**** BEGIN LOGGING AT Thu Aug 20 02:59:59 2015 Aug 20 03:19:37 hmmm the IO mapping configuration on a beagle bone - X series is handled via an EEPROM correct? (IE for the main board the HDMI connector and the on board IO is mapped in via an EEPROM on the main board)? Aug 20 04:01:53 I'm running 4.1.1 with piaware (RTL-SDR) the BBB crashes ~ every 4 days (blue blinken lights stop blinken, no screen, network lights still flash). Only logging seems to be: Aug 20 04:01:54 Aug 17 07:59:21 arm kernel: [ 40.971118] musb_ep_program 896: broken !rx_reinit, ep2 csr 0003 Aug 17 07:59:21 arm kernel: [ 40.977345] musb_ep_program 896: broken !rx_reinit, ep2 csr 0203 Aug 17 07:59:21 arm kernel: [ 40.983500] musb_ep_program 896: broken !rx_reinit, ep2 csr 0003 Aug 17 07:59:21 arm kernel: [ 40.989650] musb_ep_program 896: broken !rx_reinit, ep2 csr 0203 Aug 17 07:59:21 arm kernel: [ 40.995800] musb_ep_prog Aug 20 04:02:11 I've seen a few other posts that seem maybe related with no real solutions... Aug 20 04:02:14 any ideas? Aug 20 05:40:45 Hello Aug 20 05:41:02 Can anyone tell me how to enable 4 pwm signals on the BBB Aug 20 05:44:50 ashwin: if you look at the SRM and TRM, you'll see that the BBB has 3 (three) PWMs. If you need more you'll need to add a PWM chip and controll it e.g. over I²C Aug 20 05:53:15 or one could PRU it Aug 20 08:27:58 Hi, i'm writing a device driver for the BBB, i've correctly define my DTB with references to the GPIO for example : "gpios = <&gpio1 28 GPIO_ACTIVE_HIGH>;", i've some difficulties trying to acces gpio i've set for the device, i'm reading leds-gpio.c driver I can see : "devm_get_gpiod_from_child" can you give me some advice for documentation to retrieve the gpio data from the DTB ? Aug 20 12:03:18 any thoughts on http://beagleboard.org/approved ? okay idea? Aug 20 12:10:03 jkridner, Looks like a good idea to me. Is there any support required from the developer in order to maintain approval status? That might help keep orphaned projects at bay. Aug 20 12:12:21 Though perhaps that's covered by just having the code in the beagle-image-builder. Aug 20 12:20:32 jkridner: It would be great if you could put in a statement that requires the vendor to publish sourcecode for the hardware under some reasonable license. Aug 20 12:20:54 (Mostly to avoid products with binary only drivers) Aug 20 12:20:55 gferencz: kinda hard to remove it from packaging. :( Aug 20 12:21:43 Chrille: interesting... hadn't thought about it being for open hardware only. enclosures, for example. do they need to be open hardware? Aug 20 12:22:14 jkridner: Not an issue with cases and power supplies - but for capes and wifi adaptors it's a real issue Aug 20 12:22:19 Chrille: we can avoid approving devices that require binary blobs to function properly, though that's difficult when it comes to wi-fi adapters. :( Aug 20 12:22:34 Easy solution to that. Aug 20 12:22:42 Just make an OSHW wifi adapter. Aug 20 12:22:46 SImple, really. ;-P Aug 20 12:23:01 I haven't seen that be a real issue with capes, unless you refer to FPGA capes requiring closed source *tools*. Aug 20 12:23:57 jkridner: You should not require the vendor to publis the full hardware design, but it should be possible to build any code from source Aug 20 12:24:25 Chrille: k. Aug 20 12:36:17 Good day, i had 3.2" tft color lcd module rev 5.0 16 bits. I need to interface with BBB in parallel i need circuit with code Aug 20 12:54:26 Good Day,I have 3.2" tft ILI9325c lcd module. i need to interface with BBB in parallel 16 bits could u prefer any circuits and code Aug 20 12:55:19 victortyau: Yeah, you asked that already. Aug 20 12:55:31 Damn. Aug 20 13:00:49 HELP Aug 20 13:01:13 victortyau: Maybe stick around a bit, and give us more detail on what you want. Aug 20 13:01:23 Or vinoth Aug 20 13:01:27 Not victortyau Aug 20 13:03:10 Good Day,I had 3.2" ILI9325c TFT lcd module. Need to interface with BBB in parallel 16 bits. Could u prefer any circuits and sample codes Aug 20 13:03:20 Right, you have said that three times now. Aug 20 13:08:50 probably a script on a timer, agmlego .. suggest we ban it ;) Aug 20 13:09:14 Right? Aug 20 13:11:26 any related works on tft lcd interfaces, Aug 20 13:11:59 vinoth: What have you tried already? Aug 20 13:13:02 I've coupled a BBB to a chalkboard tft using a rgb-to-lvds serializer chip .. its not -that- hard if you have a basic grasp of electronics and linux :p Aug 20 13:13:32 but if you want somebody to hand it to you on aplate, ready-working .. go check the internet or pay someone. Aug 20 13:14:01 we can/will give you pointers, but we're not going to do the hard work Aug 20 13:14:28 I suggest a lcd cape will take some of the burden away Aug 20 13:18:14 sorted :D Aug 20 16:19:01 hi all Aug 20 16:19:33 can anyone give me some info about how to connect the beaglebone black with an external device Aug 20 16:52:19 that was vague... Aug 20 17:12:24 How does one insert a custom dtbo into the kernel? Aug 20 19:03:32 MacGyverX: it needs to be in /lib/firmware and then the CapeMgr can load it by properly specifying it to the 'slots' interface. Aug 20 19:04:00 MacGyverX: where possible, use cape-universal and config-pin instead. Aug 20 19:04:08 My issue is it loads the one in the firmware instead of mine Aug 20 19:05:20 Even when I tell it not to load it in uEnv.txt and load mine, it loads it in slot i.e. 7 instead of slot0. Aug 20 19:07:52 I found a tool that complies the dts to dbto and creates an image and mv the image in the /boot/ folder, but it continues to load the stock kernel Aug 20 19:08:05 (custom dts) Aug 20 19:08:33 The end goal is to replace the ‘stock’ dtbo with mine. Aug 20 19:16:06 MacGyverX, uboot determines the device tree that gets loaded. Check out uEnv.txt. That is where you specify uboot options. Aug 20 19:16:44 ‘dtb=“ variable? Aug 20 19:16:52 If so, Mine is blank Aug 20 19:34:19 MacGyverX: yeah, just uncomment that line and set it to the name of your binary. Aug 20 19:35:40 Has anyone had USB issues lately? I'm using a USB cell modem, an I am seeing it disconnect from USB fairly frequently. Aug 20 19:36:02 :~$ sudo tail -f /var/log/syslog|grep "USB disconnect" Aug 20 19:36:03 Aug 20 18:51:40 ironhide kernel: [ 1475.070014] usb 1-1: USB disconnect, device number 4 Aug 20 19:36:03 Aug 20 18:53:17 ironhide kernel: [ 1572.352596] usb 1-1: USB disconnect, device number 5 Aug 20 19:36:03 Aug 20 18:56:12 ironhide kernel: [ 1746.707008] usb 1-1: USB disconnect, device number 6 Aug 20 19:36:03 Aug 20 19:16:58 ironhide kernel: [ 2992.752369] usb 1-1: USB disconnect, device number 7 Aug 20 19:36:03 Aug 20 19:18:21 ironhide kernel: [ 3075.867077] usb 1-1: USB disconnect, device number 8 Aug 20 19:36:05 Aug 20 19:32:55 ironhide kernel: [ 3949.652947] usb 1-1: USB disconnect, device number 9 Aug 20 19:36:14 You mean the name of my custom dtbo? Aug 20 19:36:31 yes Aug 20 19:37:01 actually you want a dtb, not a dtbo... Aug 20 19:37:25 That’s what I thought. Aug 20 19:38:28 a dtbo is an overlay I think... Aug 20 19:38:47 it is. Aug 20 19:45:02 abferm: usb cell modems usually consume an insane amount of power, out of the usb spec Aug 20 19:50:42 Power should not be an issue, unless the BeagleBone has issues passing it through the usb v_bus. Should be a little more than 3 A available on the 5V line. Aug 20 19:54:55 So I guess I should revise the question to, how does one add a dtbo to an existing kernel image- To overwrite one that already in it. Aug 20 19:57:27 MacGyverX: I haven't done anything with overlays, I've just built full dtb files. Aug 20 20:14:14 gustavoz: I have 4 devices set up exactly the same, and only one of them is having issues. Aug 20 20:15:41 abferm: from personal experience it usually boils down to traffic, make & model Aug 20 20:18:43 power consumption on them isn't constant, they can overshoot past the standard 500 mA easily and if the board has protection (polyfuse, etc) that can trip Aug 20 20:24:56 gustavoz: like I said, I have 4 of the same model connected up to different beaglebones. They all have the same traffic, etc. so I doubt it is a power issue. I also don't think that they have fuses of any kind on-board. Aug 20 20:27:58 abferm: the BBB has a tps2051 to handle the power, which limits to 500 mA AFAIK Aug 20 20:28:23 https://github.com/CircuitCo/BeagleBone-Black/blob/master/BBB_SCH.pdf?raw=true page 4 Aug 20 20:34:29 gustavoz: according to the datasheet for that part, it should limit to ~1.18A. that should be more than enough for my modem. Aug 20 20:36:07 abferm: short-circuit output current != continuous current Aug 20 20:36:44 also which modem make/model are you talking about? and also modem consumption can vary by cell association, signal level and so on Aug 20 20:37:10 it's not uncommon for 3g/4g modems to peak at 1A easily Aug 20 20:37:24 [Sierra Wireless, Incorporated] MC7750 Aug 20 20:37:31 4g Aug 20 20:38:35 but it should be under the same conditions as the other 3 modems, which are not showing this issue. Aug 20 20:43:37 is that the minipcie variant? Aug 20 20:45:18 USB Aug 20 20:46:47 looks like minipcie, but I don't think it actually is. The carrier board I use it with just breaks out the usb pins Aug 20 20:48:12 abferm: small secret, the mini pcie modems are actually usb :) Aug 20 20:48:25 they just made them in that format for internal laptop, notebook, etc uses Aug 20 20:48:57 according to data peak current is 1.0/1.2A Aug 20 20:49:27 also they work on 3.3V so the adapter down-regulates from USB 5V so there's some loss there Aug 20 20:49:35 how much depends on the quality of the regulator Aug 20 20:49:45 s/quality/efficiency/ Aug 20 20:56:11 joy... It is looking like that may be the issue. It isn't really a problem at lower temperatures, because the tps cools off and the modem connects again, but our aplication is designed to live in an industrial environment. I was testing at 60 C and I was seeing the disconnects happen 1-2 times a minute. That makes it very hard to establish a connection and transmit data. Aug 20 20:56:47 ahh that makes sense, power consumption normally up with temperature Aug 20 20:57:28 if you want to play a little you can desecrate a bbb and just short the safety measure (the tps2051) to see if that's the culprit Aug 20 20:57:33 and the tps trip current is also affected by the temperature. Aug 20 20:58:31 hmmm, I'm borrowing these from our production stock, so I need to get permission before doing any hardware modifications. Aug 20 20:59:54 some of the minipcie-usb adapters have an ultracap to dampen the problem, which is more similar to the typical use for those modems (battery-backed) Aug 20 21:00:13 also our BBBs are conformal coated, so that might make a big mess. Aug 20 21:04:58 there is a pretty large ceramic cap on the carrier board, but I doubt it has a high enough capacitance to deal with spikes that high. Aug 20 21:10:50 <[Butch]> Sorry for barging in on a really cool discussion, but I have a fairly (I hope) simple question: How does one upgrade the debian version on a BBB _without_ reflashing? I have a custom environment and I'd like to upgrade the OS without losing everything else I've ever done. It's all on microSD, if that matters at all. Aug 20 21:26:25 [Butch] you could conceivably just change the apt sources, and do a "sudo apt-get update && sudo apt-get dist-upgrade" Aug 20 21:26:54 dist-upgrade upgrades your main packages only. Aug 20 21:27:28 I would make a backup of your system first because while this is supposed to be a smooth process it can break things. Aug 20 21:32:01 <[Butch]> abferm: What's the right repo for the latest and greatest? Aug 20 21:33:27 <[Butch]> abferm: sources.list currently looks like this: https://gist.github.com/anonymous/1cf0d2f6d0c9dfe87869 Aug 20 21:44:59 <[Butch]> abferm: I have to run -- can you please check this gist and follow up privately? Thanks! Aug 20 21:45:25 <[Butch]> abferm: https://gist.github.com/anonymous/8e7e390aefb3e9c988af Aug 20 21:45:34 <[Butch]> abferm: Thanks! Aug 20 21:45:54 abferm: use a usb extension and splice the power wire to an external 5v source? Aug 20 23:35:39 hello :) -- has anyone ever dealt with AM335x's sd/mmc controller on the BBB? i am a bit confused about how the auto-initialization sequence works **** ENDING LOGGING AT Fri Aug 21 02:59:58 2015