Back
[00:00:34] <anonnumberanon> >z64555 thinks about buck mode power
[00:00:45] <anonnumberanon> nothing wrong with that
[00:01:10] <z64555> well the way you say it makes it sound like I put a buck deer on a tradmill
[00:01:14] <z64555> :P
[00:03:54] <anonnumberanon> buck mode powaa
[00:04:32] <Snert> powaa - I like that :)
[00:04:36] <z64555> Yeah, was looking for something to replace my stupid 5V and 3.3V voltage regulators
[00:05:07] <anonnumberanon> you can either buy for cheap or make your own for fun
[00:05:48] <anonnumberanon> I have a rather small one that will give 5V (or whatever i want) and 12A.
[00:05:54] <anonnumberanon> For my biped robot.
[00:08:46] <z64555> I saw a TI chip that said it was for a power converter, but I didn't read into its specs enough to see what other circuitry it needed to work.
[00:08:55] * z64555 is teh lazy
[00:13:23] <anonnumberanon> iirc you need something that switches a FET fast (adjustably so), capacitors, an inductor. If it was me I'd use an avr for the switching. 555 can be used i think
[00:21:15] <z64555> Yeah, that's the basic idea. capacitor/inductor is the energy storage element that smooths the pulse chain down to a usable sinewave
[00:21:33] <z64555> I saw something about parallel FET's being better
[00:22:33] <z64555> better/smoother
[00:26:18] <anonnumberanon> oh but if there is an avr in the project it needs to receive a regulated voltage lmao
[00:26:32] <anonnumberanon> so no
[00:26:49] <z64555> yep
[00:27:03] <anonnumberanon> needs something that can take in voltage from 0 to like 30 then
[00:27:20] <anonnumberanon> Can 555 take 25V?
[00:27:25] <z64555> Nope
[00:27:39] <z64555> Well. it can for a millisecond
[00:28:56] <z64555> think I'll put that off as a side project for now. the voltage regulars "work"
[00:29:14] <z64555> might need a better heat sink for them, but they work
[00:30:02] <anonnumberanon> i use lm337 and 8805 or whatever for two things on my favorite project
[00:30:16] <anonnumberanon> then hook up a recuperated battery to the entire thing
[00:30:58] <z64555> I need to get the AirDragon up and flying. I was looking at R/C transmitters and possibly also a FPV camera system
[00:31:04] <anonnumberanon> 5v and 3.7
[00:31:23] <z64555> the FPV transmitters all seem to use TV frequencies
[00:31:43] <anonnumberanon> is that your project at the moment?
[00:32:07] <z64555> yes
[00:32:22] <anonnumberanon> is that a multirotor?
[00:32:26] <z64555> yes
[00:32:35] <anonnumberanon> good deal
[00:32:45] <anonnumberanon> what microcontroller for it?
[00:33:31] <z64555> I've got two of them, one of them's a HC9S12 from motorola/freescale and the other is a TI chip, the newer ARM one, not the MSP430
[00:34:38] <anonnumberanon> ah, some kind of Cortex M, ARM?
[00:35:50] <anonnumberanon> I have a plan to learn how to use one of them in the short to medium term.
[00:36:45] <z64555> LM4F120
[00:36:48] <z64555> M4
[00:37:18] <z64555> It has an FPU. :D
[00:37:23] <anonnumberanon> ah yeah, that's a good choice
[00:37:47] <z64555> Oh yeah, especially since I got it for $5 as a limited time special
[00:38:05] <anonnumberanon> nice, is that a dev board? with usb?
[00:38:13] <z64555> yup
[00:38:35] <z64555> They call it a launchpad, a competitor to the arduino, I guess
[00:39:01] <anonnumberanon> yeah it lets you just write code and get into it fast
[00:39:18] <z64555> It's got male/female headers on it, so you can stack the same board or special shields on them
[00:39:46] <z64555> Makes mounting to a robot interesting, though
[00:40:52] <anonnumberanon> oh that's weird, and cool i guess
[00:40:58] <anonnumberanon> but yeah those things are powerful
[00:41:50] <anonnumberanon> I'd like to see if they are more useful than arduino for implementing a clock that can measure a microsecond very well. (im gonna be using an stm32f407 dev board)
[00:42:22] <anonnumberanon> AND BY arduino i just mean an 8-bit micro at 16 Mhz.
[00:42:51] <z64555> right, the ATwhatevers
[00:43:04] * z64555 is very specific
[00:43:39] <anonnumberanon> Atmega328P
[00:43:57] <anonnumberanon> is what I've been using
[00:45:34] <z64555> hm, I guess I could work on the AHRS with the IMU that I have, vs. buying a new one
[00:46:02] <anonnumberanon> but yeah the way RC receivers (actually the receiving part with the little antenna and 5 or 6 slots to plug in servos and whatever you want) work is that you can connect their pinouts to your micro and you have to be able to read these signals coming in, then from those signals you increase your throttles on the 4 motors and do whatever.
[00:46:22] <z64555> yep
[00:46:45] <z64555> Already have their waveform down, since it's the same waveform I have to send to the ESC's
[00:46:49] <z64555> *type of waveform
[00:47:09] <z64555> could possibly use edge triggers
[00:47:51] <anonnumberanon> it's PWM so when it goes high start measuring how long it stays high, then that's your value
[00:48:28] <anonnumberanon> send it back to the ESCs in a "controller manner" ;) and that's your motor control part
[00:48:42] <anonnumberanon> I meant, "controlled manner".
[00:49:10] <anonnumberanon> As in, it has to be real-time, or deterministic.
[00:49:30] <z64555> yeah, R/C is "easy"
[00:50:31] <z64555> input is desired attitude (pitch and roll angle) and the controller compares it against it
[00:51:03] <z64555> Alternatively could go with rates, but that wouldn't self-centering and more... classical
[00:51:16] <z64555> *and instead be more
[00:52:22] <anonnumberanon> right now im trying to brainstorm on 2.4ghz radio links between microcontrollers
[00:52:50] <anonnumberanon> I want one that is cheap and good and that I can put a library for on GitHub.
[00:52:53] <anonnumberanon> maybe..
[00:53:26] <anonnumberanon> I just don't know how to write libraries and stuff yet.
[00:53:50] <anonnumberanon> also gotta solve the synchronous use of that one i was playing with
[00:53:59] <z64555> libraries are just C programs
[00:54:04] <z64555> .dll's that is
[00:55:10] <z64555> I think you can use C++ too with the newer compilers
[00:55:48] <anonnumberanon> maybe Ada or C. not too keen on writing C++
[00:56:52] <anonnumberanon> https://www.sparkfun.com/datasheets/Components/SMD/nRF24L01Pluss_Preliminary_Product_Specification_v1_0.pdf
[00:57:14] <z64555> C++ is flexible, although most of the times college classes say "You must use classes for everything!"
[01:00:32] <anonnumberanon> yeah, for computer programs it's great
[01:02:14] <z64555> C has a tighter control on bare-metal applications, though
[01:03:05] <z64555> although, I really really like to be able to do matrix calculations with just one call
[01:03:28] <z64555> in a format that is human readable
[01:04:55] <anonnumberanon> heh
[01:04:56] <anonnumberanon> CS
[01:05:13] <anonnumberanon> I'm implementing a binary tree as a weekend project.
[01:05:59] <anonnumberanon> but so lazy and tired
[01:07:20] <z64555> nice. what did you use as the underlying container. vector? array?
[01:08:26] <anonnumberanon> i don't have something that works yet but just structs (called records in Ada)
[01:09:00] <anonnumberanon> each node struct has a value and two pointers to the node type, so that it can point to the left node and the right child node
[01:11:41] <z64555> sounds about right
[01:15:14] <z64555> Yeah, I will need to test my AHRS software. I've got it written for the main program, shouldn't be too difficult to make a test project for it
[01:17:26] <anonnumberanon> That's the hardest part of a multicopter project imo.
[01:22:03] <z64555> it is where the bulk of the calculations are. PID and even fuzzy logic is a cakewalk compared to the calculations need to fuse the sensors together
[01:22:45] <z64555> I mean you can get away with a binary filter if you were doing a rate controlled scheme
[01:29:41] <anonnumberanon> well you have gyro first wich will make the quad fly and then you can make it better with the output of the accelerometer
[01:30:03] <anonnumberanon> they say to test it with gyro first and when it's satisfactory, to add acceleromter
[01:30:28] <z64555> yeah, that's all rate compensation
[01:31:11] <z64555> or rather,... yeah.
[01:31:48] <z64555> Gryo does the bulk of the work, the accelerometer is there to get a vector of the acceleration. Which is usually gravity but also accounts for wind
[01:32:01] <z64555> *Gyro.
[01:32:07] <z64555> Spellings. they are on the decline
[01:32:24] <anonnumberanon> my right eye is bloodshot...
[01:33:15] <z64555> probably a bit dry :/
[06:28:52] <theBear> After suffering weak gain at the poles, the National Transistor Party has been trying to energize their base ! that's right, i found a whole list of terrible technical mostly electronics/physics related, AND i will share the joy with you, and can i just congratulate on requesting the link <grin>
http://www.powerstream.com/powerhumor.htm
[06:30:54] <theBear> we got higgs partices attending church service, benchtop blowup/magic smoke jokes, we even got some of that our smart dept vs your stupid dept.for those of you into the corporate level humor, cos it makes you able to relate more easily by doing the things you do every day
[06:31:28] <theBear> back to the other room *poofsky5000effing-tiny-cloud-of-smoke"
[06:35:52] <theBear> hmm final serious note <clears throat and adjusts face as close as it can get to serious> there has been more than light rain at least a few hours already thisafternoon/evening, and while i'm too lazy and effed in the rear section to look behind the curtain way up the backside of this very desk, at least twice the distance it is to the monitors from here, yet i suspect that i may have fixed or at least somewhat improved the situation where every
[06:35:52] <theBear> night it rains and every night shortly after i got a rcd that won't turn on for maybe 6 or 10 hours, and doesn't seem to care if all the sockets/loads connected to itare disconnected before then
[06:37:52] <Jak_o_Shadows> has rained a bit recently
[06:38:39] <theBear> i even considered things like metal window frames being connected to some chassis/plug somehow being referenced to the mains voltage, cos 30mA ain't much to steal at 240, same for fone/modem, which woulda been very believable if telcom specs don't insist on isolated modems, and a pit with copper joins got a bit floody, like everywhere outside, not deep, but just so much water that any ground outdoors was covered and shiny
[06:38:49] <theBear> lawn starts to look like lake
[06:39:00] <theBear> roads/carparks too
[06:41:55] <theBear> but i think i migth got it, so ya know, check me for my next news update hopefully not before maybe 8 hours, cos it might be quick to findout i didn't get it, buuuuut bit longer the other way round, cos unlike naturewe can be held accountable for our own work, even if it's only to ourselves (and there ain't no escaping your own shame at you... egotistical nightmare !! 1hehe) later tater
[06:43:07] <theBear> and at 10 lisa will be here to tell you how weather is looking the next days... now here's ted with an interesting quiestion about leds !~
[06:48:02] <theBear> omfg ! where this receipt says i can just txt the word recharge+12digitcode to 1511 to use a credit/recharge instead of those endless slow talkey lady robot that don't let you skip much even if you know what you gonna dial
[06:48:52] <theBear> that IS new right, i'm not some old dude that just ignores all those handy hints cos he doesn't understand them :)
[11:11:45] <deshipu> theBear: can you explain the one about "sycophantic resistor"?
[11:14:45] <theBear> awww that silver tongue quick as a whip it is
[11:26:01] <z64555> "What's the perfect gift for your favorite electrical engineer?"
[11:26:04] <z64555> "Shorts."
[11:26:18] <z64555> awesome.
[11:27:02] <z64555> deshipu: 110%
[11:27:20] <deshipu> z64555: hmm?
[11:27:27] <z64555> There's an old saying in business to give it your 110%
[11:27:40] <z64555> a little bit more than your "all"
[11:28:18] <deshipu> someone thought that ohms are like percents?
[11:28:58] <z64555> nah, it's the number, although I've seen the % sign used on some resistors
[11:29:24] <z64555> which was representing the tolerance
[11:31:27] <deshipu> right, but 110% would be pretty bad
[11:32:06] <z64555> no no. not 110% tolerance
[11:36:32] <z64555> "I took my car to my mechanic for a checkup. He told me that my battery needs a new car."
[13:25:14] <rue_shop3> haha
[13:25:21] <rue_shop3> should have maintined the paint
[14:52:55] <theBear> i always liked the positive-response to something "... AND a half"
[14:53:55] <rue_shop3> theBear, do you get out of the house much latley?
[14:54:16] <rue_shop3> aside from work, I really dont go anyhwere
[14:54:18] <rue_shop3> never have tho
[14:55:35] <rue_shop3> this serial adc project is not fitting togethor
[14:56:01] <theBear> ummmmm, not last couple weeks, oh you mean socially, heh, don't be silly, in serious land tho i been getting maybe avg a night or towo just having a casual friends-house eveniing visit to have a beer and get some fresh air, which is nice
[14:56:25] <rue_shop3> indeed
[14:56:35] <theBear> you shoulda done the old parport for parasitical power and err, parallel portery
[14:56:51] <theBear> sif serial, that's too modern and tricky to interface to dumb logic ;-)
[14:57:05] <rue_shop3> its for microcontroller to microcontroller
[14:57:15] <rue_shop3> the tiny26 is THE cheapest "multichannel adc"
[14:57:41] <rue_shop3> I'm working on setting them up as low latency adcs for servo systems
[14:58:32] <rue_shop3> the current application has a tiny26 sending to a mega8, the mega8 also needs to use 5 adc channels of its own, and send out serial to a pwm genorator and do rs232 comms
[14:58:52] <rue_shop3> but one of the serial systems sits on two of the adc channel pins
[14:59:22] <rue_shop3> I dont want to do serial software
[14:59:30] <theBear> finally got my first usb-serial a couple days back, ft232 on the back of a tiny board with tiny usb hole and mega328 tiny chip size, just under 20 bucks on double sale/anniversary/coolguy-tradeprice-acct, and their actual 232 boards or cables or anything stnadalone still firmly sitting not selling at 30 bucks or more each ! i spose if i can judge by the last 6 months, they'll have to cost themselves a ton (capitalist accounting model) keeping the
[14:59:31] <theBear> little guys onn the shelf and overpriced into not selling for 3-5 years then gthey go in the next buying-dept-aren't-qualified slae :)
[15:01:07] <rue_shop3> but bear, you can get a pro-mini for about $2.50 and a cp2102 ttl-usb for about $1.50
[15:01:40] <rue_shop3> and you can use avrdude with the bootloader to put whatever hex file you want on the avr
[15:01:46] <rue_shop3> ...somehow
[15:02:11] <rue_shop3> (they must flip the boot vector when they put the bootloader on)
[15:05:54] <rue_shop3> maybe I ahve to use spi and put everything on the same loop
[15:06:51] <rue_shop3> I dont like that, it makes the queue manager on the adc chip more complex
[15:06:59] <theBear> i mean sure, in theory i could do that, but i like going to the city and ranting to peoples faces one or twice a fortnight at the moment :)
[16:03:49] <rue_shop3> hmm, I only occupy a small part of my existance
[18:11:02] <z64555> finally ordered a transmitter/reciever pair
[18:11:29] <z64555> I should focus on getting the AHRS working
[18:57:17] <z64555> hm. i need a power hub
[19:03:40] <z64555> I *could* make the chassis a ground
[19:10:23] <z64555> hm, or I could do that
[19:10:28] <z64555> Cut 8 pieces of cable
[19:10:34] <z64555> form into two squares
[19:10:45] <z64555> solder the ESC's power cable to those
[19:40:39] * z64555 tries not to stab himself
[19:55:48] <z64555> Actually, I think this was the original design, however at some point I switched to using connectors because we thought it would be easier to swap out ESC's that way
[19:55:53] <z64555> turns out, it isn't
[20:19:43] <z64555> in fact, I probably still have another pair of rings somewhere, hadn't installed them
[20:19:44] <z64555> oh well
[20:34:20] <z64555> hm, I must[ve consumed at least 3 teaspoons of honey today
[21:01:37] <anonnumberanon> what's you ahrs look like?
[21:01:43] <anonnumberanon> what RC did you get?
[21:02:56] <anonnumberanon> z64555,
[21:03:08] <anonnumberanon> sup
[21:03:15] <anonnumberanon> oX
[21:03:19] <anonnumberanon> z
[21:03:24] <anonnumberanon>
[23:45:14] <z64555> sorry, was watching TV
[23:45:43] <z64555> I got something from flysky, nothing fancy, a 4channel 2.something GHz one
[23:46:14] <z64555> Original battery pack uses 8 AA's, but I've read it's easy enough to hack in a LiPo
[23:46:22] * z64555 pings anonnumberanon
[23:46:46] <anonnumberanon> yeah
[23:47:14] <anonnumberanon> that in itself would be fun
[23:47:17] <anonnumberanon> the lipo
[23:48:30] <z64555> letsee, I've got the AHRS coded, but I need to make a test program for it to make sure it works. And probably a bunch of tweaking
[23:51:57] <anonnumberanon> could you make a program that shows the sensor in 3D in "real time"?
[23:52:56] <z64555> Yeah, did a bit of studying openGL code
[23:53:25] <z64555> the LM4F has a usb cable, and there's probably libs to allow communication via it
[23:53:54] <z64555> so I'd be able to make a program on the PC that would listen for it and display that data onto a dashboard, of sorts
[23:57:26] <anonnumberanon> yeah that would be cool. many IMUs have programs on github and stuff that let you do this
[23:57:50] <anonnumberanon> to verify it's working though you'd just need to send the data and look at it in the console
[23:59:54] <anonnumberanon> well that's what I did, showing it in 3D would have taken ages and i didn't have time