**** BEGIN LOGGING AT Wed Mar 20 02:59:58 2013 Mar 20 03:48:00 hey jkridner old buddy Mar 20 03:48:07 yo! Mar 20 03:48:45 count you in for mentoring GSoC? thought about any project ideas? Mar 20 03:49:06 how many hours a week would be required? Mar 20 03:50:10 jkridner: wash my car robot (powered by a bbb) a valid idea? Mar 20 03:50:42 sounds valid to me. Mar 20 03:51:09 so mentoring mean code mocking? Mar 20 03:51:59 I managed to get us accepted to be a mentoring organization back in 2010 with about that much scrutiny on the ideas, but we haven't gotten accepted since. :( Mar 20 03:51:59 trolling++ Mar 20 03:52:25 only the students eire should be represented in code. Mar 20 03:52:49 "enlightened trolling" Mar 20 03:53:18 i say we should have a social experiment of me, mru and _av500_ on one intern Mar 20 03:53:21 http://twitpic.com/30j8k5 Mar 20 03:53:44 kinda funny when you google for something and just find your own content Mar 20 03:53:44 jkridner: did mru sign up? Mar 20 03:54:05 he did in the past. I'm counting him in. :) Mar 20 03:54:29 jkridner: i find it funny us dropouts mentoring college students :) Mar 20 03:56:34 * jkridner can't figure out why he ever used Colloquy over LimeChat Mar 20 03:57:02 jkridner: probably the same part of the brain that picked JS over C :) Mar 20 03:57:06 * mranostay ducks Mar 20 03:57:28 don't like colloquy? Mar 20 03:57:30 JS *and* C Mar 20 03:57:51 nope, it has nasty memory leaks and looks like crap compared to LimeChat Mar 20 03:58:25 plus, it doesn't handle proxies. Mar 20 03:58:25 hmm, I run it for months at a time Mar 20 03:58:52 Do you keep logs with it and have you checked the virtual mem of it? Mar 20 03:58:54 not a lot available on mac as far as irc that I've found Mar 20 03:59:13 right, but LimeChat is actually pretty good. Mar 20 03:59:16 nope Mar 20 03:59:28 I'm open to change... just curious Mar 20 03:59:30 don't like it? Mar 20 03:59:42 yeah, took me about a day to like it better. Mar 20 03:59:49 heh Mar 20 03:59:56 won't go back now. Mar 20 04:00:19 I can see traffic in all channels without switching windows. Mar 20 04:00:30 it gives me picture previews when people post URLs Mar 20 04:00:48 one of the things I dislike about colloqy is the changing right click context menu when you click on a person and try to op them,occasionally it turns into kick ban by the time you release Mar 20 04:03:53 how was your SxSW jkridner ? Mar 20 04:05:38 good. possible highlight was meeting Chris Anderson. Or seeing some of the quadcopters. Or visiting with MakerBot and OpenROV folks. Generally a good time with good beer. Mar 20 04:06:03 The Austin Makerspace is pretty interesting. Mar 20 04:09:22 * mranostay ph34rs jkridner Mar 20 04:13:53 jkridner: you never answered the mentoring hours needed Mar 20 04:16:19 mranostay: inversely proportional to amount of "meds" you are taking Mar 20 04:17:40 does alcohol count? Mar 20 04:18:03 jkridner: beagle it is Mar 20 04:18:07 mranostay: difficult to say, but likely 2-6 hours a week. mru and koen have experience and can comment. Mar 20 04:18:40 is it 'bacon' now and not 'scripto'? Mar 20 04:19:00 is it 'bacon' now and not 'scripto' (gerald's name) Mar 20 04:19:12 (sorry for the repeat) Mar 20 04:19:47 do we really need transistors to get enough drive for the LEDs? Mar 20 04:19:58 jkridner: yea Mar 20 04:20:04 :( Mar 20 04:20:12 jkridner: no really worry they are cheap Mar 20 04:20:16 do we really need full brightness? Mar 20 04:20:20 k. Mar 20 04:20:43 jkridner: they are cheap and the "correct" way of doing it, so it is a good example Mar 20 04:20:44 hey so as far as GSoC goes, how about we get an intern to mainline some of our evil vendor tree? Mar 20 04:20:53 :-D Mar 20 04:21:00 why is there anything on P8? I thought we were going to have this stick out the side. Mar 20 04:21:08 jkridner: name it whatever, mranostay and i'll will hide the "true name" on the gerbers and copper fills Mar 20 04:21:24 alan_o: +1 Mar 20 04:21:38 jkridner: don't think we can get to all of the pwm's we need Mar 20 04:21:44 that is a perfect idea---and beneficial to other projects as well Mar 20 04:21:46 jkridner: going to look over in detail tomorrow Mar 20 04:22:01 prpplague: P9 has enough PWMs. Mar 20 04:22:28 jkridner: ok, i'll double check tomorrow and sync with mranostay Mar 20 04:22:30 and I don't think want to have the RTC interrupt on the GPIO0 bank Mar 20 04:23:00 jkridner: we can move that around Mar 20 04:23:06 can I get a thermistor and a button? Mar 20 04:23:13 jkridner: button is on there Mar 20 04:23:24 jkridner: i'd rather not the thermistor Mar 20 04:23:30 yeah, reset button isn't enough? Mar 20 04:23:36 hehe Mar 20 04:23:45 cool on the button Mar 20 04:23:46 Off the top of your heads, how much can an AM3359 sink on a gpio pin? 15mA? Mar 20 04:24:09 prpplague: what about just a place to solder one one? you concerned about the cost of one? Mar 20 04:24:21 Shadyman: rule of thumb is no more than 8mA Mar 20 04:24:28 prpplague: Cool thanks :) Mar 20 04:24:37 jkridner: it's just not that interactive Mar 20 04:24:48 jkridner: i didn't get the memeo Mar 20 04:24:51 *memo Mar 20 04:25:05 you can touch it or blow on it, but I agree the slider is more interactive. Mar 20 04:25:17 I spoke with a professor that got really excited about this board today. Mar 20 04:25:20 * prpplague hands jkridner a creeper card Mar 20 04:25:21 that's what.. .nevermind Mar 20 04:26:13 what he really wants is to be able to throw this cape, a tiny USB cable, a uSD card and a Bone into an Altoids tin for an education kit. Mar 20 04:26:32 * prpplague does a quick check Mar 20 04:26:38 an environmental sensor of some kind is needed. Mar 20 04:27:02 GIGGITY. Mar 20 04:27:27 jkridner: That'd have to be a small bone / large tin Mar 20 04:27:58 looking at even having the IP address scroll across the 7-seg LED with the image provided. Mar 20 04:28:09 jkridner: pretty :D Mar 20 04:28:56 Shadyman: the Bone fits in the Altoids tin with the tiny USB cable and a uSD already (that is something he showed me). Mar 20 04:29:01 jkridner: let me run some numbers Mar 20 04:29:02 :o Mar 20 04:29:07 it really comes down to the size of the Bacon cape Mar 20 04:29:12 mmm bacon Mar 20 04:29:27 It's tough to get a bone + cape in the tin Mar 20 04:29:31 jkridner: let me do some math and sync up with mranostay Mar 20 04:29:41 I don't care about the thermistor being populated for the Design West class so much as to be able to kill the bird of creating the education kit. Mar 20 04:29:57 I think you might could do it if you had almost no standoff. I haven't checked that use case Mar 20 04:30:03 alan_o: perhaps surface-mount pinheaders on the bottom of the cape? Mar 20 04:30:28 solder them to the fiducials Mar 20 04:30:47 could be interesting to stick the components on the same side as the pins that stick out. Mar 20 04:30:59 would save some overall depth. Mar 20 04:31:10 and LCD on the opposite side (outward facing) Mar 20 04:31:10 surface mount wouldn't save a lot of depth. Mar 20 04:31:21 no LCD, only 7-seg LED Mar 20 04:31:25 ah Mar 20 04:31:30 well, still applies :) Mar 20 04:31:40 7-seg LED is kinda tall Mar 20 04:31:54 would want it on the other side as well and just work with the board upside down. Mar 20 04:31:56 hmm. how about a 3 digit 7 seg lcd Mar 20 04:31:59 board=bone Mar 20 04:32:03 $$$ Mar 20 04:32:13 looking at something $cheap$ Mar 20 04:32:19 dirt cheap Mar 20 04:32:33 http://www.bgmicro.com/54digitlcd40pindisplay.aspx Mar 20 04:32:35 ShadowJK: jkridner with beer and thai food cheap Mar 20 04:32:42 ahh Mar 20 04:32:58 Shadyman: yea, i researched that one ages ago Mar 20 04:32:58 how many are you looking at making? Mar 20 04:33:13 Shadyman: hehe, trust me we got it Mar 20 04:33:17 ah Mar 20 04:33:22 kay :) Mar 20 04:33:34 Shadyman: we'll be doing about 2500 first run Mar 20 04:33:45 you got my buy-all-the-cheap-things urge going ;) Mar 20 04:33:56 That's a lot of mints. Mar 20 04:34:06 Shadyman: i did an early version of this for the beagle/panda Mar 20 04:34:11 neat Mar 20 04:34:16 color me intrigued on $0.52 LCD Mar 20 04:34:34 http://elinux.org/BeaconBoard Mar 20 04:34:41 * jkridner trusts prpplague to figure it out. Mar 20 04:34:43 jkridner: You'd need more than just 50 of them though Mar 20 04:34:47 jkridner: $5 worth of parts to make it work Mar 20 04:35:08 why use that horrible WS2801 Mar 20 04:35:08 Shadyman: at the time they have about 4000 of those lcd panels Mar 20 04:35:10 urgh. Mar 20 04:35:17 On a totally unrelated note... http://www.bgmicro.com/black-and-white-lcd-screen.aspx Mar 20 04:35:21 mrpacket_: because it was cheap and did what we needed Mar 20 04:35:37 yes they are cheap Mar 20 04:35:42 eink from a nook e-reader :D Mar 20 04:35:57 the problem with them, is 90% of them are clones built in a nasty factory Mar 20 04:36:02 the 10% of real ones are ok Mar 20 04:36:16 mrpacket_: Well, yeah. These are pulls though Mar 20 04:36:19 mrpacket_: hehe, well i purchased 5k with no issues Mar 20 04:36:32 I have used abotu 2M of them Mar 20 04:36:32 Oh, lcds, nevermind Mar 20 04:36:41 but we dont' any more Mar 20 04:37:18 is that a RGB Led on that board? Mar 20 04:37:24 yea Mar 20 04:37:40 I hope you are paying your royalities to Philips then Mar 20 04:37:43 The PWM tax Mar 20 04:37:49 lol Mar 20 04:37:50 "Philips wants Money" Mar 20 04:37:53 hehe Mar 20 04:38:01 mrpacket_: ugh, that makes me sick Mar 20 04:38:05 mrpacket_: i gave 1000 of them away Mar 20 04:38:10 thats the real reason we dont' use the 2801 anylonger Mar 20 04:38:25 mrpacket_: so i had to keep the costs low, hehe Mar 20 04:38:40 using a more expensive IC, saves me money Mar 20 04:38:42 prpplague: too bad that LED isn't on a USB. You could have made a fortune on kickstarter :) Mar 20 04:38:59 alan_o: hehe, never fear av500 is here Mar 20 04:39:19 http://www.youtube.com/watch?v=OHhKzCKFINQ Mar 20 04:39:30 i use 16 bit control now Mar 20 04:39:38 for what we are doing thats actually pretty useful Mar 20 04:39:44 so we can get a much better perforamnce Mar 20 04:39:51 alan_o: https://gitorious.org/av500/bacon-lunch Mar 20 04:40:30 prpplague: The whole time I was waiting for the number to change... Mar 20 04:40:32 mrpacket_: hehe, yea at the price i was willing to pay, performance wasn't an issue Mar 20 04:40:44 alan_o: hehe Mar 20 04:40:54 What's the deal on the 430 compiler? Mar 20 04:41:01 gcc, proprietary, oss? Mar 20 04:41:10 hehe got me Mar 20 04:41:16 jkridner: anyway Mar 20 04:41:22 jkridner: let me run some numbers Mar 20 04:41:29 jkridner: i'll sync with mranostay tomorrow Mar 20 04:41:39 jkridner: okie dokie crokie? Mar 20 04:41:41 rsync. lol Mar 20 04:41:58 prpplague: well he'll have to deal with community :P Mar 20 04:41:58 k. thanks for taking my extra inputs. :) Mar 20 04:42:01 prpplague: https://gitorious.org/av500/bacon-lunch/blobs/master/Makefile CC = msp430-gcc Mar 20 04:42:37 jkridner: we aren't really, we just are being nice Mar 20 04:42:53 LOL Mar 20 04:43:30 i'm totally not as nice as i seem :) Mar 20 04:43:47 On a different note, is there a pin on the beaglebone expansion header that is better suited than others to take a timepulse input from a GPS? Or are they all about the same as far as interrupts go? Mar 20 04:44:55 and on another note, i think i have a title for my ELC-E presentation: "1.21 jigabit ethernet: a case study of a perfect failure" Mar 20 04:45:20 prpplague: nobody learns anything from success :) Mar 20 04:45:26 haha Mar 20 04:46:51 prpplague: jkridner will owe us a case of beer and a bottle of rum :P Mar 20 04:47:52 any good tools for chasing down memory leaks in the kernel? Mar 20 04:47:54 *sigh* Mar 20 04:48:05 valgrind /proc/kcore Mar 20 04:49:41 alan_o: cat /dev/zero > /dev/mem should work Mar 20 04:56:09 Well, Head office has passed down a big ask Mar 20 04:56:36 *please* design us a platform that is like a beagle, but not a beagle Mar 20 04:56:58 Shadyman: I wonder if the eCAP pins might be useful to look at... Mar 20 04:57:00 unhappy that i have used a beagle on the back of our dmx player Mar 20 04:57:03 or something connected to the PRU... Mar 20 04:57:04 :-) Mar 20 04:57:08 if you want to avoid bugging the main CPU. Mar 20 04:57:49 how many hours and $$$ to design a new dog Mar 20 04:58:08 and can i hire some new engineers. Mar 20 04:58:27 jkridner: Ideally, I'd like to be able to sync the system clock/rtc with the timepulse (1/sec) Mar 20 04:58:52 shady, you doing some GPS based time keeping? Mar 20 04:58:54 mrpacket_: ??? there are other Beagle-like boards. Mar 20 04:59:03 yes, i know. Mar 20 04:59:11 i will dance their dance for a few days Mar 20 04:59:17 mrpacket_: Yep Mar 20 04:59:28 and then they will forget Mar 20 04:59:30 :-) Mar 20 04:59:45 at ~1/s, I'd just look at having it in the wake-up domain so I could put the processor to sleep. Mar 20 04:59:48 mrpacket_: didn't you buy a bunch of white-label boards from cco? Mar 20 05:00:02 no, i ended up buying them from Mouser Mar 20 05:00:07 because nobody at CCO would reply Mar 20 05:00:27 did you pin koen and now prpplague? Mar 20 05:00:38 s/pin/ping/ Mar 20 05:00:46 i talked to the boss guy Mar 20 05:00:53 jkridner: you got it right the first time :) Mar 20 05:01:00 had a meeting with him on the phone Mar 20 05:01:01 k. Mar 20 05:01:08 promised me a bunch fo thigns Mar 20 05:01:13 but nothing got delivered Mar 20 05:01:19 clint? Mar 20 05:01:50 i just figured we are too small a fish for CCO Mar 20 05:02:15 mrpacket_: how many did yo uwant? Mar 20 05:02:24 2500 Mar 20 05:02:31 that's a lot dude Mar 20 05:03:18 seriously, that's above the threshold where you're supposed to be going to them directly (and not mouser) Mar 20 05:03:25 mouser isn't supposed to even sell you that many Mar 20 05:03:26 i think that CCO is just really really busy Mar 20 05:03:32 and they are all overworked Mar 20 05:03:42 so its just fallen through the cracks Mar 20 05:03:51 victims of their own sucess Mar 20 05:03:58 jkridner: Actually, I could set the timepulse up to 1kHz Mar 20 05:04:21 0.25Hz to 1kHz Mar 20 05:05:02 i know how it happens sometimes Mar 20 05:05:08 becuase its happened to us Mar 20 05:05:13 i hate that Mar 20 05:05:23 Though really, I'm not sure if the AM3359 is made to have its clock controlled externally like that. Perhaps I'd be better just putting it on a GPIO pin and sync the RTC to the GPS every so often Mar 20 05:05:58 how much accuracy do you need? Mar 20 05:06:53 Shadyman: info on what eCAP is: http://pastebin.com/XUncDtxV Mar 20 05:07:10 I'm essentially just making a GPS shield for the lulz Mar 20 05:07:38 and as a stepping stone to my GPS/GSM combo Mar 20 05:07:54 maybe once the black is done Mar 20 05:08:06 things will be a bit easier for CCO Mar 20 05:09:05 jkridner: Thanks. So basically, it's a general-purpose high-frequency input Mar 20 05:09:48 I don't know about how "general-purpose" it is, but it is made for counting and comparing counts and helping to align phases. Mar 20 05:09:58 hmm Mar 20 05:10:44 if what you are doing is measuring a pulse-train, it'll let you do that without loading the CPU. Mar 20 05:10:57 interesting Mar 20 05:11:13 woot woot all aboard! Mar 20 05:11:17 ^ Mar 20 05:14:26 hmm, looks liek the bone already has an RTC, so I can't see trying to forcefeed it ticks. however, in the grand scheme of things, bringing it out to the eCAP pin would be useful for more people than just bringing it out to any old GPIO Mar 20 05:15:02 I think for my purposes, I'd just sync the system clock every 5 minutes or so with GPS time Mar 20 05:15:24 Shadyman: can ntp server help you here? Mar 20 05:15:38 they write about how evil it is to just sync time hard. Mar 20 05:15:44 alan_o: the time delay in NTP servers is massive compared to GPS Mar 20 05:15:56 no no Mar 20 05:16:12 hrm? Mar 20 05:16:16 I mean can ntp handle reading your GPS and syncing the system clock (gradually, as it does) Mar 20 05:16:24 hmm. Mar 20 05:16:25 sorry, ntpd Mar 20 05:16:34 perhaps. Mar 20 05:16:38 alan, yes it can Mar 20 05:16:48 you end up having a stratum-1 Mar 20 05:16:59 set up a socket on localhost that speaks ntp, and give the time over that Mar 20 05:17:15 Stratum 1, look at you :) most accurate time on your block Mar 20 05:17:18 http://www.rjsystems.nl/en/2100-ntpd-garmin-gps-18-lvc-gpsd.php Mar 20 05:17:21 lol alan_o Mar 20 05:17:40 strap your bone to your wrist and you're good to go Mar 20 05:17:48 Giggity. Mar 20 05:18:12 alan_o: Challenge accepted Mar 20 05:18:20 mrpacket_: bookmarked. thanks :) Mar 20 05:18:35 that was the irst one that popped up on that oogle thing Mar 20 05:18:45 i did this in 2001 Mar 20 05:19:02 its been a fucntion for a long time Mar 20 05:19:22 i got tired of log files that were not accurate Mar 20 05:22:09 i have my two best $/buck bits of test gear on my desk at home today Mar 20 05:27:03 :) Mar 20 05:27:17 Got my BBone ProtoCape PCBs a week or two ago Mar 20 05:27:30 self-designed Mar 20 05:28:12 They fit, so I guess I should publish the gerbers and such Mar 20 05:28:28 fitting is optional Mar 20 05:28:37 heh Mar 20 05:28:41 a big hammer is always an alternative Mar 20 05:28:49 STOP! Hammer time. Mar 20 05:37:58 * mranostay gets out his mallet Mar 20 05:45:53 :o if only I had a pinout, I would TOTALLY make capes with them. http://www.goldmine-elec-products.com/prodinfo.asp?number=G19231 Mar 20 05:54:47 Shadyman: no pinout :/ Mar 20 05:57:28 yeah :( Mar 20 05:58:19 shouldn't be too hard to find out though? Mar 20 06:00:56 you'd think Mar 20 06:04:59 they can be pretty tricksy sometimes Mar 20 06:06:52 find someone with a RNS510C and take an scope to the wires :) Mar 20 06:07:14 or a broken RNS510C where the display still shows something Mar 20 06:11:58 #@$@#$ing slide resistor with no eagle part Mar 20 06:12:10 * mranostay curses jkrider Mar 20 06:18:34 gm Mar 20 06:20:35 #@#@$ av500 Mar 20 06:20:51 i mean hi av500 Mar 20 06:22:16 yo woglinde Mar 20 06:22:22 jo Mar 20 06:22:23 welcome to the kegger Mar 20 06:22:34 trool cellar? Mar 20 06:22:47 * mranostay is in a really good mood Mar 20 06:23:01 * mranostay tosses woglinde a virutal Modelo Mar 20 06:24:01 hm hm Mar 20 06:24:06 * mranostay even hands zengweitotty a beer Mar 20 06:25:46 hi,***mranostay. Mar 20 06:26:15 the channel has no logs, what's up Mar 20 06:26:40 why must a channel always have a log Mar 20 06:27:44 mranostay hm is it not time to sleep for you? Mar 20 06:28:00 trolls don't keep normal hours Mar 20 06:28:47 hi, my lcd shows white light,can not show any content. Mar 20 06:28:59 woglinde: its only 10:30 there Mar 20 06:29:14 what maybe the problem? Mar 20 06:29:33 maybe you did something wrong Mar 20 06:29:34 it is full of stars! Mar 20 06:29:40 (just guessing) Mar 20 06:29:52 I guess its broken hw Mar 20 06:30:01 not the hw Mar 20 06:30:02 mranostay: what is that cape that prpplague and jkrindner are all excited about? Mar 20 06:30:30 can it be somethings wrong with omapfb? Mar 20 06:30:43 av500: bacon cape Mar 20 06:30:54 url? Mar 20 06:31:25 av500: in the works Mar 20 06:31:45 @#$@#$ slider resistor has no eagle part Mar 20 06:31:57 <--- doesn't feel like making oen Mar 20 06:31:58 *one Mar 20 06:32:07 doit! Mar 20 06:32:24 if i had infinite time Mar 20 06:33:32 omapfb omapfb: cannot parse default modes Mar 20 06:34:12 what does this means? Mar 20 06:34:19 are you trying to start a nuclear war! Mar 20 06:34:28 turn that system off now! Mar 20 06:36:11 oh noes we are at Defcon 5! Mar 20 06:36:20 oh, prpplague dragged my horrible msp430 code into the light.... Mar 20 06:36:23 or is that one Mar 20 06:38:09 av500: i can beat it over the head with a shovel for you.. the code or prpplague either or Mar 20 06:39:22 and even you till you forget Mar 20 06:41:36 av500 url? Mar 20 06:42:08 I must admit, I'm impressed. uBlox's GPS modules have kept backwards/fowards compatibility fairly well. Mar 20 06:42:26 ie the board I just designed, I could use a NEO 4, 5, 6, or 7 just by not placing certain components Mar 20 06:42:28 Shadyman why not? Mar 20 06:42:31 its gps Mar 20 06:42:51 nothing bitbang fancy Mar 20 06:43:00 True. Though, the neo series isn't very feature-rich. it's the lea series that's fancy pantsy Mar 20 06:43:20 blah... 6.5" Mar 20 07:22:52 Tadah. http://i37.photobucket.com/albums/e94/MasterMWL/GPS_Cape_zps67906d8b.png Mar 20 07:25:13 creative via'ing Mar 20 07:25:37 heh Mar 20 07:25:47 I hope you dont pay per hole Mar 20 07:25:57 nah Mar 20 07:26:15 friend of mine use closely spaced vias to not have to cut the PCB lots :) Mar 20 07:26:25 just break along the vias Mar 20 07:26:38 a number of boardhouses don't like that ;) Mar 20 07:26:47 I can guess :) Mar 20 07:26:49 The viaing was actually suggested by the manufacturer of the module Mar 20 07:26:56 because of RF Mar 20 07:27:13 to prevent islanding, etc Mar 20 07:27:27 well, look at how gps is done in your average smartphone Mar 20 07:27:40 heh Mar 20 07:27:47 average smartphoen doesn't use active antenna ;) Mar 20 07:28:02 right Mar 20 07:28:26 most of them use GPS to confirm thier position as given by the assistance server.... Mar 20 07:28:42 yeah Mar 20 07:28:58 hence my next board, the GPS/GSM combo ;) Mar 20 07:29:29 the ublox gsm and gps module work together quite well in that respect Mar 20 07:30:42 I'll probably have to shrink down the battery, and run the GPS antenna off of where the battery is now, to help reduce noise. Mar 20 07:30:52 er, gsm. Mar 20 07:32:55 which GPS module are you using? Mar 20 07:33:08 looks nice Mar 20 07:33:09 ublocks Mar 20 07:33:10 shady man..whats your pcb package Mar 20 07:33:17 blah Mar 20 07:33:20 its kind of plesant Mar 20 07:33:29 pheasant Mar 20 07:33:31 why ublock? Mar 20 07:33:40 maybe it has NAND? Mar 20 07:33:49 does this have a RTC on it Mar 20 07:34:11 those are $$$ Mar 20 07:34:23 I wonder if that via flood is really necessary :) Mar 20 07:34:56 or in deed helfpul Mar 20 07:35:14 and why are they all thermal pads? Mar 20 07:35:28 it is, otherwise there would be nothing to talk about. and it looks funny. Mar 20 07:36:03 yes, now i'm confused. Mar 20 07:36:16 why would you put all those vias with thermal pads Mar 20 07:36:39 maybe they are hand soldered? Mar 20 07:37:07 mmm, maybe.. Mar 20 07:37:09 :-) Mar 20 07:37:24 harsh crowd here Mar 20 07:37:37 Shadyman: do you have a rule that adds a railing of vias around every signal line= Mar 20 07:37:38 ? Mar 20 07:38:02 and I wonder if it passes the DRC checks... Mar 20 07:39:32 Shadyman: what kind of antenna is that B1? Mar 20 07:40:29 ynezz: that is a battery-holder Mar 20 07:40:43 ah! Mar 20 07:40:50 mrpacket_: RTC, yes. ynezz: Via flood, apparently, according to uBlox. mrpacket_: Because I haven't figured out how to make them straight grounded vias. av500: No rule. hand-placed. ynezz: Passes DRC. Mar 20 07:41:28 hm, but your silk is wierd :) Mar 20 07:41:39 Shadyman: why ublox? Mar 20 07:41:40 how so Mar 20 07:41:55 broken by vias Mar 20 07:42:11 ds2: The modules can be had for ~$10/ea, and they have interesting synergies with the uBlox GSM modules Mar 20 07:42:49 on U3 you've it over pads Mar 20 07:43:30 Shadyman: so it is the integration? (other modules can be had cheaper) Mar 20 07:43:50 ds2: Yeah. Featureset and integration Mar 20 07:43:51 there is always a cheaper solution Mar 20 07:43:55 ^ Mar 20 07:43:59 some are expensive to use :) Mar 20 07:44:04 ynezz: I don't see it? Mar 20 07:44:17 ynezz: There are some pads that are ground Mar 20 07:44:30 you don't see the white line going over the pads? Mar 20 07:44:41 oh, white Mar 20 07:44:43 all over them Mar 20 07:44:50 boardhouses won't print that Mar 20 07:45:07 but I guess I could fix that footprint up ;) Mar 20 07:45:10 well, I've just done 1st PCB 2 weeks ago :) but as I remember this won't pass my DRC checks Mar 20 07:45:27 Interesting. Mar 20 07:45:39 and the silk will really look weird Mar 20 07:45:42 well, I should say *most* boardhouses won't print that Mar 20 07:45:53 unreadable probably Mar 20 07:45:56 just my guess Mar 20 07:46:09 but it's v1.0 :) Mar 20 07:46:14 hehe Mar 20 07:46:32 well, they usually use some kind of inkjet printer, and just subtract the open pads from the silk layer Mar 20 07:46:55 well, $10 seems like a good price Mar 20 07:47:13 ds2: Also, [citation needed] for cheaper modules Mar 20 07:47:19 im about to buy a new pcb lab line Mar 20 07:47:50 mrpacket_: Oh? Mar 20 07:49:58 hm, which ublox is that exactly? Mar 20 07:50:30 and this is Kicad? Mar 20 07:50:34 http://www.bungard.de/index.php?option=com_content&view=article&id=109&Itemid=168&lang=english Mar 20 07:50:42 someting like this. Mar 20 07:51:13 won't fit in my flat :) Mar 20 07:51:32 ynezz: gEDA, and ublox neo-6m Mar 20 07:51:54 mrpacket_: me gusta Mar 20 07:56:02 Shadyman: hm, where one could buy them for $10? :) Mar 20 07:56:08 * Shadyman hums. Mar 20 07:56:12 1 sec Mar 20 07:56:52 ynezz: http://www.aliexpress.com/item/2PCS-LOT-Free-Shipping-UBLOX-modules-NEO-6M-GPS-positioning-module-NEO-6M-0-000-NEO/727256461.html Mar 20 07:58:44 ynezz: That's for two, with free shipping Mar 20 07:58:54 or http://www.aliexpress.com/item/NEO-6M-0-001-UBLOX-modules-NEO-6M-GPS-positioning-module/727248572.html for one with $5 shipping Mar 20 07:58:55 not fake? Mar 20 07:59:14 I've yet to test mine Mar 20 07:59:32 but, it came in a cut reel Mar 20 08:00:23 since it Mar 20 08:00:26 oops Mar 20 08:00:44 since it's quite a lot cheaper then on their store Mar 20 08:00:55 the module there costs like 85EUR Mar 20 08:01:07 yeah, their store is BS Mar 20 08:01:29 :) Mar 20 08:01:29 I think once you sign up for a credit account or somethign with them, then you get the real price Mar 20 08:01:40 I think that's more of a "We just don't want to be bothered with individuals" price. Mar 20 08:01:48 ah Mar 20 08:02:12 strange marekting strategy :) Mar 20 08:02:15 heh Mar 20 08:02:35 also, wtf DRC, y u give me strange errors Mar 20 08:04:04 too many vias maybe? :) Mar 20 08:06:17 haha Mar 20 08:06:30 nah, usually it drcs things under the ground pour, for whatever reason Mar 20 08:06:41 like the lines I use to ground the ground pads Mar 20 08:09:14 Oh Mar 20 08:09:28 just learned that the DRC centers on the middle of the problematic trace regardless of where the problem is Mar 20 08:09:28 derp Mar 20 08:18:21 I don't see that via galore here :) http://www.sainsmart.com/zen/albums/SKU/20/20-011/20-011-929/nichegeek/01.JPG Mar 20 08:23:44 hmm? Mar 20 08:23:54 hmm. Mar 20 08:24:08 Well, it's a fairly small board Mar 20 08:24:18 and not using a powered antenna, it looks like Mar 20 08:25:08 they should certainly have vias on either side of the antenna trace Mar 20 08:25:47 sure, but not 200 :) Mar 20 08:26:02 they're not following the manufactuer's recommendations Mar 20 08:26:16 The eval board for the GPS and GPS modules has over 700 vias. Mar 20 08:26:19 I think ublox has stocks in a via trader Mar 20 08:26:20 :/ Mar 20 08:26:20 so this is based on reference design? Mar 20 08:26:26 Mine is, yes. Mar 20 08:26:32 hm Mar 20 08:26:49 probably the same reference that sainsmart used Mar 20 08:27:07 or, the schematic at least. They obviously didn't read the via part, or aren't using powered antennas Mar 20 08:27:19 the clip-on ones like they use are rarely powered Mar 20 08:27:30 SMA ones, though, usually are Mar 20 08:27:59 there's a "Hardware Integration Manual" for the NEO 6 module that goes over all the best practices and such Mar 20 08:43:44 aha, figured out how to unthermal the vias and still have them connected to the plane Mar 20 09:02:34 Shadyman: http://openbsc.osmocom.org/trac/raw-attachment/wiki/osmo-lea6t-gps/osmo-lea6t_small.jpg Mar 20 09:03:00 not much vias either :) Mar 20 09:03:23 heh Mar 20 09:03:30 ynezz: they are all burried! Mar 20 09:03:48 keep the outer layers clean Mar 20 09:04:13 the line to the antenna has them Mar 20 09:04:31 sure Mar 20 09:05:37 av500: it's only 2 layer board... Mar 20 09:05:49 http://i37.photobucket.com/albums/e94/MasterMWL/GPS_Cape_zps3247ebb5.png Mar 20 09:06:23 and I've it open in eagle now, there's not much vias :) Mar 20 09:07:14 anyway, I find the vias are giving the board the special look. and they make the board lighter Mar 20 09:07:53 ynezz: http://www.u-blox.com/images/downloads/Product_Docs/LEA-6_NEO-6_MAX-6_HardwareIntegrationManual_%28GPS.G6-HW-09007%29.pdf page 43 Mar 20 09:09:12 that's it, thanks! Mar 20 09:09:38 aholler: I would write my name in vias Mar 20 09:09:50 in 5x7 pixel font Mar 20 09:09:54 yes Mar 20 09:09:54 haha Mar 20 09:10:08 I made a smiley face. Mar 20 09:12:25 hm, nice integration manual Mar 20 09:13:08 ynezz: http://www.u-blox.com/en/gps-modules/pvt-modules/neo-6-family.html Mar 20 09:13:22 Protocol Spec goes over all AT commands, Data sheet goes over the module itself Mar 20 09:14:13 Though perhaps they only wanted a rectangular area full of vias, isolated from the regular ground Mar 20 09:14:33 but then you have that ground operating at a different potential Mar 20 09:14:41 bad, rf, bad. Mar 20 09:16:33 ust don't do it like asus and put it inside some solid tinfoil case ;) Mar 20 09:19:01 wat Mar 20 09:19:59 why would they do that? :x Mar 20 09:20:13 look better Mar 20 09:20:15 +s Mar 20 09:20:20 apple tried the same Mar 20 09:21:10 they aren't bend to physic laws ;) Mar 20 09:21:48 av500: You're holding it wrong. Mar 20 09:22:30 but yes, the new one looks prettier now that I figured out how to make straight (no-clearance) vias http://s37.photobucket.com/user/MasterMWL/media/GPS_Cape_zps3247ebb5.png.html Mar 20 09:22:32 uups, s/bend/bound/ Mar 20 09:24:20 907 vias :/ Mar 20 09:24:49 do board makers charge by holes? Mar 20 09:24:58 not typically Mar 20 09:25:33 so they will start doing that ;) Mar 20 09:25:42 heh Mar 20 09:28:43 the antenna line does not need to have vias, what it needs to be is to be an impedance controlled line that is 50R Mar 20 09:29:23 KotH: I knew you would come here and spoil it Mar 20 09:29:29 sorry dude Mar 20 09:29:34 it's my EE nature ^^' Mar 20 09:29:46 i wonder if all those vias will make sense if below is a processor spraying stuff around with 1ghz Mar 20 09:30:21 the hardware integration manual explicitly says that you should not put anything below the RF part of the module Mar 20 09:30:41 depends on how you define below Mar 20 09:30:55 on the other side of the PCB Mar 20 09:31:22 so just put the bb above the cape? Mar 20 09:31:33 sure Mar 20 09:32:04 so it's an underground cape Mar 20 09:32:17 the u-cape Mar 20 09:32:21 :-) Mar 20 09:33:24 uhmm.. what ever is near the antenna trace there... i think it should not be there Mar 20 09:35:09 and R14 is probably too small Mar 20 09:35:19 (power rating) Mar 20 09:36:15 power supplies need additional blocking C's.. a lot of them Mar 20 09:36:27 a BLM15/BLM18 wouldnt be wrong either Mar 20 09:37:52 R14 just ties the reset pin high Mar 20 09:38:14 hmm? isnt that the external power for the antenna pin? Mar 20 09:38:28 * KotH does not have the datasheet open Mar 20 09:38:58 er, should be? Mar 20 09:39:05 * Shadyman eyes R14 Mar 20 09:40:07 but r14 is 3.3kohm to enable the !RESET pin and hold it high Mar 20 09:40:15 also: you might want to consider placing a low capacitance RF schottky on the antenna line, something like the HSM2852 to clamp any surge on the antenna input, probably parallel with an zenner diode to eat up the energy before it goes into the power lines Mar 20 09:40:16 as to why it's not going to the rest of the stuff... Mar 20 09:40:42 ah.. ok Mar 20 09:40:52 does the neo have a antenna power? Mar 20 09:41:02 yes Mar 20 09:41:10 ok... Mar 20 09:41:25 well, not as it is at the moment. schematic didn't connect the dots. Mar 20 09:41:45 PEBCAK Mar 20 09:41:57 put there an 1206 resistor Mar 20 09:42:11 anything smallwer will not have the power rating Mar 20 09:42:22 or prolly even a minimelf Mar 20 09:42:53 it's going through a mosfet to swtich it on/off, and a .25W R15 Mar 20 09:42:54 btw: will this be an neo6-t or one of the standard modules? Mar 20 09:43:02 before hitting the L1 Mar 20 09:43:06 just a standard one Mar 20 09:43:10 meh... Mar 20 09:43:11 ;) Mar 20 09:43:40 yes, that 6t seems interesting Mar 20 09:43:42 what's the circuitry at the top, near the antenna trace? Mar 20 09:43:50 powers the antenna Mar 20 09:43:55 hmm? Mar 20 09:44:12 you just said the neo powers the antenna? Mar 20 09:44:37 it might be easier to post the schematics also :) Mar 20 09:44:49 externally, through Q1 and R15 Mar 20 09:44:53 yeah :) Mar 20 09:44:57 you'll need a special eth-driver to avoid bit-sequences which will disturb the rf ;) Mar 20 09:45:00 ah.. Mar 20 09:45:00 ok Mar 20 09:45:20 Shadyman: make sure L1 is a low capacitance, rf type then Mar 20 09:45:27 yep Mar 20 09:45:38 using their recommended one Mar 20 09:45:40 and are you sure the antenna trace is 50R? it doesnt look like Mar 20 09:45:52 Should be. Mar 20 09:45:58 calculated? Mar 20 09:46:00 yeah Mar 20 09:46:04 hmm... Mar 20 09:46:10 dual layer FR4 Mar 20 09:46:11 ok, then it must be me Mar 20 09:46:27 how does this 50R matching work, you calculate it, produce the PCB and measure it? Mar 20 09:46:29 used the HP/Agilent AppCAD Mar 20 09:46:48 ynezz: with FR4 you hope that it's 50R+/-50% Mar 20 09:47:00 d'oh Mar 20 09:47:19 ynezz: Nah, you feed it how thick the PCB is, what material, how thick the copper is, spacing and trace width, etc, and it gives you the capacitance Mar 20 09:47:23 ynezz: FR4 has a very baldy defined \epsiolon, highly dependent on the air humidity Mar 20 09:47:31 ynezz: but for short traces it shouldnt be much of a problem Mar 20 09:48:05 for FR4, 1000 micrometer trace width and 184(?) micrometer spacing gave me 50 ohms on the dot Mar 20 09:48:43 including the ground plane underneath? Mar 20 09:48:49 Yes. Mar 20 09:48:52 hmm.. Mar 20 09:48:56 then it's me :) Mar 20 09:49:01 it was the, eh, what's it called Mar 20 09:49:03 * KotH is used to wider spacings Mar 20 09:49:08 the channel with ground on the sides and ground below Mar 20 09:50:04 * KotH cannot remember those names either :) Mar 20 09:50:40 Er, I HAD filter caps here somewhere? Mar 20 09:50:44 * Shadyman headscratches Mar 20 09:50:58 trolls took them! Mar 20 09:51:02 ^ Mar 20 09:51:13 recommendations? Mar 20 09:51:20 there's a 0.1uF on the eeprom at the moment Mar 20 09:51:27 but nothing on the neo per se Mar 20 09:51:27 uhmm.. place a 10pF 1n 100nF at the RF part and a 100nF at the baseband part Mar 20 09:51:43 uhmm.. Mar 20 09:51:51 those capacitors need to be _at_ the power pins of the neo Mar 20 09:52:02 otherwise you loose quite a lot of sensitivity Mar 20 09:52:06 yeah Mar 20 09:52:34 :/ Mar 20 09:52:44 cases of the smaller caps should be 0402 if possible (lower stray inductance) Mar 20 09:53:04 * Shadyman winces and orders the small tweezers Mar 20 09:53:46 100nF = 0.1uF, right? Mar 20 09:53:51 6AM :/ Mar 20 09:53:54 you might also want to place a 10u (if X5R >10V, otherwise tantal 6.3V) somwhere in the 3.3V line as well Mar 20 09:54:01 yes Mar 20 09:54:04 k Mar 20 09:54:58 the big cap is mostly because the power supply comes with a long trace from the bone, hence has quite a high inductance which should be compensated Mar 20 09:55:08 * Shadyman nods Mar 20 09:55:37 the ESR of it is not so critical if you have enough caps directly at the neo Mar 20 09:55:54 but ceramic is still prefered Mar 20 09:55:58 10pF 1n 100nF at the RF part all 16V? Mar 20 09:56:14 their voltage rating is not so important Mar 20 09:56:25 i doubt you'll find anything below 25V these days Mar 20 09:56:44 iirc the last 100nF's we bought for lab use were 100V's Mar 20 09:56:49 heh Mar 20 10:00:42 KotH: It'll need a bit of zooming in, but http://erroraccessdenied.com/GPS_Cape.png Mar 20 10:01:26 * KotH feels reminded of his highschool days when he used the 386 version of orcad Mar 20 10:02:41 besides the fact that all the new caps are called C1 :/ Mar 20 10:03:01 hehe Mar 20 10:03:07 gEDA :) Mar 20 10:03:13 in OMG THE COLORS mode. Mar 20 10:03:52 and yes, the symbol for the bbone is absurdly large :/ Mar 20 10:04:21 you dont want to wire ANTSHORT directly to the bone Mar 20 10:04:46 actually, you dont want to have anything at that point beside what is absolutely necessary Mar 20 10:04:55 fair nuf. Mar 20 10:05:25 should I just skip the antenna detection? Mar 20 10:05:26 if you really have to get that signal from there, decouple it using an FD302N Mar 20 10:06:28 the antenna detection isnt bad per se, you just have to be very carefull what you are doing Mar 20 10:06:33 the NEO-controlled antenna uses ANTOFF to switch the MOSFET'd VCC_RF through to R15 an dL1 Mar 20 10:06:48 that's ok Mar 20 10:07:03 so that'd nuke R16 and R18, and U4 Mar 20 10:07:26 i'd probably place a 10pF 1nF at the drain of Q1 to stabilize that point further Mar 20 10:07:36 which I'd be okay with. it was more of a "help prevent users from running it without an antenna" move Mar 20 10:07:37 the C1's should be all near the neo Mar 20 10:07:44 they have to stabilize the electronics there Mar 20 10:08:07 r16/r18 is ok imho Mar 20 10:08:46 they decouple u4 from the antenna, so it doesnt inject any noise (not that it has much to inject) Mar 20 10:09:01 and u4 decouples the noisy electronics from the antenna Mar 20 10:09:18 so ANTSHORT is nuked Mar 20 10:09:40 but, u4 should be powered from 3.3v and not vcc_rf, otherwise you'll inject noise trough that path Mar 20 10:09:49 and u4 does need it's own 100nF capacitor Mar 20 10:10:04 dammit ublox ;) Mar 20 10:10:14 why? Mar 20 10:10:33 their integration manual missed a bunch :/ Mar 20 10:11:10 well, that's standard EE procedure, anyone knows that kind of stuff ;-) Mar 20 10:11:16 hehe Mar 20 10:11:33 maybe i would connect the inverting input of u4 onto the other side of r5 to decouple it a bit from the antenna Mar 20 10:11:47 er. forgete that, doesnt work Mar 20 10:12:22 but a 1k resistor before the inverting input wouldnt be too bad Mar 20 10:12:27 Could U4 be wired to antoff_out, along with an LED of some sort, and skip the bbone altogether in that respect? Mar 20 10:12:59 so that it still disables if there's a short or open circuit, but doesn't need bbone code? Mar 20 10:13:21 sure Mar 20 10:13:40 but you have to design the circuit as a current limiter then Mar 20 10:13:48 otherwise you get an oscillator in case of a short Mar 20 10:13:53 fun Mar 20 10:13:56 +1 Mar 20 10:14:37 Now, they're saying VCC_RF is actually VCC-0.1 Mar 20 10:14:46 is that within tolerance for the op amp? Mar 20 10:15:05 read the datasheet ;) Mar 20 10:15:09 heh Mar 20 10:15:19 but it will be ok Mar 20 10:15:36 0.1V is still below the protection diode voltage Mar 20 10:15:45 but vcc_rf shouldnt be that much lower than vcc Mar 20 10:15:54 well, i would assume, as it's only comparing the two inputs, and it's actually being powered higher than its inputs Mar 20 10:15:59 if it is, you have too much resistance in the path Mar 20 10:16:06 they call for a rail-to-rail opamp Mar 20 10:16:18 yes, you are sensing near th upper rail Mar 20 10:18:11 Shadyman: if all you want is to measure the current on the line, why not use one of the power sense chips that have an integrated resistor? Mar 20 10:19:09 they are designed to sense at upper rail and you dont have to worry about that Mar 20 10:19:31 i was just going with what they suggested in the manual, but acurrent sense chip would sense above-average, and zero, current Mar 20 10:19:31 the only thing you have to do then is to place a couple of capacitors before it goes into L1 Mar 20 10:20:07 we talking current sense chips like SMSC? Mar 20 10:20:41 high side current sensors Mar 20 10:21:08 http://www.smsc.com/Products/Thermal_and_Power_Management/Power-Current_Sensors/PAC1710 Mar 20 10:23:26 too fancy Mar 20 10:23:34 you dont want to have any digital electronics there Mar 20 10:23:40 only analog, stady state stuff Mar 20 10:23:45 ok Mar 20 10:23:52 something like the MAX9610 Mar 20 10:24:00 good ol maxim Mar 20 10:24:06 doesnt have to be maxim Mar 20 10:24:15 linear, ti, national,... have similar parts Mar 20 10:24:34 just had that part number at hand as we recently used it for a project Mar 20 10:25:07 of course :) Mar 20 10:25:19 * KotH points at permasense ;) Mar 20 10:25:31 hehe Mar 20 10:25:39 you were saying that some have integrated sense resistors? Mar 20 10:25:39 this one: http://www.permasense.ch Mar 20 10:25:46 yes Mar 20 10:25:55 pretty Mar 20 10:26:07 KotH is peddling swiss chips again Mar 20 10:26:10 though i would go for the external sensors ;) Mar 20 10:26:12 do you get a commision? Mar 20 10:26:18 av500: chips? Mar 20 10:26:23 hehe Mar 20 10:26:27 av500: swiss research projects, yes, not chips Mar 20 10:26:29 Swiss Army Chip? Mar 20 10:26:36 +1 Mar 20 10:26:46 they are called FPGA I think Mar 20 10:26:50 :D Mar 20 10:27:29 once upon a time, there was an anlog equivalent of an fpga as well Mar 20 10:27:39 unfortunately, it didnt really catch on Mar 20 10:28:11 PLDs? Mar 20 10:28:16 hmm.. taht must have been 15y ago ^^' Mar 20 10:28:24 no, PLDs are still digital Mar 20 10:28:27 hmm Mar 20 10:30:16 does the neo really need some kind of serial flash/eeprom? Mar 20 10:30:31 It's ROM/RAM only Mar 20 10:30:56 it'll work on its own without flash/eeprom, but with flash/eeprom, it can store longer-term projections of ephemeris, etc Mar 20 10:31:08 ahh. Mar 20 10:31:09 especially helping boot times Mar 20 10:31:13 well... Mar 20 10:31:25 granted, bbone could feed it that Mar 20 10:31:46 the ephemeris data is only of use for a day and only if you have the exact time Mar 20 10:32:04 if you've lost time, the ephemeris data is not going to help Mar 20 10:32:14 it's got an rtc :) Mar 20 10:32:35 i assume the backup battery powers the ephemeris store as well Mar 20 10:32:41 at least it does that on the lea modules Mar 20 10:32:42 yes Mar 20 10:32:48 battery backed up ram (BBRAM) Mar 20 10:32:57 LEAs are so much fancier :D Mar 20 10:33:02 not that much Mar 20 10:33:45 so if I were using a current sense amplifier, which resistor would I current sense? R15? Mar 20 10:34:03 yes Mar 20 10:34:27 that'd nuke R16 and R18 Mar 20 10:34:46 yes Mar 20 10:35:14 you'll still need somewhere a comparator to check whether the current is too high Mar 20 10:35:37 the rent is too damn high! Mar 20 10:35:47 lol Mar 20 10:35:58 it can be a low speed, low specs comparator :) Mar 20 10:36:04 KotH: resistor divider to an adc on the bbone? Mar 20 10:36:12 also possible Mar 20 10:36:25 just make sure you have enough drive power Mar 20 10:36:26 which one would be cheaper, though? Mar 20 10:36:35 probably decouple it with a voltage follower Mar 20 10:36:40 Shadyman: did you have to make the PCM larger already? Mar 20 10:36:43 the PCB* Mar 20 10:36:43 neither? Mar 20 10:36:52 lol av500 Mar 20 10:36:54 a comparator costs 0.1 or so Mar 20 10:37:19 you can literally choose the cheapest one Mar 20 10:37:37 74lcvxx series, even? Mar 20 10:37:38 well..actually, you should use the cheapest one that both digikey and mouser have on stock ;) Mar 20 10:37:43 heh Mar 20 10:37:52 lcv or lvx, I can never remember Mar 20 10:37:55 i would use a comparator, not a schmitt trigger Mar 20 10:38:00 ok Mar 20 10:38:04 lvc and lvx ;) Mar 20 10:38:12 eh, well Mar 20 10:38:20 schmitt triggers have not so well defined switching voltages Mar 20 10:38:38 rather large tolerances Mar 20 10:38:52 what you want is to have a switching level defined to.. let say 1-5%, with some hysteresis Mar 20 10:38:58 just let KotH design the pcb, done. no exact steps needed anymore Mar 20 10:39:04 lol Mar 20 10:39:06 haha Mar 20 10:39:48 though, I'm thinking ADC Mar 20 10:40:05 as not only would I want to compare against high current, I'd also want to compare against too little Mar 20 10:40:09 it won'tgo through the tsa, but who cares Mar 20 10:40:21 Shadyman: take a douple comparator Mar 20 10:40:23 :) Mar 20 10:40:58 aholler: mranostay did get his geiger counter trough TSA, so this shouldnt a problem Mar 20 10:41:10 I would bitbang it Mar 20 10:41:13 KotH: he doesn't call for war Mar 20 10:41:40 av500: Giggity. Mar 20 10:41:48 aholler: there are FDA aproved circuits floating around in the us, that have my finger prints on, already ;) Mar 20 10:41:55 KotH: Er, Linear comparator or logic comparator? Mar 20 10:42:03 Shadyman: lol Mar 20 10:42:08 digikey y so many options Mar 20 10:42:09 Shadyman: ofc an analog comparator Mar 20 10:42:16 er, duh Mar 20 10:42:19 are there digital ones? Mar 20 10:42:38 there was some 74xx that did compare 4 bit numbers Mar 20 10:42:49 of course there are digital comparators Mar 20 10:43:01 Shadyman: or do they mean comparators with digital output? Mar 20 10:43:02 well, a XNOR Mar 20 10:43:11 exactly Mar 20 10:43:26 so a NOR is a digital maybe :) Mar 20 10:43:27 KotH: Beats me. Mar 20 10:43:29 Shadyman: and i would use mouser for that type fo search, their website is better suited for "i need something that looks like this but doesnt have high specs" Mar 20 10:43:43 er, OR Mar 20 10:43:46 gah Mar 20 10:43:53 not much logic in me today Mar 20 10:44:03 there ever was? Mar 20 10:44:06 all the logic gates perform some kind of digital comparison Mar 20 10:44:21 fact, confusing, stop Mar 20 10:44:29 haha Mar 20 10:45:24 * KotH goes back to programming his msp430 and leaves av500 alone in his confusion Mar 20 10:45:31 same here Mar 20 10:45:35 it behaves weird Mar 20 10:46:36 Shadyman: I blame you for aliexpress Mar 20 10:46:42 now I ordered something there Mar 20 10:46:43 okay Mar 20 10:48:52 KotH: I'm going to write myself notes for tomorrow and then plow into bed like a snowplow into a street full of snow Mar 20 10:49:35 * KotH wouldnt mind spring weather Mar 20 10:49:40 sleep well Mar 20 10:49:59 hehe, thanks :) Mar 20 10:50:18 av500: new launchpad? Mar 20 10:50:24 av500: or a different gadget? Mar 20 10:50:44 from ali? just cables Mar 20 10:50:55 msp430 is my kitchen light project Mar 20 10:51:05 with ALDI led chain Mar 20 10:51:49 like a true german! Mar 20 10:53:57 av500: let it show the oven temperature like those philip tvs Mar 20 10:54:24 on the wall Mar 20 10:54:32 the ambient factor Mar 20 10:55:15 it will be mostly a totally boring ON/OFF thing Mar 20 10:55:40 mind controlled? Mar 20 10:55:45 switch Mar 20 10:55:49 old school Mar 20 10:56:13 but I will add some color cycling "party mode" Mar 20 10:57:17 using a rtc it would be good as an egg-timer Mar 20 10:57:46 wont work in daylight Mar 20 10:57:59 I'd need to add a buzzer Mar 20 10:58:08 or a 1kW flash light Mar 20 10:58:12 and there are already 2 timers in the kitchen Mar 20 10:58:14 dont need a 3rd Mar 20 10:58:19 oven has one, stove has one Mar 20 10:58:31 are they that dim? Mar 20 10:59:02 they wont light the room if you dont look into that direction Mar 20 11:05:34 av500: try one of these http://www.isolicht.com/index.php/cat/c19_LED-Flex-Stripes.html Mar 20 11:06:04 KotH: sure, I have these Mar 20 11:06:16 av500: like: http://www.isolicht.com/product_info.php/info/p248_High-End-Stripe-5m---Flexibles-LED-Lichtband----14-4W---RGB-24V.html Mar 20 11:06:19 14W/m Mar 20 11:06:21 and I rather buy them from china for less Mar 20 11:06:38 yes Mar 20 11:07:46 KotH: that ALDI thing is 3led group adressable Mar 20 11:09:22 digitally adressable? Mar 20 11:09:27 yes Mar 20 11:09:34 http://dx.com/p/72w-300-5050-smd-led-rgb-light-flexible-strip-dc-12v-5-meter-114454 Mar 20 11:10:21 costs a third only :) Mar 20 11:12:58 and other being of lower quality, it seems the same :) Mar 20 11:13:14 well Mar 20 11:13:31 its only 12V Mar 20 11:13:38 24V might be better for long chains Mar 20 11:14:33 yeah.. i wouldnt use it as a 5m string with 12V, not at that power rating Mar 20 11:14:51 I like this one: http://www.aliexpress.com/item/4m-WS2811-LED-digital-strip-60leds-m-with-60pcs-WS2811-built-in-tthe-5050-smd-rgb/633124138.html Mar 20 11:15:03 240 invividually addressable leds :) Mar 20 11:16:46 av500: when you get your msp code working, i'd be interested in it ;) Mar 20 11:16:56 * KotH still needs a good bed lamp Mar 20 11:17:09 and a fully controllable led bar wouldnt be too wrong :) Mar 20 11:17:22 yeah, I'll github it Mar 20 11:17:52 the msp code? Mar 20 11:17:53 av500: 320 of those and you get a 5.5x4m 320x240 led matrix display Mar 20 11:18:13 the aldi thing was a bad buy, I thought it was single led addressable Mar 20 11:18:16 not 3-led Mar 20 11:18:23 so it goes as a kitchen light now Mar 20 11:18:31 janne: yep Mar 20 11:18:47 janne: should we start a paypal? Mar 20 11:19:10 s/paypal/kickstarter/? Mar 20 11:19:14 right Mar 20 11:19:27 that way we dont need to deliver :) Mar 20 11:19:40 risks: av500 is involved!! Mar 20 11:21:14 hmm, that led-strip has a sd-card Mar 20 11:21:51 * KotH wonders how closely one could space high power RGB leds w/o too much cooling involved Mar 20 11:22:12 get a tv Mar 20 11:22:13 one might consider making boards of those and put a full-HD display together Mar 20 11:22:24 aholler: nah! not geeky enough! Mar 20 11:23:01 aholler: just consider, you enter a dudes flat, and the first thing that greets you is a wall that doubles as a humongus TV :) Mar 20 11:23:39 we should go for the 4k version for kickstarter. that would be 64x36m display. good enough for a cinema Mar 20 11:23:46 right Mar 20 11:23:50 I rather would build me a rgb-led-cube Mar 20 11:24:05 $1000 reward: we will scroll your name once across the display Mar 20 11:24:16 aholler: floating above you, and telling you about delicious cake? Mar 20 11:24:34 $5000: reward: you can watch us watch a movie in the led wall Mar 20 11:25:06 $100 reward: a can of solder fumes from the assembly Mar 20 11:25:08 +25 Mar 20 11:25:52 $10 reward: 6 months exciting anticipation Mar 20 11:26:27 $12: we will increase the anticipation to even 12 month Mar 20 11:26:46 $15: lifetime anticipation Mar 20 11:26:49 3 million dollar seems about right, led stripes without shipping are 1.5 million Mar 20 11:27:03 janne: they do get cheaper in bulk Mar 20 11:27:40 http://hackaday.com/2011/03/20/third-times-a-charm-512-led-cube-kicks-it-up-a-notch-with-rgb-leds/ Mar 20 11:27:54 yes Mar 20 11:29:37 * KotH is hungry Mar 20 11:29:42 time to hunt some greeks Mar 20 11:30:06 +1 Mar 20 11:59:06 hey, on the bone, how do I find out at which address I can flash a new u-boot image from u-boot ? Mar 20 12:04:52 your wednesday bitbang comment of the day, "/* Set up the SPI or Microwire EEPROM for bit-bang reading." Mar 20 12:04:54 oooh yeah Mar 20 12:09:59 gqq: flash whereto exactly? Mar 20 12:10:34 the nand flash of course Mar 20 12:10:38 the one on the sd card Mar 20 12:10:49 av500, I was thinking of tftp'ing the new uImage and mmc erase/mmc write to update it ? Mar 20 12:11:08 gqq: you bone has mmc? Mar 20 12:11:14 is it, uh, black? Mar 20 12:11:22 ah well Mar 20 12:11:25 the sdcard Mar 20 12:11:33 but that uses a file system Mar 20 12:11:44 I doubt mmc write plays nicely with it Mar 20 12:13:23 ah, I was afraid it wouldn't be possible Mar 20 12:13:39 swapping the sd around is quite a hassle to test stuff Mar 20 12:14:39 well Mar 20 12:14:46 if you tftp, why do you need to write to the card? Mar 20 12:14:50 tftp the uimage Mar 20 12:14:56 nfs Mar 20 12:14:56 and nfsmount the rootfs Mar 20 12:15:02 yep Mar 20 12:15:04 you do not need sdcard at all Mar 20 12:15:11 you need it Mar 20 12:15:18 you need a kernel :) Mar 20 12:15:24 tftp Mar 20 12:15:28 I need a first stage bootloader at least Mar 20 12:15:35 h right Mar 20 12:15:36 yes Mar 20 12:15:38 well, s/kernel/bootloader/ Mar 20 12:15:42 I have mlo and uboot on the card Mar 20 12:15:44 MLO and uboot Mar 20 12:15:48 the rest is tftp and nfs Mar 20 12:15:48 right Mar 20 12:15:49 yes yes Mar 20 12:16:02 and you can then write a new uimage from linux itself Mar 20 12:16:07 (if needed) Mar 20 12:16:27 so the stupid AM335 rom code cannot net boot? Mar 20 12:16:45 ah I said uImage, sorry I want to update u-boot.img Mar 20 12:16:47 :) Mar 20 12:17:02 it's u-boot that I modify, not the kernel Mar 20 12:17:08 yes Mar 20 12:17:12 tftboot Mar 20 12:17:15 nfsroot Mar 20 12:17:25 mount partion 1 Mar 20 12:17:25 gqq: tftp uboot and jump to it Mar 20 12:17:27 copy uboot Mar 20 12:17:40 use 1 uboot as bootloader, the other uboot as what you "work" on Mar 20 12:17:58 just change the start address of the 2nd one Mar 20 12:18:07 you mean boot uboot from uboot ? Mar 20 12:18:11 yes Mar 20 12:18:24 I didn't know I could do that ! Mar 20 12:18:27 great Mar 20 12:18:29 its SW Mar 20 12:18:31 thanks Mar 20 12:23:29 one greek down, 20 kebabs made for the hungry masses Mar 20 12:23:36 \o/ Mar 20 12:23:47 lol Mar 20 12:23:53 I'm not dead yet Mar 20 12:26:10 panto: we still need some livestock for tomorrow Mar 20 12:26:31 can't you go hunt down some kurds instead? Mar 20 12:27:02 nah.. they are way to skinny Mar 20 12:45:20 av500, rom has net support Mar 20 12:47:40 but? Mar 20 12:47:48 does net have rom support? Mar 20 12:47:59 does the rom doe TCP/IP? Mar 20 12:48:01 support net rom but? Mar 20 12:49:10 but it might not work Mar 20 12:49:16 it depends Mar 20 12:49:21 on? Mar 20 12:49:28 I dunno Mar 20 12:49:32 all roads lead to rom ... Mar 20 12:49:49 it's an ongoing saga Mar 20 12:50:02 might even work on black Mar 20 12:50:03 mdp: do you think I can solder a 2nd ROM chip on top and use a jumper to select which one to use? Mar 20 12:50:07 Tartarus knows all Mar 20 12:50:19 av500, you are empowered Mar 20 12:50:20 like on my C64? Mar 20 12:50:47 yeah, just open up the case on the chip and you can white wire it in Mar 20 12:51:07 * av500 has only green wire :( Mar 20 12:51:29 fail Mar 20 12:51:39 you follow exact steps or just don't even try Mar 20 12:52:16 are you sure it's not teal? Mar 20 12:52:37 I know nothing Mar 20 12:52:45 * mdp recruits the fairer sex to identify teal vs. green for him Mar 20 12:53:08 Tartarus, lies Mar 20 12:53:17 ok, fine Mar 20 12:53:28 Tartarus, also knows how to green-wire up a custom rom to the am335x Mar 20 12:53:40 my blog only has white-wire steps Mar 20 12:53:44 av500: stupid ROM code can do network booting, but this is a "fun" argument I have with folks internally Mar 20 12:54:13 GP EVM isn't supported because of HW choices, an errata RGMII mode is required for use. Mar 20 12:54:31 and BBB? Mar 20 12:54:35 or BBW? Mar 20 12:54:35 And folks keep telling me some other errata means that beaglebone also can't be used, due to some other issue Mar 20 12:54:45 BBB should be fine, IF you can expose the sysboot pins Mar 20 12:54:52 they are exposed Mar 20 12:54:57 outer layers of the PCB Mar 20 12:54:59 I don't know if the "memory cape" which also exposes the full set of SYSBOOT pins works on BBB Mar 20 12:55:17 Tartarus: hmm... teal... german doesnt even have a word for it Mar 20 12:55:45 av500: If you can change all of 0:7 or so, then yes, you can do cpsw network booting Mar 20 12:55:46 KotH: ministry of culture can create a word Mar 20 12:56:04 KotH: it has: Krickente! Mar 20 12:57:40 av500: lieber pekingente! Mar 20 12:57:56 mdp: there is no ministry of culture Mar 20 12:58:05 that's what they tell you Mar 20 12:58:18 the ministry of truth has you confused Mar 20 12:58:28 mdp: beside those angry, heavily armed and stuborn mountain folk would strongly disagree with any such decision Mar 20 12:58:47 dwarves? ;) Mar 20 12:59:00 av500, actually I don't understand how one would "jump" to an address in uboot. All the boot commands seem to expect a kernel image Mar 20 12:59:33 gqq: read the documentation on your left Mar 20 12:59:36 bootm 0xabcdefg Mar 20 13:00:08 m = memory Mar 20 13:00:44 the memory list at the bottom is nice: http://tr.wikipedia.org/wiki/Turkuaz Mar 20 13:00:50 maybe I need to clear some env variables, but it says it expects a kernel image Mar 20 13:03:18 http://www.denx.de/wiki/view/DULG/UBootCmdGroupExec#Section_5.9.4.2. Mar 20 13:04:07 ah Mar 20 13:04:09 maybe go then Mar 20 13:04:44 well no Mar 20 13:04:51 bootm should be fine Mar 20 13:06:22 go Mar 20 13:06:32 jut go on Mar 20 13:06:35 +s Mar 20 13:06:48 hum i'f youre positive that "tftp 0x80000000 u-boot.img" followed by "bootm 0x80000000" should indeed boot u-boot from u-boot, then maybe the secure u-boot patches I applied broke that behavior Mar 20 13:07:34 go secure Mar 20 13:08:05 are you sure 0x80000000 is "free"? Mar 20 13:08:13 where does uboot itself reside? Mar 20 13:08:35 before you apply stuff, try it with the same uboot Mar 20 13:08:51 imma retry with 0x80300000, I'm sure it's free Mar 20 13:09:48 u-boot relocates to top Mar 20 13:10:05 ok Mar 20 13:10:28 av500, http://ideone.com/C572aU Mar 20 13:10:39 but some hw doesn't like it to get initialized twice Mar 20 13:10:57 not sure if that's the case with the bone Mar 20 13:13:15 ah alright, there's a paragraph about the FIT images in help bootm Mar 20 13:13:24 as said, go on. Mar 20 13:13:24 might be my problem Mar 20 13:13:41 you mean I've got to drop it ? Mar 20 13:14:09 why does it say automaic boot? Mar 20 13:14:20 over here, tftpboot only loads stuff Mar 20 13:14:21 I set autoboot to yes Mar 20 13:14:45 autostart* Mar 20 13:14:55 well, then dont Mar 20 13:14:56 go 0x80300000 Mar 20 13:16:27 aholler, "go" begins to start u-boot but hangs quite rapidly Mar 20 13:16:57 see above. chain boot doesn't always work Mar 20 13:17:52 http://www.denx.de/wiki/view/DULG/CanUBootBeConfiguredSuchThatItCanBeStartedInRAM Mar 20 13:17:53 ouch Mar 20 13:18:00 so it's assumed impossible until someone very smart come in ? :D Mar 20 13:18:30 se above link Mar 20 13:18:31 see Mar 20 13:18:39 av500, well, TIL I should read FAQs Mar 20 13:18:55 thanks for all the troubles though :) Mar 20 13:19:03 the text at the link is just there to avoid dumb questions ;) Mar 20 13:21:15 its also partly wrong Mar 20 13:21:26 since MLO loaded uboot is also in RAM already Mar 20 13:21:35 sure, u-boot changed and changes a lot Mar 20 13:21:44 gqq, it is possible Mar 20 13:21:48 so such things get outdated quit fast Mar 20 13:22:11 aholler: yes, the FAQ is for people that quit fast :) Mar 20 13:22:21 you must however remove the the early initialization and/or the copying that u-boot does from flash to ram Mar 20 13:22:33 panto: from what flash? Mar 20 13:23:03 panto, I've got enough trouble with what I'm trying in u-boot already :) Mar 20 13:23:08 29F040? Mar 20 13:23:08 normal u-boot does relocation Mar 20 13:23:25 MLO loads to RAM and start uboot Mar 20 13:23:32 uboot load to RAM and starts uboot Mar 20 13:23:37 non? Mar 20 13:23:39 ah yes Mar 20 13:23:46 so, ram argument is moot Mar 20 13:23:58 yeah MLO case is different Mar 20 13:24:28 My guess is that it's hard to replace yourself Mar 20 13:24:46 why Mar 20 13:24:49 in the MLO case should be easier Mar 20 13:24:55 I'd have thought that chain loading is quite used in actual bootloader development Mar 20 13:25:14 I've used it Mar 20 13:25:18 I would just add an extra config option to u-boot for a chain-loaded u-boot target Mar 20 13:25:19 * av500 too Mar 20 13:25:42 or configure a u-boot to be loaded elsewhere in memory Mar 20 13:25:47 I did do u-boot in RAM (but it was booting from real falsh) Mar 20 13:26:02 will the real flash please stand up? Mar 20 13:26:09 gqq: you are sure that 0x80300000 is where the code does start? Mar 20 13:26:46 which code ? all I know is it's free Mar 20 13:27:12 text_foo, can't remember the exact name Mar 20 13:27:21 it's buried in the config somewhere Mar 20 13:28:05 you have to load u-boot to where it thinks it gets loaded Mar 20 13:30:14 there are too much variables in this problem for me to understand Mar 20 13:30:25 no Mar 20 13:31:09 CONFIG_SYS_TEXT_BASE Mar 20 13:31:33 gqq: making uboot chain load is a worthy task Mar 20 13:32:21 find . -type f -exec grep CONFIG_SYS_TEXT_BASE {} + | grep am33 Mar 20 13:32:21 ./include/configs/am335x_evm.h.orig:#define CONFIG_SYS_TEXT_BASE 0x80800000 Mar 20 13:32:37 so that is where you should load it to Mar 20 13:34:21 aholler: well ... it does boot something ! just need to recompile another u-boot to confirm it's the one I just loaded Mar 20 13:34:42 but i'm not sure, might be different with spl or has changed Mar 20 13:35:55 it's at least a year ago since I've last touched u-boot ;) Mar 20 13:36:19 it runned away, full with fear Mar 20 13:36:43 use fluffy glooves next time Mar 20 13:37:01 and warm them up before Mar 20 13:37:04 I'm testing, could be that you were right on the spot :) Mar 20 13:39:11 aholler, yup, it loads the uboot I just recompiled Mar 20 13:39:18 thanks :D Mar 20 13:40:58 someone should write freely relocatble relocation code Mar 20 13:42:12 pe-u-boot Mar 20 13:43:12 aholler, or you could port http://git.pengutronix.de/?p=barebox.git;a=commit;h=a81ec0225f5a100341c20b4329c8b1d81ab025c4 to u-boot ;) Mar 20 13:46:37 someone else might do it. I'll avoid assembler (and u-boot) if possible Mar 20 13:50:08 for dev purposes having to know the load address is ok Mar 20 13:52:07 av500, being able to do just "bootm /mnt/nfs/barebox" during development keeps me from making stupid errors... Mar 20 13:53:34 one can define scripts for uboot, ya know? Mar 20 13:55:19 of course, but why bother when you don't to remember addresses for each soc Mar 20 13:56:31 can't be that hard to make the relocation code relocatable Mar 20 13:57:08 there just have to be someone who wants to do that Mar 20 13:57:43 and going through the pain to get such a patch upstream ;) Mar 20 14:00:07 s/relocatable/position idependent/ Mar 20 14:03:46 i would expect uboot to already copy itself to the text base Mar 20 14:04:01 no it copies itself on top of ram Mar 20 14:04:58 but the copying stuff is still at a fixed address Mar 20 14:05:19 + early_init I think Mar 20 14:06:59 oh Mar 20 14:28:24 lol: http://linuxundich.de/de/ubuntu/gfk-symbolisiert-raubkopiermorder-mit-redhat-logo/ Mar 20 14:29:53 haha, yes, those readhats do copy a lot ;) Mar 20 14:30:40 evil linux people Mar 20 14:33:11 here is some text for the english reading people here: http://www.reddit.com/r/linux/comments/1ahh6z/german_gfk_society_for_consumer_research_uses_the/c8xr44z Mar 20 14:35:01 knowing that company, i doubt it was an intern Mar 20 14:36:40 (side note: some of the devices gfk are using are running linux and other gpl stuff.. w/o any declaration or publicated source) Mar 20 14:37:02 morning friends Mar 20 14:37:05 and KotH Mar 20 14:37:22 KotH: it isn't forbidden to do so Mar 20 14:37:27 moin mranostay Mar 20 14:37:32 aholler: well it isnt Mar 20 14:37:58 aholler: the thing is, those boxes are shipped to participants in their studies and stay there for years Mar 20 14:38:13 my tablet uses a kernel to, no documentation or declaration Mar 20 14:38:15 aholler: beside being sold to other companies and market research institutes Mar 20 14:41:15 aholler: you can get them for that Mar 20 14:42:21 aholler: according to gplv2 sec1 distributions of the software in any form have to "appropriately publish on each copy an appropriate copyright notice" Mar 20 14:42:58 i can read myself and have read the gpl centuries ago Mar 20 14:43:47 just because you sadi "it isnt forbidden to do so" Mar 20 14:44:12 it isn't, i use linux since ages without having printed me a copy of the gpl Mar 20 14:44:52 nor did I write documentation for my devices ;) Mar 20 14:46:07 gah.. Mar 20 14:46:17 you're one of those archos guy as well? Mar 20 14:46:26 building your own tablets and stuff? Mar 20 14:47:29 he was here in the office once :) Mar 20 14:47:35 other than that, no Mar 20 14:47:49 KotH: see, you first didn't mention that they sell devices. Mar 20 14:48:24 and almost every company is using some linux Mar 20 14:48:40 that don't make them evil Mar 20 14:50:17 anyway, I just got an invitation from lego education in regard to visit an ev3 course ;) Mar 20 14:52:11 aholler: well.. gfk is evil.. i know it.. i worked for them Mar 20 14:52:13 I was begging to go visit them onsite when we were supporting them ;) Mar 20 14:52:18 er, aholler Mar 20 14:53:10 you stopped supporting them? ;) Mar 20 14:53:34 I no longer provide support, that's correct Mar 20 14:55:16 hopefully they are already happy Mar 20 14:55:43 their happiness is not my concern ;) Mar 20 14:56:06 but they won't find a better supporter Mar 20 14:56:41 they are shipping...factory support calls generally end after the 911 production stop issues Mar 20 14:57:22 luckily, all I did was non-stressful driver devel Mar 20 14:57:33 like musb? ;) Mar 20 14:57:45 lol Mar 20 14:58:14 no, we have smart people to work on that Mar 20 15:01:20 I wonder where they got my email from. they seem to think I'm a teacher Mar 20 15:03:23 be sure to request a free classroom set of EV3 kits ;) Mar 20 15:04:49 would do so if they would offer such. unfortunately they want to sell me reduced nxt kits Mar 20 15:05:07 s/reduced/discounted/ Mar 20 15:05:23 * KotH curses Ti Mar 20 15:05:43 no offense dudes... but.. some shit is brain dead.. Mar 20 15:07:40 my BB is not detected Mar 20 15:14:05 * aholler wonders where they ship the ev3s to ;) Mar 20 15:18:00 KotH: not possible, you're doing it wrong Mar 20 15:18:15 aholler: clearance sale ;) Mar 20 15:18:21 * mru curses TI Mar 20 15:18:47 * mdp curses users Mar 20 15:18:54 * panto curses cursrs Mar 20 15:18:57 *curses Mar 20 15:19:01 * mdp ncurses TI Mar 20 15:19:06 users are evil Mar 20 15:19:47 agreed, only thing more evil is a combined user/customer Mar 20 15:19:49 * mdp shivers Mar 20 15:22:02 I'm awaiting a press release from the linux foundation regarding "lego catched" ;) Mar 20 15:22:55 mdp: not you, the other TI Mar 20 15:23:04 all your kids belong to us ;) Mar 20 15:23:18 aholler: nope, av500's daughter belongs to me Mar 20 15:23:19 hehe Mar 20 15:23:38 KotH: people don't own poeple nor are they owned Mar 20 15:23:44 * mdp hands KotH 3 creeper cards Mar 20 15:24:04 aholler, can they be pwned? Mar 20 15:24:06 hey! i aquired her in a fair trade! Mar 20 15:24:37 bible says you can own slave and sell your daughter Mar 20 15:25:00 mdp: sure, I assume there is no evil drm in the ev3 Mar 20 15:25:28 supposedly open Mar 20 15:26:02 considering that it's an am180x and the brick has a uSD slot for booting...it's ready to hack Mar 20 15:27:23 I think it might be a nice device to play with. Mar 20 15:27:59 hoping the kids won't discover this channel ;) Mar 20 15:29:17 might become crowed with questions about how to control the motors Mar 20 15:29:28 using opencv Mar 20 15:29:34 istnt there SW from lego for that? Mar 20 15:29:37 and js ;) Mar 20 15:30:05 av500: there is, but kids discover almost everything Mar 20 15:30:56 wait until they discovered the shell ;) Mar 20 15:31:35 wait until they discover the hiden link to the missile launch system Mar 20 15:31:48 the sw is actually very nice to use. Mar 20 15:32:28 av500: it's a modified labview Mar 20 15:32:44 does it support IEE$((? Mar 20 15:32:47 does it support IEE488? Mar 20 15:33:07 trouble with the shift key? Mar 20 15:33:14 av500: 485? Mar 20 15:33:16 needs oiling Mar 20 15:33:40 ah.. gbip.. Mar 20 15:34:05 from back when hp was cool Mar 20 15:36:04 like ti-calculators Mar 20 15:36:05 ;) Mar 20 15:36:08 that's what you get when you place an incompetent women at the helm Mar 20 15:36:24 I'm out of cards! Mar 20 15:36:34 * panto hands mdp a creeper card Mar 20 15:36:41 it's my last one! use it well Mar 20 15:36:42 * mdp hands the card to KotH Mar 20 15:36:56 thanks! Mar 20 15:37:06 i slowly start to get a collection :) Mar 20 15:37:23 when will U-Boot be done? Will it be done before or after the kernel is done? Mar 20 15:37:28 mdp: but seriously... it could have been a man as well, it just was a woman this time Mar 20 15:37:39 3008 Mar 20 15:37:51 * mdp notes on a yellow sticky Mar 20 15:38:57 * panto facepalms Mar 20 15:39:15 read -r -t 2 != read -r -t2 Mar 20 15:39:21 wtf bash people? Mar 20 15:39:35 bash them Mar 20 15:40:17 historic stuff, you need a software archeologist Mar 20 15:40:48 panto: need a hug? Mar 20 15:40:49 or how they called correctly ;) Mar 20 15:42:33 aholler: sometimes a programmer at arms would be more usefull Mar 20 15:43:30 I prefer peaceful people Mar 20 15:43:45 you come to the wrong place Mar 20 15:43:59 * KotH throws mranostay at aholler Mar 20 15:44:02 definitly Mar 20 15:44:45 ouch Mar 20 15:50:14 http://jeelabs.net/projects/cafe/wiki/Electricity_consumption_meter Mar 20 15:52:27 koen: I'm looking at starting a nightly autobuilder of the next-gen flasher. Mar 20 15:52:44 anything you'd like to see in the script or particular place/script you think I should use? Mar 20 15:52:58 jkridner, next-gen flasher? Mar 20 15:53:00 what is that "next-gen flasher"? Mar 20 15:53:17 what colour is this flasher? Mar 20 15:53:18 goretex raincoats? Mar 20 15:53:21 10kW consumption??? Mar 20 15:53:31 whatever. Black flasher. please. Mar 20 15:53:33 that's more than 10 times than what i use Mar 20 15:54:10 KotH: dunno Mar 20 15:54:16 maybe a US household Mar 20 15:54:23 but then, the wiring is way to clean for the US Mar 20 15:54:28 hehe Mar 20 15:54:39 a 100" tv in every room uses a fair bit of power Mar 20 15:54:58 we only have 50" tvs in the bathrooms Mar 20 15:55:03 jkridner: I still don't know what you are talking about Mar 20 15:55:04 quit exagerrating Mar 20 15:55:12 so does cooling every room down to freezing Mar 20 15:55:15 jkridner: I only know of the uSD card during burn in Mar 20 15:55:27 Burn ALL THE THINGS Mar 20 15:55:35 WITH FIRE! Mar 20 15:55:43 get rid of the screen saver Mar 20 15:56:03 use a black one Mar 20 15:56:27 ethically saved screen Mar 20 15:56:40 av500: my comment on that: never let a software guy design hardware Mar 20 15:56:49 koen: yes, that one. Mar 20 15:57:06 because it initializes boards for people doing testing. Mar 20 15:58:42 KotH, there's corollary, never let a h/w guy dictate s/w Mar 20 15:58:47 I shall dual-wield soldering irons and it shall be glorious. Mar 20 15:58:55 panto: +1 :) Mar 20 15:59:40 panto: as i'm usually the one who does the software for the hw i design... i suffer from my own mistakes :) Mar 20 15:59:54 KotH: punch yourself Mar 20 15:59:55 never let a middleman handle your s/w requirements Mar 20 16:00:06 Never X your Ys Mar 20 16:00:12 mdp: outsource to india! Mar 20 16:00:13 do that with your wapon too Mar 20 16:00:32 mdp: but who will take the requirements from the customer and give them to the engineers? Mar 20 16:00:43 jkridner: you know what I'm going to say and I know you will only half follow the recommendation and then bitch and moan that it doesn't work because you skipped steps Mar 20 16:00:44 alan_o: FedEx? Mar 20 16:00:48 alan_o, we have a people person for that ;) Mar 20 16:00:49 * av500 talks to the subcontractor of the subcontractor of our chinese subcontrator Mar 20 16:00:49 alan_o: /dev/null? Mar 20 16:00:51 fun Mar 20 16:01:08 alan_o, WHAT THE HELL IS WRONG WITH YOU PEOPLE? Mar 20 16:01:16 koen: jkridner: cant you two love birds get your own channel? Mar 20 16:01:18 I'm trying to check ahead of time about some of the typical steps I would skip. Mar 20 16:01:19 Well, that escalated quickly. Mar 20 16:01:29 I'd expect to 'bitbake cloud9-gnome-image'... Mar 20 16:01:45 but, I don't know if that produces a 2GB eMMC image... though it might... Mar 20 16:01:50 mdp: empty vessels ;) Mar 20 16:02:01 and, even if it does, then I likely need to boot it and dd it back to a file... Mar 20 16:02:01 mdp: I HAVE PEOPLE SKILLS!!!! Mar 20 16:02:36 but if you've looked at automating it in anyway, then it would be helpful. Mar 20 16:02:37 http://www.youtube.com/watch?v=sCC_PxRWVI4 for anybody that forgot Mar 20 16:04:04 Somebody's got a case of the Mondays! Mar 20 16:06:12 koen: right now, I only get a rootfs .tar.(bz2|xz) Mar 20 16:06:27 correct Mar 20 16:06:35 you need root privs to make the sd card Mar 20 16:06:47 * jkridner looks at IMAGE_FSTYPES Mar 20 16:06:54 ah. :( Mar 20 16:07:04 so, are you doing anything automated? Mar 20 16:07:51 anything after the tar.xz is a bunch of scripts, moving uSD cards, pressing reset, etc Mar 20 16:14:10 jkridner: the risk of blowing away your HD is too big, kpartx' 'show me the loop dev' is mutually exclusive with 'use a loopdev' Mar 20 16:14:38 jkridner: but doing a nightly of the cloud9 image can't hurt Mar 20 16:15:18 you could check the size and type of the storage Mar 20 16:15:26 at least the size is easy Mar 20 16:15:57 aholler: it could be a netbook, less storage than the bone :) Mar 20 16:16:07 thats true Mar 20 16:16:17 these days the boundaries between tool and toy are gone Mar 20 16:16:29 but those are secured by evil drm Mar 20 16:16:39 ;) Mar 20 16:17:22 av500: No wireless. Less space than a nomad. Lame. Mar 20 16:18:59 +1 Mar 20 16:20:19 add a cc25xx and a cloud Mar 20 16:21:15 wtf is MDIO anyway? Mar 20 16:21:59 only usable at 12 o'clock Mar 20 16:23:24 (MidDay) Mar 20 16:24:07 av500, I suppose we aren't thinking of the same MDIO Mar 20 16:26:01 koen: k, guess I'll just have to learn to deal with the rootfs.tar. have you thought about having the flasher work with the tarball? I guess I'm seeing writing to a file system be slower, but not sure if you've tried anything different. Mar 20 16:26:28 koen, i need to leave now, but i wanted to point you to http://git.pengutronix.de/?p=genimage.git . we've written that for ptxdist to generate (sd/disk) images with multiple partitions without beeing root/loopdev. I've also used it with OE before. Mar 20 16:26:52 if i have some time tomorrow i can point you to beginnings of a recipe Mar 20 16:28:01 koen: given that I use virtual machines on EC2, I'm not really worried about blowing away my HDD. Mar 20 16:28:05 for the AM335x we use it to build a SD image with FAT (MLO, barebox, uImage, oftree, root.ubi for NAND) and ext4 rootfs Mar 20 16:28:16 shoragan: thanks, that solves part of the problem, the other is mkfs.ext4 Mar 20 16:28:36 thanks shoragan Mar 20 16:28:57 koen, jkridner, we use gene2fs+tune2fs+fsck.ext4 Mar 20 16:29:00 works fine Mar 20 16:29:15 shoragan: that is not an ext4 image Mar 20 16:29:31 shoragan: only newly created files will use extents, etc Mar 20 16:29:57 koen, yes, but thats a minor evil to requiring root Mar 20 16:30:04 at least for us Mar 20 16:30:18 see http://lists.linuxtogo.org/pipermail/openembedded-core/2012-June/023604.html Mar 20 16:31:08 Darren Hart worked on a pushing something like genext4fs into ext2fs progs Mar 20 16:31:10 and including a existing minimal image that then gets resize2fs'd or sthg like that is probably not 'nice' enough? Mar 20 16:31:14 but that is still ongoing Mar 20 16:31:41 jkridner: tarball would work, it isn't that much faster though Mar 20 16:32:08 koen, when you get to IMA/selinux labels, the only way seems to be to run the image once in qemu anyway :/ Mar 20 16:32:20 jkridner: I don't have much time to spend on the flasher, though Mar 20 16:32:24 is it faster at all? I guess it would have first-time boot stuff, but during burn-in it could run. Mar 20 16:32:26 k. Mar 20 16:32:44 jkridner: but I did find out that you don't actually have to boot it, you can systemd-nspawn into it Mar 20 16:32:47 koen: I'm just looking for ways to get everyone involved in testing to incorporate daily changes. Mar 20 16:32:47 sry, i've got to run. see you. Mar 20 16:33:40 jkridner: getting people to test it and report back would be a big improvement already Mar 20 16:34:10 * av500 can test stuff that requires no thinking Mar 20 16:35:16 jkridner: did you notice the tiny boris logo you get when fedora automounts the g_multi partition? Mar 20 16:37:03 nah, using OS X on a daily basis for my desktop. Linux only on servers at-the-moment. Mar 20 16:37:52 on the list of billions of things to do to look at changing the Bone desktop background. Mar 20 16:48:43 emerge -1pv bluez Mar 20 16:48:45 uups Mar 20 16:49:45 demerge Mar 20 16:49:56 dement ;) Mar 20 16:50:25 * KotH wonders whether there is an emerge equivalent of apt-get moo Mar 20 16:54:23 there was, but it it disappeared: http://www.eeggs.com/images/items/927.full.jpg Mar 20 16:55:32 looks like larry likes debian better than gentoo Mar 20 16:57:58 there even was a discussion about the type of cow. gentoo's cow was nicier than debians Mar 20 16:58:31 did it moo more melodically ? Mar 20 16:58:47 * mranostay mooos Mar 20 16:58:55 it was more rounded Mar 20 17:00:07 the debian bug about that should still exist, maybe already closed ;) Mar 20 17:00:22 you are kidding right? Mar 20 17:00:35 no Mar 20 17:00:57 * mranostay facepalms Mar 20 17:02:14 mranostay: geeks, you know? geeks Mar 20 17:13:59 phew, got the damn thing to link Mar 20 17:14:03 now if only it would run Mar 20 17:14:43 mru: you must crawl^H^H^Hlink before you can walk^H^H^H^Hrun Mar 20 17:15:03 hush: can't execute './avconv': Cannot allocate memory Mar 20 17:16:00 mru: that seems wrong Mar 20 17:16:09 of course it's wrong Mar 20 17:16:12 that's what I'm saying Mar 20 17:16:23 malloc(100GB)? Mar 20 17:16:30 it's the exec that fails Mar 20 17:16:38 uclinux Mar 20 17:16:42 yes Mar 20 17:16:42 out of buffers Mar 20 17:16:52 we need a bigger boat Mar 20 17:17:54 any suggestions for finding out _what_ allocation is failing? Mar 20 17:19:12 http://blackfin.uclinux.org/gf/project/uclinux-dist/tracker/?action=TrackerItemEdit&tracker_item_id=6940 Mar 20 17:19:19 happens on BFIN too Mar 20 17:19:58 yes, after a while memory always gets fragmented Mar 20 17:20:07 but this is on a freshly booted system Mar 20 17:20:14 did you turn it off and on again? Mar 20 17:20:16 oh Mar 20 17:20:29 and the gcc-built version runs Mar 20 17:20:34 sizes are similar Mar 20 17:20:43 ah Mar 20 17:20:50 use gcc then Mar 20 17:20:56 fuck you Mar 20 17:21:19 well now children Mar 20 17:33:08 bradfa: interns? :) Mar 20 18:12:19 exactsteps.com Mar 20 18:13:33 seems that is taken Mar 20 18:13:54 Crofton|work: welcome to the new day Mar 20 18:14:32 The domain exactsteps.com may be for sale by its owner! Mar 20 18:27:45 * ogra_ points to #pandaboard ... someone looks for exactsteps there Mar 20 18:28:09 (or a body guide ... its not really clear) Mar 20 18:40:45 pandaboard is dead, jim Mar 20 18:42:35 tell that to my build machines :) Mar 20 18:45:10 how dead? Mar 20 18:46:36 web seems to be alive Mar 20 18:46:39 I think it's beyond resting Mar 20 18:46:48 * ynezz is confused Mar 20 18:47:35 zombiepanda Mar 20 18:47:40 ynezz: it was a faux community project and TI fired all the people involved Mar 20 18:49:51 zombieboard! Mar 20 18:49:59 ++ Mar 20 18:50:08 * mranostay whois's Mar 20 18:50:44 mranostay, sadly, no Mar 20 18:51:00 koen: would a zombie board qualify for animalboard.org? Mar 20 18:51:13 heh, .de exists Mar 20 18:51:17 mranostay, was an "open source" project we chose to use, and by we I mean someone who doesn't write code picked it Mar 20 18:51:45 ah that is a mru domain as well Mar 20 18:52:01 koen: d'oh! Mar 20 18:53:21 mranostay: I've seen your g+ picture of that usb 480 protocol analyzer and I wonder if it's usable under Linux Mar 20 18:54:34 all the totalphase equipment has linux support Mar 20 18:55:01 well, I don't own any of their stuff Mar 20 18:55:09 yes Mar 20 18:55:33 I don't own any either, have used it in the past Mar 20 18:55:36 well a greek guy will be able to review it better Mar 20 18:56:28 anyway it looks like there's not much other options Mar 20 18:57:07 av500: er how do you have hawkboard.org? Mar 20 18:57:17 Hmmmm Mar 20 18:57:19 that opensizla/viszla has been kickarted like 2 years ago, but nothing yet :) Mar 20 18:57:22 er eagleboard.org rather Mar 20 18:57:35 where is the eagleboard?!!?!??!!?!?!? Mar 20 18:57:36 :D Mar 20 18:58:09 with the dragons Mar 20 18:58:22 "This is not xkcd" Mar 20 19:04:04 velociraptors? Mar 20 19:05:43 ynezz: the beagl480 is really worth it Mar 20 19:06:05 ynezz: if you are writing your own usb stack or have to debug usb hardware, there is nothing but the beagle 5000 that beats it Mar 20 19:06:39 it's required for working with musb or omap ehci Mar 20 19:06:40 ynezz: software works superb on linux (much better than on windows, but that's not totalphase fault) Mar 20 19:07:06 * KotH would not have been able to write an usb stack w/o it Mar 20 19:07:20 KotH: is that 10.41 times better? Mar 20 19:07:27 especially considering the faults in the documentation and the silicon Mar 20 19:07:41 mranostay: i havent tested it yet, so i cannot say Mar 20 19:07:51 fwiw, the spi and i2c analyzers are great too Mar 20 19:07:54 mranostay: but i assume it to be at least 5 times better Mar 20 19:08:13 wow, docs _and_ si Mar 20 19:08:22 mdp: what is their advantage over a general logic analyser? Mar 20 19:08:28 usually blaming one is enough Mar 20 19:08:32 Mmmmmm beagle Mar 20 19:08:34 none by today's standards Mar 20 19:08:47 mru: and none of the bugs i reported ended up in any errata Mar 20 19:08:53 mru: after 4 years Mar 20 19:08:59 for a lot of things, the beagle12 gets you pretty far Mar 20 19:09:00 they just have had linux host s/w for ages now...before sigrok and ols existed etc. Mar 20 19:09:16 in the meantime they made a new generation of their chips and i guess, the old bugs are still there (at least the docu bugs are) Mar 20 19:09:22 mdp: sigrok supports the beagle analyzers now? Mar 20 19:09:26 not a lot of tools in that category were like that but it has been their differentiator Mar 20 19:09:42 ds2: The software has a lot of limitations when running with the beagle12 Mar 20 19:09:42 ds2: i don't think it does Mar 20 19:10:02 ds2, I'm implying that before the hobbyist priced LA h/w and sigrok project...this was it Mar 20 19:10:03 for, example, they turn off class-level decoding (among other things) when running with a 12 Mar 20 19:10:14 sorta like bdi2000 before openocd Mar 20 19:10:18 alan_o: nothing absolutely missing except for maybe HS Mar 20 19:11:18 ds2: and that can be the crucial piece to figure out what is wrong with your software Mar 20 19:11:21 or hardware Mar 20 19:11:22 or both Mar 20 19:11:25 I'm not sure exactly what you mean..... I mean they turn off stuff in the software (PC software) that shouldn't have anything to do with hardware. (but they do it anyway, to get you to buy the 480). Mar 20 19:11:55 alan_o, value-add Mar 20 19:12:01 yeah..... Mar 20 19:12:04 efused Mar 20 19:12:12 aholler: it's not even efused Mar 20 19:12:21 it's turned off in the PC software. Mar 20 19:12:22 sfused Mar 20 19:12:23 alan_o:anything besides some decodes? Mar 20 19:12:34 ds2: trying to remember Mar 20 19:12:50 but yeah, it's all stuff you _can_ get by without Mar 20 19:13:04 of course one _can_ also get by without an analyzer :) Mar 20 19:13:20 donno about a total lack of analyzer Mar 20 19:13:26 debugers are for whimps ;) Mar 20 19:13:28 the main one I ran into has to do with decodes Mar 20 19:13:36 alan_o, if their salary is low enough, by all means Mar 20 19:13:37 ds2: another advantage of the beagle480 are the 4 digital inputs Mar 20 19:13:44 but that feature wasn't in there when I initially got it Mar 20 19:13:44 test equipment is for overpriced people Mar 20 19:13:47 ds2: helps you to debug timing related bugs in the system Mar 20 19:14:23 KotH: is that the 480 or the new 5000? Mar 20 19:14:28 ds2: 480 Mar 20 19:14:56 alan_o, your :) is duly noted...otherwise I'd call you !sane Mar 20 19:15:39 KotH: good to know. been thinking about upgrading from the at the next project. mine date back to pre480 times Mar 20 19:15:58 mdp: there's of course the difference between can and want-to :) Mar 20 19:16:01 ds2: as i said, the 480 is totally worth it Mar 20 19:16:11 ds2: wouldnt want to miss it anywhere where usb is involved Mar 20 19:16:28 I wish they made pcie analyzers :( Mar 20 19:16:30 KotH: but why not go with the latest - the 5000? Mar 20 19:16:33 ds2: especially if it's hardware from a manufacturer that stars with a and ends with tmel Mar 20 19:17:00 ds2: got the 480 in pre-5000 times, havent had a project yet that would justify the 5000 Mar 20 19:17:16 odd request.. dont' suppose anyone knows of a package that contains 4 indicator leds, + curent limiting resistors Mar 20 19:17:23 KotH: I mainly use it to see WtF with baremetal guys sending stuff up to Linux drivers that they claim is 'good' Mar 20 19:17:31 just looking at reducing placing costs, if i can replace 8 parts with 1 Mar 20 19:17:33 its a good saving Mar 20 19:17:51 ds2: eheh Mar 20 19:19:10 mrpackethead: "a package"? Mar 20 19:19:19 yeah, all in one part Mar 20 19:19:23 alan_o, btw, if you don't like their product value schemes you should start an OSHW project to replace it ;) Mar 20 19:19:33 maybe being idealisitic Mar 20 19:19:39 mrpackethead: you can buy leds with integrated current limiters Mar 20 19:19:47 mrpackethead: but that'd be single leds Mar 20 19:19:58 mrpackethead: and about 10 times the price of a single led and a resistor Mar 20 19:20:04 i can buy a resistor network Mar 20 19:20:09 alan_o, preferably not kickstarter BS Mar 20 19:20:12 thats likely what i'll do Mar 20 19:20:19 mrpackethead: why would you need a package? Mar 20 19:20:27 to reduce placement costs Mar 20 19:20:32 o_0 Mar 20 19:20:34 1 part vs 8 parts Mar 20 19:20:43 0_o Mar 20 19:20:43 all the 0.08c's add up Mar 20 19:20:45 :-) Mar 20 19:20:49 you must have millions of leds if that is an issue Mar 20 19:21:14 the optimisations end up being tens of thousands of dollars Mar 20 19:21:24 how many do you produce? 10k/y? Mar 20 19:21:39 100k+ Mar 20 19:21:39 thats why there exists hw without an rtc because the crystal is expensive Mar 20 19:21:49 mrpackethead: ok.. then it's might be worthwhile :) Mar 20 19:22:15 have optimised 46 resistors on a project Mar 20 19:22:26 aholler: and often because a crystal is way too power hungry Mar 20 19:22:27 it has an expected value of almost $80,000 pa to me Mar 20 19:22:35 for 4 hours work Mar 20 19:22:39 nice Mar 20 19:23:03 provided the sales team stick to their plan! Mar 20 19:23:08 mranostay: that's large volume production for you Mar 20 19:23:21 mranostay: you start hunting each and every rappen :) Mar 20 19:23:52 we once spend a week squeezing out 0.10usd out of a device Mar 20 19:23:57 heh Mar 20 19:24:20 $80k = half an engineer for a year Mar 20 19:24:23 KotH: sure, crystals are the reason for the global warming Mar 20 19:24:25 payback was in the range of 100k in the first year :) Mar 20 19:24:52 without them, every hw would suddenly require much less power Mar 20 19:24:55 aholler: if your device has to run from a small coin cell for over a year, then every uA counts Mar 20 19:25:16 aholler: and a crystall needs more than an RC oscillator, by a factor of 2-10 Mar 20 19:25:42 do you power stuff like the bb with a coin cell? impressive Mar 20 19:25:58 not every device we do runs linux ;) Mar 20 19:26:24 and not everything needs a rtc Mar 20 19:26:45 but every uC needs an oscillator to work Mar 20 19:27:05 mdp: hehe, I'm not complaining (well, not complaining _that_ hard), but I was confused about the software limitations when I bought it (shame on me, for assuming). I just don't want others to run into the same thing. Mar 20 19:27:27 fair enough Mar 20 19:27:55 alan_o: have you complained to totalphase? Mar 20 19:28:00 indeed crystals are the reason for global warming Mar 20 19:28:08 alan_o: they are usually quite responsive to such stuff Mar 20 19:28:11 all that entropy it must cause somewhere else... Mar 20 19:28:50 alan_o: when did you get it? from my point of view, it is more of extra features they added to other models rather then software limitations Mar 20 19:28:51 KotH: there's nothing to complain about. it's just their model. They're selling software (features) as part of the upgraded hardware price. Mar 20 19:29:09 alan_o: i would still mention that Mar 20 19:29:12 ds2: I got it 15 months ago Mar 20 19:29:19 alan_o: it might prompt them to change their model :) Mar 20 19:29:24 how come when i make an order from mouser, there is always somethign else you need. Mar 20 19:29:36 ^^^ Mar 20 19:29:42 alan_o: but on the other hand, given the fact that you get free software upgrades with your hardware... Mar 20 19:30:00 alan_o: ah... that explains it. i got it before the 480 existed so all those features are bonuses in their free sw Mar 20 19:30:24 not sue what you mean Mar 20 19:30:36 alan_o: and from the first time i used the 480 to the last time a year or two ago, the software evolved quite a bit Mar 20 19:30:37 bonuses only when you have a 480 attached Mar 20 19:31:09 those features didn't existed before the 480 was introduced Mar 20 19:31:18 ah. Mar 20 19:31:26 so I bought it w/o even considering those features (it wasn't available on any model) Mar 20 19:31:35 oh ok. Mar 20 19:31:56 whereas you got it much later so it looked like it was crippled Mar 20 19:32:01 yes Mar 20 19:32:14 like I said, I'm not complaining (very hard). It's just something I got tripped up on, that others in the market should be aware of. Mar 20 19:32:45 It'd be cool if we had a driver for it that could hook into usbmon Mar 20 19:32:51 and we could analyze from wireshark Mar 20 19:34:45 sigrok driver Mar 20 19:35:08 then output usbmon format that wireshark/tcpdump grok Mar 20 19:35:10 pun intended Mar 20 19:36:07 can be implemented with the current low/full speed protocol decoder independent of a more capable analyzer Mar 20 19:36:36 well writing that stuff shouldnt be a problem Mar 20 19:36:45 KotH: it's just software Mar 20 19:36:47 you can export the captured data and process it Mar 20 19:36:59 * KotH did a graphical timing visualization with perl once Mar 20 19:37:31 alan_o, +1 !!! Mar 20 19:37:57 better to have it within an existing de facto standard signal gathering framework Mar 20 19:39:39 mdp: what sigrok hw works well with the USB 2.0 HS signaling? massive scope captures? Mar 20 19:39:52 I said low/full Mar 20 19:40:17 but HS is the in more interesting one Mar 20 19:40:17 two separate thoughts Mar 20 19:40:32 what works well today with LS/FS? Mar 20 19:40:55 1) low/full decoding exists now in sigrok with commodity h/w. leverage that and write a new output format for usbmon protocol that tcpdump/wireshark digest Mar 20 19:41:13 2) implement hardware driver for a hs capable device (and decoder) Mar 20 19:41:56 3) win lotto Mar 20 19:42:01 ds2, iirc from list stuff...it was the Vu unit Mar 20 19:42:03 ... Mar 20 19:42:08 *) profit Mar 20 19:42:24 ds2, vague recollection Mar 20 19:42:39 mdp: then people demand you give it away for free Mar 20 19:42:50 scratch the profit part Mar 20 19:42:51 mdp: then they demand that you support them indefinitely Mar 20 19:43:06 mdp: I think we're back to 3) Win Lotto Mar 20 19:43:11 * KotH demands free swiss chocolate for everyone! Mar 20 19:43:25 alan_o, that's been the fun of FOSS forever ;) people demanding you support your free software for free on their schedule Mar 20 19:43:28 KotH: yeah!!! yes we can! yes we can!! Mar 20 19:43:38 hehe Mar 20 19:46:15 <_av500_> gm Mar 20 19:51:22 mdp: freedom isn't free Mar 20 19:51:25 * mranostay ducks Mar 20 19:51:34 <_av500_> freeducks? Mar 20 19:51:40 it's filled with morons, yes Mar 20 19:58:44 The median age of an American home is 36 years old. <- is that true? Mar 20 20:01:00 <_av500_> the home? Mar 20 20:01:07 the age Mar 20 20:01:12 sounds reasonable. they don't build houses like europeans do Mar 20 20:01:13 <_av500_> the ture Mar 20 20:01:16 <_av500_> true Mar 20 20:01:59 <_av500_> also they blow up their homes in action movies all the time Mar 20 20:02:37 * NishanthMenon wonders if his house is old enough for action movies :P Mar 20 20:02:46 * KotH does not think he has ever lived in a house that was younger than 30 years, not younger than 20 for sure Mar 20 20:03:45 * _av500_ does Mar 20 20:04:08 build one yourself? Mar 20 20:07:34 <_av500_> with my bear hands Mar 20 20:08:05 ah.. living in a hole you dug yourself Mar 20 20:08:16 <_av500_> yes, I call if "life" Mar 20 20:08:25 <_av500_> it* Mar 20 20:08:45 there is no life in IT ;) Mar 20 20:10:16 <_av500_> true Mar 20 20:10:29 * mru has (a share in) a house in sweden that's ~250 years old Mar 20 20:12:31 <_av500_> mru: I hope it does not have an 8bit age.... Mar 20 20:12:40 heh Mar 20 20:13:05 * KotH would also have a share in an old house, if the bulgarians wouldnt have "aquired" the house from our family Mar 20 20:13:06 I guess we'll find out soon enough Mar 20 20:13:07 hopefully there will be never an hurrican or tornado there Mar 20 20:14:01 like in day after tomorrow? Mar 20 20:14:13 not that soon Mar 20 20:14:33 it was allegedly built in 1760 Mar 20 20:14:57 it's an alleged house Mar 20 20:15:00 and C14 said it was probably 1860? :) Mar 20 20:15:00 mdp: that's older than your entire _country_ Mar 20 20:15:25 old and washed up Mar 20 20:15:25 and colder Mar 20 20:15:29 KotH: there's no reason to doubt it Mar 20 20:15:53 * KotH does not Mar 20 20:16:08 they always kept good records in sweden Mar 20 20:16:28 having been in buildings build before anyone knew there was another continent on the other side of the big pond... Mar 20 20:16:51 plenty of those in europe Mar 20 20:16:57 yeah Mar 20 20:17:15 KotH: you mean anybody except the people who were already there, right? Mar 20 20:17:19 being in a 700 year old turkish bath is kind of nice :) Mar 20 20:17:36 alan_o: they didnt know they were on the other side of the big pond ;) Mar 20 20:17:43 ust that it wasn't turkish then Mar 20 20:18:11 aholler: go back to your history books ;) Mar 20 20:18:21 lots of old roman stuff too if you look around Mar 20 20:18:34 KotH: Welcome to America, now speak Cherokee :) Mar 20 20:18:46 alan_o: *g* Mar 20 20:19:06 mru: and sometimes where people dont want to look ;) Mar 20 20:19:21 mru: like when digging to build a new high rise... Mar 20 20:22:29 KotH: so you bath was build by osman? Mar 20 20:22:33 <_av500_> roman stuff everywhere Mar 20 20:23:09 aholler: turkish people have been in anatolia since at least 1100 Mar 20 20:23:23 and they called themself turks Mar 20 20:23:33 yes Mar 20 20:24:07 <_av500_> bloody romand littered the whole world with theirr streets and bridges and central heating Mar 20 20:24:19 the turkish family of people is quite broad and old Mar 20 20:24:31 even the uzbeks call themselves "turks" Mar 20 20:24:46 and you left them Mar 20 20:24:49 why? Mar 20 20:25:26 av500: Romanes eunt domus! Mar 20 20:25:26 <_av500_> hmmm Mar 20 20:25:30 well... the usual hunting grounds of the turkish tribes in the west were suddenly surrounded by a large wall Mar 20 20:25:44 that made it necessary to go east to look for greener pastures Mar 20 20:25:48 <_av500_> firewall? Mar 20 20:25:58 <_av500_> or videowall? Mar 20 20:26:02 chinese wall Mar 20 20:26:27 <_av500_> chinese mall Mar 20 20:36:46 <_av500_> hmm Mar 20 20:36:53 <_av500_> XDM API now has multicore support Mar 20 20:36:58 <_av500_> makes sense Mar 20 20:37:25 something makes sense? Mar 20 20:37:27 that's a first Mar 20 20:37:53 <_av500_> well, I was asking myself how codecs on the 6678 use several cores Mar 20 20:38:04 <_av500_> since XDAIS does not define spawning threads Mar 20 20:38:31 <_av500_> I'm looking at the h264 decoder, will try to get that running in linux Mar 20 20:38:42 <_av500_> write some tiny wrapper that calls the XDAIS stuff Mar 20 20:41:16 <_av500_> ah there is a test app Mar 20 20:41:17 <_av500_> for sysbios of course Mar 20 20:42:59 * KotH should someday study the old turkish writing systems Mar 20 20:44:12 new meaning to chicken scratch? Mar 20 20:44:51 getting more meanings out of old stuff Mar 20 20:45:17 and maybe i get a clue why so many words are similar in turkish and japanese Mar 20 20:45:18 <_av500_> on multicore, a simple "i" is not enough: for (ii = 0; ii < IH264_MAXCORES; ii++) Mar 20 20:45:39 lol Mar 20 20:45:43 that should be ok on dual-core Mar 20 20:45:53 for your octocore you need iiiiiiii Mar 20 20:46:08 or is that logarithmic so iii is enough? Mar 20 20:47:49 <_av500_> i8 maybe Mar 20 20:47:52 <_av500_> or i6i Mar 20 20:59:13 hmpf.. looks like i have to learn russian in order to learn about old turkish languages... Mar 20 20:59:22 russian is hard Mar 20 20:59:41 yeah.. Mar 20 20:59:48 oh.. there a few german texts as well.. Mar 20 21:00:05 german is reasonable Mar 20 21:00:19 yeah.. especially if you already know it :) Mar 20 21:00:41 <_av500_> Copyright (C) 1996, MPEG Software Simulation Group. Mar 20 21:00:44 <_av500_> wow, old code Mar 20 21:00:48 KotH: well, so is russian Mar 20 21:03:02 _av500_: hmmm does that mean anything mentioned in there is either patent expired or unpatentable now? Mar 20 21:03:23 no, patents are forever Mar 20 21:04:38 <_av500_> ds2: its some TI dsp malloc code Mar 20 21:04:50 <_av500_> nothing really mpeg specific inside Mar 20 21:05:34 blah Mar 20 21:06:04 <_av500_> I wish TII would use consistent code formatting Mar 20 21:06:30 try to keep your wishes at least somewhat realistic Mar 20 21:06:36 you'll be less disappointed that way Mar 20 21:07:02 <_av500_> well, nothing indent cannot fix Mar 20 21:07:20 indent ;) Mar 20 21:07:37 <_av500_> I normalized a lot of indian code drops that way Mar 20 21:07:41 <_av500_> to find the 3 lines that changed Mar 20 21:08:15 <_av500_> besides each file date and "version" having been changed of course Mar 20 21:08:37 what about the pointless renames? Mar 20 21:10:10 <_av500_> that happened less Mar 20 21:14:27 jkridner: I think you helped me this weekend to build the 3.8 kernel for my beagle bone. Unfortunately I didn't get to building it that night, but i have another question. You mentioned I should copy over all the dtb files to the boot directory, do I actually need them all or just the following: am335x-bone.dtb Mar 20 21:14:58 and am335x-evm.dtb Mar 20 21:15:42 just am335x-bone.dtb Mar 20 21:16:03 And i take it the boneblack is for the next gen bb Mar 20 21:19:10 jg_: we cannot confirm or deny that sir Mar 20 21:19:35 fair enough :) Mar 20 21:19:40 but yes :) Mar 20 21:19:42 the black just means it's cloaked in secrecy Mar 20 21:20:10 mdp: black project of sorts? Mar 20 21:20:26 just one more question: do I need the uImage-dtb.am335x-bone too, or is that the uImage with the am335x-bone.dtb included? Mar 20 21:21:27 154 - 109 parts by optimising Mar 20 21:21:35 that is a postive win. Mar 20 21:24:10 <_av500_> wow, google made a shopping list app Mar 20 21:24:53 *yawn* Mar 20 21:25:27 <_av500_> I MUST HAVE IT Mar 20 21:26:07 * mranostay backs away from _av500_ Mar 20 21:26:31 throws NAND chips in front of _av500_ and unleashs ds2 Mar 20 21:26:32 * _av500_ keeps mranostay close Mar 20 21:31:54 <_av500_> there, I installed it - ARE YOU HAPPY NOW GOOGLE? Mar 20 21:44:38 what did you buy yet ? Mar 20 21:45:29 <_av500_> nothing Mar 20 21:46:15 why do you need a shopping list then ? Mar 20 21:46:28 <_av500_> becuase its from GOOGLE Mar 20 21:46:32 <_av500_> and its AWESOME Mar 20 21:46:33 <_av500_> man Mar 20 21:47:03 hah Mar 20 22:00:04 _av500_: you must be confused Mar 20 22:00:16 only apple fanboys say shit like that Mar 20 22:00:39 that was before there was a google shoppinglist app Mar 20 22:01:00 back in the days ... Mar 20 22:06:45 <_av500_> yeah Mar 20 22:06:49 <_av500_> we was blind Mar 20 22:06:52 <_av500_> now we see the light Mar 20 22:14:10 jkridner: Thanks again. Just got the kernel running (had a few hiccups but it's up) :D Mar 20 22:15:32 Anyone online, I have probably a beginner question how to start my beaglebone which I got today. The first thing I did was to download: http://downloads.angstrom-distribution.org/demo/beaglebone/archive/Angstrom-Cloud9-IDE-eglibc-ipk-v2012.01-core-beaglebone-2012.01.11.img.gz then I followed the instructions from http://downloads.angstrom-distribution.org/demo/beaglebone/ how to preformat the microSD card in Linux, I ejected it, plugged Mar 20 22:15:41 mru: i am not a apple fanboy Mar 20 22:15:49 i just own a lot of their stuff :) Mar 20 22:16:09 Then I inserted the network cable and power supply. Mar 20 22:16:17 contradiction alert!! Mar 20 22:17:17 so that brings us to now, with your completely hairless but soot covered face? :D Mar 20 22:17:46 I waited a few secs and I found out the DHCP gave it an IP address and when I tried to get onto the beaglebone IP address from the web browser chrome, nothing happens, when I try ssh: ssh: connect to host 192.168.1.102 port 22: Connection refused Mar 20 22:18:11 Could someone help me please? Mar 20 22:18:20 refa where is your serial console? Mar 20 22:18:37 nmap the range! Mar 20 22:19:26 woglinde, I didn't connect it with the USB cable Mar 20 22:20:27 you should Mar 20 22:20:37 otherwise use a crystall ball Mar 20 22:20:50 woglinde thanks for the hint I will give it a try now Mar 20 22:21:02 or just light a few of those black, ball shaped candles with a fuse on it :D Mar 20 22:21:44 mranostay, I tried nmap and it said the following: Nmap scan report for 192.168.1.102 Host is up (0.0019s latency). MAC Address: 00:18:31:8B:56:14 (Texas Instruments) Mar 20 22:22:23 <_av500_> do you see the leds blink? Mar 20 22:22:35 <_av500_> there should be at least one heartbeat led Mar 20 22:22:43 <_av500_> blinking merrily Mar 20 22:22:45 and is it counting down to zero? Mar 20 22:22:45 Yes Mar 20 22:22:53 <_av500_> well, then linux is up Mar 20 22:23:34 _av500_, I see the led0 blink fast two times, then a break and then again the same process repeated Mar 20 22:26:06 o.O (serial console) Mar 20 22:26:34 woglinde, I plugged in the USB cable, pressed the reset button Mar 20 22:27:03 woglinde, it produced a log output about the modules which were loaded and now I can log in Mar 20 22:28:10 fine than you will find out why sshd is not running Mar 20 22:28:26 woglinde, I entered: root and then enter Mar 20 22:29:40 Ooo, new stuff went up on http://www.goldmine-elec.com/ Mar 20 22:29:41 you do not have to tell me Mar 20 22:29:47 I know how to login Mar 20 22:30:16 :) Mar 20 22:33:51 woglinde, do you get the following in output as well? Mar 20 22:33:53 Uncompressing Linux... done, booting the kernel. [ 0.065817] omap_l3_smx omap_l3_smx.0: couldn't find resource [ 0.224020] _omap_mux_get_by_name: Could not find signal leds-gpio [ 0.289281] omap2_set_init_voltage: unable to get clk dpll1_ck [ 0.295379] omap2_set_init_voltage: unable to set vdd_mpu_iva [ 0.301412] omap2_set_init_voltage: unable to get clk l3_ick [ 0.307314] omap2_set_init_voltage: unable to set vdd_core Mar 20 22:34:05 sd Mar 20 22:34:34 pastein Mar 20 22:34:37 ups Mar 20 22:34:42 use pastebin please Mar 20 22:34:52 stop spamming the channel Mar 20 22:35:08 woglinde, sorry, it wasn't my intention to spam it Mar 20 22:35:43 can I send you privately the log output? Mar 20 22:35:49 no Mar 20 22:36:01 ok :( Mar 20 22:36:10 what should me the kernel output say Mar 20 22:36:15 your system is running Mar 20 22:36:27 check whats with sshd Mar 20 22:37:26 ps -A | grep sshd gives me the following message: Unknown HZ value! (87) Assume 100. Mar 20 22:38:12 if I type it, it is not found at all sshd: command not found Mar 20 22:39:12 sure because dropbear is installed Mar 20 22:39:31 or it is really none sshd installed Mar 20 22:39:37 than install one Mar 20 22:39:42 with opkg Mar 20 22:40:34 Thanks woglinde for the hint Mar 20 22:47:36 woglinde, once I tried opkg install sshd, I got the following response: Package openssh-ssh (5.9p1-r4) installed in root is up to date. I tried to type then once again sshd but it said command not found :/ Mar 20 22:49:35 /etc/init.d/sshd start Mar 20 22:49:49 I tried now with opkg update and then the same but it returned wget: server returned error: HTTP/1.1 404 Not Found Mar 20 22:50:48 woglinde, just tried that out: -sh: /etc/init.d/sshd: No such file or directory Mar 20 22:51:22 I tried to ping google.com and it works, so it is online Mar 20 22:53:06 maybe I should try with another version of angstrom? Mar 20 22:54:51 maybee its openssh Mar 20 22:55:03 have you ever used a linux before? Mar 20 22:55:48 yes, but ubuntu Mar 20 22:56:28 things haven't been that hard to set up :) but this is a chance to learn even more :) Mar 20 22:59:43 hi prpplague Mar 20 23:00:12 panto: up late? Mar 20 23:00:19 it's all purple! Mar 20 23:00:19 yeah Mar 20 23:11:17 woglinde, thanks for your help, it is late here, so I will go to bed, will continue to play with the beaglebone tomorrow :) Mar 20 23:11:48 bye to the rest of community, cya all Mar 20 23:16:57 quit Mar 20 23:17:01 exit Mar 20 23:18:06 I have a 1ma current source attached to one of the analog inputs. Do I need to worry about blowing ain if it's attached to a 3.3v rail, but limited to 1ma? Mar 20 23:18:45 lol. Mar 20 23:19:18 got an email from CCO Mar 20 23:19:22 only 6 weeks late Mar 20 23:19:26 we'ld like to quote on your job Mar 20 23:19:38 im sorry, we've almost finished the job now Mar 20 23:22:19 mrpackethead? Mar 20 23:22:58 yes. Mar 20 23:23:22 anyqay Mar 20 23:23:26 good nite Mar 20 23:55:07 depends on the compliance of the current source Mar 20 23:56:38 why would that matter? Mar 20 23:56:57 the question would be, can whatever is attached to the ain pin, that usually blows up, sink 1mA of current? Mar 20 23:57:08 safely. Mar 20 23:58:48 with^Hwithout fire Mar 20 23:58:56 er, ^W. but yes Mar 20 23:59:02 no idea :/ Mar 21 00:02:34 i'll go the safe way then. found a diode with a forward bias knee where i need it. :D Mar 21 00:38:06 KotH: Coplanar Waveguide. Mar 21 00:38:11 KotH: It just came to me. Mar 21 01:26:29 Looking for an alternative to 595-LMV358IDGKR Mar 21 01:29:04 heh ok Mar 21 01:34:57 evening all. trying to use an xm board for a class but forums just have me reading in circles Mar 21 01:40:50 stop reading circle buffers Mar 21 01:52:39 +Circle Buffers Mar 21 02:43:19 hey trolls Mar 21 02:45:22 trollface Mar 21 02:45:55 new and updated version, ala KotH: http://i37.photobucket.com/albums/e94/MasterMWL/GPS_Cape_zps4e2a5794.png Mar 21 02:46:36 er, huh. the renderer erally didn't like the 45-degree-angle resistor at R18 Mar 21 02:50:15 seems a lot of effort to hang a gps that would have been fine using tx, rx, +, - or and i2c bus + power Mar 21 02:50:29 but it looks pretty, i give it a gold star Mar 21 02:50:41 Thanks :) Mar 21 02:51:13 how is it interfacing the gps? spi, i2c or serial Mar 21 02:51:22 There. http://i37.photobucket.com/albums/e94/MasterMWL/GPS_Cape_zps965dc530.png Mar 21 02:51:42 MrCurious: I2C and/or Serial Mar 21 02:51:54 i tried i2c with my ublox neo and couldnt get it to make sense Mar 21 02:52:09 got any i2c working code? if so, i would love to peek erm, i mean blantly steal Mar 21 02:52:10 along with antenna detection Mar 21 02:52:18 not yet :x Mar 21 02:52:25 ok Mar 21 02:52:25 Checked out the protocol document? Mar 21 02:52:40 i thought i did, but i am open to looking at other docs Mar 21 02:52:51 i can see the device id, just not interpret the data Mar 21 02:53:02 http://www.u-blox.com/images/downloads/Product_Docs/u-blox6_ReceiverDescriptionProtocolSpec_%28GPS.G6-SW-10018%29.pdf ? Mar 21 02:53:07 thinking its in binary format when i want it in more of a nmea format like serial Mar 21 02:53:27 Yeah, you're not going to be able to pull a stream of nmea like that over i2c ;) Mar 21 02:53:39 though you could poll the location registers in a loop Mar 21 02:54:39 MrCurious: I put the optional EEPROM on there to let it save ephemeris projections, and gave it a massive battery Mar 21 02:55:08 ok, mine has an itty battery, and no i2c eeprom Mar 21 02:55:38 So you probably need to pull the CFG pins high/low on boot, then, to select the proper serial streams Mar 21 02:55:56 i bet i am just lacking understanding on the i2c protocol. have only done accelerometers gyrom and magnetometer before from a 10dof board Mar 21 02:55:58 ie NMEA vs UBX protocols, and speeds Mar 21 02:56:11 MrCurious: Admittedly, I'm not that experienced with i2c either Mar 21 02:56:48 Protocol in 0+1+2 – UBX+NMEA+RTCM Mar 21 02:56:48 Protocol out 0+1 – UBX+NMEA Mar 21 02:56:53 yup... that looks the key Mar 21 02:56:57 ahh no more thermal vias Mar 21 02:57:07 mrpackethead: yeah :) I figured out how to unthermal them Mar 21 02:57:07 i bet its in ubx and i am clueless about that and want nmea Mar 21 02:57:41 stilll lots of vias Mar 21 02:58:12 my pre-prod test boards fired up Mar 21 02:58:15 which is great Mar 21 02:58:16 :-) Mar 21 02:58:26 whee Mar 21 02:58:37 MrCurious: http://www.ublox.com/images/downloads/Product_Docs/NEO-6_DataSheet_%28GPS.G6-HW-09005%29.pdf page 11 Mar 21 02:58:41 section 1.15.1 **** ENDING LOGGING AT Thu Mar 21 02:59:58 2013