#avr | Logs for 2012-11-15

Back
[02:44:23] * RikusW found a 0.1F 5v5 cap :)
[02:44:27] <RikusW> 100mF
[02:44:41] <OndraSter> I have got here bunch of ollllld 50mF ones
[02:44:44] <RikusW> on the pcb of a car radio
[02:44:59] <RikusW> where did you get it /
[02:49:06] <OndraSter_> damnit
[02:49:09] <OndraSter_> <OndraSter> I took apart some old equipment :D
[02:49:10] <OndraSter_> <OndraSter> at school
[02:49:12] <OndraSter_> <OndraSter> so I kept the bigass caps
[02:49:14] <OndraSter_> and some rectifiers
[02:49:18] <OndraSter_> I found that one of them is blown :D
[02:49:42] <OndraSter_> rectifiers = diodes in this case
[03:18:37] <darsie> hi
[05:10:48] <OndraSter> I'LL RAPE SOMEBODY
[05:10:56] <OndraSter> when USB Host gives you "wLength" of FF 00
[05:10:59] <OndraSter> (aka 256 bytes)
[05:11:05] <OndraSter> you can still reply with 64B max
[05:11:11] <OndraSter> since you are USB Full Speed device
[05:11:12] <OndraSter> well
[05:11:14] <OndraSter> I didn't realize that
[05:11:25] <elektrinis> why rape
[05:11:33] <OndraSter> because my adrenalin level is up the roof
[05:12:30] <elektrinis> I will withhold all further questions
[05:12:53] <OndraSter> :D
[05:15:19] <OndraSter> the worst and saddest part is that xmega DOES have the multi packet transfers BUILT IN
[05:15:22] <OndraSter> on hardware level
[05:15:27] <OndraSter> and I was smart enough to not enable them :X
[05:33:00] <OndraSter> so appearantly Win8 x64 REQUIRES signed certs
[05:33:04] <OndraSter> Win7 x64 just shows big red warning
[05:33:19] <OndraSter> self signed certs when you import them to cert store should be enough
[05:36:57] <MrTrick> anyone know of a 'button' library? Not just debouncing, basically something to register callbacks and trigger events for 'press', 'hold', 'hold-repeat', 'click', etc?
[06:12:03] <OndraSter> there we go
[06:12:04] <OndraSter> all signed
[06:12:07] <OndraSter> self signed
[06:12:10] <OndraSter> big red warning but installs
[06:14:53] <OndraSter> unknown setup packet: 21 22 00 00 01 00 00 00
[06:14:53] <OndraSter> FIFO event: ep: 05
[06:14:54] <OndraSter> awesome
[06:15:06] <OndraSter> now what was EP5... notification channel? :D
[07:58:50] <OndraSter_> megal0maniac_afk, DOODE
[07:58:52] <OndraSter_> I HAVE JUST
[07:58:53] <OndraSter_> REcEIVED
[07:58:56] <OndraSter_> first CHARACTER
[07:58:57] <OndraSter_> over MY
[07:58:59] <OndraSter_> COMPLETELY MINE!
[07:59:01] <OndraSter_> usb cdc stack.
[07:59:27] <OndraSter_> http://pastebin.com/KSg49rd7
[07:59:28] <OndraSter_> last line
[08:02:50] <inkjetunito> congrats
[08:02:54] <Tom_itx> :)
[08:03:08] <OndraSter_> thanks.
[08:03:15] <OndraSter_> only one ripped USB connector is the victim
[08:03:20] <OndraSter_> and about 20 hours of raw time
[08:04:10] <inkjetunito> :O well done. i've zero clue about usb
[08:04:28] <inkjetunito> Tom_itx: how is updating the website ;)
[08:04:29] <OndraSter_> :D
[08:04:49] <Tom_itx> h
[08:04:50] <Tom_itx> oh
[08:04:56] <Tom_itx> i can do that i suppose
[08:05:14] <inkjetunito> Tom_itx: please do :)
[08:06:02] <Tom_itx> you order soon i can drop it on my way out this morning
[08:06:37] <Tom_itx> leaving in 30
[08:07:37] <inkjetunito> Tom_itx: yeah, i've about 10 mins. can do :)
[08:09:54] <Tom_itx> just about got it
[08:11:09] <Tom_itx> ok
[08:15:12] <inkjetunito> Tom_itx: excellent & done :)
[08:32:44] <Tom_itx> you should have a confirmation
[08:44:16] <Malinuss> I'm getting confused here. So can I run a overflow timer on timer/counter0 but at the same time use timer/counter0 for PWM? So I can just write delay(x); pwm(pinX, 0-255); delay(x); pwm(pinX, 0-255); . Using this one timer?
[08:54:28] <OndraSter_> ask arduino what timer and how do they use it for the internal delay
[08:55:32] <OndraSter_> RikusW! RikusW! RikusW! RikusW!
[08:55:38] <OndraSter_> I ditched atmel's CDC yet again
[08:55:41] <OndraSter_> and got back to mine
[08:55:45] <OndraSter_> I am receiving serial data :)
[08:56:12] <RikusW> good :)
[08:56:30] * RikusW just wrote some NMEA parsing code
[08:56:56] <OndraSter_> :)
[08:59:11] <karlp> that's a large large club
[09:03:13] <darsie> Can I use an interrupt just for wake up and without actual interrupting and calling the ISR?
[09:06:14] <RikusW> darsie: don't think so
[09:06:28] <darsie> ok. I thought so, too.
[09:06:57] <RikusW> you could add a global variable for immediately returning...
[09:07:11] <darsie> Or just a reti.
[09:07:14] <RLa> anyone built voltage regulator using avr?
[09:07:33] <RLa> i mean multi-phase synchronous buck regulator
[09:08:39] <RikusW> darsie: if that ISR won't do anything usefull except wake the AVR reti is just fine
[09:08:54] <darsie> I thought about using the atmega128 in my geiger counter to generate the signals for the 400 V flyback regulator, but I'm not sure if that's gonna save power over the 7555 I'm using right now.
[09:09:10] <darsie> RLa:
[09:09:40] <RLa> do you have feedback loop too?
[09:09:44] <darsie> Yes.
[09:10:11] <darsie> 2 200 V zeners to a transistor pulling the 7555 reset low.
[09:10:15] <RLa> what type of feedback it is, full-featured PID controller or something simpler?
[09:10:18] <RLa> oh
[09:11:14] <RLa> i assume you only have single phase?
[09:11:40] <darsie> I'm not sure what you mean by phase. It's only the 400 V output.
[09:11:41] <RLa> hm, flyback, not sure what multi-phases would mean there :)
[09:12:27] <RikusW> aask flyback :-P
[09:12:44] <darsie> I used to rectify the primary, to get 25 V to drive the piezo speaker, but that part no longer exists.
[09:12:46] <RLa> for buck regulators, multi phase is multiple pwm outputs + switching fets + coils supplying a single cap pack
[09:13:11] <RLa> pwm signals are phase-shifted, like 180 degrees when you have 2 phase
[09:13:29] <RLa> modern high-end cpu regulators use 12 phases
[09:13:49] <RLa> it allows to spread load over multiple fets and coils
[09:14:20] <darsie> Hmm, that also lowers ripple, right?
[09:14:24] <RLa> yes
[09:15:07] <darsie> So it's actually several phase shifted bucks?
[09:15:08] <RLa> and therefore needs smaller caps i think
[09:15:30] <RLa> darsie, almost, i think they need single feedback loop tho
[09:15:46] <darsie> But independent inductors?
[09:15:52] <RLa> load spreading would not be even with separated loops
[09:16:01] <RLa> not sure but i think you need
[09:16:15] <RLa> considering the current waveform in the coil
[09:16:49] <darsie> Do they use active rectification?
[09:17:34] <RLa> if you mean fet to freewheel (bottom half of output) then yes
[09:18:16] <darsie> Umm, not sure. I still have my flyback in mind :).
[09:19:29] <RLa> oh, buck converter is not isolated at all
[09:19:56] <RLa> coil-cap and output is basically low-pass filter
[09:20:01] <RLa> at*
[09:20:11] <RLa> here is IR's whitepaper: http://www.irf.com/technical-info/whitepaper/3phasepwm.pdf
[09:20:23] * darsie is back coding his geiger counter ...
[09:28:01] <Malinuss> haha are you kidding me? ask arduino? what?
[09:28:54] <Malinuss> The point in using arduino is, that you don't know which timer you use, you just write delay(x); and it waits etc.
[09:28:59] <Malinuss> lol
[09:30:17] <jadew> you don't need a timer to delay
[09:30:35] <darsie> Hmm, the unconnected port pins on an atmega168 still exist on silicon and if left floating as input could draw current, right?
[09:32:32] <asteve> holy banana apples!
[09:33:01] <RLa> darsie, enable pull-ups on them
[09:35:16] <darsie> Hmm, it's just PC7 missing, but that doesn't seem to exist, actually.
[09:35:39] <Malinuss> jadew, no, but a timer will provide the most exact delay. Also it's easier then having to figure out how many cycles each function takes etc... Not my question at all though, hehe
[09:35:56] <Malinuss> jadew, also, it's kinda like saying, that you don't need interrupts/timers for PWM
[09:36:30] <jadew> you don't need interrupts
[09:36:34] <jadew> and it's not like saying that
[09:36:51] <jadew> you can get perfect delay with out timers
[09:37:27] <darsie> With careful planning ... each time you modify the code.
[09:37:34] <darsie> more or less.
[09:38:07] <jadew> well, once you write the function/macro that does that, you don't have to modify it
[09:38:25] <Malinuss> jadew, still, just much easier to use a timer
[09:38:56] <darsie> Well, I had a 1 s main loop. When I modified the code the loop period would change.
[09:39:09] <jadew> Malinuss, that's wasted resources
[09:39:48] <darsie> Now I'm using timer 2 with 32 kHz crystal and enter power save sleep. 8 uA :).
[09:40:12] <darsie> instead of 5 mA.
[09:40:25] <jadew> ah, that's perfectly ok
[09:40:33] <jadew> but if you don't care about power, there's no reason to waste the timer
[09:40:44] <Malinuss> well my question was: If I have a "delay timer" (overflow timer), running already. I can I use a PWM in a gate that uses that same timer/counter?
[09:40:59] <jadew> Malinuss, yes
[09:41:39] <Malinuss> so I can have the timer/counter0 running a overflow interrupt, at the same time, as it runs PWM for OC0A?
[09:41:48] <Malinuss> (just wanna make sure :) )
[09:41:55] <jadew> yep, should be possible
[09:42:52] <Malinuss> okay, cool. I'll try to make the PWM work by itself first though..
[11:17:12] <AR_> hello
[11:18:22] <RLa> hey
[11:39:25] <OndraSter> hihi
[12:05:22] <megal0maniac> hi
[12:05:58] <megal0maniac> my avr won't turn on
[12:06:17] <OndraSter> you broke it!
[12:06:22] <megal0maniac> :(
[12:06:23] <darsie> wrong clock fuses?
[12:06:56] <megal0maniac> there aren't any fuses, its just a chip
[12:07:08] <darsie> read the datasheet.
[12:07:22] <megal0maniac> it didn't come with any papers..
[12:07:27] <megal0maniac> wheres that?
[12:07:30] <darsie> Download it.
[12:07:38] <darsie> www.atmel.com
[12:07:46] <OndraSter> stop trolling :P
[12:07:52] <megal0maniac> I'm sorry darsie :)
[12:08:29] <megal0maniac> OndraSter: Do you know if Rikus has done anything exciting with the xboard?
[12:08:36] <OndraSter> no idea
[12:08:48] <OndraSter> but I got my usb cdc working
[12:08:53] <megal0maniac> And xide?
[12:09:35] <OndraSter> havent worked on that today
[12:09:40] <OndraSter> usb first
[12:11:17] <megal0maniac> What, so usb stack?
[12:11:25] <megal0maniac> How big? :)
[12:12:34] <OndraSter> 2.5kb without tx code
[12:12:46] <OndraSter> will be cleaned by hand
[12:13:14] <megal0maniac> "Rip stuff out until it stops working, then put it back and carry on ripping" ? :P
[12:13:55] <megal0maniac> Heh. Speaking of... :)
[12:14:00] <RikusW> me ?
[12:14:18] <megal0maniac> megal0maniac> "Rip stuff out until it stops working, then put it back and carry on ripping" ? :P
[12:14:25] <RikusW> yep
[12:14:29] <megal0maniac> Just before you arrived
[12:14:41] <megal0maniac> OndraSter is getting there
[12:14:43] <RikusW> thats how I got CDC to fit in 800b
[12:14:55] <megal0maniac> I know, I was quoting you :)
[12:15:07] <RikusW> :)
[12:15:19] <RikusW> got my NMEA parsing code mostly working
[12:15:35] <megal0maniac> Got Debian working on my phone :P
[12:15:42] <RikusW> ye haerd
[12:15:47] <megal0maniac> With avr-gcc
[12:16:07] <megal0maniac> And can connect with VNC from a desktop machine, so I have a half-decent screen res
[12:17:15] <megal0maniac> Unfortunately, since I'm using chroot, I'm limited to the phone kernel. And it says NO to CDC. avrispmkii doesn't work either, but I didn't really look at what dmesg said about it, so might be work-aroundable
[12:17:47] <RikusW> load a driver ?
[12:17:49] <Casper> has anyone here made a DAC with an avr with PWM or r2r for slow speed? (basically 1sps is enought) and what precision did you got??
[12:18:53] <RLa> Casper, i though about making one with pwm
[12:19:12] <RLa> but only finished dc amp that works down to bottom rail
[12:20:01] <RLa> you basically need a low-pass filter and dc amp which takes load away from the filter
[12:23:44] <Casper> basically I really need 1.8-2.5V output ideally with 0.01V precision
[12:23:46] <OndraSter> i am at school atm...
[12:24:05] <OndraSter> real dac? :P
[12:24:24] <OndraSter> 12bit at 2.5vref...
[12:24:54] <Casper> I wonder how precise a 16 bits pwm would be...
[12:24:58] <OndraSter> 0.6mV resolution
[12:25:15] <RLa> Casper, i used simetrix to simulate the low-pass filter and dc amp
[12:25:37] <RLa> you will have small pwm noise on top of dc
[12:28:24] <Casper> 16 bits would be 122Hz at 8MHz o.O
[12:28:40] <Casper> I guess 1k/100u would be ok?
[12:28:53] <RLa> simulate it
[12:30:24] <Casper> when I get home, but for now I have more important things to do: finish to hack that psu :D
[12:31:02] <OndraSter> dem haxes
[12:31:07] <Casper> I'm modding an atx psu so it have 9-16VDC input instead of it's 340VDC :D
[12:32:41] <RikusW> you want low volt input for the smps ?
[12:33:04] <xjd> uhm.. what for?
[12:33:06] <Casper> I want 12V SLA or flooded input
[12:33:25] <RikusW> running it of a battery ?
[12:33:31] <Casper> yup
[12:33:40] <Casper> need buck and boost
[12:34:03] <Casper> which is a pita, the hack seems to be the easiest way beside turning to ebay
[12:34:07] <RikusW> why not boost the battery to 350Vdc ?
[12:34:15] <OndraSter> hooooooome
[12:34:40] <xjd> or connect 233x1.5 batteries in series?
[12:34:50] <xjd> or connect 233x1.5v batteries in series?
[12:35:07] <RikusW> xjd: use 12V car batteries instead...
[12:35:14] <xjd> :p
[12:35:36] <RikusW> wonder how big a spark will fly if shorting them...
[12:35:57] <RikusW> maybe vaporize the entire terminal :-P
[12:36:16] <xjd> or even the whole airport *laughing*
[12:36:30] <Casper> RikusW: I actually tought of boosting to 16 and bucking to 5 and 12, but decided against it :D
[12:37:13] <Casper> and if this hack fail, then I'm seriously considering to get an openups... but I don't like some of the missing features...
[12:37:36] <Casper> and the specs that appear to be boosted
[12:38:04] <Casper> they sell it as 10A, but 6A continuous if you keep the hottest part bellow 60C...
[12:38:20] <RikusW> heh
[12:38:28] <RikusW> PMPO ups ;)
[12:39:51] <Casper> yeah 160W PMPO usb powered speakers
[12:40:20] <RikusW> 160W for a nanosecond
[12:41:20] <Casper> Imax * Vmax *k = PMPO
[12:41:44] <RikusW> k = ?
[12:42:01] <Casper> where 1<=k>=infinite
[12:42:43] <Casper> and which k has been randomly selected from a random number of dice of random number of faces
[12:43:11] <RikusW> parse error :-P
[12:43:28] <Casper> pmpo have no real meaning
[12:43:39] <RikusW> 1<=k<=infinite
[12:43:47] <Casper> woops
[12:43:53] <Casper> where 1=k>=infinite
[12:44:01] <Casper> where 1>=k>=infinite
[12:44:10] <RikusW> errr
[12:44:14] <Casper> in other words, k is atleast 1
[12:44:45] <RikusW> your formulas go from bad to worse :-d
[12:45:15] <Casper> so yeah, if the ampli can give a max of 10V and a max of 1A, it's technically 10W, but the magic k can give it a 1000W figure
[12:45:35] <RikusW> at what THD ?
[12:45:42] <Casper> <no
[12:45:58] <Casper> thd have nothing to do with pmpo
[12:46:11] <Casper> pmpo is really a random number
[12:46:48] <specing> Casper: 1 is greater or equal to infinite? Oh?
[12:47:14] <Casper> k is atleast 1, and can be higher than infinite :D
[12:47:33] <Casper> oh my first k was right....
[12:47:36] <specing> oh?
[12:47:46] <specing> Can you prove k is larger than infinite?
[12:47:48] <Casper> no it wasn't
[12:48:01] <Casper> yes it was...
[12:48:15] <Casper> specing: sure, k is pmpo :D enought said :D
[12:48:36] <RikusW> PMPO == infinite THD ;)
[12:55:42] <RLa> Casper, you need that atx for a computer?
[12:56:02] <Casper> no
[12:56:14] <RLa> why not a push-pull converter that would give all needed voltages
[12:56:52] <Casper> network gear, which start to become important... modem, router, switch, voip ata, 2 phone bases
[12:56:55] <specing> Casper: I'd still like to see the proof that k is larger than infinity
[12:57:34] <Casper> ok then... if you insist... 1<=k<=infinite
[12:57:36] <Casper> better?
[12:58:24] <RikusW> yep
[13:01:35] <Casper> RLa: I failed at making my own push pull, so hacking the push pull of the atx psu
[13:01:54] <RikusW> rewinding the transformer ?
[13:02:06] <RLa> technically it's not a push-pull but half-bridge
[13:02:11] <Casper> basicallu, I have issue with too many factors, so I will eliminate all the control and output
[13:02:32] <Casper> but yes the psu WAS half bridge, I rewind for push pull
[13:02:38] <Casper> and will redo the drive circuit
[13:02:47] <Casper> so a minimal hack
[13:03:00] <RLa> hm, how you dismantled the transformer?
[13:03:04] <RikusW> how did you manage to open the ferrite core without breaking it /
[13:03:23] <Casper> I did not open the core
[13:03:37] <RLa> you can do that with strong liquids but i think they solve wire insulation too
[13:04:06] <Casper> you can boil the core for 15-20 mins and the glue will soften enought
[13:04:26] <RikusW> I actually managed to open a VGA screen power transformer, it wasn't glued
[13:04:30] <RikusW> and rewound it
[13:04:41] <Casper> my last attempt was not that bad actually
[13:04:58] <Casper> but things heated up more than what it was supposed to, by a good margin...
[13:05:44] <RLa> RikusW, aren't those flybacks and with an air gap?
[13:06:38] <RikusW> maybe
[13:06:56] <RikusW> I built a 12v to 350Vdc stepup
[13:07:12] <RikusW> had to use a voltage multiplier too
[13:07:15] <Casper> I succeded to make the output inductor heat more than the switch :D
[13:07:45] <Casper> oh 2pm I can eat!
[13:08:07] <RLa> anyone ordered stuff from here: http://powermagnetics.co.uk/pace-components/ferrite-cores
[13:08:09] <Tom_itx> leftover pizza
[13:08:35] <RLa> they do small quantity orders and are not very expensive
[13:08:40] <RikusW> Casper: GMT -5 ?
[13:08:53] <Tom_itx> casper is a canadian ghost
[13:09:13] <Casper> mm
[13:10:51] <Casper> but really, I cheat and make a push pull, by only doing the switch circuit and the transfo
[13:11:51] <Casper> the control and all the touchy parts should be handled by the already made circuit... I may however have to hack the proterction circuit
[13:12:09] <RLa> i'm going to get a new wheelset for my bike tomorrow :)
[13:14:31] <RLa> brb
[13:40:30] <darsie> I don't get pin change interrupts during power save sleep mode. I get them in active mode. atmega168.
[13:42:08] <darsie> Datasheet says: Pin change interrupts on PCINT23..0 are detected asynchronously. This implies that these interrupts can be used for waking the part also from sleep modes other than Idle mode.
[13:49:19] <OndraSter_> yeah I found that issue myself too
[13:49:22] <OndraSter_> not sure if I did something wrong
[13:49:24] <OndraSter_> or the datasheet lies
[13:49:39] <OndraSter_> I tried it on tiny13a and simulator though
[13:49:54] <darsie> did it work there?
[13:49:56] <OndraSter_> no
[13:51:39] <Casper> maybe you used the wrong sleep mode
[13:52:20] <darsie> set_sleep_mode(SLEEP_MODE_PWR_SAVE);
[13:53:18] <darsie> PCINT should work in any mode.
[14:00:35] <darsie> I'll write a small test program ...
[14:02:12] <Malinuss> darsie, "set_sleep_mode" is that a standard avr-c function :o?
[14:02:22] <darsie> yes
[14:02:27] <darsie> avr-libc
[14:14:07] <OndraSter_> RikusW, so, in my USB CDC I have got last issue - when I want to send some data, I put them into the IN (TX UART) endpoint and clear BUSNACK0 flag
[14:14:10] <OndraSter_> the data are sent
[14:14:35] <OndraSter_> but it does not ask for any more packets?!
[14:14:47] <OndraSter_> actually it asks for one more data
[14:14:52] <OndraSter_> so I set the CNT again
[14:14:59] <OndraSter_> and clear BUSNACK
[14:15:09] <OndraSter_> but I don't get FIFO interrupt again
[14:15:19] <OndraSter_> and the data is not transferred anymore ?!
[14:19:52] <OndraSter_> I like how atmel's udi_cdc.c is 1200 lines long
[14:20:05] <OndraSter_> my usb.c that contains both usb and cdc is only 500 lines :D
[14:35:11] <darsie> Hmm, the small program get's the interrupts in power save mode ...
[14:35:57] * DagoRed scratches head
[14:36:16] <DagoRed> Is this a programmer problem I'm having or a software problem?
[14:46:01] <RikusW> OndraSter_: do you send a short packet too ?
[14:57:45] <amee2woof> mmh... anyone got a good pointer for looking for replacement LCD panels?
[14:58:12] <amee2woof> i've got a monitor with a cracked LG LC260WXE-SBB1 panel
[14:58:28] * specing points to china
[14:58:47] <amee2woof> lol yeah
[14:58:55] <amee2woof> but where would i look for ordering these
[14:59:10] <amee2woof> just googling for "<type number> replacement" doesn't really go anywhere
[15:00:14] <amee2woof> i can buy the LVDS controller board thingy for 16$ but i can't find panels
[15:00:43] <amee2woof> http://ompldr.org/vZ2I3eA/CIMG0128b.jpeg << the electronics seem to be fine, only physical damage to the panel itself
[15:01:43] <OndraSter_> RikusW, why ZLP?
[15:02:07] <OndraSter_> when I get FIFO on ep3 (the one that is transmit MCU to PC of UART) I should ZLP it?
[15:03:04] <OndraSter_> hmm
[15:03:08] <OndraSter_> I missed the lesson about short packets
[15:03:57] <darsie> Looks like the ISR doesn't get called when the interrupt occurs during sleep.
[15:04:49] <darsie> Maybe the pin pulse is too short.
[15:06:08] <darsie> Trying 8 MHz ...
[15:09:34] <OndraSter_> RikusW, even with multipacket and autozlp enabled it does not work..
[15:10:49] <specing> CVS/SVN repos D:
[15:11:07] <specing> Can't believe avr-libc still uses SVN
[15:14:56] <RikusW> OndraSter_: you must be overlooking something...
[15:15:31] <OndraSter_> RikusW, yeah
[15:15:32] <OndraSter_> but what? :(
[15:16:35] * RikusW just finished assembling a perfboard for m644pa
[15:16:43] <OndraSter_> RikusW,
[15:16:44] <RikusW> with 40pin socket
[15:16:45] <OndraSter_> http://pastebin.com/uYhVtVNN
[15:16:47] <OndraSter_> this is the source code
[15:18:42] <specing> 8.4M avr-libc.git
[15:18:42] <specing> 44M avr-libc.repo
[15:18:43] <specing> 324M trunk
[15:18:53] <specing> 2) git svn of the repo
[15:19:00] <specing> 3) git checkout of the said repo
[15:19:10] <specing> 1) git bare repo of (3)
[15:19:32] <specing> DAFUQ DOES SVN NEED 44MB FOR?!
[15:20:28] <specing> No wait, 2) is a rsync copy of the original SVN repo
[15:20:38] <specing> the DAFUQ stays though
[15:21:50] <RikusW> OndraSter_: will have a look later, I'm not quite familiar with xmega yet
[15:22:09] <RikusW> + I need to get the GPS stuff working before monday ;)
[15:35:34] <specing> /* Handle unexpected interrupts (enabled and no handler), which
[15:35:34] <specing> usually indicate a bug. Jump to the __vector_default function
[15:35:35] <specing> if defined by the user, otherwise jump to the reset address
[15:35:47] <specing> crt1/gcrt1.S :)
[15:47:57] <OndraSter_> RikusW, sure, thanks
[15:53:38] <RikusW> just tested SPI en JTAG on my new m644pa board :)
[15:53:56] <RikusW> changed the fuses using my U2S jtag on m644 :)
[15:54:00] <RikusW> and RavrProg
[15:54:20] <RikusW> Qt
[15:55:26] <RikusW> m644 got to uarts, so it will go in between the GPS and GPRS
[15:56:46] <OndraSter_> :)
[15:57:25] <RikusW> I'll be using the SPI connected to the U2S for a third "uart" to the PC
[15:58:01] <RikusW> and can connect the dragon to jtag
[15:58:23] <RikusW> If I had my U2S PC software working I could use that for jtag too
[15:58:30] <RikusW> it should work with m644 too
[15:59:05] <specing> Oh and the u64 3000b code is only 350b on GCC gcc version 4.8.0 20121109 (experimental) (GCC)
[16:00:01] <RikusW> seems like specing is cutting edge :)
[16:00:06] <RikusW> just don't get cut ;)
[16:00:51] <specing> Anyone wants the package with it? :)
[16:03:54] <jadew> ok... so I decided to go part smd and it seems that smd component orders tend to be epic
[16:04:23] <jadew> I haven't even finished adding stuff to my chart and I already have more than 30-40k components in there
[16:04:31] <jadew> *cart
[16:05:09] <specing> 80MB
[16:07:04] <specing> Wow, 70b for a PORTB ^= PORTB; program
[16:07:16] <specing> haven't in my life seen such small gcc programs
[16:08:11] <OndraSter_> :D
[16:08:14] <OndraSter_> no interrupt jump table?!
[16:08:21] <OndraSter_> jadew, 30-40k ?"
[16:08:23] <OndraSter_> on digikey maybe
[16:09:09] <jadew> OndraSter_, no, I don't shop from digikey
[16:09:12] <specing> ++PORTB; 68 bytes
[16:09:28] <OndraSter_> jadew, where from?
[16:09:32] <jadew> tme.eu
[16:09:55] <OndraSter_> oh
[16:10:00] <OndraSter_> well digikey has ridiculously low prices on passives
[16:10:06] <OndraSter_> mouser has ridiculously high prices on passives
[16:10:13] <OndraSter_> caps and resistors
[16:10:33] <jadew> I'll have to check digikey then
[16:10:47] <jadew> I suspect this order is going to be quite expensive
[16:11:11] <OndraSter_> heh
[16:11:17] <OndraSter_> I am getting 100 parts usually
[16:11:24] <OndraSter_> the next price breaker is at 500 or 1k after that
[16:11:28] <OndraSter_> and it is not that big difference anymore
[16:11:53] <jadew> yeah, I got 100 and 200 packs
[16:12:22] <jadew> didn't meet the discount on any of them
[16:12:24] <OndraSter_> how many different parts do you have ?!
[16:12:43] <jadew> well, I went for all the values :D
[16:13:03] <OndraSter_> duh?
[16:13:07] <jadew> like 100 values for resistors
[16:13:11] <OndraSter_> huh
[16:13:12] <OndraSter_> what for?
[16:13:17] <OndraSter_> I just buy them when I need them
[16:13:20] <OndraSter_> much easier
[16:13:27] <yunta> lol
[16:13:40] <OndraSter_> I order 10 boards of something
[16:13:46] <jadew> yeah, but it sucks waiting for a week to just so you can solder a few missing parts
[16:13:58] <OndraSter_> so I see "okay, I need 20 resistors of 49R9. The best value goes for me at 100. I will get 100"
[16:13:59] <OndraSter_> huh
[16:14:10] <yunta> jadew: do you take them all in 1206 *and* 0805 ?
[16:14:16] <OndraSter_> *and* 0603 :)
[16:14:20] <jadew> only 1206
[16:14:23] <OndraSter_> duh
[16:14:24] <OndraSter_> useless :)
[16:14:30] <jadew> why?
[16:14:32] <yunta> practically :)
[16:14:35] <yunta> too big
[16:14:38] <OndraSter_> everything is 0805/0603 nowadays
[16:14:41] <OndraSter_> 0805 for the BIGGER parts
[16:14:47] <RikusW> 0402 ?
[16:14:49] <OndraSter_> such as higher capacity/voltage ceramic caps
[16:14:57] <OndraSter_> 0402 is on the edge of calling it dirt for me
[16:15:02] <OndraSter_> I found 0603 to be the best :D
[16:15:03] <jadew> I don't want to spend my whoe life soldering a freaking resistor
[16:15:07] <jadew> *whole
[16:15:33] <OndraSter_> jadew, I can build one xboard by hand soldering it in a few minutes and it has got one QFN, many passives
[16:15:38] <yunta> jadew: 0805 are still easy to solder. 1206 are just ridiculously huge
[16:15:52] <jadew> yunta, hmm
[16:16:03] <OndraSter_> aye
[16:16:09] <OndraSter_> I am doing nowadays everything I can in 0603
[16:16:13] <RikusW> 1210 :)
[16:16:19] <jadew> I'll have to find some on a few boards and see for myself
[16:16:24] <OndraSter_> 1210 is my 10uF/35V cap now
[16:16:24] <RikusW> *12:10
[16:16:27] <jadew> I just checked the sizes with a ruller
[16:16:35] <OndraSter_> :D
[16:16:36] <jadew> 1206 seemed decent
[16:16:39] * yunta lives in finland - no readily available <0805 res :(
[16:16:49] <OndraSter_> yunta, digikey/mouser/farnell?
[16:16:53] <RikusW> its 1204 here ;)
[16:16:58] <OndraSter_> 2304 here.
[16:17:07] <OndraSter_> I use 1808 once even!
[16:17:12] <OndraSter_> for 2000pF/2kV cap :P
[16:17:20] <yunta> I do digikey when I know what to buy. but for prototype work digikey shipping from us takes too much
[16:17:29] <yunta> time
[16:17:34] <OndraSter_> time or shipping $$?
[16:17:41] <OndraSter_> time = fedex.
[16:17:48] <yunta> time
[16:17:54] <OndraSter_> $$ = $60 for shipping for <$200 which is ridiculous
[16:17:54] <OndraSter_> oh
[16:18:00] <yunta> see, last package arrived from us to finland next day
[16:18:05] <yunta> and then took one week to get to me
[16:18:07] <OndraSter_> eh
[16:18:09] <jadew> damn you guys, now I'm not sure if I should keep going with this order :/
[16:18:10] <OndraSter_> fail
[16:18:29] <OndraSter_> jadew, stop getting 100 different values, just get the classic pack of 20 or so
[16:18:35] <OndraSter_> there are thousands of values :P
[16:18:47] <OndraSter_> you will *always* end up needing something specific
[16:18:50] <yunta> true
[16:18:51] <OndraSter_> like 49r9 I need
[16:18:59] <OndraSter_> for ethernet
[16:19:19] <yunta> just a basic pack for prototypes, and then buy more when you have proper, tested pcb design
[16:19:31] <jadew> told you, I hate waiting for components, not to mention that shipping adds up
[16:19:36] <OndraSter_> confession bear: I always expect my boards to work on the first try
[16:19:46] <OndraSter_> jadew, farnell = 5 GBP or so shipping
[16:19:48] <OndraSter_> at least to here
[16:20:00] <OndraSter_> reasonable price
[16:20:03] <jadew> 5gbp is expensive if you need something twice a week
[16:20:05] <OndraSter_> unlike mouser/digikey
[16:20:07] <OndraSter_> heh
[16:20:11] <yunta> OndraSter_: same here. but on my last one I counted 12 design mistakes :)
[16:20:18] <OndraSter_> yunta, I had only one :P
[16:20:27] <OndraSter_> it was small PCB though
[16:20:36] <OndraSter_> I know about one mistake on my new board - I forgot i2c pullups.
[16:20:40] * jadew goes to digg up some boards to figure out if 1206 is indeed too big
[16:21:07] <OndraSter_> I hope that I wired up the UART-SPI correctly
[16:21:10] <yunta> OndraSter_: can't you use atmegas internal pullups for that?
[16:21:14] <OndraSter_> too weak
[16:21:17] <OndraSter_> those are 50k or so
[16:21:22] <OndraSter_> i2c uses 4k7 or 10k
[16:21:25] <OndraSter_> it does not matter really
[16:21:30] <OndraSter_> those are prototypes :)
[16:21:37] <OndraSter_> revA = prototype for me :D
[16:21:47] <OndraSter_> as long as I don't lose money on it
[16:21:50] <OndraSter_> which I have so far
[16:21:59] <OndraSter_> I ordered too little parts :D
[16:23:00] <OndraSter_> I built 3 prototypes... two sent to south africa and one is here
[16:23:36] <yunta> is it... confidential thing?
[16:23:41] <OndraSter_> no
[16:23:44] <OndraSter_> http://myxboard.net
[16:23:46] <OndraSter_> this
[16:23:52] <OndraSter_> the thing I am working on the USB CDC right now
[16:23:53] <yunta> that's yours?
[16:23:56] <OndraSter_> yes
[16:24:12] <yunta> beautiful !
[16:24:18] <OndraSter_> heh thanks
[16:24:28] <OndraSter_> now if anybody knew what the hell is going on with the USB...
[16:24:37] <OndraSter_> why it does not send IN packet for second data payload..
[16:24:40] <OndraSter_> and third
[16:24:42] <OndraSter_> and fourth
[16:24:55] <yunta> toggle
[16:25:18] <OndraSter_> I will grab some older laptop with XP I have got here
[16:25:22] <OndraSter_> and use some USB sniffer
[16:25:29] <yunta> or not, wrong toggle handling usually results in every-second packet received
[16:25:38] <OndraSter_> what toggle?
[16:25:42] <yunta> atxmega?
[16:25:52] <OndraSter_> the toggle is done internally automatically
[16:26:05] <OndraSter_> and for some (for me unknown) reason it checks BUSNACK0 always, never BUSNACK1
[16:26:14] <OndraSter_> toggle is relevant only for ping pong mode
[16:26:21] <yunta> unless you use protocol that forces it to something (like during enumeration)
[16:26:32] <yunta> busnack1 is only used in pingpong
[16:26:38] <yunta> iirc
[16:26:38] <RikusW> yunta I got one coco
[16:27:43] <RikusW> xboard coco
[16:28:24] <RikusW> connected it to my dragon to see if xmega PDI finally works...
[16:28:24] <yunta> does it work? :D
[16:28:27] <RikusW> yes
[16:28:32] <yunta> nice
[16:28:39] <RikusW> haven't written anything for it yet
[16:28:48] <yunta> pdi for programming or something more?
[16:28:52] <RikusW> I did try ordering xmegas through RS
[16:28:56] <OndraSter_> RikusW, on your board I tried the PDI actually before I sent it off, I did not on megal0maniac's :D
[16:29:03] <RikusW> program & debug interface
[16:29:36] <RikusW> yunta: I hacked the dW protocol
[16:29:43] <RikusW> planning to do the same to pdi
[16:29:44] <OndraSter_> DEM HAXXORZ
[16:29:59] <yunta> what is dw?
[16:30:07] <RikusW> debugWire
[16:30:10] <yunta> ah
[16:30:36] <specing> 58 bytes :D
[16:30:38] <RikusW> I'll connect my U2S board in slave synchronous uart to PDI
[16:30:49] <RikusW> actually only the RX and XCLK lines
[16:30:54] <RikusW> and eavesdrop
[16:31:01] <RikusW> thats how I did dW
[16:31:32] <yunta> OndraSter_: "toggle is relevant only for ping pong mode" <- I think you meant "BANK" in STATUS
[16:31:42] <OndraSter_> well you started with "toggle" :P
[16:31:46] <OndraSter_> and yes, I ment BANK
[16:31:55] <OndraSter_> actually it is TOGGLE bit
[16:31:59] <OndraSter_> for data0/data1
[16:32:02] <RikusW> yunta: http://sites.google.com/site/megau2s/home
[16:32:21] <RikusW> http://www.ruemohr.org/docs/debugwire.html
[16:32:31] <yunta> what I meant is that busnack1 is used when bank=1, not when toggle=1 (but just guessing)
[16:32:40] <OndraSter_> ah
[16:32:49] <RikusW> OndraSter_: ask abcminiuser whenever you see him again
[16:33:11] <RikusW> or have a really close look at the Atmel code
[16:33:46] <OndraSter_> the atmel code is impossible to read
[16:33:51] <OndraSter_> and from what I read it does the same thing as I do
[16:34:00] <jadew> http://dumb.ro/files/IMG_0064.JPG
[16:34:01] <OndraSter_> (note that the atmel code didn't work for me actually :D)
[16:34:02] <RikusW> almost...
[16:34:10] <jadew> so correct me if I'm wrong
[16:34:16] <OndraSter_> I am SO CLOSE!
[16:34:17] <jadew> the first one is 1206, right?
[16:34:20] <OndraSter_> to getting it working
[16:34:23] <RikusW> IN is to the PC right ?
[16:34:30] <OndraSter_> yes
[16:34:43] <jadew> the second one is 0805?
[16:34:56] <OndraSter_> the top diode != 1206
[16:34:59] <OndraSter_> they have got their own package names
[16:35:12] <yunta> OndraSter_: how do you mark your IN ep as ready for sending?
[16:35:13] <OndraSter_> 1206 = 0.12" * 0.06"
[16:35:21] <OndraSter_> yunta, clearing BUSNACK0 flag
[16:35:28] <OndraSter_> it works for the first packet
[16:35:33] <OndraSter_> that one gets sent
[16:35:37] <OndraSter_> not after that..
[16:35:49] <OndraSter_> I have not figured out what is the notification channel good for :D
[16:36:04] <OndraSter_> hmm
[16:36:05] <OndraSter_> that reminds me
[16:36:06] <yunta> OndraSter_: r u sure your trncompl0 and ovf are also clear then?
[16:36:13] <OndraSter_> ovf is not relevant for output
[16:36:15] <jadew> 3.20 x 1.60
[16:36:15] <OndraSter_> nor trncompl
[16:36:22] <OndraSter_> err, not for transmissions
[16:36:22] <jadew> so 1206 is bigger than that one?
[16:36:49] <OndraSter_> 0.12" = 3mm or thereabouts
[16:37:10] <OndraSter_> hmm RTS ON
[16:37:12] <OndraSter_> DTR ON
[16:37:23] <OndraSter_> maybe those are the issue?
[16:37:32] <OndraSter_> (the "computer" sends those when I open putty window)
[16:37:38] <OndraSter_> maybe it is waiting for something?
[16:37:44] <OndraSter_> on the notification channel?
[16:37:46] <yunta_> OndraSter_: ep->STATUS &= ~(USB_EP_TRNCOMPL0_bm | USB_EP_BUSNACK0_bm | USB_EP_OVF_bm);
[16:37:48] <yunta_> that works for me
[16:38:06] <RikusW> OndraSter_: notification channel ? EP 0 ?
[16:38:14] <OndraSter_> or management channel
[16:38:23] <RikusW> is it EP 0 ?
[16:38:24] <OndraSter_> no
[16:41:51] <OndraSter_> yunta_, nope, still nothing
[16:42:01] <OndraSter_> the STATUS changes from 43 to 01 as it should
[16:42:13] <OndraSter_> I get fifo event on the ep3 channel as it should
[16:42:15] <yunta_> should it?
[16:42:18] <OndraSter_> yes
[16:42:57] <OndraSter_> then upon sending new value I get 02 => 00 (since it clears BUSNACK0 again)
[16:43:04] <OndraSter_> but no FIFO event on that channel anymore
[16:43:31] <yunta_> interrupt driven?
[16:43:35] <OndraSter_> of course
[16:44:12] <yunta_> are you aware of this? http://www.avrfreaks.net/index.php?name=PNphpBB2&file=printview&t=120080&start=0
[16:44:20] <yunta_> not sure if relevant, but it's good to know
[16:44:38] <yunta_> I've wasted entire weekend on that...
[16:44:47] <OndraSter_> yes
[16:44:52] <OndraSter_> I am using FIFO as I should
[16:45:22] <yunta_> weird :D
[16:46:41] <OndraSter_> http://pastebin.com/xpUNHenT
[16:50:15] <yunta_> comparing.....
[16:52:41] <specing> OndraSter_: I had some "fun" in crt1.S
[16:52:53] <specing> stripped out all the useless craps :)
[16:53:00] <OndraSter_> hehe
[16:53:21] <OndraSter_> like removing interrupt jump table? :D
[16:53:32] <specing> No, I didn't take it that far
[16:53:44] <specing> though one could remove the unused handlers...
[16:53:52] <OndraSter_> hehe
[16:53:54] <OndraSter_> tell me about it
[16:54:00] <OndraSter_> on xmegas the jump table has almost half a kilobyte
[16:54:05] <OndraSter_> (about 12 bytes less than that)
[16:54:35] <OndraSter_> enough with screwing with bloody USB for today
[16:54:47] <OndraSter_> I got it working through enumeration and data PC->MCU. Good day it was today
[16:54:55] <OndraSter_> except this
[16:54:55] <OndraSter_> http://www.novinky.cz/krimi/284905-auto-v-praze-ujizdelo-policistum-ukazalo-se-ze-je-kradene.html
[16:54:58] <OndraSter_> which I witnessed
[16:55:10] <OndraSter_> the guy hit the semaphores like 5 metres in front of me
[16:55:12] <OndraSter_> (I was luckily in a bus)
[16:55:25] <OndraSter_> the guy who stole the car and ran away from the cops*
[16:56:07] <specing> 46 0 0 46 2efirmware.elf
[16:56:16] <specing> http://sprunge.us/MhUj
[16:56:21] <specing> my stripped down stuff
[16:56:29] <specing> 100 0 0 100 64firmware.elf
[16:56:35] <specing> http://sprunge.us/FaSM
[16:56:37] <OndraSter_> heh nice
[16:56:42] <specing> produced by the old install
[16:56:50] <specing> 54 bytes difference...
[16:57:03] <OndraSter_> but where do you clear bss and copy .data?
[16:57:47] <OndraSter_> or did you if that around?
[16:58:01] <specing> gcc 4.8.0 is smart enaugh to know there is no data to copy ;P
[16:58:50] <OndraSter_> did they finally fix that all the constants (strings too) are stored in program memory and loaded from it directly
[16:58:55] <OndraSter_> without being copied to .data? :P
[16:58:57] <OndraSter_> and RAM
[16:59:06] <yunta_> OndraSter_: can't see a log line for STATUS change for your OUT trans (ep 03).. to accept new answer from pc. is it there, just not logged?
[16:59:07] <OndraSter_> or was that in 4.7.0 already
[16:59:21] <OndraSter_> yunta_, that is the 43 => 01
[16:59:33] <OndraSter_> and that should be IN
[16:59:39] <OndraSter_> I count them "the regular way"
[17:00:01] <OndraSter_> aka ep0 = 0, ep1 = 0 in, ep2 = 1 out, ep3 = 1 in, ep4 = 2 out, ep5 = 2 in
[17:00:05] <specing> OndraSter_: dunno
[17:00:15] <specing> Im just happy that my programs are minimal :)
[17:00:18] <OndraSter_> :D
[17:01:58] <OndraSter_> tomorrow I will grab old laptop (I have got here ibm t21 lying around without any use) which *should* have got some HDD in it with hopefuly XP
[17:02:08] <OndraSter_> USB logger will be installed
[17:02:59] <yunta_> OndraSter_: comment on that suggests IN trans: STATUS: 43 => 01 <-- STATUS of the "MCU to PC" endpoint (EP3) (clearing BUSNACK flag)
[17:03:08] * yunta_ feels lost :D
[17:03:42] <OndraSter_> yes, that is IN
[17:03:52] <OndraSter_> ep03 = in
[17:03:58] <yunta_> so, what I'm missing here is the same thing for OUT
[17:04:04] <OndraSter_> duh?
[17:04:13] <yunta_> just a sec
[17:04:17] <OndraSter_> there are three endpoints:
[17:04:26] <OndraSter_> ep3 -- IN -- data
[17:04:29] <OndraSter_> ep4 -- OUT -- data
[17:04:38] <OndraSter_> ep5 -- OUT -- NOTIFICATION (management?)
[17:04:41] <yunta_> ah, shit
[17:04:43] <yunta_> :D
[17:04:44] <OndraSter_> or was the last one IN too
[17:05:34] <yunta_> if they are sequential, last should be in, shouldn't it?
[17:05:35] <yunta_> but
[17:05:36] <OndraSter_> the notification (management?) is either ep5 or ep2. It is not important though
[17:05:57] <yunta_> in those logs there are only events for 04 and 03
[17:06:02] <OndraSter_> yes
[17:06:05] <yunta_> so 03 is IN and 04 OUT ?
[17:06:07] <OndraSter_> 04 = PC to MCU
[17:06:11] <OndraSter_> 03 = MCU to PC
[17:06:18] <OndraSter_> 03 is IN and 04 is OUT, yes
[17:06:27] <yunta_> first ok
[17:06:31] <yunta_> s/first//
[17:06:54] <yunta_> so, you first send a byte from pc, and mcu sends it back?
[17:07:05] <OndraSter_> yes
[17:07:08] <OndraSter_> it works for the first byte
[17:08:08] <yunta_> do I understand correctly that
[17:08:25] <yunta_> status change log line is missing for 04 or 03 (one of them) ?
[17:09:13] <OndraSter_> yes, it is not there, because that one works
[17:09:18] <OndraSter_> it is missing for 04
[17:09:22] <yunta_> but it does happen?
[17:09:25] <yunta_> the status change
[17:09:25] <OndraSter_> clearing BUSNACK0 there
[17:09:26] <OndraSter_> yes
[17:09:28] <yunta_> ok
[17:09:30] <OndraSter_> and it receives data just fine
[17:09:59] <yunta_> ok
[17:10:15] <yunta_> sorry, it's sometimes difficult to get unfamiliar logs :)
[17:10:24] <specing> yup its broken
[17:10:28] <specing> avr/long_arith.c:6:1: warning: return type defaults to 'int' [enabled by default] ISR (USART_RX_vect)
[17:11:59] <yunta_> OndraSter_: did you check if pc sends IN token after subsequent OUTs ?
[17:12:49] <OndraSter_> yunta_, I will tomorrow
[17:12:55] <OndraSter_> when I dust off the laptop
[17:13:13] <specing> Ah it isn't broken
[17:13:21] <specing> I forgot to #include <avr/interrupt.h>
[17:13:41] <OndraSter_> aaand it is 16th November
[17:13:46] <OndraSter_> does anyone know what that means?
[17:13:47] <specing> still it generates lots and lots of crap in prolugue and epilogue
[17:13:51] <specing> no
[17:13:55] <OndraSter_> hmm
[17:14:10] <specing> still 16 bytes shorter than gcc 4.5!
[17:14:16] <OndraSter_> that my facebook wall will be awfuly spammed.
[17:14:18] <OndraSter_> ti means
[17:14:19] <OndraSter_> it*
[17:14:28] <yunta_> birthday?
[17:14:34] <OndraSter_> aye
[17:14:51] <yunta_> congrats
[17:14:54] <OndraSter_> thanks
[17:15:02] <specing> your fault for being on assbook
[17:15:04] <OndraSter_> (or at least that is what Matrix wants me to believe that happened 20 years ago)
[17:15:06] <OndraSter_> lol
[17:15:13] <specing> some of us don't have such problems
[17:15:23] * yunta_ doesn't have fb account....
[17:15:33] <specing> yeah, me neither
[17:15:49] <OndraSter_> I like to ask people about grabbing a beer
[17:15:54] <OndraSter_> or solving some school stuff
[17:16:20] <specing> One more thing I have to do
[17:16:50] <OndraSter_> school?
[17:16:57] <specing> figure out how to make gcc put main() first
[17:17:20] <specing> so it can just flow to it from the initializations
[17:17:33] <specing> instead of doing retarded RCALLs or RJUMPs
[17:17:38] <yunta_> put it in different section and modify link script to place it wherever you want
[17:17:39] <Malinuss> how old OndraSter_ ?
[17:17:44] <OndraSter_> 20
[17:17:57] <Tom_itx> a pup
[17:18:05] <Malinuss> gratz
[17:18:11] <OndraSter_> thanks
[17:18:12] <OndraSter_> haha pup
[17:18:29] <jadew> wish I was 20 again
[17:18:33] <Malinuss> ^
[17:18:35] <specing> Im not exaclty sure why people congratulate others on their birthdays...
[17:18:46] <OndraSter_> because you made it that far without killing yourself from any cap?
[17:18:55] <OndraSter_> or am I the only one who testes charge of capacitors by hand..
[17:18:56] <specing> getting older isn't exaclty a cause for celebration...
[17:18:58] <yunta_> I congratulate patience
[17:19:15] <Malinuss> OndraSter_, I use my tongue
[17:19:23] <OndraSter_> (jk, I always short them, then remeasure with DMM to be sure I have actually shorted them and then use my fingers to be ABSOLUTELY sure)
[17:19:25] <Malinuss> OndraSter_, works on ~300uF +
[17:19:29] <OndraSter_> hehe
[17:19:32] <Malinuss> lol
[17:19:34] <OndraSter_> I shorted some bigger one
[17:19:41] <OndraSter_> it was still charged to about 30V
[17:19:45] <OndraSter_> capacity... don't remember
[17:19:47] <jadew> you know what congratulations are weird? the ones when you find out you're having a baby
[17:19:54] <OndraSter_> :D
[17:20:05] <jadew> like.. they're congratulating you for having sex
[17:20:24] <specing> and you are like "oh fuck"
[17:20:25] <OndraSter_> now the song from The Lonely Island will be stuck in my head, thanks jadew :D
[17:20:37] <jadew> (it happened to me the entire day, since last night I found out I'm gonna be a dad)
[17:20:46] <OndraSter_> CONGRATS!
[17:20:49] <jadew> hehe
[17:20:52] <jadew> thanks :)
[17:21:02] <jadew> it was great sex! thank you
[17:21:16] * yunta_ shorted 5kg of caps from weird medical device, using scalpel. metal can fly, and melts.
[17:21:31] <OndraSter_> http://www.youtube.com/watch?v=lQlIhraqL7o jadew
[17:21:45] <jadew> haha OndraSter_
[17:21:49] <yunta_> jadew: congrats on having sex
[17:21:51] <jadew> I know that one, really funny
[17:21:55] <jadew> thank you, thank you
[17:23:33] <specing> jadew: remember, you have to teach it binary and to count from 0 to INF in under 5s
[17:23:50] <jadew> you bet!
[17:24:06] <OndraSter_> specing, if we are talking 2 bits..
[17:24:13] <OndraSter_> 2 bits float!
[17:24:25] <jadew> heh
[17:24:47] <jadew> if he'll feel like learning the stuff I enjoy, sure, otherwise I won't push him
[17:24:52] <OndraSter_> because floats (or at least the ones in 387 FPU) had +inf and -inf I think
[17:24:52] <specing> OndraSter_: if you have a sign bit and a fraction part, where do you put the exponent part?
[17:25:00] <jadew> but I think that if he'll see me having fun, he'll want a piece of the action as well
[17:25:03] <OndraSter_> specing, it is either 0 or inf
[17:25:05] <OndraSter_> the second bit
[17:25:19] <OndraSter_> :P
[17:25:23] <jadew> OndraSter_, 2 bits are 4 values :P
[17:25:26] <OndraSter_> yes
[17:25:28] <OndraSter_> and first bit is sign
[17:25:34] <OndraSter_> so you have +-0 and +-inf
[17:25:36] <jadew> ah, right
[17:25:43] <jadew> yeah, I can teach him that
[17:25:50] <specing> That is one exciting float indeed
[17:26:17] <OndraSter_> fast to count with really
[17:26:17] <jadew> small error as well
[17:26:28] <OndraSter_> you can even shove FOUR into ONE AVR8 REGISTER!
[17:26:36] <OndraSter_> talk about memory savings!
[17:26:50] <OndraSter_> didn't (x)mega have some instructions to count with 8bit floats?
[17:26:54] <OndraSter_> fmul(s) etc
[17:27:00] <OndraSter_> I never checked them
[17:28:22] <specing> lol AVRs don't even have integer division or modulus
[17:28:28] <specing> atleast not the classic AVRs
[17:29:06] <yunta_> OndraSter_: this log looks totally good. mine looks basically the same (aside of some uncleaned bits). the only things coming to my mind are: 1) pc doesn't query (doesn't send IN token) 2) pc sends IN token for data1 always, never data0 (like it happens in first data packet of enumeration on ep0).
[17:29:26] <yunta_> let me know when you get to your laptop and find out what was wrong
[17:31:46] <OndraSter_> surely will
[17:31:49] <OndraSter_> thanks for tips
[17:31:55] <specing> OndraSter_: http://sprunge.us/cWNY
[17:32:13] <yunta_> well, I didn't help. but had fun trying :)
[17:33:03] <OndraSter_> specing, the last one -- not always, if user defines some .initN section
[17:33:23] <specing> yeah I put the XJMP back in :(
[17:33:47] <OndraSter_> specing, http://clip2net.com/s/2vzVT
[17:33:49] <OndraSter_> :P
[17:33:54] <OndraSter_> FMUL, FMULS and FMULSU
[17:34:49] <specing> nice
[17:34:57] <specing> does it have division and modulo?
[17:35:05] <OndraSter_> no :P
[17:35:14] <Tom_itx> does gcc support it?
[17:36:27] <OndraSter_> doubt it
[17:36:29] <OndraSter_> never seen it
[18:14:35] <static> fosscon
[18:15:15] * Tom_itx puts static in a fariday cage
[18:16:05] <yunta_> you mean Friday cage probably
[18:17:22] <Tom_itx> faraday
[18:17:33] <Tom_itx> http://en.wikipedia.org/wiki/Faraday_cage
[18:19:16] <yunta_> yes, I'm sorry, it was a stupid joke on your misspelling
[18:19:37] <yunta_> s/ a //
[18:19:46] <Tom_itx> tomorrow it may be a Friday cage
[18:20:08] <yunta_> It's already tomorrow here :)
[18:20:18] * yunta_ lives in finland
[18:20:43] <Tom_itx> oh
[18:29:40] <static> oops