**** BEGIN LOGGING AT Thu May 21 02:59:58 2015 May 21 07:03:32 woglinde: morning May 21 07:17:05 Hello. Could anyone here provide some insight on the simplest way to send single-stream data from the BBB at 10+ KHz? Is this possible to do in Python, perhaps using a GPIO pin and a built-in hardware timer? May 21 07:17:59 Ideally I would like to create a simple function capable of sending a 1 us digital pulse, high or low, and scale it up from there. May 21 07:18:24 Is there a good resource to follow for this? May 21 07:18:47 a) don't bitbang May 21 07:19:34 b) if it's not spi, but timing critical, look at the PRU May 21 07:21:47 Yes, the timing will be critical. Are there any known projects that have utilized the PRU for this? I would ideally like to avoid reinventing the wheel. May 21 07:22:41 I'd expect so. But I've never worked with the PRU May 21 07:40:44 Thank you, tbr. May 21 07:41:17 I have found a Python wrapper around a C-based PRU control library. May 21 07:41:20 http://hipstercircuits.com/pypruss-a-simple-pru-python-binding-for-beaglebone/ May 21 07:41:30 Will test it out a bit later. May 21 07:41:50 good luck, but be prepared to write some PRU ASM May 21 07:42:56 Why? May 21 07:43:22 In case this module lacks certain features that may be required? May 21 07:45:31 if it's timing critical you might need to write tight code May 21 07:45:44 Oh. Yeah. May 21 07:46:34 Really, I just need to be able to send 32-bit signals with each "bit" containing both a high and a low pulse and having a 4 us period. May 21 07:46:58 Seems incredibly simple, and probably is. May 21 07:47:42 Just a matter of finding the functions, if they exist within this module, or knowing how to control the hardware timer through the PRU. May 21 07:48:09 I usually look back after projects like this at how much time could have been saved had I known where to look. May 21 09:17:28 * www-BUKOLAY-com Click http://www.bukolay.com I Come. w.w.w.bukolay.c.o.m Click come w:)w:)w:).:)Bukolay:)C:)o:)m Come pls May 21 09:17:33 * www-BUKOLAY-com Click http://www.bukolay.com I Come. w.w.w.bukolay.c.o.m Click come w:)w:)w:).:)Bukolay:)C:)o:)m Come pls May 21 09:18:55 * www-BUKOLAY-com Click http://www.bukolay.com I Come. w.w.w.bukolay.c.o.m Click come w:)w:)w:).:)Bukolay:)C:)o:)m Come pls May 21 09:19:41 tnx May 21 09:19:48 we aim to please **** BEGIN LOGGING AT Thu May 21 09:35:43 2015 May 21 09:51:25 I want to store text files from multiple computer. Can I save the text files on BBB? May 21 09:53:38 sure, there are dozens of ways to accomplish that May 21 09:54:39 how ? please give some instruction May 21 09:57:21 explain your problem further May 21 09:59:33 the problem tbr is that your not a psychic who works does free programming on the side May 21 10:00:00 bitspill: we'll be coming to that bit shortly :) May 21 10:03:07 Actually I don't know that how to store file so that is only problem May 21 10:07:41 tbr: can I have your feed back on this: gitlab.com/nivwiz/cwvw May 21 10:08:34 tbr is a secret agent that does free programming full time May 21 10:10:52 shaggadelic May 21 10:13:18 well, this is interesting,, using a BBB, as a 6lowpan edge router May 21 10:13:46 citylight2: are you overwriting config files provided by other packages? May 21 10:14:19 mrpackethead2: works rather well with the AT96RF233 module and a 4.0 or newer kernel May 21 10:14:45 i'm usng the TI 2531 May 21 10:15:12 its a hop step jump on the way to anohter project. May 21 10:15:16 mrpackethead2: and you got that working? May 21 10:15:32 tbr, yeah, follow the instructions in the ti wiki May 21 10:15:48 http://processors.wiki.ti.com/index.php/Contiki-6LOWPAN May 21 10:18:45 mrpackethead2: that only mentions the 2531 as hardware. no details May 21 10:19:38 mrpackethead2: I've built contiki, both latest source and the contiki-cc integration tree from years ago. the cc2531 firmware was never stable May 21 10:19:45 http://processors.wiki.ti.com/index.php/Cc26xx_sw_examples May 21 10:20:06 Edge Router... May 21 10:20:10 under setup May 21 10:20:57 tbr: yes I need to load a config for the program May 21 10:21:17 I am creating a new device on the system May 21 10:21:22 when I was connect my BBB using putty, I got that "Network error: connection refused". what I do? May 21 10:21:35 tbr: how should I change the config? May 21 10:22:10 citylight2: not sure what the proper way to do this is but you need to be careful May 21 10:22:57 tbr, what issues did you run into. May 21 10:23:16 im early into testing. May 21 10:23:58 tbr: I can save the original files... May 21 10:24:33 mrpackethead2: when I built the firmware myself it was horribly unstable. it never worked longer then 2min. May 21 10:24:49 mmm, its been running all day for me. May 21 10:24:56 it looked like the SLIP data was corrupted May 21 10:24:56 when did you try? May 21 10:25:29 the only sources that worked for me were contiki-sensinode May 21 10:25:56 what were you trying to connect up? May 21 10:26:31 I have some cc2530 modules working now, also based on those ancient sources May 21 10:26:43 but cc2531 never worked stable enough May 21 10:26:46 i've got hte 2650's. May 21 10:26:57 2650's are my device.. May 21 10:27:03 that's very different hardware May 21 10:27:16 oh, 2650 is "out there" May 21 10:27:25 and the 2531 is the usb-dongle. May 21 10:27:31 yes May 21 10:27:58 any fun projects with it? May 21 10:28:23 wired up all 8 adc ports to "moisture sensors" May 21 10:28:39 Im actually hoping to use a TI CC3200 and a TI 2650 to build a Wifi / 6Lowpan gatway May 21 10:28:44 now they are nicely sending values every minute in a UDPv6 packet May 21 10:29:23 the beagle is just a stepping stone along the way May 21 10:29:28 I could not connect my bbb in windows using putty. I got " connection refused " error then What I do? May 21 10:37:09 tbr, what are you logging it too May 21 10:37:10 ? May 21 10:37:12 mrpackethead2: btw, if you want to wire this into a cc3xxx, you might want to use UART instead of SLIP over ACM over USB... May 21 10:37:27 tbr.. in deed! May 21 10:37:53 im certianly not intending on running it that way May 21 10:38:06 I'm 95% through doing the hardware design for it. May 21 10:38:24 I receive the udpv6 packets, parse the values and send them off to a postgresql db. May 21 10:38:41 interesting thign in using the 2650, is that i can do BLE, Zigabee, or 6lowpan etc etc May 21 10:38:57 once i figure out what i really want, i'll revert to the single protocol variants May 21 10:39:00 I'm also planning to evaluate a commercial "iot data munching" platform May 21 10:39:03 they are significantly cheaper! May 21 10:39:19 i've set up a MQTT cluster May 21 10:39:25 and its working really well May 21 10:39:27 i'm quite impressed. May 21 10:39:34 yeah MQTT I've used already, quite nice protocol May 21 10:40:15 I'm using mosquitto as a broker, works quite well May 21 10:40:41 IBM seem to have ditched their stuff and switched to mosquitto too at least for some areas May 21 10:40:46 i'm sending data through to a RRD database May 21 10:41:20 I don't want to degrade historic data for now, so went with a classic rdbms May 21 10:41:43 http://edison.stellascapes.com/voltage_72.png May 21 10:42:16 the data logging is more "interesting" that useful in my applicaiton May 21 10:42:30 the useful thing is knowing whats going on right now. May 21 10:43:40 whats the device your usign in your sensor? May 21 10:44:05 https://twitter.com/tbr23/status/597740519230480385 May 21 10:44:16 fancy shmancy graph using chart.js May 21 10:44:45 nice. May 21 10:44:52 I'm just using RRDgraph May 21 10:45:04 its a bit of a dinosaur, but its still good May 21 10:45:15 http://www.aliexpress.com/item/10PCS-soil-the-hygrometer-detection-module-soil-moisture-sensor-Robot-smart-car-For-UNO-R3-Mega/1207456828.html May 21 10:45:23 bought that type of stuff May 21 10:45:47 oh, and what did you plug that into? May 21 10:45:55 sorry i missed that bit? May 21 10:46:04 very important to keep the duty cycle low as they corrode within days if left on 100% May 21 10:46:22 those are wired into a cc2530 module May 21 10:46:27 oh. ok. May 21 10:46:30 it has 8 ADC channels May 21 10:46:44 yeah, these chips are all pretty capable. May 21 10:48:15 the 2650, is jsut a bit newer. May 21 10:49:55 the 8051 is a bit limited in RAM tho May 21 11:11:12 i've installed 4.0.4-bone4 and removed 3.8.13-bone70 yet uname -a still says "Linux beaglebone 3.8.13 #1 SMP Wed Nov 20 08:28:37 CET 2013 armv7l GNU/Linux", why? (running Debian) May 21 11:16:48 Kwarf: and you have rebootet the machine? May 21 11:19:48 julemand101: yep, and uEnv.txt has uname_r=4.0.4-bone4 and /boot contains "SOC.sh System.map-4.0.4-bone4 config-4.0.4-bone4 dtbs initrd.img-4.0.4-bone4 uEnv.txt uboot vmlinuz-4.0.4-bone4" May 21 11:19:54 if that helps in any way May 21 11:21:29 Kwarf: are you running from SD Card or flash? May 21 11:21:39 SD card May 21 11:21:57 and you are sure the kernel does not load from the flash? :) May 21 11:22:22 (I got the same problem last week where I forgot there was some files on the flash) May 21 11:22:52 fairly sure, the flash should contain an ancient Angstrom image if anything, but i can check just to make sure May 21 11:35:07 julemand101: wow, i erased the internal flash and now it works, this explains so much weird stuff, thanks :D May 21 11:36:40 previously i loaded my custom .dtbo from /lib/firmware, this obviously only worked because i ran the Angstrom kernel without knowing, so where should i put it so the Debian kernel will find and load it? May 21 11:55:07 Kwarf: I don't know where firmwares are stored in Debian May 21 12:01:58 julemand101: thanks anyway, it should still be in /lib/firmware but i also had to add the cape name to /etc/default/capemgr for it to load at boot May 21 12:12:32 Hello all, I got a problem with the Beaglebone Black, the HDMI audio is not working I'm using the Ubuntu rom. If you got time and want to help Thanks :) May 21 12:15:00 does it work on the official debian images? May 21 12:19:39 not tested to be honest, but we don't want to flash it because we are already to depended on Ubuntu for this school project. The HDMI video is working and the audiocard is found (I think) http://pastebin.com/UGwMhbys May 21 12:26:19 any idea @tbr May 21 16:45:41 Hi, my BBB does not show any output from kernel boot log on the serial console although I have "console=ttyO0,115200n8" on kernel command line May 21 16:45:59 anyone has an idea? May 21 17:31:21 I thought it was ttyO2 .. or perhaps I'm mistaken. May 21 17:33:52 Hello everybody May 21 17:34:17 Would anyone know how to configure the Beagle Board Black to provide i2c and SPI pins simultaneously? May 21 17:40:55 hmmm? May 21 17:41:18 just setup the mux May 21 18:02:57 I'm trying to make a eMMC backup of my BBB. I followed the instructions from the answer on http://stackoverflow.com/questions/17834561/duplicating-identical-beaglebone-black-setups but when I insert the card and power up the BBB with the boot button held, I just get a bunch of "CCCCCCCCC" from the terminal. Anyone know why i cant get it to boot? May 21 18:06:14 clockman .. that means it can't find the boot files on your SD card May 21 18:06:52 but i know they are there, because it the SD card boots another BBB that I have. the only difference is that one has Angstrom, and this one has the newest Debian distro May 21 18:08:31 I suggest there are some small differences in configuration May 21 18:10:25 well, i'm not reading thing wrong right? If the switch is pressed, the BBB first looks at the sd card for MLO and tries to boot from there. Is there some kind of software setting that could be on the BBB that tells it not to do that? May 21 18:10:49 not unless you've forced the boot pins via a cape May 21 18:11:10 afaik there have been no hardware revisions to the default boot pin setup May 21 18:11:11 nopers May 21 18:11:49 the lack of MLO would be the problem .. or the 'apparent' lack of mlo May 21 18:12:09 I know the uboot config has been changed since the angstrom days May 21 18:12:27 the person you need to ask isn't here alas. May 21 18:12:42 but jkridner may be able to shed some light on his emmc script .. May 21 18:15:14 well i'll keep searching around. last time i used linux, LILO was the bootloader of choice, so I'm a bit perplexed by u-boot et al. May 21 18:15:16 thanks May 21 18:17:53 heh I'd like to see a arm boot lilo .. but since there's no bios, you have to tell the bootloader what hardware it's looking at .. hence uboot :) May 21 18:18:30 when it works .. all is good .. when it doesn't .. its like all software/electronics .. lol. May 21 18:18:47 usually have to begin with RTFM :/ May 21 18:19:42 are your boards the same emmc? or is one the newer revision? that might be one thing May 21 18:20:50 actually .. thinking about it .. I don't think the newer configs have a mlo .. it goes straight to uboot. May 21 18:24:34 aha .. could be in luck .. May 21 18:24:52 rcn-ee .. cloning eMMC and we're getting 'CCCCCC' .. what to look for? ;) May 21 18:25:32 veremit, it's not finding the bootloader where it wants it... May 21 18:25:42 how are you cloning? May 21 18:25:49 correct. clockman .. May 21 18:26:12 process seems to be one of jason's scripts.. May 21 18:26:49 rcn-ee .. http://stackoverflow.com/questions/17834561/duplicating-identical-beaglebone-black-setups May 21 18:26:51 uboot is more common for 'embedded' linux May 21 18:27:16 or rather MLO/uboot together May 21 18:27:34 I think the bbb setups have dropped mlo May 21 18:28:20 mlo is just in the mbr, so if you do a full dd dump, it'll be fine.. May 21 18:28:25 same way kernels aren't uImage .. now zImage thank to RN's work :) May 21 18:29:47 hmm I know n othing of buildroot May 21 18:31:31 it builds stuff. ;) May 21 18:32:16 no, reaaly?! May 21 18:32:31 both boards are RevC - not sure if there's a sub version May 21 18:34:43 rcn-ee .. the mlo's likely to hav changed between angstrom and debian days? May 21 18:35:23 I see there have been sfdisk updates to your wiki pages :) I think I tripped on that one recently May 21 18:35:36 imx doesn't use mlo though... May 21 18:35:38 yeah i was thinkgin that first, but the instruction's use dd... May 21 18:35:56 mlo is just identifer the bootrom looks for. ;) May 21 18:36:08 ah .. so that'll be uboot whatever May 21 18:36:25 yeah, it's 'u-boot spl'... ;) May 21 18:36:47 was there a 3-file bootloader once? May 21 18:37:32 clockman, you did exatly as: http://elinux.org/BeagleBone_Black_Extracting_eMMC_contents May 21 18:38:13 veremit, go look at the snapdraggon..https://wiki.linaro.org/Boards/IFC6410 board setup... look at all those stages. ;) May 21 18:39:19 rcn-ee .. lol .. too many processors :p May 21 18:39:41 sorry, for the slow response, I'm providing tech support at work right now. One sec (more than one). May 21 18:41:41 rcn-ee .. I have the original pico-itx from via kickin around somewhere never got cased up :/ May 21 18:41:55 the x86 one. They've gone quad-core now :/ May 21 18:42:08 Yes. I performed those steps. I suppose I can try redownloading and reformatting. maybe a solar flare caused a bit to flip somewhere :-S May 21 18:45:10 clockman, the most common mistake, would have been using the partition vs the drive: (/dev/mmcblk1p1 vs /dev/mmcblk1) since the bootloader is in the mbr, you need to use the whole drive.... May 21 18:47:31 rcn-ee .. kodi build for he 6410 yet?! :) May 21 18:48:34 well, I'm looking at the "Steps to save" section, where all they say is to format the uSD card as FAT, copy the files in the zip to the root of the uSD card, then try to boot it in in BBB. I can't get it to boot. So I don't even get to the point of reading/writing the /dev/mmcblk0 May 21 18:50:37 clockman, ahh... i thought you had it booting... May 21 18:52:08 my bad probably .. we're a stage behind :) May 21 18:52:46 clockman, so... from the beginig, can i assume linux? May 21 18:53:10 OS X, but if need be, I can hop on my linux box May 21 18:54:03 linux would be eaiser.. as we need sfdisk... someone should create a *.img of beagelbonebalck-save-emmc, so users can just dd ;) May 21 18:55:18 okay. all set and ready May 21 18:56:38 first what version of sfdisk? sfdisk --version May 21 18:57:01 util-linux 2.20.1 May 21 18:58:00 okay, using this as reference: https://eewiki.net/display/linuxonarm/BeagleBone+Black#BeagleBoneBlack-SetupmicroSDcard May 21 18:58:37 clear out the partition table, (skip install bootloader), ... we need to modify partition for fat... May 21 19:00:24 clockman, here's a the flow: https://gist.github.com/RobertCNelson/c93b96f7c4fba548ac82 May 21 19:00:59 hum, we might want -F 32 as it a big *.img file.. May 21 19:08:17 sweet. that booted it. May 21 19:08:51 so I should be able to modify the autorun.sh on the card as long as I don't delete MLO or u-boot.img, right? May 21 19:09:40 correct, just don't delete mlo/u-boot.img. ;) May 21 19:10:27 well thanks for your help, rcn-ee. May 21 20:54:07 has anyone had success with rs485 on kernel 4.0 and beagle bone black? I went down the route of trying to use the omap-serial until i stumbled on this thread (https://www.mail-archive.com/beagleboard@googlegroups.com/msg26555.html) May 21 20:55:08 bo .. I don't think the serial driver has support to drive the DE pin for 485 May 21 20:55:19 erm .. function. May 21 20:58:22 vermeit... i got it working in kernel 3.8 but i actually never noticed which driver I was using, and i blindly applied the same technique. followed this guide http://blog.savoirfairelinux.com/en/2013/rs-485-for-beaglebone-a-quick-peek-at-the-omap-uart/#note-patch-omap-rs485 May 21 20:58:54 so if the omap-serial driver is being deprecated, and the serial driver doesn May 21 20:59:24 t have support to drive the DE pins, then what is a poor embedded developer to do? May 21 21:02:33 bo .. personally, we used a usb->uart bridge (ftdi) which provides the DE signal which you can use on a driver chip easily May 21 21:02:47 its the solution we've used for usb for years reliably May 21 21:03:13 otherwise you need a uart and drive solution which properly supports rs485 May 21 21:03:18 driver* May 21 21:03:37 Ummm get a USB serial port? or http://www.nxp.com/products/interface_and_connectivity/bridges/i2c_spi_slave_to_uart_irda_gpio_bridges/series/SC16IS740_750_760.html May 21 21:04:40 or an i2c/spi one :P May 21 21:05:09 NXP makes several of them. May 21 21:05:19 with FIFOs May 21 21:07:17 mm I/we have no need for fifo's with PCs, etc with full-featured UARTs May 21 21:07:30 but interesting to know :D May 21 21:08:01 sorry, my computer crashed a few times...silly mac. but ya. The hardware team has spun up a board for me to talk to via uart2 that uses rs485. so no, unfortunately I can't use a usb serial port. but yes to @veremit, i need to find a driver that fully supports rs485. May 21 21:08:27 hmm let me know what you find ;) May 21 21:09:14 I'm just a little confused on the whole omap-serial thingy, i don't want to spend a lot of time trying to get the omap-serial to work if its going to be deprecated. May 21 21:09:16 our product we spun up a board with a usb2504 hub chip and ft230x usb-uarts to bypass this exact problem. We used the onboard uart for rs232 though May 21 21:09:31 bo .. just stick with the old version, and maintain it yourself. May 21 21:09:36 no need to follow the 'crowd' May 21 21:10:40 i c. be independent. that would provide me job security. just curious, which driver did you use for the rs232? May 21 21:12:49 I was using an old build I rolled myself via RCN's eewiki pages May 21 21:13:09 it was probably a 3.11ish kernel ? May 21 21:13:35 somewhere about there .. mine' s a 3.13 May 21 21:15:15 so it woulda been mainline-derived with possible patches by rn May 21 21:17:41 lemmie see if I can find any sources .. May 21 21:22:27 thanks for the help veremit, I'm discussing with the team lead and i think we are gunna try to first hack the 8250 driver, and if that doesn't work, then we might try to get the omap-serial driver to work May 21 21:22:54 bo .. looks like we were on 8250 base .. I see no omap serial references at all in the 3.13 kernel image I got here May 21 21:24:33 ya, i think i was first introduced in 3.14, but can't remember. we would really like to use the 4.0 kernel, but we could roll back to 3.18 if we absolutely have to. May 21 21:24:43 it* May 21 21:25:14 I dunno how easy it would be to hack, I guess the pinmux's would be relatively trivial May 21 21:26:32 ya. the device tree is trivial. its just the driver I'm having some issues with May 21 21:26:56 ok, im gunna start poking around. ill update when I've found something if you are curious May 21 21:27:15 sure. I do remember looking at the sources and thinking .. yeahno. lol. May 21 21:27:27 we needed a usb hub for the other board that was a possible candidate May 21 21:37:29 bo .. currently looking at source in drivers/tty/serial/8250/8250_omap.c May 21 21:37:59 might be worth contacting the author/maintainer? May 21 21:40:08 veremit... so i guess RCN has released his kernel ver 4.1 that has Cape manager support. this might solve my issue. May 21 21:40:58 oo oo cool. May 21 21:47:08 veremit... im gunna move to the new kernel (got everything in buildroot so its super quick for me to bump versions.) thanks for the help. lets see how this goes May 21 21:47:54 bo .. looking at the TRM .. there's no function for Data Enable .. so it would have to be a software bodge. May 21 21:49:41 ya. so on the 3.8 kernel, i simply set the data direction pin to a gpio, so the gpio is high for receiving, it goes low once i start transmitting. is data enable and data direction the same thing...i might be confused on that May 21 21:50:28 I use 'data enable' to be tx enable ie. its for the driver enable on the max485/etc chip May 21 21:52:25 ok. so ya, as you said, it must be some software black magic going on to get the rs485 to work May 21 23:00:01 hi May 21 23:01:07 \\join May 21 23:01:48 ./quit May 21 23:02:56 thats really nice.. May 21 23:03:27 i used to create channels in MiRc since 1996 :) May 21 23:07:16 That was eons ago May 21 23:09:47 20 years ago today May 21 23:49:03 veremit: ahhh, the serial port, such a delightful peripheral May 22 02:06:52 Hi I have QT creator setup on my desktop and has a qt4 library on my BBB when I try to run my qt application on BBB i have this error cannot connect to x server May 22 02:35:08 imanpakii: how are you starting the application? unless it is started from within the graphical environment on the BBB, the "DISPLAY" environment variable won't be set May 22 02:35:54 e.g. if you're trying to start the application via SSH, first do: May 22 02:35:59 export DISPLAY=:0 May 22 02:37:40 (this is assuming an X server is actually running on the BBB) May 22 02:51:44 ok, time to write a vitriolic reply to the inane response I got from TI support -.- **** ENDING LOGGING AT Fri May 22 02:59:58 2015