Back
[00:19:12] <hackvana> rue_house: :-)
[00:26:38] <abcminiuser> SWEET JESUS I AM ON FIRE
[00:27:10] <hackvana> Old news
[00:27:42] <hackvana> I's a gonna call you "lp0"
[00:27:50] <hackvana> http://en.wikipedia.org/wiki/Lp0_on_fire
[00:37:45] <abcminiuser> Ja, would be a great name for Aus right now
[00:40:40] <hackvana> abcminiuser: When's the next MEL freaks meet?
[00:41:53] <abcminiuser> Hrm, not sure - I missed the last one
[00:42:08] <abcminiuser> Ironically, I suggested the meetup, but the consesus moved it to a date that I couldn't attend...
[00:42:17] <abcminiuser> By all means throw up a new request on the thread
[02:50:31] <amee2k> rue_house: whut? o.o
[02:51:48] <amee2k> N2TOH: its on season break because i've got my hands full with university stuff this winter, and i'm looking for a co-host who wants to stick around for more than a few episodes
[02:52:34] <amee2k> N2TOH: frankly i have an FM modulator but i don't terribly care because i don't see who would listen locally, and i'd risk being triangulated
[03:06:21] <jerkey> what is an abcmini?
[03:12:22] <BusError_> rue_bed, are you awake?
[03:16:41] <N2TOH> amee2k, what about SW band? yes triangulation is a factor, but only if you manage to piss off and upset people...
[03:17:33] <amee2k> shortwave needs a big antenna if i want to pump a practically useful amount of power into it
[03:17:54] <amee2k> and i'm not terribly keen on blowing 50USD on electricity for the SW PA every weekend
[03:18:37] <N2TOH> lets not forget that enforcement is all about pissing people off, happy people == no government agents kicking in your door :)
[03:19:52] <jerkey> amee2k that number does not sound realistic to me. electricity here is about US$0.15 per killowatt/hour ...
[03:20:11] <N2TOH> SW should be better then FM power wise, HELL you might even gather a following even following US part 15 regs...
[03:20:27] <jerkey> also the government always politely asks you to stop before they kick in the door
[03:21:29] <N2TOH> amee2k, if you could run a 100mW station how much would that cost you?
[03:21:43] <amee2k> how far would it get? :>
[03:22:29] <N2TOH> the entire world with good conditions!
[03:22:34] <amee2k> jerkey: converted to USD, .38USD/kWh around here
[03:22:54] <amee2k> and with realistic conditions and an average broadcast receiver?
[03:23:59] <N2TOH> average broadcast receiver is one thing, VSB is another...
[03:24:46] <N2TOH> correction, DRM not VSB...
[03:25:07] <N2TOH> IE Digital Radio Mondial sp?
[03:26:13] <N2TOH> in other words in the shortwave bands it does not take more then fart power to reach the world...
[03:27:28] <amee2k> http://vm197.cugnet.net/~amee2k/sandbox2/comp.png << anyone think that approach is *slightly* too ghetto for an indicator LED which one of two redundant power inputs is being used?
[03:28:22] * N2TOH I'll let you know tomorrow
[03:28:46] <amee2k> also turns off the bias current for the current sink if the output regulator drops out
[03:47:58] <Fleck> rue_house: well, I'm ... Fleck! :D
[03:50:18] * amee2k idly cleans up the Fleck :3
[03:52:04] <Fleck> thank you :D now I am Fleck without fleck :D
[03:52:17] <RikusW> Fleckless :-P
[03:52:23] <Fleck> yeah :D
[03:52:36] <RikusW> s/Fl/R :-D
[03:52:53] <Fleck> anyway - morning all! :)
[03:53:17] <RikusW> morning
[04:17:21] * amee2k innocently wags tail :3
[07:24:22] <volbus> Hi everybody! Does anyone know what is the input common mode voltage range for the ana-comp for an ATtiny-mcu?
[07:25:28] <volbus> I couldn't find this in datasheets, and I want to compare voltages close to Vcc
[07:26:49] <Tom_itx> it will be there
[07:28:24] <volbus> Tom_itx: was that an answer for me? in that case, what do you mean?
[07:28:54] <Tom_itx> the data sheet has all the information on a chip that anyone else could give you
[07:31:26] <volbus> well, I couldn't find this information in the datasheets, maybe I'm a bit blind - I'll check that again!
[07:31:47] <volbus> I just thought that someone may have expirience with this...
[07:34:01] <Tom_itx> P150 of the attiny13 has it
[07:34:06] <Tom_itx> chart
[07:34:33] <Tom_itx> .007 @ 5v
[07:45:59] <volbus> Tom_itx: was that the page no. 150 in the datasheet? I still couldn't find it
[07:46:16] <volbus> I just checked the attiny13 datasheet
[08:50:11] <dunz0r> Damn it, missed the activity.
[08:50:45] <PoppaVic> the gas-bloat will soon start
[09:10:36] <KaaK_> i've got a device whose signature will change between the expected 0x1E950F (atmega328p) and 0x000102. Anyone experience something similar before?
[09:10:56] <rue_house> KaaK_, what kinda programmer?
[09:11:12] <amee2k> stuff like that can happen if the ISP connection is faulty
[09:11:21] <KaaK_> avrdude + avrispmkII
[09:11:34] <amee2k> inadequate transfer rate and/or stuff on the ISP signal lines interfering
[09:12:42] <KaaK_> the ISP pins are shared on the board, but there is currently nothing that could screwing up the signaling, expect perhaps the traces themselves, they are rather long.
[09:16:19] <KaaK_> It is also very odd. It will often check the signature just fine, and continue to program, but I can't get any verification that the code is indeed running on the device.
[09:16:53] <KaaK_> I've put together a number of devices with atmega328ps, and I rather at a loss
[09:17:45] <amee2k> you could try slowing down the ISP transfer rate and see if that improves reliability
[09:18:54] <KaaK_> I've varied the bitrate between 1, 10, and 100, and it doesn't appear to help when it gets into the loop of reporting 0x000102 as the signature
[09:19:16] <amee2k> hmm, i see
[09:20:07] <amee2k> i'd try swapping the controller on the board. if that helps, the MCU was failing. otherwise, the issue is largely narrowed down to the board design
[09:20:15] <amee2k> also, lecture is over... bbiab
[09:20:29] <KaaK_> amee2k: thanks for the suggestions
[09:42:10] <phinxy> i got a 2.4ghz transceiver and a library for it. the "Test communication" usage code first does a 'send' and then a 'get_data'. can i assume it sends the data to itself?
[10:40:06] <awozniak> rue_house: not a specific line, but useful on AVR when doing integer maths; to avoid roundoff errors, save all your division operations until the end of your calculations
[10:42:22] <mike_papa> hello. I have a code for ADC convertion that worked perfectly stable. After an hour I didn't get any fluctuation using 9 bit precision. Even by bit. Now I want to add internal temperature and AVCC mesurement. Problem is, that if I add code changing ADMUX (Vref and Vin), results float like crazy. Datasheet says, that 1.1V needs stabilizing after it is turned off. But I never turn ADC off, so it
[10:42:22] <mike_papa> should be stable. Do I need delay after changing Vref and Vin? If yes, how long it should be? I do one "empty" conversion and discard it, after change.
[10:43:20] <mike_papa> I'm using ATmega168PA.
[10:44:34] <awozniak> mike_papa: tried throwing away a few hundred? ADC cycle time is usually stupidly fast.
[10:45:55] <mike_papa> awozniak: few hundred? Ok. I'll try. I'd like to keep AVCC mesurement as close as possible to sensor mesurement, as I need it to calculate output.
[10:46:35] <awozniak> mike_papa: Just suggesting you try an extreme, then ratchet down until it becomes more unstable than you want.
[10:47:34] <mike_papa> awozniak: I'll do that. Anyway you're saying that ADC needs to settle down after changing Vref and/or Vin, right?
[10:47:53] <awozniak> I *think* vref is the issue, but yeah.
[10:52:08] <mike_papa> awozniak: Ok. Thanks. Throwing away 300 helped. I'll experiment with lower numbers now. Actually giving it second thought... vref has cap on it. That may mess up things for first couple ms.
[10:52:44] <awozniak> mike_papa: that'll do it. there's probably a small amount of additional internal capacitance.
[10:54:17] <awozniak> mike_papa: you might be able to calculate it, if you know the two vref values and the cap value.
[10:57:43] <mike_papa> awozniak: I'm not that good in theory. Vref is more or less 5V. Usually something around 4870-4910 mV. Another Vref is internal 1.1V, which seems to be more like 1070 mV. Cap is 100 nF.
[11:00:12] <awozniak> mike_papa: that'll teach me to suggest something when i don't have a formula. I'm more digital, so i always forget stuff involving RC time constants and whatnot. I know I learned it university, but it doesn't stick because I don't use it (which is why I'll do stuff like 'try a few hundred, then ratchet down')
[11:01:04] <mike_papa> awozniak: Ok. Thanks anyway :)
[11:05:14] <awozniak> mike_papa:
http://ladyada.net/library/rccalc.html
[11:06:56] <awozniak> mike_papa: your Vc is just the starting voltage of the cap. The V is the voltage you're shooting for. The R for you will (I think) be very small.
[11:38:31] <mike_papa> awozniak: 41 kOhm gives me 1ms delay. That is about 87 adc cycles. What do you mean by very small? Anyway I'm not the only one with that problem. I've found thread on that on AVR Freaks. It ended with suggestion of not having cap on Vref at all if it's gonna be switched. But I'm affraid that will make things worse if I put everything in car. Old car with distributor large as wine bottle ;)
[11:39:20] <awozniak> mike_papa: w/o cap you'll get noise on the ADC. Do you really need to sample faster than 1KHz ?
[11:40:47] <mike_papa> awozniak: no I don't. It's just matter of time between getting AVCC and sensor value. Shorter it gets, less chance it will change, and make sensor reading inaccurate.
[11:41:34] <awozniak> mike_papa: also, how much noise and/or error is acceptable? Some noise can be eliminated with digital filtering. Amount of acceptable error will depend on the application.
[11:43:02] <mike_papa> awozniak: 8 bit precission is as low as I can go. 9 bits is ok, and I'm working on that now. I don't even try to go for 10 bits.
[11:54:05] <mike_papa> hmm... 10 seems to be a number. It get's totally stable if I throw away 10 readings. I guess larger numbers may coused inaccuracy because AVCC had enough time to change before sensor was read, and it caused random error more on calculations than actual reading.
[15:49:14] <zmo> hi!
[15:49:43] <zmo> I'm trying to use the jtagice3 with avr studio... but my avrstudio v6.1 does not offer me to choose the JTAGICE3 as tool
[15:50:43] <zmo> the only choice I got is the STK500
[15:55:16] <zmo> my goal being to debug a project I'm working on using On Chip Debug
[15:55:17] <zmo> /DebugWire
[15:55:22] <zmo> and when I try with avarice, I get `initJtagBox() failed: debugWIRE failed`
[16:14:23] <The_Coolest> Guys, do Atmels QFN chips have 'side contacts'? So that the can be soldered with a solering iron? Or is it someone one shouldn't even attempts
[16:14:28] <The_Coolest> -s +?
[17:11:41] <jerkey> the_coolest the pins on QFNs are not held in place as well as you think, so even if you were careful enough to solder to them, they would just rip out. What chip are you trying to use? you can get an atmega32u4 board ready-to-use for $5 from china (shipping included)
[17:13:57] <crib> The_Coolest: buy a cheap hot air soldering station (~100 EUR/USD)
[17:15:29] <jerkey> my mistake, i was picturing wanting to solder wires to the pins. anyway crib is right, you need hot air to do a QFN
[17:18:25] <Tom_itx> i've done them with an iron
[17:18:35] <crib> the problem with many qfn chips with side contacts is that these contacts are not plated and you can't get a connection with good conductance
[17:18:38] <Tom_itx> you're better off reflowing them
[17:18:45] <crib> we had such problems with the NXP NFC QFN chips
[17:22:54] <The_Coolest> jerkey, crib yeah I figured. thanks
[17:31:17] <The_Coolest> jerkey I'm not trying anything yet. Just thinking about giving QFN a go (like small boards :D)
[17:51:49] <Epsilon-Auriga> The_Coolest: if you look at the datasheet it will give you a physical layout of the QFN package...the N means Nolead...as in, just pads under the package, no actual leads.\
[17:52:07] <Tom_itx> qfn doesn't have leads
[17:52:17] <Epsilon-Auriga> quad flat nolead.
[17:52:43] <Epsilon-Auriga> bitch to solder without a hot air gun or reflow oven.
[17:52:58] <Tom_itx> yes
[17:53:10] <Tom_itx> especially the center pad
[17:53:16] <Epsilon-Auriga> yup.
[17:53:21] <Epsilon-Auriga> I have a solution for center pad.
[17:53:27] <Tom_itx> so do i
[17:53:39] <Epsilon-Auriga> drill hole through board and flow solder down through the hole with the board upside down.
[17:53:59] <Tom_itx> http://tom-itx.dyndns.org:81/~webpage/boards/atmega32u4/atmega32U4_3.jpg
[17:54:03] <Epsilon-Auriga> stole the idea from schmartboard
[17:54:17] <Epsilon-Auriga> that's how they do BGA boards with a soldering iron.
[17:54:21] <Tom_itx> i didn't stole it from anybody
[17:54:29] <The_Coolest> Tom_itx those are pretty
[17:54:46] <Tom_itx> http://tom-itx.dyndns.org:81/~webpage/boards/atmega32u4/atmega32u4_1.jpg
[17:54:55] <Epsilon-Auriga> yup.
[17:55:20] <The_Coolest> are those 0402 passives? done by iron or reflow?
[17:55:21] <Epsilon-Auriga> think I got one of them.
[17:55:34] <Tom_itx> not that board
[17:55:42] <Epsilon-Auriga> oh? that a new one?
[17:55:45] <Tom_itx> no
[17:55:50] <Tom_itx> but i never sold those
[17:55:54] <Epsilon-Auriga> aahh.
[17:55:56] <Tom_itx> except a coupld
[17:55:57] <Tom_itx> e
[17:55:59] <Epsilon-Auriga> I have something similar from you.
[17:56:05] <Tom_itx> 32u2
[17:56:09] <Tom_itx> tqfp
[17:56:10] <Epsilon-Auriga> yeah.
[17:58:11] <The_Coolest> yeah I'll stick to tqfp for now
[17:58:57] <The_Coolest> http://i.imgur.com/NkRWoqT.jpg?2
[18:02:02] <Tom_itx> much easier to hand solder
[18:02:35] <The_Coolest> yup
[18:02:50] <The_Coolest> but takes up so much more space :(
[18:41:39] <carabia> hello
[18:41:41] <carabia> i love smps
[18:42:01] <carabia> i am building x86 smps
[18:42:07] <carabia> halp
[18:45:22] <N2TOH> ever try to build a SMPS with a vacuum tube?
[18:46:16] <carabia> all the time
[18:46:41] <N2TOH> with the tube as the switching element?
[19:25:34] <The_Coolest> is carabia beaky?
[19:26:09] <carabia> i love vacuum tube
[19:26:48] <The_Coolest> yes, beaky loves stuff too
[22:14:48] <N2TOH> carabia, have you seen the hack a day tube amp built with a VFD display?
[23:31:01] <rue_house> who wants to be part of todays mandatory code sharing session!
[23:32:33] <Casper> ask flyback?
[23:33:24] <Casper> but what about a = b = c = d = 0x1F; ? :D
[23:42:32] <rue_house> anything ll do
[23:42:39] <rue_house> the more usefull to everyone the better
[23:43:20] <Casper> so I just installed the breaker for my isolation transformer :D
[23:43:54] <Casper> there is only 1 thing left: install a working ground... because the outlet is ungrounded and I beleive it is unsafe
[23:44:07] <Casper> the water pipe and house ground wire pass just there, so will just tap on it
