#avr | Logs for 2012-11-08

Back
[00:13:38] <rue_house> if your new to avrs, I suggest printing the index that tells you what pages the registers are explained on
[00:13:43] <rue_house> its about 1 or 2 pages
[02:16:40] <lorddelta> Hi, what is the difference between extensions and plugins in avr studio 6, and how do you write one?
[02:16:51] <lorddelta> i.e. where are the reference documents?
[02:44:34] <OndraSter_> lorddelta, visual studio plugins :)
[02:47:56] <lorddelta> OndraSter: great, so does the AVR Simulator have an API I can compile against? That's kinda what I want to do; I don't know if you're familiar with HAPSIM, but it was a great tool for AVR Studio 4 that let you connect to its Simulator to simulate periperhals for debugging purposes.
[03:01:11] <OndraSter_> I am familiar with hapsim
[03:01:19] <OndraSter_> not sure about the avr simulator in as6
[03:32:02] <lorddelta1> OndraSter_: AFAIK it appears the guy hacked the dll memory locations for various version of AvrSim per Studio...
[06:21:50] <RikusW> http://arstechnica.com/security/2012/11/crypto-keys-stolen-from-virtual-machine/
[06:32:04] <RikusW> http://www.erikyyy.de/tempest/
[06:32:14] <RikusW> CRT AM modulator
[06:32:39] <Steffanx> RikusW, read the comments as well
[06:32:45] <Steffanx> -,
[07:00:52] <OndraSter> http://keygenjukebox.com
[07:03:05] <Steffanx> chiptune-ish songs :D
[07:05:37] <Steffanx> http://heeeeeeeey.com/ is better OndraSter :P
[08:12:38] <nickoe> what do I need to be able to use the preprocessor macros __TIME__ and __DATE__ in avrgcc?
[08:22:53] <nickoe> maybe I just have to spell it with two endning underscores :)
[09:02:09] <rue_house> a system with an RTC helps
[11:15:36] <inkjetunito> Tom_itx: yoyo
[14:37:54] <inkjetunito> if you were an evil designer and could choose from 4 colors: black, green, white and orange for +5V, GND and two DATA pins, what would you choose?
[14:38:23] <inkjetunito> yes, i'm waiting for John Fluke to deliver me an assistant
[14:38:54] <asteve> inkjetunito: I'd have five wires that were +5
[14:39:10] <inkjetunito> :o
[14:57:03] <OndraSter_> inkjetunito, I would first measure each of those colours
[14:57:15] <OndraSter_> and find two which show 5V or -5V CONSTANTLY
[14:57:29] <OndraSter_> from those figure out the polarity of the supply
[14:57:32] <OndraSter_> and the rest will be the data :P
[14:58:05] <inkjetunito> sadly, for the 1970s 1980s chinese multimeter i have, nearly everything is constant
[14:58:30] <inkjetunito> anyway, i came in to conclusion that the usb port is ok and it's just the mouse that died somehow
[14:58:58] <inkjetunito> we live in a time where rats live longer than computer mouses :(
[15:03:47] <inkjetunito> Tom_itx: *poke*
[15:06:15] <Tom_itx> ?
[15:27:11] <megal0maniac> I'm back :)
[16:01:59] <DagoRed> Tom_itx: How is your stock of programmers?
[16:22:38] <os1r1s> Are there any strategies for laying out two sided boards and via placement?
[16:29:59] <amee2woof> easiest one is route everything at right angles, then place all the horizontal traces on one layer, and the vertical ones on the other
[16:30:21] <amee2woof> easy to do but quickly result in inefficient routing on larger boards
[16:44:28] <os1r1s> amee2woof: I have a pretty simple board that I'm trying to compact and turn two layer. Its this boards ... http://mounicou.com/i2cpcb/i2cpcb3sch.png
[16:45:11] <os1r1s> amee2woof: I was going to put connectors on one side (changed to jst) and the components on the other. I am looking to shrink it down primarily.
[16:45:48] <amee2woof> hmm how much smaller would it get by just using one side as a ground plane?
[16:46:55] <amee2woof> i'm tempted to say turning the 5 pin header by 90 degrees is a routing advantage
[16:46:55] <os1r1s> amee2woof: Not sure, but I could give it a shot.
[16:47:02] <amee2woof> 6* pin header
[16:47:15] <amee2woof> with R5 ending up between the header and the IC
[16:47:42] <os1r1s> That would make it more narrow for sure.
[16:47:46] <amee2woof> can you have pads on both sides?
[16:48:14] <amee2woof> you could put the three resistors at the top on the other side of the board behind the IC
[16:48:27] <os1r1s> I've etched the board manually, but I'm going to send it off to be made. Before I did so, I wanted to take advantage of the pro manufacturing.
[16:48:30] <OndraSter> why double sided if this works?
[16:48:36] <OndraSter> oh
[16:48:43] <OndraSter> double sided but parts only on one side
[16:49:00] <OndraSter> avoid having parts on both sides, it is nasty
[16:49:15] <amee2woof> manufs often can do pads on both sides, they just won't populate it for you unless you pay extra
[16:49:23] <OndraSter> yes
[16:49:26] <OndraSter> it is paid extra
[16:49:27] <OndraSter> a lot
[16:49:31] <OndraSter> it has got to go through the machine again
[16:49:43] <amee2woof> if you solder yourself it might be manageable
[16:49:49] <os1r1s> OndraSter: This is to control rgb leds on my 3d printers. I'l also have a very similar one with a mosfet. Because of that, size is a premium.
[16:50:50] <OndraSter> 0603 resistors don't make almost any difference really
[16:50:53] <OndraSter> I fit a LOT of stuff on single side
[16:50:54] <amee2woof> if you have parallel traces, you can put one on the back and route them on top of each other
[16:51:00] <os1r1s> OndraSter: Why should I avoid parts on both sides?
[16:51:13] <amee2woof> pain in the ass to solder?
[16:51:35] <amee2woof> how many of these boards are you using in that printer?
[16:51:55] <OndraSter> http://clip2net.com/page/m12051/36326467
[16:52:09] <os1r1s> amee2woof: Probably two for fans, two - four for rgb leds
[16:52:31] <amee2woof> considered putting all of that circuitry on a single board?
[16:52:50] <amee2woof> that might result in a net advantage on board area
[16:53:08] <RLa> os1r1s, building a fan controller?
[16:53:33] <os1r1s> amee2woof: I thought about for the fan board, but not for the rgb board.
[16:54:05] <os1r1s> amee2woof: If I did more than one rgb led on there, I would probably use an led driver instead.
[16:54:18] <os1r1s> RLa: Part of this, ye
[16:54:20] <os1r1s> s
[16:54:27] <OndraSter> os1r1s, the pain with double loaded boards is that if you want to reflow it in a reflow oven you might be having issues with double loaded boards
[16:54:47] <os1r1s> OndraSter: Ahh, ok. I was just gonna hit this by hand.
[16:54:48] <OndraSter> if you are hand soldering them then meh
[16:55:00] <OndraSter> but also you have to make sure that there is nothing touching the other side
[16:55:03] <OndraSter> no metallic anything
[16:55:08] <amee2woof> modular and distributed design always sacrifices board real estate efficiency ;)
[16:55:09] <OndraSter> also that the other side is not flat :P
[16:55:44] <amee2woof> a single board with two fan drivers and four LED drivers could easily work with the same total board area as two fan boards and only two led boards
[16:56:23] <os1r1s> amee2woof: That's a good point. I'll have to reconsider that.
[16:56:50] <os1r1s> The etched one came out nicely, other than my soldering issues ... http://mounicou.com/i2cpcb/i2cpcb3wled.jpg
[16:57:18] <amee2woof> and ground usually goes almost everywhere so i'd start routing just ground on the back side and optimizing the layout
[16:57:31] <RLa> os1r1s, what are you using for changing the speed, pwm?
[16:57:32] <amee2woof> then see which other traces create routing hazards and can go on the back side too
[16:57:54] <os1r1s> amee2woof: That's a great idea That and JST connectors would make things simpler.
[16:58:02] <os1r1s> RLa: Yep
[16:58:42] <os1r1s> From the attiny85
[16:59:00] <RLa> os1r1s, what frequency?
[16:59:13] <RLa> and you use timers to generate the waveform?
[16:59:41] <os1r1s> After the print is done, I want to turn it on to cool the print bed
[17:00:07] <amee2woof> also, if you're using TWI/I2C to control the MCUs there, routing these on the back could help too
[17:00:20] <os1r1s> But there are other fans that are interesting to control via the electronic such as the one that cools the stepper drivers
[17:00:32] <amee2woof> these are run from MCU to MCU in a similar fashion to the ground connection
[17:00:35] <os1r1s> amee2woof: I am using i2c
[17:02:42] <amee2woof> things that cross logical boundaries between functional blocks in the circuit tend to be a routing hazard... same way as a railway line on a roadmap
[17:02:50] <os1r1s> amee2woof: I'll give the ground plane a shot and consider consolidating the functions on one board. Thanks all for your help.
[17:03:01] <amee2woof> you're welcome :)
[17:03:24] <os1r1s> RLa: I'm not sure what frequency or the waveform to be honest.
[17:08:41] <RLa> oh
[17:09:05] <amee2woof> i'd hope the waveform is a square wave o.O
[17:09:19] <OndraSter> frequency - the highest one :)
[17:09:20] <RLa> anyway, today i worked on dc amp to convert my pwm to dc
[17:09:44] <RLa> i figured i can't generate so many pwm channels with high freq
[17:09:53] <os1r1s> RLa: Picture if you are interested ... http://mounicou.com/venom/prusa2ps1.jpg
[17:09:55] <OndraSter> eh?
[17:09:58] <OndraSter> RLa, hardware PWM
[17:10:00] <OndraSter> :P
[17:10:08] <RLa> OndraSter, i'm sick of that :)
[17:10:15] <OndraSter> lol why?
[17:10:16] <RLa> too complex circuit
[17:10:20] <OndraSter> wut
[17:10:23] <OndraSter> it is built right into the chip
[17:10:28] <OndraSter> or tiny13 has none?
[17:10:35] <OndraSter> it should have 2 PWM outputs at least
[17:10:35] <RLa> not with 8 channels :/
[17:10:39] <OndraSter> oh
[17:10:44] <OndraSter> what chip is that?
[17:10:46] <os1r1s> RLa: tiny85 has 6
[17:11:07] <RLa> os1r1s, a cnc machine?
[17:11:36] <os1r1s> RLa: That is one of my 3d printers. I also have several cnc machines
[17:16:30] <RLa> http://imgur.com/z5QRB <- this is my dc amp, works down to bottom rail
[17:20:42] <RLa> q3 works as the feedback/gain element, others are followers
[17:21:16] <RLa> rail-to-rail input opamps use something similar
[17:39:35] <OndraSter> tomorrow: syntax highliting =)
[18:36:13] <NinoScript> Is anybody here using AVR-Eclipse? I'm having problems using up avr-gdb, it says: "Failed to execute MI command: -exec-run".
[22:51:19] * rue_house sets some crickets loose for effect
[23:05:32] <lorddelta1> rue_house: you new to irc?
[23:06:27] <lorddelta1> You need either an extremely chatty/close-knit community or around 1k+ irc chat members for any sort of real noise
[23:07:14] <hackvana> rue_house is one of our regulars
[23:08:22] <lorddelta1> hackvana: ah apologies then
[23:09:04] <hackvana> lorddelta1: It's cool. Welcome to #avr :-)
[23:09:46] <lorddelta1> thnx
[23:09:52] <hackvana> Some days #avr will have forty pages or more of chat. Some days, nothing.
[23:10:08] <hackvana> It's always a good S/N ratio though
[23:12:03] <hackvana> Going now, back later.
[23:16:44] <rue_house> lorddelta1, no
[23:17:06] <rue_house> hahaha I'm a regular
[23:18:07] <rue_mohr> heh, regular
[23:20:27] <ferdna> what are the advantages disadvantages of using negative voltage..
[23:20:39] <ferdna> lets say -5v
[23:20:57] <julmuri> hi, i have static local class instance and iam getting undefined reference to `atexit'
[23:21:07] <julmuri> can i just define it and leave it empty as a workaround
[23:21:11] <rue_mohr> usually used with positive rails for analog stuff
[23:21:46] <julmuri> as we are not cleaning it up ever anyways
[23:21:50] <rue_mohr> julmae, arduino ,code??
[23:22:53] <julmuri> hmm programming 664p with avr dragon. i can make sample code if necsesary
[23:28:30] <julmuri> http://codepad.org/ShvjMiPr
[23:28:41] <julmuri> something along those lines
[23:29:14] <rue_mohr> haahah c++ on a microcontroller, I'm out.
[23:29:30] <julmuri> haha
[23:50:58] <Xark> julmuri: Make sure you are using -fno-exceptions -fno-rtti (I suspect the atexit is a "throw" for if the constructor fails).