#avr Logs

May 22 2018

12:34 AM day__ is now known as day
03:39 AM rue_bed: Emil, does the sound thats playing generate that in XY mode with L/R on a scope?
03:40 AM Emil: hm?
03:40 AM rue_bed: er, polprog
03:40 AM rue_bed: the youtube link
03:40 AM Emil: aww, you had me on speeddial <3
03:41 AM rue_bed: somehow I suspect it does
03:46 AM polprog: rue_bed: yeah
03:47 AM polprog: its one of the demos that is more pleasant to listen to, unlike Youscope
03:53 AM rue_bed: a totally new way to look for hidden messages in music!
03:54 AM polprog: yeah
03:54 AM polprog: heh. youscope sounds like noise (due to way he was generating the image from several pixels)
03:54 AM polprog: yet he displays scrolling messages
03:55 AM polprog: cool steganography technique
03:55 AM rue_bed: its a good point that the exact waveform dosn't matter ahuge amount to the ear, and that the scan rate of xy image dosn't matter
03:56 AM polprog: the nice part is that he made a couple of waveforms that draw nice stuf and sound pleasant, and then composed a track from them
03:56 AM polprog: if you have one period that draws a cube
03:56 AM polprog: then it doest matter if you play it at 440hz or 1khz on the screen
03:56 AM rue_bed: yea there is a LOT of work there
03:57 AM polprog: but 440 and 1k are two different sounds
03:57 AM polprog: so you can compose a changing melody, that draws a cube
03:58 AM polprog: drum line adds rhtym to the audio layer and creates a cool distortion on the image layer
04:00 AM polprog: :)
04:06 AM rue_bed: nonally having to generate a melody for it, but needing the vector graphics toconvert
04:36 AM nuxil: polprog, hook up an mcu , make some code and see if you can send chars to the oscillioscope from pc. so you can display single chars or strings :p
04:38 AM polprog: thats the plan ;)
04:45 AM rue_bed: then have each spin off in 3d as the new ones arrive
05:29 AM polprog: the power of C (++) is in my hands
05:30 AM polprog: just need to check how to interface pulseaudio/alsa
07:05 AM thardin: feeling a bit peeved over ADC pin leakage current being way larger than it should
07:05 AM thardin: has anyone else experienced something like that?
07:06 AM thardin: for example, measuring the 3.3V rail via two 18k resistors in series -> 9k source impedance which should be fine according to the datasheet
07:06 AM cehteh: as in?
07:07 AM cehteh: what do you use as vref?
07:07 AM thardin: yeah I see 2.0 V in the middle instead of at most 1.65+0.018 ~ 1.67 V
07:07 AM thardin: cehteh: internal 2.56V
07:07 AM cehteh: and did you measure the resistors? maybe its just tolerance
07:08 AM thardin: a bit hard in circuit. could unsolder one. but they should be 1%
07:09 AM cehteh: also, using a RC for lowpass may improve it, but i never seen this problem, are you sure you configured the port correctly? no pullup, disable the digital functions etc
07:10 AM thardin: pullup is a good point, I'll check
07:11 AM thardin: the bootloader sets PORTF = 0xFF at least, so that's suspicion #1
07:12 AM polprog: set that to zero and check again
07:13 AM thardin: woop! my code is reading the 3.3V rail as +3.15V now
07:13 AM thardin: which may well be due to tolerances
07:13 AM thardin: thanks :)
07:14 AM cehteh: yes add software calibration and done
07:14 AM cehteh: http://tinyurl.com/y8k85jgn
07:14 AM thardin: I might still add a buffer in the next revision, just to eliminate some unknowns
07:15 AM cehteh: and RC filter .. whenever applicable
07:15 AM thardin: multimeter says 1.64 V over the divider, 3.28 V on the bus
07:17 AM thardin: ripple on the rails is around 200 µV, so I'm not really concerned about aliasing
07:18 AM cehteh: a little bit (uniform) noise can actually increase accuracy when you do oversampling
07:18 AM thardin: yarp
07:19 AM thardin: 2000 µV on the 24V bus, but that's expected switching noise
07:22 AM polprog: cool, falstad works on mobile as well
07:23 AM thardin: there's no hardware oversampling, right?
08:55 AM day: oversampling means taking more samples than strictly required.
08:56 AM day: i guess you could establish the logic for that in hardware.
08:56 AM cehteh: it adds some latency (lowpass filtering) so you dont always want thant
08:56 AM cehteh: and its damn easy to implement in software
08:56 AM cehteh: few asm instructions for rolling average only
08:57 AM cehteh: (or lines in C)
09:00 AM cehteh: generating noise or ramp the vref around 1 LSB is another thing
09:15 AM thardin: ye
09:54 AM polprog: oversample.js
09:54 AM polprog: peak sample rate
09:54 AM polprog: 100 hz
09:54 AM polprog: im sure someone implemented a framework for writing JS for stm32
01:31 PM Emil: polprog: that video was super cool!
01:32 PM polprog: i should make more
01:32 PM Emil: yes!
01:32 PM polprog: maybe one about calibrating the timebase gen
02:03 PM polprog: how cool is that
02:03 PM polprog: i found a video from 2013 on an old disk
02:04 PM polprog: it shows a lego generator i made with a coil and some magnets connected to my old scope
02:14 PM polprog: shit
02:15 PM polprog: that means im doing electronics for 5 years now
02:15 PM polprog: i feel old
02:23 PM thardin: do more! :]
03:08 PM TechChristoph: hi all
03:27 PM thardin: hullo
03:34 PM polprog: looking at the evolution of my lab
03:34 PM polprog: throughout those 5 years
03:35 PM polprog: i think that a wall of scopes like Dave Jones or mikelectricstuff has is inevitable
03:35 PM Tom_L: do you wear a white lab coat too?
03:38 PM polprog: hmm
03:38 PM polprog: no, maybe i should get one
03:38 PM polprog: conductive one :p
03:38 PM Tom_L: embroider "Evil Mad Scientist" on the front
03:52 PM polprog: heh
03:52 PM polprog: i should make myself a logo
03:53 PM polprog: maybe some cool electronics company could send me one coat
03:53 PM * polprog is thinking about buying an old one with Sun or Digital logo
03:56 PM polprog: https://www.youtube.com/watch?v=45X4VP8CGtk
03:56 PM polprog: me-irl
03:58 PM polprog: i should go back to programming stm32
04:05 PM Emil: polprog: whatcha use for the toolchain yo
04:06 PM Emil: opencm3?
04:06 PM polprog: Emil: yeah. http://polprog.net/papiery/stm32/Makefile-polprog
04:07 PM Emil: absolutely disgusting
04:07 PM polprog: fu
04:08 PM polprog: :(
04:08 PM polprog: lol
04:08 PM Emil: Why or why can't it be like on avr
04:08 PM polprog: because it needs a bajilion more registers
04:08 PM polprog: i mean
04:08 PM polprog: i use part this and part that
04:09 PM Emil: not_true_dot_jpg
04:09 PM polprog: sec
04:09 PM polprog: damn not this computer
04:09 PM polprog: anyway
04:09 PM polprog: for GPIO i just write to registers
04:10 PM polprog: but for stuff like uart baudrate setup i tried to do that the AVR way but its just too much fuckery
04:10 PM Emil: but there's only like 5 registers iirc
04:10 PM Emil: of course you need to config the clock for the peripheral which is 2 iirc
04:10 PM polprog: whatever... it was my first program
04:11 PM polprog: im planning to write in an avr style with C[++] and asm
04:11 PM Emil: Perhaps I should just yield and commit sudoku with opencm3
04:14 PM polprog: commit hentai
04:26 PM polprog: srsly though
04:27 PM polprog: i think ill get some dacs
04:27 PM polprog: and write fast code for them in assembly
04:27 PM polprog: well to display stuff on a scope 20khz is enough
04:27 PM polprog: but still i wanna see how fast i can get
04:28 PM polprog: so that will probably be, using opencm3 just for register names
04:28 PM Emil: bro
04:28 PM Emil: wtf
04:28 PM Emil: just load them to dma yo
04:28 PM polprog: yeah thats the plan
04:28 PM polprog: use dma with that
04:28 PM Emil: no need for asm for that, rite?
04:28 PM polprog: its not the point
04:28 PM Emil: Or how fast you are going to go
04:28 PM polprog: i wanna learn arm assembly
04:28 PM Emil: gotta go fast
04:28 PM polprog: this is a good project
04:29 PM polprog: or ill make some asm blinkenlights
04:29 PM polprog: heh. we'll see
04:56 PM polprog: night
05:01 PM Emil: good night
05:01 PM Emil: also
05:01 PM Emil: https://www.youtube.com/watch?v=INydhyGXDHo
05:01 PM Emil: a masterpiece
05:10 PM cehteh: https://youtu.be/INydhyGXDHo?t=82 thats you?
05:16 PM Emil: cehteh: hey now
05:16 PM Emil: My taste in animu is far more refined
05:17 PM cehteh: :D
05:18 PM cehteh: madoka wasnt bad, goethes faust retelled :)
05:18 PM cehteh: note i only scolled through the vid w/o sound
05:19 PM Emil: cehteh: noooo
05:19 PM Emil: cehteh: you _must_ listen to the whole thing :D
05:20 PM Emil: Audio is 100% the point of it
05:20 PM cehteh: yes yes, now on the sofa with laptop and headphones
05:20 PM cehteh: gimme 10 minutes
05:20 PM Emil: yesh
05:22 PM cehteh: so you are single?
05:23 PM Emil: heh
05:25 PM cehteh: well somehow makes me thinking "why do i know most of the character in that video" perhaps i watch too much anime
05:25 PM cehteh: but but usually/often i do so with my wife :D
05:26 PM Emil: just watch the video yo :D the strategic timing of that self loathing implied trash-level existence is superbly humorous
05:31 PM cehteh: lol about the end
05:37 PM Emil: ikr
05:39 PM cehteh: otakus stuck in 2D :D
05:46 PM Emil: https://www.youtube.com/watch?v=0tUDW0Zm1xE super weird audio (I bet it's because of the frequency response: some frequencies are attenuated like your ears would)
06:54 PM Emil: E X T R A T U C E V A
10:29 PM toddpratt: if a datasheet shows PCINTn on a pin, would i be correct in thinking that pin supports external interrupt?
10:33 PM Thrashbarg: yup
10:33 PM Thrashbarg: Pin Change Interrupt
10:34 PM Tom_L: read a byte of pins and mask the one you want
10:35 PM Thrashbarg: yea that's what PCINT does too, but it allows the software to be interrupted
10:35 PM Thrashbarg: I'm sure you know this :P