**** BEGIN LOGGING AT Fri Sep 29 03:00:01 2017 Sep 29 03:11:50 mozzo00: probably it's an older beaglebone (before rev C) that has 2GB of eMMC instead of 4GB ? Sep 29 03:36:56 mozzo00: here's an overview of the eMMC chips I've encountered on the beaglebone black so far: https://pastebin.com/raw/3fbxPhy7 (the first one on rev A5A, the other three all on rev C) Sep 29 09:48:30 hello Sep 29 09:49:45 can i know the name whom i chat Sep 29 14:40:54 Is anyone else having issues with second USB port on PocketBeagle? Sep 29 14:56:44 it has a second usb port ? Sep 29 14:59:50 NishanthMenon: yo! I keep forgetting to ask: what's the procedure for us to obtain that omap5 trm you received on a usb stick (swpu282af) ? can we ask our TI rep to contact you? Sep 29 15:01:00 zmatt, i dont own the trm - it was handed to me by one of the product engineering colleague (the guys who actually does silicon production line test). TI FAE should know who to contact in TI to get to the TRM Sep 29 15:01:26 zmatt, worst case ask her/him to mail me and I can direct the person to the right contact internally Sep 29 15:03:05 well he/she previously failed to get us remotely recent docs, so I'm not hugely confident of their ability to locate it on their own :) Sep 29 15:03:23 zmatt, hehe.. alright ask the person to ping me... Sep 29 15:04:39 thanks :) Sep 29 15:06:24 I also recently stumbled across rev H of swcs044 (for the twl6040 audio ic) floating around on the internet... the revision we received under NDA was rev C Sep 29 15:06:32 stuff like that Sep 29 15:09:08 (neither of which actually describes its registers, for that we have some very preliminary-draft-looking "Phoenix Audio Device" specification doc from 2010) Sep 29 15:20:19 zmatt, from my recollection, there is a RM (register manual) for Phoenix device which was different document to data sheet Sep 29 15:21:20 zmatt, i must warn you however I am completely out of touch with the legacy devices now a days.. even "recent" devices such as am57.. much of the information I probably recollect is stale and mouldy.. Sep 29 15:22:45 I do also have a register doc for it, but it's older than all other TWL6040 docs I have and is unfortunately not marked with a document ID Sep 29 15:22:48 hehe Sep 29 15:23:21 ughh Sep 29 15:23:42 well, probably most docs for omap5 and associated chips are stale and mouldy too... I'm just trying to gather the freshest-looking ones ;) Sep 29 15:24:14 yeah, running pdfinfo on that register doc helpfully includes "Title: Data Sheet Template" Sep 29 15:24:18 :D Sep 29 15:24:35 heh Sep 29 15:33:57 does phoenix actually refer to the audio IC itself? I got some impression that it referred to the omap4 (i.e. "Phoenix Audio IC" means "Audio IC for OMAP4") Sep 29 15:35:14 from my memory, Phoenix was a family just like OMAP was Sep 29 15:35:28 there is probably quiet a few devices under that.. Sep 29 15:40:25 the term is definitely also used for the omap4 pmic Sep 29 15:41:16 oh well, not really important, I was just curious Sep 29 15:43:01 zmatt, that is also why we prefer NOT to use internal code names in public patches - > names sometimes get reused, some marketing guy comes on later on, hates the name, changes it, confuses everyone else... it is better to use specific part numbers.. but then, sigh.. there are complications there as well.. nothing like clean story Sep 29 15:44:14 part numbers are a mess too Sep 29 15:44:27 yeah, I know.. Sep 29 15:44:30 e.g. people easily confuse AM35xx and AM335x even though they're completely unrelated Sep 29 15:45:09 while exactly the same die can receive a bunch of completely different part numbers based on which features are efused out Sep 29 15:45:22 so I actually prefer die names whenever I know them Sep 29 15:46:27 (when I want to refer to the overall device rather than a specific part) Sep 29 15:57:46 PocketBeagle is breaking out the second port to the expansion header P1 Odd pins 7 thru 15. When I wire to a connector and plug in a Flash key, I can see it knows there is something there, but I never get anything in /dev Sep 29 16:00:49 Morantz: how do you "see it knows there is something there" ? Sep 29 16:01:03 beaglebone kernel: [ 2.177867] usbcore: registered new interface driver usb-storage <--- only entry I see in log about it... Sep 29 16:01:34 does it appear when you insert the usb flash key ? Sep 29 16:01:43 2.177 sounds like very early boot to me Sep 29 16:01:56 The devices usb0 and usb1 change to usb1 and 2 when the key is inserted on boot. Sep 29 16:02:26 you mean the network devices? that sounds weird Sep 29 16:02:27 (Those are the gadget network interfaces...) Sep 29 16:03:00 lol... that is why I am here trying to figure it out! Sep 29 16:03:30 if it's detecting the usb key in any form or fashion, there would be more log entries about it Sep 29 16:03:55 stuff with "New USB device found" Sep 29 16:04:09 can you pastebin the kernel log ? Sep 29 16:05:08 hold on Sep 29 16:07:41 I was rebooting it.... Did a boot without it plugged in. Check kern.log, then plugged it in. no entries... Sep 29 16:08:28 then it's not detecting it in any form or fashion Sep 29 16:09:38 maybe i should try an older image.... Sep 29 16:09:48 oh, there's no power switch for usb1 on the pocketbeagle Sep 29 16:10:32 can i simulate one? Sep 29 16:11:16 you can try connecting P1.3 (USB1_VBUS) to +5V but I don't know if that will work Sep 29 16:12:09 I don't think it will, but it's the only simple thing you can do Sep 29 16:12:52 and don't forget to tie USB1_ID (P1.13) to ground Sep 29 16:13:06 ID is to gnd Sep 29 16:13:09 ok Sep 29 16:14:26 jkridner: btw, the battery terminals on the pocketbeagle are probably useless since iirc the OSD335x copied its power supply design from the beaglebone black, including the 3.3v regulator bug Sep 29 16:14:56 the 3.3V regulator bug is there, but you can always address it from a system level (not use the 3.3V output). Sep 29 16:15:06 zmatt: nothing on the board uses that regulator output. Sep 29 16:15:14 zmatt: it is just provided to the header for convenience. Sep 29 16:15:19 ah, that helps Sep 29 16:16:17 zmatt: for USB1, the pass-through USB power is just one option. You can use the VOUT (5V) which has a switch from VIN, though no regulator. Sep 29 16:16:19 if it's also not used for anything in the OSD335x itself then okay :) Sep 29 16:16:37 VOUT voltage == VIN voltage (or whatever is active from the regulator). Sep 29 16:16:50 * jkridner has noticed you can power the system via VOUT, but it is "odd". Sep 29 16:17:17 yeah the problem is that iirc the usb module's state machine expects usb_vbus to be below a certain threshold while drvvbus is low and above a certain threshold when drvvbus is high Sep 29 16:17:29 with fairly tight timing between transitions Sep 29 16:17:37 when using the mikroBUS FTDI Click, the VOUT is tied to the USB power "allowing" you to power the board via the Click. Sep 29 16:18:11 I'm pretty sure that using vsys to power the system is abuse Sep 29 16:18:16 zmatt: depending on what you are trying to do... which is why I just passed through both DRVVBUS and VBUS to the end user. :-/ Sep 29 16:19:21 it is safe to power the system from 5v via vbat, but with the caveat that vsys will always remain powered Sep 29 16:19:35 zmatt: I agree it is abuse, but a side effect of the FTDI Click. I initially freaked about it, but decided you can either cut that pin or use it. Yes, there is some excess current loss if the voltages don't match perfectly, but all components should be able to handle it. Sep 29 16:20:27 probably the pmic thinks it's battery-powered in that configuration Sep 29 16:21:05 ok. So no entries powering 1.3 or 1.5 from the 5v on 1.7 Sep 29 16:21:24 Morantz: kernel log would still be interesting Sep 29 16:21:36 1.3 ? what? Sep 29 16:21:45 never ever EVER put 5V on pin 1.3 Sep 29 16:21:57 you're lucky that the processor survived Sep 29 16:22:11 drvvbus is a 3.3v *output* Sep 29 16:22:31 (used to enable an external usb power switch) Sep 29 16:22:47 vbus is the only 5V-tolerant pin on the entire processor Sep 29 16:23:46 is P1.7 a 5v output? Sep 29 16:24:16 should I tie vbus to 3v3 and try powering the whole thing? Yes. looks like tied to input from USB0 Sep 29 16:25:44 dunno, in the schematic P1.7 is labeled "USB1_VIN" which I can't seem to find anywhere else Sep 29 16:26:02 jkridner: what's P1.7 ? Sep 29 16:27:25 Morantz: vbus (P1.5) should be provided with 5v Sep 29 16:27:36 it is vin.usb according to the eagle sch Sep 29 16:28:07 okay, I'm looking at the pdf Sep 29 16:28:13 so tie p1.5 and p1.7? then boot? Sep 29 16:29:24 if p1.7 is indeed a 5v output then yes Sep 29 16:31:41 looks like that is fix. /dev shows sda and sda1 now Sep 29 16:32:28 Morantz: cool, fortunately it seems it doesn't care about vbus being always-on then (when configured as host-only port anyway) Sep 29 16:33:51 I have not tried to hotswap.... might not work then... Sep 29 16:34:03 jkridner: ah, I see octavo actually added a note about the 3v3 regulator issue to the datasheet (about two months ago) Sep 29 16:35:33 Thanks zmatt and jkridner for your help! Sep 29 16:35:51 I will be back in an hour. Sep 29 16:38:18 jkridner: although I don't really agree with their description. the issue is that the TL5209 output remains on for too long during power down (even when not battery-powered), hence using it to drive (or even pull up) AM335x I/O results in a spec violation. The issue is just made worse by battery powering since then injecting current into AM335x I/O can result in the TL5209 remaining switched on Sep 29 16:41:43 tbr: sorry for delayed response, how do i check sd card usage? Sep 29 16:46:24 zarzar: look at the partition table? Sep 29 16:47:23 tbr: ok, will research that Sep 29 16:54:41 jkridner: ehh, at least according to the schematic pdf on github, the 3.3v from the tl5209 *is* used on the pocketbeagle, for the sd card Sep 29 16:57:27 zmatt: Eagle shows the same re:3V3 Sep 29 17:35:09 tbr: i am using and IOT image i think Sep 29 17:36:02 zmatt: kern.log is at https://pastebin.com/7fzrGV31 Sep 29 17:37:20 Morantz: if it works now it's not really important anymore :) Sep 29 17:37:23 zmatt syslog is at https://pastebin.com/vuKRqc0Q Sep 29 17:38:21 yea, would be interested in getting hotswap to work also.... Sep 29 17:38:44 oh, that doesn't work yet? Sep 29 17:38:49 but isn't the hot priority getting that port to work at all was... Sep 29 17:39:08 system gives tons of errors... Sep 29 17:39:45 log? Sep 29 17:41:15 https://pastebin.com/DHSqfg2n Sep 29 17:42:11 uhh, that doesn't look at all like it's related to vbus, that looks like a bug in the driver for the wireless stick you seem to be trying to use Sep 29 17:43:23 Yea, just tried pulling and replacing the flash stick and not getting that... Sep 29 17:44:27 was just able to do 3 removals and reinserts without error.... Sep 29 17:44:53 Not too worried about that wifi. ws just using it as something OTHER than the flash.. Sep 29 17:46:59 The first time I pulled that whole connector from the PB as i didn't want to rip any wires off. Sep 29 17:47:20 So i did get errors about the vbus Sep 29 17:49:06 musb-hdrc musb-hdrc.1.auto: VBUS_ERROR in a_wait_bcon (81, yeah, so it's waiting for session end (vbus going down) and never sees it Sep 29 17:52:35 so I did remember right :) Sep 29 17:57:15 @zmatt: I posted the fix for using that second usb on Google+ and gave you credit there. Sep 29 18:12:21 proper use of that port probably belongs in the FAQ Sep 29 18:13:44 (including how to connect an external power switch ic to get a fully working host port) Sep 29 18:15:05 (preferably including the 100 μF output cap that the BBB failed to include) Sep 29 18:47:03 zmatt: Do you have a link to the proper circuit? With Power switch? Sep 29 18:49:10 zmatt: I am laying out a Scarf for the PocketBeagle with Accel, temp sensor and TypeA USB Port. Sep 29 19:19:00 see for example the "USB HOST" section on page 4 of the beaglebone black schematic Sep 29 19:19:31 except that the 100 μF capacitor should be on the output-side of the TPS2051 power switch, not the input side Sep 29 19:19:39 (see also TPS2051 datasheet) Sep 29 19:20:56 in general it's worth checking the various datasheets (AM3358, power switch, esd protection) for guidance Sep 29 19:21:15 I think TI might also have application notes for usb Sep 29 19:37:02 ok. I will check that out. Thanks zmatt! Sep 29 20:00:26 how do i find uart4 in tty*? Sep 29 20:02:16 ttyS4 Sep 29 20:09:31 found it, thanks zmatt Sep 29 20:10:01 all the uart signals on my board are weak now, worked last week, not sure what happened. signals are not clean Sep 29 20:10:10 tried uart 1 and 4 Sep 29 20:15:20 "weak" ? Sep 29 20:20:42 zmatt: its strange, some signals look clean but it will stop in the middle of transmitting mutiple characters, other times the signals on oscope are not clean Sep 29 20:22:38 zmatt: On the PocketBeagle, which I2C is reserved for eeprom for addons? 1 or 2 Sep 29 20:24:34 Morantz: no idea, check the DT ? presumably the same as on all the other beaglebones Sep 29 20:25:11 (and "used for" would be a better choice of words, "reserved for" would seem to imply you can't use it for other purposes, while you can of course) Sep 29 20:25:49 zarzar: that sounds really strange Sep 29 20:30:05 yea not sure what is going on, spi works, gpio toggle works, uart no longer works consistently Sep 29 20:30:38 even with nothing connected to the pins? Sep 29 20:31:00 (to exclude the possibility that something is going wrong on an electrical level) Sep 29 20:31:12 yea, that's what i have now, no loopback, just oscope connection Sep 29 20:41:06 zmatt: I read differing things on net as to which port it is... Why I was hoping you might have known. Sep 29 20:43:53 Morantz: I see no evidence in the DT that either of them is used for such a purpose Sep 29 20:45:46 haven't checked u-boot though Sep 29 20:59:15 ok. thanks! Night all! Sep 29 21:34:38 zmatt: so uart4 is clean without a loopback but sloppy with a loopback Sep 29 21:39:12 zmatt: will try to figure out later, gotta run Sep 29 21:51:25 :q Sep 29 21:51:35 hehe. nvm Sep 30 01:07:01 zarzar: define "loopback" ? how long is the wire that you're using and at which end of it are you measuring? what are you seeing on the scope exactly? if it's about overshoot, use a shorter wire or place a small series resistor close to the transmitting side Sep 30 01:07:20 (e.g. 50 ohm ) Sep 30 01:09:06 Hello! Thank you, thank you. Sep 30 01:09:30 Are there any persons using the Grove OLED with their BBGW? Sep 30 01:09:56 Well, are there any in here using it? Sep 30 01:13:20 hello. Im trying to turn a soda machine into a riddle. Sep 30 01:13:44 Riddle? Sep 30 01:13:51 have some questions about hardware connection on pocket beagle. Sep 30 01:13:57 I am out. Sep 30 01:14:24 darn... Sep 30 01:14:25 <<< does not have a pocket bone Sep 30 01:15:12 Maybe some people will be here later. I just asked a question. I am waiting for people, too. Sep 30 01:16:09 set_: no, you seem to be doing a poll, you didn't ask any question of a technical nature :P Sep 30 01:16:29 Oh. Sep 30 01:16:33 I was starting out. Sep 30 01:16:52 I wanted to poll and then get in depth, i.e. set_ depth. Sep 30 01:17:15 I am trying to turn the vending machine into a "game" if you will. using the soda machine's buttons as raw inputs and using a screen so that the user can see questions. Sep 30 01:17:25 set_: asking whether anyone is using specific-thing-X with specific-device-Y is not very likely to get you responses Sep 30 01:17:52 zmatt: Did you know that the Linux beaglebone 4.9.35-ti-r44 does not have python-smbus installed by default. Sep 30 01:17:53 ? Sep 30 01:17:54 wannamaker: heh, sounds creative enough Sep 30 01:17:57 Sorry. Sep 30 01:18:14 how would one connect a screen to a pocket beagle for output? Sep 30 01:18:33 set_: why would it? 4.9.35-ti-r44 is the kernel version btw, which has really zero relevance for what's installed in userspace Sep 30 01:18:48 Aw! Off to install! Sep 30 01:19:00 wannamaker: depends very much on the screen Sep 30 01:19:44 set_: apt install python3-smbus or python-smbus (for python 2) Sep 30 01:20:25 Zmat: hmmm... havent gotten a screen yet. trying to make it easy. was thinking something like one you would see on an ATM Sep 30 01:21:23 I got it. Thank you. Sep 30 01:23:21 I am close to making this OLED work. Just have to listen to what happened in here a week ago. Sep 30 01:23:28 ... Sep 30 01:23:38 i.e. that pin-config "thang." Sep 30 01:24:28 wannamaker: since the AM3358's lcd controller isn't usable on the pocketbeagle your best bet is probably some small display that can be controlled via i2c or spi Sep 30 01:25:20 set_: which grove oled module are you using? Sep 30 01:25:51 96 x 96 v2.1. Sep 30 01:27:13 any videos or forums for these types of screens? Sep 30 01:27:42 wannamaker: the oled module set_ is using is one example :) Sep 30 01:28:47 OLED! Sep 30 01:28:52 googling a bit also found a C library for some of them -> https://github.com/erazor83/libsdisp also shows one in action Sep 30 01:29:27 would you need any extra hardware for connecting the OLED to the Pocket Bone? Sep 30 01:30:45 sorry if my questions are dumb or dont make sense very very new to this. Sep 30 01:30:48 probably two pull-up resistors for the i2c bus Sep 30 01:31:14 4.7 kΩ or something like that Sep 30 01:33:25 if you find a module with spi interface, that wouldn't require any additional components (and is also much faster than i2c) Sep 30 01:34:30 Hey zmatt: Remember that stuff you wrote the other day and pushed it on gitHub.com? Sep 30 01:34:41 e.g. I'm reading the datasheet of the SSD1327 controller used in some modules, and it supports max 400 kHz via i2c versus 10 MHz via spi Sep 30 01:34:53 To see the pin configuration. Sep 30 01:34:58 set_: you mean the bbgw branch of my pin utils? Sep 30 01:35:03 Yes/ Sep 30 01:35:17 Can you please let me know where to find it again. Sep 30 01:35:20 ? Sep 30 01:35:34 https://github.com/mvduin/bbb-pin-utils/tree/green-wireless Sep 30 01:35:37 I am going to follow it this time. Sep 30 01:35:56 Thank you, zmatt. Sep 30 01:37:49 any advice on controlling a LED strip with a Pocket Beagle? Sep 30 01:38:05 what kind of led strip? Sep 30 01:40:26 e.g. the popular ws2812-based protocol often used for led strips is a bit annoying to produce, but a PRU core can produce it easily Sep 30 01:44:09 yeas they are ws2812 Sep 30 01:47:00 yeah, like i said it's a bit annoying protocol to produce, but it's popular. fortunately the PRU cores on the beaglebone excel at doing weird protocols :) Sep 30 01:47:40 searching for "ws2812 pru" should yield results Sep 30 01:48:02 !!! :) :) :) :) Sep 30 01:48:20 Anybody go to Makers Faire New york? Sep 30 01:51:21 any advice on hardware connection for WS2812 LEDs to a pocket beagle? Sep 30 01:51:54 I'm not sure if 3.3v signalling suffices for them Sep 30 01:52:53 some quick googling says no Sep 30 01:53:41 so a buffer / level-shifter would be needed Sep 30 01:53:50 or some disgusting hack like this: https://hackaday.com/2017/01/20/cheating-at-5v-ws2812-control-to-use-a-3-3v-data-line/ Sep 30 01:57:04 Dudes and gals...the pin P9.20 is not "modifyable" on the BBGW. What gives? Sep 30 02:00:00 I have an OLED and I cannot access my i2c address = no worky OLED for now! Sep 30 02:00:25 Send help if anyone has time. Sep 30 02:01:56 uhh, in what sense would you want to modify P9.20 ? Sep 30 02:02:14 it's configured for i2c by default, which is also what you want Sep 30 02:02:35 do you have pull-up resistors on the bus like I mentioned earlier? Sep 30 02:03:28 (unless the bbgw has those integrated, didn't check) Sep 30 02:03:28 I need to make my address available for me to use my i2c Grove connector. Sep 30 02:03:44 what do you mean by that? Sep 30 02:04:09 I mean...there is a Grove Connection between a BBGW and a OLED display. Sep 30 02:04:23 yes, and? Sep 30 02:04:47 I tried to configure my pins but the the pins I tried to configure are not modifyable. Sep 30 02:05:01 because they're already configured for i2c ? Sep 30 02:05:09 Right... Sep 30 02:05:12 which is what you want Sep 30 02:05:46 My error reads Error accessing 0x3c: Check your I2C address Sep 30 02:06:02 04:02 < zmatt> do you have pull-up resistors on the bus like I mentioned earlier? Sep 30 02:06:17 I will look at the docs. Sep 30 02:06:19 Please hold. Sep 30 02:07:12 I think that i2c port has one built in. Sep 30 02:07:57 It shows a resistor is there. Sep 30 02:08:15 Now...for my OLED. No clue. I will check. Sep 30 02:08:16 I don't see any in the BBGW schematics Sep 30 02:08:21 Hold. Sep 30 02:08:25 I will get it. Sep 30 02:08:26 and the display is very unlikely to have them Sep 30 02:08:32 Okay. Sep 30 02:08:40 I will get the schematics on the BBGW. Sep 30 02:08:40 so you will probably need to add them yourself Sep 30 02:08:54 I just checked the BBGW schematics, no pullups Sep 30 02:08:59 The BBGW has them onboard for the i2c. Sep 30 02:09:01 Are you sure. Sep 30 02:09:06 ? Sep 30 02:09:21 yes Sep 30 02:10:05 you need 4.7 kΩ from SDA to 3.3v and 4.7 kΩ from SCL to 3.3v Sep 30 02:10:06 I see a resistor in the schematic. Sep 30 02:10:44 Oh. It only has a 232 at 1%. Sep 30 02:10:48 if you mean R231, it's 0 Ω Sep 30 02:11:00 Damn. How did you know that? Sep 30 02:11:10 because it says "0R" Sep 30 02:11:14 Oh. Sep 30 02:11:25 and it's in series with the 3.3v and not connected to sda or scl Sep 30 02:11:46 So, I just rip apart the wiring and do it, I guess. Sep 30 02:11:58 Dang. More work on my part. Sep 30 02:12:07 life is hard Sep 30 02:12:14 I know. Sep 30 02:12:36 Believe me, I know. I thought making my bone fly, no pun intended, was easy. Sep 30 02:13:08 Okay. Please get this... Sep 30 02:13:22 Why is there even a Grove connection if the connection is faulty? Sep 30 02:13:45 I will ask Seeed or whoever produces Grove. Sep 30 02:14:14 Sorry...no offense. Sep 30 02:15:03 I don't think anyone from Seeed is here to take offense Sep 30 02:15:52 Oh. Sep 30 02:19:14 http://wiki.seeed.cc/Grove-OLED_Display_1.12inch/ Sep 30 02:19:29 ^ That is it. That is the OLED display. Sep 30 02:25:10 ... Sep 30 02:25:38 zmatt: How did you come up with 4.7 ohms? Sep 30 02:25:51 4.7 k ohms? Sep 30 02:36:44 Sorry. I cannot figure out where amps is located. Call the Pope! Sep 30 02:40:41 I am going to find out where this A rating is on the board and quickly. Just you wait. Sep 30 02:42:10 Please keep waiting. Sep 30 02:51:18 No one said I was the smartest. Sep 30 02:57:34 So, I = 0.0007 basically. I need the 0.0007, with pullup resistor included, to hit my GND and 3.3v supply. Sep 30 02:57:36 yea boy! Sep 30 02:59:49 My issue with life, BBB included, is to simplify. **** ENDING LOGGING AT Sat Sep 30 03:00:01 2017