#avr Logs

Nov 16 2017

12:00 AM day__ is now known as daey
09:09 AM amigojapan_bnc: Emil: this is my attiny85 computer https://www.youtube.com/watch?v=kfGwH2Lkdq0
09:14 AM Emil: Cool!
09:19 AM amigojapan_bnc: ty Emil
10:43 AM _abc_: Hello. Please suggest a way to instrument avr atmega and not only interrupt load. I.e. how much time / cpu load is spent in isr's. This is a thorny question. The code should be C only and lightweight, added to existing code. Any ideas, links, etc, ...
10:44 AM _abc_: ... welcome.
10:45 AM _abc_: The bad old way to do it is to run a realtime accumulator in the main thread and see how it deviates from real time. I.e. count 10 msec 'loop' ticks without using interrupts and see how much the result deviates from realtime after a while.
11:01 AM Emil: eh?
11:02 AM Emil: Why would you need such a thing
11:02 AM Emil: anycase
11:02 AM Emil: I'd start a timer at the beginning of every ISR and then add that to an accumulator at the end
11:02 AM Emil: Simple and lightweight
11:03 AM Emil: Then we should have another timer/wdt that checks the counter value every second or so
11:03 AM Emil: and clears it
11:32 AM _abc_: avr does not have an 'instruction counter' and the isr entry and exit code is significant and should be counted too.
12:16 PM Emil: _abc_ you cant win
12:32 PM thardin: I just look at the asm and count cycles
12:33 PM thardin: add what is it, 6 cycles each for entering and exiting the isr?
02:00 PM polprog_ is now known as polprog
02:41 PM NoHitWonder: i made this today https://s33.postimg.org/m5t7vx5hb/input_coord.png
02:44 PM polprog: cool
02:44 PM NoHitWonder: im starting to get bored with those default android ui components, i wanna learn how to roll my own
02:45 PM polprog: yeah the default components dont look very nice
02:45 PM polprog: boring i'd say
02:45 PM polprog: the radio buttons stand out ;)
02:45 PM polprog: in your pic
02:46 PM NoHitWonder: yeah
03:36 PM NoHitWonder: i wanna get a video camera
03:43 PM polprog: okay, im about to switch from ufw to iptables
03:43 PM polprog: my conncetion may be innterrupted
03:43 PM polprog: keep your fingers crossed ;)
03:47 PM thardin: converting coordinates for osm and aprs is a continuing pain
03:53 PM polprog: on the other hand ufw is not that bad, ill leave it
04:42 PM Emil: ufw?
04:42 PM Emil: it's super simple to use imho
04:43 PM Emil: if you mean the uncomplicated firewall
07:07 PM enh: hi
07:07 PM enh: good night!
07:11 PM enh: Ameisen? Is it possible to move a formicarium from somewhere in your house to another (portable) place?
07:13 PM Casper: enh: depend on the construction I'ld say
07:13 PM Casper: and may pose challenge in the transport...
07:14 PM enh: My question is more how to convince the ants to move to their new (mobile) home...
07:14 PM Casper: I don't know much about it, but look up ant canada on youtube or something like that...
07:15 PM enh: I believe Ameisen is an expert
07:15 PM enh: I'll look at those places too. Thanks, Casper
07:17 PM enh: I could set the focus of the laser head, and connect it to the PWM output of the CNC controller. Now I can cut 3mm thick stuff with 30% power and many passes.
07:18 PM enh: If you want I can send a video, Casper
07:18 PM enh: There are very interesting possibilities ahead.
07:18 PM Casper: nahh I'm debating between staying awake, or take a nap...
07:19 PM enh: difficult choice indeed
07:19 PM enh: 2 am there?
07:20 PM enh: I'm trying to draft a set of control laws for a tricopter. https://pastebin.com/RBBkfDp0
07:27 PM Casper: 20:28
07:27 PM Casper: at that time, I'll wait half an hour and get the bath
11:58 PM day__ is now known as daey