**** BEGIN LOGGING AT Fri Sep 20 02:59:57 2019 Sep 20 03:00:07 Last laugh, so. Gleaming in the glory, totally worth it. Sep 20 03:15:27 set_: that data looks good, but not quite what I expected so my script will fail because it's a bit too strict.. I'll fix that Sep 20 03:16:09 fixed Sep 20 03:16:42 Let me check on it. Sep 20 03:16:49 line 22 is what changed Sep 20 03:16:53 Nice. Sep 20 03:16:54 Okay. Sep 20 03:17:35 What link was it? Sep 20 03:17:45 https://pastebin.com/RbDjLFjM Sep 20 03:17:48 aw. Sep 20 03:20:53 Dang. That MaxBotix ez2 is fast. As soon as I put my hand over it, it went from 60 to seven inches. Sep 20 03:21:08 @zmatt in 'da hizzy. Sep 20 03:21:32 Well, what now? Sep 20 03:21:43 Should I put your name on it and change it? Sep 20 03:22:07 no need, just use it Sep 20 03:22:16 Or just: # w/ help on Freenode at #beagle? Sep 20 03:22:22 Okay. Sep 20 03:22:24 Nice! Sep 20 03:23:15 To be fair, it is complicated but i can learn in time what everything in the script means. Sep 20 05:07:58 Hi there. I've been planning to purchase myself a Beagleboard X15 for a little while now, but it seems to be unavailable from the linked distributors on the beagleboard.org website. Sep 20 05:08:14 Has the product been obsoleted, or are there plans to put more into production in the future? Sep 20 05:08:49 I was really excited to find that there was a dev board with all the nice IO options the x15 comes with Sep 20 05:09:17 Would be a terrible shame if it was no longer available and there wasn't something as good or better to replace it with Sep 20 05:12:19 Cancel my question, looked a little harder and it seems the distributor links on the official website, as well as the links google offered me for those distributors, are out of date. Sep 20 05:13:23 looks like they're switching manufacturer? since at mouser the original is marked EOL but this one is marked "new from this manufacturer" and has 499 units on order => https://eu.mouser.com/ProductDetail/?qs=mAH9sUMRCtul7OWq6qLH4g%3D%3D Sep 20 05:14:38 if so the links on the site should indeed be updated Sep 20 05:14:44 jkridner[m]: poke? **** BEGIN LOGGING AT Fri Sep 20 09:13:30 2019 Sep 20 13:50:34 what'd I miss? Sep 20 13:50:57 oh, yeah, switching from PTI to Seeed. Sep 20 13:51:27 taking a minute to get the new boards flowing. Sep 20 13:51:51 did the stock on the PTI boards go to 0? Sep 20 13:52:29 looks like it. Sep 20 13:53:32 dang, Digi-Key stock went to 0 too. Sep 20 13:55:47 guess folks thought that AI would take it over. Sep 20 13:56:09 Seeed should ship to Mouser w/in a month. Sep 20 13:56:30 they finished their test run and have gotten all the parts in. Sep 20 13:57:44 zmatt: help? Sep 20 13:57:54 * jkridner[m] wishes bbx15-buyer would have stuck around. Sep 20 13:58:17 though, I guess the web mostly answer the query Sep 20 14:01:14 help? Sep 20 14:01:30 yeah this is what it looked like Sep 20 16:24:53 zmatt: oh, I was just wondering if my response was useful. guess I just confirmed what you inferred. Sep 20 16:55:30 I was mostly just saying that if the product page changed at mouser than the links on the site should be updated Sep 20 16:55:36 *then Sep 20 17:10:50 hi! Sep 20 17:10:59 May you help me? Sep 20 17:11:47 I want to know how Beaglebone Black Wireless SD image is flashed Sep 20 17:13:59 I try pressing the user boot button and released after the 4 Leds turn on but after waiting for more than 45 minutes the Leds still flashing... Sep 20 17:54:35 anyone tried the alpha image for the BB AI? Sep 20 17:55:06 Looks like the one to start with if you want Debian 10 Sep 20 18:09:23 If using the ez2, would I need an inverter to work w/ PWM singals? I say no b/c I am not using the TX pin on the ez2. Sep 20 18:09:52 I am asking b/c I found a script for pw on the ez2. Sep 20 18:10:15 Wait for the paste. Sep 20 18:13:36 https://pastebin.com/amuVK05L is a rendition but I am sure I can alter it to suit my needs later. Sep 20 18:15:00 off to test it. Sep 20 18:16:22 also curious if anyone has tried getting Ubuntu on the Beaglebone AI Sep 20 18:22:00 looks like it comes with stretch. maybe no need to flash anything? Sep 20 18:22:11 unless I want buster Sep 20 18:35:34 im trying an apt upgrade to stretch, but I have a feeling this will take much longer than flashing. hehe Sep 20 18:37:24 otay. BBL! Sep 20 19:39:21 wow /usr/local is kinda crowded Sep 20 19:40:40 anyone know why there are so many files in /usr/local/bin? Sep 20 19:40:52 I'm used to seeing that pretty empty Sep 20 20:02:49 hello, i just testing my new toys BB AI... two problems appear: overheating after few minutes after booting and no JTAG path (after soldering port). Does anyone have such problems ??? Sep 20 20:39:19 set_: that code seems to make no sense, where did you get it from? Sep 20 20:40:38 set_: recall btw that your current problem isn't the inverter, but rather something being horribly wrong with powering the device (which will affect any method of getting the information out of it) Sep 20 20:40:58 the inverter seemed to be working fine Sep 20 20:41:30 so once you fix the power issue, you'll probably be able to receive the serial data Sep 20 20:42:49 using the pulse width output of the module would indeed avoid the need for an inverter, but I'm not sure if there's an easy way to measure the pulse width... the hardware has excellent capabilities but there's poor software support) Sep 20 20:43:45 I'd probably use an ecap module via uio Sep 20 20:45:19 but using the serial data is simpler from a software perspective, and you already have a working inverter, so there's no reason to try to meddle with the other outputs of the module at this point Sep 20 20:46:46 if you suspect your breadboard, then just connect ground and power of the ez module directly to the beaglebone (P9.01 or P9.02 for ground, P9.03 or P9.04 for power) Sep 20 20:48:40 and double-check that your inverter is getting proper power by checking that its output is around 3.3v when its input is disconnected Sep 20 22:09:45 Hi all - I caused a problem with my rcpy drivers. Can you help me troubleshoot? http://beagleboard.org/chat Sep 20 22:10:21 this is a gist of some of my error checking. https://gist.github.com/dmalawey/f3b5156535139cd8d5c4c77ec1d39ee2 Sep 20 22:11:36 In summary, I cannot send commands to motors on my blue. The Line 31 of the gist shows "broken packages." How can I repair these packages? Sep 20 22:12:22 Line 42 indicates a more specific problem: "ERROR: ti-pwm driver not loaded for hrpwm1 Sep 20 22:25:12 ehh, how did you get into this mess? :P Sep 20 22:26:29 might first want to resolve the package conflict... try "sudo apt-get purge roboticscape" Sep 20 22:27:03 if that succeeds then "sudo apt-get install libroboticscape" Sep 20 22:27:04 eh Sep 20 22:27:08 librobotcontrol I mean Sep 20 22:30:00 the thing about the pwm sounds like a problem with the DT, can you share your /boot/uEnv.txt ? (e.g. "pastebinit /boot/uEnv.txt") Sep 20 23:05:40 I will try the purge and reinstall. (I may have caused this mess by trying the wrong instructions installing in the first place.) Sep 20 23:11:45 Hooray it worked! Probably the uEnv.txt will not be meaningful now, but here it is Sep 20 23:11:51 http://paste.debian.net/1101814/ Sep 21 00:23:05 @zmatt: Okay. No issue. The breadboard was the issue yesterday. The inverter works and the software works. I just wanted to try w/ the PW signals. Sep 21 00:23:12 I will leave it alone for now. Sep 21 00:23:35 Well, until I find something that works relatively well. Sep 21 00:28:18 student: your uEnv.txt seems fine at least Sep 21 00:29:14 set_: serial should work fine once the ez module is properly powered Sep 21 00:33:07 id just like to say that the way you all ship the beagle bones is amazing. Sep 21 00:33:38 shows up as ethernet host on usb-c, wifi turns on, flashing to emmc super easy... Sep 21 00:33:42 serial works fine w/ your src and w/ the src I found. Sep 21 00:34:12 set_: oh right sorry I just remembered that you actually did fix the issue last night Sep 21 00:34:16 sorry Sep 21 00:34:23 Right. No issue. Sep 21 00:34:31 It was that dang breadboard issue. Sep 21 00:34:47 One of my breadboards was a rip off or something. Sep 21 00:35:02 Loose connectors. Sep 21 00:37:10 im kinda tempted to attempt an ubuntu image Sep 21 00:37:25 why? is there even one? Sep 21 00:37:39 there is for the black Sep 21 00:38:05 an unofficial/unsupported one yeah Sep 21 00:39:31 well we chose it for the LTS generally--and while Debian is similar, some of our sysadmins are a bit cut and pasty Sep 21 00:40:08 ubuntu on beaglebone has no "support", let alone long-term :P Sep 21 00:40:34 but if it works it works I guess Sep 21 00:40:36 there are more than beaglebones in the world Sep 21 00:41:34 looks like rcn has an ubuntu image Sep 21 00:41:51 not sure if it would work on the beaglebone ai though Sep 21 00:44:18 thing is, there are hardware-specific aspects, and an unofficial image has much fewer users and gets way less testing... which means the risk of issues is greater, and it'll be harder to find help for it Sep 21 00:44:24 oh cool. found a link for it Sep 21 00:44:55 that's true Sep 21 01:40:17 https://beagleboard.org/Support/BoneScript/Ultrasonic_Sensor/ Sep 21 01:40:20 Look at what i found! Sep 21 01:40:22 Nice. Sep 21 01:52:20 set_: that seems like a significantly worse way to use the sensor Sep 21 01:53:40 set_: why are you looking at alternate ways to interface the sensor when you have a working setup already? Sep 21 01:55:15 @zmatt: I am looking at another way to use the sensor b/c I want to learn more. Sep 21 01:55:43 the circuit drawn there is a very bad idea and a recipe for damaging a beaglebone Sep 21 01:56:18 I understand but the AN output of the ez2 may be an option. I have been reading. Sep 21 01:56:45 If you are bored, look here: https://www.maxbotix.com/ultrasonic-sensor-hrlv%E2%80%91maxsonar%E2%80%91ez-guide-158. Sep 21 01:57:18 I can put the correct voltage in on my power supply, 3.3v, and find my reading. Sep 21 01:57:42 jkridner[m]: https://beagleboard.org/Support/BoneScript/Ultrasonic_Sensor/ this is pretty much a guaranteed absolute-maximum-ratings violation due to power sequencing between the 5V and the 1.8V (in fact the drawing uses "vdd_5v" which can be present continuously while the bbb is powered off) Sep 21 01:57:49 set_: it is an option, it is also the worst option Sep 21 01:57:54 Oh. Sep 21 01:58:20 Why is this the worst option? Sep 21 01:58:22 the module has the information you want in digital format, you want it in digital format. converting it to analog and then having to measure that analog voltage is garbage Sep 21 01:58:28 if you are really bored - use GPIO to generate the ultrasonic signal then drive that into a transistor coupled into a piezo element. Have a 2nd one near it wired to an op-amp then feed that into the ADC. For processing ease, couple the GPIO into another ADC channel with a divider Sep 21 01:58:40 then calculate the distances with autocorrelation Sep 21 01:58:57 Oh. Sep 21 01:59:13 the adc is also noisy, the output impedance of the analog output is unspecified I think so you won't know whether the voltage divider might affect the readings Sep 21 01:59:23 and like I said it's a power domain violation hence a recipe for damaging a beaglebone Sep 21 02:00:35 passing the measurement as digital information is the best solution, and is exactly what using the serial interface offers Sep 21 02:00:46 The second set up is better from the second link. I can use 3.3v, measure the mV on my 3.3v power supply coupled w/ the multimeter, and solve the equation. Sep 21 02:01:04 But yea. You are right. Sep 21 02:01:13 I could just keep this serial output in use. Sep 21 02:01:15 No issue. Sep 21 02:02:29 ds2: Do you use ultrasonic beams? Sep 21 02:02:38 Sonar! Sep 21 02:04:52 there's also an appeal in using the pulse output because it would allow a connection with just wires and no inverter or other electronics between the bbb and the sensor, but that requires a software solution for accurate pulse width measurement (preferably using the hardware features available for this) Sep 21 02:04:55 I play with it Sep 21 02:05:11 since you struggle enough with software already, I'd suggest sticking to serial since it is simpler and already working Sep 21 02:05:31 software is easy with the PRU Sep 21 02:05:47 for me, for you, not for set_ Sep 21 02:05:56 but LIDAR is more fun then SONAR Sep 21 02:06:01 PRU seems like a waste for this actually I'd just use an eCAP module Sep 21 02:06:22 if that pin is available Sep 21 02:06:36 PRU has a higher probility of being on an available pin :/ Sep 21 02:06:46 otherwise, I completely agree Sep 21 02:08:11 the timers (P8.07-10) can also be used Sep 21 02:10:03 though not as conveniently as eCAP Sep 21 02:10:26 but yeah PRU is always a option too Sep 21 02:11:09 I found that source for the rpi earlier and was trying to alter it to work w/ the pw pin on the ez2. No go so far and yes, you are right, I struggle w/ software stuff. Sep 21 02:11:39 I have not practiced enough or gained the knowledge to handle most things. Sep 21 02:11:51 <<<< still reading stuff Sep 21 02:13:20 rpi software to do pulse width measurement is never going to work on a bbb if it makes use of hardware-specific features, and is not going to be accurate or reliable if done without making use of hardware-specific features Sep 21 02:13:35 Oh. Sep 21 02:14:02 Okay...I thought they just used the pw pin and I could transfer that software to the BBB w/ some alterations. Sep 21 02:14:52 no Sep 21 02:15:07 for instance, instead of rpi-gpio or whatever they use, I could use Adafruit_BBIO.PWM as PWM. Sep 21 02:15:17 there's absolutely no use for pwm here Sep 21 02:15:49 Oh? But the pw pin on the ez2 is for pulse width. Sep 21 02:16:16 pwm refers to a pulse-width modulated output of the bbb, not pulse width measurement Sep 21 02:16:20 which is a completely different thing Sep 21 02:16:25 Ouch. Sep 21 02:16:52 See. I am still reading. Okay. No issue. Probably another analog pin on the ez2, heh? Sep 21 02:16:59 Sheesh. Sep 21 02:17:19 like I said a few times already the bbb has fine hardware capabilities for it but afaik no kernel driver or convenient ready-to-use software solution afaik Sep 21 02:17:43 Okay. I am done. You win. I made my mistakes but what works so far is as good as I could have hoped. Sep 21 02:17:47 so I suggest sticking to serial Sep 21 02:17:54 serial, it is! Sep 21 02:18:41 unless someone offers a working ready-to-use software solution for pulse width measurement (using PRU or eCAP) Sep 21 02:18:57 Aw! Sep 21 02:19:05 Okay. I was wondering why you said eCAP. Sep 21 02:19:21 pulse width measurement <<<<< off to look this up Sep 21 02:20:54 because that's a hardware module capable of being used for timestamping signal edges (with 10ns resolution) which can be used e.g. for measuring pulse width or pulse interval Sep 21 02:21:05 (the same module can alternatively be used as pwm output, which is how it's more commonly used on the bbb) Sep 21 02:21:37 Oh. So, I could use the sample I find to use PWM output on my BBB? Sep 21 02:21:47 ???? Sep 21 02:22:41 Do not fret. Sep 21 02:22:55 I am not going to make you understand. Sep 21 02:23:03 I am most likely incorrect. Sep 21 02:32:24 https://os.mbed.com/questions/1433/Pulse-Width-measurement-trouble/! Sep 21 02:32:27 Check it out. Sep 21 02:32:44 I found a lot of articles on this subject. Sep 21 02:32:58 unless they're for the beaglebone, they're of no relevance Sep 21 02:33:24 Oh. Sep 21 02:33:26 Dang. Sep 21 02:34:23 Otay. Up, up, and otay! Sep 21 02:36:37 just use serial, it already works Sep 21 02:48:34 think of it this way, you have a counter Sep 21 02:48:39 it is driven by a fixed clock Sep 21 02:48:51 on edges, it saves the value of the counter Sep 21 02:48:59 which edge, etc is somehting you config Sep 21 02:50:48 Alright. **** ENDING LOGGING AT Sat Sep 21 03:00:00 2019