#avr | Logs for 2013-05-22

Back
[00:17:45] <thetruthisoutthe> theBear <= yeah, but fpgas are more suited to this kind of thing
[00:18:25] <theBear> maybe they are, but if you already got a micro, chances are the traditional approaches will be trivial to implement
[00:32:33] <Roklobsta|2> what's trhe minimum number of guaranteed flash cycles of an avr?
[01:03:37] <theBear> Roklobsta|2, depends on the series, as usual very clearly stated in the datasheet, from memory the older ones were 1000 and the newer ones (tiny/newer megas) are 10,000
[01:04:32] <Roklobsta|2> yeah derrr, "– Write/Erase Cycles:10,000 Flash/100,000 EEPROM"
[01:05:27] <theBear> :)
[01:05:57] <theBear> it's only REALLY old ones that are 1000, like 90sXXXX and atmega8515
[01:09:39] <Roklobsta|2> it's all coming back , 10k seems to be the norm for most kinds of cpus with flash
[01:09:47] <Roklobsta|2> or any kind of flash.
[01:10:03] <Roklobsta|2> nono i retract that last statement
[01:10:49] <thetruthisoutthe> Roklobsta|2 <= 1000 - 10000
[01:11:36] <theBear> Roklobsta|2, heh, just in time
[01:11:53] <Roklobsta|2> funny, my powers to debug my statemachine are already improving with BEER
[01:12:52] <theBear> BEER eh ? that some kind of err, Bit Error something Ratio ?
[01:13:27] <Roklobsta|2> no i have a big ol' state machine to debug (it compiles fine) now.
[01:14:02] <theBear> lol, a big ol' state machine like that guy with the traffic lights and 900 lines of code the other day ? <grin>
[01:21:21] <iCopter> whats the best avr kits these days for a beginner? i want to work with attinys ideally, and the arduino AVR's, but would like to get into FGPA's at a later date
[01:24:18] <theBear> iCopter, there isn't much difference between most basic devboards, 90% of just about all small avr devboards is just the chip and maybe a reg/crystal/reset circuit, and MAYBE a ft or max or similar chip for basic serial comms, then just variations between which pins are exposed and how the programming connector is wired
[01:24:53] <theBear> and arduino don't make avrs, they are just a selection of basic boards like i just described with a few different models of avr on them
[01:58:28] <GuShH> theBear: my god... that epoxi with steel in it is pure crap, at least the one I got...
[01:58:36] <GuShH> it crumbled at 250°C and began to smoke
[01:58:59] <GuShH> I even used a digital scale (good for 0.01 gram resolution, repeatable) to weigh the parts
[01:59:25] <GuShH> surfaces where cleaned with the dremel tool using a stone bit and then cleaned up with alcohol (in case there was any grease etc)
[01:59:30] <GuShH> so it's certainly not my fault it sucks.
[01:59:41] <GuShH> let it cure for 30hs, says 24hs on the package
[02:00:11] <GuShH> I can push it off with a screwdriver once it reaches 250°C, it just falls apart. BS marketing or faulty batch?
[02:01:13] <GuShH> I wonder if the one Loctite sells is any better
[02:01:23] <GuShH> At this point I just wasted my time and money
[02:01:51] <GuShH> Any ideas on how to fasten a cylinder to torque out a 50mm nut then? :>
[02:02:12] <GuShH> the cylinder is 50cm in diameter.
[02:04:49] <GuShH> granted the radial expansion of the stainless steel coupler could've been the reason the epoxi failed, but since it degraded instead of just cracking I think it's a fault on the composition or just bs marketing like I said....
[02:10:02] <GuShH> mixture was perfectly even, it looks discoloured so it clearly decomposed with the heat... what a piece of crap, and to think mechanics use this stuff to fix manifolds!!!
[02:43:35] <iCopter> theBear what about tom_itx's version 1.1 miniusb AVR programmer? (i think thats the right name)
[02:44:54] <iCopter> btw when i said arduino AVRs, i meant the chips utilized in the arduino, pardon me
[02:56:15] <theBear> GuShH, err, i'd have to know what the marketing said to answer that
[02:58:04] <theBear> iCopter, i haven't owned one of toms programmers so i got no idea what the various ones are called, but i do know they're technically good and everything i seen from him is well designed and put together and can program any avr i've heard of in any mode/connection... not to mention a very reasonable price and made by someone i like :)
[02:59:43] <GuShH> theBear: the marketing says it's used to fix manifolds and injection moulds, among other things
[02:59:56] <GuShH> which implies it can handle the temperature and compression/force
[03:00:29] <GuShH> iCopter: go with Tom's, don't waste money on Faguinos.
[03:02:16] <theBear> i suppose, depends what kinda manifolds and injection moulds they mean :) both can apply to things that don't ever see anything more than a few degrees above ambient :)
[03:02:28] <GuShH> pfft
[03:02:31] <GuShH> regular engine manifolds
[03:02:41] <GuShH> and injection moulding often involves heat
[03:03:01] <GuShH> why are you defending them! I even used a scale to get the proper amounts and everything, I'm not at fault! :(
[03:03:05] <theBear> iCopter, and while we don't advocate it, you can use the arduino software stuff on just about any avr if you REALLY want to, but we don't advocate it
[03:03:40] <GuShH> pft you can do whatever you want, just don't waste money by being misinformed!
[03:03:55] <theBear> i was impressed with the effort you went to for accuracy, and not defending them, just reminding you to be careful not to assume stuff :)
[03:04:23] <theBear> hehe, was that last comment for me or him ?
[03:05:06] <GuShH> for everyone I guess
[03:05:13] <GuShH> well, I rarely get to use the digital scale
[03:05:19] <GuShH> I bought it for that kind of stuff though
[03:06:40] <GuShH> once I mixed a similar adhesive in an uneven manner and it never cured properly... so it must be important to have an even distribution hehe
[03:06:48] <GuShH> actually when I do it by eye I'm always off.
[03:07:01] <GuShH> that's why they make the double syringe type
[03:07:35] <GuShH> if only it mixed it for you as well... (patent pending?)
[03:10:12] <theBear> yeah, epoxy is much more sensitive to slight mix errors than a lot of things
[03:12:43] <xorm> i find that i really need to use the mixer nozzles
[03:12:57] <xorm> the ones with the gigantic corkscrew things inside the tube that mix for like 6 inches
[03:13:02] <theBear> nah, professional epoxy products mix, you get double-tube glue-guns, you know, like you use for silicone or building adhesives, and a special double mixer nozzle that uses some standard vortex or some such theory, you can see it happen at the start of use (there is a black and a white ingredient in the 2 tubes, but after a few moments it all looks grey inside the mixer) ... for reference always get extra mixer-nozzles if you plan on not using it all in on
[03:13:03] <theBear> e go, that stuff sets ROCK HARD, as in stronger in at least several regards/possible measurements of strength than almost any material i could name
[03:13:04] <xorm> it's hard to get them to blend properly
[03:13:23] <theBear> pfft, you mighta beat me to enter, but my one was better :)
[03:13:26] <xorm> yes
[03:34:41] <twnqx> hm
[03:35:08] <twnqx> will a 15V zener diode die if asked to drop ~700V, but within the total power dissipation?
[03:37:59] <Fleck> suggest a good asm tutorial/book for avr
[03:39:59] <theBear> zeners don't drop voltage, they clamp it, you gotta worry about whatever you use for the shunt part of "zener shunt reg"
[03:40:28] <theBear> i mean, they CAN drop a voltage equal to their rating, but that's not what you're asking
[03:40:31] * theBear is psychic :)
[03:41:48] <twnqx> ok. let's assume i have a DC voltage of 380V, and an AC voltage of 230VRMS (~ +-325V). the 0 point for the 380VDC is derived froma diode recifier connected to the AC
[03:42:47] <twnqx> now add a 1M resistor to the 380VDC, with the Z diode on the other end. the other terminal of the z diode is connected to the AC
[03:43:20] <twnqx> which should give a waveform of AC+15V at the point between resistor and Z diode if i am correct
[03:44:01] <theBear> the what why ?
[03:44:13] <theBear> and you mean single diode or bridge ?
[03:44:22] <twnqx> bridge
[03:44:48] <twnqx> and i need a voltage of ac +15V to drive mosfets
[03:45:21] <twnqx> but i can't just use the 380V since Vgs shouldn't exceed +-30V
[03:46:50] <theBear> huh ? that doesn't sound right, plus you're doing it wrong, whatever you are doing
[03:46:58] <twnqx> :/ why?
[03:47:56] <theBear> depends what yer doing, but there's no good reason to need to generate a dc voltage floating on top of mains ac pretty much ever.... but i CAN say that the zener should be cool, lemme just math it for a moment
[03:48:27] <Fleck> twnqx: isn't 5V standby for PC PSU doing that?
[03:48:31] <twnqx> PFC does it all the time
[03:48:53] <twnqx> AC => 380V DC => flyback to whatever voltages you want
[03:49:19] <theBear> well, still not enough info to math, what does the 0v ref of the 380vdc look like in relationship to the ac neutral ? or i might say, what is the greatest voltage the zener+resistor combo will ever have at any given moment
[03:49:35] <twnqx> my expectation is 705V
[03:49:45] <twnqx> but could be ummm
[03:49:52] <twnqx> 260*sqrt(2)+400
[03:50:00] <theBear> twnqx, active pfc chops the ac, not dc
[03:50:11] <twnqx> eh?
[03:50:22] <theBear> lets go with 700 for ballpark
[03:50:29] <twnqx> all active pfc circuits i have seen ever rectify the ac, then step it up to 380-400V
[03:50:32] <twnqx> primary side
[03:51:11] <theBear> hmmmm, i suppose you halve the circuitry if you turn all cycles into only positive cycles
[03:51:18] <twnqx> the problem is
[03:51:26] <twnqx> i want to replace the diode rectifier with active mosfets
[03:51:32] <twnqx> to reduce the diode losses.
[03:51:50] <theBear> hmm, you'd want at least a 1watt zener to be safe in that example
[03:52:09] <twnqx> yeah, i ended up with 0.476 with my own calculations
[03:52:19] <twnqx> (690V)
[03:52:49] <theBear> yeah, but it'd not wise to put a .5watt zener when you expect .476, specially in a mains circuit
[03:52:56] <twnqx> indeed
[03:53:08] <theBear> lookup synchronous rectifier i think they call it
[03:53:26] <twnqx> didn't find any primary side ones
[03:53:33] <theBear> unless i'm going crazy again i'm sure you can make a 'mosfet bridge rectifier' without any control circuit
[03:53:40] <theBear> the side is irrelevant
[03:54:07] <twnqx> the voltage IS a problem
[03:54:40] <twnqx> for n-channel to conduct you need to be above source voltage
[03:54:50] <theBear> either way, by the time you gonna have mosfets connected to the mains, it'll all be simpler and use less fets if you just chop/active pfc the ac direct, only reason to pre-rectify it would be so you only need one chopper instead of a pair, and you're gonna have at least 2 big enough to chop everything anyway
[03:55:24] <twnqx> welcome back to diode losses
[03:56:02] <twnqx> also, doubling the circuit is only an option for the upper diodes, bot for the bottom ones
[03:56:36] <twnqx> and then you run into nice problems with detecting 0 current detection through your boost inductor
[03:56:52] <twnqx> since you now have at least 2 (4 in case of certain topolgies)
[03:57:05] <theBear> what diodes ?
[03:57:09] <theBear> i didn't say any diodes
[03:57:28] <twnqx> you can replace the top diodes from your classic bridge with mosfets
[03:57:34] <twnqx> you can't leave out the bottom ones
[03:57:39] <theBear> i'm not saying to use a bridge/rectifier at all
[03:57:49] <twnqx> you never know the polarity
[03:57:52] <theBear> i said chop the mains, both sides of the cycle
[03:58:10] <theBear> and like any mains switcher, you have a VERY tiny trivial supply to bootstrap the control circuit
[03:58:20] <twnqx> yeah
[03:58:23] <twnqx> that's irrelevant
[03:58:30] <theBear> no it isn't, you're doing it wrong !
[03:58:36] <twnqx> what
[03:58:44] <twnqx> bootstrap is irrelevant performancewise
[03:59:01] <theBear> you don't need to rectify anything
[03:59:13] <twnqx> have an example for that?
[03:59:15] <theBear> in fact i think you could avoid any diodes
[03:59:20] <theBear> nup, i just invented it
[03:59:27] <twnqx> i read optimiced circuits for the past two weeks
[03:59:36] <twnqx> optimised*
[03:59:49] <twnqx> noone managed to skip on the bottom diodes
[04:00:34] <theBear> correction, avoid any rectification, use active fets on raw ac live, one each half of the cycle, 2 inductors, directly feeding err, i guess around 750vdc main resovoir cap
[04:00:38] <theBear> now THAT is optimised
[04:00:43] <twnqx> yes
[04:00:49] <twnqx> that's a standard way
[04:00:53] <twnqx> you still have a return
[04:01:01] <twnqx> and you need diodes there
[04:01:04] <theBear> and if you don't use n fets, you don't need anything higher than ac
[04:01:14] <twnqx> wait
[04:01:15] <theBear> mmmm, why ?
[04:01:18] <twnqx> 750V?
[04:01:24] <theBear> that's what i said
[04:01:30] <theBear> give or take a few v
[04:01:46] <twnqx> uh, because L is +-325V around N ?
[04:01:51] <theBear> you end up with higher voltage resovoir but you only need 2 big fets and 2 inductors
[04:02:07] <twnqx> and you can't predict which wire is L and which is N?
[04:02:08] <theBear> yes it is, plus AND minus
[04:02:35] <theBear> you can in this country, some countries have stupid and dangerous lack of rules re: that kind of thing tho
[04:02:45] <twnqx> yes. like all of europe.
[04:02:48] <theBear> tho it doesn't really matter
[04:02:53] <twnqx> 180° rotatable connectors
[04:03:01] <theBear> this is all floating and ac, the circuit won't care
[04:03:22] <twnqx> so what do you reference yout 750V to
[04:03:24] <theBear> stupid europe, wait, all of eu doesn't have grounds ?
[04:03:32] <twnqx> of course
[04:03:36] <twnqx> line, return, ground
[04:03:38] <twnqx> 3pin
[04:03:39] <theBear> twnqx, to nothing but itself
[04:03:44] <twnqx> aha
[04:03:48] <theBear> and you can still reverse the mains plug ?
[04:03:53] <twnqx> yes
[04:03:57] <theBear> so you got 3pin plugs and 4pin sockets ?
[04:04:02] <twnqx> dual ground on top and bottom
[04:04:08] <twnqx> right and left are mains and return
[04:04:13] <theBear> that's the stupidest fucking thing i've heard this week
[04:04:37] <theBear> but i'm sure this will still work, from the circuit p.o.v the 2 ac lines are still just 2 ac lines
[04:04:55] <twnqx> https://www.google.de/search?q=schuko+stecker&safe=off&client=firefox-a&hl=en&rls=org.mozilla:en:official&tbm=isch&tbo=u&source=univ&sa=X&ei=4YacUbWOC4mJtAbZpoHYBg&ved=0CEEQsAQ&biw=989&bih=617
[04:05:06] <theBear> i believe yer, it's just stupid
[04:05:46] <theBear> most likely tho it only came about 'cos they had to deal with old mainly us designed/built stuff that had non polarised AND non grounded plugs originally
[04:07:01] <theBear> i do like having a strict electrical code in some ways, like the capability to be SURE which is gonna be live or neutral all the way from the wall socket to wherever the mains gets used or psu'd to not mains
[04:07:21] <twnqx> and then someone for one reason or another switched it, and you produce smoke?
[04:07:24] <twnqx> yeaaaaah
[04:07:45] <theBear> well, there's still the odd stupid or non electrician here and there, but i can't think of anything that won't work either way
[04:08:12] <theBear> including the circuit i just designed for you ! the awesome diodeless active pfc circuit !
[04:08:27] <twnqx> well, i would like to see how your one-wire fed circuit works, since i have no clue how you connect the return to it
[04:08:46] <twnqx> and how you get 700V from.. umm
[04:08:51] <twnqx> one wire and itself
[04:09:00] <theBear> hmmm, without drawing anything, i'm gonna say 2 caps either side of the 2nd ac line
[04:09:08] <theBear> 1 either side of
[04:09:32] <theBear> each active pfc fed with a single chopper off either side of the wave
[04:09:39] <theBear> heh, i'm brilliant !
[04:09:52] <twnqx> sigh
[04:10:25] <theBear> i said heh on the same line for a reason, but it's a solid concept, if yer really wanna get away from losses and diodes, i think it's by far the most sensible
[04:10:42] <theBear> and by think, i mean cannot see any reason that any other approach makes more sense
[04:11:45] <twnqx> until you put it into a schematic
[04:11:51] <twnqx> really.
[04:12:58] <theBear> seriously, you aren't following ? my final version description was pretty complete there
[04:13:54] <twnqx> all i can see is that i can't figure out the current flow in that
[04:14:00] <theBear> i don't think i got a schematic editor installed anywhere right now
[04:14:24] <twnqx> http://img189.imageshack.us/img189/6111/br1y.png this is around the best so far
[04:15:16] <theBear> ok, think two 400v or bigger resovoir caps in series, the join between them is one ac line, for sanity lets just call it neutral, altho being ac it is irrelevant.... the other line is chopped alternately by two different fet+inductor pairs, each of which feeds one of the two caps
[04:16:53] <theBear> that's actually not far off the concept of what i just said, except needs the diodes on the output of the two choppers, and the two choppers fed off opposite ac lines, because they don't want two resovoir caps and 700v to deal with
[04:17:05] <twnqx> the diodes are the boost diodes
[04:17:08] <twnqx> can't cut them off
[04:18:57] <twnqx> yeah well, there's no control circuitry atm that would survice 700V
[04:18:57] <twnqx> also not sure how much more flyback transformer losses you'd have
[04:18:57] <twnqx> going from 700+ to 12V :P
[04:18:57] <theBear> ahh, it's true, ok, we'll keep those 2, then feed the two inductor/choppers off one line of ac, only the 2nd fet/diode are reversed, each booster alternately feeds a "plus" and a "minus" side resovoir which are both centered around the other side of ac
[04:18:57] <theBear> transformers don't care about ratio
[04:19:13] <twnqx> perfect transformers don't
[04:20:51] <theBear> and that's a good point about both control and secondary (post pfc/resovoir) chopper ratings, but if you really wanted to, you could do a symmetrical chopper off the symmetrical filter caps, of course then you'd be adding more components cancelling the savings on the pfc side... but i still like the idea, i'll have to draw it... ooh, i got a circuitlab acct too, i thought i'd drawn some stuff somewhere recently
[04:21:37] <theBear> simulators are even good for getting things like 1000v power fets that may or may not exist :)
[04:22:08] <twnqx> 900V do, at least
[04:22:15] <theBear> hmmm, 1000v i say ? remind me, are igbt's like a base with a fet switching side, or like a gate with a transistor switching side ?
[04:23:12] <twnqx> the point of basing both mosfets against the rectified returns
[04:23:26] <twnqx> is that you can have a stable voltage against the source
[04:23:36] <twnqx> and source always tied to 0V, so it's easy to control
[04:23:53] <twnqx> (gate voltage i mean=
[04:24:44] <twnqx> i am not sure how you can build what you mean without assumptions to polarity (you will invert them on a switched AC)
[04:24:58] <twnqx> and having problems with controling the gate voltage
[04:25:36] <twnqx> i'm curious to learn, though :P
[04:26:01] <theBear> no i won't, it's ac.. i'll have to think about driving them, but i'm sure its possible without any dc floating on the ac wave
[04:26:28] <theBear> i like to keep things neat and simple, and dc floating on mains just feels horribly wrong
[04:26:54] <Tom_itx> keeps you on your toes
[04:27:01] <twnqx> floating.
[04:27:14] <theBear> oooh, way back to your original question, you seen those little new dirtcheap ics that both isolate and handle their own control/bias voltage generation/regulation for driving fets ?
[04:29:20] <twnqx> nope
[04:31:15] <theBear> times like this, it'd be handy if i remembered what they were called or who made them <grin> but a little searching will likely find them
[04:31:29] <theBear> pretty sure it's someone like ti or ad that makes them
[04:31:52] <twnqx> if you could guarantee polarity and use one p-channel and one n-channel fet
[04:32:26] <twnqx> you could base the gate voltages against the capacitor center taps
[04:32:35] <theBear> it's ac, there is no polarity ! your control circuit just needs to be aware of this and watch the ac for polarity at any moment/half cycle
[04:32:58] <theBear> it's already gonna be a fairly 'intelligent' and active control circuit
[04:33:09] <twnqx> right... there would just be a 180° phase shift
[04:33:13] <theBear> not hard to monitor the ac
[04:33:23] <twnqx> hm
[04:33:54] <twnqx> in fact
[04:34:09] <twnqx> having stable +-400V
[04:34:15] <twnqx> with a center tap
[04:34:30] <twnqx> could back for an improved transforer drive with no DC component...
[04:34:33] <twnqx> make*
[04:34:50] <theBear> haha ! now yer coming around :)
[04:39:25] <twnqx> wagh.
[04:39:44] <twnqx> i'll take a deeper look after lunch, i'll have to draw that
[05:23:01] <Fleck> suggest a good asm tutorial/book for avr please?
[05:47:09] <RikusW> he instruction set reference ?
[05:47:44] <RikusW> http://www.atmel.com/Images/doc0856.pdf
[05:48:08] <RikusW> https://sites.google.com/site/megau2s/home/datasheets
[05:48:35] <twnqx> theBear: hm, there's a lack of p-channel power mosfets... in fact, only one vendor :S
[05:49:25] <theBear> huh ? that doesn't sound right... tho i suppose i never looked in that kinda voltage/power range
[05:49:36] <theBear> and p are much less common in general in mosfetland
[05:49:40] <twnqx> indeed
[05:49:46] <twnqx> there are TONS of n-channel
[05:50:43] <theBear> Fleck, if you haven't done any asm before, reading any asm beginner guide is important, and the basic concepts are the same for any asm/chip/arch, just like the concepts of variables and conditional tests and loops is the same for any programming language
[05:52:48] <Fleck> I think I need beginners guide... no clue what __vectors: mean, is it a function or what? :D
[05:53:06] <Fleck> how do they execure... just line by line from top?
[05:53:11] <RikusW> yes
[05:53:22] <RikusW> interrupt vectors
[05:53:24] <Fleck> then there is __vectors_4:
[05:53:39] <Fleck> *__vector_4:
[06:16:18] <Fleck> RikusW, theBear are you good at ASM? :D
[06:17:41] <theBear> i dunno i'd say good... i'm definately outta practice, and not incapable, i understand how to do things nice/neatly, but i haven't exactly put a large % of my overall programming experience into practicing it
[06:18:24] <theBear> and done VERY little in avrland, so don't be quizzing me on opcode names and things
[06:18:32] <Fleck> :D
[06:18:58] <Fleck> out CLKPSR, r16 << this stores CLKPSR address into r16?
[06:19:31] <theBear> mmm, out would be a opcode <grin> it definately does that or the reverse of that :)
[06:20:11] <Fleck> Stores data from register Rr in the Register File to I/O Space (Ports, Timers, Configuration Registers etc.)
[06:20:24] <Fleck> RikusW`s link helped there
[06:20:31] <theBear> so opposite of what you said
[06:20:41] <Fleck> yeah
[06:20:46] <RikusW> Fleck: I've done quite a bit of AVR asm coding
[06:20:59] <RikusW> > 6000 lines :)
[06:21:08] <theBear> i've done >6 :)
[06:21:09] <Fleck> nice ;)
[06:21:47] <Fleck> few years ago I played with video mem and banks etc...
[06:21:56] <RikusW> and reverse engineered > 10k lines of asm...
[06:22:06] <Fleck> uhh
[06:22:24] <Horologium> http://www.avr-asm-tutorial.net/avr_en/ avr assembly tutorial for you.
[06:22:33] <RikusW> Fleck: on the jtag tab on my site you'll find the reverse engineered jtagice mki fw
[06:22:39] <RikusW> and a clone in asm
[06:22:44] <Horologium> it's old but still good.
[06:23:10] <RikusW> you'll need AVRStudio to assemble though
[06:23:13] <Horologium> many people these days just use C for AVR
[06:23:19] <Horologium> RikusW, not with the assembler on that link.
[06:23:24] <Fleck> thx Horologium
[06:23:34] <Horologium> it just doesn't support the most current AVRs though.
[06:23:41] <RikusW> I used the AS4 asm for my coding
[06:24:05] <RikusW> I also have a partly working assembler I coded
[06:24:21] <RikusW> can take instructions and assemble them into hex values
[06:24:26] <RikusW> and disassemble too
[06:24:43] <RikusW> don't have symbolic support yet...
[06:25:07] <Horologium> aahh disassemblers...gotta love em.
[06:25:15] <RikusW> I make my own :)
[06:25:19] <Horologium> wrote one many moons ago for the commodore vic-20 and the 6502.
[06:25:25] <RikusW> I did one for x86 too many years ago
[06:25:30] <RikusW> and PIC
[06:25:33] <RikusW> and 8051
[06:25:34] <Horologium> then I got the vicmon cartridge that did that and more.
[06:26:05] <RikusW> writing an asm/disasm isn't that hard either
[06:26:25] <RikusW> though x86 addressing modes is a nightmare to parse...
[06:26:36] <Horologium> I really need to take the time to go through the vic-20 basic assembly and convert it to avr.
[06:27:00] <Horologium> RikusW, wrote that one for the vic-20 in basic.
[06:27:31] <Horologium> I also wrote a pseudo basic compiler for it too, after I added 16K of ram to the thing.
[06:30:09] <Horologium> it never did work fully though.
[06:30:28] <Horologium> was attempting to write a basic compiler, in basic, that would compile itself.
[06:30:47] <Horologium> I ended up going into the military before I finished it and kind of forgot about it for many years then never went back to it.
[06:32:06] <RikusW> that happens...
[06:32:24] <Fleck> where are you from Horologium?
[06:35:18] <Horologium> I'm from earth
[06:35:24] <Horologium> and still live there.
[06:35:39] <Horologium> unfortunately
[06:37:35] <Horologium> 42° 10' N Latitude / 93° 5' W Longitude if you must know geospacial coordinates.
[06:38:38] <theBear> that's not geospacial ! at least give a 3rd dimension :)
[06:38:48] <theBear> i might go there and not be able to find you <grin>
[06:38:54] <Horologium> oops...you are correct...
[06:39:06] <Fleck> :D
[06:39:24] <Horologium> atitude 42.173 and longitude -93.094.
[06:39:24] <Horologium> The elevation is 1,010 feet.
[06:39:27] <Horologium> better?
[06:39:30] <Fleck> and tell the country name was just... too simple, right? :D
[06:39:41] <Horologium> of course.
[06:39:44] <theBear> cheers :) i'll come visit one day
[06:40:00] <Horologium> any time.
[06:40:10] <Horologium> we have a spare bedroom and spare tent
[06:40:31] <theBear> awesome ! i'll take the tent, camping is fun :)
[06:40:36] <Horologium> and I think there is still some room in the chicken coop but you might have some duckies snuggle with you in there.
[06:41:04] <Fleck> :D
[06:41:22] <Horologium> Fleck, Bangor, Iowa, USA, Earth
[06:42:12] <Fleck> Horologium: yeah, I know how to use google maps, thx :D
[06:42:25] <theBear> hehe, there's a bit in the cleveland show, cleveland has a cup of change, walks past a homeless dude who asks him if he can spare some change, he politely says no and apologizes, then starts tipping them into a change -> note machine in the wall next to the dude, one rolls in front of the dude "little help here please ?" so the guy passes it back to him, a nice fresh note comes out of the machine and he says to himself "hooray ! now i can buy a tent and s
[06:42:26] <theBear> leep outdoors for fun !" hehehe, it makes me laugh
[06:42:37] <Horologium> Bangor has a population of approximately 20 humans, 4 dogs, 45 chickens, 3 rabbits, and 9 ducks.
[06:42:46] <Fleck> :D
[06:43:04] <Horologium> theBear, hehe
[06:43:08] <twnqx> how many roosters?
[06:43:19] <Horologium> twnqx, 1, but kinda counted him in with the chickens.
[06:43:30] <twnqx> place is dead to me.
[06:44:03] <Horologium> we might have 2 but we aren't sure of the sex of the baby chicken hatched out last week named Nugget.
[06:44:05] <Fleck> chicken-rooster? :)
[06:44:12] <theBear> ahh, bangor maine... i heard of that, IN EVERY STORY STEVEN KING EVER WROTE ! seriously, that dude needs to be just a LITTLE more creative in that regard
[06:44:18] <Horologium> rooster is a male chicken
[06:44:22] <Horologium> hen is a female chicken.
[06:44:27] <RikusW> chicken-roaster ? :-P
[06:44:31] <Horologium> theBear, he likes that area.
[06:44:39] <Horologium> RikusW, no roasters..just layers..
[06:44:46] <Fleck> ;p
[06:44:51] <Horologium> we don't have the gumption to butcher them and it's expensive.
[06:45:20] <Horologium> although the neighbor wants to butcher a couple of the ducks for christmas dinner.
[06:45:23] <RikusW> I can help with that ;)
[06:45:35] <RikusW> I've actually done it
[06:45:39] <Horologium> I have too.
[06:45:44] <Horologium> and it is a horrid mess.
[06:45:45] <theBear> Horologium, i like his books, specially when i was younger, but seriously, EVERY SINGLE ONE !
[06:45:59] <Horologium> theBear, isn't he from there?
[06:46:12] <theBear> i dunno, somewhere in that general area
[06:46:24] <Horologium> there's always the Xanth books....Xanth being an alternate universe Florida.
[06:46:34] <Horologium> 20 or 30 of them in fact.
[06:46:34] <RikusW> Horologium: lately I've been lazy when it comes to butchering stuff....
[06:46:48] <Horologium> RikusW, yeah. I haven't done it for about 15 years myself.
[06:46:52] <theBear> but you know, he's creative enough to write about aliens and ufos and all kinds of crap, but he can't manage a different town
[06:46:54] <Horologium> last one was a deer someone gave me.
[06:47:12] <RikusW> btw speaking of ducks have you seen the "true facts" series on youtube ? :-D
[06:47:19] <Horologium> nope.
[06:47:33] <theBear> nup
[06:47:48] <RikusW> you won't look at ducks quite the same again :-P
[06:49:06] <theBear> heh, you don't know how i look at them now :)
[06:50:50] <RikusW> https://www.youtube.com/watch?v=6k01DIVDJlY
[06:50:55] <RikusW> that one
[06:51:00] <Horologium> http://buttersafe.com/2009/05/26/apple/
[06:56:24] <theBear> i'll never look at that researcher lady the same way again
[06:56:33] <RikusW> heh
[06:57:28] <theBear> hehe, i like the morgan freeman one, and he's got the perfect voice for it
[07:00:56] <tzanger> what researcher lady? I watched the entire thing, saw way more duck penis than I was comfortable seeing and all I saw that was female (human) was a pair of hands around a duckling
[07:02:18] <theBear> didn't you listen at the same time ?
[07:02:26] <tzanger> no. had audio muted
[07:02:51] <theBear> she's spent most of her professional life checking what a duck penis can force its way through
[07:43:37] <theBear> WHAT THE ! ZE KNOWS DWIGHT !
[08:01:14] <beaky> hello
[08:01:47] <beaky> I neglected to connect avcc, aref, reset, and the second GND (the first is connected), and my MCU still works! why is that?
[08:02:27] <theBear> probably mostly the protection clamp diodes on all the io pins
[08:03:00] <beaky> ah
[08:04:28] <beaky> I also found out why my mcu was doing things 2 times faster than how I specified it in my code: my professor clocked it to 8 MHz, but my build settings for my program had #define F_CPU (4000000lu)
[08:04:57] <beaky> as always there's always a perfectly logical reason why things go wrong in microcontrolleristan :D
[08:05:19] <theBear> hmmm, was it you with the 3 dead avrs ? did that sort itself out logically ?
[08:05:22] <jadew> beaky, avcc and aref are used for analog stuff (reading the DAC, comparators, things like that)
[08:05:36] <jadew> so if you don't use them, it wouldn't affect things too much
[08:05:48] <beaky> theBear: yeah; my professor just stuck them into a jtag and clocked them from 1MHz to 8MHz and now my usb programmer thing worked on them :D
[08:05:55] <jadew> the second gnd is internally connected to the other gnd, but it's there so it minimizes the ground loop
[08:06:12] <jadew> the RESET pin is internally pulled up, so it would work even if it's not connected
[08:06:25] <beaky> ah
[08:06:45] <jadew> however, the pullup on reset is quite weak, so it's a good idea to have it externally pulled up
[08:07:18] <beaky> right; I guess I should stick a 10k ohm -> Vcc on the reset pin
[08:07:59] <jadew> yeah, and connecting avcc and all that, even if you're not using them, is still a good idea
[08:08:14] <beaky> so avcc goes to vcc, and aref goes to gnd?
[08:08:34] <jadew> aref goes to whatever your voltage reference is (for ADC input)
[08:08:47] <beaky> ah
[08:08:48] <jadew> however it can't be above avcc
[08:08:57] <jadew> and avcc can't be above vcc IIRC
[08:09:06] <jadew> you'll have to check the datasheet
[08:09:16] <jadew> either way, most of the time, avcc will get connected to vcc
[08:09:28] <beaky> Vcc is awesome
[08:09:44] <beaky> it's what every electrical component runs on
[08:10:06] <beaky> electronic*
[08:10:45] <beaky> how did the industry standardize on Vcc so that everything is consistent?
[08:10:58] <jadew> what do you mean?
[08:11:43] <jadew> there's nothing standard about it, it can have any range of voltages
[08:11:54] <twnqx> vcc is just a name
[08:11:57] <beaky> ah; I guess Vcc is just some old naming convention
[08:12:06] <jadew> it comes from cmos stuff IIRC
[08:12:12] <beaky> which stands for (Voltage of the collector collector)
[08:12:42] <twnqx> fun fact: mosfets don't have a collector ;)
[08:12:43] <theBear> beaky, huh, that doesn't make sense, surely you don't have a programmer so shit it can't program a brand new chip ?
[08:13:22] <beaky> yeah I guess the usb programmer sucks; I have no idea why it needs the new chips to be configured by a JTAG
[08:13:23] <twnqx> theBear: oh, you can have fun with those crappy avrs with bootloader
[08:13:39] <twnqx> cannot program them without external clock, since the fuses default to them
[08:13:47] <beaky> ah
[08:13:54] <beaky> so my avrs suck? :(
[08:13:58] <theBear> not bootloader, PROGRAMMER
[08:14:06] <theBear> ISP
[08:14:12] <twnqx> theBear: you *cannot* program the ISP without connected clock
[08:14:15] <theBear> no, don't listen to him
[08:14:35] <twnqx> a friend of mine had to trash >100 fabbed PCBs because of that
[08:14:50] <jadew> sure you can, if they're using the internal osc
[08:14:57] <twnqx> like i just said
[08:14:58] <theBear> twnqx, i know what i can and can't do, but this guy bought a programmer with a whole chip builtin to drive it, and it won't program a new chip
[08:15:06] <twnqx> they are shipped with the fuses set to external clock
[08:15:23] <twnqx> so cou cannot even fix the fuses without external clock
[08:15:36] <theBear> if i'm gonna buy a programmer, i kinda expect it to be able to program a chip i buy at the store
[08:15:41] <twnqx> was a fun discovery for him that cost him upward of a grand
[08:15:46] <theBear> that's the whole idea of programmers
[08:15:49] <twnqx> no
[08:16:00] <twnqx> ISP programmers don't have an OSC pin
[08:16:21] <twnqx> it's not even meant to be at the ISP connector, just the SPI-like clock is there
[08:16:39] <twnqx> jtag programmers work
[08:17:02] <twnqx> no idea about the other protocols, but ISP does not work withut a clocked chip
[08:17:40] <edmont> hi
[08:18:02] <theBear> if the programmer has a zif socket on it, it should be able to program a new chip, and if it can't, IT IS STUPID
[08:18:34] <twnqx> if there's a zif socket
[08:18:45] <twnqx> there's a chance beaky didn't set up clocking correctly.
[08:18:59] <jadew> I agree with theBear, especially one that came embeeded with a chip (which means it has a socket for new chips or a target board)
[08:19:02] <beaky> I am glad that it turned out that my USB programmer sucks rather than my chips being dead :D
[08:19:05] <edmont> i'm using a modified version of arduino bootloader for atmega128rfa1
[08:19:18] <beaky> I guess I can trust the local electronics dude
[08:19:38] <edmont> at the beggining it checks if a wdt reboot was performed:
[08:19:43] <edmont> if (! (ch & _BV(EXTRF))) // if its a not an external reset...
[08:19:43] <edmont> app_start(); // skip bootloader
[08:20:26] <edmont> the thing is that when condition matches my mcu starts rebooting all the time
[08:20:28] <twnqx> jadew: didn't see that part about zif sockets
[08:20:38] <twnqx> but it can be problematic anyway
[08:20:40] <edmont> like it doesnt reach the application code
[08:20:47] <edmont> any idea?
[08:22:27] <edmont> void (*app_start)(void) = 0x0000;
[08:24:49] <RikusW> beaky: I might have an idea, your programmer's ISP clock is 500kHz or above, thats why it can't program the 1MHz mega16's
[08:25:18] <RikusW> fyi avr clock needs to be 4 (preferably 5) times higher than ISP clock for programming to work
[08:25:27] <RikusW> same goes for JTAG or SPI slave
[08:25:49] <RikusW> so your professors programmer used a lower clock for programming
[08:26:16] <RikusW> beaky: you might be able to change your programmer's clock though
[08:26:26] <RikusW> lower it to make it work with new AVRs
[08:28:04] <RikusW> twnqx: AVR jtag needs the target chip to be clocked too...
[08:33:24] <Badaboom> morning
[08:35:28] <beaky> ah thanks
[08:36:38] <theBear> hangon tho, his programmer can program the avr now, and nothing has changed in the programmer, therefore IF you need a clock to ISP program, the programmer must supply one
[08:36:55] <theBear> oh, you beat me to it
[08:38:10] <thetruthisoutthe> h
[08:39:24] <theBear> w
[08:39:29] <theBear> :)
[08:42:57] <thetruthisoutthe> λ
[08:44:13] <beaky> what does the standard call this thing? []() -> foo {}
[08:44:27] <beaky> is it a lambda function, an anonymous function, a lambda, a closure?
[08:44:42] <beaky> or is it anonymous :D
[08:44:54] <beaky> oops wrong channel
[08:45:14] <Badaboom> I need wake up juice,,brb
[08:48:52] <theBear> erm, random punctuation
[08:53:10] <RikusW> https://www.youtube.com/watch?v=ritaljhhk7s
[08:53:32] <RikusW> no ducks involved :-P
[08:53:59] <theBear> then i'm not interested !
[08:54:06] <theBear> wait, any hedgehogs ?
[08:54:16] <RikusW> only weird chemistry
[08:54:34] <theBear> pfft
[08:54:39] <RikusW> very weird
[08:54:54] <theBear> nah, i already got it qeued queued ? i Q-ed it
[08:55:25] <RikusW> more interested in zefrank's vids ? ;)
[08:55:34] <theBear> i may have watched a couple
[08:55:47] <theBear> but i just snacked, so a real movie is on
[08:56:26] <Badaboom> its alive
[08:56:34] <RikusW> seems so :)
[08:56:47] <Fleck> how do you compile asm file in linux for avr?
[08:56:48] <theBear> it is ?
[08:56:50] * theBear looks down
[08:57:20] <RikusW> Fleck: there is avrasm or similar
[08:57:32] <RikusW> look in your package manager for avr
[08:57:32] <Badaboom> bear,, thats just wrong
[08:57:45] <theBear> i'm not the one that claimed it was alive !
[08:57:57] <theBear> but you're right, it doesn't appear to be moving
[08:57:58] <Badaboom> sigh
[08:58:04] <theBear> hehe, sorry :)
[08:58:09] <Badaboom> lol
[09:01:20] <twnqx> RikusW: avr-gas :P
[09:01:34] <RikusW> I still prefer the AS one
[09:02:39] <RikusW> And I'm running AS inside Linux now :)
[09:03:02] <RikusW> well not directly... Inside XP inside a VirtualBox on Mint..
[09:03:17] <RikusW> even connects to the dragon just fine
[09:03:36] <RikusW> had to install the virtualbox extension pack to make usb work
[09:03:44] <theBear> mmm, vbox and vmware seem pretty darned good at talking to usb stuff these days
[09:03:57] <beaky> can you install linux on an avr?
[09:04:06] <RikusW> I got a win7 native install to
[09:04:06] <theBear> not a 8bit one
[09:04:07] <beaky> I wanna boot ubuntu on it
[09:04:09] <beaky> ah
[09:04:14] <RikusW> guess I won't be using it much,
[09:04:43] <theBear> not a 8bit anything that i can recall, maybe some 16bit things, well uclinux
[09:04:44] <megal0maniac_afk> Ubuntu on an m328 :)
[09:05:05] <RikusW> so far I like mate or cinnamon better than aero....
[09:05:15] <megal0maniac> It's okay..
[09:05:29] <RikusW> megal0maniac: paid for my MT order a little while ago, should be posted today :)
[09:05:46] <megal0maniac> Cool :)
[09:06:05] <megal0maniac> I'm trying to get my Logic to work on the Dockstar
[09:06:11] <RikusW> seems they are very busy these days...
[09:06:28] <jadew> mt? logic? dockstar?
[09:06:37] <theBear> i know the last one
[09:07:08] <theBear> it's a hd enclosure with way more cpu/ram than could possibly be needed
[09:07:09] <megal0maniac> jadew: Mantech(.co.za) Seagate Dockstar and Saleae Logic
[09:07:15] <RikusW> mantech.co.za
[09:07:19] <theBear> runs linux good
[09:07:25] <jadew> got it
[09:07:28] <megal0maniac> theBear: That it does :)
[09:07:44] <RikusW> Its been quite a while since I ordered from mt
[09:07:56] <RikusW> even their bank account changed...
[09:07:57] <megal0maniac> 1.2ghz arm5 with 128mb RAM, 256mb flash, 4x USB and 1x gigabit LAN
[09:07:58] <jadew> mantech.co.za => server not found
[09:08:04] <megal0maniac> jadew: www.
[09:08:16] <RikusW> http://www.mantech.co.za
[09:08:18] <jadew> who uses www. anymore...
[09:08:23] <jadew> fucking stupid subdomain
[09:08:24] <megal0maniac> Mantech :)
[09:08:38] <RikusW> megal0maniac: they moved from ABSA to STD
[09:08:43] <jadew> ah, so they have stuff
[09:08:52] <megal0maniac> Typically there's a CNAME record which link the two
[09:09:08] <megal0maniac> jadew: Yeah. Horribly marked up most of the time
[09:09:27] <RikusW> jadew: they're partnered with mobicon and farnell
[09:09:40] <jadew> I see
[09:09:46] <jadew> aren't they expensive?
[09:10:01] <RikusW> many things are cheaper at them than RS
[09:10:08] <jadew> whenever I think "farnell" I think overpriced
[09:10:18] <RikusW> however RS have better pricing and variety of AVRs
[09:10:23] <megal0maniac> The alternative is importing, which is worse. Digikey charges $60 flat rate for shipping
[09:10:33] <megal0maniac> And then there's still customs
[09:10:40] <theBear> they're fine for regular everyday components, sometimes good, but anything unusual or tool or god forbid test equipment or devboards they aint so special
[09:10:40] <RikusW> and that....
[09:10:51] <twnqx> megal0maniac: regardless how much you order?
[09:11:00] <twnqx> here it's free shipping from 65€ order value
[09:11:05] <megal0maniac> twnqx: Might go up, but definitely doesn't come down
[09:11:10] <RikusW> they sell reels for around $16
[09:11:24] <Fleck> RikusW: avra, but that fails to compile
[09:11:28] <theBear> for several years until very recently all web orders from farnells were free delivery, regardless of value
[09:11:44] <theBear> in this country that is
[09:11:44] <RikusW> Fleck: my code was made for the AS4 (or 5/6) assembler
[09:11:52] <twnqx> i guess it depends on the country
[09:12:13] <RikusW> new toys arrived :)
[09:12:25] <twnqx> that 65€ limit is the reason why i have tons of electronics i don't really need ;_;
[09:12:35] <Fleck> RikusW: no clue about your code... but ok, i found avr-gcc does this...
[09:12:47] <jadew> haha twnqx
[09:13:11] <twnqx> otherwise it's 18€ shipping!
[09:13:12] <jadew> I always overstock so I wouldn't have to pay for the shipping fee in the future
[09:13:19] <megal0maniac> twnqx: RS used to have free shipping on everything. So I used to order 20x 0805 resistors or something stupid, then add some stuff so I didn't feel so bad
[09:13:26] <jadew> so far, I can't say I did a mistake by overstocking
[09:13:39] <twnqx> well, i now have three of my dual CAN boards
[09:13:41] <twnqx> and need... 1
[09:13:46] <megal0maniac> I have eeprom chips
[09:14:08] <theBear> i could do with some chips.. pref. hot, but salt and vinegar would do
[09:14:26] <twnqx> at least they helped me to stop being scared of tssop.
[09:14:33] <theBear> oooh ! i should buy some frozen crinkley ones tomorrow ! i haven't had chips in the fridge forever
[09:14:46] <twnqx> fozen.. chips
[09:14:55] <thetruthisoutthe> who uses hamming code for data tx/rx ?
[09:15:16] <twnqx> theBear: according to the datasheet cooling mosfets to -40°C helps with RDSon as well!
[09:15:33] <theBear> mmmm, even cooling them a little bit doesn't hurt
[09:15:54] <jadew> hot chips are better
[09:15:55] * twnqx grabs a liquid nitrogen compressor
[09:15:56] <thetruthisoutthe> twnqx <= but they get faster with increasing temperatures :)
[09:16:11] <jadew> actually, I don't think that's true
[09:16:38] <jadew> the more heat, the higher the resistance (in most materials)
[09:16:46] <jadew> so you might get shittier signals
[09:16:49] <thetruthisoutthe> how does that relate to speed?
[09:17:08] <jadew> well, if it's a current based chip
[09:17:11] <jadew> (not voltage)
[09:17:20] <jadew> the time for a cap to get charged will increase
[09:17:25] <megal0maniac> jadew: Is that logic analyser project on the shelf?
[09:17:28] <jadew> therefore => lower speed
[09:17:45] <jadew> megal0maniac, for now, but I'm gonna get back to it soon
[09:18:04] <jadew> right now I have some other personal issues
[09:18:19] <thetruthisoutthe> i was just thinking about an actual use of the mosfet in a power supply, not a theoretical speculation
[09:18:47] <jadew> well, heat never relates to faster anything
[09:18:50] <jadew> heat is noise
[09:19:11] <jadew> so best case scenario: nothing happens
[09:19:12] <thetruthisoutthe> it also lowers the V_G_TH param.
[09:19:19] <jadew> worst case scenario: shit stops working
[09:19:55] <theBear> raises gain on transistors tho
[09:20:06] <jadew> well, that's not always desirable
[09:20:17] <thetruthisoutthe> IGBT
[09:20:33] <theBear> in a given example means a transistor will tend to thermally runaway while a mosfet limits itself, to an extent
[09:20:55] <RikusW> got a Altera MAX II EPM240T100C5N on a breakout for $10
[09:21:13] <thetruthisoutthe> theBear <= mosfet will happily suicide
[09:21:29] <theBear> thetruthisoutthe, if you push them, but my statement stands
[09:21:35] <thetruthisoutthe> if it cannot unsolder itself
[09:21:45] <jadew> lol
[09:21:55] <theBear> i never seen a fet that'll unsolder itself before it dies
[09:22:00] <theBear> not at any kind of power
[09:22:12] <thetruthisoutthe> i have seen some unselder itself, but it does not die
[09:22:21] <theBear> tho it seems if a fan dies that my 50w led module can
[09:22:29] <jadew> they just love freedom, if they can't unsolder themselves, they suicide
[09:22:40] <thetruthisoutthe> :)
[09:22:44] <beaky> If my avr dies, is there a way for my circuit to know when it dies and to do something when it does?
[09:22:49] <theBear> i seen a LOT of fets AND transistors die once they get past the 100-110C kinda zone, that's a long way off melting solder
[09:23:07] <thetruthisoutthe> beaky <= if you arm a wdt, it may restart
[09:23:21] <thetruthisoutthe> if it dies then it will do nothing
[09:23:24] <beaky> ah
[09:23:24] <jadew> well, it depends on how it dies
[09:23:39] <theBear> beaky, depends what your circuit is, but chances are it can't do anything too useful without a ... wait, do we mean die, or freeze/crash/not die
[09:23:42] <jadew> you could constantly output something on a pin
[09:23:49] <jadew> when that stuff stops happening, the chip died
[09:23:52] <beaky> also how do I know if my chip is dead?
[09:23:59] <megal0maniac> theBear: Sent the new mb in, came back "no fault found", despite identical problems to the previous one
[09:24:00] <thetruthisoutthe> beaky <= maybe take 2 interactively(tm) and checking on each other?
[09:24:02] <beaky> e.g. when I feed it 15V
[09:24:03] <jadew> however it could die in the sense of a burnt port
[09:24:26] <megal0maniac> theBear: Should I try and fix it with this? http://www.corsair.com/us/power-supply-units/vs-series-power-supply-units/vs-seriestm-vs650-650-watt-power-supply.html
[09:24:29] <theBear> megal0maniac, yeah, just cos it doesn't do what it should doesn't mean it's faulty :(
[09:24:44] <thetruthisoutthe> beaky <= you use a voltage regulator, and transient suppressor diodes to avoid that
[09:25:05] <megal0maniac> (I have this now: http://www.gigabyte.co.za/products/product-page.aspx?pid=3047)
[09:25:41] <theBear> megal0maniac, it was something about disk errors or something right ? if you really suspect power, run a parallel supply just for the disks, that should confirm it... if the disk controller at the mobo end was screwing up due to power you'd have more noticable problems than disk errors
[09:25:43] <beaky> AH
[09:25:53] <beaky> ah*
[09:27:02] <megal0maniac> theBear: The solid state "disappeared" while in use. It didn't happen until you were actually using the machine. So I'm suspecting some kind of brown-out. I don't have another supply, but I can effectively upgrade from the one to the other for $30
[09:27:45] <megal0maniac> So might be worth it, given that I'm running a hungry graphics card + 3 drives + solid state (3 drives weren't connected during testing though)
[09:27:47] <theBear> megal0maniac, run the disks on a seperate supply, like i say, if the mobo is having power issues you'll know about it from much more serious issues than disk errors
[09:28:22] <theBear> meh, i'm out... it's late and stuff
[09:28:29] <megal0maniac> It's 4pm!
[09:28:31] <megal0maniac> :)
[09:32:15] <megal0maniac> RikusW: What did you still have to get from Manly Technologies?
[09:37:01] <thetruthisoutthe> beaky <= 7805, or a low dropout thing like the LD1117
[09:38:34] <RikusW> megal0maniac: quite a bit of stuff R1900 worth of...
[09:39:00] <megal0maniac> Wow...
[09:39:10] <RikusW> well R100 is shipping
[09:39:38] <RikusW> there is some 5V 2.5A power supplies, 2000mAh li batteries
[09:39:59] <RikusW> and those female headers aren't cheap either, got 160 of them...
[09:40:41] <RikusW> amazing how quickly stuff adds up...
[09:40:49] <RikusW> about half of it is for work
[09:40:59] * RikusW will have to do some accounting....
[09:41:27] <megal0maniac> Haha :) Yeah, I know. I really can't wait until I get some work in the holidays. Funds are low. Student life :P
[09:42:13] <RikusW> megal0maniac: so do you still want the extender boards ?
[09:42:39] <megal0maniac> Probably this holiday
[09:43:10] <megal0maniac> Need to send out at least one invoice first :)
[09:45:56] <megal0maniac> Looks like upgrading my PSU is a waste of time (at least with that one). Only the 12V rails can push more current. The rest are weaker :/
[09:47:43] <jadew> I don't think you need too much power on the 5V one
[09:49:18] <megal0maniac> jadew: I suspect that my solid state is drawing short bursts of high current (which my PSU can't deliver). It "disappears" under load. i.e. same symptoms as if you'd unplugged it. Comes back after power cycle
[09:50:11] <jadew> put a big cap in there :P
[09:50:16] <megal0maniac> Lol
[09:50:25] <jadew> it should work :)
[09:50:35] <megal0maniac> I have a 22000uF somewhere...
[09:50:39] <megal0maniac> 150V I think
[09:50:57] <jadew> well, not that big
[09:51:06] <megal0maniac> That might hurt my PSU :)
[09:51:09] <jadew> yeah
[09:51:10] * twnqx wonders if he should build this circuit IRL
[09:51:16] <megal0maniac> Next biggest is 470uF
[09:51:22] <RikusW> megal0maniac: is it an old or new psu ?
[09:51:36] <twnqx> it's always a scary thought, putting electronic components straight on mains :X
[09:51:43] <megal0maniac> RikusW: New-ish. Worked fine with 3 physical drives
[09:51:57] <megal0maniac> s/physical/spinny
[09:52:06] <RikusW> probably not bad caps then...
[09:52:13] <RikusW> do you have onboard display ?
[09:52:21] <RikusW> try without the graphics card ?
[09:52:31] <megal0maniac> Was considering that
[09:52:34] <twnqx> gfx mostly uses 12V
[09:52:45] <megal0maniac> The mb is still in my bag :) Need to build it again
[09:52:48] <twnqx> at least mine has two extra connectors for 12V :X
[09:53:00] <jadew> same here
[09:53:23] <megal0maniac> twnqx: Mine even has two 12V rails. But it's only 470W
[09:53:31] <twnqx> "only"
[09:53:50] <twnqx> i am running a box with 10 spinning drives + 2SSDs from 400
[09:54:03] <megal0maniac> Okay power is not the problem.
[09:54:37] <twnqx> (i suspect the psu runs over spec during spinup, but that happens once in a blue moon)
[09:55:30] <megal0maniac> This is the graphics... http://www.asus.com/Graphics_Cards/ENGTS250DI1GD3/
[09:55:40] <twnqx> gts250?
[09:55:44] <twnqx> is that even high power?
[09:56:16] <megal0maniac> I don't know. But it's damn noisy. More so when running 2 displays (1920x1080 + 1440x900)
[09:56:34] <megal0maniac> "up to 150W1 additional 6 pin PCIe power required"
[09:56:37] <jadew> twnqx, 400W for the system you say you have, doesn't seem enough
[09:56:48] <twnqx> sure is
[09:56:49] <jadew> the spikes will be more frequent than you think
[09:56:58] <megal0maniac> Drives go to sleep
[09:56:59] <twnqx> runs stable for months
[09:57:11] <twnqx> megal0maniac: but with raidz2 they all come back
[09:57:14] <jadew> megal0maniac, letting drives go to sleep is a bad idea
[09:57:14] <twnqx> and spindown is disabled
[09:57:25] <twnqx> though hm
[09:57:41] <twnqx> i think one of the ssds disappeared two weeks ago, suddenly
[09:58:08] <twnqx> while the spinning disks happily ran
[09:58:35] <jadew> my pc uses about 240W in idle (only spinning disks) and only 5 of them, not 10 as you said you have
[09:58:45] <jadew> however, I have 2 graphics cards
[09:59:01] <megal0maniac> jadew: Where do you measure that? On the AC side?
[09:59:09] <jadew> megal0maniac, yeah
[09:59:15] <megal0maniac> I'm going to do that
[09:59:19] <twnqx> this one doesn't have a gfx card
[09:59:24] <twnqx> and is <100W idle
[09:59:37] <jadew> ah... then you're probably safe with 400W
[10:00:22] <jadew> I got a 700W one, but the output kinda sucks, there's some weird ringing on the USB's VCC
[10:00:35] <jadew> I can't imagine what's on the 12V one
[10:00:50] <twnqx> you don't give it enough load for regulation
[10:00:55] <twnqx> common problem with large PSUs :P
[10:01:05] <jadew> oh, it has plenty of load
[10:01:13] <jadew> 240W sounds decent to me
[10:01:18] <twnqx> 30% in idle
[10:01:21] <jadew> but then again, maybe it's not enough on the 5v, right?
[10:01:23] <twnqx> that's "low"
[10:01:44] <jadew> interresting
[10:01:47] <twnqx> depends how it is regulated... modern PSUs only regulate the 12V and leave the rest to autobalance
[10:02:01] <twnqx> (balanced via transformer windings)
[10:02:13] * megal0maniac gets out cable with crocodile clips and AC plug
[10:02:24] * twnqx expects ping timeout
[10:02:32] <megal0maniac> :D
[10:02:50] <jadew> lol
[10:03:23] <megal0maniac> Need to build it first
[10:06:00] <megal0maniac> If you have two rails, can there be continuity between 12V and 12V if they are on different rails?
[10:06:13] <twnqx> the one i disassembled
[10:06:19] <twnqx> hat ONE regulator
[10:06:27] <twnqx> and three current senses. for three reails.
[10:06:37] <megal0maniac> So no :)
[10:06:40] <megal0maniac> I mean yes
[10:08:49] <tzanger> megal0maniac: on what, a PC supply?
[10:08:58] <megal0maniac> yes
[10:09:43] <RikusW> megal0maniac: I have such a plug + clips...
[10:09:53] <RikusW> aka suicide clips :-P
[10:10:29] <RikusW> useful for applying power to disassemble psu's
[10:10:39] <megal0maniac> Nipple clips :)
[10:10:51] * megal0maniac imagines everyone cringing
[10:11:07] <twnqx> you can do whatever with your nipples
[10:11:17] <asteve> uh
[10:11:38] <tzanger> megal0maniac: a PC supply has ONE 12V rail most times
[10:11:53] <tzanger> there are a dozen 12V wires coming out but if you take it apart you can see they all connect to the same place
[10:11:55] <asteve> so, I had an mri this morning, it sounded like transformers were banging on top of my face for 20 minutes
[10:11:56] <tzanger> *most* times
[10:12:09] <tzanger> and yes, you are a grown man, you can do whatever you want with your nipples
[10:12:25] <megal0maniac> tzanger: And if it claims to have two rails?
[10:12:39] <tzanger> megal0maniac: I'd be suspicious
[10:12:58] <megal0maniac> Well I'm getting continuity on everything 12V
[10:14:54] <twnqx> asteve: close, just electromagnets :P
[10:15:18] <twnqx> interesting. why does this make such a difference :S
[10:16:07] <twnqx> drawing 30mA from mains to get 1mA at +.15V is not particularly efficient, is it...
[10:16:13] <twnqx> +-15V*
[10:19:04] <megal0maniac> twnqx: At mains frequency?
[10:19:09] <twnqx> yes
[10:19:36] <megal0maniac> As my electronics lecturer likes to say, "it is not a magic"
[10:19:46] <megal0maniac> But you have done something funny
[10:19:56] <twnqx> "funny"
[10:20:08] <twnqx> i just connected a linear regulator straight tomains
[10:20:28] <tzanger> twnqx: why?
[10:20:33] <twnqx> virtually.
[10:22:11] <tzanger> ah virtually
[10:24:10] <twnqx> and here i was wondering why the (virtual) circuit didn't work if i revered the AC connections
[10:24:18] <twnqx> it's just a matter of the reference :P
[10:25:50] <twnqx> guess i'll order some high power diodes
[10:26:10] <twnqx> 1n4148 doesn't sound like they can handle 325V
[10:26:49] <twnqx> hm 1n4001
[10:26:58] * megal0maniac hasn't done this before
[10:27:03] <megal0maniac> Measure current over live?
[10:27:05] <twnqx> hm, nope
[10:27:57] <twnqx> 1N5400...
[10:28:10] <twnqx> still 50V, damn
[10:28:24] <megal0maniac> Anyone?..
[10:28:50] * twnqx just uses a wattmeter
[10:29:05] <megal0maniac> I'm using a multimeter
[10:29:16] <megal0maniac> Current over live>
[10:29:18] <megal0maniac> ?
[10:29:25] <twnqx> does it matter
[10:29:27] <twnqx> current is current
[10:29:35] <megal0maniac> I don't know. That's why I'm asking :)
[10:29:44] <twnqx> if you current in one direction on one wire
[10:29:51] <twnqx> guess what you'll have on the other wire
[10:30:21] <megal0maniac> Voltage?
[10:30:25] * megal0maniac is trolling :)
[10:30:32] <megal0maniac> Okay, so live it is
[10:30:47] <megal0maniac> If I haven't said anything in the next 10 minutes...
[10:33:17] <megal0maniac> Lol
[10:33:28] <megal0maniac> 570mA
[10:33:35] <twnqx> rms?
[10:34:20] <twnqx> and now you have VA
[10:34:30] <twnqx> only need the power factor to convert to watts!
[10:34:56] <megal0maniac> P=VI
[10:35:47] <twnqx> yes, for DC
[10:35:50] <twnqx> not quite for AC
[10:36:36] <megal0maniac> Oh. Yeah, because you take frequency and inductance into consideration
[10:36:55] <tzanger> P=VI cos theta for AC, no?
[10:38:15] <Badaboom> twnqx: 85v?
[10:42:21] <megal0maniac> 500mA - 600mA while installing Windows. Graphics card fan going crazy.
[10:42:29] <twnqx> Badaboom: ?
[10:42:47] <Badaboom> was it 85v you were looking for?
[10:43:01] <twnqx> at least 300-400
[10:43:06] <Badaboom> ok
[10:43:12] <twnqx> but i was looking thorugh what i have :P
[10:43:17] <twnqx> 1n4007 will do
[10:43:22] <Badaboom> rgr
[10:43:44] <Badaboom> <---waiting on Atmel Delivery
[10:43:52] * twnqx looks atreamins of PC PSU
[10:44:04] <Badaboom> \lol
[10:44:07] <Badaboom> you killed it
[10:44:12] <twnqx> by desoldering parts, yes
[10:44:32] <twnqx> HER208
[10:44:33] <Badaboom> i love doing that,, ive spent whole days recycling parts
[10:44:34] <twnqx> perfect.
[10:44:41] <twnqx> 1kV, 2A
[10:45:37] <Badaboom> Think its gonna storm here
[10:45:59] <Badaboom> brb,, gonna check rad
[10:46:00] <twnqx> stormy here all day long
[10:47:49] <Badaboom> http://www.baynews9.com/content/news/baynews9/weather/klystron-9-radar.map.html/Pinellas.html
[10:48:04] <Badaboom> look at the cell building in my area
[10:49:16] <twnqx> guess i'll just be happy that we don't have tornadoes
[10:50:28] <megal0maniac> 370mA at idle. I don't think power is my problem
[10:50:31] <Badaboom> yep,, we get them here sometimes,, nothing like the on in OK
[10:52:17] <beaky> what programming languages can I program avr in?
[10:52:41] <twnqx> C
[10:52:43] <twnqx> asm
[10:52:44] <twnqx> C++
[10:53:22] * twnqx wonders if he could compile the gcc fortran frontend
[10:53:26] <twnqx> for avr.
[10:53:45] <beaky> I wish I can use a higher-level language
[10:54:01] <beaky> like python, js, ruby, haskell, or lisp
[10:54:22] <beaky> asm, C, and C++ are great, but too low level :D
[10:55:04] <twnqx> the other sound more like STDs than programming languges to me...
[10:55:10] <Badaboom> lol
[10:56:15] <beaky> I did saw colleagues programming an arduino in js and python, but I guess arduino is different from avr
[10:56:23] <beaky> did see*
[10:56:30] <Badaboom> damn,, the cel is moving off,, i put my tomato plant outside :(
[10:57:03] <beaky> I guess C++ is king in MCU development
[10:57:24] <twnqx> no
[10:57:30] <twnqx> C is
[10:57:54] <twnqx> you don't always have enough memory for C++
[10:58:05] <twnqx> with 256 bytes ram in small systems
[10:58:15] <Badaboom> speaking of ,, time to crank up studio
[10:59:28] <twnqx> and that's 2x HER208 scavenged :3
[10:59:35] <jadew> if you have memory for C, you have enough for C++ too :) it's just that you might not have enough memory to waste on OO programming :P
[10:59:45] <twnqx> well, ok
[11:00:08] <beaky> C++ takes a lot of memory?
[11:00:09] <twnqx> i still haven't found true benefits of c++ anyway
[11:00:13] <beaky> ah
[11:00:15] <jadew> beaky, no
[11:00:16] <twnqx> objects do
[11:00:20] <twnqx> c++ itself not
[11:00:39] <twnqx> in fact, only objects with virtual function tables
[11:00:47] <jadew> twnqx, come on... templates, objects, optional parameters
[11:01:02] <twnqx> objects are things i hate with a passion
[11:01:04] <beaky> I tend to handroll vtables in C, so I guess C++ is alright for avr :D
[11:01:27] <twnqx> i never needed optional parameters, except varargs
[11:01:37] <jadew> twnqx, why would you hate objects?
[11:01:48] <twnqx> they bloat the amount you have to write
[11:02:00] <jadew> that's really not true
[11:02:14] <twnqx> and probably my very first contact with oop was..
[11:02:15] <twnqx> well
[11:02:18] <jadew> if you have a complex program, it would most likely be bloated and crappy with out objects
[11:02:34] <twnqx> "oh hey, it's all inherited, so you have to go 5 levels up, then manually instatantiate 4 levels back down"
[11:02:47] <twnqx> you know
[11:02:51] <twnqx> that's wrong
[11:02:58] <jadew> well, that's not what OOP is about
[11:03:08] <beaky> OOP is about message-passing
[11:03:13] <twnqx> bullshit
[11:03:21] <twnqx> but your data into a struct
[11:03:22] <twnqx> and pass it
[11:03:27] <twnqx> you have the same
[11:03:35] <twnqx> (except for private/public distinction)
[11:03:46] <twnqx> and compiler type checks
[11:03:53] <beaky> and virtuality
[11:03:57] <jadew> and templates
[11:04:13] <jadew> and destructors
[11:04:20] <jadew> and inheritance
[11:04:22] <beaky> well, templates and dtors aren't really OO :D
[11:04:27] <jadew> yep, c == c++ :P
[11:04:37] <jadew> beaky, destructors are
[11:04:40] <beaky> ah
[11:04:42] <twnqx> implicit ctors/dtors are expressively oop
[11:04:50] <jadew> template objects are a huge advantage in c++ tho
[11:05:16] <beaky> imagine how tiny and efficient avr programs would be if we used templates :D
[11:05:16] <twnqx> for code obfuscation i will happily agree
[11:05:27] <twnqx> beaky: the binary will be larger
[11:05:32] <jadew> FALSE
[11:05:36] <jadew> on both counts
[11:05:41] <twnqx> while the source will probably be roughly the same
[11:05:47] <jadew> first of all, it wouldn't obfuscate anything
[11:05:58] <twnqx> ok, that depends on nesting and use
[11:06:07] <jadew> second of all an instantiated template will have the same code as a specifically designed object
[11:06:14] <twnqx> (though most template magic i have seen causes obfuscation as side effects)
[11:06:18] <megal0maniac> < beaky> imagine how tiny and efficient avr programs would be if we used templates :D
[11:06:29] <megal0maniac> Arduino ;)
[11:06:57] <beaky> my avr-g++ supports C++11 :D
[11:06:59] <twnqx> jadew: i am a lost cause for oop, sorry
[11:07:00] <jadew> twnqx, the obfuscation you're talking about comes from the complexity of the code
[11:07:06] <jadew> the complexity of the definition
[11:07:26] <jadew> it's something you can't even do in C, you'd have to start write the final result from scratch
[11:07:50] <jadew> where the "obfuscated" teamplates help you write that same crap, with specific traits, properly defined and written, in one go
[11:07:57] <twnqx> i wouldn't necessarily agree, but preprocessor magic in C isn't less obfuscating
[11:08:12] <jadew> twnqx, I'm sure you'd get around if you actually found a good use of OOP
[11:08:29] <twnqx> the only OOP i ever saw
[11:08:36] <twnqx> was either totally misused
[11:08:57] <beaky> I'm also looking for a good use of OOP myself
[11:08:59] <twnqx> or twice the sourcecode size the same thing would have taken in C
[11:09:55] <jadew> let me write some code
[11:09:59] <beaky> OOP has been overhyped too much by the "enterprise methodology software engineering" folks who chant buzzwords about "dependency injection" and "design patterns" all day
[11:10:03] <twnqx> really, don't try to convinve me
[11:10:15] <twnqx> what, design patterns are good
[11:10:37] <twnqx> they stop people from reinventing the wheel
[11:11:02] <jadew> http://pastebin.com/EvnTjX9N
[11:11:11] <jadew> tell me how exactly is the C code better
[11:11:21] <jadew> if anything, it's worse, because it has an additional fail point
[11:11:40] <jadew> second of all, if well written, the C++ code can guarantee it's leak free
[11:11:58] <twnqx> how do you guarantee against memory fragmentation?
[11:12:08] <twnqx> especially with so little ram as avrs have?
[11:12:14] <jadew> twnqx, that's something NOTHING can guarantee it
[11:12:19] <twnqx> oh sure
[11:12:22] <jadew> you'd have to write your own memory manager to do it
[11:12:26] <twnqx> i can perfectly manually manage my ram
[11:12:34] <jadew> come on...
[11:12:56] <jadew> you can obviously override new or malloc and move shit around in the ram
[11:13:02] <jadew> but that's very... not efficient
[11:13:18] <beaky> avr has tiny ram? so no hope for running java on it? :(
[11:13:24] <twnqx> you can't, there might be dangling pointers
[11:13:39] <jadew> yeah, but we're not talkign about c++ vs C on an avr
[11:13:42] <jadew> that thing is covered
[11:13:49] <jadew> we're talking about C++ and objects in general
[11:14:47] <beaky> objects are a really poorly-defined concept in general
[11:15:02] <beaky> what is an object anyway?
[11:15:04] <twnqx> like i said, i'm a lost cause for it
[11:15:31] <jadew> beaky, it's an instance of a deffinition, it contains data and has associated methods
[11:15:52] <twnqx> it's a struct that binds varaiables and function pointers together, yey
[11:16:00] <jadew> you'd obviously have to look it up if you have to ask that question
[11:19:08] <jadew> anyway, it's clear that you can make anything with out objects, it's just messier and harder to maintain
[11:20:04] <beaky> programming in C++ is a real pleasure thanks to templates and dtors
[11:20:24] <beaky> (and a real pain when templates go wrong :D 50 page errors are normal)
[11:20:48] <beaky> alright I will port my traffic light thing to C++
[11:20:58] <jadew> dude...
[11:21:01] <jadew> for AVR
[11:21:23] <jadew> I still use C, even tho I hate it
[11:21:31] <jadew> well, I don't hate it
[11:21:38] <jadew> but it's lacking so many cool stuff from C++
[11:21:50] <twnqx> so why don't you use c++
[11:21:51] <beaky> why use C over C++ when C++ is an option on AVR?
[11:22:15] <jadew> beaky, because most shit on an AVR are signleton
[11:22:25] <beaky> (I guess C++ is pretty rare on resource-constrained platforms like MCUs)
[11:22:29] <beaky> ah
[11:23:02] <jadew> and since memory is an issue, having a pointer to an object means more memory use
[11:23:22] <twnqx> so use static objects?
[11:23:38] <beaky> yeah idiomatic C++ doesn't have pointers everywhere :D
[11:23:40] <twnqx> and if you don't use virtual functions, the overhead should be nil
[11:23:56] <jadew> twnqx, I think methods of static objects will still take a pointer as a parameter
[11:24:05] <jadew> that's again wasted cycles
[11:24:08] <kobsu> o.O
[11:24:13] <twnqx> they still will have self, yes
[11:24:26] <jadew> on an AVR you have to save as much as you can
[11:24:41] <jadew> so code niceness is trade off for more available space
[11:24:51] <twnqx> i think you _always_ should
[11:25:41] <twnqx> but i am complaining that the several thousandfold increase in resources on my PC compared to the c64 did nowhere translate to benefits in the same order of magnitude
[11:26:11] <twnqx> except for tightly packed math like video encoding
[11:26:18] <jadew> well, sometimes 10 times bigger code is only twice faster
[11:26:21] <jadew> that's fine in my book
[11:26:25] <twnqx> not in mine
[11:26:34] <jadew> we're not in the 80's anymore
[11:26:38] <twnqx> so what
[11:26:44] <twnqx> i am still paying for programmer's lazyness
[11:26:49] <twnqx> and billions others with me
[11:26:56] <jadew> space is not an issue anymore and all PC's have huge RAMs
[11:27:01] <twnqx> yes
[11:27:04] <twnqx> which i have to buy
[11:27:08] <jadew> that's not programmer's lazyness
[11:27:10] <twnqx> thanks to programmers lazyness
[11:27:11] <jadew> but just the other way around
[11:27:18] <jadew> it's optimization for something else
[11:27:22] <twnqx> "it's there, let's waste it"?
[11:27:29] <jadew> yes!
[11:27:33] <jadew> it's there, let's USE IT
[11:27:35] <jadew> not waste it
[11:27:55] <twnqx> in fact it's only there
[11:27:57] <jadew> why do something in 10 bytes and take 5 seconds, when we can do it in 200 bytes and take only 1
[11:27:59] <twnqx> because programmers demand it
[11:28:00] <jadew> which one do you prefer?
[11:28:07] <jadew> laggy program or big program?
[11:28:07] <twnqx> well
[11:28:11] <twnqx> modern programs seem to me
[11:28:24] <twnqx> like take 10x the memory to do some half as fast on a 4times as fast platform
[11:28:54] <jadew> yeah, that's not really an issue
[11:29:03] <twnqx> what
[11:29:15] <twnqx> being slower while using 40 tiems the resources is not to you?
[11:29:15] <jadew> you'd be wrong if you wouldn't take advantage of the higher resources of the modern PCs
[11:29:31] <jadew> oh, "half as fast"
[11:29:32] <twnqx> see, and this is where i only can disagree
[11:29:36] <jadew> I don't know wherey ou got that
[11:29:45] <twnqx> just about everywhere i look
[11:30:09] <beaky> what is the best way to make a led blink every .5s on avr?
[11:30:16] <jadew> I think you might mistake proper code with fast, imporper code
[11:30:35] <LoRez> beaky: "best" is dependent upon what your beliefs and other requirements are.
[11:30:48] <jadew> every code will get slower once it matures, because it gets fixed and shit the origianl code didn't take into account, is not accounted for
[11:30:52] <beaky> ah
[11:30:54] <twnqx> with a users definition of code (effect), or a programmers definition (ease of maintenance)?
[11:31:10] <beaky> what ways have you guys discovered to blink leds, and what works best?
[11:31:11] <jadew> so more checks, more code running, better result => slower program, obviously
[11:31:20] <beaky> atm I am blinking a led using the 8-bit timers on my avr
[11:31:37] <twnqx> i normally register a timer callback for that
[11:31:46] <twnqx> obviously only works in my code.
[11:31:47] <beaky> ah how do I do that
[11:32:04] <LoRez> timers and interrupts would work fine if you're using a chip that has them
[11:32:09] <twnqx> you write an interrupt handler and structures to hold variables delays and callback functions...
[11:32:15] <beaky> ah that's what I do atm
[11:32:30] <LoRez> it also lets you sleep the chip while doing it
[11:32:53] <twnqx> beaky: i once used an fpga for that.
[11:32:59] <beaky> yeah sleep is awesome
[11:33:04] <beaky> wow an fpga seems overkill :D
[11:33:37] <twnqx> jadew: if the checks are there to validate your own code that's a debug build, not something the user should have
[11:33:39] <beaky> I thought fpga were for specific applications that needed high-performance like market trading
[11:33:51] <beaky> or figher-jet navigation
[11:33:54] <twnqx> you can also build LED blinkers with them.
[11:34:27] <jadew> twnqx, obviously there's not there to validate your own code
[11:34:37] <jadew> they're there to validate shit that the "fast code" did not
[11:34:42] <jadew> take xfce for example
[11:34:46] <jadew> it's a piece of garbage
[11:34:52] <beaky> but I use xfce :(
[11:35:13] <twnqx> piece of garbage from which perspective?
[11:35:22] <jadew> if I trigger the menu, while my mouse is on a different monitor, it shows up on that monitor, instead to show up above the "start" button
[11:35:27] <beaky> ah
[11:35:39] <jadew> twnqx, from the perspective of not being fully developed
[11:35:42] <twnqx> i would consider that how it should be
[11:35:44] <jadew> which is why it's fast
[11:36:00] <jadew> it's cutting corners => fast
[11:36:21] <twnqx> why the hell yould you want the menu on a different monitor than the mouse currently is at :S
[11:36:50] <twnqx> i would sy your example is more like "works as designed"...
[11:36:54] <jadew> twnqx, if you hit Alt + F
[11:37:16] <twnqx> i use e17
[11:37:18] <jadew> would you expect the "File" menu to show up under "File" or over the contents of some movie you're running on a separate monitor?
[11:37:19] <twnqx> so i can't comment
[11:37:29] <twnqx> i would expect it
[11:37:36] <twnqx> to open in the window that my mouse is in
[11:37:41] <jadew> come on...
[11:37:44] <twnqx> since i use sloppy focus
[11:37:49] <jadew> you obviously don't believe that
[11:37:53] <twnqx> what
[11:37:55] <jadew> what if you have another program there?
[11:38:00] <jadew> one that launches nukes
[11:38:06] <jadew> or better yet
[11:38:07] <twnqx> the window my mouse is over is the active one
[11:38:17] <jadew> what if the program that launches nukes is the one with the menu
[11:38:21] <twnqx> if the mouse is not over it, it's inactive
[11:38:29] <jadew> and the program on top of which the menu shows up is mine sweaper
[11:38:37] <twnqx> so i expect it to open the file menu on the active window
[11:38:37] <jadew> omg...
[11:38:38] <jadew> dude!
[11:38:46] <jadew> the menu has to appear where it belongs
[11:38:50] <twnqx> why the hell should it choose random windows
[11:38:52] <jadew> it shouldn't appear over other crap
[11:38:53] <twnqx> yes!
[11:39:01] <twnqx> EXACTLY
[11:39:05] <edmont> hi
[11:39:18] <jadew> and it doesn't belong on a separate display, if the point of origin for the menu is on another
[11:39:21] <jadew> it's as simple as taht
[11:39:27] <twnqx> ...
[11:39:37] <twnqx> ok, i might be misunderstanding you
[11:39:44] <twnqx> it's the only thing i can assume
[11:39:52] <edmont> is it possible to prevent avrdude from erasing the whole flash when writting only the bootloader?
[11:39:53] <twnqx> since my use of WMs is not the usual one
[11:40:02] <jadew> twnqx, let me see if I can find a screenshot
[11:40:13] <twnqx> like, i won't accept the WM using any space on the desktop
[11:40:48] <twnqx> (ok, i give it one px height on the top, atm
[11:41:07] <twnqx> but menus belong just in the center of the active display for me
[11:41:19] <twnqx> and that's what win+r does for me
[11:42:08] <twnqx> edmont: afaik it uses chip erase, not any possibly available sector erase
[11:42:32] <edmont> ok :(
[11:42:51] <twnqx> you have to set a fuse so it doesn
[11:42:57] <twnqx> 't erase the eeprom while it's at it
[11:44:13] <edmont> that's nice
[11:44:21] <edmont> i also wanted this :)
[11:54:49] <beaky> how do I build a makefile on linux for avr?
[11:56:59] <twnqx> with an editor of your choice
[11:57:11] <twnqx> but best, take a reference makefile, and edit that
[11:57:21] <beaky> ah where can I get one of those :D
[11:57:45] <twnqx> dunno, i got mine years ago :X
[12:00:24] <twnqx> https://www.google.de/search?q=%23+WinAVR+makefile+written+by+Eric+B.+Weddington%2C+J%C3%B6rg+Wunsch%2C+et+al.&ie=utf-8&oe=utf-8&aq=t&rls=org.mozilla:en:official&client=firefox-a
[12:00:34] <twnqx> this should get you a ton of those Makefiles :P
[12:04:09] <beaky> hmm there was this avr-project program :(
[12:04:20] <beaky> when I was on a mac
[12:48:42] <inkjetunito> beaky: avr-gcc ships with examples
[13:05:05] <beaky> ah thanks
[13:07:13] <beaky> http://ideone.com/M93p9G vs http://ideone.com/j6U0YD vs http://ideone.com/LeSe9J, which version is best?
[13:13:34] <inkjetunito> beaky: not this one http://ideone.com/M93p9G
[13:14:10] <beaky> hahaha I figured that :(
[13:14:22] <beaky> that was my attempt at doing a FSM of that thing
[13:14:45] <beaky> http://www.youtube.com/watch?v=srhC-1_fXeQ (not my video, but the target device is something similar)
[13:21:19] <beaky> his also makes the following lights go amber when the current one is amber, and rings a bell
[13:37:25] <beaky> why is most avr code not elegant?
[13:37:37] <thetruthisoutthe> sup beaky ?
[13:38:21] <beaky> hello
[13:54:36] <thetruthisoutthe> anybody know how does the ps2 keyboard initialization goes?
[13:54:55] <thetruthisoutthe> (at pc start time)
[13:55:25] <OndraSter> first is RESET
[13:55:26] <OndraSter> :)
[13:55:46] <thetruthisoutthe> so i should wait for reset? don't try sending BAT_OK ?
[13:56:02] <OndraSter> check online
[13:56:13] <thetruthisoutthe> online says various conflicting information
[13:56:20] <thetruthisoutthe> too many combinations
[13:56:27] <thetruthisoutthe> impossible to guess
[13:56:29] <OndraSter> I was guessing and usually RESET is always the first thing on almost all the stuff
[13:56:49] <OndraSter> http://sourceforge.net/projects/ps2avr/
[13:56:51] <OndraSter> check source
[13:56:59] <thetruthisoutthe> that source fails btw
[13:57:09] <OndraSter> define fails?
[13:57:16] <thetruthisoutthe> won't even compile
[13:57:21] <thetruthisoutthe> it is not done
[13:57:24] <OndraSter> if you have got a different chip..
[13:57:33] <thetruthisoutthe> no i maen it won't compile on gcc
[13:57:52] <OndraSter> what is the error?
[13:58:31] * RikusW just removed a sot23-5pin using a J shaped wire and iron
[13:58:33] <thetruthisoutthe> avr-objcopy:keymain.cof: Invalid bfd target
[13:58:44] <OndraSter> and I said check source, not compile source :P
[13:58:47] <OndraSter> you can check what the init goes
[13:58:48] <OndraSter> how*
[13:58:49] <thetruthisoutthe> lol
[13:58:52] <thetruthisoutthe> i see
[13:59:19] <OndraSter> RikusW, hotair ftw :P
[13:59:25] <OndraSter> on everything except PTH
[13:59:33] <RikusW> solidcore wire ftw :_
[13:59:35] <OndraSter> :D
[13:59:44] <RikusW> I used 0.5mm wire
[13:59:57] <OndraSter> you have no idea how much work it is to desolder a bunch of 74 series chips.. through hole stuff... from the 70s/80s
[14:00:03] <RikusW> had to cut a track underneath the regulator... (en pin)
[14:00:05] <OndraSter> chips that are not made anymore
[14:00:07] <OndraSter> 74 141
[14:00:20] <OndraSter> aka BCD nixie tubes
[14:00:21] <OndraSter> drivers
[14:00:43] <RikusW> OndraSter: I have stripped 8088 motherboard, quite hard to remove from plated holes...
[14:00:44] <OndraSter> I sold the big nixies I had here, still have the smaller ones
[14:00:47] <OndraSter> yep
[14:00:53] <RikusW> especially if the pins are bent...
[14:00:57] <OndraSter> I still have got some Z80 based boards from... plotter I think it was
[14:00:58] <OndraSter> yep
[14:01:06] <OndraSter> I took one chip off, one lost half of its legs
[14:01:08] <OndraSter> so I gave up
[14:01:12] <RikusW> though a heatgun works nicely
[14:01:14] <OndraSter> and dumped the EPROMs that were in slots :D
[14:01:24] <OndraSter> no readable ASCII :(
[14:01:36] * RikusW hand OndraSter a soldersucker and iron to remove a bunch of 74 ics
[14:01:38] <OndraSter> but I have got a buttload of old school EPROMs
[14:01:40] <OndraSter> haha
[14:01:43] <OndraSter> soldersucker
[14:01:45] <OndraSter> have one
[14:01:52] <OndraSter> but my tin does not bond with the existing one
[14:01:56] <beaky> how do i do something useful with my avr?
[14:02:05] <OndraSter> beaky, power it up and programm it
[14:02:10] <beaky> ah ok
[14:02:20] <RikusW> OndraSter: soldersucker is almost useless with plated holes
[14:29:28] <stanreg> [attiny44]. ISR INT0 Set up to fire on any logical change. It fires twice instead of once. Trying to figure why. http://pastebin.com/0A6fSAWZ -- Note: Oddly, when line 6 is enabled, ISR INT0 no longer executes. Any ideas?
[14:30:49] <ambro718> in my experience you can't rely on the interrupt firing on every single change, if it changes fast it could "compress" the events
[14:31:17] <ambro718> so check the pin status after every interrupt
[14:32:28] <ambro718> in your case that probably means remembering the old state
[14:32:40] <stanreg> ambro718, any clue as to why disabling the "Pin Change Interrupt Enable" also disables INT0 from firing?
[14:33:06] <thetruthisoutthe> yo stanreg
[14:33:14] <stanreg> thetruthisoutthe, hola :)
[14:33:15] <ambro718> hm no, I've only ever used the PCIs
[14:33:25] <stanreg> PICs, I suppose.
[14:33:44] <thetruthisoutthe> stanreg <= because the button bounces, you need to debounce it digital filter, or delay
[14:33:51] <ambro718> huh, no, PICs, Pin Change Interrupt
[14:33:54] <ambro718> PCI
[14:33:57] <OndraSter> PCINT :)
[14:34:01] <stanreg> ambro718, ah :)
[14:34:30] <stanreg> thetruthisoutthe, if you're suggesting delaying for the ISR not to fire twice, I did. There's a 500ms delay in the ISR.
[14:34:53] <ambro718> not a good idea, don't wait in interrupts
[14:35:25] <stanreg> Another (possibly related?) problem is, as soon as I disable Pin Change Interrupt, INT0 seems to get disabled as well.
[14:35:30] <ambro718> AVRs have memory... use it
[14:35:39] <tzanger> set a flag and catch it in the main routine to delay and check
[14:35:46] <tzanger> make sure the flag is volatile
[14:35:53] <thetruthisoutthe> stanreg <= don§t do delay in ISR
[14:36:39] <thetruthisoutthe> setup a timer, and count time since pressed instead
[14:36:52] <stanreg> hm
[14:37:11] <cart_man> RikusW: Hey Rikus
[14:37:17] <thetruthisoutthe> you can increment a variable (or decrement) with a timer overflow interrupt
[14:37:36] <thetruthisoutthe> and 16ms is fine
[14:37:37] <stanreg> That's interesting.
[14:38:28] <stanreg> Or simply a double verification (slightly delayed) before execution, I suppose.
[14:39:00] <stanreg> Still doesn't answer why disabling Pin Change Interrupts criples the code.
[14:39:08] <thetruthisoutthe> your interrupt may trigger in 200ns...
[14:40:59] <cart_man> Do you guys think its possible to make RX and TX go through 2n2222 tarnsistors and still come out the other side without the voltage drop? becaues the voltage drop is over BE of the Transmitter right?
[14:41:00] <RikusW> hi cart_man
[14:41:26] <cart_man> RikusW: Im going to buy myself that GPS IC that you linked me
[14:41:48] <cart_man> RikusW: they sell them quite close to me..little past Communica Midrand
[14:41:53] <RikusW> ah
[14:42:09] <RikusW> Otto in Randburg also sells then for R165 ex
[14:42:14] <cart_man> RikusW: Then id just need to figure out how Seriel work on ATMELS O.o
[14:42:15] <RikusW> www.otto.co.za
[14:42:25] <RikusW> very easy actually
[14:42:29] <thetruthisoutthe> cart_man <= use inverting mode, and saturate the transistor
[14:42:50] <cart_man> Yes saturation was the thought but what do you mean with invert?
[14:43:02] <cart_man> Invert RX TX signals?
[14:43:03] <thetruthisoutthe> ground source, and have a pullup resistor
[14:43:54] <thetruthisoutthe> *emitter
[14:48:19] <RikusW> cart_man: I used this for a level translator between two uarts http://imagebin.org/258594
[14:48:41] <RikusW> also, one device can be off without any current flowing through the serial lines
[14:48:52] <twnqx> really
[14:48:53] <RikusW> thats the actual reason I made this
[14:48:56] <twnqx> 4 resisoros
[14:49:00] <twnqx> 2 transistors
[14:49:02] <twnqx> wow
[14:49:14] <RikusW> and it works at 115200bps too
[14:49:29] <twnqx> cute
[14:49:33] <RikusW> thanks
[14:49:38] <OndraSter> why level converters? What voltage is the IO on the module?
[14:49:41] <twnqx> i usually use 120mhz chips
[14:49:56] <RikusW> 2v8 on both actually
[14:50:03] <OndraSter> use xmega then :)
[14:50:06] <OndraSter> 2.8V -> 32MHz
[14:50:12] <RikusW> but one or the other device might be off
[14:50:17] <OndraSter> so? :P
[14:50:26] <OndraSter> oh
[14:50:28] <OndraSter> I see
[14:50:33] <OndraSter> clamping would power the second chip up
[14:50:34] <RikusW> I don't want current running through the serial wires...
[14:52:33] <RikusW> err actually I used that from 5V -> 2v8
[14:53:28] <RikusW> and back again
[14:53:31] <RikusW> rx+tx
[14:54:01] <RikusW> http://www.chunx.com/e2a/images/Chaos%20Field.jpg
[14:57:13] <Badaboom> whats the chances of me finding pinouts of some newer lcd displays from cell phones i have (parts)?
[14:57:52] <RikusW> for nokia its possible
[14:58:00] <Badaboom> and samsung?
[14:58:21] <RikusW> no idea
[14:58:24] <RikusW> google ?
[14:58:35] <Badaboom> hmm,, am i better off just getting a module ?
[14:58:46] <Badaboom> yeah tried google
[14:58:51] <RikusW> might be easier, just make sure you get docs for it too
[14:59:04] <Badaboom> yup,, was thinking crystalfontz
[14:59:23] <Badaboom> they seem reasonable
[15:01:05] <twnqx> the problem is
[15:01:14] <twnqx> i haven't found a source of nice, small, highres displays
[15:01:23] <twnqx> except phone spares >_>
[15:01:54] <Badaboom> very true,, i have alot of them here ,, very good res but no data:(
[15:02:47] <Badaboom> and seeing as i now have 43 atmels here,,lmao,, i need projects
[15:03:11] <stanreg> Well, well.. It turns out that (a) GIMSK &= (0<<PCIE) and (b) GIMSK &= ~(1<<PCIE) -- compile differently and produce different results. (a) stops INT0 from working, (b) is good. Weird.
[15:04:33] <cart_man> RikusW: nice stuff ..so that way I can work mulitple Seriel devices with 1 chip using pins to switch on and off....EPIC!
[15:04:37] <Badaboom> Arnie & Vern would be so proud
[15:05:41] <cart_man> RikusW: BTW. do you pulse the GPS unit if you want it to send the Co-Ords?
[15:05:55] <RikusW> cart_man: it sends them by itself
[15:05:55] <cart_man> RikusW: or does it just spam it on it sonw
[15:06:06] <RikusW> you can set how many ms apart though
[15:06:13] <twnqx> stanreg: in the first case you wrote gimsk &= 0, which can be shortened to gimsk = 0, what do you expect
[15:06:28] <RikusW> using PMTK220,1000 for 1s
[15:06:36] <cart_man> RikusW: ok one last question... The circuit in the datasheet...is that really all it takes? couple of caps and resistors? not much of a Datasheet unless I have the wrong one
[15:06:50] <RikusW> http://imagebin.org/258594
[15:07:03] <RikusW> or just and AVR running at 3v3 works fine too
[15:07:11] <RikusW> 2v8 isn't that far away
[15:07:26] <stanreg> twnqx, doesn't (a) simply put AND gimsk with gismk in which '0' has been shifted to position PCIE?
[15:07:32] <RikusW> the PA6H does recommend putting 330 Ohm resistors on its serial lines though
[15:07:41] <twnqx> no, 0 shiftleft something is still 0
[15:07:41] <RikusW> I used 470 for my tracker
[15:08:08] <twnqx> so you and with overall 0
[15:08:14] <RikusW> cart_man: the imagebin one is good for 5V -> 2v8 and 2v8 -> 5V
[15:08:17] <stanreg> twnqx, understood. Thanky.
[15:08:28] <RikusW> I tested up to 115200bps
[15:08:44] <RikusW> had a bit of trouble with that, used too large resistors...
[15:09:06] <RikusW> and didn't have the top 10k one
[15:09:25] <RikusW> without it, it don't work too well.....
[15:09:42] <RikusW> or at all
[15:11:09] <cart_man> Well Kudos figuring that out!
[15:11:12] <cart_man> thanks Rikus
[15:11:30] <RikusW> I _had_ to make it work ;)
[15:11:43] <RikusW> its for interfacing my U2S to the SIM900
[15:11:47] <cart_man> nice catalyst that
[15:11:52] <RikusW> heh
[15:12:21] <thetruthisoutthe> RikusW <= even a 2k resistor and a parallelport pin high powers an atmega168pa :)
[15:12:21] <cart_man> stress is my catalyst to learn python right this very moment :s
[15:12:45] <thetruthisoutthe> at works with even a red led on, 1.8V
[15:12:53] <RikusW> I should learn some python sometime too...
[15:13:12] <RikusW> thetruthisoutthe: I'm fairly certain my circuit doesn't leak..
[15:13:21] <RikusW> if either device is off
[15:17:27] <thetruthisoutthe> RikusW <= clamp diodes power the chip through any io pin
[15:17:46] <RikusW> thats what http://imagebin.org/258594 will prevent
[15:18:22] <RikusW> if you don't care that the logic level is inverted a single NPN will do
[15:18:32] <RikusW> for uart thats not an option
[15:20:51] <thetruthisoutthe> i don't care about backfeed
[15:38:16] <Badaboom> twnqx: i gonna work on a source for displays
[15:39:16] <Badaboom> crystalfontz is a little high when u get into larger displays
[15:45:27] <megal0maniac> abcminiuser: You get that window to dock yet? :)
[15:45:44] <abcminiuser> Ja
[15:45:56] <abcminiuser> Lemme just upload it...
[15:46:16] <megal0maniac> And I'll uninstall it
[15:46:21] <megal0maniac> cart_man: 'Lo
[15:46:34] * abcminiuser is building a fresh copy, stand by
[15:46:46] <abcminiuser> I had to use an internal IDE hack to force it to MDI dock
[15:46:49] <abcminiuser> Not pretty, but works
[15:46:51] <abcminiuser> In theory
[15:47:01] <megal0maniac> But it was working before :?
[15:47:36] <megal0maniac> Saleae Logic doesn't run on armel :(
[15:47:46] <abcminiuser> Old version uses DTE.ItemOperations.Navigate, using the embedded web browser
[15:47:54] <megal0maniac> SHUT UP LADYBIRD
[15:47:59] <megal0maniac> I don't want to update
[15:48:10] <abcminiuser> Which is slow, limited in styling and prevents me from making crazy links like one direct to the LUFA integrated help
[15:48:17] <abcminiuser> New version uses native XAML
[15:48:54] <megal0maniac> Ah, so that's what broke it
[15:49:19] <megal0maniac> Okay, so undocking was a side effect of something more awesome :)
[15:49:42] <abcminiuser> http://fourwalledcubicle.com/files/temp/LUFA-TESTING-13.05.22-22.35.51.vsix
[15:49:56] <abcminiuser> Well, more stable and easier to extend in interesting ways
[15:50:13] <abcminiuser> For example, I could make a different screen pop up with relevant drivers when a LUFA device enumerates for the first time
[15:50:16] <abcminiuser> That sort of thing
[15:51:02] <megal0maniac> Because it's more native and less hackish
[15:52:02] <abcminiuser> Ja
[15:52:09] <megal0maniac> abcminiuser: Are you at GMT-4?
[15:52:35] <megal0maniac> Can't be...
[15:52:51] <OndraSter_> he is same as me
[15:52:54] <OndraSter_> which should be the same as you
[15:52:57] <OndraSter_> or is he +1 from us?
[15:52:58] <megal0maniac> Oh. Idiot. Sorry :)
[15:53:00] <thetruthisoutthe> abcminiuser <= do you know anything about ps2 keyboard?
[15:53:02] <megal0maniac> GMT+2
[15:53:06] <abcminiuser> I'm at 22:40 whatever that is
[15:53:12] <OndraSter_> same as us then
[15:53:21] <abcminiuser> thetruthisoutthe, made a decoder years and years and years and months ago
[15:53:35] <thetruthisoutthe> abcminiuser <= cany ou tell me how shoudl a keyboard init ?
[15:53:38] <megal0maniac> Stupid me was looking at yesterday's download.
[15:53:55] <megal0maniac> OndraSter_: How's studies going?
[15:54:02] <abcminiuser> It's been way too long since then I'm afriad
[15:54:06] <OndraSter_> except math great :D
[15:54:13] <abcminiuser> But I'm sure there's PS/2 protocol explanations all over the web
[15:54:31] <megal0maniac> Heh. Me too :) Wrote a maths test today and my nose started bleeding on my paper :/
[15:54:37] <thetruthisoutthe> abcminiuser <= i have a 200% completed code, and unable to init, just getting keyboard error
[15:54:45] <megal0maniac> I'm convinced it was from concentration
[15:55:20] <twnqx> Badaboom: if you find a 3-4" 840x480 or near that, tell me :P
[15:55:35] <thetruthisoutthe> i have no choice now but to build a signal analyser and log everything from a keyboard plugged in
[15:56:43] <abcminiuser> Or use USB :P
[15:56:56] <thetruthisoutthe> i must use ps2
[15:57:01] <Badaboom> twnqx: definitly,, same if u find:)
[15:57:04] <thetruthisoutthe> and it even has to work on windows
[15:58:00] <twnqx> Badaboom: replacement display on ebay is all i found
[15:58:02] <thetruthisoutthe> so i should send F1 key after BAT_OK ?
[15:58:06] <twnqx> Badaboom: we should talk to hackvana!
[15:58:10] <thetruthisoutthe> is this ps2 standard?
[15:58:11] <thetruthisoutthe> lol
[15:58:54] <Badaboom> twnqx: yeah,, i need a good source or at least find info on the ones here,, i have a ton of them im willing to share
[15:59:26] <Badaboom> there from various samsung phones and others
[16:00:28] <abcminiuser> So megal0maniac on a scale from sexy to OH MY GOD, the new version?
[16:00:35] <megal0maniac> abcminiuser: 100% for you!
[16:00:40] <abcminiuser> Woo
[16:01:14] <megal0maniac> Front page docks, help installs, xmega examples are there and there's a separate EULA for xmega examples
[16:01:35] <abcminiuser> Shweet, I win at integration
[16:01:37] <Badaboom> abcminiuser: you just reminded me of Fluffy,,lol]
[16:01:42] <abcminiuser> Helps I wrote some of the backend bits...
[16:01:48] <abcminiuser> Badaboom, who now?
[16:02:12] <Badaboom> when u said OH MY GOD,, Fluffy,, Gabriel Iglasies,, comedian
[16:02:13] <abcminiuser> megal0maniac, next version/release should have an extra section "Watch the Video"...
[16:02:39] <megal0maniac> Cool :)
[16:02:40] <abcminiuser> Ah that guy
[16:02:43] <Badaboom> lol
[16:02:47] <Badaboom> yeah him
[16:02:47] <megal0maniac> I will test the shit out of it
[16:03:26] <turutk> hi i am trying to control pwm by adc data but whenever i activate pwm, adc gets noise. how i can prevent it?
[16:03:44] <megal0maniac> abcminiuser: Only thing is that it always compiles 9999999 things regardless of the project open
[16:03:56] <megal0maniac> Is it supposed to?
[16:04:28] <twnqx> turutk: more decoupling, better pcb layout
[16:04:56] <twnqx> add some filter caps particularly at your pwm frequency, etc
[16:05:00] <abcminiuser> Yeah, it's hard to split up the internal stack unfortunately
[16:05:02] <megal0maniac> Inductor too
[16:05:09] <abcminiuser> I can look at cutting up the various classes I guess
[16:05:10] <abcminiuser> Hr,
[16:05:13] <abcminiuser> *Hrm
[16:05:33] <megal0maniac> It's no biggie. Just checking that it's expected behaviour :)
[16:06:02] <megal0maniac> Successfully created a new project, changed the device and compiled without errors
[16:06:14] <megal0maniac> And you comment like a MAN
[16:06:21] <megal0maniac> should comment
[16:06:26] <megal0maniac> :D
[16:07:33] <megal0maniac> Which I suppose is expected if the code is predominantly for the use of others. What is the difference between low level and class level API?
[16:07:51] <abcminiuser> I comment the examples since I have to :P
[16:07:51] <turutk> i have got some caps and an inductor. i am working on a breadboard. can it also affect?
[16:07:58] <abcminiuser> The core isn't so commented
[16:08:02] <twnqx> of course
[16:08:41] <twnqx> i would not expect any precision on a breadboard
[16:10:56] <megal0maniac> abcminiuser: Why is there an option of class/low level API for most examples and why?
[16:11:03] <Badaboom> twnqx: this is what i have when i can obtain a datasheet A 4-inch Super AMOLED display 800x480-pixel (WVGA)
[16:11:09] <megal0maniac> i.e what does it mean?
[16:11:11] <Badaboom> and i have 4
[16:11:34] <twnqx> Badaboom: i would be very interested.
[16:11:55] <abcminiuser> The low level APIs are for making your own custom USB class implementations and the like
[16:12:04] <abcminiuser> The class drivers are the more friendly ones with it all wrapped up
[16:12:14] <RikusW> like CDC
[16:12:21] <abcminiuser> The low level versions are there as a base if you need them, but most people want the class driver APIs
[16:12:23] <Badaboom> twnqx: i did a few port repairs and that was the payment,, my friend had several around due to color changeout
[16:12:33] <abcminiuser> Which reminds me, I should add that to the getting started page
[16:12:44] <Badaboom> twnqx: also comes with the touch element
[16:13:01] <twnqx> cool
[16:13:02] <Badaboom> so there brand spaking new
[16:13:22] <Badaboom> - the dust collecting on them:)
[16:13:36] <abcminiuser> Double commas?
[16:13:40] <twnqx> for which phone?
[16:13:42] <Badaboom> no,, i have them wrapped up
[16:13:48] <Badaboom> samsung focus2
[16:13:56] <Badaboom> at&t
[16:14:51] <megal0maniac> Got it :) So essentially, you want API level unless you know that you want low level :)
[16:15:19] <Badaboom> twnqx: ive seen where ppl have used them with avr's but no details
[16:15:42] <twnqx> linky?
[16:15:51] <Badaboom> yup,, one sec
[16:16:35] <abcminiuser> megal0maniac, how about this added to the class driver project descriptions: "This demo uses the user-friendly USB Class Driver APIs to provide a simple, abstracted interface into the USB stack."
[16:16:49] <Badaboom> i never bm it so i have to search the blackberry lcd i have and then it shows the others,,lol
[16:17:03] <abcminiuser> For the low level demos "This demo uses the low level LUFA APIs to manually implement the USB class functionality."
[16:19:14] <megal0maniac> Sounds good. As long as it points less-advanced users in the API direction. Those who want to dabble with low-level USB stuff probably won't have to read the description
[16:20:51] <Badaboom> twnqx: This is the guy ive been trying to email,, only prob is of course he wants $$ http://www.flickr.com/photos/maan2k
[16:21:28] <Badaboom> look at all of his toys:)
[16:21:56] <abcminiuser> Right, I'm off to bed
[16:21:57] <abcminiuser> Night all
[16:22:07] <megal0maniac> 'Night
[16:22:17] <OndraSter_> gn
[16:22:38] <Badaboom> gn
[16:23:06] <Badaboom> twnqx: notice the LCD hackboard?
[16:23:12] <twnqx> yes
[16:24:07] <Badaboom> I think hes probably a good bet,, but then again $$:(
[16:25:00] <megal0maniac> I can has segfault :(
[16:27:48] * megal0maniac sleeps
[16:31:25] <twnqx> Badaboom: do you have a part number for your displays?
[16:31:50] <Badaboom> yeah,, lemme open one,, i was taking pics for u,,lol
[16:31:56] <twnqx> :)
[16:32:28] <Badaboom> you know what,, its sealed bihind the casing,,,hmm
[16:32:56] <Badaboom> ok,, surgery time
[16:34:02] <Badaboom> lmao,, this could be tricky
[16:40:44] <Badaboom> http://tinypic.com/view.php?pic=30vj0a1&s=5
[16:41:01] <Badaboom> I think ill attempt it later,, it needs heat to seperate it
[16:41:16] <Badaboom> that or risk breaking the glass
[16:41:23] <Badaboom> which ,, no
[16:42:26] <twnqx> uh.
[16:43:14] <Badaboom> see the black where it meets the white?
[16:43:22] <twnqx> yeah
[16:43:32] <Badaboom> that comes off and the screen is under it
[16:43:57] <Badaboom> but its sealed on there with a sort of double sided tape
[16:44:07] <Badaboom> heat is the best method
[16:44:11] <Badaboom> and patients
[16:45:05] <Badaboom> Im sure the part# is under there:(
[16:48:13] <Badaboom> Dont worry,, im gonna seperate the screens when i find data on them,, it actually comes out from the fron now that i recall
[16:48:29] <Badaboom> then the glass seperates from the digitizer
[16:49:15] <Badaboom> front
[16:50:45] <Badaboom> twnqx: heres another screen i have thats out,, a little smaller but ive removed it so u can see what im talking about
[16:51:05] <Badaboom> lemme photo it
[16:53:30] <beaky> hello
[16:54:03] <OndraSter_> heloo
[17:01:21] <Badaboom> twnqx: so it looks something like this when extracted http://tinypic.com/view.php?pic=2ikxyck&s=5
[17:02:49] <Badaboom> twnqx: numbers on the back of that are 31115P00152FE
[17:05:19] <Badaboom> 3 other numbers all come up with 0 results
[17:05:49] <Badaboom> anyways,, i have alot:)
[17:07:00] <twnqx> hm
[17:07:57] <Badaboom> and i have access to many more,, some newer that the focus2
[17:11:06] <Badaboom> seems like samsung is very guarded with there data/pinouts
[17:14:25] <beaky> what is a good name for a function that operates on a led?
[17:14:37] <twnqx> operate_on_led
[17:17:33] <RikusW> led_xxx
[17:18:59] <Badaboom> twnqx: i may have my friend ask his distb if he can get ahold af data,, doubtful but ill try
[17:19:26] <twnqx> set_led_state
[17:19:38] <beaky> ah, set_led_state seems good
[17:19:59] <beaky> naming is the hardest part of programming
[17:20:35] <Badaboom> lol,, i saw a video of where a guy was naming with all kinds of crazy things
[17:20:50] <Badaboom> like bipolarmood
[17:22:01] <Badaboom> Gonna ger supper started,, bbl
[17:22:05] <Badaboom> get
[17:22:48] <twnqx> gonna head to bed soon :P
[17:26:34] <Badaboom> ok,, cya tomorrow
[17:52:08] <beaky> How do i make my program smaller/
[17:52:16] <beaky> atm it is 838 bytes
[17:52:30] <twnqx> does it fit?
[17:52:51] <OndraSter_> that's what she said, twnqx
[17:52:52] <OndraSter_> :P
[17:53:01] <beaky> it does
[17:53:06] <twnqx> so why do you care
[17:54:06] <beaky> my avr only has like 1KB of SRAM
[17:54:19] <beaky> so I guess I should offload some bits to progmem
[17:56:00] <OndraSter_> wait, RAM
[17:56:20] <twnqx> your program doesn't run from ram
[17:56:21] <OndraSter_> your .text or .data segment is 838B?
[17:56:36] <twnqx> lol
[17:56:39] <OndraSter_> overflowed
[17:56:42] <twnqx> there she goes
[20:52:55] <vectory> lol, _delay_ms expects integer constant in my avr-gcc, but not in AS 6?
[20:53:39] <vectory> thought AS 6 uses gcc, too. might have set it up to use the atmel compiler instead?
[20:53:52] <Tom_itx> it does but it's a newer one
[20:54:04] <Tom_itx> you can actually download it separately
[20:56:22] <vectory> im not sure i understand. AS 6 likely doesnt use gcc in the version my school has installed?
[20:56:49] <vectory> or is it that a newer gcc supports _delay with variables?
[20:57:03] <vectory> a) or b)? :D
[20:57:31] <Tom_itx> 6 uses gcc and it's a newer version
[20:57:38] <Tom_itx> you can download it separately
[20:57:43] <Tom_itx> or bundled
[20:57:55] <Tom_itx> i dunno about changes to delay
[21:03:36] <vectory> hmmmm, alright. ubuntu obviously suffers from outdated packages :/