**** BEGIN LOGGING AT Thu May 14 02:59:57 2020 May 14 04:10:39 Well, I did it. My first real break-in. I broke into the chromebox and put Debian on it! May 14 08:52:44 does anyone understand what the "clear-capture" feature means for the 1PPS interface? It seems to be missing/broken on pps-gpio May 14 08:59:06 hello, i have a new bbb. I have insert the SD card with debian and then i inserted the usb. The power led is flashing and the others leds of usb are off. Someone know why please? May 14 09:52:26 Nacho: sounds like maybe your usb port can't supply enough power? May 14 09:53:11 so it's trying to power on, draws too much current, your usb port power temporarily shuts off, and this cycle repeats May 14 09:55:36 thinkfat_: pretty capture clear and capture assert mean it'll timestamp the moment the signal is asserted or cleared (deasserted) respectively May 14 09:55:40 *pretty sure May 14 09:57:46 thinkfat_: and yeah it looks like the pps-gpio driver supports it but only if enabled in platform data (with no option to do so via DT).. really weird May 14 09:59:29 using a gpio for timestamping sounds like it would suck anyway, but I guess noone has written a driver to support the AM335x eCAP (10ns resolution) or gptimer (41.67 ns resolution) instead May 14 10:14:48 eCAP sounds like it could be a good idea May 14 10:15:01 it's designed for the purpose May 14 10:15:49 it even has a 4-entry fifo, so it can measure the width of a pulse, or the time between two pulses, stuff like that May 14 10:19:40 well, the 1PPS signal right out of a GPS receiver jitters +/- 10ns or so anyway, May 14 10:20:29 and you only need to capture the edge that is UTC aligned May 14 10:20:52 yeah for this purpose I'm sure either eCAP or gptimer would do May 14 10:22:11 you'd kinda want to ensure the same timer is being used as the one used for the raw monotonic time anyway, otherwise you have the additional issue of having to figure out how to translate timestamps made using different hardware timers May 14 11:11:55 zmatt: I'm guessing there is some kind of driver for the eCAP already, only it cannot server as a 1PPS source, right? May 14 11:14:35 there is an ecap driver but I think only for using it as pwm output, but not sure May 14 11:15:25 (and I've made a clocksource driver for it, which I should probably try to get into mainline) May 14 15:05:26 note to self: applying a 10MHz signal to a GPIO with an interrupt attached is a good way to completely stall any CPU May 14 15:06:31 it really shouldn't, that sounds terrible May 14 15:07:03 it's spending 10ms (10 million cpu cycles) per interrupt? May 14 15:07:17 what the fuck is it doing, rendering a mandelbrot fractal for each irq May 14 15:07:35 oh wait May 14 15:07:36 I misread May 14 15:07:37 10 MHz May 14 15:07:43 yes May 14 15:07:45 :-) May 14 15:07:49 100ns cycle time May 14 15:07:56 yeah that will be a bit more problematic May 14 17:08:44 hey zmatt so got the encoder working fairly well I think. However, when I reverse the motor, the encoder counts still go up instead of down. Is this standard for the eqep lines? May 14 17:08:57 nope, that sounds like it's misconfigured May 14 17:09:19 maybe it's just in pulse-counting mode instead of quadrature decoding mode? May 14 17:10:08 maybe, i'll check in a few, going to put kiddos down for bed May 14 17:10:16 thought i'd ask and wait since typically people are busy :D May 14 17:19:04 @zm May 14 17:19:25 zmatt here's what I'm doing https://pastebin.com/QVB0sH8Y (encoder link is shown at the bottom) May 14 17:19:35 I've got it setup as rotary encoder...so I think that's right? May 14 17:22:21 I don't know anything about the adafruit stuff, the eqep counter is just a sysfs attribute that could be read with 1 line of pure python, eQEP configuration can't be done by userspace code in the first place since it's determined by DT May 14 17:23:21 I'm fully open to suggestions on a different way to read :D May 14 17:24:25 That was just the way that I found that made sense May 14 17:24:28 it does seem like cape-universal configures the eQEPs into quadrature mode by default May 14 17:25:09 So eli5: what is cape-universal? May 14 17:29:00 a better name for it would be the universal overlay, it's an ugly but convenient hack, an overlay that enables most peripherals (but without any pinmux declared) and exports all gpios (but without any pinmux declared), and creates special "pinmux helper" devices that allow pinmux to be changed at runtime May 14 17:29:47 instead of declaring your actual hardware configuration in DT, which is how it's supposed to work May 14 17:29:49 I see; I dont' think I've done anything like that. Unless it's in the Adafruit thing? May 14 17:30:06 it's enabled by default on all beaglebone.org images the past several years May 14 17:30:17 Gotcha. May 14 17:30:24 you've used it unless you've created your own DT overlay to configure the hardware May 14 17:30:34 So you dont' see a reason why it wouldn't be counting down? And no...I'm not smart enough yet for that :) May 14 17:31:34 no... that sounds really bizarre. like, it's not something that can be caused if one connection is bad or something like that, since then it wouldn't count up nor down May 14 17:32:12 It does count up, just not down if it's reversed (it still counts up) May 14 17:32:36 yes, there's absolutely nothing I can imagine that could cause that with a quadrature-encoded signal May 14 17:32:54 I'm inclined to suspect the adafruit library May 14 17:32:56 My only thing is maybe I have something hooked up wrong going to the diff to single ended chip? May 14 17:33:20 yeah no, I don't see any plausible connection issue that could cause it May 14 17:33:21 I've got a+ and a- going to the same input and b+ and b- going to the other? May 14 17:33:24 it just wouldn't count at all May 14 17:33:32 ehh May 14 17:34:00 I'm assuming your description is just a bit confusing here May 14 17:34:16 https://www.omc-stepperonline.com/download/17HS13-0404D-PG27-E1000.pdf May 14 17:34:19 Encoder bottom right May 14 17:34:53 I have EA+ and EA- going to one set of inputs on the quad interface. EB+ and EB- going to another input set, then the outputs going to the BBB May 14 17:35:00 yep, that's correct May 14 17:35:20 Wasnt' sure if EA+ and EB+ needed to go to the same input or not May 14 17:35:35 i.e. EA+ and EB+ going to one set and the minuses going to the second set of inputs May 14 17:35:53 nonono, (name)+ and (name)- form a differential pair May 14 17:36:03 I figured...I'm just baffled May 14 17:36:19 I may just get a 5V to 3.3V level shifter and try running single ended instead and see if that works May 14 17:37:02 can you confirm that /sys/bus/platform/devices/48304180.eqep/mode is "0" and /sys/bus/platform/devices/48304180.eqep/position shows the same behaviour of only counting up? May 14 17:37:30 I can try, just type that into terminal? May 14 17:37:38 cat FILENAME May 14 17:37:47 where FILENAME is one of those above May 14 17:38:35 Sure gimme a sec May 14 17:38:38 Yes the first shows up as 0 May 14 17:40:55 or if you want to read the position in python: https://hastebin.com/qogopitadi.py May 14 17:41:14 (not the most efficient way to read it but hey) May 14 17:42:13 Ok, let me see if while running it the terminal and cloud9 thing give the same answer (yes; using cloud9 for a bit) May 14 17:42:35 there is no conceivable way that matters May 14 17:43:16 i'm runnign this in cloud9 then have putty pulled up in a different window May 14 17:43:27 So I can compare between the Adafruit and the cat/position readings May 14 17:43:31 Shoulda clarified May 14 17:45:36 oh lol, I just noticed how awful the tieqep driver is May 14 17:46:01 (I'm browsing its source code) May 14 17:48:55 Both count up both ways May 14 17:49:11 I'll dump data on what I'm doing May 14 17:50:20 I don't suppose you have access to a scope to check the actual signals? May 14 17:51:27 I do! May 14 17:51:34 I'll check that here in a sec May 14 17:51:43 But yes; they're both counting up May 14 17:51:54 Wonder if my 1khz refresh rate is off May 14 17:52:21 eQEP has no "refresh rate" May 14 17:52:31 Update frequency May 14 17:53:10 it updates its counter on each edge of either signal, and when you read the "position" in sysfs the driver directly reads that counter, so you always have a fresh value May 14 17:53:52 (the "period" property only applies to relative mode, not to absolute mode) May 14 17:56:07 gotcha May 14 17:56:18 sec, i'm writing data to a text file so I can import it easier May 14 17:56:22 I suggest checking the signals going into eQEP for both motor directions May 14 17:56:22 They'rereading very close May 14 17:57:01 I mean, yeah I'd assume adafruit is just reading the "position" property and not doing anything weird with it May 14 17:57:13 so you don't really need to collect data on that if it looks like that's the case May 14 17:57:32 Yea; I'll go check signals May 14 17:57:41 I've got a 2 probe scope, so should be good enouh May 14 17:59:46 Will update you in a bit. Will probably post to an imgur or something to double check my brain. May 14 19:18:03 zmatt still working on getting the oscope hooked up, probably wont' be until tonight. Got pulled off getting a few other fires put out at work. Do you havea good suggestion for a 5V to 3.3V level shifter that would work well with the encoders? You haven't led me wrong yet May 14 19:25:30 something liek this might make it quick and easy (unless I'm missing something) zmatt May 14 19:25:31 https://www.digikey.com/product-detail/en/sparkfun-electronics/BOB-12009/1568-1209-ND/5673795 May 14 20:13:49 would need a schematic for it first obviously. May 14 20:22:01 biggi_: I really don't think the issue is with your receivers May 14 20:22:15 I was just going to run single ended to see May 14 20:22:18 biggi_: there's no plausible way you'd get it to count *at all* if that were the case May 14 20:22:19 And can't with it being 5v May 14 20:24:11 biggi_: I'd prioritize capturing the signals coming out of the receivers May 14 20:26:54 uhh hum May 14 20:27:27 "The main difference between the stepper motor with encoder and the closed-loop stepper motor is the receiving mode of encoder signal. The encoder signal of the encoder stepper motor is accepted by the control system, and the encoder signal of the closed-loop stepper motor is accepted by the stepper motor driver. May 14 20:27:36 oh my... gpsdo is working ;) May 14 20:28:05 biggi_: the datasheet you linked earlier says it's a "closed-loop" stepper (in the PDF title), not a stepper with encoder May 14 20:28:41 it definitely outputs. May 14 20:28:48 yes but what? May 14 20:28:57 Pulses May 14 20:29:23 very descriptive :P May 14 20:29:38 The aggervating part is I have this running on LabView May 14 20:29:49 I'll post screenshots this evening May 14 20:30:08 Once I get the oscope setup and tested since it got an oil bath at work (dont' as....possibly $2k scope ruined...will find out) May 14 20:31:39 ah well, could have been a $20k scope... May 14 20:31:48 Yea; we don't need one that nice. May 14 20:31:53 Not for what I do May 14 20:32:11 it's just really annoying that this company doesn't seem to have any documentation (that I've been able to find) regarding what the fuck the signals are May 14 20:32:20 Most of our stuff runs on an 8 bit 8mhz micro XD May 14 20:32:25 I agree. May 14 20:33:33 I'll capture it and let you know. I just hooked it into LabView and it worked no issue (running single ended into a counter module) so didn't have to do much troubleshooting. May 14 20:33:53 using what hardware? May 14 20:34:01 NI 9411 May 14 20:34:41 https://www.ni.com/pdf/manuals/373506a_02.pdf May 14 20:35:01 yeah that sounds like it's designed to support pretty much anything in software May 14 20:35:32 Yup; just popped in the daqmx reader and let it go to town XD May 14 20:35:58 By the way, not sure if you do much labview, but they're releasing a community version that's free for home hobbiest. May 14 20:37:08 I've used it a little bit a few decades ago May 14 20:37:40 what's the lure with labview? May 14 20:37:56 lazy\iness May 14 20:38:09 quick prototyping, tons of hardware support (from the same company), etc May 14 20:38:09 that is "it might be easier to do it this way" May 14 20:38:14 it's kind of a neat graphical data-driven programming language May 14 20:38:15 Easy to do state machines/etc May 14 20:39:00 it is good for one off things or just testing something. May 14 20:39:21 and spaghetti code in labview is far more literal spaghetti... https://thedailywtf.com/images/201104/labview.jpg ;-) May 14 20:43:24 it's also interesting that labview at heart is independent of evaluation order, like a pure-functional programming language, except impure elements have crept in due to pragmatism (and at some point they even added a way to introduce explicit sequence points to control when side-effects happen when using impure blocks) May 14 22:58:38 Right but one thing you need to remember, don't use Labview for production anything. It's only good for testing things. Namely their is little to no backwards compatibility when they change versions. Don't expect any either. It's a poor choice for anything that requires maintenance unless you have static hardware for development. It's really a lab tool. May 14 22:59:17 hence lab view. May 15 00:56:48 Hello...what is the easiest way to make sure my BBGW starts from SD when I uncommented that line in /boot/uEnv.txt file already and put in the SD Card? May 15 00:57:28 Could I? Could I commnet that line and re-insert the SD Card? May 15 00:57:51 I am trying to keep my data. May 15 01:09:07 Do not fret. I ruined everything. May 15 01:09:40 Hey. How did you do that? May 15 01:09:55 I just ruined my data on my SD Card. May 15 01:10:30 I am going to try some web dev. for "beginners" and I am already eight steps behind. May 15 01:10:37 Oops! May 15 01:10:51 Never fear, the BBGW has images. Hey! May 15 01:11:09 Did the Seeed Studio people come out w/ a new Green w/ Wireless and Ethernet? May 15 01:13:45 https://www.seeedstudio.com/SeeedStudio-BeagleBone-Green-Gateway-p-4586.html <<< Look! May 15 01:14:07 I knew I was not sleeping w/ the BBGW on my mind. May 15 01:14:17 Did you get that info I sent? I sent a message to the BBB Enhanced folks. Never heard back. Like is the serial port 3.3 or 5V, Can I power off USB OTG port, is Relay cape compatible, If case available for the BBB Enhanced WiFi External Antenna 1GB RAM.. May 15 01:16:04 I am not sure about the BBBE. May 15 01:16:10 I never once used it. May 15 01:17:38 So why is the BBGW called a gateway? And what are those two connectors on the back for? May 15 01:17:55 Oh. THat is one Grove UART and one Grove I2C. May 15 01:18:28 WHy it is called a gateway, I am not completely sure. There must be a reason. Maybe b/c of it being accessible to Bluetooth and Wifi enabled products? May 15 01:20:40 So the BBGG has WiFi , Ethernet and USB. Are the USB client or guest? May 15 01:21:30 I am not good w/ technical specificities. Ha. But...USB onboard is for external use. May 15 01:21:34 guest? May 15 01:22:41 KenUnix: Did you know you can change a Chromebook/box into a linux computer? May 15 01:23:01 i.e. Ubuntu, Debian, and etc. May 15 01:24:36 No. Tell me. May 15 01:25:24 Yep. If you crack one of them open, you can find a button or a screw. Push the power button on while the screw is removed or the button is held down. May 15 01:26:07 I set mine up for a 30 second delay in case I want to give it another go. May 15 01:26:42 You know. I just removed all my data and it is a bummer. Boo! May 15 01:28:26 Removed it to what a flash drive May 15 01:28:41 No. I lost all my data somehow on my BBGW. May 15 01:28:59 BBGG dosen't have a case yet May 15 01:29:14 I am not completely sure what exactly happened but I was fooling w/ images a while back and got disoriented. May 15 01:29:26 OH. The BBGG. May 15 01:29:31 I am going to get it. May 15 01:29:40 And put a bunch of neat stuff on it and make it work. May 15 01:29:54 Where is/was your data? May 15 01:29:59 Oh SD. May 15 01:30:15 I already overwrote the data w/ a new image. May 15 01:30:22 See. May 15 01:30:36 So, mount SD on P.C. and browse the folders May 15 01:30:52 I thought I had kept the data on this SD. It would not boot in the BBGW. May 15 01:31:10 I tried to boot it in eMMC mode but it would not boot. May 15 01:31:24 Anyway, no looking back. It is done. May 15 01:31:37 "Back to the ole drawing board." May 15 01:32:32 That's why I run off SD and only boot eMMc in an emergency. And make backups copies of SD. May 15 01:32:55 I always do. Just not always, i.e this time! May 15 01:33:14 zmatt to a good procedure for duplication SD May 15 01:33:31 I once knew of a script to back up the entire image. May 15 01:34:10 If you want I can email it to you os pastbinit it May 15 01:34:41 pastebin! May 15 01:34:48 ok, wait. May 15 01:35:16 Okay. May 15 01:37:10 Look https://pastebin.com/1TuqBuvD May 15 01:39:29 Oh. That is a bit more complicated than what I am used to now. May 15 01:39:45 Maybe I can try it one day once I figure out what exactly that does to the system. May 15 01:46:27 zmatt finally got the scope shots. Here is an album with them all https://drive.google.com/drive/folders/1NrGKgovlHJQXln2VcAOdT3Gg1qv5-vtF?usp=sharing May 15 01:46:44 the file names are what I'm probing in the order of the channels from top to bottom May 15 01:47:14 by looking at it, if i had to guess; I'd say there's an issue with the single ended outputs not being pulled to ground? **** ENDING LOGGING AT Fri May 15 02:59:57 2020