#avr Logs

Oct 22 2017

#avr Calendar

01:55 AM rue_bed: the magnetivity drops off with heat
01:58 AM rue_bed: ..
02:33 AM day is now known as daey
04:23 AM day is now known as daey
04:44 AM day is now known as daey
05:28 AM day is now known as daey
05:54 AM day is now known as daey
06:48 AM day is now known as daey
06:58 AM day__ is now known as daey
08:49 AM LiaoTao_ is now known as LiaoTao
08:50 AM renn0xtk9: WARN src/common.c: unknown chip id! 0xe0042000 when wanting to upload on ST link v2 on stm32 does anyone know something about it?
09:00 AM renn0xtk9: anyone who develop for STM32 on linux?
09:06 AM Lambda_Aurigae: looks like the chip IP is wrong.
09:07 AM Lambda_Aurigae: or
09:07 AM Lambda_Aurigae: your chip is in sleep mode?
09:07 AM Lambda_Aurigae: or
09:07 AM Lambda_Aurigae: read this?
09:07 AM Lambda_Aurigae: https://github.com/texane/stlink/issues/62
09:08 AM Lambda_Aurigae: https://github.com/texane/stlink/issues/107
09:08 AM Lambda_Aurigae: Short the BOOT0 pin with VDD
09:08 AM Lambda_Aurigae: Reset the board
09:08 AM Lambda_Aurigae: st-flash erase
09:08 AM Lambda_Aurigae: google is your friend.
09:09 AM Lambda_Aurigae: that second link was the first hit in a google search for your error.
09:19 AM renn0xtk9: I have read those but they don't solve it there
09:19 AM renn0xtk9: apart from pulling a windows computer which I don't have
09:28 AM renn0xtk9: okay in my case it seems to work after mass erase but weir though
09:28 AM renn0xtk9: and don't know how to avoid that
09:28 AM Lambda_Aurigae: erase the whole chip before writing.
09:28 AM Lambda_Aurigae: it's kinda standard for flash based devices.
09:43 AM renn0xtk9: Lambda_Aurigae: no way to avoid the "short the boot pin" stuff ? that constraint me to use the jumper every time:S
09:48 AM Lambda_Aurigae: dunno.
09:48 AM Lambda_Aurigae: never had the problem...that's just what I see when I search for your error.
09:53 AM renn0xtk9: Lambda_Aurigae any idea why this would not work? https://github.com/renn0xtek9/USB2I2C-STM32/blob/master/Src/main.c I am trying to toggle a pin as they do in https://letanphuc.net/2015/02/stm32f0-tutorial-gpio-blinking-led-cubemx-keil-source-insight/ but nothing happens
09:55 AM Lambda_Aurigae: no clue.
09:55 AM Lambda_Aurigae: would have to study your code and figure out how everything is setup.
09:55 AM Lambda_Aurigae: and I'm not up to code review on a sunday morning
09:56 AM Lambda_Aurigae: you might check with the people in ##stm32
11:27 AM Ristovski: I shorted one of the unused pins to ground (via an LED) on my attiny45 and now its output current is very low. Mind this all happened on 3.3V
11:28 AM Ristovski: How is this even possible
11:36 AM Jartza: Ristovski: maybe you burned the protection diode from the pin
11:37 AM Ristovski: Jartza: At 3.3V through a resistor?
11:39 AM Jartza: depends of the resistor, but if you went over the max. load one pin can handle, possible
11:40 AM Jartza: I think it's around 40mA on attiny
11:40 AM Ristovski: Hmm, interesting
11:41 AM Jartza: Absolute Maximum Ratings in datasheet says "DC Current per I/O Pin ............................................... 40.0 mA"
11:42 AM Jartza: and to whole chip, 200mA
11:43 AM Jartza: and if I remember correctly, the current one pin can source is even smaller
11:43 AM Ristovski: Its weird tho, it still works perfectly
11:44 AM Jartza: "The sum of all IOL, for all ports, should not exceed 60 mA."
11:45 AM Jartza: internally, each I/O pin has protective diode to both GND and VCC
11:46 AM Ristovski: I tested all of them, and they all seem to provide very low output current
11:46 AM Jartza: plus some capacitance and internal pull-up that can be enabled/disabled
11:46 AM Jartza: how do you measure the output current is low?
11:47 AM Jartza: are you absolutely sure you are driving them as output, and not input with pull-up resistor enabled :)
11:48 AM Ristovski: I was testing PWM with an LED when this happened, and after I accidentally touched one of the non connected floating pins with the resistor connected to the LED, the whole chip seems to be outputting low current
11:48 AM Jartza: the internal pull-up resistor is 20..50kOhm
11:49 AM Jartza: which would mean you get somewhere in the vicinity of 0.1mA
11:49 AM Ristovski: Right now its just setting PB0 to HIGH and the led is really dim
11:49 AM Ristovski: happens on all pins
11:49 AM Jartza: setting PB0 of DDRB to high and setting PB0 of PORTB to high?
11:50 AM Jartza: dim led is not proof of small current, though
11:50 AM Jartza: maybe you burned the led itself
11:50 AM Ristovski: The LED was the first thing I tested
11:50 AM Lambda_Aurigae: does the chip work otherwise?
11:51 AM Ristovski: Lambda_Aurigae: Appears to
11:51 AM Lambda_Aurigae: if so, you have damaged something in the pin drive somewhere...throw it away and buy a new one.
11:51 AM Ristovski: I even checked the whole 8k of flash
11:51 AM Lambda_Aurigae: or use it for other things.
11:51 AM Lambda_Aurigae: do you actually get 3.3V out of it when set high?
11:51 AM Ristovski: let me re-check
11:52 AM Ristovski: ~2V
11:54 AM Ristovski: 0.4mA according to my meter
11:56 AM Lambda_Aurigae: and you are certain your supply voltage is good?
11:56 AM Ristovski: Absolutely
11:58 AM Lambda_Aurigae: then you got a bad chip
11:59 AM Lambda_Aurigae: I've shorted AVR pins to GND and VCC before and never damaged one.
11:59 AM Lambda_Aurigae: always a first time.
12:00 PM HighInBC: I have
12:01 PM HighInBC: they are sensitive little souls
12:02 PM Lambda_Aurigae: must be the newer ones.
12:02 PM Lambda_Aurigae: I haven't bought any new ones in 7 years, give or take.
12:02 PM Lambda_Aurigae: not since I got 50 atmega1284p
12:04 PM HighInBC: mine was the mega328
12:04 PM Lambda_Aurigae: if I want a 28 pin chip I go for the pic32mx270f256b.
12:08 PM HighInBC: if I wanted to go 32 bit I would probably reach for an stm32
12:08 PM Lambda_Aurigae: I find pic32 to be a bit easier to use than stm32.
12:08 PM Lambda_Aurigae: mainly because I know pic peripherals already.
12:08 PM Lambda_Aurigae: and I like the mips core.
12:11 PM Lambda_Aurigae: I also like it for the single chip solution in a dip package that's easy to play with.
12:12 PM Lambda_Aurigae: used to be a dip arm chip but that went the way of the dodo.
12:17 PM HighInBC: I like dip for prototyping but I never use them in final designs
12:17 PM HighInBC: even for prototyping I tend to just use some kind of breakout board
12:17 PM HighInBC: surface mount is just easier, no drilling, faster to solder etc
12:18 PM Lambda_Aurigae: I don't make much in the way of final designs.
12:18 PM Lambda_Aurigae: it's mostly just hobby playing these days.
12:18 PM Lambda_Aurigae: and teaching.
12:27 PM HighInBC: I am just a hobbyist too, I tend to go through the motions of making a proper PCB for anything I want to keep though. If only for the experience
12:29 PM Lambda_Aurigae: haven't actually made a keeper project in a few years.
12:29 PM Lambda_Aurigae: mostly I futz around with new stuff and learn.
12:29 PM Lambda_Aurigae: physical work time lately has been more woodworking than electronics though.
12:33 PM HighInBC: I just got a CNC so I am going to start playing with wood
12:33 PM HighInBC: no innuendo intended
12:38 PM polprog: i dislike dip for anything but breadboard
12:38 PM polprog: dip pcbs are yugee
12:38 PM thardin: it's easy to route between dip pads :]
12:38 PM polprog: one pcb i made with smd components and tqfp avr was the size of two dips of that avr
12:39 PM polprog: thardin: yes, it is
12:39 PM polprog: dip and tht sure looks way more easy to route
12:39 PM thardin: your typical 50 mil SO package is quite nice
12:40 PM polprog: those are my favourite
12:40 PM thardin: could maybe to down to 1 mm spacing and still be friendly to probes, bodge wires and such
12:40 PM thardin: 25 mil is right out
12:41 PM polprog: i managed to bodge some (thick) wires on a tqfp
12:41 PM polprog: once
12:41 PM polprog: i think i have pics
12:41 PM polprog: single cat5 wire
12:42 PM HighInBC: I have some 30awg kynar wire I use to bodge onto small pitched pins
12:42 PM polprog: ^^
12:42 PM polprog: kynar <3
12:42 PM HighInBC: http://images.highinbc.com/GRBL_to_DB25_Rev_A.jpg
12:43 PM thardin: noice!
12:43 PM polprog: nice
12:43 PM thardin: and yes, kynar
12:43 PM polprog: i have red and violet kynar
12:44 PM polprog: old project with bodgewire: https://puu.sh/y4qzn/e6289b3de1.jpg
12:44 PM thardin: I managed to order wrong diameter both kynar and ptfe wire recently. annoying
12:44 PM polprog: i dont have a choice, where i buy they just have different colors :D
12:44 PM thardin: clean your boards man :]
12:44 PM polprog: this one is like.. 1 year old
12:45 PM polprog: laid in the drawer and that was the time i still used rosin for smd
12:45 PM polprog: now i use no clean flux
12:46 PM polprog: HighInBC: what was that db25 nano supposed to be? printer controller?
12:46 PM HighInBC: it is for my CNC router
12:46 PM HighInBC: the nano turns gcode text into stepper motor pulses
12:46 PM polprog: cool
12:47 PM HighInBC: the driver for the CNC uses db25 and optical isolation
12:47 PM HighInBC: I am driving it from the parallel port of an old computer right now, but want to convert it to a raspberry pi->usb->nano->db25
12:48 PM HighInBC: once I change some pins around(I mapped some things wrong) I will make a PCB
12:48 PM thardin: uhm, don't pi:s have enough gpio? or maybe the OS gets in the way of doing the stepping
12:49 PM thardin: do people do no-OS stuff on the pi, like just use it as a micro?
12:49 PM polprog: its possible to program pi w/o any os
12:49 PM polprog: but nobody does it
12:49 PM polprog: its easier to make a realtime avr communicate with pi
12:50 PM polprog: anything that needs accurate timing is undoable under linux on a pi
12:50 PM polprog: un-doable*
12:51 PM thardin: maybe a real-time os would be appropriate
12:51 PM polprog: maybe
12:51 PM thardin: feels like a bit of a waste having to have two modules
12:52 PM polprog: not really, you can have pi running networking and controlling a swarm of small devices via spi/uart/i2c
12:52 PM polprog: or rs485 <3
12:53 PM thardin: rs-485 = win
12:53 PM LeoNerd: '485 is nice
12:53 PM LeoNerd: Though a little bit of work to have a multimaster bus arrangement out of it
12:54 PM thardin: unicast is a bit tricky to deal with sometimes. unless you go duplex rs-485 of course
12:54 PM thardin: s/unicast/simplex
12:54 PM HighInBC: thardin: the atmega328 is single threaded and real time
12:54 PM polprog: any atmega is technically :D
12:54 PM thardin: the pi is too, if you don't bother with the OS
12:54 PM LeoNerd: My usual arrangement for multimaster-ish stuff over 485 is basically token passing... at any moment in time exactly one node has transmit permissions
12:54 PM HighInBC: the pi is going to be doling out cycles to my software at its leisure
12:54 PM thardin: and conversely, an AVR isn't if you toss an OS on it :]
12:54 PM HighInBC: well I want an OS
12:55 PM LeoNerd: A master node holds it normally and periodically polls all the others, one at a time, to ask if they have anything interesting to say
12:55 PM HighInBC: let the pi use its strengths and the avr its
12:55 PM polprog: this is an example of a very-non-realtime os: https://www.youtube.com/watch?v=nm0POwEtiqE
12:55 PM HighInBC: but yes you can probably run it directly from the GPIO
12:55 PM HighInBC: may need to adjust the optoisolation board to deal with 3.3V signals
12:56 PM polprog: something bothers me for a longer time now
12:56 PM thardin: polprog: still probably faster than a pdp-8
12:56 PM LeoNerd: I tend to use magnetics or capacitive-based isolator chips these days. Far nicer and easier to work with than traditional optoisolators
12:56 PM thardin: or whichever machines it was the unix was originally written for
12:56 PM polprog: how do you make two-way isolation or lever conversion (real level conversion)?
12:56 PM Tom_L: LeoNerd, how so?
12:56 PM Tom_L: i've never tried them
12:56 PM Lambda_Aurigae: kinda miss my pdp-11/73
12:57 PM LeoNerd: Tom_L: things like the ADuM14xx series chips
12:57 PM LeoNerd: They just act like a logic buffer
12:57 PM LeoNerd: E.g. it might just feel like one of those TTL buffer chips like a 7404
01:01 PM Tom_L: huh
01:01 PM LeoNerd: Some of the chips even have things like hiZ-capable outputs with an enable pin
01:01 PM LeoNerd: E.. the 1401 is lovely; 3 signals in one direction and 1 in the other. That sounds like an SPI bus. The OE pin can be used then on the inbound pin, controlled from SS, to make it behave sensibly
01:02 PM HighInBC: love those chips that share buses
01:02 PM LeoNerd: Muuuuuch faster capable than an optoisolator, too. 10Mbit/sec is easily doable
01:04 PM LeoNerd: Though yaknow really annoyingly, the OE lines are active-high, so if you want to easily use it as an SPI chip you basicially ignore the real OE line and stick a 741G125 on it instead
01:04 PM polprog: am i thinking right if i use two buffers both way, or would they just lock up? im looking at pca9521 datasheet
01:05 PM LeoNerd: Hm?
01:05 PM LeoNerd: Oh, for bidirectional stuff, e.g. for an I²C bus you want to use one of the dedicated I²C-capable chips
01:05 PM polprog: yeah
01:06 PM LeoNerd: ADuM1251 I -think- if memory serves
01:06 PM LeoNerd: Those usually rely on the slight voltage range difference that is permitted by I²C devices
01:06 PM polprog: what if i wanna use one for non i2c or one to interface low level logic to high voltae logic - two level standards that are not compatible (like 1.2V and 5V logic)
01:07 PM LeoNerd: Different voltage levels are totally fine
01:07 PM LeoNerd: almost universally these chips have no requirement to use the same voltage level on one side as they other. They're independent
01:07 PM polprog: yeah, but what if i wanna make one using, say, some noninverting buffers
01:08 PM LeoNerd: That's not galvanic isolation then
01:08 PM polprog: if i connect the output of one to the input of one will it work or will it lock up?
01:08 PM LeoNerd: Oh... most probably yes
01:08 PM polprog: not talking about galvanic isolation nhere
01:08 PM LeoNerd: Ah; well if you do'nt want galvanic isolation there's much simpler ways to do that sort of level-shifting
01:08 PM polprog: probably yes meaning lock up?
01:49 PM Ameisen: Hello.
01:49 PM polprog: .olleH
01:49 PM Ameisen: Got sent here by #reprap, as I'm doing AVR work on a fork of gcc, trying to get better AVR functionality into g++
01:50 PM Ameisen: and in the future, Clang, though Clang's AVR support is rather weak
01:53 PM Ameisen: Also, I pushed this bug report regarding AVR into the gcc bugzilla: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=82658
01:54 PM Ameisen: If anyone could validate it exists/doesn't exist on older toolchains (I think the atmel toolchain uses a modified GCC 4 which I haven't tested on) that'd be swell
01:54 PM Ameisen: as regression is useful, and so far I've found the bug on as far back as I have
02:04 PM rue_mohr: whats clang?
02:04 PM polprog: another compiler
02:05 PM rue_mohr: ah, dont use it, just use gcc
02:05 PM rue_mohr: FIXED!
02:05 PM rue_mohr: next...
02:05 PM polprog: would like to help but i have avr-g++ 7.2.0
02:05 PM rue_mohr: furthermore, dont program 8 bit microcontrollers in c++
02:07 PM thardin: Ameisen: amazing that avr-gcc can be considered superior
02:07 PM thardin: it often seems to generate rather daft code
02:08 PM Ameisen: thardin - it's confusing more in that gcc and g++ have the same backend
02:08 PM Ameisen: for that test case, the output should be 100% the same
02:08 PM Ameisen: for whatever reason, g++ is ignoring the built-in logic for logical 8-bit right shift for any shift amount <= 3
02:09 PM Ameisen: rue_mohr - clang is the C language family frontend for LLVM
02:09 PM thardin: strange that it behaves differently
02:09 PM Ameisen: honestly, clang tends to be better than gcc at most things, but it's AVR support is very rough still
02:09 PM Ameisen: it was only added to mainline from a branch earlier this year, and it's still experimental
02:09 PM Ameisen: However, I suspect that it will end up replacing gcc for AVR work once it gets its feet on the ground
02:09 PM thardin: you'd think the intermediate format is sufficiently platform agnostic
02:09 PM thardin: maybe llvm would be better at that come to think of it
02:10 PM Ameisen: mainly because it ties in better, is more flexible, and their maintainers don't block things like adding __flash to the C++ frontend
02:10 PM Ameisen: thardin - both the gcc IL and LLVM should work fine for this
02:10 PM thardin: ye
02:10 PM Ameisen: I have no idea yet what's different on the the C++ side. Since I'm trying to get other AVR features into g++, I pushed hte bug report
02:10 PM Ameisen: if nobody looks at it in a few days, I will take another gander at it
02:11 PM Ameisen: It's a rather serious performance bug, so
02:11 PM thardin: should've taken a compiler theory course at uni. llvm got me slightly motivated a while back, but I couldn't make too much sense of it
02:11 PM Ameisen: right now I'm trying to add a constexpr lpm builtin to g++
02:11 PM thardin: wanted to write stuff for targeting 6502
02:11 PM polprog: i get a feeling that super timecritical stuff should be hand written
02:11 PM Ameisen: I'm still not super comfortable with either gcc's source _or_ AVR
02:12 PM Ameisen: and my AVR chips I have to test have LPMX, so I'm unsure what the arduino SDK is doing in the non-LPMX branch
02:12 PM thardin: polprog: yes, asm tends to become a requirement on all platforms
02:12 PM Ameisen: like, it adiw's on r30 when reading 4 bytes from program memory... but I'm unsure why
02:12 PM thardin: there's been plenty of wtf:ing at both gcc and clang in #ffmpeg-devel
02:13 PM Ameisen: asm becomes necessary because nobody ever implements proper intrinsics, and sometimes they miss optimizations when they are writing platform overloads
02:13 PM Ameisen: Worse in that often the intrinsics were written for C, and never had C++ functionality like constexpr-awareness added
02:13 PM Ameisen: so you get suboptimal C++
02:14 PM polprog: what do you expect on an 8 bit 20mhz micro
02:15 PM polprog: some of them dont even have enough memory
02:16 PM Ameisen: polprog - I expect the compiler to work correctly.
02:16 PM Ameisen: Regardless of target.
02:16 PM polprog: yeah
02:16 PM polprog: but it's just a computer program
02:17 PM polprog: im not saying it's not an error
02:17 PM polprog: but im not surprised either
02:17 PM Ameisen: I am doing almost pure C++ for AVR at the moment, and overall my C++ outperforms C (you can do amazing things with templates and constexpr in deriving better algorithms at compile-time), but the lack of ability to call/use specific instructions hurts
02:17 PM Ameisen: the builtin intrinsics exist, but they lack contextual information for C++ to know that the value might be known at compile-time
02:19 PM Lambda_Aurigae: sounds like a lot of programming keyword bingo to me.
02:19 PM Lambda_Aurigae: my algorithms are created when I write the code. I don't let the compiler create them.
02:19 PM Ameisen: Lambda_Aurigae - I am writing code that is designed to work on multiple chips, with multiple sets of data from tables depending on configurations
02:20 PM Lambda_Aurigae: sounds like ardweeny
02:20 PM Ameisen: so, I autogenerate certain functions at compile-time based upon values from the tables, and the chips, using different types
02:20 PM Lambda_Aurigae: I write code to work on the chip I'm using to build the project I'm working on.
02:20 PM Ameisen: However, if the table is in program memory, it's a PITA to do that due to g++ having very limited support for flash memory
02:20 PM Ameisen: __flash doesn't work, and the other way is to use inline asm which obviously isn't knowable at compile-time
02:21 PM polprog: what if i told you you can always put asm volatile to read from a block of the memory
02:21 PM polprog: alsmost as fast as possible
02:21 PM Ameisen: polprog - good luck deriving compile-time values from an asm block
02:21 PM Lambda_Aurigae: polprog, it's not intrinsic in the programming language.
02:21 PM polprog: linker symbols?
02:21 PM Ameisen: constexpr.
02:21 PM Lambda_Aurigae: so, rather than knowing the hardware you are working on, use this huge table to create a bunch of code you don't really understand.
02:21 PM Ameisen: I cannot derive language constructs without values being known at compile-time.
02:22 PM Lambda_Aurigae: it's called C++
02:22 PM polprog: that's what the bloody linker is for
02:22 PM polprog: i have no idea where my linker puts the binary blob i linked in
02:22 PM Ameisen: The linker doesn't generate code unless you're using LTO, and even then it will only generate code based upon the IL that comes in.
02:22 PM Lambda_Aurigae: the rest of us just write code we understand and link in libraries.
02:22 PM polprog: but i can define extern and use it
02:22 PM Ameisen: Lambda_Aurigae - 3d printers
02:22 PM Lambda_Aurigae: Ameisen, my 3d printer has no C++ in it at all.
02:22 PM Ameisen: I literally cannot know all the hardware in advance, and don't want to write N potential branches for each chip, and each thermistor, etc
02:23 PM Ameisen: Lambda_Aurigae - I mean, I could write the firmware in C or asm
02:23 PM Lambda_Aurigae: and never will.
02:23 PM Ameisen: I just don't see the benefit
02:23 PM Lambda_Aurigae: this is how we end up with software like windows 10
02:23 PM polprog: i think your problem is that you wanna write code for everything, and stuff that universal usually sucks in one way or another
02:24 PM polprog: in the most extreme way you have two different pieces of code separated by preprocessor directives in one C file
02:24 PM polprog: cpp*
02:24 PM polprog: and that's just absurd
02:24 PM Lambda_Aurigae: take something simple and write a bloated wrapper for it so it handles every processor and motherboard and video card on the planet....but doesn't do any of it particularly well so we have to add special drivers for certain chipsets or video cards or printers....
02:24 PM Ameisen: Possibly. Sort of past the point, though - the g++ maintainers won't add __flash to g++, so it limits a lot of capabilities
02:25 PM polprog: write that piece in regular c
02:25 PM Ameisen: Lambda_Aurigae - compile-time derived code doesn't bloat
02:25 PM Lambda_Aurigae: you are just bloating it before hand.
02:25 PM Ameisen: it generates one instance of the function depending on the values presented to it.
02:25 PM Ameisen: though I'm not particularly willing to get into a C vs C++ argument
02:26 PM Ameisen: since that argument has been done for the last 30 years, and I suspect we aren't going to cover any new topics :D
02:26 PM Ameisen: At the moment, I'm just trying to extend g++'s avr support
02:26 PM Lambda_Aurigae: ultimately, your question is kinda lost in this channel for the most part as we tend to use C for microcontrollers and don't bother with templates and intrinsic dohickies in C++
02:27 PM Lambda_Aurigae: there are some here who use C++
02:27 PM Ameisen: I prefer intrinsics (in C as well) - cleaner and easier to read
02:27 PM Ameisen: than a large asm block
02:27 PM polprog: speak for yourself :>
02:28 PM Lambda_Aurigae: those awake/alive/conscious currently tend to use C the old fashioned way, the way K&R meant it to be used!
02:28 PM Ameisen: polprog - I mean, I did say "I prefer"...
02:28 PM Ameisen: so I pretty much was
02:28 PM polprog: ah, my bad
02:28 PM Ameisen: K&R? Real men write BCPL.
02:28 PM Ameisen: and we like it.
02:28 PM Lambda_Aurigae: K&R wrote C
02:28 PM Ameisen: I know, I'm just going further back.
02:29 PM Lambda_Aurigae: but we don't have a BCPL compiler for AVR, far as I know.
02:29 PM Ameisen: K&R also developed B, which was the predecessor to C, and based upon BCPL
02:29 PM Ameisen: Lambda_Aurigae - could write one
02:29 PM Ameisen: make a BCPL frontend for LLVM
02:29 PM Lambda_Aurigae: naaa.
02:29 PM Ameisen: not sure why you'd want to
02:29 PM Lambda_Aurigae: I already have a C interpreter in the works for AVR.
02:30 PM Lambda_Aurigae: well, C subset...but,,
02:30 PM Ameisen: ... why...
02:30 PM Lambda_Aurigae: for the fun of it.
02:30 PM Lambda_Aurigae: C instead of BASIC
02:30 PM * Ameisen ports his MIPS32r6 emulator to AVR
02:30 PM Lambda_Aurigae: bah...I already have someone else's ARM emulator for AVR.
02:30 PM Lambda_Aurigae: not particularly useful but an interesting toy.
02:31 PM polprog: i linked an x86 emulator for avr
02:31 PM Lambda_Aurigae: 32 bit arm, with thumb support, running on an 8-bit avr.
02:31 PM polprog: the linux video
02:31 PM Lambda_Aurigae: I even connected Jartza's octapentaveega to it for the fun of it.
02:32 PM Ameisen: I was going to port my MIPS32r6 emulator to MIPS32r6
02:32 PM Ameisen: then I thought to myself "what the hell am I doing with my life?"
02:32 PM Emil: Jartza: mate
02:32 PM Emil: You know what you should do next
02:33 PM Emil: A HDMI generator
02:33 PM Emil: With some integrated functions
02:33 PM Emil: I'm dangerously close to entering the world of trusted devices and that
02:33 PM Emil: 's one of the required things for presentations
02:33 PM Lambda_Aurigae: hdmi, from an avr?
02:33 PM Emil: No
02:33 PM Lambda_Aurigae: hmmmmm.
02:33 PM Emil: Not from AVR :D
02:34 PM Lambda_Aurigae: would need something slightly faster.
02:34 PM Emil: Thought that'd be _wicked_
02:34 PM Lambda_Aurigae: I would rather see someone bitbang sata.
02:34 PM Lambda_Aurigae: but
02:34 PM Lambda_Aurigae: ok...gotta go recompile some wood into useful devices.
02:34 PM Ameisen: oh, derp
02:35 PM Ameisen: that's why they're using adiw; lpm in that case reads from r30
02:35 PM Ameisen: now I feel dumb
02:35 PM polprog: hdmi from avr? it has a pixel clock of minimum 25MHz so you'd have to use liquid nitrogen cooled avr :DD
02:36 PM Emil: polprog: nooo
02:36 PM Emil: polprog: you don't
02:36 PM polprog: or many chips with their clocks different in phase, that would be sick
02:36 PM Emil: You can overclock to 25MHz just fine
02:36 PM Emil: "just fine"
02:36 PM Emil: to reach like 40MHz you need dat LN
02:36 PM thardin: LN2 is fun
02:37 PM Emil: Sure is
02:37 PM Emil: and cheap
02:37 PM polprog: if i ever wanted to shatter a fruit/some meat i'd go for it
02:37 PM thardin: most boards deal with it surprisingly well. so long as they don't use electrolytics
02:37 PM polprog: instead of cutting tomatoes, cool them down to LN temp, and break into pieces with a hammer
02:37 PM polprog: then unfreeze
02:39 PM thardin: I prefer just to keep my chef knife sharp
03:05 PM polprog: if i had a sig gen i'd overclock avrs :D
03:09 PM Lambda_Aurigae: up to 25MHz things ran nicely except for the ADC.
03:11 PM Lambda_Aurigae: liquid nitrogen is good for liquefying propane in a test tube.
03:11 PM polprog: will experiment
03:11 PM Lambda_Aurigae: rubber stopper with hole and glass tube in it...stuck in test tube....connect to propane supply from bench and submerge test tube in liquid nitrogen and turn on propane supply.
03:12 PM polprog: fun until you freeze your fingers off :P
03:12 PM Lambda_Aurigae: give it a few minutes and shut off propane supply, remove from liquid nitrogen, remove connection from propane supply, watch propane boil.
03:12 PM Lambda_Aurigae: light top of glass tube.
03:12 PM Lambda_Aurigae: watch pretty flame.
03:12 PM Lambda_Aurigae: run finger down test tube and watch pretty flame grow dramatically,
03:13 PM Lambda_Aurigae: tip test tube over and pour liquid propane onto floor, along with lots of flames....in the process, freezing then burning the wax on the floor and pissing off school janitor.
03:13 PM polprog: lol
03:13 PM Lambda_Aurigae: yes, it took 3 times doing that before we were told not to do it anymore.
03:14 PM polprog: we used to make smears on wooden bench elements with H2SO4
03:14 PM polprog: the teacher was pissed
03:15 PM Lambda_Aurigae: had a bottle of high concentration ammonia once and 4 of us were standing around pretending to sniff from it.
03:15 PM Lambda_Aurigae: class druggie came over, grabbed the bottle, and took a big deep sniff...
03:15 PM polprog: lol
03:15 PM Lambda_Aurigae: turned cocacola red.
03:15 PM polprog: cruel
03:15 PM polprog: friend poured ethanol all over his bench then lit it up
03:15 PM polprog: fun time
03:15 PM polprog: this schools chemistry lab sucks
03:15 PM Lambda_Aurigae: a couple of days later the school had to be evacuated due to ammonia smell...took days to get the smell out.
03:16 PM Tom_L: acy in a balloon wakes the shop up in the morning too
03:16 PM polprog: and since this year we dont have chemistry in the timetable at all - maths/phys profile.
03:16 PM polprog: that sucks
03:16 PM polprog: acy?
03:16 PM Lambda_Aurigae: our school was one big monolithic building built around the central gym..3 stories plus basement under the gym.
03:16 PM Tom_L: acetylene
03:16 PM polprog: ah
03:17 PM Tom_L: add a little oxy for added effect
03:17 PM polprog: guess it makes a nice bang
03:17 PM Lambda_Aurigae: one biiig honking AC/furnace unit to feed the whole school.
03:17 PM Lambda_Aurigae: someone took that ammonia and poured it into the air filters for the AC unit.
03:17 PM Lambda_Aurigae: polprog, serious bang.
03:17 PM polprog: i used to bang away hydrogen baloons
03:17 PM Lambda_Aurigae: 30 gallon trash bags full of oxy/acet works well too.
03:17 PM Tom_L: hah i bet
03:18 PM Lambda_Aurigae: dozen or so of them in a big empty field
03:18 PM Lambda_Aurigae: shoot at them with flaming arrows.
03:18 PM Lambda_Aurigae: kaWOOMP!
03:18 PM polprog: i guess if you done that indoors windows would fly
03:18 PM Lambda_Aurigae: yeah, pretty much.
03:18 PM polprog: safety first!
03:18 PM Tom_L: not ballons
03:18 PM Lambda_Aurigae: safety...bah.
03:18 PM Tom_L: bags, yes
03:18 PM Tom_L: don't do it when the osha guy is there
03:18 PM polprog: once drunk i threw a deodorant into a campfire
03:18 PM polprog: at 3am
03:18 PM Lambda_Aurigae: our chem/physics teacher always had an oxy/hydrogen splitter in the classroom by his desk.
03:19 PM polprog: luckily noone woke up
03:19 PM Lambda_Aurigae: he would bubble it into a bowl of soapy water then toss a match into the bowl.
03:19 PM polprog: and noone apart from some friends remembers that
03:19 PM Lambda_Aurigae: kerbang!
03:19 PM Lambda_Aurigae: wake up the classroom.
03:20 PM Lambda_Aurigae: as kids we found a shed full of half used spraypaint cans.
03:20 PM polprog: our chem teacher would put a soab bubble full of propane-butane and light it up
03:20 PM Lambda_Aurigae: made a launcher for them.
03:20 PM Lambda_Aurigae: long tube with nail in the bottom.
03:20 PM polprog: holding fire in your hands ;) that was cool, i did that too
03:20 PM Lambda_Aurigae: drop can down tube, puncture bottom of can, SPFFFFFFFFFFFT! and away it would fly.
03:20 PM Lambda_Aurigae: like a mortar.
03:20 PM Lambda_Aurigae: we punched a hole in the tube and held flame to it to make flaming mortars.
03:21 PM Lambda_Aurigae: safety,,,bah.
03:21 PM Tom_L: made a bb gun of sorts in shop class that used firecrackers as propellant about 8" long tube
03:21 PM polprog: classic, we did somethign loke that too
03:21 PM Lambda_Aurigae: we got caught because of the paint in our hair.
03:21 PM polprog: lol
03:21 PM Tom_L: back when we had steel trashcans, went thru 4 cans
03:21 PM Lambda_Aurigae: golf ball cannon was fun too.
03:21 PM Tom_L: tennis ball
03:21 PM Tom_L: and lighter fluid
03:21 PM Lambda_Aurigae: naaa.
03:22 PM polprog: tennis balls make nice smoke flares
03:22 PM Lambda_Aurigae: estes model rocket igniter, black powder, threaded pipe with cap on one end that juuust fit a golf ball.
03:23 PM Lambda_Aurigae: line several of them up along the fairway of the 15th hole on the local golf course...with long wires back to a battery.
03:23 PM Lambda_Aurigae: wait till some guy is lining up a shot and bambambam!
03:23 PM polprog: rofl
03:23 PM Lambda_Aurigae: hit the old fucker in the ass and knocked him on his face.
03:23 PM Lambda_Aurigae: we ran like hell and hid for hours while the cops looked for us.
03:24 PM polprog: vietnam flashbacks
03:24 PM Lambda_Aurigae: my father was a vietnam vet...when he heard he laughed his ass off....I told him 5 years after the fact.
03:25 PM Lambda_Aurigae: ok...gotta go pick up my woodworking tools.
03:25 PM polprog: have fun
03:43 PM Jartza: well. attiny85 runs just fine at 40MHz, fed 5.5V :)
03:47 PM Tom_L: warm?
03:50 PM polprog: i think i have a 14 mhz gen, and then a 32mhz gen
03:50 PM polprog: s/gen/oscillator
03:50 PM polprog: i lack the ones in between
04:27 PM Emil: I hate the ESP8266
04:27 PM Emil: I wish there was an alternative as cheap
04:27 PM Lambda_Aurigae: newer version isn't there?
04:28 PM Emil: I just want a stupid fucking chip that handles setting up the TCP/UDP connection by host or ip
04:29 PM Lambda_Aurigae: use it as a usb-wifi bridge rather than programming it for anything else?
04:29 PM Emil: Lambda_Aurigae: yes, if only it was that simple
04:29 PM Emil: Also that'd be wasting its potential
04:30 PM Lambda_Aurigae: https://hackaday.com/2016/07/28/new-chip-alert-rtl8710-a-cheaper-esp8266-competitor/
04:31 PM Lambda_Aurigae: there's also the esp32
04:32 PM Lambda_Aurigae: the esp32 has bluetooth.
04:35 PM Lambda_Aurigae: that esp32 looks awesome.
04:36 PM Lambda_Aurigae: works with external QSPI(aka SQI) flash and sram chips.
04:36 PM Lambda_Aurigae: might have to grab one or three of those myself.
05:17 PM enh: hi
05:27 PM Lambda_Aurigae: iH
05:32 PM Lambda_Aurigae: Jartza, can you use 20MHz crystal and the internal pll to run it at 40MHz?
07:31 PM eszett: Hi! anyone up for a digikey group buy!
08:48 PM Ameisen: well, I've given myself a migraine trying to work on GCC a bit
08:50 PM Casper: Ameisen: then you still has not worked enought yet, continue for another 4 hours
08:51 PM Ameisen: Casper - trying to do a simple thing, too
08:51 PM Ameisen: It's just trying to figure out how expression objects and such get passed around... and then running into random crashes as a result
10:18 PM rue_ is now known as rue_mohr
11:36 PM e is now known as deadk