**** BEGIN LOGGING AT Tue Dec 15 02:59:57 2020 Dec 15 11:06:53 Hello guys, I want to connect BBB to Keithley Dec 15 11:07:11 Thanks Dec 15 11:09:29 Connecting to Beaglebone black via GPIB in Python Dec 15 11:57:31 gpib? ew Dec 15 11:58:39 ethernet not an option? Dec 15 11:59:12 my one keithley instrument has it Dec 15 11:59:21 older ones might not Dec 15 12:00:09 yeah, I only know about GPIB because I've done some stuff with controlling an oscilloscope via network (VXI-11), and that protocol is essentially GPIB tunneled over TCP Dec 15 12:00:22 iirc Dec 15 12:00:45 makes sense that it would be Dec 15 12:01:25 the GPIB standard also remains vividly in mind because it contained the most inane statement I've yet seen in an IEEE standard... "The Input Buffer shall be equal to or greater than zero in length." Dec 15 12:01:33 modern tek and keithley instruments also talk something they call LXI Dec 15 12:01:58 oh no, why can't it be negative? Dec 15 12:05:13 LXI is a bit vague to me Dec 15 12:05:45 I've never had cause to look at it closely Dec 15 12:05:55 just seen it in the setup screens Dec 15 12:07:30 @zmatt yes GPIB Dec 15 12:07:55 keithley and ILX Dec 15 12:08:21 could you please help me ? Dec 15 12:08:27 mru: like, it doesn't actually seem to define a protocol to control devices... mostly just with synchronization and events, along with general requirements on a device... but I've only glanced at it Dec 15 12:10:26 Do I need to add some libraries? like import visa in raspberry? Dec 15 12:10:52 my main concern would be the electrical interfacing Dec 15 12:12:34 if that's taken care of I'm sure the beaglebone (and more specifically PRU) is capable of dealing with this interface, but it seems unlikely that there will be any existing code for it Dec 15 12:16:12 it seems like you could either put in the time to implement this yourself (design a suitable electrical interface along with PRU code to support it), or just get an off-the-shelf GPIB-Ethernet adapter Dec 15 12:20:43 (do note that the Prologix thing that's the top search result for that appears to implement a custom command interface via TCP rather than VXI-11.. that could either be a benefit or a downside depending on what you want to do with it I guess) Dec 15 12:22:11 there are also usb-gpib adapters Dec 15 12:22:11 TCP/IP you mean? Dec 15 12:22:52 no I meant TCP. the term "TCP/IP" is used to refer to the entire networking stack, not one specific protocol Dec 15 12:23:38 The Transmission Control Protocol Dec 15 12:24:09 I can talk to keithley just via GPIB Dec 15 13:02:36 oh wow, gpib actually has dedicated transceiver ICs, one for data and one for control signals (SN75160B (data) + SN75161B (control), or the newer but mostly identical SN75ALS160 + SN75ALS161) Dec 15 13:06:28 good Dec 15 13:07:55 And do you have any idea how can I read GPIB via BBB? Dec 15 13:10:42 I have to admit, the sort of questions you're asking don't inspire confidence that you'll have the ability to make this working (short of getting some off-the-shelf interface and hooking that up) Dec 15 13:11:15 do read up and follow some tutorials on the PRU unit's Dec 15 13:11:40 they're great little processors inside the sitara processors Dec 15 13:13:43 I still need to experiment with the EDIO peripheral in pruss, for bidirectional gpios (without the latency and jitter of accessing the SoC's GPIO controllers) Dec 15 13:14:44 latency, jitter, and race conditions I should say (the GPIO controllers don't support atomic direction change) Dec 15 13:15:39 hmm, not sure if there is enough room, but external directon control? one pin out, another pin in? Dec 15 13:16:34 that would require additional external hardware and more pins than available Dec 15 13:42:57 Hey zmatt, I have the following in my dts: https://pastebin.com/zJ9DLpYY . but it seems that usb1_en(DRV pin) is still high. Is there a simpler way to disable it without writing raw bit to mem location as : https://e2e.ti.com/support/processors/f/791/t/270060?USB1-power-control-with-DRVVBUS Dec 15 13:43:46 huh, DRVVBUS definitely shouldn't be high when configured to peripheral mode Dec 15 13:46:28 oh, hm Dec 15 13:47:13 just test it without my board, ... seems low... Dec 15 13:47:36 oh right, pocketbeagle, not bbb Dec 15 13:48:01 you're leaving the ID pin floating right? Dec 15 13:49:24 yea, but yea, it's low when i power without my prototype. Dec 15 13:50:28 that sounds weird, wtf is your board doing? Dec 15 13:50:41 btw, for more insight into what musb is doing you can enable its debug messages at runtime: Dec 15 13:50:43 bah!!! it's high on the prototype. Dec 15 13:50:44 echo 'file musb* +p' >/sys/kernel/debug/dynamic_debug/control Dec 15 13:51:12 .. but DRVVBUS is an output of the am335x, it should be high-impedance on your board Dec 15 13:51:50 actually Dec 15 13:52:04 haha, it is! no idea what is going on, my only guess is that by registering the fusb302 that is attached to the other end of the port, it enables vbus somehow Dec 15 13:52:30 ohhhhh Dec 15 13:52:42 I completely forgot you were the one doing stuff with that usb c thing Dec 15 13:52:51 yup, it me. Dec 15 13:54:14 I'm a bit confused though... is the port supposed to be dual-role? if so, why did you configure it as peripheral, if not then why are you doing anything with drvvbus Dec 15 13:55:27 Peripheral to get to mvp, and drvvbus drives an "otg/vap en" input of the bq25710 Dec 15 13:56:36 if you're forcing musb into peripheral mode you'll probably need to do the same with the fusb302 Dec 15 13:57:50 hmm, will try. Have https://pastebin.com/fFCwmr23 now Dec 15 13:58:16 also, what's your workaround for the lack of ID-pin output on the fusb302 ? is there some kind of software connection between the two drivers? does musb even support overriding the ID pin software? (I think it does but I'm not sure) Dec 15 13:59:22 I think it does, not sure, but in worse case scenario, i might be able to just hook up the id pin to a gpio Dec 15 14:00:19 this is a wip, There are a lot of things left to figure out. for now, I am focusing on getting peripheral mode working. Dec 15 14:01:26 hmmm, maybe i should just reconfigure the pin to gpio and control via software. Dec 15 14:01:38 that will just make the driver very upset Dec 15 14:02:39 since drvvbus being high means it has configured the port into host mode. if you override drvvbus and no +5V appears on VBUS as a result, the driver will think there's an electrical fault and start to flail Dec 15 14:03:22 the question is what the driver is doing and why Dec 15 14:03:34 the vbus should be okay, going to have the chg_ok pin pulled up to 5v and fed to the vbus line Dec 15 14:03:59 for which my first suggestion would be to use dynamic debug, which is exactly why I recommended it 15 minutes ago ;) Dec 15 14:04:02 uhhh what do you mean? Dec 15 14:04:15 vbus should not be high unless you're connected to a host Dec 15 14:04:26 (or you're a host yourself) Dec 15 14:04:29 yea, have dynamic debug enabled now, does it persist between powerups? Dec 15 14:04:55 no, but you can configure it to be enabled via kernel parameter Dec 15 14:05:00 yup, that matches the use-case of the CHG-OK pin of the bq25710 Dec 15 14:05:01 https://www.kernel.org/doc/html/v4.19/admin-guide/dynamic-debug-howto.html Dec 15 14:06:03 ok then I don't understand what you're saying Dec 15 14:06:58 the problem isn't drvvbus (the physical line) being high, the problem is that musb is switching into host mode (and thus drives drvvbus high) Dec 15 14:07:45 that's a good point, i was focusing too much on the line itself. Dec 15 14:07:59 normally musb does that based on the ID pin Dec 15 14:08:27 (floating = peripheral mode, grounded = host mode) Dec 15 14:09:47 hmm, i have the pin floating but it seems to be low when i probe the id pin on the prototype Dec 15 14:30:36 woot, "data-role = "device"; // or host, device, or dual" seems to have done it, great insight zmatt Dec 15 14:30:59 oops, scratch that, still high Dec 15 14:31:45 it probably doesn't tie in with the fusb connector settings of tcclass Dec 15 14:33:43 is fusb even relevant here? how would musb currently know about fusb's opinion on things? Dec 15 14:34:31 I didn't see any reference from fusb to musb or vice versa in your DT, and there's no ID pin interconnection, so now that I think about it I don't actually see how fusb could be influencing things Dec 15 14:35:36 through the port{ } correlation between the fusb and musb, but i don't think that works. Dec 15 14:35:47 oh, was I blind? Dec 15 14:36:14 oh it's referencing the phy Dec 15 14:36:24 or something? not sure Dec 15 14:36:24 still though, i dont think musb driver has any understanding of that, though i don't get errors from it. Dec 15 14:37:04 https://pastebin.com/J43YAv2f Dec 15 14:39:52 there's documentation/example to suggest the ti,am335x-usb-phy supports this? Dec 15 14:43:08 not that i saw from here: https://github.com/beagleboard/linux/blob/5.4/Documentation/devicetree/bindings/usb/am33xx-usb.txt. but i get warnings from fusb that i'm missing ports, so i added it. Dec 15 14:43:34 didn't see an error from the musb driver, I think it just ignores it. Dec 15 14:44:27 ok this is really weird, setting dr-mode to "peripheral" should actually cause the ID pin to be ignored Dec 15 14:44:52 anyway, it's a bit pointless to do blind speculation.. get debug messages Dec 15 14:45:10 maybe it is, I didn't try driving it to any state. Dec 15 14:45:31 still working on that. found the register dump so that's a start Dec 15 14:47:07 don't use musb-hdrc.*.dsps/regdump .. it causes a bus error Dec 15 14:47:09 :D Dec 15 14:47:40 ..... Dec 15 14:47:44 haha Dec 15 14:48:15 well, i'm looking at the musb-hdrc.1/regdump now Dec 15 14:49:37 what's the value of DevCtl ? Dec 15 14:50:05 19 Dec 15 14:50:58 it's in device mode and thinks it's attached to a host Dec 15 14:51:48 are you sure drvvbus is high? because that notion is kinda incompatible with this value of DevCtl Dec 15 14:51:57 3.44v Dec 15 14:52:05 P1_3 Dec 15 14:52:57 Could it be in Source-Open state ;) Dec 15 14:53:24 ehm, why the fuck is the voltage that high? Dec 15 14:53:34 magic? Dec 15 14:53:43 are you driving a voltage into that pin? Dec 15 14:54:39 shouldn't be, but good point, 3v3 rail is not that high Dec 15 14:54:45 and no, it should be a push-pull output Dec 15 14:56:03 note that the 3.3V rail on the expansion headers is not the same as the one internally used as I/O supply.. the external one can be slightly too high I think while the internal one should be more accurately 3.3v Dec 15 14:58:09 alright, other prototype starts at 3.3 and drifts up to 3.41 Dec 15 14:58:45 for comparison measure some gpio driven high on the pocketbeagle itself Dec 15 15:00:08 also 3.41 Dec 15 15:00:15 u0 tx Dec 15 15:01:57 weird Dec 15 15:02:20 pmic specifies ±2% accuracy on its LDOs Dec 15 15:02:45 I am driving higher v into the U0rx Dec 15 15:03:03 4.09v it seems Dec 15 15:03:08 wtf Dec 15 15:03:30 nvm, not the issue. removed and it's still 3.41 Dec 15 15:03:48 ehhh, it's a cheap cp2102 converter Dec 15 15:03:49 regardless of whether it's "the issue", it's definitely not something you want to be doing Dec 15 15:03:57 it will destroy the am335x Dec 15 15:04:00 yea, disconnected that. Dec 15 15:04:03 and frankly I'm surprised it hasn't yet Dec 15 15:04:14 I doubt that, it's a weak pull i believe Dec 15 15:04:23 it's 4.09V Dec 15 15:05:03 the silicon oxide layers get destroyed by static overvoltage, no matter the current Dec 15 15:05:31 isn't that what the steering diodes prevent up to a certain current limit? Dec 15 15:06:48 you mean the esd protection diodes? evidently they don't protect against continuous injection of whatever amount of current you're injecting since 4.09V generously violates the absolute maximum ratings Dec 15 15:07:00 esd protection diodes are designed to protect against esd, not against static overvoltage Dec 15 15:07:58 nagh, i meant the steering diodes... but not sure if the sitara has those. Dec 15 15:08:06 I don't know what you mean by that Dec 15 15:08:37 that sounds like the esd protection diodes Dec 15 15:08:38 anyway, update, i re-disconnected the tx from cp2102 and it did indeed drop to 3.29v Dec 15 15:08:58 okay so you were actually charging up the 3.3v supply? Dec 15 15:09:01 via the protection diode Dec 15 15:09:47 ahh, esd is generic name, tvs in parallel is esd protection, shottky to v rails is steering and also esd protection. Dec 15 15:10:00 yup Dec 15 15:10:08 also, you were talking about uart rx on the pocketbeagle so that will be a push-pull output on the cp2102, not a "weak pull" Dec 15 15:10:24 you were clearly in fact injecting a ton of current Dec 15 15:10:53 poor I/O cell Dec 15 15:12:18 not sure how much exactly i was injecting but it was under 20ma Dec 15 15:12:36 certainly not 2000-pounds ;) Dec 15 15:15:20 20 mA is 40% of the max current per _supply_ pin (for the I/O supplies) :P Dec 15 15:16:14 re: esd maybe not, Clamping/Steering diode don't seem to be considered esd, makes sense i think, you wouldn't want to push voltage spikes to any rails rather than just absorb it which a tvs would do. Dec 15 15:16:41 haha, alright, i would agree it was a lot. Dec 15 15:17:08 they're clamping diodes intended to protect against esd Dec 15 15:17:28 i.e. they are esd protection diodes, as in diodes whose purpose is esd protection Dec 15 15:18:20 anyway, steady state absolute maximum voltage on an I/O is vddio+0.3V Dec 15 15:19:06 ... how could a clamp protect against esd? wouldn't that just dump that voltage spike onto the rails? Dec 15 15:19:30 .. i guess if the pin is more sensitive than the rails that would make sense. Dec 15 15:19:32 which is the fastest path out of the chip Dec 15 15:19:50 and probably also has a lot of capacitance to absorb it Dec 15 15:20:02 at least compared to the I/O Dec 15 15:20:04 Alright, yea that makes sense. Dec 15 15:21:03 high-speed I/O need to have very low capacitance, but that also maens very little charge is needed to get a very high voltage Dec 15 15:21:21 yea, makes sense. Dec 15 15:23:41 hmm, re: devctrl, where can i find that bit meaning in that manual? Dec 15 15:23:43 anyway, still no clue what's going on with your drvvbus, except that the musb devctl seems to imply it thinks it's in peripheral mode (ID floating) and vbus went high on its own (i.e. a host connected) Dec 15 15:24:31 which also means it doesn't drive drvvbus high Dec 15 15:25:18 Konsgn: http://dl.linux-sunxi.org/users/tom/musbmhdrc_pspgUSB.pdf Dec 15 15:25:35 section 3.2.12 on pages 35-36 Dec 15 15:25:36 hmmm. perhaps i shouldn't have connected the vbus line to 5v_sw.... Dec 15 15:25:47 what's 5v_sw ? Dec 15 15:26:11 the pb 5v output Dec 15 15:26:25 vout Dec 15 15:26:55 correct, that makes you sense. but you said earlier that it was controlled by CHRG_OUT ? Dec 15 15:27:00 *makes no sense Dec 15 15:27:07 Thanks, that manual is a bit hard to look through. Dec 15 15:28:13 .. the vbus should be controlled by chg_ok from the bq charger, i haven't rewired it to that yet though. need to do some more circuit surgery. Dec 15 15:29:41 pocketbeagle vbus (P1.05) should be around 5V whenever you're connected to a host (as peripheral) or you're in host mode and sourcing power (at request of drvvbus) Dec 15 15:30:40 bah, going to need to do some surgery on layer 2.... Dec 15 15:30:47 of a 4 layer... Dec 15 15:34:03 id floating, vbus < 0.2V: peripheral mode, not connected to host Dec 15 15:34:19 id floating, vbus > 4.5V: peripheral mode, connected to host Dec 15 15:35:26 oh btw, the am335x will pull the ID pin up to 1.8V, it is not 3.3V-compatible Dec 15 15:36:23 or at least, its abs max voltage is 2.1V so I'm inferring it uses the usb phy's 1.8V supply Dec 15 15:36:39 "This terminal is connected to analog circuits in the respective USB PHY. The circuit sources a known current while measuring the voltage to determine if the terminal is connected to VSSA_USB with a resistance less than 10 Ω or greater than 100 kΩ. The terminal should be connected to ground for USB host operation or open-circuit for USB peripheral operation, and should never be connected to any ... Dec 15 15:36:45 ...external voltage source. Dec 15 15:37:47 so should you end up controlling it with a gpio you'll want to use a transistor or mosfet or something Dec 15 15:37:50 so open-drain on tied pin would be necessary. Dec 15 15:38:09 wouldn't open-drain be suitable Dec 15 15:38:33 brb, in surgery Dec 15 15:41:28 I don't exclude it might work but it wouldn't make me very comfortable. if you make a small mistake and accidently drive it high (easy eonugh to do especially since the am335x gpio controller has no real "open drain mode", you'd just be switching direction between input and output), and moreover you'd be holding that gpio at 1.8V continuously, i.e. an intermediate level between logic-low and ... Dec 15 15:41:34 ...logic-high (for that gpio), which it may not appreciate Dec 15 15:41:57 I also don't know if the AM335x GPIOs are sufficiently high-impedance (even with internal pull disabled) to avoid disturbing this analog measurement Dec 15 16:22:31 baaaaah Dec 15 16:23:14 forgot there is a diff between vbus and usb vin Dec 15 16:23:27 have vbus floating and usb vin was 5v. Dec 15 16:41:05 Konsgn: https://pastebin.com/nJCbVsaA better description of the pocketbeagle power/usb pins Dec 15 16:43:24 .... Dec 15 16:43:54 bah... if only there was an undo button for work done. Dec 15 18:08:33 hello BBB enthusiasts Dec 15 18:08:50 this is my first time here... looking to hire a BBB expert for automation with PLCs Dec 15 18:09:32 excited to be here... Dec 15 18:19:52 BBBuser: hi! do note that although there are generally plenty of knowledgable people here, activity may vary and it can take time to get a response (if any), especially for a question like this which has a rather limited audience Dec 15 23:38:46 hey everybody Dec 15 23:41:11 been some time since i've been here. i just updated my beaglebone black with BBB-blank-debian-9.5-iot-armhf-2018-10-07-4gb Dec 15 23:41:18 since when has root a password, Dec 15 23:41:19 ? Dec 15 23:52:18 Hello. Dec 15 23:52:36 Baeyens...the newer images are debian:temppwd Dec 15 23:52:53 debian is the name and temppwd is the passcode. Dec 15 23:55:18 yeah that i know, but i used to be able to just root in Dec 15 23:55:28 now i have to go the long way Dec 15 23:56:55 Right. Dec 15 23:57:27 I think people want to have security w/out the hoopla. I think you can erase the passcode for use if you do not care about the security issue. Dec 15 23:57:43 Let me see if I can find that link. Dec 15 23:58:37 https://elinux.org/Beagleboard:BeagleBoneBlack_Debian#i_take_full_responsibility_for_knowing_my_beagle_is_now_insecure is the elinux wiki ideas behind erasing the passcode *I think*. Dec 15 23:58:54 I have not tried it in a while. Dec 16 00:03:05 i only use it locally. so am not to worried about security Dec 16 00:03:16 normally i'm offgrid with my beaglebone Dec 16 00:03:39 but thanks :-) Dec 16 00:06:35 No issue. Hey, I am sure someone might be around later to better direct you if needed. Dec 16 00:06:43 just out off curiosity, the AM5729 Debian 10.3 2020-04-06 16GB eMMC IoT TIDL Flasher on a beaglebone black, would that work? i have tried, but came out on a 2015 version? Dec 16 00:07:11 Oh. try am335x only for the BBB. Dec 16 00:07:18 Not the am5729. Dec 16 00:07:37 The am5729 is for two, particular boards they produced. Dec 16 00:07:39 oh sorry, copied the wrong title. but still, got the same issue Dec 16 00:08:06 Ha. Try beagleboard.org/latest-images for the correct images they made available. Dec 16 00:09:25 "same issue" means what? Dec 16 00:09:32 i've downloaded bone-eMMC-flasher-debian-10.3-iot-armhf-2020-04-06-4gb.img - about 5 days ago Dec 16 00:09:56 when installing, got that cyclon leds, so assumed all was good Dec 16 00:10:04 Oh. I just use the SD Card versions so far. Maybe your BBB is a 2GB version? Dec 16 00:10:14 but on reboot, i found out i got a 2015 version Dec 16 00:10:28 Oh. Dec 16 00:10:59 Probably your bootloader. Try to go to /opt/scripts/ and git pull. Dec 16 00:11:18 ok thanks Dec 16 00:11:27 Then, go to /opt/scripts/tools/developers/ and perform... Dec 16 00:12:16 I think it is the sudo ./update_bootloader.sh file to update your bootloader. Hey. You are using the eMMC right? Dec 16 00:12:24 yeah Dec 16 00:12:52 Okay. I had some trouble w/ this many years ago before this fellow @zmatt told me about the instructions. Dec 16 00:13:01 Now, if something goes awkward, please just reply. Dec 16 00:13:32 Wait! Dec 16 00:13:34 ok, give me a moment, am running an update on the 2018 version Dec 16 00:13:46 There is a 2020 version. Dec 16 00:13:48 i'll try the update_bootloader Dec 16 00:14:01 I thought something was perculiar. Dec 16 00:14:23 Try the SD Card image if you have a SD Card. Dec 16 00:14:49 This is what i know... Dec 16 00:15:00 so, there are 2 versions with the same name? a 2015 version and a 2020? cos i thought i downloaded the 2020 version Dec 16 00:15:26 There is a SD Card image and a separate eMMC (flasher) image. Dec 16 00:15:40 The flasher image I think is older. Let me go and check. Dec 16 00:16:54 Here. If you go to flasher images at, https://beagleboard.org/latest-images, and search for am335x, you can then download the correct image. Dec 16 00:17:05 It is 2020! Dec 16 00:17:10 This year! Dec 16 00:17:48 I must have misunderstood you. I thought that you were trying to update/upgrade and needed a new image but found older images. Dec 16 00:17:59 the second image on the page? Dec 16 00:18:09 No. Let me go and count how many down. Dec 16 00:18:19 or the 6th? Dec 16 00:18:25 cos i have the 6th Dec 16 00:19:13 Here: https://debian.beagleboard.org/images/bone-eMMC-flasher-debian-10.3-iot-armhf-2020-04-06-4gb.img.xz is the exact image and that link should allow you to download it. eMMC, right? Dec 16 00:20:04 2020-04-06! Dec 16 00:20:34 Then, you will need to update/upgrade to get all the updated versioning, i.e. 10.7 Buster and so on. Dec 16 00:21:23 Yea. The sixth. Dec 16 00:21:51 hm, ok, i'll delete the image on my laptop and redownload it Dec 16 00:22:00 hopefully i don't get the same result Dec 16 00:22:09 What was the result? Dec 16 00:22:23 when running uname -s Dec 16 00:22:30 it only stated Linux Dec 16 00:22:31 uname -a too? Dec 16 00:22:32 nothing more Dec 16 00:22:40 Hmm. Dec 16 00:22:43 Try uname -a Dec 16 00:22:53 when looking for the date or version, it said 2015 Dec 16 00:23:12 So, did the board power down itself after the image was written? Dec 16 00:23:58 So, it basically does this... Dec 16 00:24:43 yeah it did Dec 16 00:24:59 The bbb, after the sd card is inserted when etcher has been used to write the image to sd card, does its "night rider" pattern, shows all LEDs lit, and then turns off. Dec 16 00:25:14 One would then take the SD Card out and power up the board. Dec 16 00:25:42 And that shows an image from '15? Dec 16 00:25:44 yeah did that Dec 16 00:25:45 Baeyens: so you reflashed it and it still said 2015 ? did you use a flasher image or did you manually edit /boot/uEnv.txt ? Dec 16 00:25:55 flasher image Dec 16 00:26:26 i tried it 3 time, with the same result. so went for the 2018 version, which i'm installing now Dec 16 00:26:29 without issues Dec 16 00:26:31 :-) Dec 16 00:26:35 ehh what Dec 16 00:26:53 it's possible that i have done something wrong Dec 16 00:26:59 human error and all Dec 16 00:27:03 Are you guys pulling my leg? Dec 16 00:27:29 but it's isn't that hard to use the boot button after inserting the sd card and booting up the bbb Dec 16 00:27:30 well the most likely candidate for weird things like this is an ancient bootloader on eMMC being unable to boot the new system on sd card Dec 16 00:27:36 If not, I will wait to see how this unfolds. Dec 16 00:27:54 ok, so to update the bootloader? Dec 16 00:28:04 well, no, that normally happens while reflashing Dec 16 00:28:10 Right. Dec 16 00:28:14 That is what I thought. Dec 16 00:28:29 I'm still trying to make sense of what exactly you did to identify what could have happened Dec 16 00:28:39 since even with bootloader issues, that would just result in not flashing at all Dec 16 00:28:44 he said uname -s Dec 16 00:28:56 or she or whomever. Dec 16 00:29:12 Right Baeyens? Dec 16 00:29:17 and the bootloader on eMMC should be bypassed if you power up while holding the S2 button (the boot button, the one closest to the sd card slot) Dec 16 00:29:22 he, and i could be wrong, it's 1h30 in the morning here. it could have been -a Dec 16 00:29:36 look, i'll try again, with the newly downloaded image Dec 16 00:29:39 Gotcha. Dec 16 00:29:39 if you want to know the image date of the installed system, use "cat /etc/dogtag" Dec 16 00:29:46 ok Dec 16 00:30:08 Baeyens: bone-eMMC-flasher-debian-10.3-iot-armhf-2020-04-06-4gb.img.xz is the correct image to use Dec 16 00:30:17 damn, kinda sucks, have just installed the 2018 version Dec 16 00:30:24 but ok, upgrade it is Dec 16 00:30:29 please don't install ancient stuff :P Dec 16 00:30:53 hey, that ancient stuff made it possible for the newer stuff to exist Dec 16 00:31:03 yes, it has its place... namely in 2018 Dec 16 00:31:11 :P Dec 16 00:31:13 hahaha Dec 16 00:31:16 :-) Dec 16 00:31:48 one way to be absolutely sure the bootloader on eMMC is bypassed: power on the beaglebone with the S2 button held (you can let go once the power led turns on) and _no_ SD card present Dec 16 00:32:07 the S2 button bypasses eMMC, so with no SD card the result should be that it doesn't boot at all... no activity other than the power led Dec 16 00:32:25 then insert the sd card and (without power-cycling) press the reset button Dec 16 00:32:39 ok Dec 16 00:33:08 reset! Dec 16 00:33:09 (this is an excessively paranoid procedure, but it can help in case the little S2 button turns out to be fiddlier than expected) Dec 16 00:33:35 That is a nice, new trick to figure things out. Dec 16 00:34:02 normally the S2 button isn't necessary at all... usually it's fine to just insert the sd card and turn the BBB on Dec 16 00:34:44 So, yep. uname -a and cat /etc/dogtag should give the two relative ideas about your "new" image. Dec 16 00:34:57 Baeyens: but, you did not at any point edit /boot/uEnv.txt (as mentioned in some instructions) as part of your attempts? Dec 16 00:37:03 since the only way that I've ever seen people make a puzzling mess of reflashing is if they do that but accidently edit the one on eMMC instead of the one on SD card Dec 16 00:38:14 What about me? I have done some silly stuff! Dec 16 00:38:30 Putting motors in action w/out interrupts! Dec 16 00:38:32 Oops. Dec 16 00:39:01 By the way, thank you GenTooMan. I needed some Global Stops and some interrupts. Dec 16 00:39:41 if memory serves i did edit /boot/uEnv.txt the first 2 times Dec 16 00:40:07 Baeyens: you probably managed to flash the contents of eMMC onto the SD card, overwriting the new image you put on it Dec 16 00:40:18 and then turned that into a 2015-image-flasher Dec 16 00:41:03 hm, ok, weird, i just download it, tried to flash it with etcher and got twice a corrupt file error Dec 16 00:41:34 redownload in progress Dec 16 00:41:36 "corrupt file error" ? what does it say exactly? Dec 16 00:42:28 nothing really. just failed writing, corrupt file Dec 16 00:43:20 unzipping with 7zip, no issue Dec 16 00:43:26 no need to unzip it Dec 16 00:43:34 with etcher? Dec 16 00:43:35 etcher will do that on the fly Dec 16 00:44:00 just download, use etcher? Dec 16 00:44:03 and if it compresses successfully then the file isn't corrupt Dec 16 00:44:46 after decompression there's not actually any way for etcher to detect a corrupt file, since an .img is just raw data to be written onto sd card (so it has no way to tell whether that data is "right" or not) Dec 16 00:45:23 oh ok, not bad. Dec 16 00:45:32 decompressing with etcher Dec 16 00:45:35 hold on Dec 16 00:45:44 that's why I asked what its error was exactly.. since if it fails to flash an .img file onto sd card then it can't possibly be complaining about a corrupt file, more likely it would be about a corrupt sd card Dec 16 00:46:00 "decompressing with etcher" ? Dec 16 00:46:18 flashing sd card Dec 16 00:47:28 i did an unzip with 7zip but since you said etcher does that on the go, gave etcher the downloaded xz file Dec 16 00:48:07 so far, so good. no errors yet Dec 16 00:48:09 another way to double-check that the downloaded file is intact is by computing its SHA256 hash (on windows you can use certUtil for that: https://www.shellhacks.com/windows-md5-sha256-checksum-built-in-utility/ ) and comparing it with the value shown on the website below the download link, which for bone-eMMC-flasher-debian-10.3-iot-armhf-2020-04-06-4gb.img.xz is ... Dec 16 00:48:14 ...e339459077b83f6458cb3432494954582aedad897b9f3b62fa390dfdb010a9df Dec 16 00:48:39 yeah i always forget to do that Dec 16 00:48:52 etcher is validating Dec 16 00:49:13 yeah I'm also usually too lazy :P Dec 16 00:50:28 You...lazy? No. Dec 16 00:50:30 and .xz includes integrity protection anyway Dec 16 00:50:42 so if the file is damaged it should get detected during decompression Dec 16 00:51:14 And/or...the sd card has been set to ext when it needs to be fat on a Win sys. Dec 16 00:51:26 ok, bypassed eMMc bootloader without SD card. like you said Dec 16 00:51:27 set_: that is totally irrelevant Dec 16 00:51:33 now running the installation Dec 16 00:51:34 set_: the filesystem gets overwritten Dec 16 00:51:45 Okay. But, I have had some issues w/ SD Cards b/c of that idea. Dec 16 00:52:38 Switching systems w/out checking has got my SD Card reader in a couple of binds. Dec 16 00:52:50 set_: not relevant in this context, etcher overwrites the entire contents of the sd card, making any previous formatting irrelevant Dec 16 00:53:01 Okay, okay. Dec 16 00:54:12 currently, I am going to try to use pyqt to set up a method for controlling my motors. Dec 16 00:54:33 Blah is all I can say. I need a book. Who knows of a good one about pyqt? Dec 16 00:55:29 unfortunatly, can't help you there. most of my knowledge comes from searching online Dec 16 00:55:36 and lots of testing Dec 16 00:55:48 Okay. No issue. Just throwing that out there in case. Dec 16 00:56:30 The system is done but I cannot control the interface w/ respect to LEDs or motors. Dec 16 00:57:29 I guess a .service file, some other type of _______, and some python3 might help more. I am drawing a blank. Seriously. Dec 16 00:58:42 Off to search! Dec 16 00:59:52 back Dec 16 00:59:59 firefox crashed Dec 16 01:00:00 lol Dec 16 01:00:10 think i need to reinstall my laptop Dec 16 01:01:12 You mean hardware on the laptop? Dec 16 01:01:17 Or the OS? Dec 16 01:02:04 Once in a blue moon, I get the air compressor out to blow out the encasing. That thing attracts more crud than expected. Dec 16 01:02:17 Heat! Dec 16 01:04:39 the OS, the hardware get's a yearly check Dec 16 01:04:47 Smart. Dec 16 01:06:05 laptops need a regular check never know what will go wrong with them but mostly when. Dec 16 01:06:08 sweet, the 2020 version is installed Dec 16 01:07:03 Phew and hello GenTooMan... Dec 16 01:07:58 I opened an issue in github for the bbb and motors I purchased. They do not want to be that open on github.com. Odd but I am going to contact them to see what they want from a consumer like me. Dec 16 01:10:15 Who knows? I may make a neat creation for the board and these people and their motors! Dec 16 01:12:17 same stuff applies as before for the /sbin/route add gateway Dec 16 01:12:33 and /etc/resolv.conf nameserver Dec 16 01:17:19 the fewer explosions the better I always say. Dec 16 01:18:55 Right! No explosions and just motor movement w/ a global stop, some interrupts, and a couple or more of switches. Dec 16 01:27:27 I have a html page serving pyqt. Right? So, would I use view.setHtml() for adding specific html radio buttons for use w/ Flask? Dec 16 01:27:54 The Flask app is running w/in pyqt5. Dec 16 02:48:46 It seems pyqt is far more complicated than it is worth but it does offer some neat displays. Dec 16 02:49:19 I hope the best for them but it would have been neat to manage my motors via their nice display. Dec 16 02:49:51 It always says, and I quote, "missing plugin." Dec 16 02:49:53 Boo. Dec 16 02:50:07 you have wxpython also Dec 16 02:50:30 The flask app displays in qt5 but the connection is not there yet. I am just one step away and I can feel it. Dec 16 02:50:33 wxpython. Dec 16 02:50:34 Okay. Dec 16 02:50:52 now for reality if you are making a UI that's function it won't be easy. Dec 16 02:52:05 GenTooMan: oh. UI is the thing right now. A nice one w/ a starter button is all I am calling. I have one that can reply w/ the button being pressed and returns output to the console. Dec 16 02:52:28 Not esp. helpful right now but it is a start w/ buttons and flask and and and. Dec 16 02:52:54 so flask is to migrate QTpy apps into websites? Dec 16 02:53:18 No, server for now or an html page at best. Dec 16 02:53:51 I just grabbed some source from a tutorial and tested it. I then changed some items to fit my flask app into it. Dec 16 02:54:17 the render_template function is called in the pyqt5 app. Dec 16 02:54:24 from flask. Dec 16 02:54:58 So, I got like from pyqt5.WebKit import * && from flask import Flask. Dec 16 02:55:35 Here. Let me show you the script. Well, are you really interested? Dec 16 02:56:33 I got a python3 file. This python3 file has a bunch of pyqt5 imports and flask imports. Dec 16 02:57:43 Then, in a server from vnc for testing it, I just call python3 MyFile.py and it runs opening my template/MyFile.html file along w/ it being inside of HTML page from pyqt5. Dec 16 02:58:03 Does that make sense? Dec 16 02:58:51 Source on the way. Please hold, i.e. just in case you are not too busy. **** ENDING LOGGING AT Wed Dec 16 02:59:56 2020