**** BEGIN LOGGING AT Sat Sep 22 03:00:00 2018 Sep 22 08:20:01 hello.. I need some help interfacing this LCD ( https://www.digikey.ch/product-detail/en/kyocera-international-inc/F-51320GNB-LW-AEN/73-1285-ND/969373 ) to bbb. Any similar projects that you are aware of? Sep 22 10:24:50 thats a pretty unique screen to interface Sep 22 10:25:45 it uses a memory like protocol for an interface and i don't believe there is a driver for that Sep 22 10:25:48 and its monochrome Sep 22 10:26:12 whereas the BBB is generally better suited for RGB interface, 3 color TFTs or something like that Sep 22 10:26:16 or hdmi, obviously Sep 22 10:26:39 (sorry, talking to you umbaman) Sep 22 10:27:28 to get this interfaced to the bbb, i've never seen anyone do anything like this, you have to do some lowlevel work somewhere Sep 22 10:27:35 how do you intend to program/drive it? Sep 22 10:27:51 like software wise Sep 22 10:45:44 probably low-level i/o manipulation Sep 22 10:47:04 I will start from there. Setting high and low various gpios, that is Sep 22 10:48:31 and I will make a class out of it. Sep 22 10:50:25 Maybe check how I will check how a microcontroller is driving these things, and make the coresponding actions Sep 22 10:52:32 I can't think of anything else Sep 22 13:16:53 ayjay_t: actually, the lcd controller peripheral does have a mode to interfacing with displays that use (one of several) bus protocols, so at least you wouldn't have to manually bit-bang it using gpios Sep 22 13:19:30 he'd still either need to write a kernel driver or use uio to use lcdc directly from userspace Sep 22 13:31:23 (py-uio includes a class for lcdc already, albeit untested) Sep 22 19:10:18 interesting Sep 22 19:10:33 umbaman i would use zmatts suggestion Sep 22 19:10:51 i knew the peripheral itself was very flexible but TI's LCD driver seems to support mostly RGB protocol Sep 22 19:11:08 and that uses four timing signals instead of the bus-like write/hold signals Sep 22 19:11:46 py-uio is probably your best bet for the low-level work and zmatt is very helpful, and would probably be very happy at this point to help some people using his work and also making some effort themselves :P Sep 22 19:12:08 umbaman isn't here Sep 22 19:12:18 that's why I replied to you and not to him Sep 22 19:16:06 we don't have a bot that supports !note do we.... Sep 22 19:16:18 !note Sep 22 19:16:21 -shrug- Sep 22 19:57:48 cu Sep 22 21:52:07 hello. Is there in the meantime now a free GPU drive available for the beaglebone black? Sep 22 21:59:18 *driver Sep 22 22:11:09 hen_: as far as I know there are no serious efforts to reverse engineer the drivers or otherwise document the gpu, let alone make an open source driver for them Sep 22 22:12:19 thats sad. is there any plan to make a beaglebone black with a NXP i.mx8 cpu + vivante gpu? Sep 22 22:12:34 then i could finally have a free driver gpu output Sep 22 22:12:47 ehm, no? that wouldn't be a beaglebone, it would be some random board Sep 22 22:13:34 is there any new TI-CPU with a free-driver capable GPU chipset? Sep 22 22:14:09 why do you need 3d graphics on a board like this, which has pretty minimal graphics capabilities to begin with? Sep 22 22:16:09 for a propper working desktop environment in embedded usecases as a standalone machine with usb-keyboard+mouse Sep 22 22:16:42 x11 doesn't require opengl Sep 22 22:17:05 you mean vesa? Sep 22 22:17:06 and a desktop environment runs pretty awful on the bbb anyway Sep 22 22:17:28 because with the vesa driver you cant really run anything Sep 22 22:17:36 ? Sep 22 22:17:53 vesa is a thing on x86 PCs Sep 22 22:18:13 ok, maybe i am not up to date. what is the recent state of the gpu inside the bbb when using the latest kernel? Sep 22 22:18:32 https://xfce.org/ sudo apt install xfce4. Sep 22 22:18:34 (without any closed sourc drivers enabled) Sep 22 22:18:46 This works better than lxqt. Sep 22 22:18:57 maybe I should clarify something: on PCs it's common for the GPU to be responsible for the display output. this is not the case on the beaglebone or any other TI SoC Sep 22 22:19:09 the display driver is completely open source and mainline Sep 22 22:19:10 It just takes up a bunch of room. Sep 22 22:19:15 the gpu is purely for 3d rendering Sep 22 22:21:18 so when you use xfce/lxde with and without binary blobs, there is completely no difference in the speed or something else because nither of those use opengl for rendering? Sep 22 22:21:35 the proprietary drivers are not even compatible with x11 Sep 22 22:21:57 Oh and can I please talk smack about this city and how they do not have a 100W power supply in any store. Sep 22 22:21:58 ... Sep 22 22:21:58 (although someone is working on that) Sep 22 22:22:08 Aw! Sep 22 22:22:37 hen_: xf86-video-modesetting is the appropriate x11 video driver to use Sep 22 22:23:26 so this image here dont even include the proprietary driver, because it wont have any use of it? http://debian.beagleboard.org/images/bone-debian-9.4-lxqt-armhf-2018-06-17-4gb.img.xz Sep 22 22:24:16 the gpu drivers are not included on any of the beagleboard.org images as far as I know Sep 22 22:27:32 i think its also logical so say that when the proprietary driver that dont support x11 also dont support wayland(that require opengl support by design), right? Sep 22 22:27:46 they do support wayland Sep 22 22:28:48 (also, just to be precise, the drivers don't implement OpenGL but OpenGL ES (1 and 2), along with EGL) Sep 22 22:28:56 That makes sense because wayland is the future. Are those drivers recently developed? Sep 22 22:29:47 basically TI doesn't care about x11, or desktop users whatsoever. they're focussing on industrial and automotive applications Sep 22 22:30:27 are there any iso images to test that make use of this driver together with wayland? Sep 22 22:30:48 TI's Processor SDK for Linux Sep 22 22:31:22 probably this here: https://www.ti.com/tool/PROCESSOR-SDK-AM335X Sep 22 22:31:26 http://software-dl.ti.com/processor-sdk-linux/esd/AM335X/latest/index_FDS.html Sep 22 22:31:29 yeah Sep 22 22:33:25 thats not really thrustworhty when they dont know how to even configure a webserver... https://software-dl.ti.com/processor-sdk-linux/esd/AM335X/latest/index_FDS.html Sep 22 22:34:11 why are they even trying to use the akamai.net certificates on a ti-website with completely different domain name?? Sep 22 22:34:12 it's only a configuration mistake if they actually use http links to that page Sep 22 22:35:49 the akami cert is probably a fallback when the server doesn't have an ssl cert for that domain? Sep 22 22:36:03 note that the server is run by akamai, not TI Sep 22 22:36:42 maybe you know the relevant people better to inform about this issue Sep 22 22:37:55 because i cant even download the software with tls enabled. Downloading a software you run on your computer from an insecure connection is nothing to do any more in 2018 Sep 22 22:37:56 https://software-dl.ti.com/processor-sdk-linux/esd/AM335X/latest/exports/ti-processor-sdk-linux-am335x-evm-05.00.00.15-Linux-x86-Install.bin Sep 22 22:38:42 I agree, but my voice carries no more weight than yours Sep 22 22:39:34 i dont even know who to inform and the relevant email address for that Sep 22 22:43:51 do i still need 0001-am335x_evm-uEnv.txt-bootz-n-fixes.patch when building u-boot? Or have u-boot now propper bbb support? Sep 22 22:44:01 https://e2e.ti.com/support/site_support/ would be my best guess Sep 22 22:45:16 you mean for the u-boot or the ssl part? Sep 22 22:45:43 to recreate u-boot as used on beagleboard.org images you should use both 0001-am335x_evm-uEnv.txt-bootz-n-fixes.patch and 0002-U-Boot-BeagleBone-Cape-Manager.patch Sep 22 22:45:49 that was about the ssl thing Sep 22 22:48:13 mainline u-boot has proper bbb support, but that is not sufficient to be compatible with rcn's images since the boot script is different (and that is not going to change due to the need for backwards compatibility). moreover the latest images use u-boot overlays Sep 22 22:48:30 the patch may also add support for recent beaglebone variants that might not be in mainline yet Sep 22 22:49:49 you can probably make a working system using a mainline u-boot and mainline kernel, but you'd basically be on your own for that Sep 22 22:52:48 i posted the tls issue. i got this: You have posted to a forum that requires a moderator to approve posts before they are publicly available. If the administrator has configured this forum to support email notifications you will receive an email when your post is either approved or denied (if you have emails enabled in your profile). Sep 22 22:54:28 spam protection no doubt. the message is badly worded since it's not the forum in general that requires it (most likely) but more likely because it's your first post, especially since it probably contains a link Sep 22 22:55:06 if someone wants an anonymous account to post issues on the ti-websites: username/email: bruxxtjw@sharklasers.com PW: bruxxtjw Sep 22 22:55:46 that's asking for a ban on e2e Sep 22 22:56:49 i am not posting spam or something other crap like that and other users who are here in the irc probably are also not bots crawling for usernames and passwords ;) Sep 22 22:57:11 there are bots that publish irc logs on the internet Sep 22 22:58:12 I also don't see why anyone would care for your account information on e2e Sep 22 22:59:06 every account i create is for the public and not for me. The more people use the same account, the more private each user of this account gets. Sep 22 23:00:22 the more likely the account will get banned and its complaints ignored Sep 22 23:03:48 lets hope not. but probably noone would start using the account because the few people who reads that in most cases wont need it. but maybe there is this one user who needs it. Sep 22 23:04:33 there is none free gpu for any PowerVR GPU? https://en.wikipedia.org/wiki/Free_and_open-source_graphics_device_driver#Imagination_Technologies Sep 22 23:06:03 nvidia, amd, vivante, arm, matrox, VIA, qualcomm, broadcomm, ... they all have some sort of a free driver for some or even most chipsets. For PowerVR there are 0? Sep 22 23:07:42 I think broadcom sponsored an open source driver Sep 22 23:08:14 yes Sep 22 23:08:32 they pay a full time gpu driver developer now Sep 22 23:09:40 eric unhold (dont know if i wrote his name right) Sep 22 23:09:41 imgtec has no interest in doing so apparently, powervr is not very widespread (at least on hardware where potential devs have a personal need to get 3d graphics working), and its architecture (tile-based deferred rendering) also requires a high initial investment of effort to get *any* result afaik Sep 22 23:10:23 there was a powervr based Intel Atom cpu - that was so terrible with linux... Sep 22 23:11:32 i see even few of those are with powervr: https://en.wikipedia.org/wiki/Atom_(system_on_chip) Sep 22 23:21:17 ah, i get the information about imagination. the reason why no developer have spend time into development of those drivers is probably because the company would die out soon Sep 22 23:22:23 here the explaination: https://cdn.media.ccc.de/contributors/berlin/datengarten/webm-hd/datengarten-88-eng-Datengarten_88_webm-hd.webm Sep 22 23:23:31 would new TI chipset be released in near time with other gpu chipset? Sep 22 23:26:50 very unlikely, especially since the GPUs they use right now seem to be adequate for their needs and they have already expended the development investment into them Sep 22 23:29:10 and I doubt that open source drivers would have any impact on the company. they make hardware, they don't sell drivers Sep 22 23:29:35 *design hardware Sep 22 23:36:37 what about 64bit cpus and so on? Wont they continue the development of the cpus? Or even make risc-v licence free cpus? Sep 22 23:37:35 I'm not privy to any inside information about what TI is planning or developing Sep 22 23:40:35 ok, your irc support is always great here. Every single time i had a question about the bbb you had an answer. Thats really geat and helped me a lot in my bbb projects Sep 22 23:41:02 :) Sep 22 23:47:02 maybe you or someone else could add the missing information here to the beaglebone part: https://en.wikipedia.org/wiki/Comparison_of_single-board_computers Sep 22 23:47:44 things like speed of the memory and so on are missing. and i know many people using this wiki page for choosing their board for a project Sep 22 23:48:44 beaglebone wireless is fully missing, other things are also missing Sep 23 01:29:37 How can I calculate what wattage I will need from a AC/DC power supply if I do not know the wattage I need for a project? Sep 23 01:29:38 ... Sep 23 01:30:15 The only info. I have is that I need 100w over "60w" b/c of some additions and that "60w" may not be correct. Sep 23 01:30:28 This is off subject but my books are failing me. Sep 23 01:34:13 For instance...the ac/dc power supply that supplies the board w/ power needs to now be 100w higher than it used it be. Sep 23 01:34:50 My machine puts out an amount energy. This energy cannot be measured (in my mind). Sep 23 01:38:52 this is maybe offtopic: my desktop pc makes wired noises when `cat /dev/zero` Sep 23 01:39:52 Coffee: What are you doing? Did you not read my exhaustive idea of a question? Sep 23 01:40:07 I have no info. and I have to make decisions. Sep 23 01:40:41 Sorry, I was trying not to forget such wired thing Sep 23 01:40:55 Oh. Sep 23 01:41:31 I have an entire machine that promotes energy in watts. Right? So, how can I measure this if my board is producing the outputs? Sep 23 01:42:44 oscilloscope? Math? both? Multimeter? Sep 23 01:45:39 two multimeters, one for current and the other for voltage (if you want to be precise) Sep 23 01:46:21 Oh. Sep 23 01:46:23 Okay. Sep 23 01:47:23 CoffeeBreakfast: that's caused by cavitation due to the rapid flow of bits Sep 23 01:49:15 zmatt interesting, I thought it was armonics from the PCB traces Sep 23 01:49:25 Oh. Sep 23 01:50:41 To put the idea into perspective: I have "every" math calculation at my reserve but my AC/DC power supply is 60W. I need 100W over 60W. I just cannot measure how much wattage is produced by the board. Sep 23 01:50:42 ... Sep 23 01:50:54 So, putting the idea you presented into action works? Sep 23 01:51:02 CoffeeBreakfast: my actual guess would be that it's causing power fluctuations made audible by inductors and/or capacitors in the power circuit Sep 23 01:51:50 ooohh piezoelectric effect Sep 23 01:53:22 from the decoupling caps? Sep 23 01:55:36 oh I hadn't considered that option... it might be possible? Sep 23 01:56:20 i'm not sure Sep 23 01:58:52 I suppose if the sound is coming from caps, decoupling caps would actually be the likely candidate Sep 23 01:59:11 scope time! Sep 23 02:00:06 The only reason I say an oscilloscope is popular is b/c of a 1979 organ I have in my possesion. It works only so much. Sep 23 02:00:07 ... Sep 23 02:00:22 I am still testing all forms of the machine. Sep 23 02:00:52 Luckily, the Kawasaki people, the mfg, gave out a How-To book on testing the motherboards. Sep 23 02:01:09 What to test for each connection and how to test it. Sep 23 02:01:14 That is the idea. Sep 23 02:01:26 Older boards are easier. Sep 23 02:01:34 Not layered... Sep 23 02:02:26 I was thinking of promoting a BBB into the project. Sep 23 02:02:41 But...I ran out of resouceful books on the subjects. Sep 23 02:02:52 synths and linux? Sep 23 02:10:39 set_: testing motherboards? Sep 23 02:11:35 Yes. Sep 23 02:11:40 and fixing them. Sep 23 02:18:51 So, if I had a board, say the BBB, w/ six (or whatever number) outputs, how could I measure the amount of watts produced by these outputs? Sep 23 02:33:48 I don't know the maximum ratings Sep 23 02:34:44 Coffee: Are you chatting w/ me? Sep 23 02:35:42 set_: the maximum current ratings is what you must be aware for Sep 23 02:37:08 If I know the input in volts (AC) is 120V and the output of the power supply is 12V at some wattage, is there a way to configure my device (the target board) to read... Sep 23 02:37:09 Oh. Sep 23 02:37:18 I need to know what watts the target board can take? Sep 23 02:38:05 So, like this... Sep 23 02:38:55 I have a board, the board has input power at 12v some Ah and then at some ____ watts. The target board then in return produces several outputs. Sep 23 02:39:37 The four steppers are then outputs that can be tested. Sep 23 02:40:09 oh, you mean the BBB plus the motorcape? Sep 23 02:40:46 And...there is another output...no, not the motorcape. The output is a heated bed and thermister (100k). Sep 23 02:41:09 It is some junky 3D printr project. Sep 23 02:41:54 zmatt: already told me how to test for volts w/ the multimeter at specific points. Sep 23 02:41:59 This is all find and dandy. Sep 23 02:42:14 ... Sep 23 02:42:29 It is not that simple, I think. Or...is it/ Sep 23 02:42:30 ? Sep 23 02:44:00 So, on motorOne for example. Would I just test volts and ohms and work from there? Sep 23 02:44:08 Then... Sep 23 02:44:34 Mtwo, Mthree, Mfour, and so on? There are only so many outputs. Sep 23 02:45:33 I think this would help me find my total or "maximum current ratings." Sep 23 02:45:34 Right? Sep 23 02:47:11 do you know the current consumption of the heated bed? Sep 23 02:47:18 Not yet. Sep 23 02:47:38 I can find out, though. Sep 23 02:47:57 I still need to find my wattage value before attempting that venture. Sep 23 02:49:15 So, for instance, my power supply from "MeanWellUSA.com" states something on it. Sep 23 02:49:31 It states 12v xxAh 60W. Sep 23 02:50:41 Now, I need 100W over so said "60W" but that is only if my the "60W" is a real value and not some pseudo-value. Sep 23 02:51:14 "60W" could mean anything. It could be over the rate of power I need or just about at it. Sep 23 02:51:15 ... Sep 23 02:51:21 I would like to know for sure. Sep 23 02:52:52 I mean, I have an overkill-360W power supply. Sep 23 02:53:14 I could use this for sure. The output is 12v still. Sep 23 02:53:30 I would rather get close to the range of watts that are needed. **** ENDING LOGGING AT Sun Sep 23 02:59:59 2018