**** BEGIN LOGGING AT Sat Dec 16 03:00:02 2017 Dec 16 15:04:31 hey guys i want to read xy2 100 messages to save x and y coordinates in a database Dec 16 15:04:53 and now im searching for hardware which is capable to do that Dec 16 15:05:05 im a kind of newbie Dec 16 15:05:22 so i thought that maybe someone of u can help me Dec 16 15:06:06 these xy2-100 messages are the standard protocol for controlling laser machines Dec 16 15:12:40 the beaglebone has a fair variety of interfaces, and almost any custom protocol can be implemented using the PRU subsystem Dec 16 15:12:53 without more information on this protocol it's hard to say more Dec 16 15:22:52 Someone here says it's very much like SPI: https://www.raspberrypi.org/forums/viewtopic.php?t=98908 Dec 16 15:23:13 hm, ok, you just want to capture this data? I'd probably use mcasp for it myself (it might accept 1 slot/frame even though the docs say no, and otherwise burst mode should work), but using PRU is also a fine option Dec 16 15:23:21 no, it's not spi-compatible Dec 16 15:24:48 you might be able to get away with using both spi controllers in slave mode (one for x, one for y) if you don't mind losing the parity bit, but it feels like an awkward approach Dec 16 15:26:16 its this protocol Dec 16 15:26:17 http://www.newson.be/doc.php?id=XY2-100 Dec 16 15:26:26 mcasp supports a pulse-based frame sync signal, and natively has the ability to capture multiple data lines (up to 4) Dec 16 15:26:44 what is mcasp? Dec 16 15:26:57 multichannel audio serial port Dec 16 15:27:20 this isn't audio data, but obviously it doesn't actually care :) Dec 16 15:28:02 ok but the problem is that the clock of the xy2 100 protocol has a frequency of 2mhz Dec 16 15:28:08 it supports up to 50 MHz Dec 16 15:28:15 (bit clock) Dec 16 15:29:24 ok that sounds interesting Dec 16 15:29:50 the main compatibility issue is that each "frame" (the period delineated by sync pulses) only contains a single "slot" (data value), while audio normally always uses 2 or more slots Dec 16 15:30:58 ok Dec 16 15:31:05 but it could simply work when number of slots is configured to 1 ... I see no plausible reason for the restriction of having 2 or more slots when pulse-sync is used (as it is in this case) Dec 16 15:33:08 is there no possibilty to plugin a normal dsub 25 cable and parse the bitstream so that i can pack the messages in structures in save them in a database? Dec 16 15:33:34 im coming more from the software side instead of the hartware side :/ Dec 16 15:33:39 hardware Dec 16 15:34:18 you're trying to connect to an obscure hardware interface. I'm afraid dealing with the hardware side is pretty unavoidable Dec 16 15:34:59 although I've actually only discussed the software aspect above... proper electrical interfacing is a separate issue that you'd need to deal with Dec 16 15:36:46 the pdf you linked doesn't mention the voltage used, but I've seen 5V mentioned while googling, which means you'd need a level shifter to convert it to the 3.3v required by the beaglebone Dec 16 15:37:34 I would suggest this: There appear to be commercial products that already do what you're trying to do. Presumably, they were designed by people with some amount of expertise in the field, and presumably, their pricing reflects the amount of work it took to implement them. Considering your level of expertise, ask yourself what your time is worth, and how long it's likely to take you to achieve the same goal. Dec 16 15:37:36 hmm ok Dec 16 15:37:40 Dec 16 15:37:56 so can u maybe sum up what i need? Dec 16 15:38:31 what 'myself' just said is pretty decent advice Dec 16 15:39:13 if you want to go for it yourself (i.e. if you have a lot of time and little money, or just want the experience), expect to have to do a lot of learning Dec 16 15:39:15 I always used to get caught on the "Always build EVERYTHING" side of the build-vs-buy decision, and I eventually came upon a better way to make the decision: I now ask myself "If I build this myself, I'll acquire certain skills along the way. Are those skills I want to have? Would I pay for a class to learn them?", and if the answer is yes, if I value having those skills, then yes, I'll build. Otherwise I'l Dec 16 15:39:21 l buy. Dec 16 15:40:09 so, for instance, I'll pay someone to tile my bathroom, because I don't value that skill enough to spend messy weeks learning it myself. But I won't pay someone to set up a satellite receiver for me, because that's interesting and I want to learn it anyway. Dec 16 15:42:02 xy2: the beaglebone can definitely do this, with probably a level shifter ic needed but with some luck no other external electronics. Dec 16 15:42:20 do you think i can learn it in 2 months? im a really determined person Dec 16 15:42:27 So if you want to learn the hardware anyway, yeah, build this XY2 thing! But you're gonna need some tools you don't have (a decent oscilloscope, for one) to make sure your level-shifter is doing sane things and the signal is making it to the 'bone intact. And you're gonna need to learn a fair bit about electronics to troubleshoot that part, make sure the powering and grounding and shielding aspects are all Dec 16 15:42:33 sane. And then there's the software side.. Dec 16 15:43:04 I mean, there's a chance that the first level-shifter you try will Just Work and that would require no further consideration. But noobs are rarely so lucky. Dec 16 15:43:16 the software side is probably not hugely difficult, but I'm saying that because I already have a deep understanding of the AM335x (the processor on the beaglebone) and I've dealt with mcasp plenty enough Dec 16 15:43:58 going the mcasp route (which is one of two options, the other being PRU) would probably require patching the mcasp kernel driver, although it would be a small patch Dec 16 15:44:28 I'm a hardware guy and I consider the software to be deep scary voodoo, but then I also mentor noobs all day in both my work and hobby lives, and I've seen the mistakes people make on the simplest things like grounding. Dec 16 15:45:01 xy2: I have no way of estimating how long it might take you to learn, especially since I have no idea what your current skill level is Dec 16 15:45:17 but even if I did, I still wouldn't feel comfortable about guessing such things Dec 16 15:45:28 hmm ok Dec 16 15:45:45 can i maybe have ur email adresses for further questions? Dec 16 15:46:02 and my consulting rate-card ;) Dec 16 15:46:07 thanks for ur help guys you're really helpful Dec 16 15:46:14 even "whatever you *think* and then go two orders of magnitude up" might not cut it ;-) Dec 16 15:46:15 :) Dec 16 15:46:55 xy2: no, I'm not your personal tutor Dec 16 15:47:35 but asking smart questions ( http://www.catb.org/esr/faqs/smart-questions.html ) here and sticking around for an answer can often be fruitful Dec 16 15:48:06 (do be patient, people are in different timezones and may simply be busy, so it can easily take hours to get an answer. just leave your irc client connected) Dec 16 15:48:42 For me, put it this way: If you're gonna release the final result as OSHW with full docs and schematics and patches and a writeup, so anyone can make their own, and bring this crappy protocol within reach of more experimenters, then I'll try to contribute a reasonable amount of help. But if you're not gonna share back the results of what you learn, then my consulting rates are very expensive. Dec 16 15:49:31 realizing the difference between asking questions here, where many others might contribute to answers, and private chat/mail/etc is a very important distinction Dec 16 15:49:34 i have no problem to publish my results Dec 16 15:50:13 yes but i think its not bad to have a contact person Dec 16 15:51:51 https://lasermarkingsystems.eu/e1701.php Dec 16 15:52:13 i think this controller looks like a beagle board setup? Dec 16 15:52:23 do u also agree? Dec 16 15:53:08 that's definitely a beaglebone Dec 16 16:01:48 Personally if I were taking this on, the mcasp would be the hard part, and I'd start by replicating some other well-documented mcasp-based project, preferably one that also uses it in a non-standard (which is to say, non-audio) way, to see what hoops they had to jump through to make it work, and prove that I can jump through them too. Having a known-working implementation of something sorta-close, as a sani Dec 16 16:01:54 ty-check, is something I find invaluable as an experimetner. Dec 16 16:02:31 myself: actually I think you can probably use the standard driver with very little modification, if you don't mind receiving the captured data via alsa Dec 16 16:03:01 which is also something I've never done, so simply working through an alsa hello-world would be important to me :) Dec 16 16:03:03 but yeah, even then, getting that driver to do what you want is far from trivial... I speak from experience Dec 16 16:03:41 so I wouldn't be surprised if for some people going the PRU route might be simpler Dec 16 16:04:14 although then you need to worry about offloading a stream of data from PRUSS to the linux side Dec 16 16:05:16 Once I'd proven that I could make the software do the thing, I'd sit down with a few different level-shifting ICs, an oscilloscope, and XY2 signal source of some sort. I'd make sure I was getting a nice square signal into the bone, then spend a little while packaging the hardware side so it's physically robust and unlikely to fall apart during my software work. Dec 16 16:08:20 Depending on what sort of connector the source machine uses, that might require nothing more than a protocape, which I would recommend everyone have a handful of anyway. Dec 16 16:52:19 hello all Dec 16 16:52:39 I wanted to know if the robotics cape library uses the PRUs Dec 16 16:53:20 hello all Dec 16 16:53:30 is there any easy way to use the PRUs on the beagle bone blue Dec 16 16:53:46 does the robotics cape library use them? Dec 16 17:10:51 I wonder if we could charge a tiny fraction of a bitcoin to join chat, refundable after 1 hour of waiting for your answer... :P Dec 16 17:19:09 do you mean me? Dec 16 17:20:16 nah, confused :) Dec 16 17:20:39 drive-by questions are the bane of IRC Dec 16 17:21:38 yes now im a little bit confues :/ XD Dec 16 17:21:45 confused Dec 16 17:28:26 myself: that means you'd incur at least 80€ or so of transaction costs (iirc around 40 per transaction at the moment) Dec 16 17:28:33 * tbr approves Dec 16 18:05:47 does anyone have any tutorial on the bbblue Dec 16 18:05:56 for using the PRUs to generate PWM Dec 16 18:06:56 and how to create programs that use the functions of reading imus Dec 16 18:07:11 any AM335x tutorial for PRU + PWM will apply Dec 16 18:07:42 so also any beaglebone Dec 16 18:08:04 which tutorial would that be Dec 16 18:08:12 no clue Dec 16 18:08:16 also Dec 16 18:08:22 in the robotics cape lib Dec 16 18:08:32 does the ppm function utilise the PRU Dec 16 18:08:35 pwm* Dec 16 18:11:59 any ideas how to use the library in programs Dec 16 18:30:35 I need something very very simple and like making it in my own way so I am making my own code. I don't want to get through all the trouble of understanding the whole ardupilot code. I will be using the roboticscape library for reading the IMU and I could be using it to generate the pulses for the ESCs at 400Hz but I would prefer if they are automatically generated at a given frequency in the PRUs, e.g. 400 Hz.. Dec 16 18:30:50 Could someone tell me what would I need to do to simply get the RC signals from the PRUs and send ESC (PPM or PWM) signals to the PRUs, without the trouble of the all the AP_HAL libraries and all the ardupilot code. Dec 16 18:34:04 Without all the trouble of using the code that already does that? Okay hotshot, if it's that simple why are you asking? Dec 16 18:35:30 ....he'll be back. Dec 16 18:38:44 lol, that nickname Dec 16 18:41:31 I mean yeah we all need a little help sometimes, but this isn't ##psychiatry :) Dec 16 18:43:34 lol Dec 16 20:12:56 hello all Dec 16 20:13:09 what serial port does the microUSB use Dec 16 20:15:58 what serial port does the microUSB use Dec 16 20:19:06 I'm new to beaglebone, which tool chain do you recommend to install? Thanks, **** ENDING LOGGING AT Sun Dec 17 03:00:00 2017