**** BEGIN LOGGING AT Fri Sep 21 03:00:02 2018 Sep 21 03:00:17 did you have the multimeter in the wrong mode? Sep 21 03:00:26 voltage...no. Sep 21 03:00:54 I call it spitfire. Sep 21 03:02:10 For instance w/ dead batteries and live batteries. It is this way: dead + to live +. And then...live - to dead -. Sep 21 03:02:12 ehm, then I have no idea what you manage to do with your equipment to cause that Sep 21 03:02:31 Me neither. Sep 21 03:03:07 It happened twice to me and it was surprising. Sep 21 03:03:09 Maybe that order you mention set_, + first, is for safety in battery cars Sep 21 03:03:27 car's batteries* Sep 21 03:03:30 Yes. For cars, trucks, batteries when you jump them. Sep 21 03:04:10 hmm, what's the reason for it in that situation? Sep 21 03:04:37 So, electric shock from 12v battery does not erase the memory. Sep 21 03:04:46 "Human mind memory." Sep 21 03:06:43 I have this book about the BBB and it deals w/ diagnostics w/ trucks and OBD-II tech. Sep 21 03:06:52 ahh, I found the explanation in the car battery situation Sep 21 03:06:56 makes sense Sep 21 03:07:01 I cannot wait to test it out. Sep 21 03:07:03 Where? Sep 21 03:07:19 https://www.quora.com/Why-do-we-connect-the-positive-terminal-before-the-negative-terminal-to-ground-in-a-vehicle-battery Sep 21 03:08:07 Yep. Sep 21 03:08:21 A electric car, pardon the pun, is not good. Sep 21 03:08:31 ? Sep 21 03:08:31 bzzzt. Sep 21 03:09:34 And zmatt: Do you trust that site? Sep 21 03:09:59 That is the first time I have read off that site. Sep 21 03:10:10 makes sense Sep 21 03:10:15 Oh well. I guess we could grab a bunch of ideas from different sources. Sep 21 03:10:27 set_: the first explanation makes perfect sense Sep 21 03:10:32 Fine. Sep 21 03:11:08 Interesting, thanks for the input zmatt. Any idea if there's an on board regulator that might've failed before the IC? Sep 21 03:13:05 A lot of people on that page had different opinions. Sep 21 03:13:26 Who can you trust? Sep 21 03:13:31 yassyass: I don't think I've ever heard of such a thing. the pmic is pretty robust and moreover if it detects a problem (e.g. a short-circuit) it will simply cut the power automatically. in contrast, the processor is relatively fragile and unfortunately people manage to destroy it fairly regularly. at work we also have a small cemetary of beaglebones ;) Sep 21 03:13:58 sad Sep 21 03:14:11 Fare-the-well. Sep 21 03:14:21 set_: the first (highest-rated) answer makes perfect sense, I didn't bother reading any further Sep 21 03:14:48 Oh. I like the cat lovers' ideas but he is not correct. Sep 21 03:14:57 the who what? Sep 21 03:15:21 Further down, there was a feline lover or something like that. "66 years of feline servitude." Sep 21 03:15:37 That was one of the headers near his name. Sep 21 03:16:44 that "answer" doesn't even answer the question Sep 21 03:16:53 I know. That is why it is fun to read. Sep 21 03:17:02 Just ideas. Sep 21 03:17:09 zmatt: https://datasheets.maximintegrated.com/en/ds/MAX14851.pdf what do you think for protecting I/O ? Sep 21 03:17:56 brb Sep 21 03:18:12 CoffeeBreakfast: for what application? Sep 21 03:20:28 protecting the MPU agains short circuits Sep 21 03:21:19 a series resistor would do that too and is a lot cheaper Sep 21 03:22:02 short-circuits aren't usually the biggest worry though Sep 21 03:22:31 zmatt: I feel pretty bad for the dead board I have. I wonder what one could do with it. That isolator might not be a bad idea. Wouldn't a series resistor cause voltage drop issues? Sep 21 03:23:13 I just asking opinions, not suggesting anything Sep 21 03:23:41 overall this thing seems like an expensive chip for a weird mish-mash of IOs, and the sort of isolation it performs (which is the reason it's expensive) is rarely needed Sep 21 03:24:10 and it may still need overvoltage/ESD protection Sep 21 03:24:59 yassyass: what protection is needed and appropriate depends on the situation Sep 21 03:27:09 I think what might've killed my board was the use of pogo pins to program LED driver chips via i2c. I think the pins might've come in contact with a 24V rail on the PCB. Even though it was not seen to happen and we were successful in programming a bunch of chips until we realized the BBB was off. Sep 21 03:28:08 yep, that sounds like an easy way to destroy a beaglebone Sep 21 03:31:26 I wonder if the use of an opto-coupler of some sort would prevent this from occuring Sep 21 03:33:58 well you'd need something more complicated than an optocoupler since i2c is bidirectional. plus you'd probably end up destroying it if you have a similar accident, but I guess that's better than destroying the beaglebone Sep 21 03:38:44 there are i2c isolators, but, the usual way is paying attention Sep 21 03:40:29 you don't strictly need isolation for this anyhow, you need overvoltage protection Sep 21 03:40:51 although I guess isolation may be relevant if there's a risk the target is at a different ground potential Sep 21 03:41:29 I have to confess, I'm a bit paranoic when playing with I/O of the BBB =P Sep 21 03:44:52 maybe was useful, never burned out any I/O in my life Sep 21 03:55:28 I think I burned out an ADC pin before. Sep 21 03:55:35 I forgot to test it out any further. Sep 21 03:55:50 "Heat of the battle" and all. Sep 21 03:56:02 1.8v only! Sep 21 03:59:35 yep, the digital I/O use black magic to be 3.3v resistant, but that's not practical for analog inputs Sep 21 04:00:25 Right-o Sep 21 04:25:06 Have you powered your BBB via power supply? Sep 21 04:25:08 Variable? Sep 21 04:28:36 I have this knock off power supply. Cheap-o, Depot. For some reason, it states this in the instruction manual: power effect ion: CV <= 0.05% + 1mV. I look around online for "power effect ion." I came up short. Sep 21 04:28:52 I am afraid of what this power supply will do if I connect things to it. Sep 21 04:28:55 Forget it. Sep 21 04:28:59 I am going to chunk it. Sep 21 04:31:00 it might be some sort of voltage ripple spec with a description in broken english translated from chinese Sep 21 04:31:11 Aw! Sep 21 04:32:11 Maybe...since it does say ripples and noises: CV <= 10 mV (RMS)... Sep 21 04:32:20 But that is the third line. Sep 21 04:32:56 The one thing that makes sense is this: Output Technical Parameters. Sep 21 04:33:00 That is about it. Sep 21 04:33:02 does this thing have a part number or something? Sep 21 04:33:09 yes. Sep 21 04:33:17 It is unlisted. Sep 21 04:33:38 I looked and looked. I went through years of purchasing to find this specific part. Sep 21 04:34:10 I looked online at the retailers I probably purchased it from. I put in the alpha-numeric values and hoped for rain. Sep 21 04:34:11 ... Sep 21 04:34:17 No go...so far. Sep 21 04:35:07 Model# 3010D Sep 21 04:35:20 O - 30V and 0 - 10A Sep 21 04:35:44 I found many but none that looks like mine. Sep 21 04:36:32 it is a generic. Sep 21 04:38:41 no brand? Sep 21 04:38:50 ok Sep 21 04:40:21 Nope. Sep 21 04:40:23 No brand. Sep 21 04:42:01 If you read the instruction manual, you would flip out. Sep 21 04:42:39 The manual is full of nonsensical ideas. Sep 21 04:42:48 Typos and etc... Sep 21 04:42:58 ... Sep 21 04:44:10 I read one line and I felt way far from smart. I read the same sentence over and over until I realized that none of what I read over was sensible. Sep 21 04:44:11 ... Sep 21 04:44:13 lol Sep 21 04:46:58 well, i'm sure it's still fine as a power supply for many uses Sep 21 04:47:04 just don't leave it alone in case it catches fire Sep 21 04:47:18 ;) Sep 21 04:52:38 Right! Sep 21 04:53:21 I even have one of those easy-to-clean extinguisher items. Sep 21 04:54:10 It just sprays and the foam is easy to clean up unlike that dusty chemical in the "regulated" extinguisher items. Sep 21 04:54:32 Anyway... Sep 21 04:54:59 I just updated/upgraded all my boards. Sep 21 04:55:15 The BBBs are nice and fresh. I had to apply that fix on one of them. Sep 21 04:55:28 The fix about openCL or whatever that was about. Sep 21 04:57:11 zmatt: I really appreciate the extra support w/ the Python software you gave out earlier. Sep 21 04:57:12 ... Sep 21 04:57:30 The software has not been tested yet but I called it MotoDeluxe.py. Sep 21 04:57:56 I cannot wait to try that reverse feature you made. Sep 21 05:01:38 hopefully I didn't make some stupid mistake, it's obviously untested Sep 21 05:02:00 No issue. I can fix the small stuff. Sep 21 07:56:46 I'm a beginner but im really enthusiastic about the open source network and would like to contribute my best to this. I would like to get started but dont have anybody to mend me so please quide me through the ways i can contribute. Sep 21 07:58:15 Rosy: 1) pick and idea you want to explore and carry out 2) if you get stuck, ask 3) share the results. Sep 21 07:58:18 :-) Sep 21 08:11:24 I'm an enthusiastic beginner i don't know much about gsoc but im really interested in learning and contributing to open source network .But dont know what im actually supposed to do. Im eager to learn about open source. It would be really nice if you could help me with what i im actually supposed to do Sep 21 08:25:44 you might want to try #beagle-gsoc Sep 21 08:26:18 people in here don't necessarily know anything about gsoc either Sep 21 08:52:56 is it correct that the 3,3V power output pins (pin P1.14 and P2.23 of the pocketbeagle) are powered from the internal ldo TL5209 inside the OSD335x? Sep 21 09:00:58 dg7sgq: yup, and that supply is also used for various other stuff on the pocketbeagle (sd card, various pull-ups) Sep 21 09:01:13 hence the battery terminals are useless Sep 21 09:01:34 (in case that's why you were wondering) Sep 21 09:10:04 i measured in real 3,4V (as expected agains the datasheet of the OSD335x) Sep 21 09:11:21 there is also described what you said that not to use when battery terminals used Sep 21 09:12:29 therefore when using the SYS_VDD1_3P3V (+3,3V power output at PB) no bat can be used... Sep 21 09:12:48 yeah it sucks, it would have been better to ignore that ldo entirely and use the pmic's ldo4 output as single unified 3.3v supply for everything Sep 21 09:13:03 i wondering why its called 3,3V when in real 3,4 cames out?!? :-) confused Sep 21 09:13:19 I didn't know it was that inaccurate Sep 21 09:14:26 I don't think it is supposed to be Sep 21 09:14:51 datasheet of OSD335x says: "The OSD335x-SM contains a dedicated 3.4 VDC supply to power external circuitry. This is supplied by the TL5209, powered by the TPS65217C SYS output, and enabled by the TPS65217C LDO4." Sep 21 09:15:38 Footnote: "The LDO has an accuracy of 1 – 2% depending on the ambient temperature which will also affect the nominal voltage. See the TL5209 datasheet for more information." Sep 21 09:15:50 So all is okay... :-))) Sep 21 09:16:16 But need to know that on the +3,3V pins came 3,4V out... Sep 21 09:16:44 huh, they actually set it at 3.4v ? that's... weird Sep 21 09:17:28 jupp Sep 21 09:19:12 it really sucks that octavo basically copied the beaglebone design as-is instead of thinking about it carefully Sep 21 09:21:33 although the 3.4v setpoint is something they introduced, it should be closer to 3.3v on the beaglebone (based on schematics and tl5209 datasheet) Sep 21 09:28:07 weird Sep 21 09:28:52 with the 2% of the TL5209 it could be 3,468 in worst case... uuuuh Sep 21 09:29:10 thats hell for some chips Sep 21 09:31:39 yeah I don't know what the hell octavo was thinking Sep 21 09:46:19 at least on the beaglebone it's pretty easy to remove and bypass the tl5209 and create a unified 3.3v, but obviously that's impossible for the pocketbeagle. it doesn't even look like there's any way to solder onto vdd_3v3a at all Sep 21 09:47:25 only way I can think of to recreate a good 3.3v supply would be something like a voltage-tracking regulator whose vref is provided by a gpio that's driven high Sep 21 11:02:34 Hello, i am trying to connect to my 4-digit 7-segment display with a python file. I downlaoded all librarys but if i say i2cdetect i am not getting my 70 address which is default. How can i solve that ? Sep 21 11:12:09 zmatt: i will measure later VOH on a pin... will report; Sep 21 12:27:45 HansPeterKonrad_: does your i2c bus have pull-ups? Sep 21 12:29:43 hmmm, WTF... measured now the SPI1 interface, running at 3,4 too?!? wondering and scary; Sep 21 12:30:34 are the complete 3,3V rail is on 3,4V? Sep 21 12:55:07 dg7sgq: uhh no Sep 21 12:56:31 the I/O cells of the am335x are powered from "SYS_VDD3_3P3V", which is the LDO4 output of the PMIC Sep 21 12:58:17 are you sure your measurement is accurate? Sep 21 13:06:05 zmatt: measuring at lab environmental, think is accurate ;-) Sep 21 13:18:14 zmatt: try getting out a scope shot, moment... Sep 21 13:29:51 zmatt: https://pasteboard.co/HEUZRWm.png Sep 21 13:30:13 zmatt: C1 is power rail 3,3V out of the PB Sep 21 13:30:36 zmatt: C2 is CS line from PB to W5500 chip Sep 21 13:32:26 dg7sgq: isn't the external chip simply pulling up the line to its supply voltage? Sep 21 13:33:13 with no chip present on the lines, the voltage at the PB pins are equal to the 3,3Vout voltage Sep 21 13:33:27 measured with volt meter Sep 21 13:33:44 so i think no pull-up artefact from the W5500 Sep 21 13:34:09 btw... measuring the same voltage level at the unconnected SPI0 Sep 21 13:36:01 then I don't really have an explanation, the I/O supply of the am335x should be 3.3v ±2% Sep 21 13:38:55 i think so too, datasheet confirms that, scary Sep 21 13:43:49 btw 3,4V are in the range of the recommended op conditions for the W5500 AND the OSD335x, so i keep in mind and notice that.... puh! :-))) Sep 21 13:44:43 I don't think the osd335x specifies 3.4v anywhere other than as output voltage for the tl5209 ? Sep 21 13:45:04 the pmic outputs can't even be configured higher than 3.3v Sep 21 13:49:07 zmatt: yeah i think so too, but the datasheet says "Input/Output voltage range (with respect to VSS) All pins unless specified separately -0.3 to 3.6 V" so the 3,4V is safe, and i think the clamps or protection diodes will not break through or drain crrent at 3,4V; Sep 21 13:49:47 yeah it should be safe Sep 21 13:50:18 I think it might still leak current? Sep 21 13:53:27 I guess it probably doesn't (other than insignificant leakage) Sep 21 13:54:27 i think not, some uA e.g. Sep 21 13:55:07 not more than with wet fingers ;) Sep 21 14:25:25 hi BeagleFans ! i m working on beaglebone black since few days and i have a problem to configure some pins. I'd like to set them to "CAN" (=> "config-pin P9.20 can" and it works well ) but each time I reboot, pins are in default mode.. do i forget to do something ? thanks for your help and your time :) Sep 21 14:26:06 config-pin is not persistent Sep 21 14:26:43 so the described behaviour is the expected behaviour Sep 21 14:28:32 ok ty =) so i need to change the pin in a boot file or something ? Sep 21 14:29:20 for example Sep 21 14:42:51 m Sep 21 14:50:23 For those of you using a beagleboard x15, is your Chromium broken? Sep 21 14:52:59 https://pastebin.com/ZfWXNDqd Sep 21 15:06:14 Downgrading to 63.0.3239.84-1~deb9u1rcnee0~stretch+20171227 works. Sep 21 15:07:56 maybe report it to rcn? Sep 21 15:17:35 Just did. Sep 21 15:47:10 zmatt: damn, i made a mistake while measuring the voltage on the IO pins; the IO voltage on the lines are in range of what the PMIC suggest to have... sorry; but 3,3Vout stays at 3,4V! :-) Sep 21 16:36:23 dg7sgq: as expected **** ENDING LOGGING AT Sat Sep 22 03:00:00 2018