**** BEGIN LOGGING AT Mon Nov 05 03:00:00 2018 Nov 05 03:08:47 Too many people want to run IP cameras Nov 05 03:08:56 those things really saturate the 2.4G band Nov 05 03:09:43 Oh. Nov 05 03:10:09 Like...for security and things like chatting w/ face-to-face? Nov 05 03:10:15 so in a residentual area, if all your neighbors run IP cameras, it just sucks Nov 05 03:10:21 mostly for securitying as far as I can tell Nov 05 03:10:22 Aw. Nov 05 03:10:27 Makes sense. Nov 05 03:10:37 then you got the people doing wireless headphones over BT Nov 05 03:10:42 same pool of bandwidth Nov 05 03:11:00 I know someone who just caught a person digging through his car on camera. Nov 05 03:11:04 ... Nov 05 03:11:16 He is ruining your gigs man. Nov 05 03:11:19 Ha! Nov 05 03:12:21 I was going to try to set him up w/ a website and security system for fun w/ the BBB. Nov 05 03:12:22 ... Nov 05 03:12:39 He is used to some fancy amazon stuff. Nov 05 03:13:20 He could start a motor to scare them off too! Nov 05 03:13:25 vroom! Nov 05 03:14:10 One online button and invaders beware. Nov 05 03:15:43 vroom, vroom, bzzzt! Nov 05 03:17:06 Or...the fellow could ground the car at night. Ha! Nov 05 03:17:20 bzzt, bzzt, bzzt. Nov 05 03:19:59 <\dev\cache> so no one has any information about a case for an BB X15? Nov 05 03:21:02 I bet you can fabricate one! Nov 05 03:21:39 Aluminum is easy to bend w/ a brake and most are fairly cheap. Nov 05 03:23:44 <\dev\cache> i would prefer to buy one or look for some designs and get someone else to manufacture it Nov 05 03:24:04 Oh. Let me look online real quickly. I think I found something not too long ago. Nov 05 03:25:55 https://www.thingiverse.com/thing:1330724 is a cheaper, acrylic case. Nov 05 03:26:02 Do you want metal indefinitely? Nov 05 03:27:08 just mill it out of a solid block of aluminum! Nov 05 03:27:20 https://www.yeggi.com/q/beagleboard+x15/ is another site I found a while back. Nov 05 03:27:30 They have 3D printable parts and guides. Nov 05 03:27:49 mills are nice! Nov 05 03:28:14 parts = files Nov 05 03:28:16 Sorry. Nov 05 03:29:19 Oh and if you have a laser engraver, 100% on cutting power will cut that acrylic nicely. Nov 05 03:30:47 yes..mills are. It'd be a good exercise to get started with a bridgeport Nov 05 03:31:06 bridgeport! I am off to look at one. Nov 05 03:32:21 Dang. I knew a fellow that had one. He forgot to fix the sideward movement w/ stepper interface. Nov 05 03:33:09 That would be cool to diddle w/. I always wondered why they needed to be so heavy. Nov 05 03:34:00 brb Nov 05 03:36:42 you want it to cut not bend Nov 05 03:36:51 mass makes it harder for it to bend Nov 05 03:40:51 Oh! Nov 05 03:42:16 I just found some very cheap angle iron. I am going to use it to produce something. Nov 05 03:42:41 I think that my BBB needs to be better equipped w/ heavier items. Nov 05 03:43:49 I found some motors that seem to be perfect but they are expensive. I need to take some time to keep searching. Nov 05 03:46:34 Did anyone see that contraption from Maker Faire in NY? Nov 05 03:46:56 It seemed like a large, extra equipped dragon of some kind. Nov 05 03:47:47 It seemed to be three stories high. Yikes! Nov 05 03:53:22 Oh and there was that Hackaday Con thing, too. Nov 05 04:07:02 ... Nov 05 04:07:41 Hey...I thought this may be interesting, although out of bounds, to people in here: https://www.nbcnews.com/mach/science/scientists-are-trying-bottle-solar-energy-turn-it-liquid-fuel-ncna930676?cid=sm_npd_nn_fb_ma. Nov 05 04:07:46 ... Nov 05 04:07:53 I thought this was a cool subject. Nov 05 04:38:30 That thing in NY did not happen. It was in France and it was a Minotaur. Sorry for that bad info. Nov 05 15:36:36 m Nov 05 17:54:11 morning Nov 05 17:54:21 hw do ya'll do sysconfig management? Nov 05 17:54:25 puppet or something like that? Nov 05 18:05:23 READMEs everywhere Nov 05 18:05:30 READMEs for READMEs Nov 05 18:07:31 also libdill is cool Nov 05 20:53:03 I am looking for help with I2C communication on the beaglebone black. Nov 05 20:54:25 Address 0x41 and 0x43 do not register with my I2C device, are there any drivers to help with this? Nov 05 20:56:23 what do you mean? Nov 05 20:57:33 your problem description is too unclear. what kind of device are you trying to interface, and what do you mean with "Address 0x41 and 0x43 do not register with my I2C device" Nov 05 21:03:36 I am using the controleverything opticly seperated I2c moduals PCA9536 and they will not show with the 12cdetect Nov 05 21:04:42 these units have the address of 0x41 and 0x43 in the data table Nov 05 21:05:32 the usual checklist: double-check you got the right pins, double-check pinmux, make sure your i2c bus has pull-up resistors of an appropriate value Nov 05 21:06:29 pinmux? Nov 05 21:09:29 pin configuration... no need to worry about that if you're using the i2c-2 bus on P9.19 (scl) + P9.20 (sda), for which pinmux is set up by default Nov 05 21:10:53 I am using 19 and 20. I can detect the other 8 cards in the system but not these 2 cards. Nov 05 21:11:07 also, ensure that your i2c device is *supposed* to be detectable using i2cdetect. In general i2c is not a discoverable bus, so i2cdetect merely guesses per address what sort of transaction to perform to try to detect a device at that address, but there's no guarantee Nov 05 21:11:28 okay, so you actually have a working i2c bus and just don't discover these devices? Nov 05 21:12:05 correct Nov 05 21:12:49 have you tried to simply assume the devices are reachable and interact with them in the way specified in their datasheet? Nov 05 21:13:43 like I just said, i2c is not a discoverable bus and i2cdetect is not a reliable tool. in fact, some devices get so confused by i2cdetect's attempt at probing them that they go entirely unresponsive until they're power-cycled Nov 05 21:15:57 No, but when I used the Pi3 they show up just fine hence the confusion. I will try to work with them as if they are present and see if that works. Nov 05 21:16:09 okay that is weird Nov 05 21:16:53 I would attach an osci to the pca Nov 05 21:17:30 yeah, it's starting to sound implausible that it's a software-issue Nov 05 21:18:39 I have contacted the manufacture and they say the problem is with the BBB. I love finger pointing. Nov 05 21:18:57 I don't see how Nov 05 21:19:20 but regardless, putting a scope on the i2c bus may be informative to figure out what's going on Nov 05 21:20:06 is the PCA9536 powered at 3.3v or 5v ? Nov 05 21:20:45 datasheet says the 9536 is 5v tolerant on the i2c pins btw Nov 05 21:20:48 even cheap scopes nowadays can trigger on an I2C NACK Nov 05 21:20:58 Thanks for the advice. 5v Nov 05 21:21:02 but the BBB is 3.3V I/O Nov 05 21:21:07 Humpelst1lzchen: not useful in this situation Nov 05 21:21:49 more importantly, this device (unlike some other gpio expanders) will not support 3.3v i2c when powered at 5v Nov 05 21:22:31 clemansta: ehh, 5v? Nov 05 21:22:46 clemansta: datasheet says V_IH(min) = 0.7*vcc Nov 05 21:22:52 0.7 * 5v = 3.5v Nov 05 21:23:12 so 3.3v is an indeterminate logic value for this device Nov 05 21:23:22 Sorry the power supplied is 5vdc but the comunications is at 3.3v Nov 05 21:23:33 tested Nov 05 21:23:41 yes, that's not supported by the pca9536 according to its datasheet Nov 05 21:24:31 so whenever that works, you're basically relying on luck Nov 05 21:26:07 and a power supply voltage that's slightly higher or lower, or things like increased bus speed and/or increased rise time could end up making the difference between whether it works or not Nov 05 21:28:36 Okay on the board there is a power feed that can be used at 12 vdc, and a communications input for 5 vdc, both are being used. However the board is geeting its power from the 12 vdc source and the communicaions are from the 5 vdc source. When just the PCA9536 cards are connected the communications is at the 3.3 vdc the power input to the power terminals is 12 vdc and the power to the board I2C terminal for 5vdc is at 5vdc. Nov 05 21:29:37 ehm, okay, and? Nov 05 21:30:45 no communication Nov 05 21:31:13 have you read what I just said? Nov 05 21:31:42 yes Nov 05 21:32:49 the pca9536 is not designed to support 3.3v i2c when it is powered at 5v. you should either power it at 3.3v or pick a different chip Nov 05 21:33:39 okay there is a misunderstanding. the board is powered by 12vdc through the external power bridge. Nov 05 21:34:08 why do you think this is relevant? Nov 05 21:35:13 The communications has 5 vdc on the +5 vdc terminal and is reading 3.3 vdc on the sda/scl terminals Nov 05 21:35:24 which is exactly what it does not support Nov 05 21:36:55 you can check the electrical characteristics section of the pca9536 datasheet yourself if you don't believe me Nov 05 21:38:32 minimum V_IH (HIGH-level input voltage) for SDA/SCL pins is specified to be 0.7 * supply voltage Nov 05 21:39:31 0.7 * 5V is 3.5V, which means this minimum is not satisfied Nov 05 21:41:02 you can try lowering the bus speed, maybe if you're lucky it will work, but you *will* be depending on luck rather than on specifications Nov 05 21:41:28 please refer to the image on the following website: https://store.ncd.io/product/4-channel-i2c-pca9536-optical-isolated-digital-input-with-i2c-interface/ Nov 05 21:43:16 yep, they say it too: "NCD I2C devices communicate 5V I2C data" Nov 05 21:43:27 so you cannot connect that directly to the beaglebone, which uses 3.3V Nov 05 21:43:45 you'll need an i2c level shifter Nov 05 21:44:04 e.g. https://store.ncd.io/product/i2c-shield-for-beaglebone-with-outward-facing-i2c-port/ Nov 05 21:44:28 the unit is inplace already Nov 05 21:48:14 btw, be careful to *not* enable the pull-up resistors on this NCD board (or any other pull-ups to 5V) if the beaglebone is connected since this can damage it Nov 05 21:49:13 None of them are. Nov 05 21:50:44 anyway, there's not much more to say... if you want to use these modules with the beaglebone, you'll need to use an i2c level shifter from the beaglebone's 3.3v to the 5v required by these devices. it's unfortunate if you overlooked this, but their site is pretty clear about it I think Nov 05 21:51:28 same goes for the rpi btw. if it worked without level shifter then that's pure luck Nov 05 21:51:31 Thanks for the information. Nov 05 21:55:11 I fell into this trap on the rpi already Nov 05 21:55:22 hmm? Nov 05 21:55:52 blindly copied some schematics from another project, overlooked that the DS1307+ RTC is spec'd for 5V I/O Nov 05 21:56:01 heh Nov 05 21:56:25 worked sometimes Nov 05 21:56:38 actually that's not true Nov 05 21:56:49 it requires 5v supply, but supports 3v i/o Nov 05 21:57:04 pull up are usually the killer Nov 05 21:57:08 VIH(min) 2.2v Nov 05 21:57:39 btw what's the VOH(min) on the AM335x? Nov 05 21:58:00 check my pins spreadsheet, it has an overview of electrical characteristics too :P Nov 05 21:58:16 its a google sheet right? Nov 05 21:58:38 yeah the problem might have been something else Nov 05 21:58:40 2.85 V @ 6 mA (or 4 mA for some pins) Nov 05 21:58:48 I think it was something with the backup battery Nov 05 21:59:08 oh nice Nov 05 21:59:52 and yeah it's a google sheets doc, https://goo.gl/Jkcg0w Nov 05 22:00:43 need to avoid it... donno which set of cookies is loaded Nov 05 22:00:56 incognito tab? :P Nov 05 22:02:26 just in case: https://cdn.sparkfun.com/tutorialimages/BD-LogicLevelConverter/an97055.pdf Nov 05 22:08:21 using a bidirectional level shifter ic is probably usually more convenient than using discrete fets Nov 05 22:09:01 mosfet level shifters are bidirectional Nov 05 22:10:33 noone is disputing that? Nov 05 22:11:16 yeah sorry Nov 06 00:29:34 wow that kinda sucks... by default timers assert the output-enable of their pin, so if you want to use them in capture mode instead you have to configure the peripheral *before* its pinmux is done (to avoid risking a drive conflict) **** ENDING LOGGING AT Tue Nov 06 03:00:00 2018