Back
[00:21:10] <Tom_itx> http://tom-itx.dyndns.org:81/~webpage/temp/quad.mpg
[02:24:58] <marloshouse> inflex: you in?
[03:22:55] <inflex> Doooooh.... no wonder my logger wasn't terminating at the anticipated time... I was waiting for it to reach 4M-BYTES, not 4M-bits
[03:23:05] <mrfrenzy_> sa
[03:23:33] <inflex> it was more like trying to get to 19.5hrs, rather than 2hrs :D
[03:24:18] <mrfrenzy_> oops
[03:25:04] <Casper> man replacing the speakers in a laptop is quite labor intensive...
[03:25:35] <inflex> Casper: probably why you don't... and just use USB powered ones :D
[03:25:50] <Casper> inflex: it was the last thing to do on this one
[03:25:51] <inflex> still a bit odd that my SPI code didn't pick up the error code
[03:26:01] <Casper> now I have a fully functionning laptop!
[03:26:42] <Casper> inflex: did I told you that the laptop battery pack was finally ok? the problem was that the charger inside the laptop had a floating diode... factory problem... one side had basically zero solder
[03:26:55] <inflex> yep, you had mentioned that
[03:27:01] <inflex> what laptop is it?
[03:28:24] <Casper> hp nc8230
[03:29:23] <inflex> ooh, a Pentium-M chip, nice
[03:30:08] <Casper> that laptop is really fast
[03:30:15] <Casper> specially since it now have 2G of ram
[03:30:23] <Casper> radeon x600 too
[03:30:27] <inflex> yeah, it's a good combo
[03:30:44] <Casper> sadly, I only have bad disks
[03:30:53] <Casper> this one have 47 remapped sectors
[03:30:57] * inflex has an Acer D250 (Atom), which I use for the odd task, but mostly upstairs in the electronics room for flashing chips and controlling the vinyl cutter
[03:31:06] * inflex has a drawer full of drives :\
[03:31:23] <Casper> 2.5" ?
[03:31:26] <Casper> ide?
[03:31:51] <inflex> yeah, quite a few of them are the exposed-pin IDE type
[03:32:09] <inflex> too bad you don't live near by
[03:32:15] <Casper> yeah :/
[03:33:30] <Casper> maybe I should visit the nearest reno depot... (it's like home depot)... their advertisement on tv... is funny, their warehouse is so big that it go underground... and poke throught the other side of the earth :D
[03:33:50] <inflex> basically I've got a big pileup of all manner of drives in this place... because I don't want to discard them without destroying them properly
[03:34:16] <Casper> thermite!
[03:34:57] <inflex> or just pour some potassium-permagnanate + glycerine over them
[03:35:42] <inflex> ""The bodies of three women and four men, aged between 18 and 60, were found with gunshot wounds near a Christmas tree in the apartment living room."" ---- that's just a crazy sick thing
[03:40:33] <Casper> good night
[05:15:29] <vectory> why is atmels numbering for datasheets/app-notes and related binaries that odd?
[05:16:45] <vectory> avr101.zip :D i get that, small number and avr only, thats nice, but why not avr2526, as thats the app note it comes with
[05:16:56] <vectory> doc2526 thats it
[05:17:39] <vectory> and comes doc0932 with avr100, where is the order? designing microchips, they should have a thing for order >:|
[06:21:13] <pc_magas> cu
[06:32:01] <mapee> morning
[10:07:59] <sparr> is there an AVR library for decompressing gz files? I'm having a hard time coming up with search terms that narrow it down beyond unzipping parts of the AVR stk and other libraries
[10:08:27] <jacekowski> you can't do that on avr
[10:09:09] <sparr> why?
[10:09:30] <jacekowski> because you need more ram to do it
[10:09:57] <jacekowski> gzip is using 200k of ram to decompress
[10:10:04] <jacekowski> on avr you've got 1-2k
[10:10:36] <sparr> gzip inflate can be done in 1300 bytes of ram
[10:11:18] <jacekowski> who said that
[10:11:21] <sparr> http://atariarea.krap.pl/x-asm/inflate.html
[10:12:09] <jacekowski> that's assuming you use small block sizes
[10:12:13] <jacekowski> gzip uses 32k blocks
[10:13:04] <sparr> that's configurable
[10:13:29] <jacekowski> and that asm code assumes you are running on von neuman arch
[10:14:05] <jacekowski> hmm
[10:14:06] <jacekowski> maybe not
[10:14:20] <sparr> just an example, I'd probably start from scratch if I was writing an AVR implementation
[10:14:26] <sparr> I'm just hoping someone else already has
[10:15:27] <jacekowski> do you need gzip or you can use any compression?
[10:17:25] <sparr> undetermined
[11:05:30] <pc_magas> hello
[11:28:15] <vectory> hi pc_magas
[11:28:21] <pc_magas> Hello
[11:28:31] <pc_magas> Marry Xmas
[11:28:40] <pc_magas> And A happy new year
[11:28:40] <vectory> thx, you too
[11:28:53] <vectory> new year? this one is rather old
[11:31:13] <pc_magas> The next one that it will come
[11:31:16] <pc_magas> ;P
[12:03:18] <vectory> i have a LUT, where do i want to save it, eeprom or flash?
[12:04:50] <vectory> btw, it takes me about an hour to read 120 pages of datasheet, feels slow :(
[12:04:57] <vectory> 10 pages*
[12:05:16] <OndraSter> quite difference lol
[12:05:52] <RikusW> vectory: how often does it need to change ?
[12:06:07] <RikusW> if loaded with the program each time put it in flash
[12:06:12] <vectory> the data is hardwired
[12:06:22] <vectory> needs to read quickly, like 2000hz
[12:06:43] <RikusW> use flash
[12:06:52] <OndraSter> use RAM :)
[12:06:55] <RikusW> or flash and copy to ram if you feel like it
[12:07:07] <OndraSter> if it will be changed, use EEPROM... I'd say
[12:07:14] <OndraSter> and cache it in RAM
[12:07:18] <RikusW> char lut[] = { 0x44, 0x55, .... };
[12:07:36] <RikusW> will use flash and copy to ram, using gcc magic ;)
[12:07:51] <vectory> .db 0b01001000 :P
[12:08:04] <RikusW> vectory: using asm or gcc ?
[12:08:29] <vectory> well, gcc understands .asm but im using studio4 to assemble
[12:08:47] <vectory> which is quite annoying since i dont know if the programmer works in xp for me
[12:08:51] <RikusW> in C / gcc ?
[12:08:58] <RikusW> mine ?
[12:09:02] <vectory> yes
[12:09:07] <RikusW> it works with AS4
[12:09:10] <vectory> i dont write avr c yet
[12:09:17] <RikusW> just put it in bootloader mode
[12:09:33] <RikusW> or stk500 if you want to program another avr
[12:09:41] <Tom_itx> morning
[12:09:46] <RikusW> and connect as stk500 to COM3 (usually)
[12:09:48] <RikusW> morning
[12:09:53] <vectory> evening :)
[12:10:13] <RikusW> more like evening here too ;)
[12:10:20] <OndraSter> ey Tom_itx
[12:10:50] <RikusW> vectory: your program is using the CDC com port ?
[12:11:01] <vectory> no
[12:11:06] <vectory> why?
[12:11:45] <RikusW> if you do you'll need to use U2Sfw as a starting point...
[12:12:02] <RikusW> whats the project you're trying to do ?
[12:12:09] <vectory> hm, for 2nd avr?
[12:12:37] <RikusW> so you're programming anoter avr ?
[12:13:26] <vectory> yes
[12:13:30] <RikusW> you'll need to use the u2s stk500 mode then
[12:13:33] <vectory> i just started with the m8 i got and want a nice led matrix with rtc and buzzer melody playback, an alarmclock basicly.
[12:13:53] <vectory> bad idea, seeing it will be hard to get the crystalerror down to a few ppm
[12:13:55] <RikusW> got the isp connected yet ?
[12:14:11] <vectory> nope, but i wrote half the code >_<
[12:15:31] <RikusW> you could use the U2S_GUI app to switch to stk500 mode, its a little more convenient ;)
[12:15:47] <RikusW> for programming your m8
[12:15:49] <vectory> or u2scli :)
[12:16:12] <keenerd> Wait, why can't you just use a 32k crystal? I've had no problems with error using them for rtc.
[12:16:44] <RikusW> of course..
[12:17:12] <RikusW> what crystal does the m8 use ?
[12:17:19] <vectory> keenerd: +/-20ppm is like 10 minutes a year, afaict
[12:17:29] <vectory> RikusW: hm?
[12:18:03] <RikusW> what clock source does the m8 have ?
[12:18:08] <vectory> ill run it with internal rc, probs 1mhz and use 32khz crystal for async timer
[12:18:11] <keenerd> vectory: So you want an alarm clock good enough for rip van winkle's 40 year nap? Whyever?
[12:18:51] <keenerd> You'll probably have a brownout or a dead battery before clock drift becomes an issue.
[12:18:55] <vectory> keenerd: no, batteries wont even hold for a year x)
[12:19:28] <RikusW> vectory: battery powered ?
[12:20:22] <vectory> yes, i still consider to switch to tiny2313 because of lower voltage, but where do i get 1.8V batteries
[12:21:18] <RikusW> use 3v ?
[12:21:45] <vectory> thats the current plan
[12:22:01] <keenerd> I've heard it is actually more effecient to use an LDO and run the micro as low as possible.
[12:22:13] <keenerd> There is an msp430 app note about it, iirc.
[12:22:21] <OndraSter> how do you do "idle" mode? NOPs = idle mode?
[12:22:23] <OndraSter> or some kind of sleep mode
[12:22:55] <RikusW> avr got sleep mode
[12:23:01] <RikusW> hlt instruction
[12:23:08] <OndraSter> okay, so they mean sleep mode
[12:23:08] <RikusW> need an irq to wake up
[12:25:33] <vectory> there is actually a sleep instruction with 4-5 different modes
[12:25:38] <keenerd> vectory:
http://www.ti.com/lit/an/slyt356/slyt356.pdf No reason it won't work for AVRs too.
[12:30:54] <vectory> thx keenerd, reading it later
[12:32:36] <vectory> that figure on the first page, does it mean msp430 needs less than 1<mA, what ever active-mode is?
[12:33:10] <keenerd> Pretty sure that is the sleep mode.
[12:33:53] <keenerd> Actually
[12:34:16] <keenerd> They say 200uA at 1MHz at 2.2V, so maybe not.
[12:35:07] <vectory> that sounds to good to be true
[12:35:55] <vectory> have the msp430launchpad here, will get a digital meter soon
[12:36:02] <vectory> can try then :D
[12:36:04] <keenerd> Yeah, I know. I got that number of the first page of the datasheet. You can never trust the first page.
[12:36:11] <keenerd> *off the
[12:38:07] <keenerd> Digger further, they are pretty adament about those numbers. That is the fully active. Of course you'd be pulling like 100x more at the full 8Mhz.
[12:38:18] <keenerd> *fully active current.
[12:38:34] <vectory> only need 1mhz
[12:38:54] <vectory> for a clock i only need 1 hz :'D
[12:45:12] <vectory> with only 500µA it would work way longer than 300h on 3V AAA batteries
[12:45:21] <vectory> two 1.5 AAAs
[12:46:43] <vectory> or not, dont know how many mAh they had
[12:46:57] <vectory> anyways, i'll save it for later
[12:49:17] <RikusW> why worry about avr current when you wan a led matrix ? :-O
[12:55:32] <vectory> true, that wories me more
[12:56:23] <vectory> but it only need to switch on the leds for fractions of a second
[12:57:15] <vectory> and with the method you proposed, having two ports to drive a 8x8 matrix, will only draw current from the avr
[12:58:07] <vectory> still need to experiment, how low leds can go in terms of switching frequency and hold time
[12:58:20] <keenerd> There is no hold time in LEDs.
[12:58:24] <OndraSter> "fractions of a secpmd"
[12:58:25] <OndraSter> second
[12:58:30] <OndraSter> but they will be always lit up :P
[12:58:32] <vectory> second, yes :)
[12:58:32] <OndraSter> if it is 8x8 matri
[12:58:33] <OndraSter> x
[12:58:45] <OndraSter> there will be, let's say, 160mA per row tops
[12:58:49] <OndraSter> now you need to hold that...
[12:58:59] <OndraSter> okay, so make it 40mA constant current in average?
[12:59:12] <RikusW> battery killing 160mA...
[13:00:12] <keenerd> High capacity NiMH is like 2500mAh.
[13:00:43] <vectory> alkaline, too
[13:01:00] <RikusW> maybe make the leds very dim and brighten up at the push of a button ?
[13:01:16] <keenerd> If you are going through batteries every 15 hours you don't want to use alkalines.
[13:01:18] <vectory> thought of 4 amps per led, 256 times for 1µsec each second
[13:01:31] <vectory> that will be very dimm?
[13:02:03] <RikusW> 4 A ?!
[13:02:06] <OndraSter> 4Amps?
[13:02:09] <OndraSter> it will not be dimm at all! :D
[13:02:14] <vectory> milli*
[13:02:26] <vectory> OndraSter: it will be, once the leds blow
[13:02:26] <RikusW> better ;)
[13:02:28] <OndraSter> u sure about 1usec per second?
[13:02:51] <vectory> for one instruction cycle at 1mhz
[13:02:56] <keenerd> 1uS pulses are barely visible without current limiting resistors.
[13:02:58] <OndraSter> I am afraid you will need it to be lit up more than 1uS
[13:03:16] <vectory> as i said, should experiment
[13:03:39] <keenerd> My vote is to plug it in and don't worry about anything.
[13:03:44] <vectory> there will be limiting resistors
[13:04:27] <OndraSter> you say 4mA?
[13:04:31] <OndraSter> check how much can atmega source
[13:04:36] <keenerd> You don't want any resistors in a battery powered design if you can help it, and you don't need them if you've got good PWM at higher voltages.
[13:04:38] <vectory> keenerd: that will burn a lot of energy, think about the environment ;)
[13:04:48] <OndraSter> oh, you have it on battery
[13:04:49] <OndraSter> I forgot
[13:06:47] <vectory> OndraSter: source as in supply?
[13:06:51] <OndraSter> ye
[13:07:10] <vectory> 40
[13:07:17] <vectory> mA
[13:07:28] <OndraSter> really? I remember having problems with 10mA on some pins actually
[13:08:05] <vectory> oh, i read the maximum ratings
[13:08:23] <vectory> well thats all i can find about I/O pin current
[13:14:00] <vectory> laters
[13:49:59] <papo> hello. I just grabbed an arduino duemilanove board, but I am not using any arduino software. I was trying to read the fuses to figure out what clock settings are active. This is what I get:
http://pastebin.com/WX7SWBqW
[13:50:22] <papo> does that actually mean that all fuses are set to 0?
[13:50:55] <papo> the table is from the avrdude output by the way
[14:32:36] <OndraSter> d00des, I am so smart
[14:32:40] <OndraSter> I just managed to find a way how to save one chip :P
[14:32:42] <OndraSter> http://clip2net.com/s/1qYgP
[14:32:54] <OndraSter> address one chip with xxxx10 and the other one with xxx01
[14:33:03] <OndraSter> aka connect one to QF and the other one to QE
[14:36:59] <amee2k> hmm i just had an inspiration... the input power of a switching regulator is (more or less) output power plus conversion losses
[14:37:40] <amee2k> and a switching LED (down converter) driver is really just a buck converter with current feedback
[14:38:02] <amee2k> so why can't i PWM a power LED by putting a big FET in parallel and shorting out the LED?
[14:39:07] <amee2k> while the LEDs are off (i.e. shorted out), the driver is just circulating current in the inductor and doing nothing
[14:40:17] <Kevin`> that's nasty
[14:40:19] <Kevin`> why would you do that
[14:40:38] <j4cbo> you can
[14:40:38] <amee2k> higher PWM resolution
[14:40:45] <j4cbo> in fact, i've seen that as the recommended way of doing it
[14:40:59] <j4cbo> in some app note somewhere
[14:41:16] <amee2k> PWMing a driver's "enable" input, even if it is made for doing that, limits your resolution to whole switching cycles (plus slopes when the driver turns on and off
[14:41:21] <Kevin`> interesting
[14:41:23] <rue_house> put the led backwards and use an inductor in series with the power line, with a bypass diode,
[14:41:31] <rue_house> buckboost converter
[14:42:02] <rue_house> extra heat on the led cause of the AC across it
[14:42:53] <amee2k> what is wrong with AC across it? the shitty reverse voltage ratings of most LEDs will mean you need another blocking diode i suppose
[14:43:15] <rue_house> capacitance of the led will generate heat
[14:43:27] <amee2k> hmm okay
[14:43:33] <rue_house> this is an obscure conversation
[14:43:42] <amee2k> it is?
[14:43:45] * rue_house sprinkles passive components all over it
[14:44:18] <amee2k> well, what made me think about it is the fact that 16 bit PWM with a switch mode driver doesn't really make much sense
[14:45:12] <amee2k> being limited in resolution to full switching cycles means at 150Hz PWM frequency i need 150kHz to barely get 10 true bits of PWM out of it
[14:45:17] <rue_house> do a buckboost, then you can put as many leds in series as the raw supply can take and still have power when the batteries are dead
[14:45:46] <rue_house> /power/voltage/
[14:46:20] <amee2k> my current design is based on an LM3404 down converter which looks dead easy to use
[14:47:27] <amee2k> but for 8 bit gamma-correct PWM i need a bare minimum of 10 bit linear, preferably 12 or more
[14:47:47] <OndraSter> "Can't backannotate this operation. Please do this in the schematic!"
[14:47:50] <OndraSter> this was always driving me crazy
[14:48:56] <amee2k> rue_house: hmm got any recommendations on a 1A LED driver IC?
[14:49:10] <rue_house> 1A?!
[14:49:19] <amee2k> yeah
[14:49:29] <amee2k> thats not too much, really
[14:49:31] <rue_house> current mode simpleswitcher
[14:49:41] <OndraSter> 1A per LED?
[14:49:44] <amee2k> lol
[14:50:04] <amee2k> i made a shitty 500mA driver with an MC34063
[14:50:08] <OndraSter> that is 50 LEDs
[14:50:08] <OndraSter> oh
[14:50:25] <amee2k> OndraSter: 1A per led or for all leds is the same since they're in series
[14:51:05] <OndraSter> series?
[14:51:08] <amee2k> at first i was considering a 3A driver which is probably more useful, but then i settled for the small one
[14:51:29] <OndraSter> series would mean that there would be 1A through each diode :o)
[14:51:45] <amee2k> yeah, thats the whole point
[14:51:58] <amee2k> putting LEDs in parallel is a huge no-no. like, exxon valdez sort of bad vibes
[14:52:17] <OndraSter> wait, how do you plan on cooling diodes where is constant 1A through them
[14:52:31] <OndraSter> I hope it is some better LED :)
[14:52:32] <OndraSter> Cree
[14:52:37] <OndraSter> with built-on heatsink
[14:52:55] <amee2k> a ton of vias under the thermal pad, then ground plane on the bottom layer. then bolt the whole board against a heatsink
[14:53:39] <amee2k> cree has a very interresting appnote where they demonstrate and benchmark different cooling arrangements
[14:54:38] <amee2k> i've got some XR-C LEDs at home, but i want to try XM-L series LEDs for the final design for higher efficiency. they're considerably more expensive though
[14:55:43] <amee2k> rue_house: what is putting me off about generic SMP controllers is the high reference voltage
[14:56:05] <amee2k> 1.25V or 2.5V typical vs. 200mV for the 3404 LED driver
[14:56:32] <amee2k> that means either active current monitor or relatively large value (and wasteful) shunt
[14:56:50] <OndraSter> what do you say on this
[14:56:51] <OndraSter> http://clip2net.com/s/1qYyh
[14:57:00] <OndraSter> 2Amps for few usec per second :)
[14:57:21] <amee2k> what is that? strobe light?
[14:57:32] <OndraSter> LED field
[14:57:36] <OndraSter> 32x96 :)
[14:57:41] <OndraSter> 32x48 bi-color
[14:57:43] <amee2k> nice
[14:57:51] <OndraSter> programmable through USB
[14:57:59] <rue_house> hmm
[14:58:04] <OndraSter> atmega128a with xmem, ftdi ft232rl
[14:58:12] <amee2k> i'm more interrested in lighting applications rather than displays
[14:59:36] <amee2k> i'll probably just finish this board with the 3404 driver to see how it works
[14:59:58] <amee2k> it'll work for testing the cooling arrangements and the control circuitry and its firmware
[15:00:18] <amee2k> then make a second revision with a higher performance driver
[17:02:03] <amee2k> hmm i'm working on a replacement for the shitty interior light delay circuit on my car. my circuit will always draw power for the lamp from the permanent rails. how do i know how much sense current to pass through the door switch for reliable operation?
[17:10:48] <amee2k> mmh no ideas? :)
[17:11:01] <OndraLappy> test it out? :)
[17:11:07] <OndraLappy> how does the switch work?
[17:11:13] <OndraLappy> magnetic field?
[17:11:31] <amee2k> a rubber nipple that the door pushes against when it is fully closed. so probably purely mechanical
[17:12:00] <amee2k> i don't like "trial and error" anymore
[17:12:10] <amee2k> at least not for reliability critical parameters
[17:14:17] <Valen> amee2k: whats your concern?
[17:14:47] <OndraLappy> if it is pure mechanical with (in theory) 0R
[17:14:55] <OndraLappy> then any current that can be detected by the chip
[17:14:56] <Valen> also your door switches will likley go to other places in the car, IE the body computer to do "door is open dumbass" chimes and such
[17:14:57] <amee2k> that a medium current switch doesn't work reliably with just a few microamps passing
[17:15:30] <amee2k> or that leakage current causes false closed readings
[17:15:34] <Valen> I'd just look to tap off the existing thing and read its voltage
[17:16:09] <amee2k> either the wire goes straight to the door switch, or at least it is an open collector / open drain style output
[17:16:43] <amee2k> the light module just has the door contact wire and the manual on switch OR-wired
[17:16:59] <Valen> it'll probably just have 12v applied to it that gets pulled to ground in the existing setup
[17:17:43] <Valen> oh, well thats different so it makes 12v happen when the door is opened?
[17:17:48] <amee2k> when the door opens, the switch closes and current runs through the lamp and out through the door switch wire to ground. except for a shitty 3-wire light delay circuit the light module is dumb
[17:18:35] <amee2k> i.e. permanent positive -> lamp -> ... a) manual switch -> ground .... b) wire to the door switch
[17:18:46] <Valen> I C
[17:19:01] <amee2k> the delay module is wired to the negative end of the lamp, permanent ground, and door switch ground
[17:19:03] <Valen> given its in a car put something like 200 ohms in i reckon
[17:19:34] <Valen> enough that the resistor won't get too hot
[17:19:50] <Valen> but you have plenty of power and it'll only be flowing when the door is open
[17:20:03] <amee2k> hmm yeah, thats true too
[17:20:31] <amee2k> i've managed to reduce the standby current on paper to under 100uA otherwise
[17:20:43] <Valen> thats kinda worthwile
[17:20:49] <amee2k> i'll build a prototype when i get home, which is probably shortly after new year
[17:21:06] <Valen> though the "standby" current of most cars is pretty insane
[17:21:12] <Valen> I have seen up to an amp or two
[17:21:21] <amee2k> i'm not sure if i could do zero-standby if i tried
[17:21:26] <amee2k> what the fuck???
[17:21:51] <Valen> some fancy sports car, known issue would kill the battery in a few days
[17:22:08] <amee2k> at that rate you need a huge battery and it won't be enough to start after maybe 12-15 hours on a cold day
[17:22:32] <Valen> they had a huge battery and it ate it in 3 days or so
[17:23:36] <amee2k> for most cars i'd say everything >60Ah counts as large. and if you get below 60% charge state cranking ability will be compromised
[17:23:49] <Valen> dont shoot the messenger
[17:23:53] <Valen> just telling the story
[17:23:58] <amee2k> i know >_>
[17:24:02] <amee2k> but that is ridiculous
[17:24:23] <amee2k> someone deserves to lose his fertility over this. for the good of the gene pool
[17:24:23] <Valen> it was a sports car so it was mostly driven on weekends
[17:24:37] <Valen> so it'd die before each weekend
[17:24:39] <amee2k> lol
[17:24:49] <Valen> manufacturer suggested putting a battery charger on it
[17:25:14] <amee2k> when you start shipping cars with a float charger as stock accessory, you know you're doing it wrong :P
[17:26:35] <keenerd> Or right. See Leaf and MiEV.
[17:27:31] <Valen> those arent "float" chargers ;-P
[17:27:58] <amee2k> lol
[17:29:00] <amee2k> i'd seriously pull the ground strap on the battery terminal
[17:29:21] <amee2k> fuck the radio station memory in a sportscar
[17:29:38] <Valen> how about the engine map?
[17:29:42] <Valen> kinda important that
[17:30:06] <amee2k> if the ecu loses its engine map when the battery runs out, i'd send it back for repair
[17:30:14] <Valen> they all do
[17:30:21] <amee2k> not really
[17:30:31] <Valen> they start from scratch relearning the engine from the stock map
[17:30:41] <amee2k> then i'd have had to get mine reprogrammed a few times allready
[17:30:43] <Valen> well a decent % of them anyway
[17:31:04] <Valen> no they come with a stock map, then they "learn" your engine and driving style
[17:31:17] <amee2k> oh, the on the fly correction data is back after a few minutes usually
[17:31:35] <keenerd> I miss cars where the smartest thing was an eccentric cam.
[17:31:55] <Valen> i'm of two minds
[17:32:04] <amee2k> i miss my old car. not quite the part where it dropped its exhaust on the road, but still
[17:32:07] <Valen> if the car came with a schematic and source code i'd be ok with it
[17:32:17] <amee2k> erm
[17:32:21] <amee2k> BHAHAHAAAA
[17:32:27] <amee2k> erm, sorry
[17:32:34] <Valen> it is a bit like that
[17:32:40] <amee2k> but take a look at the car industry
[17:32:50] <Valen> though i am thinking of putting a new engine with a megasquirt into my car ;->
[17:32:55] <amee2k> you can't find a fucking datasheet for anything at all
[17:33:21] <keenerd> I've been thinking you could do quite a bit with bus sniffing/injecting.
[17:33:35] <Valen> my car has an ECU that runs the spark and injectors and not much else and its not integrated into everything else, its about where i like it
[17:33:48] <Valen> keenerd: you can do that with your car lol
[17:34:01] <amee2k> a while ago i tried to find a datasheet for an ignition coil. not even aftermarket suppliers have public datasheets as it seems
[17:34:22] <Valen> coil might be "too simple" for a datasheet
[17:34:31] <amee2k> umm
[17:34:48] <Valen> i mean they only really need a few numbers
[17:34:49] <amee2k> there are datasheets for bolts and wires even
[17:34:54] <Valen> yeah i know that
[17:34:58] <Valen> but these are car people
[17:35:04] <Valen> most of them probably can't read
[17:35:05] <Valen> ;-P
[17:35:08] <amee2k> thats what i just said :P
[17:35:43] <amee2k> in a time far far away when i'm a rich fuck and need to kill some time, i'll start an open source car project
[17:35:59] <Valen> megasquirt
[17:36:51] <amee2k> starting with a welder and some steel beams. documentation includes full blueprints for all parts, complete schematic, functional description and firmware
[17:37:07] <OndraLappy> firmware
[17:37:11] <OndraLappy> will it run on AVR? :D
[17:37:13] <keenerd> There is that guy who made an open source tractor. A minute for link...
[17:37:42] <amee2k> OndraLappy: i didn't look to closely because they are fairly expensive, but the automotive AVRs look pretty cool
[17:37:45] <keenerd> http://opensourceecology.org/wiki/LifeTrac
[17:37:51] <OndraLappy> yeah
[17:37:53] <OndraLappy> CAN bus 'n stuff
[17:38:00] <amee2k> i did consider using a CANbus based bus to control my lighting modules
[17:38:34] <keenerd> Top speed is a little limited by the hydraulic motors :-)
[17:38:34] <OndraLappy> adding your device to the bus that ECU is connected to?
[17:38:41] <OndraLappy> you know what happens when you do some fault :P
[17:39:17] <amee2k> yeah, some of that stuff will need thorough debugging
[17:39:34] <amee2k> i'd avoid all the fly by wire crap wherever possible, tbh
[17:40:07] <amee2k> old style throttle and clutch cables, and a real steering linkage
[17:40:45] <OndraLappy> hehe
[17:41:26] <OndraLappy> will be there some #ifdef for automatic clutch? :P
[17:41:41] <OndraLappy> I ain't "stupid american" but I like auto a bit better than manual gearbox
[17:41:48] <amee2k> double a-frame suspension is simple to build, easy to analyse and if done right has good handling characteristics
[17:43:18] <amee2k> for the engine, i'd try to buy some existing engine design from the 90s or early 00s as a starting point, fit my own control systems
[17:43:34] <OndraLappy> would it pass certification?
[17:43:38] <OndraLappy> what is all required?
[17:43:46] <OndraLappy> emissions, safety
[17:43:50] <Valen> i hate auto gearboxes
[17:43:53] <OndraLappy> hehe
[17:43:57] <Valen> i hate how they keep pushing
[17:43:57] <OndraLappy> I am doing car school
[17:44:05] <OndraLappy> I tried auto for few minutes
[17:44:05] <amee2k> safety regulations are probably the most difficult to comply with i imagine
[17:44:07] <Valen> i have terrible difficulty driving with them
[17:44:09] <OndraLappy> and it felt better in city
[17:44:23] <Valen> i also avoid driving in traffic like the plague lol
[17:44:27] <OndraLappy> you know... 1 2 stop. 1 2 stop.
[17:44:41] <OndraLappy> when you are stopping on every single red light, it gets really boring
[17:44:49] <Valen> Ondra you mean put it in first and idle along at the average speed of traffic?
[17:44:50] <Valen> ;-P
[17:44:56] <amee2k> i grew up on stick, and my first car's gearbox was about to fall apart
[17:45:03] <OndraLappy> :D
[17:45:19] <amee2k> so i pretty much had to learn how to shift properly. now i do it without really thinking about it anymore
[17:45:23] <mrfrenzy_> if you are doing "1 2 stop. 1 2 stop. 1 2 stop. 1 2 stop. 1 2 stop. 1 2 stop." you are doing it wrong
[17:45:28] <OndraLappy> I don't know how many of you know czech skoda fabia
[17:45:30] <mrfrenzy_> just find the proper speed to flow along ;)
[17:45:36] <Valen> i notice because my car has a heavy clutch :-<
[17:45:49] <OndraLappy> mrfrenzy_: if you have to go quickly before the red comes on, you can't go avg speed really
[17:45:58] <amee2k> then you have light legs :P
[17:45:59] <Valen> http://www.vapourforge.com/jake/general/hotness_touched.jpg
[17:46:13] <OndraLappy> looks good
[17:46:37] <OndraLappy> http://www.celysvet.cz/fotky/auto-skoda-fabia_1.jpg
[17:46:39] <OndraLappy> skoda fabia
[17:46:40] <OndraLappy> classic CZ car
[17:46:43] <OndraLappy> I am doing school in that
[17:46:49] <amee2k> Valen: in most cars you don't need to push the pedal to the floor in order to disengage the clutch
[17:46:50] <OndraLappy> it is either 1.2 or 1.4
[17:46:53] <OndraLappy> with LPG
[17:47:09] <OndraLappy> you need here go pedal to the metal to make it accelerate a bit lmao
[17:47:17] <OndraLappy> I am doing school in this
[17:47:23] <OndraLappy> classic czech cheap car
[17:47:32] <amee2k> and every car sort of has its "rythm" for lack of a better term... probably some time constant that governs how quickly engine speed drops when you lift off and put in the clutch
[17:48:00] <amee2k> when you get a feel for how quickly or slowly to shift, you can let the clutch out really fast without a jolt
[17:48:55] <amee2k> in my old car, getting first in while moving was pretty much impossible without a double clutch. upshifting into second worked fine, but downshifting was a lot easier with double clutch too
[17:48:59] <OndraLappy> my position to cars is similar to other people's position to computers - "as long as it works fine"
[17:49:04] <amee2k> and i didn't have a tach in that car >_>
[17:50:24] <amee2k> OndraLappy: well, to me a car is a tool. but as with every tool you can have much more fun if you treat it respectfully and get a feel for how to use it smoothly and decisively
[17:50:37] <OndraLappy> yeah, that sure
[17:50:50] <OndraLappy> but I don't really enjoy getting to know how it works inside
[17:51:00] <OndraLappy> too much mechanic stuff :P
[17:51:23] <OndraLappy> it took me a long time before I learned how to accelerate with clutch... because with LPG, you have to go a bit harder on the gas pedal
[17:51:27] <OndraLappy> which I was afraid
[17:51:38] <OndraLappy> once the boss disabled LPG and woot, worked first time fine
[17:51:47] <OndraLappy> just pure gasoline
[17:51:51] <amee2k> there is a difference between getting into all the greasy details, or understanding involved concepts to better put its potential to use
[17:51:52] <OndraLappy> petrol*
[17:53:45] <amee2k> you only really need the latter to understand some basic handling aspects, like when you'll get a jolt when you get off the clutch too fast
[17:54:14] <OndraLappy> yeah
[17:54:19] <OndraLappy> I learned that the hard way :D
[17:54:54] <amee2k> hehe
[17:55:47] <amee2k> i never took a gearbox apart, but i know the general concept of how it works. same as, say, a torque wrech or a scope or whatever
[17:55:53] <OndraLappy> yeah
[17:56:01] <OndraLappy> I know how it _should_ work lol
[17:56:02] * amee2k . o O ( much unlike the exhaust, but that doesn't count... it took itself apart >_< )
[17:56:11] <OndraLappy> lol
[17:57:04] <amee2k> why rev matching works is pretty basic theory, and you can do a vestigial form of it by just timing it right when to let the clutch out
[17:57:55] <amee2k> i once drove a couple friends around in the back seat (my flatmate was sitting in the passenger's seat)
[17:58:26] <amee2k> and when they got out one of them went like "huh, its a stick shift? it kinda felt like automatic. you're driving a lot, eh?"
[17:58:37] <OndraLappy> eh
[18:00:01] <amee2k> when i'm driving normally most people apparently don't feel my shifts anymore... only when i got distracted or something and had to rush a shift
[18:01:04] <mrfrenzy_> haha, americans talking about stick shifts like something exotic is hilarious
[18:01:46] <amee2k> good that i'm not american then. otherwise i'd be tempted to feel insulted ;)
[18:01:56] <OndraLappy> :))
[18:01:58] <Valen> most people drive manuals here, and the only person who i know that isn't good at it is my missus little sister
[18:02:13] <OndraLappy> yeah, people have here mostly manual too
[18:02:17] <Valen> (thats probably because she hasn't gotten here licence yet ;->)
[18:02:44] <amee2k> the only person i know driving auto here is my flatmate's grandpa, and he isn't driving a lot anymore at all
[18:08:20] <amee2k> keenerd: the site with that tractor article is quite interresting
[18:09:27] <amee2k> keenerd: too bad they don't have a working welder project yet
[18:09:56] <keenerd> That is actually really easy.
[18:10:10] <OndraLappy> welder? high current transformer
[18:10:14] <OndraLappy> :P
[18:10:25] <amee2k> yeah, made me wonder why not as well
[18:10:26] <OndraLappy> or SMPS for high currents
[18:10:27] <keenerd> I've got several books from the '60s that talk about winding your own high current transformer and making a welder.
[18:10:49] <OndraLappy> I've seen some trafos for hundreds/thousands of amps.... thicker than thick!
[18:11:40] <amee2k> keenerd: people rewind big MOTs with 10sqmm single stranded wire and apparently it works for small to medium welds
[18:12:17] <amee2k> keenerd: you don't by chance have any links to these books?
[18:12:49] <keenerd> I
[18:12:57] <keenerd> I'll just tell you the titles?
[18:13:19] <amee2k> mmh, okay :)
[18:14:05] <amee2k> finding MOTs or other high power mains frequency cores around here is not exactly easy, and i could use a small welder sometimes
[18:14:39] <OndraLappy> MOTs?
[18:14:40] <Valen> i saw an indian welder that was made out of a stack of bean cans for the laminations and some wires wrapped around it
[18:14:43] <OndraLappy> old microwave :)
[18:14:50] <amee2k> and all the online stuff is either rewinding existing cores or needs special materials with no real information how to adopt to different ones
[18:15:24] <amee2k> OndraLappy: well, try finding them here. almost all electronic waste recycling here is handled by private companies
[18:15:36] <OndraLappy> oh
[18:15:39] <OndraLappy> we have some public stuff
[18:15:43] <OndraLappy> if you come at the right time
[18:15:46] <OndraLappy> they will give you :P
[18:15:50] <amee2k> and since they make money off the junk, dumping stuff is for free. that implies dumpster diving is not
[18:16:37] <amee2k> most companies won't let you near the dumpsters unless you can show something to dump
[18:16:51] <keenerd> amee2k: Of the books I have in easy reach, "Home Built Power Tools", published in 1963, has plans for a spot welder and a butt welder.
[18:16:53] <OndraLappy> oh, different country :)
[18:17:07] <amee2k> germany >_>
[18:17:17] <OndraLappy> some people come there and grab old PCs that people threw away :)
[18:17:25] <OndraLappy> and other small electronics stuff
[18:17:31] <OndraLappy> they actually take it apart there too usually
[18:17:34] <OndraLappy> like CRTs etc
[18:17:44] <OndraLappy> I'd love to work there as a part time job lol
[18:17:51] <keenerd> amee2k: There is a copy on Ebay right now for $2
[18:17:51] <OndraLappy> flybacks, MOTs, everything
[18:17:53] <OndraLappy> for fun
[18:18:49] <amee2k> OndraLappy: about two years ago, i managed to divert some computer junk from the university to my place before it reached the dumpster
[18:18:55] <OndraLappy> wow
[18:18:57] <OndraLappy> cool :))
[18:18:58] <amee2k> keenerd: ooh, with shipping across the pond?
[18:19:18] <OndraLappy> btw, who would have thought that flybacks have built in capacitor?
[18:19:36] <amee2k> got me a huge pile of passives, including lots of foil and ceramic caps and some power resistors. and also 5 FBTs from 22" color CRTs
[18:19:43] <OndraLappy> wow
[18:19:49] <OndraLappy> I have 1 from 17 and one from 19
[18:19:51] <OndraLappy> and then one from TV
[18:20:14] <amee2k> i salvaged virtually everything. including the power transistors from the mainboard regulators
[18:20:29] <OndraLappy> :)
[18:20:45] <amee2k> back then my junk box was a small cardboard box with some passives and LEDs and 74HC00 series ICs
[18:21:30] <amee2k> now it is a pretty large white plastic box full of these boxes with little compartments to sort stuff
[18:21:52] <amee2k> like 2/3rd of that was from that huge salvage run i guess
[18:22:08] <keenerd> amee2k: Not sure if the seller is open to that. You could try contacting them directly. I'd offer to scan mine, but it is 17 pages total and I am a little busy right now.
[18:22:26] <keenerd> *17 pages just for the welders
[18:22:38] <amee2k> keenerd: i can poke you sometime later again, if you want
[18:23:15] <amee2k> but still quite interresting that there used to be books about that kind of stuff. i didn't know that before
[18:23:20] <keenerd> The books is 160 pages total, you should get it just for the pure awesomenes.
[18:23:37] <amee2k> i'll have a look around and see what i can find on european ebay or amazon too
[18:24:09] <keenerd> They printed them annually, dunno what tools are in the other editions.
[18:25:13] <amee2k> i did read some stuff about mains transformer design a while ago, but finding suitable core materials to roll my own is the worst issue
[18:25:22] <keenerd> Anyway, the internet is probably a better source than 50 year old books. For example,
http://diyspotwelder.blogspot.com/
[18:27:26] <amee2k> spot welding sounds interresting as a tool but not really my thing. i did stick and mig/mag before
[18:28:13] <OndraLappy> you know you can get welders for almost 0 bucks/euro/pounds on some kind of garage sells :)
[18:28:16] <amee2k> as i said lots of projects i've seen before are interresting but not exactly easy to rebuild
[18:28:27] <OndraLappy> or on copies of ebay or on ebay in germany
[18:28:35] <OndraLappy> if you have luck, they are from your city :)
[18:29:00] <amee2k> german craiglist basically doesn't exist from what i've seen
[18:29:18] <amee2k> i tried to get a scope that way and after over a year i had idly saved enough money to get a new one
[18:29:45] <OndraLappy> ds11052 or whatever was the number
[18:29:48] <OndraLappy> like 300 bucks
[18:30:10] <amee2k> i got the better one and over two years ago it was a bit more expensive
[18:30:18] <OndraLappy> 50MHz flashable to 100MHz :)
[18:30:55] <amee2k> but still. scoring used stuff and at a useful price is pretty hard here
[18:31:07] <amee2k> but i've ranted about that to great length before >_<
[18:33:32] <keenerd> Well yeah. It is german. The stuff never wears out and never needs replacing. :-)
[18:37:51] <amee2k> and sometimes i think the old farts keep hanging on to the good stuff until it goes to the trash and never sees the used market anymore
[18:38:29] <amee2k> anyway, i'm off to bed now
[18:38:41] <OndraLappy> gn
[20:04:33] <Tom_itx> lo abcminiuser
[20:46:12] <abcminiuser> Yahoyoy
[20:46:47] <Tom_itx> hey
[20:47:53] <Tom_itx> rue's frustration was over a typo
[20:48:17] <Tom_itx> it was looking at PIND instead of PINB
[20:55:57] <abcminiuser> Ah
[20:56:08] <abcminiuser> Well, at least I got a bit of practise in
[20:56:19] <abcminiuser> Currently packing, not a whole lot of room for three months worth of stuff...
[20:57:06] <Tom_itx> you gonna be doin the same thing when you return?
[20:57:22] <abcminiuser> I'll chuck most of my clothes, get new ones when I get back
[22:44:46] <CapnKernel> abcminiuser: I know all about the packing for 3 months caper :-)
[22:46:19] <CapnKernel> I find coming home is lighter. Four days ago I was 13.5kg all up :-)
[22:48:42] <abcminiuser> Yeah, not taking everything back
[22:48:57] <abcminiuser> But I've filled my suitcase and I've still got tons of stuff I need to take
