**** BEGIN LOGGING AT Mon Feb 15 02:59:58 2021 Feb 15 03:27:06 lets try some alliteration, FLIR flicking fun? Feb 15 03:27:58 set_: thanks Feb 15 03:29:19 Liter? Feb 15 03:29:32 CCFL_Man: No issue. Feb 15 03:29:55 It is fun to start a project and then move onto the project depth later on. Feb 15 03:32:15 I will fix it. Feb 15 03:44:51 GenTooMan: I bolted through the instructions b/c they are just instructions on gathering, attaching, and performing. Feb 15 03:45:32 In any context, that is the start of a beautiful relationship of electronics and human interfaces. Am I right? Feb 15 03:46:08 set_ should I have said frolicking FLIR fun instead? :D Feb 15 03:46:14 I will paraphrase some items later. I just wanted to throw that out there...Blah. Feb 15 03:46:30 set_: awesome. i will be using the BBB in a seismic data acquisition application Feb 15 03:46:53 It is fun, not only fun, there are many additional things one can do outside of what that app. given is capable of doing for now. Feb 15 03:47:04 Nice. Feb 15 03:47:24 GenTooMan: Yes. Frolicking. Feb 15 03:48:01 I was dancing through the forest on that typing up of the instructions. Feb 15 03:48:25 I understand that things are more complicated than I annunciate. Feb 15 03:48:29 via typing! Feb 15 03:48:39 But...it is a start. Feb 15 03:50:01 CCFL_Man, are you using any of TI's ADS1281/2/3/4/7 converters for that or more DIY? Feb 15 03:57:45 Mardi Gras is canceled. Feb 15 03:57:53 My BBB is doing okay...and! Feb 15 03:58:24 GenTooMan: It is Cupid's B-Day! Feb 15 03:58:57 I saw a dude make a heart out of metal w/ LEDs on the inside of it. Feb 15 03:59:00 you do realize why the celebrate valentines day don't you? Feb 15 03:59:05 No. Feb 15 03:59:09 Not a clue. Feb 15 03:59:47 Zeus and Cupid made babies? Feb 15 04:00:28 OpenCV in a comp. takes forever on the BBG. Feb 15 04:00:39 Blah. I should have not started this tonight. Feb 15 04:01:48 GenTooMan: I got my first set of ice sickles ever on this house. Feb 15 04:02:19 It is going to be 19 degrees tomorrow. The entire city shutdown. Trash pickup, Library, and so on... Feb 15 04:02:22 All shut. Feb 15 04:02:33 Ha. Feb 15 04:02:43 seriously? Valentine was a Christian martyred during the time in which Nero ruled Rome. Valentine was asked to teach a deaf and dumb Aristocrat's daughter to read, write, and speak. Apparently successful before execution he left the woman a note. Wishing her well and signed "Your Valentine" Feb 15 04:03:21 Oh. Feb 15 04:03:22 Nice. Feb 15 04:03:28 That is a nice story. Feb 15 04:03:31 How. Feb 15 04:03:33 I never knew. Feb 15 04:04:38 you would be amazed at many holiday that modern people ignorantly celebrate actually have a meaning. Feb 15 04:05:14 I would celebrate but I am sitting here on the ole 'puter. Feb 15 04:05:59 Yeah my rents said it was -3 out Saturday they aren't much further north than Louisiana Feb 15 04:06:18 See! Feb 15 04:06:36 This is unheard of for this area. Sickles! Feb 15 04:06:50 that's -13C (for other people) Feb 15 04:07:44 bleah -19C (heh) Feb 15 04:08:04 ha. Feb 15 04:09:08 I am not checking up on your work. Feb 15 04:09:51 well at least it wasn't a meter vs yard issue. Feb 15 04:10:15 My replicape is dying out...the 12v power for specific hardware no longer shows its presence. Feb 15 04:10:49 I got a $4.00 coin cell 3v lithium battery today. Feb 15 04:11:07 I placed it in the DMM and checked continuity of the sensor. Dead. Feb 15 04:11:57 I checked voltage on the Replicape for the sensor, none. Feb 15 04:12:13 I do not understand how to actually figure things out just yet. Feb 15 04:12:33 Send rations, GenTooMan. Feb 15 04:14:17 * GenTooMan tosses pancakes. Feb 15 04:14:43 does the replicape have a schematic publicly available. Feb 15 04:16:11 which pcb do you have? Feb 15 04:17:01 A B C Q? Feb 15 04:17:33 No. Feb 15 04:17:38 B3a. Feb 15 04:17:54 Well, it may. Let me check. Feb 15 04:18:00 I believe it does not. Feb 15 04:18:22 No public schematic that I have found. Let me check again. Feb 15 04:19:05 https://wiki.thing-printer.com/index.php?title=Replicape_rev_B and I think TI represented them for a bit too. Feb 15 04:19:08 Let me check TI. Feb 15 04:19:35 https://bitbucket.org/intelligentagent/replicape/src/master/ Feb 15 04:19:40 I already found it Feb 15 04:19:53 https://www.ti.com/tool/TIDEP0007 Feb 15 04:19:56 OH. Feb 15 04:19:59 they used eagle of all the dumb things. Feb 15 04:20:07 Oh. Feb 15 04:20:16 Do you have eagle? Feb 15 04:20:29 in any case I will see if I can move the schematic to KiCAD and make sense of them. Feb 15 04:20:41 I am going to look at it now. Feb 15 04:20:42 there is a PCB B3A Feb 15 04:20:52 Yes, okay. Feb 15 04:28:17 interesting Feb 15 04:28:37 What? Feb 15 04:28:58 I looked online and it seems there are many files but not just one file. Feb 15 04:30:31 heh ... Feb 15 04:30:45 https://bitbucket.org/intelligentagent/replicape/src/master/PCB_rev_B3A/ <-- Feb 15 04:31:11 .pcb and .sch files are it. Feb 15 04:31:58 Oh. Okay. Feb 15 04:35:44 so is I2C2 functioning and SPI1? Feb 15 04:35:55 Yes. Feb 15 04:36:25 The only thing not functioning is P9_18 or P9_15 or P9_13. Feb 15 04:36:27 I think. Feb 15 04:36:29 Let me check. Feb 15 04:37:40 I'm so glad the connectors are labelled Feb 15 04:37:49 * GenTooMan was being sarcastic of course. Feb 15 04:39:16 The only thing I know is that 12v, 5v, 3.3v, and GND show no pwr to the pins for endstop_z2 which is also the "inductive sensor" pin(s). Feb 15 04:39:33 So, I think those are controlled via the P9_18 pin. Feb 15 04:39:50 I am not sure what the use of P9_18 is in their world, though. Feb 15 04:40:01 It could easily be GPIO or something else. Feb 15 04:41:29 I would need to get my board, plug it in, and find out (I think). Feb 15 04:41:53 W/out the overlay and schematic, I am busted so far. Feb 15 04:49:06 * set_ Is wanting to know how to install the saved .sch format in the Eagle source! Feb 15 04:52:08 umm Kicad->File->Import Project->EAGLE CAD Feb 15 04:53:10 I found that but in Kicad but I cannot save the file from BitBucket. Feb 15 04:53:32 I C & P the file instead. Feb 15 04:57:51 Hmm if you want to save the files try this click on the file wait a while for it to load then click the "..." button next to edit click Open raw then save the page as a file. repeat for brd and .sch Feb 15 04:58:55 Okay. Feb 15 04:59:00 Sheesh. Feb 15 05:01:35 I got it now. Feb 15 05:02:21 be sure the two files .sch and .brd are in the shame directory. Then try the umm ... sequence on the .sch file. Feb 15 05:03:56 Oh. Feb 15 05:04:21 Okay. I saw what happened when that was not available. I installed one at a time and the bricks hit the fan. Feb 15 05:06:43 noisy that would be Feb 15 05:08:44 What? Feb 15 05:09:01 My radio is playing! Yes? Feb 15 05:11:17 Okay. I got it. Feb 15 05:11:26 I got it on Eagle and on KiCAD. Feb 15 05:11:28 Neat. Feb 15 05:11:39 Now, I can configure whatever is needed. Feb 15 05:12:01 I want to ask you something about the three pins on the top of the Cape. Feb 15 05:12:06 Here goes. Oops. Feb 15 05:12:13 First, let me find it in the schematic. Feb 15 05:15:09 Anyway... Feb 15 05:15:26 There are three gold circles on the top of the Cape. Feb 15 05:15:43 One is labeled 12v, the other 5v, and the last one is labeled 3.3v. Feb 15 05:15:50 I cannot find it on the schematic right now. Feb 15 05:16:56 'gold circles' are probably marked TP for test points Feb 15 05:17:16 Let me go and check. Feb 15 05:17:35 or just open the pcb layout and click on them... Feb 15 05:18:24 Right. It is undistinguishable on the pcb layout for now. Feb 15 05:18:30 SOrry. they are silver. Feb 15 05:18:37 turn off some layers Feb 15 05:18:41 Aw. Feb 15 05:19:54 Okay. I found them. Feb 15 05:20:20 I found them on the pcb layout. Feb 15 05:38:03 It seems there is a 'floating layer' of a line that crosses over the Ethernet cable connector on the pcb view. Feb 15 05:38:16 This is the 12v line. Feb 15 05:38:19 Supposedly. Feb 15 05:38:28 So, who knows? Feb 15 05:38:37 B/c...I do not. Feb 15 05:38:55 I do not see a floating wire or anything over the ethernet connector. Feb 15 05:40:14 Anyway, it is late. Later for now. Feb 15 12:07:02 set_ the PCB is messed up because it's a conversion from eagle. Although the layers might exist etc. Not everything is correct. I suggest examining the schematic. One thing I noticed is that the schematic has 6 pages yet the people who made it only show 3 on the web site. They used PNGs and provided no pdfs. This may indicate they just want to show something but not actually show details. Feb 15 12:32:26 oh, just open it in EAGLE then Feb 15 12:55:24 1 Eagle is Dead 2 finding a useful Free copy is hard 3 it's not for Linux Feb 15 15:22:43 GenTooMan: seems to be available for Linux too: https://www.autodesk.com/products/eagle/free-download Feb 15 15:23:43 though this version seems to be more limited than required since you mentioned 6 schematic pages (unless the free version can open these even if it cannot save them) Feb 15 16:25:41 <_austin> morning/afternoon all. hope everyone is staying safe with this wonky winter storm! Feb 15 16:26:33 <_austin> is there a way with the adafruit lib in python to have a counter count pulses when a pwm output is going?  or would i have to just make a timer and estimate using the timer and my frequency? Feb 15 16:28:09 in principle every pwm output will be synchronous to the main oscillator hence to the raw monotonic time, which can presumably be obtained in python, though the exact ratio may be annoying to determine Feb 15 16:28:41 what exactly are you trying to do? Feb 15 16:30:51 <_austin> I'm trying to determine if my stepper is "close enough" without reading an encoder.  We know about how far we need to move the stepper, so I'm trying to count the pulses (I have the ratio of input pulses:encoder output) Feb 15 16:31:03 <_austin> If that makes sense Feb 15 16:33:30 <_austin> i.e. If I know my frequency is 1hz and I move for 10 seconds, I should have 10 pwm pulses.  I can start a timer and do it that way (10 seconds/1hz).  But I was wondering if there was an easier way to have Python just count for me. Feb 15 16:35:06 there are definitely ways to give an exact number of pulses, but not something easy using python Feb 15 16:35:16 I'd also assume you need a higher frequency than 1 Hz in practice? :P Feb 15 16:36:27 <_austin> Haha yes, that was just an easy example Feb 15 16:36:40 <_austin> We're going to be swtiching between like 5k and 10k Feb 15 16:37:11 PRU is a great way to generate arbitrary waveforms, even though it has some learning curve Feb 15 16:37:46 <_austin> Yea, might just be easier to do the timer and derive that way Feb 15 16:38:18 yeah but at 10 kHz you may end up being significantly wrong Feb 15 16:38:21 <_austin> I'm writing this and then going on vacation for a week (goodbye midwest snow storm, hello Orlando!) and a mechanical guy with beginnerish coding skills may have to troubleshoot Feb 15 16:38:54 <_austin> Could I set it up in parallel with an input pin and count rising edges? Feb 15 16:39:46 I mean I guess, but dunno if you can hope to react to that with 0.1 ms precision in python running on a BBB Feb 15 16:40:02 in fact I very much doubt that Feb 15 16:40:23 (but I might just be pessimistic) Feb 15 16:40:36 <_austin> I don't need super precision. Feb 15 16:40:44 <_austin> We're going very slow. We have it geared WAY down Feb 15 16:41:31 I'm not sure what the easiest way would be to do this without using PRU... in no small part since I'd be using PRU ;) Feb 15 16:41:53 since generating a sequence of pulses with it is a particularly easy application Feb 15 16:41:54 <_austin> That's why we're looking at this in the first place.  I doubt we need the precision of the encoder, so we're testing to see if there's an easy way that we don't have to use it.  It would cut back on cables in our system drastically Feb 15 16:42:32 <_austin> Currently I'm just using the Adafruit PWM.setfrequency for what I'm doing (as they're just square waves at a fixed frequency) Feb 15 16:42:43 <_austin> I'm not sure what PRU is to be honest. Feb 15 16:43:44 PRU stands for Programmable Real-time Unit, they're simple microcontroller cores embedded in the SoC with very simple deterministic timing (nearly all instructions are 1 cycle) and a bunch of IOs wired directly into registers Feb 15 16:44:14 so they can very easily generate arbitrary signal waveforms with 5ns precision Feb 15 16:44:32 <_austin> Ah, so not doing the Python state-machine thing that I'm doing then eh? Feb 15 16:44:59 <_austin> At the end of the day, I'm going to have a fun statemachine that receives a message, sets all the I/O, then responds to the main computer. I'm just using the BBB as a slave device Feb 15 16:45:13 I've made a python library that can load code onto the PRU cores (there are two of 'em in the PRU subsystem) and interact with it: https://github.com/mvduin/py-uio Feb 15 16:46:11 so you could e.g. have a tiny PRU program that generates a specific number of pulses with configurable timing, and then just run that program whenever you need to move the stepper Feb 15 16:46:40 the PRU program is probably a dozen lines of code or so Feb 15 16:47:37 <_austin> Interesting. I'll take a look at it Feb 15 16:47:57 <_austin> I may keep going the timer route for now to see if it's "close enough" and then if not, try the PRU route before going back to encoders. Feb 15 16:47:59 <_austin> Seems reasonable to me Feb 15 16:48:46 I can probably add this as an example to py-uio, since it seems like it would be useful generically Feb 15 16:52:05 <_austin> Possibly, that might help Feb 15 16:52:16 <_austin> I'm very intermediate at coding in anything other than embedded c Feb 15 16:52:40 <_austin> Before the past 18 months, 95% of my programming was embedded c on 8 bit micros XD Feb 15 16:53:17 I suppose this could be done in C too, but I'm not even sure that would make this simpler :P Feb 15 16:53:42 like, it makes generating the pulses *very* slightly simpler while making the interaction with the python code harder Feb 15 16:54:10 <_austin> Oh my state machine is all in python Feb 15 16:54:20 <_austin> That's why I'm using the Adafruit lib for all the IO/ADC/PWM Feb 15 16:54:38 which is why making the program simpler but interaction with python code harder doesn't sound like a win to me Feb 15 16:55:02 <_austin> For your thing? Feb 15 16:55:17 for the send-pulses thing Feb 15 16:58:26 <_austin> The send to PRU thing? Feb 15 16:58:36 <_austin> I'm so confused if you're saying I'm making mine harder or using yours might be harder haha Feb 15 16:59:18 I was saying that prorgamming PRU in C instead of assembly is an option but is probably not overall easier Feb 15 17:00:23 I'm writing the example currently (but was distracted for a bit) Feb 15 17:03:17 <_austin> Ah gotcha. Feb 15 17:03:21 <_austin> Yea I'm not doing any C on this Feb 15 17:03:32 <_austin> I was just giving a view into my background more than anything Feb 15 17:04:48 ok I think I've written the example Feb 15 17:05:09 the PRU part of it I mean, need to write a python example to go with it Feb 15 17:08:34 the program sends a configurable number of pulses on one of the PRU code's direct outputs with programmable on-time (in 10ns units, min 10ns) and off-time (in 10ns units, min 20ns) Feb 15 17:08:49 that's presumably accurate enough for you ;-) Feb 15 17:10:51 and sends a completion event when done (so if you request a fairly long pulse-sequence you don't have to poll for completion but can get an event instead, e.g. via asyncio or whatever main event loop you have) Feb 15 17:41:22 _austin: I've added an untested pulse-train example: https://github.com/mvduin/py-uio/commit/be439fc26ce2a899da691f6f75e4c4a9df92ebce Feb 15 17:42:41 the output, currently hardcoded to be pru0 output 0, which is P9.31, needs to be configured to pruout mode for this Feb 15 17:43:38 hmm, maybe not the ideal pin to use as example since it conflicts with HDMI audio Feb 15 17:48:00 Hi i have beaglebone green. is there a protocol to give more confidentiality with web traffic? Feb 15 17:53:44 david19: ehh that's a super vague web development question that has absolutely nothing to do with the beaglebone Feb 15 17:54:21 in general however the way the web standards have been designed they've made it extremely difficult to get confidentiality with devices on the local network Feb 15 18:06:52 GenTooMan: Good news. Feb 15 18:07:13 The Cape is not faulty. I needed another test point for GND on the Cape. Feb 15 18:07:22 So, it is indeed the sensor being kaput. Feb 15 18:28:18 so you got desensitized? Feb 15 18:28:51 *sound of crickets* Feb 15 18:31:09 Hello! Feb 15 18:31:49 The GND on the input needs to be the test point and not on the pin that is GND on the terminal. Feb 15 18:32:31 Hello @zmatt...how are things going? Feb 15 18:32:45 We have ice and snow on the ground here in LA. Ha. Feb 15 18:32:48 Weird days. Feb 15 18:34:01 here too, it's currently thawing but some snow/sludge remains Feb 15 18:34:25 Hmm. Nice. Feb 15 18:34:36 Yea but you are up north. Feb 15 18:34:43 Way north. Feb 15 18:36:18 The ole sludge. Black ice too! Feb 15 18:36:44 snow is still pretty rare here, at least in recent times Feb 15 18:36:49 Oh. Feb 15 18:36:53 I was unaware. Feb 15 18:37:11 Hmm. Feb 15 18:37:31 Okay. So, my thoughts of a winter wonderland are completely incorrect. Feb 15 18:37:35 like, we got through this winter without any freezing until the jetstream thing happened last week Feb 15 18:37:42 Oh. Feb 15 18:38:07 Just three days ago, it was 76 degrees fahrenheit here. Feb 15 18:38:23 It was like a cool front in June. Ha. Feb 15 18:39:30 weather is a mysterious thing, still. Feb 15 18:40:01 Usually, i only have to worry when I see black or green in the sky. Feb 15 18:40:18 Now, we have this hardened ice slicks too. Odd... Feb 15 18:41:53 Hi Feb 15 18:42:19 okay my previous statement seems to be an exaggeration, apparently the temperature did dip slightly below freezing at night occasionally Feb 15 18:43:19 I figured it had to get colder than that... Feb 15 18:43:22 Hi Feb 15 18:43:37 Anyway... Feb 15 18:43:38 Ha. Feb 15 18:43:48 hi - bye? Feb 15 18:44:57 @zmatt: Do you know of any vision libraries for the FLIR cameras? Feb 15 18:45:33 I am asking b/c OpenCV will not fully compile on the BBG and I have not started to compile it on the Debian desktop. Feb 15 18:45:44 why would you need to compile it? Feb 15 18:45:54 instead of just apt-get install Feb 15 18:45:58 B/c...I want the C++ lib. Feb 15 18:46:03 Not the python3 version. Feb 15 18:46:10 so apt-get install the C++ lib Feb 15 18:46:15 Oh. Feb 15 18:46:21 Is that possible? Feb 15 18:46:24 Hmm. Feb 15 18:46:31 Okay. Off to look that up. Feb 15 18:46:42 libopencv-dev Feb 15 18:46:48 is what you'd presumably want Feb 15 18:47:19 Oh. Feb 15 18:47:21 Okay. Feb 15 18:47:40 I found the opencv in debina buster but it will not install. Feb 15 18:47:48 ? Feb 15 18:47:52 online. Feb 15 18:47:58 ?? Feb 15 18:47:58 debian. Sorry. Feb 15 18:48:21 I thought if I can find it in the debian.org site, I could just use apt install to get it. Feb 15 18:48:40 Let me try again. Feb 15 18:48:55 sudo apt update && sudo apt install libopencv-dev Feb 15 18:49:12 I was trying, sudo apt install opencv Feb 15 18:49:17 I will try that idea. Feb 15 18:49:28 there's no package named "opencv" Feb 15 18:49:55 you can list all packages with "opencv" in the name using: apt-cache search -n opencv Feb 15 18:49:56 You are right. I just saw it. Feb 15 18:50:25 They just have a bunch of libs. below the opencv heading in Debian Buster. Feb 15 18:50:31 generally for C/C++ development you want a -dev package Feb 15 18:50:37 Oh. Feb 15 18:50:39 Okay. Feb 15 18:50:41 Nice. Feb 15 18:50:45 Good info. Feb 15 18:50:45 so apt-cache search -n opencv.*-dev Feb 15 18:51:01 will list all -dev packages with "opencv" in the name Feb 15 18:51:07 its been rather cold for the last 3 months here actually. it only once in a while got above freezing. Feb 15 18:51:12 This is where I was looking: https://packages.debian.org/source/stable/opencv Feb 15 18:51:26 libopencv-dev is a metapackage that installs all the others Feb 15 18:51:33 GenTooman is in Alaska! Feb 15 18:51:38 Okay. Feb 15 18:51:42 Thank you, phew. Feb 15 18:51:44 if you only need specific functionality you can install specific components instead Feb 15 18:51:56 to save space Feb 15 18:53:26 Wowzzers. Feb 15 18:53:28 Nice. Feb 15 18:53:33 That was way easy. Feb 15 18:53:54 Now, to install the contrib libs. Feb 15 18:58:07 <_austin> zmatt thanks i'll look at it in a bit!  Had some meetings that took me away for a bit Feb 15 18:58:44 _austin: it could definitely still be improved.. like, there's no reason to hardcode the pin number in the PRU program like it is right now, I ought to fix that Feb 15 18:59:43 <_austin> yea, but for now it'll work.  i'm not using hdmi on this at all anyways, so conflicting with audio isn't an issue Feb 15 19:00:10 <_austin> we're even sending our messages with rs485 XD Feb 15 19:11:28 set_: here's what I meant by "no freezing" until last week... like sure, we've technically dipped below 0͏°C but kinda negligibly: https://docs.google.com/spreadsheets/d/e/2PACX-1vTDWCkEjgZ3HIDbMCJEG5o150U5__AjjswhZiSCc-QpYtBBY8outCiB2qMQyixp0VUUWuZzockzx-yU/pubchart?oid=873820207&format=image Feb 15 19:13:00 Right...just freezing during the nights mostly if at all. Feb 15 19:13:37 and even then with little enthausiasm Feb 15 19:14:13 Right. So, 0 C is equal to 32 here, GenTooMan! Feb 15 19:14:19 and then the jetstream said "SURPRISE! Would you like some polar wind?" Feb 15 19:14:25 Aw. Feb 15 19:15:50 It is -4 C right now here. Feb 15 19:15:54 and now it's back to normal, at least once the remaining snow sludge has melted Feb 15 19:16:15 Aw. We are in for a real cool night. 17 F. Feb 15 19:16:35 I do not remember when the last time it got below 29 F. Feb 15 19:17:07 weather can indeed be funny Feb 15 19:17:21 That sludge stuff is messy. Dirt and slow and mud and whatever else. Feb 15 19:17:31 slow = snow Feb 15 19:17:33 Ha. Feb 15 19:18:26 yeah the sludge sucks Feb 15 19:19:06 Yep. Weather endeavors. I still have not found time to make the weather thing a reality here w/ my BBB. I need to purchase a new board still. I am going to get a BBBW next (again). This way, I do not have to have an Ethernet cable dangling from the BBB. Feb 15 19:19:24 you're on the wrong side of the channel Feb 15 19:19:36 no slush over here Feb 15 19:20:03 Oh. I have frozen, hardened ice on the ground but not slush or sludge. Feb 15 19:20:11 there was a light dusting of snow one morning last week, but it didn't remain long Feb 15 19:20:37 mru: on the other hand, on this side of the channel you're not in the UK ;-) Feb 15 19:20:56 but our fish are much happier! Feb 15 19:21:00 I took a short walk. to the paver stones and back, slipped w/ one foot and returned away from the paver stones. Feb 15 19:21:02 how's brexit going? Feb 15 19:22:09 I haven't noticed much difference personally Feb 15 19:22:46 I know we've had some problems due to a supplier in the UK but I don't know the details Feb 15 19:23:09 shipping stuff needs a lot more paperwork Feb 15 19:23:09 I think they ended up moving their production out of the UK Feb 15 19:23:29 but I'm not in shipping Feb 15 19:24:00 neither are we.. unfortunately one frequently does need things shipped even if you're not in shipping :P Feb 15 19:24:24 yes, but then it's somebody else's problem Feb 15 19:24:58 I ordered something from lenovo recently Feb 15 19:25:07 it was billed by some irish company and shipped from the netherlands Feb 15 19:25:52 ups tracking had a pageful about customs events but it still arrived in 2 days Feb 15 19:56:31 <_austin> we've got ~12" of snow coming tonight Feb 15 20:00:15 a little dusting right? Feb 15 21:42:11 Hello! I have a BBB that I haven't tried to use in a long time. I recall being able to connect it to a Windows computer with a mini-USB cable when I first got it and the BBB mounted as a removable drive. That doesn't happen anymore. I tried creating a flasher SD card last night, then plugging in a 5V power supply while holding down S2, but I do not Feb 15 21:42:11 see any activity. All I see is a single solid blue LED directly next to the 5V power input. Any thoughts as to what I can try? Feb 15 21:42:40 hmm Feb 15 21:43:02 hard to say what might be going on... having just the power led turn on but nothing else sounds like a weird failure case Feb 15 21:43:33 I was hoping that wasn't going to be the answer. :-) Feb 15 21:43:34 my first thought is to connect the serial console Feb 15 21:43:43 How might I do that? Feb 15 21:43:51 my first thought would be measure the 3.3V supply Feb 15 21:44:16 also a good idea Feb 15 21:44:17 I'll go grab my multimeter. BRB Feb 15 21:44:43 I wouldn't have even powered it on without the console attached Feb 15 21:44:49 since the only case where I've seen a beaglebone power on but unable to boot from eMMC or SD card, something was dragging down the 3.3V supply... and also the eMMC was getting quite hot and starting to smell toasty :P Feb 15 21:45:13 at least you got toast Feb 15 21:46:00 Well the board is indeed getting toasty. Feb 15 21:46:12 this is possible because the 3.3V comes from a normal LDO instead of from the PMIC, so it's not being monitored for overcurrent Feb 15 21:46:12 Do I need a JTAG to connect to the serial console? Feb 15 21:46:18 no Feb 15 21:47:15 serial console is just 3.3V (but 5V-tolerant) "TTL" serial, using an ftdi-compatible 5-pin header Feb 15 21:48:12 5.138V Feb 15 21:48:23 3.3V can be measured on P9.03-04 while ground is in a bunch of places including P9.01-02 Feb 15 21:48:30 uhh, what are you measuring? Feb 15 21:48:47 the input 5V voltage isn't that interesting Feb 15 21:48:49 my ftdi headers have 6 pins Feb 15 21:48:56 6-pin sorry, not 5-pin Feb 15 21:49:02 my bad Feb 15 21:49:05 Sorry - not sure what you were asking me to check Feb 15 21:49:25 This is my first time ever using a BBB Feb 15 21:50:05 Bought it a couple of years ago for a project and the project never came to fruition, so it's been sitting in the box. Feb 15 21:50:07 JeffH: ah, if the board is oriented with ethernet connector facing up, the P9 header is on the left and pin 1 is topright corner Feb 15 21:50:53 this is a nice image: https://www.mathworks.com/help/examples/beaglebone/win64/xxbeaglebone_black_pwm_pinmap.png Feb 15 21:51:04 that shows where you can measure the 3.3V Feb 15 21:51:24 OK and you want me to check 01 and 03 to determine voltage out? Feb 15 21:51:29 yep Feb 15 21:51:32 K Feb 15 21:52:43 Nthing Feb 15 21:52:48 Nothing Feb 15 21:53:00 ehh that has to be a measurement error Feb 15 21:53:33 Give me a minute - will try again Feb 15 21:53:34 or I guess it's possible the 3V3B might actually be completely shorted to ground somehow somewhere Feb 15 21:53:45 but I don't think a damaged IC could do that Feb 15 21:55:12 I've seen damaged ICs do the strangest things Feb 15 21:56:04 someone once managed to fry a usb port so badly that the current limiter shorted vbus to ground Feb 15 21:56:28 the limiter still worked though, so it quickly turned off the voltage Feb 15 21:56:34 and then back on ... Feb 15 21:56:39 lol... overvoltage or something? Feb 15 21:56:46 no idea how it happened Feb 15 21:56:58 the usb phy was also partially damaged Feb 15 21:57:12 worked in full speed but not high (or the other way around) Feb 15 21:57:36 that was an allwinner H3 chip Feb 15 21:57:52 we had a customer fry the beaglebone's ethernet... it still kinda worked if ethernet was plugged in at power up, but it was no longer able to detect link changes Feb 15 21:58:05 (they plugged it into a phone line) Feb 15 21:58:17 did the phone ring? Feb 15 21:59:00 I'd expect the transformer to survive the normal DC on a phone line Feb 15 21:59:44 probably, but maybe the phy didn't appreciate the brief pulse when the phone line was plugged in Feb 15 21:59:50 and removed again Feb 15 22:00:49 3.3 Feb 15 22:01:01 JeffH: ok hmm, well that eliminates one suspect Feb 15 22:01:07 Probes weren't making good enough contact Feb 15 22:01:55 just to confirm, you've getting the same lack of led activity when you just power it on without SD card and without using the S2 button? Feb 15 22:02:04 *you're Feb 15 22:02:07 Correct Feb 15 22:03:05 A friend was trying to get Ubuntu running off eMMC at one point when he had the board. Is it possible he did something catastrophic that would prevent it from booting again? Feb 15 22:03:11 isn't that what you'd see if there's no valid boot image anywhere? Feb 15 22:03:30 JeffH: okay so you don't know if it was in a bootable state last time it was used? Feb 15 22:03:48 in which case the serial console would print 'C' characters Feb 15 22:03:49 and no, not really.. the only thing that could prevent it from booting anymore is hardware damage Feb 15 22:04:24 Correct. It was when I bought it. I know that for certain. But then he had it and gave it back to me a couple of months later. This was all roughly 1.5/2 years ago...so it' just been sitting. Feb 15 22:04:33 so if it's possible your friend left eMMC is a non-bootable state, it may be possible you failed to properly prepare the SD card... what did you do exactly? Feb 15 22:04:34 Where can I connect the FTDI serial to? Feb 15 22:04:51 there's a 6-pin header right next to P9 in the middle Feb 15 22:05:00 pin 1 (black wire) is marked with a dot on the PCB Feb 15 22:05:34 All I did last night was try to create a flasher SD card using the official image. Wrote it to a 16 GB micro SD card using Balena Etcher. Inserted the SD card, held down S2, and connected power. Feb 15 22:05:53 Didn't see any LEDs light up aside from the one next to power. Feb 15 22:05:55 what's the name of the image you used? Feb 15 22:06:09 bone-eMMC-flasher-debian-10.3-iot-armhf-2020-04-06-4gb.img Feb 15 22:06:28 yeah that should be fine, and using Etcher eliminates most risk factors Feb 15 22:06:47 Pretty much all I use unless I'm on a Linux box Feb 15 22:07:33 OK - let me see what I can accomplish with the serial console. Might have to do that later and come back. What are the console settings? 9600 8N1? Feb 15 22:07:47 115200 more likely Feb 15 22:08:00 OK Feb 15 22:09:58 Just found my USB TTL adapter Feb 15 22:10:48 Of course...no wires. Feb 15 22:12:21 ah you don't have one with a 6-pin connector? if you need to wire individually, it's pin 1 = gnd, pin 4 = txd (to BBB), pin 5 = rxd (from BBB) Feb 15 22:15:22 Yes - it is a 6-pin but it doesn't have leads attached. I just found a set. Feb 15 22:27:09 OK - if I hold down S2 while powering up, I see CCCCCC on the serial console Feb 15 22:27:32 so the CPU is working but it's unable to load the bootloader (u-boot) from SD card Feb 15 22:27:50 Well as of this moment, I don't have an SD card inserted. Feb 15 22:27:58 Let me try putting the SD card back in Feb 15 22:28:06 then this is the expected result if powered on with S2 held Feb 15 22:29:02 So how long should I hold S2 down for when I power on with the sd card? Feb 15 22:29:12 you can let go once the power led turns on Feb 15 22:29:22 OK Feb 15 22:29:23 or technically 20ms after the power led turns on iirc Feb 15 22:30:35 Going to try to write the image to SD again. I'll grab the non-flasher image for now Feb 15 22:30:48 S2 is sampled at power-on and changes the boot order from { eMMC, μSD, uart-xmodem, usb-rndis } to { spi flash, μSD, usb-rndis, uart-xmodem } ... that applies only to the order for bootrom to find u-boot, not for u-boot to find linux Feb 15 22:31:26 the CCC you're seeing is the uart-xmodem method Feb 15 22:31:57 OK Feb 15 22:32:14 Well I'm happy I'm seeing some activity - I figured it was just dead Feb 15 22:32:35 yeah the processor seems fine and the 3.3V supply seems fine, so there's no obvious reason for it to not boot Feb 15 22:33:04 Problem exists between chair and keyboard Feb 15 22:33:06 unless maybe the sd card connector is broken and your friend managed to get eMMC wiped in his efforts to flash Ubuntu Feb 15 22:33:32 That's what I'm wondering - perhaps the SD card connector pins are bent Feb 15 22:34:06 a broken SD card connector would also make the latter much easier (e.g. he thought he had booted from SD card while in fact he had booted from eMMC, and then he tried to flash that onto eMMC, resulting in a crash after the "wipe eMMC" step) Feb 15 22:34:51 Lovely - and at that point, it's toast, right? Feb 15 22:35:10 no, but it does make reflashing eMMC a bit more exciting Feb 15 22:35:17 LMAO Feb 15 22:35:39 This is getting interesting. Feb 15 22:36:40 Grab your popcorn Feb 15 22:36:42 something I've done once is take advantage of the fact that the eMMC is wired in parallel to expansion header pins, which means that if the board is powered but the CPU is held in reset you can access the eMMC directly via the expansion header pins Feb 15 22:37:11 a broken sd card connector doesn't look that difficult to replace Feb 15 22:37:21 OK - well just flashed a new SD card - used this image instead bone-debian-10.3-iot-armhf-2020-04-06-4gb.img Feb 15 22:37:23 No dice Feb 15 22:37:44 JeffH: what happens if powered on without S2 ? Feb 15 22:37:47 @mru depends on how good you are at soldering - which I am not. Feb 15 22:37:59 Standby Feb 15 22:38:10 JeffH: another possibility is netbooting the BBB and then reflashing it via the network Feb 15 22:38:23 Oh - that might work Feb 15 22:38:48 it falls in the category of "definitely possible, firmly non-trivial" Feb 15 22:39:05 OK - serial console still connected. Removed power & sd card. Powered on - no activity. Just the blue power LED is on. Feb 15 22:39:13 Nothing written to serial console Feb 15 22:40:12 Ahh - looks like I can do a TFTP/NFS boot Feb 15 22:40:27 in theory yes Feb 15 22:41:26 I've gotten as far as getting u-boot loaded via the network and then tftp-loading a kernel and dtb and then booting that, failing at the "no root fs" step Feb 15 22:42:00 so if you're already comfortable with using nfsroot from that point on, then you'd have a successful netboot Feb 15 22:42:03 Well this looks like something I want to try - just not going to have the time to do it today. Feb 15 22:42:13 you'll need to custom-build an u-boot though Feb 15 22:42:15 Once booted, I should be able to flash the eMMC though? Feb 15 22:42:19 yep Feb 15 22:42:45 https://bootlin.com/blog/tftp-nfs-booting-beagle-bone-black-wireless-pocket-beagle/ Feb 15 22:42:57 http://linuxkernel51.blogspot.com/2015/08/boot-beaglebone-black-with-nfs.html Feb 15 22:43:13 I figure with those two links, I should be able to get it going Feb 15 22:43:20 here is a known-good u-boot for booting via ethernet: https://github.com/dutchanddutch/u-boot/tree/am335x-v2017.11-netboot Feb 15 22:43:31 lemme see what those say Feb 15 22:43:54 Awesome! Thanks so much for your time and patience. Feb 15 22:44:46 the BBB can netboot via usb(rndis) or via ethernet... for the latter you need to use two wires to change the boot config to one that includes ethernet in the boot order Feb 15 22:45:22 I have no idea how you have figured that out. TRM? Feb 15 22:45:31 I've personally always used dnsmasq to serve BOOTP and TFTP Feb 15 22:45:52 Looks like there's a way to do a PXE boot too Feb 15 22:46:26 that would be an u-boot feature I presume? Feb 15 22:46:54 the only thing I have experience with is just serving a FIT to u-boot, which contains a kernel + dtb + initramfs Feb 15 22:47:26 which is of course sufficient... the initramfs can contain a busybox, you don't need much to reflash eMMC Feb 15 22:47:50 Well - doesn't look like the guy was very successful with PXE boot. There's no reason I can't try TFTP or NFS. Feb 15 22:48:02 Right Feb 15 22:48:19 OK - well this will have to wait until the weekend. I'm out of time today. Feb 15 22:48:25 I wonder if I still have a link with notes about my netboot efforts somewhere Feb 15 22:48:40 Was hoping it would be something I could resolve quickly and that I was just doing something stupid. Feb 15 22:48:42 Thanks again! Feb 15 22:50:07 JeffH: I'll probably see you again then ;) Feb 16 02:22:44 GenTooMan: No! **** ENDING LOGGING AT Tue Feb 16 03:00:33 2021