**** BEGIN LOGGING AT Mon Feb 12 03:00:01 2018 Feb 12 03:00:11 any idea why the measured signal from mosi would differ from what is expected? Feb 12 03:04:13 MotorBridgeCape? Feb 12 03:09:11 basically the mosi should be sending a command repeatedly to get yaw pitch and roll, however the measured signal varies both in length and values Feb 12 03:10:04 Is a mosi analog? Feb 12 03:10:13 Hold please...I will look it up. Feb 12 03:11:14 I give up. You win. Sorry to interject. Feb 12 03:11:27 I am sure someone will be around soon to see your question. Feb 12 03:12:35 SPI? Feb 12 03:12:46 Hey...are you using your SPI to talk to it? Feb 12 03:16:45 I am no SPI champion but I would say interference. Feb 12 03:17:56 ...Just for a query, I am basically a motor dude that makes machines go in a specific order w/ repetition. Feb 12 03:18:22 So...I got software A, then B, then C, and then A starts again. Feb 12 03:18:30 all good i am a comp dude who got thrown on this project Feb 12 03:18:36 Oh. Feb 12 03:18:46 What project? Feb 12 03:19:29 interfacing beaglebone w/ an IMU Feb 12 03:20:10 Oh. Feb 12 03:20:23 BeagleBone Blue has an IMU built-in. Feb 12 03:20:35 Good then...have some fun. Maybe I was misunderstanding the question. Feb 12 03:20:38 this is a very precise IMU Feb 12 03:20:47 k Feb 12 03:21:04 BongoShaftsbury: hmm? Feb 12 03:21:12 Come on, let me try. Feb 12 03:21:42 BongoShaftsbury: your description of what issue you're seeing with the spi data is rather vague Feb 12 03:21:59 well basically the signal does not match what i am intending to send Feb 12 03:22:07 I don't suppose you have a scope picture? Feb 12 03:22:18 Scope photo! Feb 12 03:22:19 i do not Feb 12 03:22:25 Dang... Feb 12 03:22:26 i can get one tomorrow though Feb 12 03:22:58 I got this cheap-o, depot scope. What should I do? I can test things, too! Feb 12 03:23:08 other than that, check you're using the right spi mode, try lowering the clock rate to determine if it could be related to signal integrity issues (your description of it doesn't sound like it though) Feb 12 03:24:04 i guess i can try lowering the clock to like 1khz go really slow and have it loop at like 1 second intervals Feb 12 03:24:10 and see what i get then Feb 12 03:24:31 note that in between transfers (when clock is idle and cs deasserted), mosi will be undriven and will just get slowly pulled up (assuming pull-up is enabled). this can look weird on the scope but is not harmful Feb 12 03:24:58 i don't think the mode would impact how long the message is, but would impact the content of the message i will double check that as well but i am pretty sure we have it set correctly, spi mode 1 Feb 12 03:25:21 that is a good point zmatt Feb 12 03:27:58 ... Feb 12 03:28:03 Hey. Feb 12 03:28:27 I did it. I almost made a loop but instead made a back and forth. Feb 12 03:28:28 other than this, you'll need to get more specific information otherwise I don't see much chance of getting it diagnosed Feb 12 03:28:47 while moving left... Feb 12 03:29:00 rgr that Feb 12 03:29:22 <<<< on break Feb 12 03:33:29 ... Feb 12 03:33:37 Here is an idea (idea only). Feb 12 03:34:14 Has anyone made a New Orleans' style Trollie Car w/ their BBB so they could mow grass? Feb 12 03:34:46 ... Feb 12 03:35:08 I am asking b/c I was thinking that batteries are getting driven down too quickly w/ all this tech. Feb 12 03:35:15 ... Feb 12 03:37:10 I could set up a line, AC/DC, so that I could run my BBB on a line on poles that powers the BBB so my damn mower knows where to go. Feb 12 03:37:11 heh? Feb 12 03:38:40 straight, u-turn right, straight, u-turn left, and restart the program... Feb 12 03:40:05 The program only runs four different intervals if you include both "straight" functions. Feb 12 03:41:19 >>>> ] <<<< [ and repeat. Feb 12 03:52:41 I2C issues...what changed? Feb 12 03:52:53 4.9.x and I2C issues... Feb 12 03:56:02 Are we using I2C pins for other things outside of I2C devices now? Feb 12 03:58:30 ehm, what? Feb 12 03:59:16 I do not know how to ask anymore. Let me try again. Feb 12 03:59:29 I missed three months of change. Feb 12 03:59:38 here is some of the data from the oscope Feb 12 03:59:40 https://ghostbin.com/paste/7kz4m Feb 12 03:59:49 P9.19 (scl) + P9.20 (sda) are configured to i2c (bus 2) by default Feb 12 03:59:57 Got it. Feb 12 04:00:08 you can see the 01 36 00 F0 53 D5 00 00 E0 D8, i think that is correct, but you can see it varies Feb 12 04:00:53 BongoShaftsbury: I have no idea what I'm looking at Feb 12 04:01:17 the last bit in hex is the data the oscope captured being transfered from mosi Feb 12 04:01:30 there should be a repeating pattern, but it varies a lot Feb 12 04:03:04 this is no substitute for a picture of the actual signal Feb 12 04:03:17 I get some impression that it's picking up noise and/or somehow gets misaligned Feb 12 04:03:39 but it's impossible to tell from this what's going on Feb 12 04:04:02 rgr that Feb 12 04:05:14 Hey...what should I do if my I2C address is not being sensed? Feb 12 04:05:28 BongoShaftsbury: it looks like packet 10 started right in the middle of a transfer Feb 12 04:05:56 set_: debug the problem Feb 12 04:06:13 (either software or hardware) Feb 12 04:06:56 Error accessing Ox4b? Feb 12 04:06:57 could it be that the software might be requesting a new transmission before the prior transmission is complete Feb 12 04:07:06 BongoShaftsbury: no Feb 12 04:07:28 BongoShaftsbury: there's no way to abort a transfer in progress Feb 12 04:08:01 maybe the transmission buffer is getting full Feb 12 04:08:24 and parts of the transmissions are lost when it is full Feb 12 04:08:34 no Feb 12 04:08:45 I mean, that would require a serious driver bug Feb 12 04:09:04 we make heavy use of spi ourselves and I've never seen any such issue Feb 12 04:09:25 so if the buffer is full the entire tx is rejected? Feb 12 04:09:43 uh no? Feb 12 04:09:57 as in, if i try to queue a new tx, that new tx would be rejected Feb 12 04:10:01 when the buffer is full Feb 12 04:10:04 no it will wait Feb 12 04:10:18 what buffer do you mean? Feb 12 04:10:29 i'm assuming there is a buffer for spi for tx waiting to be sent Feb 12 04:10:55 if you mean the hardware fifo, the driver makes sure data is written to it only when space is available Feb 12 04:12:16 pinmux file not found! Cannot read pinmux file: ... Feb 12 04:12:47 the data buffer in which the data is held for transfer is allocated by whichever client in the kernel requests the transfer. if using the spidev interface to userspace, the spidev driver allocates a fixed buffer (size can be configured using a module parameter) Feb 12 04:13:44 when doing a transfer from userspace, the driver copies the data to that buffer, performs the transfers (waiting synchronously for it), and copies the received data back to userspace (if applicable) Feb 12 04:14:07 there's a mutex held over all this so if you'd try to do transfers from multiple threads/processes, they'd just block to wait their turn Feb 12 04:14:19 oh i see Feb 12 04:14:32 so the case i am proposing should never happen Feb 12 04:14:45 correct Feb 12 04:15:33 it is true that data would be lost if you'd write data to the spi controller peripheral while it's transmit fifo is full, but like I said this would require a serious driver bug. it would also trigger an error interrupt Feb 12 04:18:06 makes sense Feb 12 04:18:11 i will get some images of the wave forms tomorrow Feb 12 04:18:19 when i am back in the facility Feb 12 04:19:38 if there's a cable of significant length between the beaglebone and the device, it may also be wise to include series resistors to suppress overshoot, especially on the clock signal. (place them near the side that's driving the line) Feb 12 04:20:30 the line is < 6 inches Feb 12 04:20:42 < 15.24 cm Feb 12 04:21:06 that's not a negligible length I think Feb 12 04:22:02 series resistor as in beaglebone ---/\/\/\//\----/\/\/\/\----VN100? Feb 12 04:22:17 w/ the resistors closer to the beagle bone Feb 12 04:22:17 you want to avoid getting ringing on the clock line since it may trigger false clock edges Feb 12 04:23:05 note I used plural "resistors" because there are multiple lines, not multiple resistors on one line :P Feb 12 04:23:12 what ohm resistors would you recommend? Feb 12 04:23:13 it's less important on the other lines though Feb 12 04:23:30 oh, you said series, so i thought you meant two resistors in series Feb 12 04:23:57 no I meant series as in, inline in the signal line Feb 12 04:25:06 would 5kohm be good? Feb 12 04:25:21 the other way to terminate the line is parallel termination, which places it from signal line to ground (or some other fixed voltage) near the other end of the line, but that consumes a lot of power and is not appropriate here Feb 12 04:25:56 5k is very large, typical is 120 ohm or less Feb 12 04:26:02 updating images. Aw! Feb 12 04:26:04 optimal value actually depends a bit on the cable Feb 12 04:26:17 okay i'll start w/ 220 and work my way lower then Feb 12 04:26:32 optimal would be characteristic impedance minus the driver's output impedance Feb 12 04:26:48 less than that results in overshoot Feb 12 04:27:18 more than that means the signal will rise/fall more slowly, so that's usually less a problem Feb 12 04:28:00 using excessively large series resistance would limit the max clock speed Feb 12 04:28:23 (and also make the line a bit more sensitive to picking up external noise) Feb 12 04:31:22 rgr that Feb 12 04:52:18 Hi new here Feb 12 04:53:38 Hi Feb 12 04:54:17 Yep...late night bites. Feb 12 04:54:18 Hello. Feb 12 04:54:32 Motor Bridge Cape? Feb 12 04:54:46 Hey good people. Is it possible to use a 9v battery, with no lipo, to power a BBblue? Feb 12 04:55:46 resistors might work. Feb 12 04:55:53 Do not take my word for it. Feb 12 04:56:00 The thing about 9v batteries is the active plate area is so small, and thus the internal resistance is so high, they can't source much current Feb 12 04:56:12 Been on it's before like 20 years ab Feb 12 04:56:28 indeed ignore set_ here Feb 12 04:56:30 :P Feb 12 04:56:35 Yea boy! Feb 12 04:56:36 Ago Feb 12 04:57:02 <<<< has terrible advice Feb 12 04:57:20 So it would depend on the battery, but in general, 9v batteries are the most expensive and awful way to buy energy for a project. Do it in a pinch, but prefer a 4xAA holder or a lipo or something if you have the option. Feb 12 04:57:35 https://www.baldengineer.com/9v-battery-energy-density.html Feb 12 04:57:42 So if I wanted to power a couple motors and servos I should not be using a 9v. Only thinking 9v since I have lots laying around. Feb 12 04:58:16 I am getting a newer image to debug. I could not upgrade my Adafruit_BBIO on my BBB. Feb 12 04:59:07 Can I ask is this a doggie chat or did I get it wrong Feb 12 04:59:31 BBB! Feb 12 04:59:36 Way: also note that you'd have to connect it to the barrel connector (vcc), not the battery terminals Feb 12 04:59:53 Telephone: this channel is related to the beagleboard and beaglebone ARM-based embedded systems Feb 12 05:00:03 sorry :) Feb 12 05:00:10 doggie? Feb 12 05:00:45 What is a doggie chat? Feb 12 05:00:52 Way: A couple motors and servos really deserve a real power supply. Running motors and logic from a high-source-impedance supply like a 9v is just asking for pain. Feb 12 05:00:55 Telephone: the logo of beagleboard.org is a cute doggie though ;) Feb 12 05:01:05 Aw! Feb 12 05:01:19 Haha nothing to do with dogs but sounds z Feb 12 05:01:27 You need a H-Bridge. Feb 12 05:01:40 set_: the blue has H-bridge drivers Feb 12 05:01:43 Awesome Feb 12 05:01:44 Yea boy! Feb 12 05:01:45 that's not relevant here Feb 12 05:01:55 Pain in terms of what? I don't have a strong EE background Feb 12 05:01:57 they don't magically improve your power supply Feb 12 05:02:25 I know but they are gates of sorts which allow specific current to go and stop and go again. Feb 12 05:02:36 Magic Gates. Feb 12 05:02:51 My Motor Bridge Cape has an H-Bridge. Feb 12 05:03:10 That sucker lets my motors know when and where. Feb 12 05:03:45 ... Feb 12 05:03:49 What would you recommend more: Lipo or AA batteries Feb 12 05:04:18 LIPO for the BBBLUE. Feb 12 05:04:29 They have a barrel jack. Feb 12 05:04:36 and a LIPO connection. Feb 12 05:04:46 Ì will leave thanks it's u guys and girls that make the future reality Feb 12 05:05:20 Bye for now. Feb 12 05:05:29 Wish I knew what you wd Feb 12 05:05:37 Telephone, change your mind one day and get into the BBB. Feb 12 05:05:54 Not space exploration yet but interesting. Feb 12 05:06:01 What's bbb Feb 12 05:06:11 BeagleBone Black! Feb 12 05:06:12 The Blue is made to run from a lipo, I would recommend that you get exactly what they use in the examples. Feb 12 05:06:25 MCU! Feb 12 05:06:49 Ì love the new space Feb 12 05:07:06 Lithium is generally happier with high current (has lower internal resistance) compared to other chemistries, so it'll have the easiest time keeping up with motor starting current without browning out the logic Feb 12 05:07:10 w/ GPIOs and SPI, I2C, UART pins, and pruss. Feb 12 05:07:38 So much dark soppppp Feb 12 05:08:03 Space wow Feb 12 05:08:21 my new image is one second away! Feb 12 05:08:22 Ding! Feb 12 05:08:46 That was a tylo Feb 12 05:09:02 Goodnight Feb 12 05:09:20 gotcha, so would a 1.6A lipo be a good choice or would you recommend higher? Feb 12 05:10:20 that depends on so many things... what's the locked-rotor stall current on your largest motor, and how many of those will be moving simultaneously? Feb 12 05:10:41 what's your mass budget? Physical space constraints? Monetary budget? Feb 12 05:11:35 Heavy Metal and Blades! Get that BBB telling heavy machinery what to do! Feb 12 05:11:52 myself: lol, I tried to look up how the 5v regulator used on the blue responds to line transients Feb 12 05:12:25 but it has almost no performance specs other than (mediocre) efficiency. not even stuff like output ripple Feb 12 05:14:34 1. Motors are not very powerful, not sure about the exact numbers but the overall goal is for them to provide enough force to a propeller/impeller to lift a couple pounds of hovercraft off the ground. 2. Max 2 motors and 1 servo moving at a time. 3. mass budget will be built around battery so no issue there. 4. Cost is sponsored so no issue there either but preferably low cost Feb 12 05:17:29 I am currently looking at 1.6 vs 2.2 Feb 12 05:19:15 ... Feb 12 05:19:35 use Adafruit_GPIO.I2C instead of Adafruit_I2C. Feb 12 05:19:59 This is what this machine is telling me. Off to learn more. Feb 12 05:25:30 I used it but something is odd. I am getting new issues. Feb 12 05:29:29 or just use any i2c library for python that's *not* adafruit Feb 12 05:29:31 e.g. smbus2 Feb 12 05:44:25 I have to use the dang Adafruit one b/c my Motor Bridge Cape is based on it. Feb 12 05:44:51 I have not scripted everything over yet but I just may have to do it. Feb 12 05:45:11 I changed some of the software. I am updating right now. I will let you know what happens. Feb 12 05:53:36 cannot debug...turning into sugar or salt. BBB! Feb 12 05:59:21 zzzz. Feb 12 08:08:45 hey guys! can anyone suggest on how does one make the kernel think of a touchscreen as a keyboard? I've found something like adding ID_INPUT_KEYBOARD=1 to udev for touchscreen, but I'm really not sure where do I do that :-) Feb 12 08:08:54 can anyone help me to enable an interrupt from uart1 in C programming Feb 12 08:42:22 Hello, I'm having some weird issues with basic linux commands on the Debian distribution (PuTTy) on my BBB. Feb 12 08:43:33 I created a user and under that user, I've tried running commands like mkdir or ls -l > test.txt and they ran perfectly well. Feb 12 08:43:51 Later, I deleted that user and created another use. Feb 12 08:43:54 user* Feb 12 08:44:40 As this user, when I try and run the same commands before, I get back "Permission Denied". I'm able to run some of the commands as sudo and unable run others such as ls -l > test.txt Feb 12 08:44:59 Not sure what the issue is. Can anyone help? Feb 12 08:45:55 abhi_: that may depend on where you are in the filesystem and a few other things Feb 12 08:46:48 okay..I'm in the home directory in my user account...is that what you mean Feb 12 08:47:32 yes Feb 12 08:48:03 so what happens if you "touch test" Feb 12 08:50:22 I typed in touch test under home dir as user - nothing happened - just went onto the next line with no error or result Feb 12 08:52:37 although you've deleted your user, user id is changed, so you can't just use the same home dir with another user Feb 12 08:53:05 okay - so how do I fix this? Feb 12 08:53:46 if it helps, i deleted the home dir related to the first user Feb 12 08:56:14 please run these commands: whoami ; cd; pwd; ls -ld . ; cat /etc/passwd | grep $USER Feb 12 08:56:22 to get needed info Feb 12 08:59:33 user /home/user drwxr-xr-x 3 user user 4096 Feb 11 17:38 . user:x:1001:1001:username,101,,:/home/user:/bin/bash Feb 12 08:59:48 this is what i got for those commands Feb 12 10:00:20 hello Feb 12 10:01:01 i have query regarding beaglebone black Feb 12 11:15:12 Hi **** BEGIN LOGGING AT Mon Feb 12 12:28:34 2018 Feb 12 13:37:16 need help https://askubuntu.com/questions/1005462/internet-access-beagleboard Feb 12 13:37:48 Trying to access internet to update packages etc , on Beagleboard, tried a couple of things but did not help Feb 12 13:38:30 is there a way to bypass this connman service, it is so annoying Feb 12 13:38:35 well, what did you try Feb 12 13:38:46 also which distribution and version are you running on it? Feb 12 13:39:18 i tried changing /etc/network/interfaces on the beagle board Feb 12 13:39:27 this is my lsb_releas -a output Feb 12 13:39:36 No LSB modules are available. Distributor ID: Debian Description: Debian GNU/Linux 9.2 (stretch) Release: 9.2 Codename: stretch Feb 12 13:40:19 i tried changing the /etc/network/interfaces on my guest OS Ubuntu but that did not help either Feb 12 13:40:30 so you want to route out via the USB network interface between BBB and your PC? Feb 12 13:41:16 Okay, i have a Windows host, i have an Ubuntu guest, i can SSH from my Ubuntu to beagleboard, also from Windows host Feb 12 13:41:43 I can access Internet from Windows host and Linux guest, but not from beagleboard Feb 12 13:41:56 long answer short: there is no straight forward way to configure this on windows. Feb 12 14:05:49 Yea i guess route out is the right term, i am clueless about netwroking Feb 12 14:05:49 you'd need to modify network configuration scripts on the BBB and enable internet connection sharing on your windows Feb 12 14:05:49 okay, the network config scripts are in /etc/network/interfaces right, or in /var/lib/connman Feb 12 14:05:49 haven't looked at that in ages. probably somewhere in /opt Feb 12 14:05:49 getting this wrong will break your network connectivity to the BBB **** BEGIN LOGGING AT Mon Feb 12 14:06:10 2018 Feb 12 14:36:26 connmanctl should be the way Feb 12 14:44:32 this might be more of linux question.. but.. i created a quick and dirty module that is ouputing some debug messaging to the kernel via printk, output is shown on the serial port console, and in dmesg, I would like to be able to have these messages go to the LCD that is connected.. Feb 12 14:46:01 There is no input device, so I would need to set this up over the serial port, or a startup script. Feb 12 14:46:27 sounds like you'd want a framebuffer console Feb 12 14:48:25 I have a LCD that is showing the beagle bone log in, and my serial connection.. Feb 12 14:48:51 ok.. finding good google resules on bbb famebuffer console.. will read Feb 12 14:49:18 if it already shows a login prompt, then it's set up Feb 12 14:49:59 figure out which device node it is and add it to your kernel command line as another console Feb 12 14:51:46 ok, that should help.. I will try out Feb 12 14:56:09 how do I figure out which device node it is? Feb 12 15:02:36 there is probably a process running for it. you should be able to see it in the output of 'ps auxf' along with the others bound to their /dev nodes Feb 12 15:04:16 would this be it? /sbin/getty 38400 tty1 Feb 12 15:05:46 ok.. I can send a message >echo "hello" 1>/dev/tty1 Feb 12 15:12:49 updated uEvn.txt to pass tty1, it's doing what I want it to do.. thanks tbr Feb 12 15:13:55 Eric_T: yay :) Feb 12 15:33:56 Anyone using a bbblue and have programmed the Mode and Pause buttons ? Feb 12 15:48:40 Pause and mode buttons, anyone programmed them on the blue? Feb 12 16:03:26 I've just tried the Debian 9.3 2018-01-28 4GB SD LXQT image on my bbb Feb 12 16:04:05 the usb network is not coming up like normal, and sudo ifup usb0 results in cannot find device "usb0" Feb 12 16:04:31 what am I missing? Feb 12 16:05:25 those images seem to have TWO usb network modules loaded (for increased compatibility) Feb 12 16:05:57 iow check usb1 too Feb 12 16:06:47 I can see in /etc/../interfaces that it's meant to be usb0 Feb 12 16:07:11 also I can see usb_ether is loaded via lsmod Feb 12 16:07:24 (sudo ifup usb1 results in unknown interface) Feb 12 16:07:58 ok, something else going wrong then Feb 12 16:08:32 dmesg looks ok Feb 12 16:09:02 though I do see this: musb-dsps: probe of 47401400.usb failed with error -16 Feb 12 16:18:50 so is the image broken, or is it incompatible with the bbb hw I have? Feb 12 16:24:17 sounds like the image might be broken? Feb 12 16:24:23 looking at the kernel boot messages, after uboot finished I can see this message: Failed to start Load Kernel Modules. Feb 12 16:24:29 hmmm Feb 12 16:24:40 might explain things Feb 12 16:24:45 I'll try a new sd card just in case Feb 12 16:24:46 not really Feb 12 16:24:50 no? Feb 12 16:24:55 or, well Feb 12 16:25:05 no, not really Feb 12 16:25:10 actually yes you're right, when i do lsmod I see usb_ether laoded Feb 12 16:25:15 yeah, that Feb 12 16:25:19 it's still odd though Feb 12 16:26:21 maybe check what's up with that service using: systemctl status systemd-modules-load Feb 12 16:28:16 service is running ok Feb 12 16:28:33 no wait Feb 12 16:28:50 I'm logged in over serial and there are escape characters everywhere! Feb 12 16:29:02 Active: failed Feb 12 16:29:37 uhh, maybe your terminal type is set to something that supports colors while you're using something that doesn't? Feb 12 16:29:59 (although normally it shouldn't matter whether you use serial or ssh) Feb 12 16:30:47 jsut a standard xterminal Feb 12 16:31:36 what are you using for serial communications? maybe it "helpfully" intercepts escape chars instead of letting your terminal interpret them Feb 12 16:31:46 miniterm.py Feb 12 16:32:01 never heard of it Feb 12 16:32:09 huh, screen works much better Feb 12 16:32:12 I always use screen without any problems Feb 12 16:32:12 yeah Feb 12 16:32:23 it gets installed with the python serial libs, I quite like it Feb 12 16:32:28 well, I used ot;) Feb 12 16:32:52 just remember to use ctrl-A ctrl-K if you want to close the serial port instead of just detaching from the screen session Feb 12 16:33:11 or ctrl-A K Feb 12 16:33:24 thanks Feb 12 16:35:35 so do you think the systemd-modules-load might have something to do with the missing usb0? Feb 12 16:36:13 dunno, what does it say about why it failed Feb 12 16:36:21 nothing in the journal Feb 12 16:36:40 I usually use this wrapper around screen when using it for serial: https://pastebin.com/JRtKwsNw maybe it's useful to you too Feb 12 16:36:51 journalctl -u systemd-modules-load.service Feb 12 16:36:56 no entries Feb 12 16:37:36 does kernel log mention a failure to load a module? Feb 12 16:38:22 the failure to probe musb-dsps does sound like potentially the root problem though Feb 12 16:38:25 lemme look up -16 Feb 12 16:38:56 just rebooting with a new sd card Feb 12 16:39:00 -EBUSY Feb 12 16:39:04 hm Feb 12 16:40:55 the only thing in the kernel messages is that line about module failing to load, even suggests to run the systemctl command to see the results as you recommended Feb 12 16:41:18 that's not actually a kernel message though Feb 12 16:43:49 just trying on another bbb Feb 12 16:43:57 I don't see any obvious possible source of EBUSY for the probe of musb_dsps Feb 12 16:44:18 it must be passed up from somewhere, but no idea where Feb 12 16:44:24 same message about Failed to start Load Kernel Modules Feb 12 16:44:36 yet no error in journal? Feb 12 16:44:37 same missing usb0 Feb 12 16:44:52 no message in journalctl -u systemd-modules-load.service Feb 12 16:45:23 which modules are in /etc/modules-load.d/* ? Feb 12 16:45:36 though just running journalctl and paging to the end I can see Jan 28 14:26:03 beagleboard sh[401]: g_multi: waiting for /sys/class/net/usb0/ Feb 12 16:46:03 yeah g_multi won't fail on that, it'll just wait for it to show up Feb 12 16:46:22 so maybe it's explicitly trying to load musb_dsps (directly or indirectly) hence failing on that Feb 12 16:46:32 what happens if you try to restart systemd-modules-load.service ? Feb 12 16:46:40 good idea, Feb 12 16:47:08 btw, modules-load.d contains an empty modules.conf and some stuff in robotics_cape_modules.conf Feb 12 16:47:21 pwm, pru and spi stuff Feb 12 16:47:35 weird Feb 12 16:48:14 it's normally almost never necessary to load modules this way Feb 12 16:48:27 at least not drivers Feb 12 16:49:02 failed to restart the service, but no useful messages as far as I can see Feb 12 16:49:13 nothing in kernel log either? Feb 12 16:49:38 I guess you could try increasing the kernel log level to debug and then try again Feb 12 16:49:41 can you remember how to disable pager on the journal? It's truncating some lines Feb 12 16:49:54 --no-pager Feb 12 16:50:10 you can use left-right arrow keys in the pager btw Feb 12 16:50:42 it's fialing on loading those modules reuqired by the robotics cape config Feb 12 16:50:51 I'll get rid of it Feb 12 16:50:58 i.e. broken image Feb 12 16:51:28 this is a release image, not even a testing image? Feb 12 16:51:32 yep Feb 12 16:51:41 ok, now the service starts ok Feb 12 16:51:43 I'd suggest yelling at rcn Feb 12 16:51:50 but a reload doesn't get me usb0 Feb 12 16:51:54 I'll tryanother reboot Feb 12 16:52:16 what's the best way of giving that feedback? the google group? Feb 12 16:52:19 no like I said, I don't really expect systemd-modules-load.service to have much to do, if anything at all Feb 12 16:53:16 yeah, dang same probelm Feb 12 16:53:25 ok maybe I'll try an older image Feb 12 16:53:54 I think the lxqt images don't get anywhere near as much testing as the iot images Feb 12 16:54:32 do you know what the differences are? I think I instictively avoid things with iot in the name! Feb 12 16:54:53 lxqt = iot + desktop environment Feb 12 16:55:09 i.e. nearly 2G of gui crap Feb 12 16:55:28 but what is the iot bit mean? Feb 12 16:55:47 it's just called that because it doesn't include a gui Feb 12 16:55:51 ok Feb 12 16:57:09 thanks for the help; Feb 12 16:58:02 i.e. most suitable for devices on the Internet of Things-that-shouldn't-be-on-the-internet Feb 12 16:58:05 ;) Feb 12 16:59:22 aka Idiots on TCP, aka Internet of Terror thanks to the DDoS attacks they make possible Feb 12 17:01:19 ;) Feb 12 17:06:53 ok, well usb0 and usb1 are both available in the iot release Feb 12 17:11:57 anyway, yeah you can try the google groups, or even just mail him directly Feb 12 17:12:20 ok Feb 12 17:13:03 I think a clearly broken image is enough reason to mail him directly Feb 12 17:13:37 hmm Feb 12 21:08:46 hello i am looking forward to contribute to pypruss , which mentor should i refer? Feb 12 21:17:50 I am plugging in my BBG and trying to make my software run w/out the cape attached to see if the software is collective. Feb 12 21:48:16 Motor Bridge Cape? Feb 12 21:55:46 Are there any people out there working on a Moto, Motor, H-Bridge Cape for the BBB/BBG? Feb 12 21:57:01 <<<< lazy or unknowledgeable Feb 12 21:57:23 If I am both, work w/ me. Feb 12 21:59:31 Motors and flailing grass and Trollie style ac/dc electricity. Come one, come all. If not for me, do it for the BBB. Feb 12 22:02:05 ... Feb 12 22:02:08 Seriously. Feb 12 22:02:56 I have dealt w/ Seeed and, by all means, their Motor Bridge Cape is cool/collective but I want POWER and MORE of it. Feb 12 22:03:02 ... Feb 12 22:03:23 I checked the Moto Cape from Circuit Co. I think they do not mfg. it anymore. Feb 12 22:05:26 heh? Feb 12 22:42:45 we are looking to bring some capes back. github.com/beagleboard/capes Feb 12 22:43:56 CCo doesn’t make anything at all anymore, as far as I know. Feb 12 22:47:36 +1 on bringing capes back Feb 12 22:47:50 maybe sparkfun or adafruit? Feb 12 22:48:34 well, lots are still out there, but nice to have some basic ones. Feb 12 22:49:19 starting with GHi. Adafruit still makes one or two. Feb 12 22:49:24 yes indeed Feb 12 22:49:49 well, E14 makes the official Robotics Cape and will do more bb.org official capes soon too. Feb 12 22:50:02 okay, time to get off the plane. Feb 12 22:53:38 Okay. Feb 12 22:54:39 I like motors. Things change and the Capes must be updated or debugged. I say updated! Feb 12 22:58:48 Okay...while I work on ideas and fail, I will wait for other, more knowledgeable, people to pick up. Feb 12 22:58:56 ... Feb 12 22:59:02 <<<<<< on break Feb 12 23:58:38 I just ordered all the shit required to build a Cryptocape, was gonna leave some in the drawers at i3. Feb 12 23:59:26 why a cryptocape? Feb 13 00:00:35 Because smartcards and other secure credential storage / authentication chips have always fascinated me Feb 13 00:00:57 ok Feb 13 00:00:59 and on the one hand, I'm not the software genius to fire all that up and do interesting things with it Feb 13 00:01:31 but on the other hand, I can make sure the hardware is available and easily integrated in case someone more software-clueful wants to tinker with it, the barrier to entry is as low as possible Feb 13 00:03:02 most of what I do (in my role as zone-coordinator of the electronics lab at a hackerspace) is enable other people's explorations and try to put interesting pieces in readily-accessible places so others can put the pieces together Feb 13 00:03:03 of course its usefulness is somewhat limited... while it may provide a secure environment for the secrets it holds, being able to freely perform crypto operations with those secrets is pretty much just as useful Feb 13 00:04:26 Ahh, another Slackbot that's gonna burn through its allocation of API tokens in 8 minutes flat :D Feb 13 00:06:03 lol Feb 13 00:23:08 My motor Bridge Cape for the BBG does not work right now. Is it my fault...most likely. Will I ask for help...yes. Feb 13 00:23:13 Help! Feb 13 00:28:45 Here I go explaining things again... Feb 13 00:28:47 So... Feb 13 00:29:38 I tried this: import Adafruit_BBIO.GPIO as GPIO Feb 13 00:30:04 I tried this afterwards: import Adafruit_I2C as I2C Feb 13 00:30:50 Issue Numeral Uno: no mas Adafruit_I2C. Now, it is import Adafruit_GPIO.I2C as I2C instead. Feb 13 00:30:53 ... Feb 13 00:31:17 I did this idea and changed the software on my .py library and the file that calls my library. Feb 13 00:31:20 So... Feb 13 00:32:36 Now...I have to configure a bunch of stuff that is out of my realm of understanding. C/C++ stuff is not for me. Feb 13 00:36:02 ... Feb 13 00:36:06 Off to cry in a corner. Feb 13 01:51:38 https://i.imgur.com/TUyagC0.png this is the signal i am getting Feb 13 01:58:50 let's see Feb 13 01:59:56 okay this is really difficult to analyze with the clock and data graphs so far apart... Feb 13 02:00:02 * zmatt opens gimp to overlay them Feb 13 02:08:13 uhh, why is your scope decoding the data in a nonsense way? Feb 13 02:08:39 or, at least the bar immediately below data is Feb 13 02:09:13 the selected line also doesn't match what I'm seeing here Feb 13 02:10:02 the first three bytes are is 00 00 36 Feb 13 02:12:06 Bongobong: https://liktaanjeneus.nl/spi-decoded.png (the vertical lines separate the bytes) Feb 13 02:13:37 00 00 36 01 Feb 13 02:14:34 yeah Feb 13 02:16:00 https://i.imgur.com/Gkccs9B.png Feb 13 02:16:57 so what problem are you seeing here? this looks like a fine spi signal to me Feb 13 02:18:55 it shouldn't be varying Feb 13 02:19:00 it should all be the same Feb 13 02:19:52 the same every 4 bytes? Feb 13 02:20:14 I mean, debug the software I guess? I see nothing indicating a hardware issue here Feb 13 02:20:53 alright Feb 13 02:23:05 I still have a pre-load library lying around to trace spi traffic done by a program, maybe it can be of use: https://liktaanjeneus.nl/spi-trace.so.tar.gz Feb 13 02:23:25 usage would be LD_PRELOAD=/path/to/spi-trace.so ./your-program Feb 13 02:26:02 awesome thank you **** ENDING LOGGING AT Tue Feb 13 03:00:02 2018