#avr Logs
Aug 02 2017
#avr Calendar
12:20 AM Casper: does anyone know if you can change the clock source/speed from in software?
12:44 AM rue_more: most chips I think you can
12:44 AM Casper: didn't saw anything about that in the tiny48
12:46 AM Casper: _ami_: prescaller yes, but that ain't the clock select
12:47 AM rue_more: oooh, sorry I thought you meant speed, not source
12:47 AM rue_more: no your boned
12:47 AM Casper: also, trying to figure out the best pin connections to this... kinda annoying
12:48 AM Casper: need to share the programming pins with something else...
12:48 AM rue_more: GROUND THEM ALL
12:48 AM rue_more: trust me, it makes routing traces easier
12:49 AM Casper: routing ain't an issue
12:49 AM Casper: my issue is: which program pin do I share with something else!
12:50 AM rue_more: :( they discontinued my credit card and the new one has a new number, gonna be a pain in the ASS!
12:50 AM rue_more: its gonna cut down my aliexpress spending an insane amount
12:50 AM rue_more: not being able to just remember everything off the top of my head
12:50 AM Casper: hehe
12:50 AM Casper: but you know there is a "save my credit card info" checkbox right?
12:51 AM rue_more: nOOOOOOT gonna do it!!!!
12:51 AM rue_more: the day after their database would get broken into
12:51 AM rue_more: id have to abandon all hope of remembering my card at all
12:54 AM Casper: man the power compagny is really slow
12:54 AM rue_more: the evil part is that I need to stop spending for a day while I switch over
12:55 AM rue_more: Casper, they didn't deliver enough electrons today?
12:55 AM Casper: like 45 minutes to go cut the power to an electrical power, emergency call from the firefighters...
12:55 AM rue_more: yea
01:32 AM Casper: does anyone know of a simple tx only ring buffer software uart lib? one that work on the tiny85? I don't have pin left for rx... and don't need rx anyway
01:33 AM rue_more: you bit banging the serial anyhow?
01:33 AM Casper: no choice with the tiny85
01:39 AM rue_bed: well, you can use the internal shift register, but still have to manually handle it all
01:40 AM rue_bed: why do you need to buffer up what your sending?
01:40 AM Haohmaru: non-blocking transmit ;P~
01:41 AM rue_bed: you have to manually send the bits, why not just use a blocking send?
01:41 AM rue_bed: its not like you have a uart thats gonna send the byte for ya
01:42 AM * Haohmaru actually doesn't know what is being discussed exactly
01:42 AM rue_bed: it would be funny to have an spi uart tho
01:42 AM Haohmaru: that smells doable
01:42 AM rue_bed: it is
01:42 AM Haohmaru: and i think i've thought about it
01:43 AM Casper: I actually do consider it...
01:43 AM rue_bed: slave tiny 13
01:43 AM Haohmaru: you'd have to transmit bytes all the time to get a constant clock
01:43 AM rue_bed: ?
01:44 AM Haohmaru: the downside would be that i think SPI reception is not as fool-proof as a USART receiver.. which typically samples a lot finer (16 times or so)
01:44 AM Casper: trying to figure out what I need to send, so I know if I can do blocking...
01:45 AM rue_bed: your blocked by the fact that you need to manually handle the bits
01:45 AM rue_bed: maybe with a low baud rate you could run a state machine to send bytes as reentrant code
01:45 AM rue_bed: on a timer
01:46 AM rue_bed: with an irq
01:49 AM Casper: maybe 120 bytes per line... hmm
01:49 AM Casper: 1200 bits... 9600... 0.125 sec... not too bad, could go blocking...
01:50 AM Casper: will most likelly implement the transmit itself interrupt based however, would be simpler...
01:51 AM rue_bed: want my awefull t13 tx code?
01:51 AM Haohmaru: transmition and writing is always simpler than reception and parsing
01:51 AM Haohmaru: hahaha
01:53 AM Casper: rue_bed: is it a ring buffer interrupt based? :D
01:53 AM rue_bed: no buffer
01:53 AM rue_bed: blocking send, seeing as the bit time is a delay lop
02:18 AM rue_bed: are you regulating something during the serial?
02:26 AM Casper: not really, want like a 1 second "per cycle" so still lots of time left
02:36 AM Thrashbarg_ is now known as Thrashbarg
02:41 AM Casper: I always forget... is it static or volatile for global variables?
02:44 AM * Haohmaru whispers util/atomic.h
03:04 AM Casper: found it
03:28 AM noHitW_work: all global variables are static, but you dont have to put the "static" keyword in them?
03:31 AM Casper: you do have to put a keyword...
03:32 AM Haohmaru: you can use extern on a global variable
03:33 AM Haohmaru: but if you declare the global variable "static" i think extern doesn't apply then
03:33 AM Haohmaru: i think that's the diff
03:34 AM noHitW_work: oh i was wrong, if you dont put any storage-class specifier in global varables it defaults to "extern"
03:35 AM noHitW_work: just like local varables defaults to "auto", you dont have to put it there
03:35 AM noHitW_work: https://www.tutorialspoint.com/cprogramming/c_storage_classes.htm
03:36 AM Haohmaru: well, in my head i remember it this way "you can access a global var from another .c/.cpp by declaring it extern" "if you declare a global variable as static - it can only be "seen" within this .c/.cpp file"
03:36 AM Haohmaru: [/my_own_interpretation]
03:37 AM noHitW_work: §6.2.2 Linkages of identifiers
03:37 AM noHitW_work: 3) If the declaration of a file scope identifier for an object or a function contains the storage-class specifier static, the identifier has internal linkage.
03:37 AM noHitW_work: 5) If the declaration of an identifier for a function has no storage-class specifier, its linkage is determined exactly as if it were declared with the storage-class specifier extern. If the declaration of an identifier for an object has file scope and no storage-class specifier, its linkage is external.
03:45 AM Casper: I'll figure that out tomorrow
03:49 AM noHitW_work: Haohmaru you need to use "extern" only in the files you use it, if you declare a variable in main.c you dont out the "extern" there, only to other files
03:49 AM noHitW_work: *put
03:50 AM Haohmaru: that's exactly what i said, didn't i?
03:50 AM noHitW_work: oh
03:50 AM Haohmaru: you declare it in some .c file, you also set its initial value there
03:51 AM Casper: yikes... my compilation environnement seems very borked...
03:51 AM Haohmaru: in other .c files you just declare it as extern, without assigning value.. this way the compiler knows it should look outside for that variable
03:51 AM Haohmaru: compiler or linker, i don't know
03:51 AM noHitW_work: yeah
03:51 AM Haohmaru: Casper, what's broken about it?
03:52 AM * Haohmaru has duct tape
03:53 AM Casper: /usr/avr/include/avr/pgmspace.h:1067:26: error: unknown type name ‘uint_farptr_t’; did you mean ‘uint_fast8_t’?
03:53 AM Casper: extern size_t strlen_PF (uint_farptr_t src) __ATTR_CONST__; /* program memory can't change */
03:53 AM Casper: :D
03:54 AM Haohmaru: what version of avrgcc is that?
03:54 AM Casper: 7.1.0
03:55 AM Haohmaru: o_O
03:55 AM Haohmaru: where did you stole that from?
03:55 AM Casper: nowhere
03:55 AM Casper: why?
03:55 AM Haohmaru: TELL ME
03:55 AM Haohmaru: or imma arrest you
03:55 AM Casper: me won't say!
03:56 AM Haohmaru: that's it, ur goin to jail, son
03:56 AM Haohmaru: i'll take your avrgcc as "evidence" ;P~
03:56 AM Casper: you need more than that
03:56 AM Haohmaru: i'm the law!
03:57 AM Casper: I will say that I obtained it legally, from a legal source, from an official way!
03:57 AM Haohmaru: where from?
04:00 AM Casper: portage
04:00 AM dunz0r: AVR-GCC is version 7.1.0 in Arch at least
04:00 AM Haohmaru: ah, 1337 h4x0r linux..
04:00 AM Haohmaru: why didn't i think about it
04:01 AM dunz0r: I think you can cross-compile it for windows with MingW
04:01 AM Haohmaru: i've tried building avrgcc myself..
04:01 AM Haohmaru: gcc 5.something
04:01 AM dunz0r: Tadaaa http://blog.zakkemble.co.uk/avr-gcc-builds/ :>
04:01 AM Haohmaru: but i hit a wall when building avrlibc
04:02 AM dunz0r: Haohmaru: http://blog.zakkemble.co.uk/avr-gcc-builds/ Some dude has done all the work for you already it seems :)
04:03 AM Haohmaru: i think i've seen that site
04:03 AM Haohmaru: but when i saw it back then, there was no win32 version
04:03 AM Haohmaru: while right now i'm on linux
04:04 AM dunz0r: Haohmaru: If you're on linux it's quite a lot easier though
04:06 AM Haohmaru: synaptic gives me avrgcc from atmel.. 4.9.2
04:06 AM Haohmaru: that's not bad
04:07 AM Casper: ow well, sleep time, I'll fix the issue tomorrow
04:07 AM Casper: nite
04:10 AM dunz0r: Haohmaru: You can probably find a debsrc-package with a newer version and build from that
04:11 AM Haohmaru: uhm.. but doesn't that come from atmel?
04:11 AM dunz0r: No, it doesn't. AVR-GCC 7.1.0 isn't from atmel
04:11 AM Haohmaru: cuz it's the same version as atmel has on their site (last time i checked)
04:11 AM Haohmaru: i meant the package in debian/synaptic
04:12 AM dunz0r: Debian is famous for having really old versions of stuff for ages :>
04:12 AM Haohmaru: thus i don't think there will be a newer version than that one
04:12 AM dunz0r: That's why I use it on all my servers, stability for the win
04:45 AM Emil: dunz0r: well
04:45 AM Emil: dunz0r: it leads to hacky things
04:45 AM Emil: like backports
04:45 AM Emil: and installing from repos
04:45 AM Emil: so you don't actually want too old software
04:46 AM dunz0r: Indeed, that's why I make sure to run the latest stable version of debian
04:46 AM dunz0r: The few times I need something that isn't included in the stable version I use backports or make my own packages.
04:49 AM Emil: "testing" as a name is pretty unfortunate
04:49 AM Haohmaru: i use debian9 nao
04:49 AM Emil: I should upgrade to 9
04:50 AM Haohmaru: i wiped my nasty kubuntu..
04:50 AM Haohmaru: NAASTEH
04:50 AM Emil: Excellent decision
04:50 AM Emil: :D
04:50 AM Haohmaru: yes, i feel so freed
04:51 AM dunz0r: Then you try to use modern software :((((
04:51 AM dunz0r: Debian's great for servers, drives me mad on workstations though
04:52 AM Haohmaru: well, it requires a configuring this and that
04:52 AM Haohmaru: what doesn't
04:53 AM dunz0r: Indeed it does. I've done it so many times I can basically do it with my eyes closed now though :)
04:53 AM Haohmaru: i also did it enough times that i know what search terms to use in duckduckgo to find the webpage with the answer
04:54 AM Haohmaru: like for disabling mouse acceleration ;P~
04:54 AM dunz0r: man xset?:)
04:55 AM dunz0r: I think it's xset mouse acceleration-multiplier/acceleration-division threshold
04:56 AM Haohmaru: nah.. the answer is creating a file somewhere
04:56 AM Haohmaru: in xorg.d or some such place that doesn't exist by default
04:56 AM Haohmaru: because debian9 uses "libinput"
04:58 AM dunz0r: Hehe, I'm old school, I do all those things in my .xinitrc :D
05:00 AM Haohmaru: https://wiki.archlinux.org/index.php/Mouse_acceleration
05:00 AM Haohmaru: the first example there is what i used
05:01 AM Haohmaru: ah no, not the first one.. it's down near the end.. the one that does "flat"
05:03 AM dunz0r: I've got my acceleration set really high, but I'm a trackball-user :)
05:03 AM Haohmaru: i need it completely off
05:03 AM Haohmaru: linear response for me
05:41 AM kdmal: hello, guys
05:51 AM noHitW_work: hi
07:44 AM robotron: why is it so silent in here
07:44 AM LeoNerd: Perhaps nobody has any questions right now
07:44 AM LeoNerd: That one aside ;)
07:44 AM Haohmaru: i will have a whole bag of questions.. after some time
07:45 AM * Haohmaru ties robotron with a rope
07:45 AM Haohmaru: >:)
07:45 AM north_: robotron has left the building
07:45 AM Haohmaru: no he ain't!
07:46 AM Haohmaru: uh uh!
07:46 AM robotron: I'mright here
08:53 AM Snert__ is now known as Snert
11:02 AM enh: hi
11:02 AM enh: good afternoon
11:03 AM enh: I wonder what is the difference between channels that start with # and channels that start with ##
11:04 AM Emil: enh: shitnode policu
11:04 AM cehteh: freenode policy
11:04 AM Emil: enh: # for real projects
11:04 AM Emil: ## for anything else
11:04 AM enh: What is real?
11:04 AM Emil: some # were grandfathered
11:04 AM dunz0r: enh: "official" channels
11:04 AM enh: ok
11:05 AM Emil: and they dont actively seek out channels that are not #project channels
11:05 AM enh: math, for example is unnoficial?
11:05 AM Emil: it's just general discussion
11:05 AM enh: ok
11:05 AM dunz0r: It's not an open source project, so it can't be unofficial.
11:05 AM enh: avr should be ##?
11:05 AM Emil: "yes"
11:05 AM Emil: but dont rock the boay
11:06 AM dunz0r: avr isn't an open source project either :>
11:06 AM Emil: boat*
11:06 AM Emil: The mods already know about us
11:06 AM polprog: what's the difference between # channels and ## channels on freenode?
11:06 AM enh: hum. not important. Just curiosity
11:06 AM Emil: polprog: I just explained :D
11:06 AM Emil: polprog: it's just a naming convention
11:06 AM polprog: kk
11:08 AM enh: IRC is very different from other social networks. We do not know lots of info about who the channel members are. Anonimity is normal here. In other networks it is not. What is the average age here, for example? No idea.
11:08 AM dunz0r: Who cares?
11:08 AM enh: I don't.
11:09 AM dunz0r: You could be intergalactic, genderless spacecowboys for all I care
11:09 AM polprog: after, say a month on a channel you start to know some people
11:09 AM enh: My mind asks for numbers, just that. Pure curiosity.
11:09 AM Emil: Of course statistics are interesting
11:10 AM polprog: enh: 42
11:10 AM enh: But it is good to keep chatting on the topic
11:10 AM polprog: btw why did you shorten your nick? :P
11:10 AM enh: I'm lazy.
11:10 AM enh: It was taking too long to type it. :)
11:10 AM Emil: short nicks are nice
11:10 AM polprog: not having nickserv auth on connect as auto cmd? :P
11:11 AM enh: Nope. Joking. My initials are enh. The rest were letters from the last name. Class conflict.
11:11 AM polprog: Emil: i second
11:11 AM enh: by the way, I'm 44.
11:12 AM enh: I wonder if young people use IRC.
11:12 AM polprog: no idea
11:12 AM Emil: polprog: lolo
11:12 AM Emil: polprog: :D
11:13 AM enh: These days all young people in my country seems to need a mobile phone with whatsapp, facebook, etc. They have no idea about IRC
11:13 AM Emil: so?
11:13 AM Emil: There are not many clients with good ux
11:13 AM polprog: hmm, i think my phone's broken then... i have twitter, reddit and SSH client
11:13 AM Emil: The one I'm using is really a ssh client
11:14 AM Emil: And the ux on irc is not that great
11:14 AM polprog: for android, so far the best one was AndChat.
11:14 AM Emil: enh: stop being a luddite ;)
11:14 AM polprog: now i just use SSH client since i have a server
11:14 AM enh: I'll google luddite
11:14 AM Emil: polprog: reddit is fun
11:14 AM Emil: best client
11:14 AM Emil: hmm
11:14 AM polprog: i mean Relay for reddit
11:14 AM polprog: :P
11:15 AM Emil: I wonder if I should so a telegram group for this channel
11:15 AM Emil: polprog: try rif
11:15 AM Emil: it's nice
11:15 AM enh: I should set up a IRC client on the raspberry pi, so it will be aliv full time.
11:15 AM enh: I use lime chat, on OSX
11:16 AM polprog: my username on irc is 31415 cos this setup was copied from a headless raspi :P
11:17 AM enh: Emil: Just learnt a new word
11:17 AM enh: Thanks
11:17 AM twnqx: are there LDOs that can go ~40V -> 3.3V?
11:18 AM twnqx: preferably 60V or more on input.
11:19 AM polprog: eww
11:19 AM polprog: use a buck
11:19 AM twnqx: nah, too expensive
11:19 AM polprog: it's more of a heater than LDO
11:19 AM cehteh: how much current?
11:19 AM cehteh: and why *L*DO .. you dont need low drop then :D
11:20 AM twnqx: 2mA
11:20 AM cehteh: for very little current (tens of mA) there are some linear regulators which do the job
11:20 AM twnqx: foudn one, but it's more than 1€, too
11:20 AM cehteh: of course .. bit special
11:21 AM twnqx: i just want to drive a LED as an indicator that power is there :X
11:21 AM polprog: for minimal cost i'd do a divider and an emitter follower/opamp
11:21 AM twnqx: i have no defined input
11:22 AM cehteh: when you know your currents you can some some pre regulation with zener/fet
11:22 AM twnqx: 12V-60V
11:22 AM cehteh: and only use a linear regulator for stabilizing the last few bits
11:22 AM polprog: a resistor and a zener will do
11:22 AM cehteh: yes
11:22 AM twnqx: mh
11:22 AM twnqx: that could do
11:22 AM twnqx: good idea, thanks
11:22 AM cehteh: but resistor + zehner + fet might be better
11:23 AM polprog: ^ or that
11:23 AM enh: I should learn electronics.
11:23 AM polprog: I should
11:24 AM enh: I know what I learnt on physics course. Far too little.
11:25 AM polprog: on physics we had only ohms law, voltmeter and ammeter
11:27 AM enh: More or less the same
11:28 AM Emil: polprog: you can change your username iirc
11:29 AM Emil: twnqx: >ldo
11:29 AM Emil: twnqx: fropping over 50 volts
11:29 AM Emil: How much current are you asking for, exactly? :D
11:29 AM Emil: oh 2mA
11:29 AM Emil: yeah there are
11:30 AM Emil: 100mW over an ldo is nothing
11:30 AM polprog: Emil: i know, let it stay
11:33 AM Emil: I would personally probably use some high vf diode with a more normal voltage regulator
11:33 AM Emil: Shouldn't be that expensive
11:34 AM polprog: if you connect a zener across it and power both through a resistor, say 30k, it's gonna be ok
11:43 AM enh: I heva a physics/electronics question: how many 12V batteries can be serially linked together before breakdown occurs, and on which battery breakdown will occur first?
11:44 AM enh: Maybe it should be rephrased with 9V batteries because they are physically smaller.
11:45 AM polprog: enh: you mean this? https://www.youtube.com/watch?v=8hwLHdBTQ7s
11:47 AM Emil: polprog: oh yeah zeners might work, too
11:48 AM polprog: :)
11:48 AM enh: more or less, polprog. I mean't dielectric breakdown
11:49 AM polprog: well, that depends on the dielectric
11:49 AM enh: air
11:49 AM polprog: i bet 3kV
11:49 AM polprog: ah
11:50 AM enh: there is a point where you cannot add more batteries
11:50 AM polprog: it's gonna be always 9V between the terminals of each battery
11:50 AM polprog: because HV is between the ends of the chain
11:51 AM enh: getting back to the positive terminal
11:51 AM enh: Sorry. Imagine a 2kV electron getting back to the positive terminal
11:53 AM polprog: hmm, this case...
11:53 AM enh: I have no ideai, but I believe this is not the same case as a flyback based PS
11:55 AM enh: Another ignorant question: if I join the positive terminal of a battery serial network with the negative one and suppose it can handle infinite current, will the voltage increase indefinetely?
11:56 AM polprog: hmm
11:56 AM Emil: ?
11:56 AM Emil: wat
11:56 AM polprog: afaik a perfect battery gives 0V and infinite current when shorted
11:56 AM Emil: polprog: lolno
11:56 AM polprog: lol i have no idea
11:57 AM Emil: polprog: a perfect battery would give infinite current and retain the short circuit current
11:57 AM polprog: s/battery/current source
11:57 AM Emil: s/short circuit current/nominal voltage
11:57 AM Emil: a perfect voltage source retains it's voltage
11:58 AM enh: So... What would happen to the circular battery network?
11:58 AM enh: Supposing perfect batteries
11:58 AM Emil: infinite current
11:58 AM enh: and what would the voltage be?
11:58 AM polprog: perfect battery = current or voltage source?
11:58 AM Emil: polprog: voltage source of course
11:59 AM enh: voltage
11:59 AM polprog: well we have that cleared up now
11:59 AM polprog: :)
12:00 PM Emil: enh: if you attach batteries in a circular fashion you can use Kirchhoffs laws to see what happens
12:00 PM enh: Does not a battery add voltage to the circuit?
12:00 PM Emil: matey
12:00 PM enh: Or... Does not a battery add energy to the electrons that flow to it?
12:00 PM Emil: Don't make it too hard for yourself :D
12:01 PM enh: If electrons already come with energy...
12:01 PM enh: I'm just adding food for the brains.
12:01 PM Emil: They don't add more than once
12:02 PM enh: Why? Electrons have counters?
12:02 PM polprog: https://www.ibiblio.org/kuphaldt/electricCircuits/DC/index.html
12:02 PM polprog: this was a nice book for me before i got The Art of Electronics
12:02 PM Emil: enh: if you want to ponder yourself, go ahead
12:02 PM enh: I'm reading Practical electronics for inventors. I hope it is a good book.
12:03 PM polprog: never read it
12:03 PM enh: amazon
12:03 PM polprog: somebody threw it in pdf on some uni site
12:04 PM polprog: BTW US navy released their electronics course handbook for self learners
12:04 PM enh: Emil: Physicists always go beyond...
12:04 PM polprog: called NEETS
12:04 PM enh: Do you have a lnk, polprog?
12:05 PM enh: Engineers stop in a prctical point. Physicists don't.
12:05 PM polprog: enh: sure
12:06 PM polprog: https://maritime.org/doc/neets/
12:06 PM enh: Thanks!
12:06 PM polprog: https://hackaday.com/2017/06/09/electronics-education-courtesy-of-the-us-navy/ contents here
12:06 PM enh: Thanks a lot.
12:07 PM polprog: np :)
12:07 PM enh: Sitesucker is a nice app.
12:08 PM polprog: i used down them all extension
12:08 PM polprog: good ol soft
12:09 PM enh: Oh, lord... It is downloading 350 books...
12:10 PM polprog: but there are only 24 chapters in neets
12:10 PM enh: It went one stet above
12:10 PM enh: Ops. 1500 books.
12:10 PM enh: I stopped it.
12:12 PM enh: I believe somebody should tell them not to put restriced documents on that networks...
12:12 PM polprog: whops :P
12:12 PM enh: For the sake of my safety, I'm wiping out evrything.
12:14 PM learath: uh... how restricted?
12:14 PM enh: https://maritime.org/doc/
12:15 PM enh: This restricted.
12:17 PM enh: It said restriced in the textbook covers, but the page says: Below are several full text declassified Navy documents
12:17 PM enh: So, in principle they can be downloaded.
12:26 PM enh: Amazing titles
12:26 PM enh: I wish I had time to read them.
12:27 PM learath: They are *supposed* to strike out the restricted markings when declassifying them.
12:28 PM enh: They should
12:29 PM enh: Well. All that technology should not be so outdated. I would never declassify such things
12:34 PM enh: polprog: Another picky question: Do you know why precession happens?
12:35 PM polprog: precession?
12:35 PM polprog: never heard of that
12:35 PM enh: hum... Rotating stuff under a torque
12:35 PM polprog: yeah i googled that
12:35 PM enh: pull them down, they turn to the side
12:36 PM polprog: didn't have that yet
12:36 PM enh: Which year are you in?
12:36 PM polprog: hmm, that would be 5th starting now
12:37 PM polprog: different school system keep in mind
12:37 PM enh: major in physics?
12:37 PM polprog: no, highschool
12:38 PM enh: you will see precession soon, I believe
12:38 PM enh: When you see, ask your teacher how it happens
12:39 PM enh: or why.
12:39 PM polprog: ill ask, but now, holidays :P
12:39 PM enh: soo good, holidays..
12:40 PM polprog: i will be attending some local "programming in electronics" course in the last 2 weeks of tis month
12:40 PM enh: seems nice
12:40 PM polprog: it's eiter gonna be something nice or arduino basics
12:40 PM enh: if too basic you can ask advanced questions
12:41 PM polprog: anywa, free messing around,
12:41 PM polprog: nah, i dont wanna be that person :P
12:41 PM enh: suck knowledge from everywhere
12:41 PM enh: good knowledge sources are very rare
12:53 PM Jartza: evening
12:54 PM polprog: o/
12:54 PM APic: Heya
12:54 PM APic: \o
12:54 PM polprog: evening
12:55 PM enh: hi
12:57 PM kdmal: guys, are have ideas how to wake atmegaX8 by analog comparator from power down/save ?
12:58 PM polprog: comparator interrupt
12:58 PM kdmal: ac interrupt only in idle mode
01:08 PM Lambda_Aurigae: will pin change interrupt wake it up?
01:09 PM LeoNerd: There's usually a table somewhere in the power management section of the datasheet, showing which kinds of interrupts can wake the chip up from which kinds of sleep
01:09 PM kdmal: voltage level only ~0.5v
01:11 PM LeoNerd: Weird. It seems that the anacomp doesn't count as asynchronous wakeup. Perhaps it uses a clocked synchroniser on the interrupt line
01:12 PM Lambda_Aurigae: probably screwed then.
01:55 PM enh: Does anybody here come from outer space? Just for statistical purposes.
01:56 PM enh: :)
02:33 PM polprog: im gonna send him to outer space, find another race
02:39 PM Lambda_Aurigae: well, I've always claimed to not be human as far back as I can remember and my parents say I've claimed such since I could talk.
02:58 PM enh: That makes two of us.
02:59 PM enh: If I just could remember how to build an antigravity device...
03:03 PM antto: with some antigravitons, duh
03:03 PM antto: and carbon fibres!
03:03 PM antto: cuz it's 2017
03:06 PM day is now known as daey
03:18 PM Casper: yay weird compilation issue is gone :D
03:19 PM Casper: gentoo changed the location of the avr/include ....
03:19 PM Tom_L: lovely
03:19 PM Casper: yeah
03:19 PM Casper: from /usr/local/avr/avr/include to /usr/avr/include
03:20 PM Casper: /usr/avr/include/avr/pgmspace.h:1456:25: error: unknown type name ‘uint_farptr_t’; did you mean ‘uint_fast8_t’? <=== that :D
03:28 PM enh: So easy to find those errors...
03:28 PM enh: I usually spend days
03:29 PM APic: 😎
03:35 PM Casper: I googled, and found a thread that hinted to a wrong include
03:36 PM Casper: now... to fix the rest of the code
03:36 PM Casper: I kinda wrote some code without a great plan on what I'm doing, kinda more like a scratch pad... now I'm filling the blanks
03:37 PM Casper: now... I get more sane error like: main.c:62:42: error: ‘usart_putchar’ undeclared here (not in a function); did you mean ‘putchar’?
03:38 PM enh: usart _putchar seems like not standard library code. You wrote it?
03:38 PM Casper: yes, just had to move the line from top to under the function...
03:38 PM JanC is now known as Guest87139
03:38 PM JanC_ is now known as JanC
03:39 PM enh: I've got a usart class for atmega328p. If you wish, it is working fine
03:41 PM Casper: need software uart, tx only
03:41 PM Casper: not enought pins for rx
03:43 PM Casper: better and better... just...
03:44 PM Casper: 69 lines of errors
05:28 PM Casper: is there anyone here good at weird bug hunting?
05:28 PM Casper: gcc throw many errors that I can't figure out :/
05:29 PM Casper: https://paste.pound-python.org/show/IYmc9YvZASqIpXMFVBME/ <=== code and errors ===> https://paste.pound-python.org/show/U4iVpQA8GObLVr5b3Rnh/
05:47 PM NoHitWonder__: casper youre missing a "{"
05:47 PM NoHitWonder__: line 82
06:16 PM Casper: ...
06:17 PM Casper: NoHitWonder__: thanks... I was looking everywhere :D
06:17 PM Emil: Damn it feels filthy making pcbs like I'm making now
06:17 PM Casper: still not compiling, but that's expected...
06:18 PM Emil: https://emil.fi/jako/kuvat/2017-08-03_01-48-18_7w3kJSld.png
06:18 PM Emil: https://emil.fi/jako/kuvat/2017-08-03_01-48-50_b8v27Avs.png
06:18 PM Emil: Single side boards with so_many_manually_drillable_holes
06:19 PM Emil: Out of huge throughhole parts
06:19 PM Emil: Like
06:19 PM Emil: _THIS_ is what I want!
06:19 PM Emil: https://emil.fi/jako/kuvat/2017-08-03_01-49-57_DR64fFhb.png
06:20 PM Emil: But it takes like 40 days to get those pcbs from oshpark :D
06:20 PM Casper: oh great... the error is a path error.... fun...
06:21 PM Emil: Well, I'll try the diy solder mask tomorrow
06:21 PM Casper: yay!
06:22 PM Casper: I need to see if I can make a board again soon, I have some etchan and some acetates, but no more program....
06:24 PM Lambda_Aurigae: Emil, at least you aren't doing 4 layer boards the way I used to do them.
06:24 PM Emil: But Filthy, I say! FILTHY!
06:24 PM Emil: Lambda_Aurigae: oh you mean manually stacking?
06:24 PM Lambda_Aurigae: two single sided boards and a two sided board...all etched and stacked.
06:24 PM Lambda_Aurigae: yeah
06:24 PM Emil: nice
06:25 PM Emil: That's one extra thicc board
06:25 PM Lambda_Aurigae: etch and stack and drill.
06:25 PM Emil: https://emil.fi/extrathicc
06:25 PM Lambda_Aurigae: yeah...used wire wrap sockets with the long pins to go all the way through.
06:26 PM Emil: But you get blind and buried vias with that technique
06:26 PM Lambda_Aurigae: and a dremel drill press to drill.
06:26 PM Emil: so it has it's upsides
06:26 PM Lambda_Aurigae: it is possible to do, yes.
06:26 PM Emil: its*
06:26 PM Lambda_Aurigae: had plenty of vias...
06:26 PM Lambda_Aurigae: took some serious preplanning to get everything connected up.
06:27 PM Lambda_Aurigae: 3 layer was fairly easy to do though.
06:27 PM Emil: I wonder if I could cut fr4 with a laser
06:28 PM xentrac: no
06:28 PM Emil: I could have nice clean edges
06:28 PM Emil: xentrac: whatcha mean no?
06:28 PM xentrac: I mean, in theory, sure, you can cut anything with a laser
06:28 PM xentrac: but cutting FR4 or other glass composites with a laser is so difficult that nobody does it
06:28 PM Emil: hmmm
06:28 PM Emil: Challenge accepted
06:29 PM Emil: Does it produce toxic fumes?
06:29 PM xentrac: sure, but that's not the problem
06:29 PM Emil: What is?
06:29 PM xentrac: the glass
06:29 PM Emil: Hmm
06:29 PM Emil: Well I'll try anyways
06:30 PM xentrac: you probably want to be thinking about the kinds of lasers people use for cutting steel and aluminum — room-sized fiber lasers of several kilowatts — rather than the kinds of lasers people use for cutting wood, cardboard, and plastic (30 to 100 watts)
06:33 PM xentrac: there's a paper reporting some success with a 200W laser at http://www.industrial-lasers.com/articles/print/volume-26/issue-5/features/laser-machining-fiber-reinforced-composites.html
06:40 PM Emil: fuq
06:41 PM xentrac: it looks like pretty good success really
06:41 PM xentrac: it's not a very good paper. more like an advertisement or something really
06:48 PM Casper: grrr my software uart don't work...
06:49 PM tpw_rules: i think they call those "white papers"
06:58 PM Lambda_Aurigae: the glass fibers in fr4 will spread the heat so you need LOTS.
07:01 PM xentrac: also, it's a lot harder to boil glass than to boil plastic
07:01 PM xentrac: I mean it requires a higher temperature
07:01 PM xentrac: tpw_rules: I don't think this quite rises to the level of a white paper
08:08 PM Casper: it's ALIVVE! ... damn I'm so rusty...
11:55 PM day_ is now known as daey