#avr | Logs for 2011-10-30

Back
[00:13:58] <jd_ce> boom!
[00:14:01] <jd_ce> lufa compiled
[00:14:06] <jd_ce> external makefile to the rescue
[00:14:11] <abcminiuser> Vunderbar!
[00:14:17] <jd_ce> hah!
[00:14:22] <jd_ce> you're here!
[00:14:47] <jd_ce> I spent all yesterday getting nowhere with lufa
[00:14:54] <abcminiuser> :(
[00:15:16] <jd_ce> not your fault. I just don't understand the protocol too well
[00:15:41] <Tom_itx> ok i'm setting up an arm env
[00:15:45] <Tom_itx> slowly
[00:59:37] <jd_ce> :/
[03:24:37] <inflex> lo folks
[03:34:43] <CapnKernel> hi inflex
[03:34:51] <CapnKernel> My first day back in Shenzhen :-)
[03:34:53] <inflex> hiya
[03:35:02] <inflex> Good thing you're not flying around with Qantas atm
[03:36:01] <CapnKernel> Haha I just read The Age, and yes, that thought had occurred to me..
[03:36:17] <CapnKernel> I flew China Southern to Guangzhou, then took the train to Shenzhen.
[03:36:21] <CapnKernel> But I have to go now.
[05:09:37] <amee2k> from what i remember from my digital clock projects, mains as frequency reference has pretty good long term stability
[05:10:20] <amee2k> i'm trying to measure the stability of a frequency reference circuit i made for a frequency counter, but for that i need an even better reference, obviously
[05:10:43] <amee2k> so how long counts as long term in terms of mains frequency stability?
[05:20:46] <h4x0r`> suh
[05:22:10] <amee2k> mmh?
[05:23:09] <h4x0r`> hows it goin?
[05:23:15] <h4x0r`> havent been here in a while :)
[05:23:23] <h4x0r`> hi Tom_itx :)
[05:23:29] <amee2k> mildly bored
[05:23:33] <h4x0r`> hi theBear :)
[05:23:46] <h4x0r`> ahh, well i can help u then lol
[05:23:47] <amee2k> idly working on a frequency counter project
[05:24:06] <h4x0r`> lol sounds fun :p
[05:24:09] <h4x0r`> http://www.telit.com/en/products/gsm-gprs.php?p_id=12&p_ac=show&p=109
[05:24:10] <amee2k> fixing my computer... first time i had a corrupted file system in a long time
[05:24:26] <h4x0r`> in the image of the chip, it has 2 x 2 mm dia balls
[05:24:27] <amee2k> actually i think first time ever since i decided to ditch windows >_>
[05:24:48] <h4x0r`> in the specsheet, section 4.1.2, it has one, and page 79 seems different again
[05:24:54] <h4x0r`> amee2k, which os?
[05:25:08] <amee2k> ubuntu 11.4
[05:25:09] <mrfrenzy> amee2k: from what I understand, over 24h the mains frequency is a very good reference
[05:25:13] <h4x0r`> gentoo is my fave nix flava
[05:25:16] <h4x0r`> np
[05:25:23] <mrfrenzy> if you need short time stability, maybe use a gps clock source
[05:25:41] <amee2k> first time i decided to give ubuntu a shot since the 9.4/9.10 disaster
[05:26:16] <amee2k> mrfrenzy: nah, i can let it run for a few days to get a stability reading
[05:26:16] <h4x0r`> ahh, thats around the time i used to call it nubuntu/spewbuntu/pubuntu
[05:26:30] <amee2k> ubunpoo is my favourite one
[05:26:53] <h4x0r`> LUL
[05:27:35] <amee2k> the 9.4->9.10 update went pretty bad for me and any of the later install CDs refused to boot on my box so i went over to debian 5 for a year or so. 11.4 was the first i could get to work
[05:27:52] <h4x0r`> ahh i think i remember u doing that :D
[05:28:17] <amee2k> yeah, i think i was ranting about it to great length :)
[05:28:52] <amee2k> it was kinda meh but it worked i suppose. now wondering if i should give deb6 a shot or try 11.10
[05:29:51] <amee2k> ditching gnome was a dick move IMO. they still have the "classic desktop" but it looks like not too much effort is going into maintaining that
[05:30:08] <h4x0r`> gnome has its goodside hey
[05:30:22] <h4x0r`> personally i prefer bash
[05:30:24] <amee2k> i used to use gnome for like 4-5 years or so
[05:30:29] <h4x0r`> yea
[05:30:53] <amee2k> didn't try gnome3 yet though... everyone seems to like bashing it :P
[05:31:06] <h4x0r`> heh, anything released under gpl gets bashed lol
[05:31:15] <amee2k> well, that too
[05:31:16] * inflex is still with Fluxbox
[05:31:20] <inflex> saves a lot of memory and CPU :)
[05:31:22] <amee2k> but even the gplfags are basing gnome3
[05:31:32] <h4x0r`> lol
[05:31:38] <h4x0r`> hey inflex
[05:31:43] <inflex> hiya h4x0r`
[05:32:08] <h4x0r`> hows the country treating ya ? still out bush mate?
[05:32:15] <amee2k> inflex: i did try xfce on my laptop. its pretty cool, but most other WMs make me miss nautilus and gvfs
[05:32:25] <inflex> yeah, still out in the bloody bush.... trying hard to get out of here still.
[05:32:53] <inflex> amee2k: yeah, I suppose it comes down to what you're doing/wanting... for me everything is setup in Fluxbox, a lot of key-combo starts and such
[05:33:26] <inflex> amee2k: I don't need much in the way of fancy - so it works well and if something special comes up I find that the GNOME/fancy backend is still operating and as such apps do work fine
[05:33:47] <amee2k> might just give fluxbox a shot. i've also tried awesome3 last summer and liked it. application support sucks pretty badly though
[05:34:23] <inflex> amee2k: they all take getting used to, that's for sure
[05:34:56] <inflex> amee2k: FB works best if you put your preferred apps on the primary menu that comes up when you right-click
[05:35:00] <amee2k> well, what i can't stand is that lots of applications have funny quirks when used outside of the DE of choice of the developer :/
[05:35:19] <inflex> yeah, that can be quite a PITA
[05:35:56] <amee2k> kde apps in awesome3 were quite offensive when i tried it because they don't seem to use the window hints (or whatever its called) in any consistent way
[05:36:38] <amee2k> so a save as dialog popping up pushes all the windows aside and messes up your layout, and shit like that
[05:38:43] <h4x0r`> a quick avr question, ive been using the mega162, dual usarts, i like it.
[05:38:46] <amee2k> last time i checked gentoo was /so/ not my thing. any other distro recommendations for medium to long time debian user?
[05:39:22] <h4x0r`> hmm, any recommendations for a similar chip, that might provide fair flash and some usb ?
[05:40:14] <amee2k> nothing usb on my end, but my personal favourites are tiny24 and mega8/mega88
[05:40:49] <h4x0r`> yeah i love those things lol :)
[05:40:57] <amee2k> with noone wanting serial ports anymore, i've been looking into usb ones too
[05:41:03] <h4x0r`> not sure why id be switching to arm yet either..
[05:41:09] <amee2k> but i'd want some that i can get in easy to use DIP package as well
[05:41:45] <h4x0r`> ahh, preffered dip, for replacement zapps?
[05:42:05] <amee2k> "zapps"?
[05:42:31] <h4x0r`> smoked chips
[05:43:14] <amee2k> oh no, for my own projects. smd stuff is bit of a pita because i don't have a useful procedure for making homemade boards yet besides a sharpie
[05:43:49] <amee2k> and despite other people here masturbating vigorously, ordering from a board house is too expensive for my needs
[05:44:10] <h4x0r`> np+agreed
[05:44:26] <h4x0r`> understood at least.
[05:44:49] <h4x0r`> xmega... wtf 8 usarts+usb....
[05:45:00] <amee2k> for my usual purposes, the board would cost more than the rest of the project usually. especially since i'd only need one board
[05:45:46] <h4x0r`> do you have some pics or examples of your work?
[05:45:47] <amee2k> xmegas are cool, but iirc the least antisocial packages they have are tqfp
[05:46:15] <amee2k> not right now, all my photos are on the computer with the corrupted ubuntu system >_<
[05:46:38] <h4x0r`> :D
[05:46:42] <h4x0r`> HAHAHA
[05:46:44] <h4x0r`> soz
[05:46:44] <amee2k> what kind of work are you interrested in?
[05:46:46] <h4x0r`> :D
[05:46:58] <h4x0r`> all sorts, i like checking out how people design :D
[05:47:32] <specing> amee2k: tqfps are easy
[05:47:32] <h4x0r`> the xmega sounds ok, might check out a tqfp hmm
[05:47:52] <amee2k> specing: with a sharpie?
[05:47:54] <h4x0r`> yep theyre replaceable to in case of smoking they just plug in right ?
[05:48:00] <specing> amee2k: sharpie?
[05:48:13] <amee2k> http://ompldr.org/vYXJ6Ng/CIMG8650b.jpeg << not related to pcbs at all, but still a project
[05:48:33] <amee2k> going through my irc for old links
[05:48:58] <h4x0r`> what is it?
[05:49:04] <h4x0r`> looks good :p
[05:49:33] <h4x0r`> and that attiny is so koot :p
[05:49:43] <amee2k> ghetto isolation transformer. two 300VA 21V transformers back to back
[05:50:20] <amee2k> thats not an attiny, its a tiny bridge rectifier to run the LEDs in the voltmeter :)
[05:50:57] <amee2k> when i printed the schem to wire it up i noticed it fit right into the top of the case, so i left it there
[05:52:00] <h4x0r`> oh that? yeah that looks cool.. i was talkin about the attiny in the bb tho
[05:52:12] <amee2k> ah, okay
[05:53:35] <amee2k> i couldn't find a proper case for it... the transformers are pretty big. so i got the cheapest toolbox they had at the hardware store :P
[05:53:50] <amee2k> http://ompldr.org/vYXhpMw/CIMG8662b.jpeg << ghetto speaker made from an ATX PSU case and a transducer i got out of an old TV
[05:54:34] <h4x0r`> lolpimpin
[05:54:37] <h4x0r`> nice :D
[05:55:31] <amee2k> http://ompldr.org/vOXh2cQ/CIMG8408b.JPG << half-finished prototype for a random number generator the fuzzball hanging over the breadboard is a noise generator using noise from a transistor junction
[05:56:05] <amee2k> the board allready in the case is the power supply. the meter shows the 24V bias voltage for the noise generator
[05:56:31] <h4x0r`> wow, kool. looks involved.
[05:56:57] <amee2k> the circuit is supidly simple, actually. figuring out how to do it took the most time :)
[05:58:42] <amee2k> http://ompldr.org/vOWszMQ/RNG.png << the output of that is fed through a 1u coupling cap and biased at 1.25V dc offset with two 100k resistors and a 2.5V reference
[05:59:38] <amee2k> then sampled by a mega8's ADC using the 2.5V ref as external ADC reference. generates a nice gaussian distribution and has enough bandwidth to saturate a 115kbaud uart easily
[06:00:11] <h4x0r`> jesus, wish i was that good lol.. itll take time for me to fully understand how to refine current like that mate ;)
[06:00:20] <h4x0r`> niiiiice work lol
[06:00:22] <amee2k> the noise output is far from white but extremely wideband, far into the MHz region
[06:00:34] <amee2k> thanks :)
[06:03:16] <amee2k> chances are you could extract a few Mbit/s worth of entropy from that 50 cent noise generator, and probably more with a bit of filtering to make it a bit more white
[06:03:47] <h4x0r`> hm, uc3 chips seem nice.. but not much point.. the site wont help me find an ethernet candidate.. id like to be fully equipped if making the change..
[06:05:18] <amee2k> aren't UC3s 32bit ones?
[06:05:40] <amee2k> i don't think i have any more interresting project pic links handy right now
[06:06:30] <h4x0r`> The AVR UC3 L Series runs on 165 µA/MHz in active mode, 600 nA with RTC running, and down to 9 nA with all clocks stopped.
[06:06:31] <h4x0r`> wtf.
[06:06:55] <h4x0r`> id need to translate to talk to most ic's.
[06:07:03] <h4x0r`> or step..
[06:08:01] <amee2k> http://ompldr.org/vOTgycA/cimg7676b.jpg http://ompldr.org/vOHRmcw/cimg7224b.jpg << series lightbulb rig i made... its got series taps, an NTC and a PFC choke in it too
[06:09:56] <amee2k> first pic shows the phase indicator... shows which side of the plug has phase and which has neutral. and as a safety feature, if both are lit then your PG sucks and you should stop doing whatever you're doing and fix that first :)
[06:10:11] <amee2k> "translate"?
[06:12:07] <h4x0r`> cmos logic levels
[06:12:54] <amee2k> http://ompldr.org/vOTdybQ/cimg7662b.jpg << last one i can find is this power supply for a high voltage diode tester
[06:13:23] <h4x0r`> lol cool
[06:13:51] <amee2k> most of the stuff i do "still" uses 5V logic. when i'm out of max232s i'll probably switch to usb and hyrid 5V/3.3V logic
[06:14:10] <h4x0r`> yup
[06:14:14] <h4x0r`> smicko
[06:14:16] <amee2k> making a point of buying parts that can run on both
[06:14:32] <h4x0r`> 2.8v for a lot of gsm stuff too
[06:14:52] <h4x0r`> ftd232rl can communicate at that level tho
[06:15:04] <h4x0r`> so its kind of a lot easier using ftdi
[06:15:22] <h4x0r`> than max232 if u want to save circuitry space
[06:15:23] <h4x0r`> iric
[06:15:25] <amee2k> i'm not too much into extremely low voltage fetish. and neither extreme miniaturization fetish >_>
[06:15:31] <h4x0r`> ahh
[06:15:40] <h4x0r`> tad bit of an smd freak atm
[06:16:03] * CapnKernel 's over through-hole
[06:16:16] <amee2k> SMD sounds cool for homemade boards because it saves drilling, but thats about it
[06:16:40] <amee2k> for most of my purposes i don't see much of an advantage
[06:17:02] <CapnKernel> Because it's getting to the point where the cool parts only come in SMD
[06:17:11] <amee2k> yeah
[06:17:13] <h4x0r`> pffft who needs this uc3 junk anyway :p
[06:17:26] <h4x0r`> not feelin it
[06:17:36] <amee2k> SOIC and QFP are okay i suppose, but i find BGA or QFN fairly offensive to my sanity
[06:17:43] <h4x0r`> soic8
[06:17:45] <h4x0r`> <3
[06:18:01] <h4x0r`> 3mm 12v vreg ftw.
[06:18:33] <amee2k> i made a USB switch with a QFN switch IC... ended up super-glueing it upside down to the board
[06:18:52] <amee2k> then pulling strands out of a piece of multistranded switching wire for the connections
[06:19:04] <h4x0r`> kewl :p
[06:19:04] <amee2k> then put a drop of hot glue over it to prevent shorts >_<
[06:19:20] <amee2k> sort of a ghetto "epoxy blob" style IC
[06:19:29] <h4x0r`> works :p
[06:19:35] <amee2k> yep, it does :)
[06:19:52] <h4x0r`> hmm, i hope eagle has a tqfp mega162 print out there.. couldnt stand to make a new part for it lol
[06:21:29] <qiau> greets.. I'm getting "storage.c:111: ERROR: address out of bounds: 0x4000" running simulavr and I can't figure out what the problem really is..
[06:22:41] <qiau> sounds like my program is too big.. but running avr-size shows me that it's 3650 text, 34 data and I'm trying to simulate an atmega16
[06:23:10] <h4x0r`> dunno the specs, sounded like mem issue at first.
[06:23:46] <amee2k> i'm tempted to point at a runaway pointer... out of bounds RAM access?
[06:24:03] <amee2k> but never used simulavr... i have real ICs :)
[06:24:09] <qiau> ;)
[06:24:40] <h4x0r`> ditto ^
[06:25:01] <h4x0r`> runaway pointer is what i was thinking.. cant be sure tho.
[06:25:01] <qiau> the program is trivial and the difference to make it crash is to link in an extra file so I don't think it's runtime related
[06:25:27] <qiau> I havn't even included the header for that object file, just added it to the makefile
[06:25:46] <h4x0r`> 0x4000 reminds me of a base address for 32bit applications in windows.. hmm
[06:25:51] <qiau> :)
[06:25:56] <h4x0r`> :)
[06:26:02] <amee2k> lol
[06:30:42] <h4x0r`> sweet, got me a mega162 tqfp footprint.. goin to work!
[06:30:48] <h4x0r`> (on the pcb)
[06:31:00] <h4x0r`> bah! i need a socket now amirite?
[06:31:07] <h4x0r`> they have sockets dont they lol ?
[06:31:59] <h4x0r`> or am i thinkin of qfp, where you just drop the avr into the socket.. (not pdip)
[06:32:32] <amee2k> tqfp sockets are fairly expensive because they're complex and noone uses them
[06:32:38] <h4x0r`> lol
[06:32:51] <h4x0r`> ill look into it
[06:33:21] <amee2k> they're mostly used on test equipment and their specs suck compared to the parts specs
[06:33:52] <amee2k> there are sockets for j-lead packages lik used to be popular for bios eeproms on old mainboards
[06:33:58] <amee2k> like* even
[06:34:00] <h4x0r`> cant imagine (nor would I have any clue) on how to solder a tqfp chip without a socket package
[06:34:14] <h4x0r`> the pins are too small.
[06:34:24] <h4x0r`> and theres like 64 of them.. or whatever..
[06:34:30] <amee2k> tin the pads or put solder paste on them, put the part on top, then heat gently with hot air gun
[06:34:43] <h4x0r`> can do that np ;)
[06:34:57] <h4x0r`> cheers :>
[06:35:04] <h4x0r`> ill design accordingly :D
[06:35:27] <amee2k> if you do lots of SMT stuff, you may want to invest in a hot air rework station
[06:36:32] <h4x0r`> got one.. its an ebay one.. but it should see the prototype out ;)
[06:37:04] <amee2k> i spend a lot of time fixing stuff so i got a cheap rework station some time ago. its quite handy for reworking smt boards
[06:37:09] <h4x0r`> it came with a spare element just in case the other one blows.. which was as reassuring, as it was daunting...
[06:37:19] <amee2k> lol
[06:37:29] <h4x0r`> lol
[06:37:45] <amee2k> yeah, the heating elements are wearing parts, same as on regular soldering irons actually
[06:38:04] <h4x0r`> cool
[06:38:12] <h4x0r`> glad ive got one on hand just in case
[06:38:38] <amee2k> my hot air station is still on the first element, but i've had failed soldering irons before
[06:38:45] <h4x0r`> was a little scary.. its like "here, heres a spare part, coz hey face it.. its gonna blow up cheap china junk.. good luck and thanks for buying!"
[06:38:58] <h4x0r`> lol
[06:38:58] <amee2k> lol
[06:39:13] <h4x0r`> it came with one for the soldering iron too :p
[06:39:21] <amee2k> nice :)
[06:39:37] <h4x0r`> not realy.. feels like cheap aluminium foil and 3 wires..
[06:39:45] <h4x0r`> lol .. scary lol..
[06:39:45] <amee2k> i bought the spares for mine separately. the heating elements aren't exactly cheap :/
[06:39:52] <h4x0r`> ahh
[06:40:23] <amee2k> 12EUR-ish for a heating element and 3 for a spare pump membrane
[06:41:31] <h4x0r`> darn
[06:41:38] <h4x0r`> but cheaper than a new station, so WIN!
[06:41:59] <amee2k> yeah. i'm quite glad selling replacement parts for these is still normal
[08:15:38] <qiau> aargh!€#&
[08:17:42] <h4x0r`> ?
[08:18:25] <h4x0r`> suuuuup ;/
[08:19:36] <qiau> still same stupid memory problem
[08:19:58] <qiau> tried to change simulated device from atmega16 to atmega32 then it works all fine
[08:23:29] <h4x0r`> define the mega16 charachteristics into your code
[08:23:38] <qiau> ?
[08:23:48] <h4x0r`> clockspeed.. err, mhz..
[08:23:50] <h4x0r`> ?
[08:23:58] <qiau> 8MHz
[08:24:09] <h4x0r`> that depends on the settings
[08:24:14] <qiau> yeah :)
[08:25:05] <qiau> the error messages tells me that "storage.c:111 address out of bounds: 0x4000".. that's inside "storage_writeb" so obviously I'm trying to store something at 0x4000
[08:26:50] <qiau> looking in the list file, the last instruction is located at 0xdc0
[08:26:55] <qiau> not even near 0x4000
[08:27:41] <amee2k> there are other addresses involved besides the instruction pointer
[08:28:07] <qiau> amee2k: where can I find information about what others are involved?
[08:28:18] <amee2k> in the program code :)
[08:28:25] <qiau> :P
[08:28:38] <amee2k> if its not the instruction pointer, then some other pointer must have a runaway
[08:29:20] <qiau> the entire code is a c program with 3 floats, 3 ints and 3 longs.. where I'm adding, multiplying and dividing each kind.. i.e. float1 = float2 * float3
[08:30:19] <qiau> http://pastebin.com/RaGXuLE7
[08:30:52] <amee2k> is that all of your program?
[08:30:56] <qiau> yes
[08:31:12] <qiau> that's why it's so annoying :)
[08:31:15] <amee2k> put while(1); at the end of main and try again ;)
[08:31:46] <qiau> ehm
[08:31:49] <qiau> sure thing
[08:31:54] <qiau> mean while, why? :)
[08:32:09] <amee2k> because you're returning from main()
[08:32:20] <qiau> bad thing?
[08:32:30] <amee2k> wheres that supposed to go on a micro with no OS
[08:32:48] <qiau> cyberspace!.. yeah, get your point
[08:32:59] <qiau> but it should load into the simulator anyway :)
[08:33:21] <amee2k> i've never used that sim, so the return from main could do anything, really
[08:33:21] <qiau> execution hasn't even started yet
[08:34:26] <qiau> ooooh.. I might have a lead here.. the simulator runs the .elf file
[08:34:44] <qiau> and according to the list file, the elf file contains debug sections
[08:35:13] <qiau> last section size is 0x50, located at 0x4467
[08:35:58] <amee2k> if the simulator can run off a hex file, you could try that instead
[08:36:37] <qiau> yeah, the elf is 21890 bytes, the hex is 9938 bytes.. but simulavr won't run hex files :(..
[08:36:43] <amee2k> that should eliminate the debug stuff or any funny offsets that might be in the elf headers
[08:37:05] <amee2k> okay
[08:37:06] <qiau> I'm developing on OSX so I don't know if there are any other simulators
[08:38:09] <amee2k> last idea i've got is removing all sections from the elf file except for .flash
[08:38:37] <amee2k> and checking the load address (or whats it called again) of that section
[08:39:17] <amee2k> if that doesn't help, you'll have to find someone else with more simulator exp than me
[08:40:33] <qiau> I'll give it a try to find a osx simulator first, thanks for the help :)
[09:58:36] <district> if you return from main it'll just infinitely loop anyway
[10:10:58] <qiau> ah
[10:59:02] <robertjw> Is this the right place to discuss issues with avr-gcc, avr-libc and avr-binutils?
[10:59:38] <Tom_itx> i suppose
[11:00:30] <robertjw> I'm seeing what looks like a problem with the avr tool chain as packaged in Fedora.
[11:01:34] <robertjw> Details are documenting in this thread:
[11:01:39] <robertjw> http://forums.adafruit.com/viewtopic.php?f=25&t=23439&p=123688#p123688
[11:02:11] <Tom_itx> might bring it up on avrfreaks too
[11:02:36] <Tom_itx> may get more 'official' results there
[11:03:25] <robertjw> thanks for the tip
[11:03:45] <Tom_itx> someone may reply here eventually
[11:04:35] <robertjw> I'll continue monitoring this channel today for responses.
[11:06:05] <Kevin`> robertjw: what's the actual problem there?
[11:06:48] <Kevin`> the question is about avrdude stuff I see, which is obviously not very related to the compiler
[11:07:04] <Kevin`> but everyone is talking about the compiler
[11:07:08] <Kevin`> so, clear it up for me :)
[11:09:20] <Tom_itx> or more an issue with the uno chip they decided to replace the ftdi with
[11:09:48] <robertjw> A simple program like the Blink.cpp example will build a correct hex file on windows. The same build on fedora results in a no-error build, but the resulting behavior on an Uno is invalid.
[11:10:15] <Tom_itx> did anybody compare the output files?
[11:10:31] <robertjw> Yes, they are different in both size and content.
[11:10:36] <Kevin`> robertjw: works for me on ubuntu. if you can reproduce the problem, comparing the generated assembly listing (or even preprocessed c) would be interesting.
[11:11:06] <Kevin`> robertjw: also, confirm that the toolchain works with something simple
[11:11:22] <Kevin`> like a non-arduino-lib blink app
[11:11:32] <Tom_itx> rue had an issue recently with delays being optimized out but i don't know if that's a related thing
[11:11:39] <Kevin`> it could just be totally broken in some way
[11:12:15] <Tom_itx> rue_mohr dammit! wake up
[11:12:48] <robertjw> If you have a non-arduino-lib hex file handy, I would be happy to load it for a test.
[11:12:56] <Tom_itx> see if the delays are being screwed with
[11:13:14] <Tom_itx> i've got several on my site
[11:13:20] <robertjw> link
[11:13:43] <Tom_itx> http://tom-itx.dyndns.org:81/~webpage/avr/blinkled/delay_loop/mega8_168/
[11:13:46] <Tom_itx> it's for a 168
[11:13:58] <Tom_itx> internal osc
[11:14:16] <robertjw> Uno has a 328
[11:14:46] <Tom_itx> i can recompile it real quick
[11:14:53] <Tom_itx> booting...
[11:15:15] <Tom_itx> it might work anyway
[11:17:02] <Kevin`> does your led blink thing use eeprom?
[11:17:12] <Kevin`> oh, I see, the bin file is 0 bytes
[11:17:33] <Tom_itx> the make just generates the files
[11:17:37] <Tom_itx> not really used
[11:18:10] <Tom_itx> what clock does it have?
[11:18:32] <Kevin`> iirc most of them use a 16mhz crystal
[11:18:49] <Kevin`> not that blinking twice as fast/slow is a problem
[11:18:49] <robertjw> The Uno uses a 16MHz clock.
[11:20:32] <Tom_itx> http://tom-itx.dyndns.org:81/~webpage/avr/blinkled/delay_loop/mega328/
[11:22:29] <Tom_itx> that should toggle the whole of portb
[11:23:42] <Tom_itx> yeah, and ignore the eeprom stuff
[11:27:06] <Kevin`> Tom_itx: why don't you use the avr-libc delay functions?
[11:27:29] <Tom_itx> because it's an instructional tutorial
[11:27:36] <Tom_itx> i get asked that all the time :/
[11:27:42] <Kevin`> heh :)
[11:27:59] <Tom_itx> i wanted the very basic i could get
[11:28:21] <Tom_itx> there is an int ver too
[11:28:39] <robertjw> Your blink_led.hex loads and functions correctly.
[11:28:52] <Tom_itx> sounds like a broken toolchain
[11:29:02] <robertjw> that's what it sounds like to me
[11:29:22] <robertjw> possibly a problem with the fedora packaging
[11:30:06] <robertjw> I just loaded your hex file. Now I'll try building with my fedora tool chain.
[11:30:18] <Tom_itx> compare the output
[11:31:03] <Kevin`> the output might be different from other random reasons. unless you actaully look at the difference in instructinos
[11:33:51] <robertjw> I see your makefile is set for a 32 bit system, I'm on 64 bit.
[11:43:29] <robertjw> can you see what's wrong with this: http://pastebin.com/spKhDELG
[11:50:40] <Tom_itx> zlog
[11:50:42] <rue_bed> I warn you tho, I'm still asleep
[11:50:45] <Tom_itx> take a peek rue
[11:51:11] <rue_bed> give me a few mins, this things browser is hopeless
[11:52:00] <Tom_itx> it was from a post on adafruit about the arduino uno but the problem seems to be the compiler
[11:53:30] <rue_bed> cpp?
[11:53:36] <rue_bed> use C on an avr
[11:54:17] <rue_bed> is this for an avr striaght or an arduino?
[11:54:29] <sedres> How can I use ATmega168 to supply clock to ATmega16?
[11:54:41] <Tom_itx> clock out
[11:54:42] <robertjw> arduino Uno
[11:54:56] <Tom_itx> sedres read the sheet about it
[11:55:10] <Tom_itx> but be careful with the fuses
[11:55:14] <rue_bed> k, so are you using arduino tools or avr-gcc and avrdude?
[11:55:55] <rue_bed> ddra=0xff; while(1){PINA=0xff;};
[11:55:59] <robertjw> the test used the arduino ide on fedora and windows to make a comparison
[11:56:01] <sedres> Tom_itx: I'm new to it. So not sure.. do you mean PC0-PC5? (ADC)
[11:56:20] <Tom_itx> no
[11:56:23] <rue_bed> then connect and pin from port A to the clockin of the chip
[11:56:31] <Tom_itx> read about clock sources in the data sheet
[11:56:56] <Tom_itx> xtal2 i think can be used as an output
[11:57:01] <rue_bed> robertjw, I cant help with arduino stuff, from what I'v seen its java, and dosn't work worth a damn
[11:57:55] <Tom_itx> rue i just thought it may be a similar problem as you were having with your delays etc
[11:57:58] <robertjw> i've verified that the tool chain calls were identical in both environments, that suggests that the ide is operating the same in win/fedora
[11:58:20] <rue_bed> do you have avr-gcc on the system?
[11:58:26] <amee2k> can anyone recommend an easy to build generic oscillator for testing xtals?
[11:58:35] <rue_bed> amee2k, yes
[11:58:44] <robertjw> yes, from the fedora 14 rpms
[11:58:49] <rue_bed> amee2k, can you give me an hour to wake up?
[11:58:51] <amee2k> then by all means, please do :)
[11:59:03] <amee2k> rue_bed: sure, whenever you're ready
[11:59:07] <rue_bed> its a 1 transistor deal
[11:59:15] <rue_bed> lights a led if the crystal is good
[11:59:17] <amee2k> ooh, i love these :)
[11:59:36] <Casper> hey rue_bed have you made some fast linear actuators?
[11:59:36] <rue_bed> robertjw, have you tried compiling C tests with it?
[11:59:44] <Casper> fast and small?
[11:59:46] <rue_bed> Casper, yes, air muscles
[11:59:52] <Casper> electric
[12:00:05] <rue_bed> do they have to be strong?
[12:00:17] <Casper> 10lbs+ pull/push
[12:00:23] <rue_bed> I did that thing wh.... hnmm
[12:00:35] <robertjw> rue_bed: if you have a link, I can run that
[12:00:42] <rue_bed> not especially fast like that no
[12:01:03] <Casper> motor and theaded rod would provide the small and power, but not the speed... I was aiming at 5"/sec travel
[12:01:33] <amee2k> bigger motor and longer twists?
[12:02:00] <rue_bed> robertjw, hmm I need to post you my new led flaher code
[12:03:11] <Tom_itx> rue_bed i gave him mine
[12:03:22] <Tom_itx> you know.. the one i stole from you???
[12:03:44] <rue_bed> that disn't work on my new gcc
[12:03:48] <rue_bed> I have a rewrite
[12:03:52] <rue_bed> need to post
[12:04:49] <rue_bed> give me 10-15 mins
[12:04:55] <rue_bed> maybe 2-
[12:05:07] <Tom_itx> 10
[12:05:10] <Tom_itx> 9
[12:05:12] <Tom_itx> 8
[12:05:14] <Tom_itx> ...
[12:05:22] <rue_bed> those are seconds you ninny
[12:05:34] <Tom_itx> no you're just in a time warp
[12:05:55] <rue_bed> time dosn't go THAT fast when I'm lying in bed in the morningn
[12:10:08] <amee2k> lol
[12:12:00] <rue_bed> hmm, wonder how I do a house with two 10' cielings and keep the roof lower than a 1 story 8'
[12:12:07] <rue_bed> ok got toget up
[12:12:09] <rue_bed> ugh
[12:12:37] <rue_bed> ugh, its only 10
[12:12:39] <Tom_itx> go down
[12:12:41] <Casper> amee2k: the problem is that threaded rods have too many TPI unless you go with expensive ACME one...
[12:12:45] <Tom_itx> think cave
[12:12:52] <rue_bed> Casper, you said small
[12:13:05] <rue_bed> how small
[12:13:17] <Tom_itx> double lead
[12:13:20] <Casper> as small as possible :D
[12:13:43] <amee2k> depending on the length you need, if you know someone with a lathe he could cut custom ones for you
[12:13:43] <Tom_itx> will double your pleasure
[12:13:55] <JanneP> what do you want to move with your linear dive
[12:13:59] <JanneP> drive even
[12:14:03] <Casper> JanneP: stuff :D
[12:14:12] <amee2k> ghetto CNC project?
[12:14:24] <Casper> nahh, not this time
[12:14:42] <amee2k> Casper: how about the kind of linear drive with a belt that printers or scanners use?
[12:15:05] <amee2k> if you need stronger you could use a car timing belt or a bicycle chain
[12:15:15] <Casper> amee2k: that would need a geared motor
[12:15:33] <JanneP> what kind of precision and force are we talking about
[12:15:36] <Casper> which I don't have and is kinda hard to find around here
[12:15:39] <amee2k> or a really high torque one. taking one from a cordless drill comes to mind
[12:15:50] <Casper> JanneP: as I said, 10+lbs pull-push
[12:15:50] <JanneP> does it need position feedback etc..
[12:16:02] <Casper> that can be added if needed
[12:16:08] <JanneP> ah sorry i missed the 10lbs part
[12:16:23] <Casper> currently I don't care about positionning
[12:16:29] <rue_mohr> ok a cat insisted I get up
[12:16:55] <JanneP> http://www.servocity.com/html/12v_linear_actuators.html
[12:16:56] <Casper> could go with linear pot or optical
[12:16:58] <rue_mohr> schematic for amee2k
[12:17:02] <JanneP> would these kind of thigs be too much overkill
[12:17:02] <amee2k> \o/
[12:17:07] <rue_mohr> test code for... who?
[12:17:32] <rue_mohr> where was that schematic from
[12:17:40] <Casper> JanneP: you also missed "build"
[12:17:40] <rue_mohr> why dont I ahve voice control on the lights yet
[12:17:54] <amee2k> crystal tester... you said you have a simple one but didn't tell from where
[12:18:53] <JanneP> yeah i noticed, it just might be a bit hard to build without machine tools.. especially with that kind of speed requirement
[12:19:13] <rue_mohr> not on this bookshelf, I'll check the livingroom
[12:21:05] <JanneP> i think I'd also go towards a scanner type belt drive, like amee2k suggested
[12:21:43] <rue_mohr> wow, thats not small at all
[12:21:50] <rue_mohr> how much stroke?
[12:22:01] <rue_mohr> use a rack drive from a cdrom
[12:22:10] <Casper> I'ld like 5-6" travel
[12:22:15] <rue_mohr> oh
[12:22:25] <rue_mohr> use a rack drive from a photocopier
[12:22:37] <Casper> no access to those
[12:22:37] <rue_mohr> (which is cables and steppers)
[12:22:42] <rue_mohr> why not?
[12:22:45] <rue_mohr> hu?
[12:23:01] <rue_mohr> do you want dc or stepper?
[12:23:08] <Casper> it's hard to find stuff here...
[12:23:10] <Casper> anything
[12:23:21] <Casper> I'll have to code an AVR anyway for it
[12:23:21] <rue_mohr> your just not good at asking
[12:23:35] <rue_mohr> ok
[12:23:58] <rue_mohr> so photocopiers use a stepper with a spool and a cable
[12:24:09] <amee2k> i'd go with a bicycle chain and two of the smallest sprockets you can find
[12:24:23] <rue_mohr> the cable is lighter, easier to start/stop
[12:24:26] <amee2k> and the DC motor from a power drill
[12:24:36] <amee2k> on that note...
[12:24:41] <rue_mohr> but the synchrobelt from a inkjet is just as good
[12:24:55] <rue_mohr> 10lbs...
[12:24:57] <amee2k> Casper: how about getting the cheapest power drill and salvaging motor and gearbox from that? its AC then though
[12:24:57] <rue_mohr> hmm
[12:25:10] <rue_mohr> how fast it need to move?
[12:25:24] <Casper> rue_mohr: as I said, 5"/sec
[12:25:35] * rue_mohr finds a ruler
[12:25:39] <amee2k> 5"/sec travel iirc
[12:25:44] <rue_mohr> thats not too bad
[12:25:59] <Casper> not too bad, but too fast for threaded rod
[12:26:05] <amee2k> yeah
[12:26:15] <rue_mohr> vacuum cleaner motors get up to about 10000rpm at about 30V
[12:26:24] <rue_mohr> they have a lot of torque to
[12:26:30] <Casper> and power drill have too much momentum...
[12:26:34] <Casper> and big
[12:26:35] <rue_mohr> you can reverse them if you pull a trick with your H bridge
[12:26:38] <Casper> and noisy
[12:26:53] <rue_mohr> do you need to control position?
[12:27:05] <amee2k> cordless power drills have pretty strong dc motors
[12:27:10] <Casper> yes, but that is a detail
[12:27:22] <amee2k> if you can find small sprockets, chances are it'll work fine without gearbox
[12:27:35] <rue_mohr> 6-32 rod and a shopvac motoro would pull 10lbs easy and move that fast
[12:27:45] <Casper> and noisy
[12:27:49] <rue_mohr> no
[12:27:53] <amee2k> the one i've got is impossible to stop at 3.3V allready
[12:27:58] <rue_mohr> not without the vac blades
[12:28:15] <amee2k> and i reckon it can break your hand at 12V no probs
[12:29:19] <Casper> but no spare shopvac, and people replace the shopvac due to burned out motor
[12:29:19] <amee2k> needs stall current limiting though or it'll trip the overcurrent shutdown on my hack-o-atx when it starts
[12:29:48] * amee2k nudges rue_mohr
[12:30:05] <Casper> here I'ld abuse my ups to power it up :D
[12:30:33] <amee2k> make sure you don't lose any important bodily appendages to it 0.0
[12:31:38] <rue_shop> sorry two transistors
[12:31:43] <rue_shop> (found it)
[12:31:49] <amee2k> i think i can afford that much :)
[12:34:36] <rue_mohr> hahah I took a video of a book, I'll ahvae to post it on youtube now :)
[12:34:51] <amee2k> lol
[12:35:13] <Casper> I wonder how much the charger in the ups can be abused...
[12:35:34] <Casper> as in what if the battery is drained to 10.7V and stay there for hours?
[12:35:42] <Casper> would the charger overheat?
[12:35:42] <amee2k> as long as it has a battery on the output, probably considerably
[12:35:46] <Casper> (it's a smps)
[12:36:12] <amee2k> the charger has to have current limiting and the battery will provide the surge capability
[12:36:31] <amee2k> maybe put a nice cap bank parallel to the battery to catch any inductive kickback from the motors
[12:36:33] <rue_mohr> I'm just trying to figure out how to post the images
[12:36:44] <amee2k> omploader.org?
[12:36:51] <Casper> yeah, but can the current limit work indefinitelly? or it's thermal limited? the original battery is 7AH, I plan to put like... 60AH on it :D
[12:36:56] <rue_mohr> no I have a site, and there is a directory and a script
[12:37:19] <rue_mohr> 60Ah is liek 2 U1 batteries
[12:37:23] <rue_mohr> in parallel
[12:37:36] <amee2k> Casper: an SMP charger should be able to feed a short indefinitely. a linear one probably won't
[12:37:50] <Casper> I have 6 half dead 12V 20AH batts
[12:38:02] <Casper> ranging in the 5-15AH
[12:38:05] <amee2k> smp chargers being more complex by design, they'll probably detect a short and shut down
[12:38:32] <amee2k> 60Ah sounds like a medium sized car battery really
[12:38:50] <Casper> at worse I'll just add a thermally controlled fan I guess...
[12:39:09] <amee2k> ummm
[12:39:26] <amee2k> don't underestimate the amount of cooling that a linear high current regulator needs during a short
[12:39:49] <amee2k> it can easily be 3-5 times the dissipation during regular operation
[12:40:27] <Casper> I know
[12:40:31] <Casper> but it's SMPS
[12:41:26] <Casper> ow well, I'll check the motors I have on hand one day...
[12:41:38] <Casper> first, need to do another cleanup...
[12:41:39] <rue_mohr> !assist circuits
[12:41:40] <tobbor> Possibly http://eds.dyndns.org/~ircjunk/circuits
[12:43:41] <rue_mohr> amee2k, http://eds.dyndns.org/~ircjunk/circuits/crystalTester.jpg
[12:44:15] <rue_mohr> works great
[12:44:26] <amee2k> nice
[12:44:31] <amee2k> any resistor for the LED?
[12:44:38] <rue_mohr> yea, 1K
[12:45:16] <amee2k> hmm can i feed the emitter output to a 74HC14 so i can feed the output into a frequency counter?
[12:45:19] <rue_mohr> mine is on a 1cmx1.5cm pcb with pads to push the crystal up against
[12:45:44] <rue_mohr> amee2k, I wouldn't rely on the freq output of that circuit, it might work tho
[12:45:48] <amee2k> yeah, thats how i want to build mine too. i've got a bunch of small plastic boxes here to put the board in
[12:46:31] <amee2k> stability won't be as good as a "proper" one, i just want to get an idea if its on the right frequency or spewing out random crap
[12:47:04] <rue_mohr> I'v never used mine to check freq, but the light works for between 1.35 and 80Mhz crystals
[12:47:23] <amee2k> hmm ever tried it with watch xtals?
[12:47:44] <amee2k> i think these are the only ones i have that are slower than 1MHz
[12:47:55] <rue_mohr> everything to the right of the 1K resistor is a ac->dc voltage multipler, its dc from there
[12:48:01] <rue_mohr> 32Khz
[12:48:07] <rue_mohr> I cant recall if it worked for those
[12:48:18] <rue_mohr> I keep it in my crystal drawr
[12:49:04] <amee2k> will a 4148 work in place of the 1n914?
[12:49:21] <rue_mohr> prolly yea
[12:49:29] <amee2k> i've got some small shottkys too if they're better... BAT54 or something like that
[12:50:00] <rue_mohr> 914 is in just about everythign tho...
[12:50:27] <rue_mohr> its a forgiving circuit or I could't have got it working
[12:50:40] <amee2k> hehe
[12:50:59] <amee2k> right, i'll prop it up and see how it goes. thanks for the schem :)
[12:51:05] <rue_mohr> I used to write "IT WORKS" in my books beside the schematics I could get going
[12:51:10] <rue_mohr> welcome
[12:52:43] <amee2k> if you're interrested i'll make sure to drop you a note how it went when i got around to try it (maybe tonight or tomorrow)
[12:56:26] <rue_mohr> yes pls
[12:56:32] <rue_mohr> who needed test code?
[12:56:40] <rue_mohr> shower, biab
[12:56:58] <shadewind> hey guys... can someone tell me if there's anything wrong with the fuses 0xC7FF for an 8 MHz external crystal?
[12:59:23] <Tom_itx> shadewind, http://www.engbedded.com/fusecalc/
[13:01:19] <shadewind> Tom_itx: yeah... should be correct...
[13:02:04] <shadewind> Tom_itx: but I get absolutely no reading on my scope when probing either of the osciallator pins
[13:02:06] <sedres> In this schematic - http://tom-itx.dyndns.org:81/~webpage/how_to/atmega168/mega168_led_sch.png - RESET is directly connnected to 5V without any switch. If I'm right, since there is no bar above RESET it will RESET when the pin is high ...
[13:02:12] <jd_ce> hey folks. im having problems integrating a lufa makefile into avr studio 4. I clicked 'use external makefile', and added the sources and changed the board, fcpu, and fclock. I actually dont call any USB functions yet, so it compiles to the same program size as before. but the hex doesnt do anything after I program it to my ch\ip
[13:02:46] <sedres> er, sorry its inverted
[13:02:58] <sedres> Noticed the bubble later.
[13:06:26] <amee2k> shadewind: are you sure the loading caps are correct?
[13:07:09] <amee2k> i've had issues with avr xtal drivers in some mega series parts before because the loading was off... seems that they're a bit touchy this way
[13:17:07] <rue_mohr> shadewind,
[13:17:15] <rue_mohr> hmm
[13:17:17] <rue_mohr> !assist avr
[13:17:18] <tobbor> Possibly http://eds.dyndns.org/~ircjunk/avr
[13:19:03] <rue_mohr> shadewind, the avr resets on low, there is a bar, see the / before RESET?
[13:19:23] <rue_mohr> RESET would be high state reset but /RESET means low state
[13:19:28] <rue_mohr> mkay?
[13:19:49] <rue_mohr> sedres, ^^^ sorry
[13:20:31] <rue_mohr> who needed the led flash code?
[13:20:38] <rue_mohr> I need breakfast
[13:22:36] <rue_mohr> http://eds.dyndns.org/~ircjunk/tutorials/elex/breadtut/main.html
[13:23:47] <rue_mohr> hmm, toms page looks a might like rifrafs
[13:25:41] <robertjw> I need the led flash code
[13:25:48] <rue_mohr> http://eds.dyndns.org/~ircjunk/tutorials/elex/breadtut/main.html
[13:25:53] <rue_mohr> ^^ that code will work
[13:26:10] <rue_mohr> tell me if you need more help with it
[13:26:18] <robertjw> thanks
[13:26:23] <shadewind> amee2k: dunno... tried two crystals and I tried different fuses (both full swing and non-full swing)
[13:26:31] <shadewind> amee2k: must be a board fault or something
[13:26:34] <rue_mohr> robertjw, it flashes a lamp on port B
[13:26:50] <rue_mohr> shadewind, did you change your avr fuses?
[13:27:02] <shadewind> yes
[13:27:16] <rue_mohr> do you have an external crystal?
[13:27:27] <shadewind> yes
[13:27:50] <rue_mohr> what avr and what did you set the fuses to tell me what you set the high and low fuses to
[13:28:54] <shadewind> ATMega8A
[13:29:00] <shadewind> 0xC7FF
[13:29:08] <rue_mohr> fuses are only 8 bits
[13:29:12] <rue_mohr> thats 16 bits
[13:29:13] <shadewind> also tried 0xD7FF
[13:29:20] <rue_mohr> what command line did you use to set them
[13:29:23] <shadewind> no, fuses are 16 bits
[13:29:27] <shadewind> high and low fuses
[13:29:34] <shadewind> each is 8 bits
[13:29:49] <rue_mohr> you set high to d7 and low to ff?
[13:30:06] <shadewind> yeah
[13:30:13] <shadewind> also tried c7 for high and ff for low
[13:30:15] <shadewind> avrdude -c avrispmkii -P usb -p m8 -U hfuse:w:0xc7:m -U lfuse:w:0xFF:m
[13:30:59] <rue_mohr> what crystal speed do you have on it?
[13:31:07] <robertjw> rue_mohr: /usr/lib/gcc/avr/4.5.3/../../../../avr/bin/ld: cannot find crtm328.o: No such file or directory
[13:31:27] <robertjw> avr-gcc -mmcu=atmega328 -I/usr/include/avr/avr/include blink.c -o blink.elf
[13:31:37] <shadewind> rue_mohr: 8 MHz but I also tried one with 16 MHz
[13:31:43] <sedres> rue_mohr: Hi! I followed the example code - http://tom-itx.dyndns.org:81/~webpage/how_to/atmega168/mega168_led_blink_delay_index.php - and it worked for me. I'm wondering if I can use PC0 now as a clock source ?
[13:31:59] <rue_mohr> EEEEEEEEEEEEEEKKK!
[13:32:15] <rue_mohr> I cant tripplestream well right now
[13:32:41] <rue_mohr> robertjw, what code you trying to compile?
[13:32:53] <rue_mohr> (from where)
[13:32:58] <rue_mohr> shadewind, ...
[13:32:59] <robertjw> your example blink.c
[13:33:11] <robertjw> from your link
[13:33:28] <rue_mohr> why is it trying to include crtm328
[13:33:41] <robertjw> that's what I would like to know
[13:33:51] <rue_mohr> one min
[13:34:10] <rue_mohr> ah you used the command line from the page right?
[13:34:19] <robertjw> correct
[13:34:20] <rue_mohr> my library path is different than yours
[13:34:33] <rue_mohr> where is your avr/include directory?
[13:34:54] <rue_mohr> shadewind, how do you have the crystal connected?
[13:34:58] <robertjw> /usr/avr/include
[13:35:20] <rue_mohr> avr-gcc -mmcu=atmega8 -I/usr/avr/include blink.c -o blink.elf
[13:35:25] <shadewind> rue_mohr: http://cl.ly/401l3I0F1b3Q04252H2l
[13:35:27] <rue_mohr> er, change the avr tyep
[13:35:51] <rue_mohr> shadewind, can you still change the fuses on it?
[13:35:57] <rue_mohr> aka program it?
[13:36:02] <shadewind> rue_mohr: nope...
[13:36:13] <rue_mohr> what was the last fuse settings you set
[13:36:15] <shadewind> rue_mohr: I've ruined two boards so far trying different options
[13:36:24] <shadewind> rue_mohr: 0xC7FF
[13:36:29] <rue_mohr> you must not "just try things" with fuses
[13:36:44] <shadewind> rue_mohr: I obviously did not "just try"
[13:36:50] <rue_mohr> express answer in hfuse and lfuse, if you get them backwards you pooch it
[13:37:17] <rue_mohr> shadewind, do you have access to an OLD vga card you can strip parts off?
[13:37:28] <shadewind> hfuse = D7, lfuse = FF
[13:37:42] <shadewind> should be non-full swing 8Mhz crystal
[13:37:49] <shadewind> slow startup et.c.
[13:37:53] <shadewind> SPI enabled and reset enabled
[13:38:10] <shadewind> hfuse=C7 lfuse=FF
[13:38:23] <shadewind> that's with CKOPT set to 0, that is programmed
[13:38:31] <shadewind> and that should be full swing crystal
[13:38:36] <rue_mohr> shadewind, do you have access to an OLD vga card you can strip parts off?
[13:38:43] <shadewind> rue_mohr: not really
[13:39:00] <rue_mohr> what old computer stuff do you have you can wreck
[13:39:10] <shadewind> rue_mohr: non, really
[13:39:16] <shadewind> an old Dell perhaps
[13:39:28] <rue_mohr> ok, that might be workable
[13:39:35] * rue_mohr thinks
[13:39:47] <shadewind> rue_mohr: what are you planning?
[13:39:55] <rue_mohr> you need a clocksource to recover your chip
[13:40:05] <rue_mohr> are you running it on 3.3 or 5V
[13:40:15] <shadewind> 5V
[13:40:18] <rue_mohr> ok
[13:40:50] <rue_mohr> does the dell have any cards in it? or is everything onboard?
[13:41:06] <shadewind> but recovering the board is not really a priority, I have parts for another one
[13:41:52] <shadewind> the weird thing is that the previous revision worked in this aspect but the drive level of the crystal was too high which is why I changed it
[13:44:02] <amee2k> rue_mohr: mmh, how about generating a 1kHz sine wave with a soundcard and feeding that into the clock input?
[13:44:19] <rue_mohr> iirc you need something 1Mhz or faster
[13:44:26] <amee2k> eeew
[13:44:51] <rue_mohr> shadewind, do you have one of toms programmers? they have an unbricking feature
[13:45:21] <rue_mohr> shadewind, you need to get your chips working to make sure their ok, and progress differently with this change your trying to make
[13:45:24] <amee2k> i've got an attiny here with the fuses set to 8MHz that just does "PORTB++" in a loop
[13:45:33] <rue_mohr> :)
[13:45:44] <amee2k> that gives me 8 different frequencies to try and the highest is just over 1.5MHz
[13:45:45] <rue_mohr> I have about a bazillion 4 pin crystal modules
[13:45:57] <amee2k> i've got some too but they're all 40MHz :/
[13:46:05] <rue_mohr> which you can collect easy from all sorts of stuff
[13:46:07] <amee2k> pulled them off old scsi cards
[13:46:09] <specing> Does anyone happen to have a simple smps circuit (with an AVR)?
[13:46:26] <rue_mohr> specing, you dont want to control a smps with an avr...
[13:46:41] <rue_mohr> or any microcontroller
[13:46:49] <specing> Why not?
[13:46:53] <amee2k> when i tried that, i had EMI resetting the controller >_<
[13:47:02] <amee2k> admittedly it was an FBT but still
[13:47:08] <eadthem> do you wnat to mod the feedback loop with the avr or accualy run the smps with the avr
[13:47:17] <rue_mohr> the responce time isn't garunteed enough
[13:47:24] <eadthem> the latter can be hard due to the long reaction time
[13:47:29] <specing> eadthem: the latter
[13:47:59] <rue_mohr> shadewind, do you see my point?
[13:48:12] <sedres> rue_mohr: I tried what I was asking you before. http://pastebin.com/c6Hki8p1 -- Here, considering the F_CPU 8000000UL , the /* ... */ code if uncommented won't work, right?
[13:48:13] <eadthem> in gneeral its alot better to use a DAC for adjusting current and adjusting voltage controls
[13:48:22] <eadthem> and a full analog loop
[13:48:44] <rue_mohr> sedres, F_CPU means NOTHING in that code
[13:49:33] <sedres> rue_mohr: So, I can uncomment and have Port C as clock source for mega 168?
[13:49:44] <sedres> from
[13:49:57] <shadewind> rue_mohr: yes I do
[13:49:58] <amee2k> mmh, you don't have a function generator handy by chance?
[13:50:05] <rue_mohr> sedres, what are you trying to do?
[13:50:19] <shadewind> I have no function generator handy
[13:50:25] <shadewind> but I do have an arduino
[13:50:42] <eadthem> http://en.wikipedia.org/wiki/Resistor_ladder thats how to do a dac sutable for adjusting a SMPS loop with a AVR
[13:50:47] <sedres> rue_mohr: I'm trying to use Port C as clock source with that commented code in the link ...
[13:50:58] <rue_mohr> shadewind, you can use the arduino to generate a clock
[13:51:06] <rue_mohr> lots of people have done that before
[13:51:23] <rue_mohr> sedres, then just remove the delay(100)
[13:51:27] <shadewind> rue_mohr: yeah
[13:51:35] <rue_mohr> why is everyone bricking their avrs all the sudden?
[13:51:58] <shadewind> rue_mohr: I don't know... I've done this lots of times before
[13:52:05] <rue_mohr> bricked it?
[13:52:05] <shadewind> at it worked
[13:52:11] <shadewind> no, set the fuses correctly
[13:52:21] <rue_mohr> I think you did the high and low fuses backwards
[13:52:31] <rue_mohr> cause it causes exactly this problem
[13:53:00] <sedres> rue_mohr: without having to use { DDRC = 1; while(1) { PORTC = 1; PORTC = 0; }; return 0; } and consider output from Port B?
[13:53:11] <shadewind> rue_mohr: I hope
[13:53:21] <rue_mohr> sedres, yes just remove the delay
[13:53:29] <shadewind> rue_mohr: that would make my life easier since I can just make another board and be done with it
[13:53:32] <rue_mohr> shadewind, ah, on that avr
[13:53:47] <rue_mohr> getting the fuses backwards sets it to external RC osc
[13:54:11] <rue_mohr> but it also changes the boot vector
[13:54:11] <shadewind> rue_mohr: besides, I didn't get it backwards: mackbok:egoriser shadewind$ avrdude -c avrispmkii -P usb -p m8 -U hfuse:w:0xc7:m -U lfuse:w:0xFF:m
[13:54:44] <shadewind> but that wouldn't make it unresponsive to ISP though
[13:55:02] <rue_mohr> and that was the last thing that worked
[13:55:14] <rue_mohr> only screwing up the clock settings stops isp from working
[13:55:25] <shadewind> yeah
[13:55:43] <rue_mohr> in anycase, if so, you can recover it with what sedres is doing
[13:56:02] <rue_mohr> shadewind, do you have a working avr you can program without screwing with the fuses?
[13:56:13] <rue_mohr> :)
[13:56:15] <shadewind> rue_mohr: yeah, an Arduino Uno
[13:56:31] <rue_mohr> set that to toggle an output as fast as it can, connect that signal to...
[13:56:34] <shadewind> also, I have a couple of ARM boards but they're 3V3
[13:56:43] <shadewind> OSC in I assume
[13:57:18] <sedres> rue_mohr: I would like to have Port B for led blink and Port C as external clock source ...
[13:57:47] <rue_mohr> to pin 9 (dip)
[13:58:10] <shadewind> rue_mohr: I use QFN
[13:58:12] <rue_mohr> inject squarewave into pin 9 of a dip atmega8
[13:58:36] <rue_mohr> PB6
[13:58:54] <shadewind> yeah
[13:59:09] <rue_mohr> next I need breakfast
[13:59:20] <shadewind> rue_mohr: thanks very much for the help!
[13:59:39] <jacekowski> sedres: it won't be 50% duty cycle square wave
[13:59:50] <jacekowski> with that code
[14:00:21] <jacekowski> sedres: and there are dedicated delay functions in avrlibc
[14:01:24] <sedres> jacekowski: I just need something to confirm that m168 is working, howsoever silly it may sound
[14:02:10] <grummund> sedres: in your pastebin, it's not going to get past the first while loop
[14:02:31] <jacekowski> get HV programmer
[14:03:04] <sedres> yes, but I can change it to use both Port C and Port B.. one as xtal source and other blink?
[14:03:26] <grummund> within the same loop, yes.
[14:03:40] <sedres> yes but
[14:03:50] <specing> So does anyone have a schematic
[14:04:06] <specing> I have a mosfet hooked up to a 100mH inductor
[14:04:09] <jacekowski> sedres: connect two leds in opposite directions
[14:04:13] <sedres> grummund: as jacekowski said it won;t be 50% duty cycle square
[14:04:18] <specing> with a 20A schottky inbetween
[14:04:25] <sedres> ?
[14:04:27] <specing> and 100 uF caps around
[14:04:28] <jacekowski> sedres: between PB1 and PB2
[14:04:32] <sedres> ah, I see it now
[14:04:42] <jacekowski> sedres: and then just have PORTB++; in your loop
[14:04:47] <jacekowski> sedres: and get your clock from PB21
[14:04:49] <jacekowski> PB1*
[14:05:01] <jacekowski> if both LEDs look lit then your avr is working
[14:05:19] <jacekowski> if one stays lit or none then something is wrong
[14:06:46] <sedres> jacekowski: Two led's between PB1 and PB2?!
[14:07:07] <jacekowski> yes
[14:07:12] <jacekowski> in opposite direction
[14:07:18] <jacekowski> and then run PORTB++ in loop
[14:07:29] <sedres> or you mean PB1 - resistor - led - GND and PB2 - res - ~led - GND?
[14:07:34] <jacekowski> no
[14:07:39] <jacekowski> between
[14:07:59] <sedres> but both will be pin high at the same time and then pin low at other time
[14:08:01] <jacekowski> PB1-LEDs-PB2
[14:08:10] <jacekowski> not if you do PORTB++
[14:10:05] <sedres> Ok, so the while will contain PORTB++ and PORTC code ...? Port B for blink test as you suggested and PortC for xtal source
[14:10:54] <sedres> jacekowski: Thanks!
[14:10:54] <jacekowski> no
[14:10:58] <sedres> got it
[14:11:00] <jacekowski> you just use portb for everything
[14:11:09] <sedres> yep
[14:11:32] <jd_ce> god im an idiot. I optimized 2 for loops into one, but forgot to change the type of the integer to 16 bits.
[14:11:35] <jd_ce> there goes several hours
[14:12:40] <grummund> sedres: http://pastebin.com/mmv9TCCF
[14:14:41] <sedres> Yes, thank you.
[14:37:30] <shadewind> rue_mohr: that worked... now I'll have to figure what causes the malfunction
[14:38:24] <rue_mohr> will it work if you set the fuses back to default?
[14:42:11] <shadewind> rue_mohr: what are the defaults?
[14:43:06] <rue_mohr> it wasy n that page
[14:43:32] <shadewind> rightr, I'll check
[14:45:00] <shadewind> weird thing... when I changed the fuses on one of the boards from hfuse=d7 lfuse=ff to hfuse=c7 lfuse=ff it stopped working altogether
[14:45:08] <shadewind> even with the external clock
[14:45:40] <rue_mohr> you need an rc clock to recover from that
[14:46:02] <shadewind> from what specifically?
[14:46:34] <rue_mohr> a resistor and capacitor, see the clock options section of hte datasheet
[14:46:49] <shadewind> recover from what?
[14:47:01] <shadewind> what in those fuses is it that needs an rc clock to recover?
[14:47:10] <rue_mohr> oh I thought you screwed up the avr, evermind
[14:55:17] <shadewind> rue_mohr:avrdude: Device signature = 0x0f0102
[14:55:18] <shadewind> avrdude: Expected signature for ATMEGA8 is 1E 93 07
[14:55:31] <shadewind> rue_mohr: when I try the other board
[14:55:56] <vectory> i cant wrap my head around a 4 gate NAND logic circuit for a simple a xor b
[14:55:59] <rue_mohr> is it really an atmega8?
[14:56:47] <vectory> just found it on the internet, but i would never ever xome up with it by my self
[14:57:15] <shadewind> rue_mohr: yeah, says so
[14:57:16] <vectory> tried to form it using boolean algebra, is there a good channel for people interested in stuff like this, perhaps mathematics?
[14:57:38] <rue_mohr> shadewind, you running the avr on 12V or soemething?
[14:57:51] <shadewind> nope
[14:58:42] <shadewind> 5.00 V says my multimeter
[14:58:55] <shadewind> 5.000 even
[14:58:59] <rue_mohr> hmmm
[14:59:29] <rue_mohr> your failures ahve me proplexed
[14:59:32] <amee2k> rue_mohr: i just tried the xtal tester. i can't seem to get it to oscillate :/
[14:59:33] <devilsadvocate> shadewind: you need to do HVP or something and restore the fuses to a functional set
[14:59:44] <rue_mohr> amee2k, you set it up with the led first?
[14:59:55] <shadewind> I starting to doubt the fuses themselves are really the problem
[15:00:01] <amee2k> "led first"?
[15:00:06] <rue_mohr> shadewind, I think you have an electrical problem
[15:00:10] <shadewind> rue_mohr: me too
[15:00:19] <amee2k> i built it ugly, err... manhattan style on the back of a piece of stripeboard
[15:00:22] <rue_mohr> amee2k, "verbatum" as per schematic, did the crystal light the led?
[15:00:23] <devilsadvocate> shadewind: if after you change the fuses it started behaving this way, i'm pretty sure the problem is the fuses themselves
[15:00:39] <shadewind> devilsadvocate: not if the fuses are correct
[15:00:47] <amee2k> i tried several 12, 16 and 20MHz crystals, no reaction from the LED whatsoever
[15:00:54] <rue_mohr> hmmm
[15:00:59] <rue_mohr> may I see the circuit?
[15:01:00] <shadewind> if the fuses are set like they are supposed to be set and it doesn't work, then it's likely an electrical problem
[15:01:03] <amee2k> i even turned off the desk lamp to make sure i'm not missing any
[15:01:09] <amee2k> sure, lemme up a pic
[15:01:12] <shadewind> rue_mohr: my schematic or amee2k's?
[15:01:21] <devilsadvocate> shadewind: that doesnt take into account pebkac
[15:03:06] <shadewind> devilsadvocate: no... but I read the fuses back and they were correct.
[15:03:43] <devilsadvocate> shadewind: i've had fuse related issues where the problem manifests itself only after reset
[15:04:41] <shadewind> I don't know what you are really trying to say but I checked the fuses
[15:05:11] <shadewind> I'm thinking it might be a soldering issue... maybe the QFN ATMega8A had some bridging somewhere or something
[15:05:29] <amee2k> rue_mohr: http://ompldr.org/vYjFxcg/CIMG8681b.jpeg
[15:06:13] <shadewind> rue_mohr: you saw the way I connected the crystal before, right?
[15:06:14] <amee2k> you can see some random xtals i tried with it, and i tried almost my entire baggy of new 16/20MHz xtals
[15:06:56] <amee2k> only component substitutions i made are 4148 for the diodes and 47k for the 33k
[15:07:07] <amee2k> but i tried 10k, 22k and 100k as well, no change
[15:10:20] <rue_mohr> http://www.youtube.com/watch?v=XSGBVzeBUbk&feature=related
[15:11:31] <rue_mohr> amee2k, tracing this is gonna be a challange
[15:12:11] <rue_mohr> your capacitors look WAY too big
[15:12:32] <rue_mohr> oh I just found the small ones
[15:12:44] <amee2k> the reddish-brown ones are 100V types
[15:12:46] <rue_mohr> I need another angle
[15:12:58] <amee2k> the two small amber ones are ceramics
[15:13:11] <amee2k> forther from the right? or from the left?
[15:13:14] <amee2k> further* even
[15:13:40] <rue_mohr> hmm, if you take a 1K resistor and go from +5 to the base of the led transsitor, does the led light?
[15:14:12] <rue_mohr> amee2k, I dare you to build one like this with SMT parts
[15:15:23] <rue_mohr> also amee2k I dont see your power led on
[15:15:30] <rue_mohr> might be the angle
[15:15:37] <amee2k> i made the shot with power turned off
[15:15:47] <amee2k> with the 1k resistor the LED lights
[15:15:53] <amee2k> not very bright but visible
[15:16:12] <shadewind> rue_mohr: I'll solder a third board... maybe I had bad luck the two first times...
[15:16:32] <rue_mohr> ok, if you repeatedly short the CE of the first transistor does the ledblink a bit?
[15:16:49] <rue_mohr> shadewind, na, there is something daftly wrong somewehre
[15:17:07] <rue_mohr> shadewind, test the next avr WITHOUT changing ANY of the fuses
[15:17:14] <shadewind> rue_mohr: yeah sure
[15:17:30] <shadewind> rue_mohr: it would be stupid to do that a third time
[15:18:08] <amee2k> rue_mohr: yep, but its barely visible. i had to turn off the lights to see it
[15:26:03] <amee2k> rue_mohr: http://ompldr.org/vYjFyNA/CIMG8682b.jpeg http://ompldr.org/vYjFyNQ/CIMG8683b.jpeg << here are two more angles, respectively from the right and left compared to the first one
[15:27:27] <amee2k> (and no, the green led isn't lit, thats just a reflection from the flash)
[15:27:56] <Steffanx> Waow, amee2k
[15:28:13] <amee2k> wut
[15:28:27] <Steffanx> Fancy soldering :)
[15:28:27] <rue_mohr> have a shakey hand?
[15:28:40] <amee2k> me? no, why?
[15:28:43] <amee2k> Steffanx: thanks :)
[15:28:49] <rue_mohr> can you "scratch" a short against the CE of the first transistor and see if the led comes on?
[15:29:04] <rue_mohr> needs to have a lot of bounce
[15:29:42] <shadewind> firing up the hot air now...
[15:30:38] * rue_mohr grabs the nearest polititian and challanges shadewind to a hot air faceoff
[15:30:52] <amee2k> 0.0
[15:31:11] <amee2k> thats biological warfare
[15:32:16] <rue_mohr> amee2k, hows the scratching?
[15:32:16] <amee2k> rue_mohr: yeah, rubbing a piece of wire roughed up with some coarse sand paper against the pin makes the led flicker
[15:32:26] <rue_mohr> ok
[15:32:42] <rue_mohr> so, the problem is right in the first few components
[15:32:43] <amee2k> still not very bright though
[15:33:05] <rue_mohr> !assist circuits
[15:33:06] <tobbor> Possibly http://eds.dyndns.org/~ircjunk/circuits
[15:33:46] <rue_mohr> 2N2222 is EBC?
[15:34:12] <amee2k> cbe in my datasheet
[15:34:21] <rue_mohr> could be
[15:34:37] <amee2k> i remember "pin 1 is collector"
[15:35:00] <rue_mohr> 103
[15:35:03] <rue_mohr> .01uF
[15:35:09] <rue_mohr> 10nf
[15:35:30] <amee2k> ?
[15:35:32] <rue_mohr> 10000pf
[15:35:40] <rue_mohr> the ceramic, I'm trying to work out who is who
[15:35:53] <rue_mohr> looking for the 1000pf (1nf) and 100pf
[15:35:57] <amee2k> wtf
[15:36:05] <amee2k> yeah, that one is wrong >_<
[15:36:13] <rue_mohr> so there should be a "101" and a "102"
[15:36:14] <sedres> http://yuki-lab.jp/hw/avrdude-GUI/error-e.html .. What does this mean - "Not Connect Programmer with Target circuit" ?
[15:36:28] <amee2k> thats supposed to be 102... the 1n on the oscillator transistor
[15:36:33] <rue_mohr> I cant tell who is who, so I'm just trying to match up counts
[15:36:37] <amee2k> lemme fix that
[15:37:18] * rue_mohr questions the servicability of amee2ks circuit, but dosn't say anything to be polite
[15:37:29] <amee2k> lol
[15:37:45] <amee2k> its temporary for the moment :P
[15:37:54] <rue_mohr> 47k....
[15:38:09] <amee2k> thats what i substituted for the 33k one because i didn't have any
[15:38:17] <rue_mohr> fiar enough
[15:38:29] <rue_mohr> I have to go shopping, 1hr trip...
[15:38:39] <rue_mohr> still here when I get back?
[15:38:41] <amee2k> i tried 20:58 < amee2k> only component substitutions i made are 4148 for the diodes and 47k for the 33k
[15:38:44] <amee2k> 20:58 < amee2k> but i tried 10k, 22k and 100k as well, no change
[15:38:57] <rue_mohr> yup
[15:38:57] <amee2k> and the 10n that i got wrong. working on changing that one right now
[15:39:24] <shadewind> rue_mohr: now we shall se...
[15:39:24] <rue_mohr> 100pf and 1nf
[15:39:25] <shadewind> see*
[15:39:36] <rue_mohr> "101" and "102"
[15:39:55] <rue_mohr> hahah C4 is the ONLY labeled component
[15:39:58] <rue_mohr> biab
[15:40:19] <sedres> I get fourth error message mentioned here - http://yuki-lab.jp/hw/avrdude-GUI/error-e.html . I've checked the connection from ISP to AVR chip. The connections are right and I used the same connection to transfer hex to mega 168 which worked. Regarding the third possibilty mentioned there of target circuit turned off -I've checked with DMM, its getting right voltages. I'm not able to understand what it means by - 'Not Connect Programmer
[15:41:10] <shadewind> rue_mohr: avrdude: Device signature = 0x1e9307
[15:41:11] <sedres> Can anyone guide me on this?
[15:41:14] <shadewind> rue_mohr: so far so good
[15:41:21] <shadewind> sedres: responds to ISP
[15:41:59] <sedres> ?
[15:43:56] <shadewind> rue_mohr: so, any idea about what I should try=
[15:43:58] <shadewind> *=
[15:43:59] <shadewind> *?
[15:44:01] <shadewind> crap
[15:45:25] <amee2k> o.O
[15:45:37] <amee2k> changed the 10n for the proper 1n cap
[15:45:53] <amee2k> now its even deader than before... not even the wire scratching thing works anymore
[16:00:02] <sedres> How reliable it is to use internal RC oscillator over external crytals?
[16:00:17] <Tom_itx> not as accurate
[16:00:24] <Tom_itx> if you need critical timing use the crystal
[16:00:31] <Tom_itx> if you're playing around either is ok
[16:01:12] <sedres> Critical timing, so how inaccurate RC oscillator can be?
[16:01:27] <sedres> any example ..
[16:02:29] <shadewind> rue_mohr: woohoo! IT WORKS :)
[16:03:53] <sedres> shadewind: :)
[16:04:39] <shadewind> I have no idea what was wrong though
[16:04:51] <shadewind> maybe my fuses actually were wrong
[16:05:05] <shadewind> or the two first boards had botched soldering
[16:05:16] <shadewind> but I'm not sure I care all that much really right now :)
[16:06:05] <amee2k> sedres: from my experience you're lucky if its within 10% of the nominal frequency
[16:06:34] <amee2k> without calibration, that is. the datasheet has it you can reasonably get 1% with calibration
[16:07:00] <amee2k> a crystal will be pretty much as stable as a swiss watch for most purposes.
[16:07:26] <sedres> Interesting
[16:08:00] <Ikarus> given that most are run by an ordinary 32.687kHz crystal, yes
[16:08:55] <amee2k> if you need even better stability, you can make a ghetto temperature controller for like 2$ in parts
[16:09:22] <amee2k> dual opamp, temperature sensor, medium sized power transistor
[16:10:50] <amee2k> http://www.genesisradio.com.au/G59/phase11.html << something like this will be enough to keep the the xtal within a few degrees to minimize drift
[16:17:19] <sedres> Yes, thanks amee2k
[16:17:49] <amee2k> you're welcome
[16:18:20] <Tom_itx> sedres pm
[16:43:01] * amee2k nudges rue_mohr
[16:44:19] <rue_mohr> I'm back
[16:44:40] <rue_mohr> did it work?
[16:44:46] <amee2k> mmh what work?
[16:44:57] <rue_mohr> after the capacitor correction
[16:45:01] <amee2k> i changed the 10n cap to 1n, then it didn't do anything anymore
[16:45:15] <amee2k> not even blinking when i scratch it with the wire
[16:46:07] <rue_mohr> you know there are supposed to be 2 1n caps in there?
[16:46:26] <rue_mohr> tell me the markings on the caps you have there
[16:46:43] <amee2k> yeah, the coupling cap is the right size
[16:46:56] <amee2k> i only got the one on the oscillator wrong
[16:47:22] <rue_mohr> ok, may I inquire to the markings on your caps?
[16:47:43] <amee2k> the two on the oscillator are 102 and 101, the coupling cap is .001, and the smoothing cap is 562
[16:48:53] <rue_mohr> huh, ok
[16:49:03] <amee2k> mmh?
[16:52:51] <rue_mohr> I'm confused, I'll take some pics of mine in a min for ya
[16:53:34] <amee2k> okay
[16:58:42] <ziph> abcminiuser: What's the LUFA announcement with respect to NXP?
[17:00:02] <abcminiuser> It's supposed to be a launch of their new library
[17:00:20] <ziph> Based on LUFA?
[17:00:34] <abcminiuser> Couldn't say that...
[17:01:33] <ziph> I would find it surprising given that their LPC controller is about as incompatible with LUFA as you could get. :)
[17:01:47] <abcminiuser> Yes, that would certainly be interesting
[17:06:07] <Tom_itx> abcminiuser you seen the sparkfun thread about uno and fedora?
[17:06:42] <abcminiuser> Tom_itx, no?
[17:06:49] <abcminiuser> Link?
[17:06:56] <Tom_itx> zlog
[17:07:01] <Tom_itx> earlier today
[17:09:31] <Tom_itx> http://forums.adafruit.com/viewtopic.php?f=25&t=23439&p=123688#p123688
[17:09:41] <Tom_itx> err adafruit i guess
[17:11:46] <abcminiuser> Tom_itx, that's not my code
[17:11:51] <abcminiuser> I don't think
[17:12:00] <Tom_itx> oh
[17:12:29] <Tom_itx> i think it's a fedora problem
[17:12:34] <abcminiuser> That looks like the bootloader's no good
[17:12:44] <abcminiuser> It's not a USB error, but a higher layer error
[17:12:53] <Tom_itx> we tried some test programs here that worked
[17:13:39] <rue_mohr> I'm not havin much of a seuccess rate today helping people am I?
[17:13:46] <Tom_itx> me either
[17:13:53] <Tom_itx> wanna take over on mine?
[17:14:01] <Tom_itx> sedres
[17:14:22] <Tom_itx> i'm gettin tired
[17:17:11] <amee2k> rue_mohr: hey, but at least you're trying :)
[17:20:18] <rue_mohr> amee2k, oops I lied
[17:20:29] <Tom_itx> rue_bed
[17:20:31] <Tom_itx> <sedres> http://pastebin.com/EcT5DjUa
[17:20:34] <Tom_itx> check it out
[17:20:39] <amee2k> lol
[17:20:40] <Tom_itx> i gotta go carve pumkins
[17:20:41] <amee2k> :P
[17:20:52] <Tom_itx> rue_mohr !! FIX IT DAMMIT!!
[17:20:53] * amee2k idly passes the used shotgun on to Tom_itx
[17:21:08] <rue_mohr> 100pf .001uF .001uf .0047uf
[17:21:09] <oohy9n> amee2k does he have shells?
[17:21:33] <amee2k> is that edible?
[17:21:42] <oohy9n> no, for the shotgun
[17:21:42] <Tom_itx> sedres, rue promised he would help you to the bitter end!
[17:21:44] <rue_mohr> transistors are MPS3392 for the first one, and 2N5210 for the second
[17:21:49] <amee2k> oh, i see
[17:21:58] <rue_mohr> led resistor 330 ohms
[17:22:13] * oohy9n passes amee2k some money, conicidently it is enough to buy 20 shotgun shells
[17:22:13] <rue_mohr> 33K and 1K for the other
[17:22:23] <amee2k> hmm i see
[17:22:47] <rue_mohr> honestly, I'm guessing they are random npn transistors from the bin
[17:23:02] <amee2k> i'd be considerably surprised if either the bias resistor or the smoothing cap are this critical
[17:23:10] <amee2k> i've used some 2222s i had around
[17:23:33] <amee2k> i like them enough to buy them
[17:23:36] <ziph> Where's the circuit you're talking about?
[17:24:28] <amee2k> ziph: http://eds.dyndns.org/~ircjunk/circuits/crystalTester.jpg http://ompldr.org/vYjFxcg/CIMG8681b.jpeg http://ompldr.org/vYjFyNA/CIMG8682b.jpeg http://ompldr.org/vYjFyNQ/CIMG8683b.jpeg << here, sans schem
[17:24:30] <rue_mohr> yea 2222 is what I stock new
[17:25:14] <amee2k> i get 3904/6 and 2222 new, plus whatever i can get good deals on when i need some
[17:25:15] <rue_mohr> 2 mins for my images
[17:25:34] <rue_mohr> I think I use 06 for pnp new stock
[17:25:48] <rue_mohr> 100 for like $5?
[17:25:57] <oohy9n> that circuit looks like it could fry pancakes
[17:25:59] <amee2k> i got lots of bc337-16 and -25 that i got cheap a while ago
[17:26:05] <rue_mohr> maybe its $15
[17:26:13] <ziph> amee2k: Well, the first section is an oscillator, so those components could be critical. :)
[17:26:15] <rue_mohr> I have enough of those from vcr's
[17:26:46] <amee2k> ziph: i've got all except for the 33k bias resistor the same, and for that i tried a bunch of different values. none seemed to work
[17:27:26] <ziph> amee2k: You're using a 2n2222 as a transistor?
[17:27:27] <amee2k> i got some strips of belted bc5xx that i don't remember the exact types right now
[17:27:32] <amee2k> ziph: yep
[17:27:44] <sedres> Hi! Can anyone help me fix this - http://pastebin.com/EcT5DjUa ?
[17:28:21] <amee2k> ziph: symptoms is its not doing anything at all, whatever i put in as xtal
[17:28:24] <rue_mohr> amee2k, ok, extra high rez
[17:28:32] <rue_mohr> !assit projects
[17:28:36] <rue_mohr> !assist projects
[17:28:42] <rue_mohr> !time
[17:28:46] <rue_mohr> pff
[17:28:47] <amee2k> o.O
[17:29:04] <ziph> amee2k: Those emitter follower oscillators actually rely on the tank circuit formed by the C's and crystal to have an AC gain exceeding 1, maybe varying the ratio of the 1000pf to the 100pF will help. And what frequency crystals are you using?
[17:29:38] <rue_mohr> !time
[17:29:39] <tobbor> My watch says its 15:22 Sun Oct 30 2011
[17:29:43] <rue_mohr> !assist projects
[17:29:43] <tobbor> Possibly http://eds.dyndns.org/~ircjunk/projects
[17:29:56] <amee2k> ziph: i tried a some new 12, 16 and 20MHz ones, the type you'd want for sticking on an MCU. plus various used ones ranging from 6 to 14MHz
[17:31:22] <rue_mohr> amee2k, http://eds.dyndns.org/~ircjunk/projects/crystaltest/
[17:31:30] <rue_mohr> pls wait
[17:32:20] <amee2k> nice work :)
[17:32:55] <ziph> amee2k: Try it with both C's 100pF
[17:33:30] * amee2k dives back into the junk box
[17:35:38] <ziph> amee2k: And if you have a radio that can receive HF/Shortwave you could use it to determine if the oscillator is running (and the detector is faulty) or if the oscillator just isn't running.
[17:36:03] <amee2k> i've got an 11m set but it it currently being ghettojammed :/
[17:36:19] <ziph> Ghettojammed? :)
[17:37:40] <rue_mohr> amee2k, see anything that helps?
[17:37:46] <amee2k> ziph: i suspect someone is operating some extremely rf offensive device for a few hours every other evening
[17:38:27] <amee2k> the noiseband seems to have lot of drift and shows several peaks. when its smack dead on, all channels read S9+
[17:38:40] <ziph> amee2k: Maybe you have a spark gap enthusiast in the neighborhood. ;)
[17:39:16] <oohy9n> lol
[17:39:18] <amee2k> i want to use his c**k as spark gap >_<
[17:39:28] <amee2k> this is exceedingly annoying
[17:40:10] <amee2k> ziph: are two 120p close enough?
[17:40:15] <ziph> amee2k: Yeap.
[17:43:36] <amee2k> still no cookie :/
[17:44:08] <ziph> Is there any DC on the base of the second transistor?
[17:45:23] <amee2k> hmm this is strange
[17:45:33] <amee2k> i'd say no since the LED isn't lighting up
[17:45:42] <amee2k> but there is voltage on the base o.O
[17:45:58] <amee2k> lemme pull that second transistor out and check it
[17:46:11] <ziph> What voltage?
[17:46:25] <oohy9n> 1.81 Gigawatts
[17:46:33] <amee2k> almost 5
[17:46:36] <amee2k> V
[17:46:42] <oohy9n> lol
[17:46:58] <h4x0r`> yo :)
[17:47:02] <ziph> 5V? That should be sufficient. :)
[17:47:13] <h4x0r`> 5v 10A
[17:47:17] <h4x0r`> ROFL
[17:47:25] <amee2k> wow
[17:47:30] <amee2k> tranny is a slug
[17:47:38] <amee2k> >_<
[17:49:53] <h4x0r`> has anyone used usb 32005 irl?
[17:50:05] <h4x0r`> just wondering if the eagle footprint makes drillholes
[17:50:46] <ziph> You can't tell by looking at it?
[17:50:54] <h4x0r`> nup
[17:50:56] <h4x0r`> im not sure
[17:51:05] <jd_ce> o.O
[17:51:12] <jd_ce> hide all the layers besides drill
[17:51:20] <h4x0r`> if the eagle footprint will make holes if i send it off to a fabhouse
[17:51:21] <ziph> Time to switch to DipTrace then. ;)
[17:51:22] <h4x0r`> just sayin
[17:51:29] <h4x0r`> lol :p
[17:51:43] <ziph> h4x0r`: Make Gerbers and check them there?
[17:51:52] <amee2k> weee
[17:52:04] <amee2k> i put a bc337 in its place and now it works :)
[17:52:16] <amee2k> something seems to have killed that 2222
[17:53:42] <ziph> If you BC337 dies eventually try a resistor on the base. :)
[17:53:46] <ziph> your
[17:53:54] <amee2k> lol
[17:54:21] <amee2k> hmm on second thought that probably has a point
[17:56:55] * h4x0r` presumes altium is overkill
[17:57:54] <ziph> It's a good product if your Mandarin is good. ;)
[17:58:00] <h4x0r`> rofl
[17:58:27] <h4x0r`> not sure i get it but i presume most good support is in chinese ?
[17:58:35] <h4x0r`> pffft anyway
[17:58:43] <h4x0r`> gday.. ziph ;)
[17:58:49] <ziph> No, they were an Australian company. They fired all of their Australian staff and moved to China.
[17:58:56] <h4x0r`> ahh
[17:59:02] <h4x0r`> sounds like quantas
[17:59:04] <h4x0r`> lol
[17:59:36] <amee2k> ziph, rue_mohr, thanks a lot, i think it works now :)
[18:09:10] <sedres> I'm getting this error - http://pastebin.com/EcT5DjUa . Should I go ahead with -F force flag?
[18:10:03] <JanneP> -F probably wont help
[18:10:22] <JanneP> try adding a -B switch for slower ISP clock mayby it will help
[18:11:12] <sedres> Yeah, that's what the action method says. I asked for -F flag 'coz it helped someone here - http://www.avrfreaks.net/index.php?name=PNphpBB2&file=viewtopic&t=62715
[18:11:59] <sedres> JanneP: How do I decide the bitclock?
[18:12:08] <robertjw> rue_hohr: your simple test case builds and uploads fine with the Fedora tool chain, but a bigger test case bringing in the arduino library fails.
[18:12:32] <JanneP> sedres: add a -B 16 switch to the command line
[18:12:45] <robertjw> rue_hohr: your simple test case builds and uploads fine with the Fedora tool chain, but a bigger test case bringing in the arduino library fails.
[18:12:53] <JanneP> the bigger the number after -B the slower it should be..
[18:13:11] <robertjw> rue_hohr: your simple test case builds and uploads fine with the Fedora tool chain.
[18:13:28] <sedres> JanneP: Could you explain how slower ISP helps?
[18:13:42] <robertjw> http://forums.adafruit.com/viewtopic.php?f=25&t=23439&p=123722#p123722
[18:13:53] <robertjw> test results in the last post on this thread
[18:14:33] <JanneP> sedres: if it's something related to connection it might help. or if the targer is running below 2MHz it is required with USBasp
[18:15:05] <Tom_itx> You can use the -B option to specify the ISP speed. By default the value is 10 which means 100KHz clock, this is good for target clock speeds > 500KHz. If you want the high speed clockrate (400KHz) for target frequencies > 4MHz you can use "-B 1" to speed up programming
[18:15:34] <Tom_itx> In general, the clock frequency should be at least 4 times larger than the target clock frequency. Try "-B 32" if you're having clocking issues, that should handle even 128khz clocks.
[18:15:52] <Tom_itx> SPI clock frequency in KHz = 1000/(1.5+B) where B is the delay value
[18:16:01] <JanneP> sedres: http://www.avrfreaks.net/index.php?name=PNphpBB2&file=printview&t=107339&start=0
[18:16:09] <JanneP> seen that one?
[18:16:34] <sedres> yep
[18:17:42] <JanneP> ok so old firmware is not a problem?
[18:19:27] <sedres> I'm using avrdude package version 5.10. Ubuntu 11.10 linux. http://pastebin.com/e4NcfyDS
[18:20:19] <sedres> Regarding firmare,
[18:21:04] <robertjw> sedres: if you add the -F option to avrdude's command line, then it will ignore the signature bug and move on
[18:21:54] <JanneP> though, if the SCK speed was too high, it would not be able to even enter programming mode
[18:23:13] <JanneP> why both switches D & e ? disable auto erase and then do a chip erase..
[18:24:18] <sedres> ok
[18:26:49] <sedres> http://pastebin.com/PTjzke8v Please scroll to bottom for the last command output
[18:28:30] <JanneP> seems like a hardware problem or too high ISP speed to me..
[18:28:33] <sedres> Thanks, even flags order helped to get atleast the device signature... What next JanneP ?
[18:28:41] <sedres> umm
[18:28:43] <JanneP> try the slow mode jumper if you already haven't
[18:29:47] <sedres> Too high ISP speed, how can I confirm that? No, I've not tried slow mode jumper before ...
[18:32:45] <wrongle> hey guys, do you know where i can buy those header pins as such here : http://imakeprojects.com/Projects/avr-tutorial/
[18:32:49] <wrongle> i didn't see them online?
[18:33:22] <ziph> Look for 0.1" headers.
[18:34:57] <wrongle> searched, they are all in a "row"
[18:35:02] <wrongle> vs. 2x3
[18:35:26] <ziph> You can buy them in break-off strips and snap off any length you want.
[18:38:11] <wrongle> sweet, thx -
[18:38:25] <sedres> JanneP: Could you guide me regarding slow mode jumper?
[18:49:51] <sedres> JanneP: No, I'm not sure how to check that.. linking any reference for the same, I can try
[18:58:13] <rue_mohr> amee2k, ok, so?
[18:58:33] <rue_mohr> do you think you threw a pooched 2222 back in the bin?
[18:59:17] <rue_mohr> amee2k, http://eds.dyndns.org/~ircjunk/projects/crystaltest/ you see all 3? is there something I missed?
[19:12:54] <rue_mohr> amee2k, whats odds of new pics?
[21:39:32] <rue_mohr> I think their low rue
[21:39:35] <rue_mohr> I agree
[21:55:24] <Tom_itx> rue_mohr
[21:55:26] <Tom_itx> what's up?
[21:56:10] <rue_mohr> I just pasted in the adc code to the air-muscle control stuff
[21:56:14] <rue_mohr> so I'll test that
[21:56:23] <rue_mohr> still have the lcd and kbd hooked up
[21:56:25] <Tom_itx> got time to look at a makefile?
[21:56:29] <rue_mohr> ggonna lose that soon
[21:56:43] <rue_mohr> I'll see if I can close the loop with prop feedback
[21:57:38] <rue_mohr> I suppose if I have a timed interrupt going off I can use it as a clock source for PID calcs
[22:24:31] <rue_mohr> --> rue is going to his shop now <--
[22:25:55] <rue_shop> I'm over here now
[22:27:14] * Casper throws rue in bed?
[22:28:16] <rue_shop> no I have a little time left
[22:28:20] <rue_shop> about an hour
[22:28:31] <Tom_itx> didn't want him to get discusted with me
[22:28:39] <rue_shop> I shall take over the world and have some time left to settle for a good nights sleep
[22:29:05] <Tom_itx> and it wasn't avr related
[22:29:45] <rue_shop> this sucks, my lcd is plugged into my adc port
[22:29:59] <rue_shop> I HAVE to rewrite the lcd lib now...
[22:30:04] <rue_shop> damn
[22:30:16] <rue_shop> see how fast I can do this
[22:30:24] * rue_shop puts on some loud techno
[22:30:35] <atom1> couple clicks in the .h file if you used mine
[22:44:39] <rue_shop> fixed
[22:44:44] <rue_shop> lots of time left
[22:47:30] <Tom_itx> now eclipse is all screwed up
[23:07:35] <rue_shop> whats eclipse for?
[23:07:49] <rue_shop> wow, I can move the muscle within 1 adc count
[23:12:20] <rue_shop> this is a slow system I think that one controller can handle a lot of channels
[23:16:40] <rue_shop> having an avr with a lcd and keybaord set up on it is kinda fun
[23:17:21] <Tom_itx> for the c env i think
[23:17:42] <Tom_itx> the batch was his to tie it all together i guess
[23:17:47] <Tom_itx> err make
[23:18:03] <Tom_itx> but now eclipse complains it can't find it's editor
[23:18:13] <rue_shop> whats eclipse?
[23:18:25] <Tom_itx> linux gui
[23:19:29] <rue_shop> gui? why do you need a gui?
[23:19:50] <rue_shop> ok , here goes, here's hoping the feedback isnt reversed
[23:20:14] <atom1> http://www.eclipse.org/downloads/
[23:20:33] <Tom_itx> you'll love it. it's java
[23:21:11] <Tom_itx> it's fubar right now
[23:21:26] <rue_shop> nowonder, its java
[23:21:36] <rue_shop> the feedback dir is right!
[23:23:02] <rue_shop> hmm how to deal with 80 units of deadband
[23:23:06] <atom1> k, the meter has expired here
[23:23:29] <rue_shop> pwm0 = (target - Analog(0));
[23:43:01] <rue_shop> the math is NOT behaving
[23:43:11] <rue_shop> avrs dont like mixing signed and unsigned numbers
[23:48:05] <ziph> rue_shop: What are the values?
[23:49:09] <rue_shop> washing machine and dishwasher
[23:49:14] <rue_shop> they can do this, I know they can
[23:49:18] <rue_shop> my code just isn't right
[23:55:31] <rue_shop> I think I have a bunch of things to account for here
[23:55:54] <rue_shop> min times for hte valves to break (large deadband) and I have to limit the on time of them too