**** BEGIN LOGGING AT Sat Apr 24 03:00:29 2021
Apr 24 03:29:35 <mattb0000ne> anyone have a quick test python program for steppers
Apr 24 04:32:28 <GenTooMan> trying to get a step in the right direction?
Apr 24 04:38:04 <mattb0000ne> just step at all
Apr 24 04:38:27 <mattb0000ne> i have not had much experience driving pins from python
Apr 24 05:17:04 <zmatt> mattb0000ne: https://pastebin.com/1UCnbVNT
Apr 24 05:17:21 <zmatt> these are optimized for simplicity, not for performance
Apr 24 05:18:16 <zmatt> oh, and they don't include any way to change direction (input <-> output) since that wasn't required for the person for whom I wrote this example (who use an overlay to setup his gpios)
Apr 24 05:18:36 <mattb0000ne> ok I can do that too
Apr 24 05:18:41 <mattb0000ne> just set to output
Apr 24 05:19:02 <zmatt> it also wouldn't be hard to add that functionality
Apr 24 05:40:31 <mattb0000ne> how can I check a level shifter
Apr 24 05:40:45 <mattb0000ne> would touching for continuity between high and low make sense
Apr 24 05:40:51 <mattb0000ne> been looking at this for about 2hours
Apr 24 05:41:05 <mattb0000ne> I have good signal coming from the encoder up to the level shifter
Apr 24 05:42:54 <zmatt> how you test a level shifter: send a test signal to the input (i.e. either 0V or whatever the high voltage is on that side), measure the output
Apr 24 05:43:09 <zmatt> no form of continuity testing makes sense
Apr 24 05:43:24 <mattb0000ne> ok
Apr 24 05:43:29 <zmatt> (for pretty much anything other than wires)
Apr 24 06:30:12 <mattb0000ne> are logic shifters only rated to take 5V on the high side
Apr 24 06:30:45 <zmatt> logic shifters are rated to whatever the logic shifter is rated
Apr 24 06:37:22 <mattb0000ne> not sure what the hell the thing did but it took out my board
Apr 24 07:09:24 <mattb0000ne> so
Apr 24 07:28:40 <zmatt> in case it's of interest, simple python gpio library with support for changing direction: https://pastebin.com/cVHktYPn
Apr 24 07:29:43 <zmatt> wouldn't be too hard to convert to an object-oriented variant (which would probably also be step 1 towards a version with better performance, so that state and open file descriptors can be cached in that object)
Apr 24 07:32:47 <mattb0000ne> when you say improved perofrmance
Apr 24 07:32:50 <mattb0000ne> what are we talking
Apr 24 07:33:19 <mattb0000ne> may not be relevant for my purposes
Apr 24 07:33:35 <mattb0000ne> sounds more like under the hood optimization
Apr 24 07:33:45 <mattb0000ne> question for you before you depart
Apr 24 07:33:58 <mattb0000ne> the two sids of the logic shifter should be isolated right
Apr 24 07:34:54 <mattb0000ne> IF I only have power on the high side I should not detect a voltage on my low side with my multimeter
Apr 24 07:35:06 <mattb0000ne> I am not sure what the hell happened
Apr 24 07:35:51 <zmatt> I have no idea what you mean by that
Apr 24 07:36:12 <mattb0000ne> I had power coming out of my LV side
Apr 24 07:36:25 <mattb0000ne> I think that is what killed the board
Apr 24 07:36:48 <mattb0000ne> i was doing like you said plugged in my ground and I was going to screw in the LV wire to power the low side
Apr 24 07:36:55 <mattb0000ne> got a spark then poof
Apr 24 07:37:11 <mattb0000ne> i am trying to figure out why
Apr 24 07:37:14 <mattb0000ne> this is for my encoder
Apr 24 07:37:20 <mattb0000ne> and I have to step it down
Apr 24 07:37:37 <zmatt> a _spark_ ?  surely you must have wired something wrong somehow
Apr 24 07:37:47 <mattb0000ne> its just a logic shifter
Apr 24 07:37:50 <mattb0000ne> nothing to wire
Apr 24 07:38:04 <mattb0000ne> GND and HV
Apr 24 07:38:08 <zmatt> like, that is kinda only possible if you connected the LV power supply to the HV power supply somehow
Apr 24 07:38:23 <mattb0000ne> and my two wire signals
Apr 24 07:38:53 <zmatt> what voltage did you put on the HV side?
Apr 24 07:39:44 <mattb0000ne> 5V
Apr 24 07:39:47 <zmatt> ok
Apr 24 07:39:54 <mattb0000ne> but my encoder has 24V on it
Apr 24 07:40:02 <zmatt> uhh
Apr 24 07:40:08 <zmatt> on its outputs?
Apr 24 07:40:14 <mattb0000ne> input
Apr 24 07:40:34 <mattb0000ne> actually has 5-30 in/out
Apr 24 07:40:42 <mattb0000ne> so I guess it is comign in way over
Apr 24 07:40:46 <mattb0000ne> 5V
Apr 24 07:41:24 <mattb0000ne> that is probably it
Apr 24 07:42:07 <mattb0000ne> i was so proud of myself systematically checked everything with the oscilloscope
Apr 24 07:42:42 <mattb0000ne> everything looked good checked the lines coming out of the shifter too. I was just plugging into the beagle
Apr 24 07:42:48 <zmatt> so you have an encoder that supports being powered by 5-30V, you powered it with 24V, and then stuck that 24V signal into a level shifter whose high side you powered at 5V ?   I'm baffled by what you were even thinking
Apr 24 07:43:01 <zmatt> uhh
Apr 24 07:43:31 <zmatt> how did you check the lines coming out of the shifter if you hadn't wired up the LV supply yet?
Apr 24 07:43:41 <mattb0000ne> well I thought the shifter would max at whatever the HV was which was5V
Apr 24 07:44:09 <mattb0000ne> so my encoder needs everything powered at 5V
Apr 24 07:44:22 <mattb0000ne> so i do not fry everything
Apr 24 07:44:53 <zmatt> or a level shifter that tolerates a 24V high side
Apr 24 07:46:02 <zmatt> if a level shifter has a separate supply for each side (which is typical) then that needs to be the I/O supply of that side, and will be the output level on that side (for outputs) and the max tolerated input level on that side (for inputs)
Apr 24 07:46:53 <mattb0000ne> i see
Apr 24 07:46:59 <zmatt> unidirectional level shifters might just have a supply for the output-side, in which case the datasheet should be consulted for the electrical specs on the input side
Apr 24 07:47:13 <mattb0000ne> yeah this was bi directional
Apr 24 07:48:09 <zmatt> some bidirectional level shifters have a single supply, which should then be the LV side
Apr 24 07:49:53 <zmatt> (and require pull-ups on both sides to their respective power supplies)
Apr 24 07:51:10 <zmatt> didn't you have the encoder working in the past already?
Apr 24 07:52:19 <mattb0000ne> i did
Apr 24 07:52:35 <mattb0000ne> and I was not getting a signal so I was checkign the wiring
Apr 24 15:43:20 <vaibhav> need help to buy Beagle bone black for linux device driver study personalise
Apr 24 16:10:36 <vaibhav> high
Apr 24 16:16:54 <buzzmarshall> give it a bit of time as i am sure one of the guys here will see you and respond
Apr 24 16:18:28 <buzzmarshall> i use a few bbb's myself but probably am not the one to answer your questions as i kinda do my own thing and am off the beaten track for whats official per here
Apr 24 16:24:37 <demostanis[m]> can a BBB be suited for desktop usage (
Apr 24 16:25:08 <demostanis[m]> * can a BBB be suited for desktop usage (i.e. web browsing, coding)?
Apr 24 16:25:28 <demostanis[m]> which are the best resellers?
Apr 24 16:25:46 <demostanis[m]> and does it require proprietary firmware?
Apr 24 16:30:49 <buzzmarshall> grabbed mine from Mouser just because i get a fair bit of other stuff from them but i quess it kinda depends on where you are
Apr 24 16:30:52 <buzzmarshall> im in canada
Apr 24 16:31:38 <buzzmarshall> as far as propriatary TI's pretty good with info so everything you need is easily found
Apr 24 16:32:05 <buzzmarshall> as far as desktop i quess it kinda comes down to what exactly you want to do
Apr 24 16:32:21 <buzzmarshall> the bbb is a bit light for gui based stuff tho it can be done
Apr 24 16:32:56 <zmatt> demostanis[m]: I wouldn't want to run a modern web browser on a BBB
Apr 24 16:33:24 <buzzmarshall> heres the man that can answer you questions.  :)
Apr 24 16:33:58 <zmatt> demostanis[m]: the SoC on the BBB is designed for industrial embedded systems, not desktop computers.   it has some minimal graphics capability mainly to support simple touchscreen interfaces (e.g. it doesn't even support a hardware mouse cursor)
Apr 24 16:35:25 <zmatt> also it has 512MB of (pretty slow) ram, which is more than plenty for embedded systems, but rather tight for modern desktop environments
Apr 24 16:36:15 <zmatt> especially web browsers, which nowadays can easily consume hundreds of megabytes of ram
Apr 24 16:39:28 <zmatt> so like, could it maybe be used for the purpose?  in theory perhaps, but I wouldn't want to work on it.   the specs that one might care about for a desktop computer are all weak points of the BBB (ram size/performance, cpu performance, storage options, graphics, networking), while using it a desktop pc will use none of the strengths of the BBB
Apr 24 16:39:54 <zmatt> so it's just entirely the wrong fit for the application
Apr 24 16:40:53 <demostanis[m]> too bad. i read somewhere that BBB revision C can have up to 4GB of RAM?
Apr 24 16:41:25 <zmatt> no, 4GB of eMMC
Apr 24 16:41:37 <demostanis[m]> <buzzmarshall "grabbed mine from Mouser just be"> their website says that if I buy it today, it will be delivered in november
Apr 24 16:41:56 <demostanis[m]> <zmatt "no, 4GB of eMMC"> = SD card?
Apr 24 16:42:16 <zmatt> sort of but integrated on board (and faster than the SD card slot)
Apr 24 16:42:34 <zmatt> (but still not fast)
Apr 24 16:43:31 <zmatt> there is a beaglebone variant with 1GB of ram and gigabit ethernet (beaglebone enhanced), though it still has only 4GB of eMMC
Apr 24 16:44:13 <zmatt> and it's still not magically going to make it a good desktop pc
Apr 24 16:44:35 <demostanis[m]> well, definitely not very suitable for web browsing, that'd fill up the entire storage...
Apr 24 16:44:35 <demostanis[m]> what about the coming soon beagle V?
Apr 24 16:45:17 <zmatt> dunno, some chinese SoC with no documentation yet
Apr 24 16:46:49 <buzzmarshall> jeese thats a long time...
Apr 24 16:46:52 <zmatt> I doubt you're going to see spectacular performance there, I think its main purpose is putting RISC-V into the hands of people who want it :P
Apr 24 16:51:10 <buzzmarshall> mouser here in canada shows over 5k bbb's in stock for delivery
Apr 24 16:51:40 <buzzmarshall> anyways as zmatt says i am not sure desktop is the best thing for the bbb's
Apr 24 16:53:37 <buzzmarshall> even the ai version i have with its larger memory isn't something i would use for desktop
Apr 24 16:54:14 <buzzmarshall> personally i'd use something more like a Rockchip pro or one of the other many sbc's that are out there
Apr 24 16:54:55 <buzzmarshall> the issue tho with pretty much of them will be the proprietary gpu stuff
Apr 24 16:55:55 <buzzmarshall> rockchip, amlogic, allwinner, broadcom there all kinda the same when it comes to propriatary hardware acceleration on the gpu's as well as some of the media codecs
Apr 24 16:56:19 <buzzmarshall> open source is getting better on some of them but still struggles in some areas
Apr 24 16:57:14 <zmatt> I mean, isn't that pretty much always the case with GPUs?  the ones on TI SoCs also need closed-source firmware and closed-source userspace libraries....  it's mostly irrelevant though since it is entirely optional to use and most people don't (in fact they're a pain in the ass to get them installed and working if you *do* want them)
Apr 24 16:57:50 <buzzmarshall> as for the newer RISC-V goes, until there is enough of them floating around out there its not something i would bother with at this stage
Apr 24 16:58:08 <buzzmarshall> ya your right there pretty much all in the same boat
Apr 24 16:58:51 <buzzmarshall> if your just after desktop without a lot of media playing a few of them are in better shape desktop wise then others when it comes to linux
Apr 24 17:00:11 <buzzmarshall> most these days are trying to play media which is really where the biggest issues are routed in as projects like ffmpeg and v4l2 still struggle with the newer codecs
Apr 24 17:00:51 <buzzmarshall> allwinners and rasp4's are probably the furthest ahead in those areas as stateless is ahead of stateful codec wise
Apr 24 18:34:10 <GenTooMan> zmatt one of the most frustrating things about many of the GPU of many of TI's parts. Seems to make the GPU useless for the majority of applications.
Apr 24 21:52:14 <zmatt> GenTooMan: I'm particularly annoyed by the fact that ImgTec got OpenCL working to some degree on the SGX530 (https://university.imgtec.com/fun-with-beagle-gpu/) but downloading them requires registration, approval, and the materials can only be used for "self-study and company training purposes, as long as these are not-for-profit activities"
Apr 24 21:53:05 <zmatt> blog post about it: https://www.imaginationtech.com/blog/fun-with-powervr-and-the-beaglebone-black-low-cost-development-made-easy/
Apr 25 00:05:47 <GenTooMan> zmatt, I am uncertain why they did it, however often in commercial entities there are a large number of conflictingly stupid things going on.  Mentor Graphics had(has?) a package called PADS it was quite messed up circa 2015, in 2017 the "subscribers" found out what had been going on is the person in charge of development didn't like it so that person basically purposefully made sure it couldn't be fixed. Seems that suddenly
Apr 25 00:05:47 <GenTooMan> they had "fixes" to the numerous issues we had suffered from prior to that. It was too late though the company began using Ultium (which has it's own messed up crap but was usable).
Apr 25 01:39:23 <mattb00000ne> we are back with board number 5!!!!
Apr 25 01:39:29 <mattb00000ne> i think this will be the last one boys
Apr 25 01:39:42 <mattb00000ne> no more $80 dollar lessons learned
Apr 25 01:39:43 <mattb00000ne> lol
**** ENDING LOGGING AT Sun Apr 25 02:59:56 2021