there's a sleepy little street next to where i live that gets about one car per minute on average. today it's backed up all the way to the avenue. the highways must be pure fun.18:02
wpwrak(xmas and new year mark the beginning of the summer holidays here. so lots of people head out for the end of december and all/most of january)18:04
apeletesounds like traffic hell18:05
wpwraktotally. kinda like germany vs. italy around easter18:07
wpwrak(if they still do that and don't disperse a little more)18:07
larscnobody drives cars anymore ;)21:01
--- Sun Dec 22 201300:00

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