**** BEGIN LOGGING AT Wed Aug 13 02:59:58 2014 Aug 13 03:55:02 anyone know i2c interfacing for beaglebone, im not able to see device im trying to connect too.. Aug 13 08:19:27 tawr: I did that for the original beagleboard Aug 13 08:20:57 the bug itself won't draw much, so just pay attention to how much you drain through devices you connect Aug 13 08:21:06 bug? hub! Aug 13 08:25:41 Hi leiflindholm Aug 13 08:25:53 thank you very much for getting back to me. I've been sitting here trying to figure out if I should risk it or not Aug 13 08:26:26 The hub came with a 5v 3A power supply (it's a 4 port RadioShack I got on clearance for $2.99) Aug 13 08:28:26 5V 3A should be okay Aug 13 08:36:27 woglinde, Aug 13 08:37:00 What I was trying to do, was to cut the wire from the 5v 3A suppy, and make a "Y" cable, with 2 male barreljacks. One barrel jack powering the BBB, and one barrel jack powering the powered USB Hub attached to the BBB Aug 13 08:37:41 But I was unsure if having the same supply powering the BBB and it's powered hub may cause issues of the magic smoke variety -- I didn't want to run 2 AC adapters within 3 inches of each other (one to the BBB, one to the hub) Aug 13 08:45:00 y cable should be fine Aug 13 08:56:44 I built a separate y-cable instead Aug 13 08:57:04 the copper in those power supplies tend to be really thin and fiddly Aug 13 08:59:06 leiflindholm, what do you mean a separate y cable? Aug 13 09:00:47 tawr: a female connector to two males Aug 13 09:01:23 as opposed to cutting the power supply cable Aug 13 09:01:55 Gotcha. This thing had one of those "mini" 1.5mm inner pins, it's a tiny connector Aug 13 09:02:31 I don't have a female/male laying around to do a "real" adapter, although I'd like to truthfully. maybe i'll run back to ratshack and see if they have a male/female pair to do it "right", though I have no doubt I'd be able to solder it Aug 13 09:06:34 Hello, what is the difference between BB-BBLK-000 and BBONE-BLACK-4G? Aug 13 09:06:52 I see, that second one has 4gb flash included Aug 13 09:06:58 theres a 4G bbb? :O Aug 13 09:07:03 what about parameters, chips? same revision? Aug 13 09:07:04 nvm Aug 13 09:07:07 day: revC Aug 13 09:07:26 older ones had 2G eMMC Aug 13 09:07:35 the one im using right now :p Aug 13 09:07:54 I don't expect that there are any left for sale, proably only 4G ones reaching the market Aug 13 09:08:40 2G is to small. Well its not to small. Its just that the official images suck Aug 13 09:08:51 too* Aug 13 09:08:58 well Aug 13 09:09:00 1.9GB debian iirc :P Aug 13 09:09:12 from my understanding rev C (4gb) started shipping in May Aug 13 09:09:25 and they've all been out of stock pretty much everywhere Aug 13 09:10:06 where "out of stock" - selling faster than new stock arrives Aug 13 09:10:38 yeah - very high demand for them. the topic talks about it slightly Aug 13 09:10:45 adafruit got a batch in, and I snagged one the other day Aug 13 09:11:25 for most tasks the smaller one is perfectly fine tho Aug 13 09:11:33 i have 900MB free Aug 13 09:11:59 the debian image contains a whole desktop environment etc. Aug 13 09:12:08 debian is not exactly an embedded distro Aug 13 09:12:11 yeah complete bullshit :D Aug 13 09:12:22 so, 4G is never one? I'm confused, because i was compairing parameters on farnell site, and I have noticed, that on 4G model is, Silicon Core Number: AM3358, on 000 model we have Silicon Core Number: AM3359 Aug 13 09:12:25 i wanted it because of packages Aug 13 09:12:29 day: opinions are bound to differ between users ;) Aug 13 09:12:32 so in general AM3359 is never Aug 13 09:12:36 apt-get and done :p Aug 13 09:12:43 day: plus, there's a console only image out there nowadays too. Aug 13 09:13:00 LetoThe2nd: for the 2gb one? Aug 13 09:13:01 on a fresh debian flash, and a apt-get update / upgrade Aug 13 09:13:02 Bogis: BBB is BBB, whatever you buy is 99,9% going to be the latest hardware revision Aug 13 09:13:03 i'm right at 49% used Aug 13 09:13:23 day: yes. i'm just not sure if its already linked on bb.org Aug 13 09:13:50 ok, one more thing: Aug 13 09:14:12 hm i looked a few weeks ago and couldnt find it. i found one on a private page iirc. Aug 13 09:14:16 BB gives power output only 3,3V? Aug 13 09:14:33 theres a 5V header iirc Aug 13 09:14:34 day: http://elinux.org/BeagleBoardDebian#eMMC:_BeagleBone_Black Aug 13 09:14:40 day: mentions it, for example. Aug 13 09:14:41 logic outputs are 3.3 iirc Aug 13 09:17:32 well Aug 13 09:17:48 Bogis: you might want to read the SRM, it's linked on the website Aug 13 09:17:49 i bought one last week (came in today), i have the am3358 processor, but it's rev c 4gb Aug 13 09:18:46 3358 vs 3359 is a change that's almost impossible to spot IIRC Aug 13 09:19:26 tbr: hey, its "ONE MORE"! Aug 13 09:19:55 * tbr demands one that goes to ELEVEN! Aug 13 09:26:24 i bet the 3359 does 2Ghz on ln2 Aug 13 09:27:28 the beaglebone could really be a bit faster though :/ Aug 13 09:27:40 it can barely keep my irc bouncer alive Aug 13 09:27:44 that woudl be a beagle-cat Aug 13 09:27:50 cats are faster than dogs Aug 13 09:28:01 i dont think so :p Aug 13 09:28:37 but cuter. and more evil :3 Aug 13 09:31:08 best situation ive witnessed so far was. our cat on the outside of the window. 50kg black hunting dog on the inside. Barely separated by 1cm of glass. Dog totally losing it....Cat starts yawning and cleaning herself while touching the window :D Aug 13 09:31:57 rip little kitty :( Aug 13 13:36:04 Hi ! Aug 13 13:39:41 Quick question: what's the proper way to use my own uboot on a BBB? Aug 13 13:42:41 Ecco: compile, put on sd card, and force BBB to always use that one by placing a 10K-ish resistor P8-43 - P9-46 Aug 13 13:42:58 Ecco: mainline uboot should give you fitting spl and u-boot binaries. Aug 13 13:43:06 Thanks a lot Aug 13 13:44:06 I'm trying to use this: http://elinux.org/U-boot_musb_gadget_support Aug 13 13:44:13 LetoThe2nd: what with the resistor? Aug 13 13:44:55 (I've reordered a USB-TTL cable, but in the meantime I'd like to play with custom kernels and I don't see any other way to get kernel logs) Aug 13 13:45:39 is that just a SYSBOOT thing? Aug 13 13:45:39 Other question: how should I format the SD card and name the bootloader? I couldn't find any docs regarding the early boot sequence of the BBB Aug 13 13:46:23 Ecco: read the TRM for the SoC and also the SRM for the BBB Aug 13 13:46:38 ok Aug 13 13:46:39 thanks Aug 13 13:46:39 also look at how sd-card images are structured Aug 13 13:46:51 will do. Thanks for the pointers Aug 13 13:46:59 yeah, thats the easy way. grab an existing image and just replace the files. Aug 13 13:47:17 the usual™ way is to put the MLO and U-Boot on a small FAT partition first in order on the device Aug 13 13:47:24 I can do that. You don't get the warm fuzzy feeling of handmaking it all though :-) Aug 13 13:48:03 if it tickles your fancy you can put the MLO also at certain offsets, regardless of partitioning Aug 13 13:48:37 if you want to play with MLO/U-Boot, then UART is pretty much a hard requirement Aug 13 13:48:56 tbr: it seemed to want to read u-boot from fixed offser too if you do that Aug 13 13:49:01 offset Aug 13 13:49:17 leiflindholm: yeah, quite likely Aug 13 13:49:25 * tbr has _never_ tried it Aug 13 13:49:47 I think the Archos devices that I own do that though, but they had their own loaders Aug 13 13:49:49 tbr: I summarised my experiments, and the ugly patch I used to work around that in http://blog.eciton.net/uefi/bbb_raw_sd_boot.html Aug 13 13:49:56 \o/ Aug 13 13:50:01 So basically I'm out of luck trying to run my own u-boot without a real serial port? Aug 13 13:50:29 Ecco: well, if you are super lucky it will boot. but there is a really good chance it just won't Aug 13 13:50:32 Ecco: https://github.com/ungureanuvladvictor/BBBlfs check this, you can usb boot whatever MLO/u-boot you want Aug 13 13:50:43 so no more sd-card magic or whatever Aug 13 13:50:50 oh, interesting Aug 13 13:51:01 if you need help using it ping me, I need testers :) Aug 13 13:51:20 Thanks Aug 13 13:51:26 i developed it for flashing boards but can be used for just bootloading also Aug 13 13:51:30 Oh so the SoC supports booting off of USB? Aug 13 13:51:45 yes, it simulates an RNDIS interface over usb Aug 13 13:51:49 Nice Aug 13 13:51:53 and expects some bootp magic to be sent to it Aug 13 13:51:58 And this is in the ROM of the SoC, right? Aug 13 13:51:59 chech the AM335x TRM Aug 13 13:52:05 yes, but not what people think "booting from usb" means, at least not what most people think. Aug 13 13:52:06 yes, this you annot change Aug 13 13:52:13 excellent Aug 13 13:52:19 s/annot/cannot Aug 13 13:52:25 yeah sure so it's more like netboot-over-usb Aug 13 13:52:30 yep Aug 13 13:52:43 as in you tftp-in your bootloader and get going Aug 13 13:52:56 yep Aug 13 13:53:09 sweet Aug 13 13:53:13 you can also do it the regular way with a tftp server and and so Aug 13 13:53:14 I have to get my hands on the right doc Aug 13 13:53:24 but i took matters in my own hands with libusb Aug 13 13:53:46 that's pretty sweet Aug 13 13:54:36 that wouldn't let me get kernel logs though, would it? Aug 13 13:54:45 you do not have serial cable ? Aug 13 13:54:47 I mean, this way I can send a modified uboot Aug 13 13:54:51 Not yet :( Aug 13 13:54:56 no :( Aug 13 13:55:09 I could use my logic analyzer, but that's somehow overkill Aug 13 13:55:17 plus it only reads from the bus Aug 13 13:55:19 you can just make your rootfs and put ssh server there Aug 13 13:55:22 yeah Aug 13 13:55:25 and then just ssh into it Aug 13 13:55:30 check how I generate the FIT file Aug 13 13:55:31 Ecco: Got an Arduino or similar around? Aug 13 13:55:43 agmlego: Well yeah, I have a few AVRs lying around Aug 13 13:55:45 i am making a basic busybox with nothing on it Aug 13 13:55:52 I guess I could build my own USB<->ttl interface with it Aug 13 13:55:56 Ecco: Actually, this time I meant Arduino proper. Aug 13 13:56:02 Ecco: Like, the dev board. Aug 13 13:56:08 agmlego: oh, well, then no Aug 13 13:56:11 Ah. Aug 13 13:56:23 Because those can also be used as a USB-TTL interface. Aug 13 13:56:25 Well, my USB-TTL cable shouldn't take too long Aug 13 13:56:55 But actually being able to log into the BBB over USB was the reason I got this board instead of a RPi Aug 13 13:57:20 (which already works wonderfully once the Linux kernel has booted) Aug 13 13:59:23 (my understanding being that the Linux kernel emulates an FTDI chip on the USB-otg port and exposes a serial console on top of it) Aug 13 13:59:33 Ecco: it uses g_multi Aug 13 13:59:54 when it boots Aug 13 14:00:09 and offers ethernet and massstaorage on the usb bus to the host Aug 13 14:00:46 and serial as well, right? Aug 13 14:01:01 It's nice to see all this Aug 13 14:01:05 don't think about serial but u can tweak the script to do it Aug 13 14:01:08 lemme find it :) Aug 13 14:01:16 Oh it does serial too Aug 13 14:01:23 Ecco: THis is the beaglebone black? Aug 13 14:01:26 Yes Aug 13 14:01:35 https://github.com/RobertCNelson/tools/blob/master/scripts/beaglebone-black-g-ether-load.sh Aug 13 14:01:46 IIRC, out of the box, the only serial availble is the FTDI header. Aug 13 14:01:55 agmlego: same here Aug 13 14:01:58 The white one had serial on the USB port. Aug 13 14:02:05 ok you guys lost me :) Aug 13 14:02:28 Ecco: after you ssh to the board check if you can do sudo modprobe g_serial Aug 13 14:02:36 sure Aug 13 14:02:45 but I don't even ssh into the board Aug 13 14:02:46 does bbb rev c come pre installed with opencv? Aug 13 14:02:50 I minicom into it :-) Aug 13 14:02:57 doh_: lemme see Aug 13 14:03:13 root@beaglebone:/home/debian# modprobe g_serial Aug 13 14:03:13 ERROR: could not insert 'g_serial': No such device Aug 13 14:03:33 doh_: yes it does Aug 13 14:03:45 nice! thanks Ecco! Aug 13 14:03:46 (mine did at least. It came with Debian 7.4 and includes OpenCV) Aug 13 14:03:54 Ecco: http://hipstercircuits.com/serial-over-usb-on-beaglebone/ Aug 13 14:04:06 doh_: You're welcome. Note that it's super easy to install if OpenCV is missing. Aug 13 14:04:26 oh, easy to install? how? Aug 13 14:04:48 apt-get install maybe :) Aug 13 14:04:57 doh_: The standard Debian way (if you're on debian). Something along the lines of "apt-get install opencv-dev" Aug 13 14:05:16 ("aptitude search opencv" would help you find the actual package name) Aug 13 14:05:41 Or apt-cache search opencv Aug 13 14:05:54 vvu: Ok, wait. I thought the Beaglebone *black* had no hardware USB-to-serial chip (the FTDI one mentionned in this article) Aug 13 14:06:03 Ecco: thanks, i haven't bought bbb yet, so i assume it will be preinstalled if it's alrready on yours Aug 13 14:06:07 Ecco: Correct. Aug 13 14:06:22 doh_: It should be :-) Aug 13 14:06:33 Ecco: yes, it does not have but you emulate it using g_serial Aug 13 14:06:42 So this article (http://hipstercircuits.com/serial-over-usb-on-beaglebone/) doesn't apply to the beaglebone black, does it? Aug 13 14:06:55 Ecco: So, for boot messages, you need the FTDI serial header, not the USB port. Aug 13 14:07:06 agmlego: Ok, that was my understanding so far Aug 13 14:07:08 Ecco: the kernel can emulate a serial device Aug 13 14:07:15 tbr: Yeah, I got this running already :-) Aug 13 14:07:28 I'm logging into my BBB via an USB port Aug 13 14:07:34 but there is no dedicated ftdi chip for serial and jtag like on the BBW Aug 13 14:07:41 Once the kernel is running, it works perfectly Aug 13 14:08:21 But I'm trying to build a custom kernel Aug 13 14:08:33 And I'm afraid I won't be able to get the early logs with this technique Aug 13 14:08:36 we're back at UART, yes :) Aug 13 14:08:47 Ecco: like this you cannot get logs before kernel boot Aug 13 14:08:50 That's what I thought :-) Aug 13 14:09:09 *but* I've seen a special feature in u-boot that exposes a console over USB too Aug 13 14:09:18 you _could_ redirect the console to the gs0 device, but output won't show up before kernel has booted enough AND it's been configured Aug 13 14:10:10 also it's dangerous as you can shoot yourself in the foot. the kernel will crash hard once the console buffer is full, IIRC Aug 13 14:10:18 Ooh, interesting Aug 13 14:10:30 And regarding that uboot feature Aug 13 14:10:50 (http://elinux.org/U-boot_musb_gadget_support) Aug 13 14:11:32 it is for some old omaps as Google says, never heard that for the BBB Aug 13 14:11:52 that is for beagleboard not beaglebone Aug 13 14:11:55 oh Aug 13 14:12:23 maybe it works even in the latest u-boot for BBB, at least rndis gadget works Aug 13 14:12:28 so g_serial has a chance Aug 13 14:12:32 go to #u-boot and ask Aug 13 14:12:39 I'm gonna Aug 13 14:12:49 Is there a way to just try it? Aug 13 14:13:07 I mean, boot the "stock" u-boot but prevent it from starting the linux kernel afterwards? Aug 13 14:13:57 It's too bad they removed the hardware USB-TTL converter though Aug 13 14:14:01 that's handy :) Aug 13 14:14:44 ftdi cables are much cheaper than putting the chip on the board though AFAIU Aug 13 14:15:06 You have a point Aug 13 14:15:38 but one could argue that people buying that kind of device are more likely to do some low-level stuff with it Aug 13 14:16:00 (I mean one could just buy a random cheap-ass Android phone instead) Aug 13 14:16:12 so including an FTDI chip could make sense Aug 13 14:16:23 well, I guess that was thought over though since the first version had it Aug 13 14:16:48 Ecco: http://git.denx.de/cgi-bin/gitweb.cgi?p=u-boot.git;a=blob;f=include/configs/am335x_evm.h;h=35ae0e6fb76b5fd976f486f593f20ba21d290560;hb=HEAD#l289 Aug 13 14:16:52 plus like you said, USB-TTL cables are 1.80 € on Amazon, *including* shipping (I wonder how that's even possible) Aug 13 14:17:09 Ecco: have no clue if g_serial support is there Aug 13 14:17:16 the first version had a complicated expensive chip as it also had JTAG and somehow muxed in the SoC USB Aug 13 14:17:18 BTW, I got my hands on the Sitara TRM. Thanks a lot for pointing me to it Aug 13 14:17:19 but there is some gadget support Aug 13 14:17:38 Ecco: look at the boot sequence and sysboot pins Aug 13 14:18:02 WHOA, page 4096. Damn you don't want to print that thing :-D Aug 13 14:18:14 Ecco: Yes, but the FTDI chip is ~$3/1000, and takes board area and placement time and testing time and another slot on the pick and place. Costs add up, especailly when you are driving the price of a board down. Aug 13 14:18:32 Makes sense Aug 13 14:18:57 It was probably one of the first, fast cost optimisations done to the board. Aug 13 14:19:23 Plus, it gives you a convenient logic-level serial port *not* on the headers, which can be useful. Aug 13 14:19:34 Indeed Aug 13 14:20:01 How comes the chip is $3 when they sell cables at €1.8? Aug 13 14:20:14 mass purchase ? Aug 13 14:20:17 What is the Euro-USD conversion these days? Aug 13 14:20:19 (Well, it's a PL2303-based cable, but still) Aug 13 14:20:24 There you go. Aug 13 14:20:24 Ecco: different chip Aug 13 14:20:28 1.3smthg last time I checked Aug 13 14:20:31 PL2303 is literally worthless. Aug 13 14:20:58 I would rather use a cable packed withy nothing but beach sand--it would work more consistently. Aug 13 14:21:07 Is it really that bad? Aug 13 14:21:10 Crap I just bought one :-D Aug 13 14:21:39 Yes. Aug 13 14:21:53 :-D Aug 13 14:21:55 But as long as you are not expecting much performance out of it, you should be fine. Aug 13 14:22:08 ok, I'll keep that in mind Aug 13 14:22:43 Like, the hardware flow control is defective. Speeds higher than 38400 are defective. Power consumption is like 8x the FTDI. Parity is broken. Aug 13 14:22:54 :-/ Aug 13 14:22:56 Software flow control does not always get through. Aug 13 14:23:19 Word lengths different from 8 bits sometimes work and sometimes do not. Aug 13 14:23:40 Ecco: THis is why everyone on the planet uses FTDI for USB-TTL--theirs suck a lot less. Aug 13 14:23:50 ok :) Aug 13 14:24:00 And there's a driver for it in recent OS X releases :) Aug 13 14:24:15 And everything else for a decade ago. ;-P Aug 13 14:24:19 :-D Aug 13 14:24:45 Well it's true that drivers isn't where OS X shines… Aug 13 14:25:36 ok Aug 13 14:25:42 Amazon hasn't shipped my cable yet Aug 13 14:25:51 lemme see if they have anything with an FTDI chip Aug 13 14:26:21 Meh, i've seen this one often recommended : http://www.adafruit.com/products/954 Aug 13 14:26:26 uses a PL2303 as well… Aug 13 14:27:39 ::shrugs:: It is your time and hair. Aug 13 14:27:46 I use something like this one: https://www.adafruit.com/products/70 Aug 13 14:28:04 And the myriad versions of the FTDI Friend / USB-BUB laying about the hackerspace. Aug 13 14:31:30 WHat about CP2102? Aug 13 14:33:23 I have not had the length and breadth of time wasted and frustration with the CP2102 chipset as I have with the PL2303, simply because both of the CP2102 devices I have ever touched failed to work at all. Aug 13 14:33:32 :-D Aug 13 14:33:41 Take that as you will. Aug 13 14:33:46 ok FT232RL it is then Aug 13 14:34:05 found one for $6 Aug 13 14:34:13 My philosophy is to buy good test equipment, because then you can safely assume that the faults you are seeing in the hardware you are debugging are not just the shitty test equipment. Aug 13 14:34:22 Others have different ideas. Aug 13 14:34:25 That makes an awful lot of sense Aug 13 14:34:43 The way I see it is "how much of my time is worth saving $2" Aug 13 14:34:50 Yup, that too. Aug 13 14:34:53 Fortunately, not so much :) Aug 13 14:35:46 ok, so back on the uboot topic Aug 13 14:36:04 apparently uboot supports USB CDC ACM too Aug 13 14:44:09 I understand that no power should be applied to the GPIO on the BBB Rev C. until the SYS_RESETn signal is HI. I have a loop powered sensor that is applying voltage at all times. With that said what are some recommendations on IC's to disallow power unitl the SYS_RESETn signal is HI? Aug 13 14:44:17 I am trying to flash the eMMC on my beaglebone black. It has been over 45 minutes and it hasn't signaled that it is complete. How do I know if it is actually flashing? Aug 13 14:44:45 Paul_10: Does the sensor have a reset line? Aug 13 14:44:51 bobonthenet: How are you flashing? Aug 13 14:45:17 agmlego: No the sensor does no have a reset line. It is on all of the time. Aug 13 14:45:29 Ecco: I have an sd card with the debian image I got from the beagleboard.org website. Aug 13 14:45:46 ok. How are you doing the MMC flashing? Aug 13 14:45:58 Paul_10: Then youy probably want to put FET(s) on the pins connected to the bone, and have your program turn the FETs on when it starts. Aug 13 14:47:40 Okay good idea. What would the limitation be on the number of FETs that the SYS_RESETn pin could drive? So bascially I am asking how much current it can supply. Aug 13 14:48:08 I would nopt drive it off the SYS_RESETn line. Aug 13 14:48:32 Ecco: I am holding down the user boot button while the board is plugged in. I release the button when the 4 lights that are on at the same time go off. Aug 13 14:48:37 But since FETs are not driven by current, it would probably be OK to have a few on it if you must use that pin. Aug 13 14:48:37 https://learn.adafruit.com/beaglebone-black-installing-operating-systems/flashing-the-beaglebone-black Aug 13 14:49:04 I'm providing power thru P1 Aug 13 14:49:13 Oh, ok. Then I guess you won't have any progress indication this way Aug 13 14:50:36 It has been on for over an hour. Do you think it is going to finish or should I try again? Aug 13 14:50:42 : So instead of using the SYS_RESETn pin, I should just drive the FET's off a GPIO after SYS_RESETn is Hi? Aug 13 14:50:53 bobonthenet: I have no idea… Aug 13 14:51:18 I'll try again. Aug 13 14:51:27 ok :) Aug 13 14:53:02 Well at least it gets enumerated Aug 13 14:53:13 Paul_10: I would just drive the GPIO you choose when the program you are using to read the sensors starts up--just set those GPIO pins high as part of the initilization routine. Aug 13 14:53:28 When I boot my BBB without any SD card and bypassing the eMMC Aug 13 14:53:35 it gets enumerated as an "AM335x USB" Aug 13 14:53:43 (on the host PC it's connected to) Aug 13 14:55:28 agmlego: Thank you for the information. Aug 13 14:58:14 Paul_10: No problem. Aug 13 14:58:33 Oh, I forgot to thank you guys as well Aug 13 14:58:37 thanks a lot for all your help!! Aug 13 14:59:31 Ecco: anytime Aug 13 15:00:03 be careful, I might take your word for it :-p Aug 13 15:35:23 has anyone in here had any problems with the i2c bus not waiting for a slave device? Aug 13 15:36:07 i am struggling to interface a adxl345 which will answer correctly for like 3 reads of its data registers and just send bullsh... for the next reads Aug 13 15:36:43 if i use sleep() it wont help. if i close and open the i2c bus it works Aug 13 15:37:19 my theory is that the bbb is not waiting for the device to be ready and keeps reading its own buffer Aug 13 15:37:23 but im not too sure about this Aug 13 15:37:37 does that have an interrupt signal? Aug 13 15:39:00 tbh i implemented my reads using the i2c-dev.h included in the ubuntu distribution. i did not expect to have to check for interrupts myself ... Aug 13 15:39:53 the adxl itself stores 32 values in its own fifo Aug 13 15:40:10 so if it would not be able to sample fast enough this should happen after 32 reads ... Aug 13 15:41:07 (it itself has some kind of an interrupt system. i can kinda controll it using register writes, but its still some kind of a blackbox Aug 13 15:42:41 theese interrupts inside of the adxl tell the chip when to read new data Aug 13 15:42:51 currently its set to "8 of theese 32 values have been read" Aug 13 15:43:03 when to sample new data* Aug 13 16:00:39 Oh gosh! I'm a noob, I was trying to flash the non flashing version. Aug 13 16:06:52 bobonthenet: you can use the non-flashing version to flash the eMMC... Aug 13 16:06:57 you just need to run a script on it. Aug 13 16:07:18 in fact, there is only a single text file difference between the two images, though I don't recall the details. Aug 13 16:11:29 jkridner: for some reason it wasn't working so I'm trying the other one. I'll report back what happens. Aug 13 16:23:47 hey , anybody have faced random reboot in beaglebone white Aug 13 16:26:05 hi Aug 13 16:26:43 somebody know how to use uarts on beaglebone black on linux Aug 13 16:33:47 <_av500_> google does Aug 13 16:33:50 debian-installer should support serial console installs of BeagleBone Black out of the box, though may require some manual u-boot tweaking: https://lists.debian.org/debian-devel-announce/2014/08/msg00005.html Aug 13 16:34:03 for installing jessie Aug 13 16:35:13 * vagrantc needs to get in sync with rcn regarding u-boot patches... Aug 13 16:44:21 vagrantc: all the patches are mostly regarding the bootcmd including kernel location and uenv.txt locations, right? Aug 13 16:44:46 vagrantc: I think we should be getting close to something Debian friendly and that could be pushed upstream. Aug 13 16:46:25 jkridner: yeah, basically uenv.txt, kernel locations, etc. Aug 13 16:46:36 jkridner: if you get it pushed to u-boot, i'll have no choice but to merge it :) Aug 13 16:47:20 jkridner: also, in the near future, it'll need to support bootscripts, rather than uenv.txt, at least for it to be integrated "the Debian way" with flash-kernel and all that. Aug 13 16:48:43 i've only briefly glanced at rcn's extensive u-boot env patches, although they looked ok to me for the most part. Aug 13 16:49:31 not sure if there's been any major updates since mid-july or so Aug 13 16:54:05 just a few small fixes.. If the july u-boot worked, then the current one will to.. ;) Aug 13 17:14:23 hi rcn-ee... Aug 13 17:14:37 I'm still drafting a kernel location blog post and e-mail. :( Aug 13 17:15:11 rcn-ee: I got side tracked trying to announce an easy way to boot new kernels to test using vvu's BBBlfs. Aug 13 17:16:24 i saw you changed the default branch on bb.org/linux so i pushed a newer v3.14.x the am335x will suspend. ;) Aug 13 17:17:06 yeah, I'm hoping all of this encourages testing/patching. Aug 13 17:18:06 Most users (well as long as they are using one of the later then july testing releases) will be able to apt-get install that kernel too. ;) Aug 13 17:20:23 right now i'm debugging my lcd4.. it works on v3.16.x.. Otherwise spi & uart's are working on v3.14.x for cape people. Aug 13 17:21:40 trying to keep things simple: https://github.com/beagleboard/linux/blob/3.14/arch/arm/boot/dts/am335x-boneblack.dts#L30 (just uncomment what you want. ;)) Aug 13 17:22:23 i fee like i decided to check out this channel at just the right time Aug 13 17:22:30 hides. ;) Aug 13 17:22:54 lol Aug 13 17:24:02 i'm excited to see a newer kernel out for the beaglebone black. I play with this thing everyday Aug 13 17:24:51 rcn-ee: is there a list of supported/non-supported functionalities of the new kernel? Aug 13 17:25:29 vmayoral|pc, yeah i need to start a wiki page to compare the two.. Aug 13 17:25:42 if your using a cape/overlay, there are 'workarounds'... Aug 13 17:26:12 rcn-ee: :) Aug 13 17:28:05 jkridner: rcn-ee thx for the announcement :) Aug 13 17:28:29 when i reach home i will modify stuff to just support bootloading and stuff like that...lack of hardware now is the main issue Aug 13 17:46:59 I am interested in recompiling the Debian kernel for the BBB to remove GPIO driver support (so that I can use an mmap without risk of conflict). Can anyone point me in the right direction about which kernel I should target and the build process? I am quite new to kernel compilation. Aug 13 17:59:57 sheridp: I am not sure what is in the debian kernel. If it is using device tree, you could just mark gpio entries with status = "disabled" in the DT to have them not instantiate. That said you probably want to try using sysfs entries for GPIO use. Aug 13 18:07:00 Hi jeol, thanks for the tip. Debian does use device trees, so I'll look into disabling the driver from there. The main problem with sysfs access is (1) speed and (2) I want to be able to read a whole 12 bit register at once, and as far as I know, sysfs only allows export of single pins (please let me know if I'm wrong--it would be so much easier) Aug 13 18:16:17 write a proper driver Aug 13 18:16:22 the registers are in the TRM Aug 13 18:16:59 mod the gpio driver to ignore a GPIO bank and then have at it Aug 13 18:17:20 but for things requiring a full bank like a parallel port, do look into GPMC instead Aug 13 18:19:01 Hi ds2, when you say mod the gpio driver, would you suggest doing this with a device tree overlay? Aug 13 18:22:10 No, modify the driver. Aug 13 18:22:16 device tree != driver. Aug 13 18:23:04 right, the problem is basically that the driver itself is compiled into the (debian) kernel, unless I am mistaken Aug 13 18:23:35 which is why I began by asking " Can anyone point me in the right direction about which kernel I should target and the build process? I am quite new to kernel compilation" Aug 13 18:24:16 that's a whole different problem that I can't help you in the next 2 minutes (have a lunch appt) Aug 13 18:24:26 it is debian Aug 13 18:24:40 and the kernel is GPL... Aug 13 18:25:22 now with angstrom...you can run the recipe for the kernel and... Aug 13 18:26:08 with debian, unless you've done it before, I wouldn't say it's trivial Aug 13 18:28:59 or apply the sysfs gpio block export patches to kernel Aug 13 18:46:09 Hi guys Aug 13 18:46:25 I need to configure the bbb to be a rs485 slave Aug 13 18:46:34 woudl you suggest any code/readings/documentation ? Aug 13 18:46:36 thank you Aug 13 18:48:29 oussemah: http://lmgtfy.com/?q=beaglebone+black+rs485 Aug 13 19:12:01 Xra, that doesn't really help; I already did that; am actually asking about the specifi case of configuring bbb as rs485 slave, not generic rs485 usage Aug 13 19:16:30 any beagleboard expert looking for a freelance project on beaglebone ? Aug 13 19:19:10 i have a beaglebone related project and willing to give good price , is there anybody ? Aug 13 19:24:47 Guest71889: there's probably a few around in various states of attentiveness, maybe a little more detail would draw some attention Aug 13 19:45:55 On the BBB Rev C., P9_11 is High on boot. I have a python script using the Adafruit GPIO library that sets P9_11 low. As soon as the script completes the P9_11 goes high again. Is there somewhere in the BBB where I can default this pin to Low? Aug 13 20:23:37 trying to connect beagle to a ftdi device, dmesg shows device recognized, but simple echo commands dont come back, anyone know how to test for connection? Aug 13 20:25:02 and you've shorted tx and rx? Aug 13 20:30:05 unfortunatly its a codec, i have no way of shorting tx and rx Aug 13 20:31:44 the codec uses a ftdi driver also, do you suppose if I monitor the usb data that goes down from windows where it gets a response and i use stty to send it from beagle i should get a response that way? Aug 13 20:32:22 also the codec program only shows baud rate, no stop bits parity etc, would ttyusb not have these either? Aug 13 20:32:52 im not sure if codec program isnt showing the parity stop bits info or its not needed at all Aug 13 20:33:34 basically im trying to send command and startup scrips that windows does and emulate it on linux Aug 13 20:33:44 scripts Aug 13 20:35:55 that's not really beagle specific then though Aug 13 20:37:22 im wanting to send the same usb ftdi based commands that come from config program on windows and do those commands thru stty over usb / ftdi, i just dont know if things listening Aug 13 20:39:47 if there is only a usb-to-serial endpoint then just sniffing the serial port data is enough to get an idea Aug 13 20:40:07 if there are other end points, then you might want to look into the whole usb bus activity Aug 13 21:16:34 crofton: you've ofcourse seen the classic: http://www.well.com/~cynsa/newbutt.html Aug 13 21:17:04 oops. wrong channel. Aug 13 21:53:38 anyone having difficult with hdmi? i have tried some stuff but only get 'cant open display' with xrandr Aug 13 21:54:47 on the bbb ? Aug 13 21:55:16 do you have more of the logs that you can post to pastebin? Aug 13 21:55:16 Err, notserpe, heh? Aug 13 21:55:18 yes, on the BBB Aug 13 21:55:51 rtmc, unless you are in an active x11 session: xrandr -display :0.0 -q Aug 13 21:56:49 ah... I made the assumption it was in X11 Aug 13 21:57:13 it just says "No protocol specified. Can't open display: 0.0" Aug 13 21:57:30 which means your in ssh/serial... use above ^ Aug 13 22:03:13 i think there is something missing with /var/run/gdm Aug 13 22:03:21 because i dont have "gdm" Aug 13 22:04:16 did you install gdm? Aug 13 22:05:09 err... no.. i just found out the gdm is the gnome, i will try install Aug 13 22:57:18 hmmmmm Aug 13 23:44:17 thanks, it worked. i updated to newest armstrong and added " video=HDMI-A-1:e" to the end of uEnv.txt file Aug 14 00:11:04 My beaglebone black keeps crashing. Is there a log file or something I can look at to see what it was doing right before it goes down? Aug 14 00:14:14 if you're in the debug console using ftdi it might give some indications Aug 14 00:15:35 If I have a beagle / basset mix in real life, do I get extra support in the channel? :P Aug 14 00:17:04 No. Aug 14 00:17:40 woof. Aug 14 02:23:22 hello, goodmorning guyz Aug 14 02:24:52 How can I deploy my qt4 application to beagleboneblack Aug 14 02:26:25 :D Aug 14 02:32:45 ok, been at this 2 days now. Got a lapdock for beaglebone black. Got some special cables made just for this project found here: https://specialcomp.com/beaglebone/ I plug everything up, bbb gets power, but no display Aug 14 02:32:55 no xorg is installed, I just want display for console first Aug 14 02:44:50 is it possible to expand the beaglebone black to support multiple ethernet ports? Aug 14 02:46:01 hdmi_force_hotplug=1, hdmi_drive=2, hdmi_mode=86, and disable_overscan=1 uncommented. Aug 14 02:46:20 i need your help.. for my beaglebone. Aug 14 02:46:24 these are some settings a rpi user had to set in a file labelled "config.txt" for a problem very similar to mine Aug 14 02:47:19 Jesperhead, I'm having a similar problem but with overscan issues (clipped monitor output) Aug 14 02:47:31 at least you get a display Aug 14 02:48:19 well, not that I can do anything with it Aug 14 02:48:24 I can draw boxes on the desktop.. that's about it Aug 14 02:59:20 yay I got Yona's fork of LEDSCAPE working on these leds on 3v **** ENDING LOGGING AT Thu Aug 14 02:59:59 2014