Back
[00:59:32] <Flipp_> does the avr-gcc flag -Os optimize away uncalled functions like O2/3 would?
[01:03:20] <Casper> it should, afaik any optimisation level does, and I think even unoptimised one does too
[01:16:25] <aandrew> optimizing away functions/data is the linker's responsibility (with the compiler's help)
[01:16:50] <aandrew> -ffunction-sections and -fdata-sections for the compiler and -Wl,--gc-sections for the linker IIRC
[01:47:19] <Casper> hmmm
[01:47:27] <Casper> will have to redo some tests...
[01:47:43] <Casper> maybe the Makefile I used when I did my test back then did it...
[01:47:57] <Casper> but that might explain why some of the code I did was bigger than I expected..
[03:10:38] <Flipp_> do most I2C devices support a WHO_AM_I command? and if so, is it "standardized" across components?
[03:53:14] <DO9XE> Flipp_, I never heard of such a thing.
[03:53:53] <DO9XE> In I²C you have to know, to who you are talking
[03:54:00] <Flipp_> DO9XE: ah, okay. It's something I found in a ST datasheet for a gyro. response to I2C command
[03:54:29] <Flipp_> yeah. I can figure out what's addresses have things hanging off the bus, just don't know if there was a way to tell what was where
[03:54:48] <Flipp_> (other than knowing the chips beforehand)
[03:54:53] <Flipp_> just curious
[04:35:49] <less> Hi
[04:36:00] <less> I attached external crystal to obtain a 16mhz clock speed
[04:36:32] <less> and example in datasheet of atmega328p doesnt work with two capacitors 22pF
[04:36:53] <less> If I use only crystal without capacitors and connecting it to GND it works
[04:37:10] <less> and without connecting crystal to the GND it works *
[04:37:27] <less> someone could explain me why it is like this ?
[04:44:42] <less> prototype board capacitance ?
[04:56:41] <Tom_itx> less, very likely yes
[04:58:44] <malinus> less: I've used a 16mhz crystal with my atmega328p on a prototype boards, a lot of times. Never had those problems. Not sure if this is helpfull :)
[04:59:55] <less> malinus: not really :) and you used datasheet configuration for crystal ?
[05:00:03] <malinus> Yes, exactly.
[05:00:48] <less> if I connect capacitors and connect it to GND it doesnt work :)
[05:01:21] <less> So I have special abilities or i have luck of beginners :)
[05:03:40] <Tom_itx> you certain the values of the capacitors?
[05:04:44] <malinus> less: yeah double check the values with the datasheet, and with the values on the caps. It's easy to read them wrong
[05:06:18] <Tekkkz> Hello!
[05:07:17] <Tekkkz> main.c :
http://ix.io/hc3 main.h :
http://ix.io/hc4 => main.c:180 the if condition istn true (led doesnt become high) but why? any ideas?
[05:10:35] <rue_bed> Tekkkz == or =
[05:11:13] <Tekkkz> huh? what do you mean?
[05:11:26] <rue_bed> I cant tell line numbers
[05:11:39] <Tekkkz> ahh yes wait im putting it into pastebin ok?
[05:11:46] <rue_bed> could I get you to post it on a code site like codepad?
[05:11:51] <Tekkkz> y
[05:11:53] <rue_bed> k
[05:12:37] <Tekkkz> main.c :
http://codepad.org/hTe3hxwT
[05:12:42] <rue_bed> your not writing grbl are you?
[05:12:52] <Tekkkz> grbl?
[05:12:59] <Tekkkz> main.h
http://codepad.org/sl3ZY2b0
[05:13:23] <Tekkkz> ahh i see
[05:13:27] <Tekkkz> no, jut my own
[05:13:41] <rue_bed> heh
[05:14:03] <Tekkkz> so now the line number plss?
[05:14:06] <rue_bed> I wish you would write an open source cam program instead :)
[05:14:14] <rue_bed> yup, so
[05:14:35] <rue_bed> what data type is program_read_byte
[05:14:46] <Tekkkz> program_read_byte?
[05:14:46] <rue_bed> pgm....
[05:14:50] <Tekkkz> you mean pgm?
[05:14:58] <rue_bed> pgm_read_byte(&x_array[run])
[05:15:11] <Tekkkz> look main.h
[05:15:13] <rue_bed> does that return an unsigned int or something else
[05:15:19] <Tekkkz> const uint32_t PROGMEM x_array[360]
[05:15:37] <Tekkkz> yes, unsigned int
[05:15:45] <less> malinus: 22pF capacitors like in datasheet
[05:15:59] <rue_bed> read byte
[05:16:08] <rue_bed> you know a byte is 8 bits?
[05:16:15] <Tekkkz> yes
[05:16:17] <Tekkkz> ..
[05:16:19] <Tekkkz> yayaya
[05:16:24] <Tekkkz> so i need to use read_ähh
[05:16:58] <rue_bed> I'm not familiar with the function, but seems to me that would be a problem
[05:17:36] <Tekkkz> ahh
[05:17:44] <Tekkkz> uint32_t pgm_read_dword 4
[05:17:50] <Tekkkz> i need to read a dword
[05:17:55] <rue_bed> that sounds better
[05:18:11] <rue_bed> its 3am, good luck/night
[05:18:19] <Tekkkz> man that i didnt payed attention that im just returning a byte..
[05:18:25] <Tekkkz> yeah,ty, byebye good night
[05:32:47] <less> IS there a way that my clock is working with 16mhz ?
[05:32:51] <less> to check*
[05:33:11] <less> I have 16mhz crystal connected
[05:34:19] <less> generaly spekaing lcd is working leds also but I cannot force my tsop4836 to work
[05:34:28] <less> using this lib
https://github.com/pinkeen/avr-rc5
[05:37:03] <rue_bed> flash an led
[05:37:14] <_abc_> Guys, where is avr-gcc writing to program memory documented, if at all? As in, writing to arrays in program memory?
[05:37:28] <rue_bed> set it to internal clock to get an idea, then change to the crystal and see if its about 16x faster
[05:37:59] <rue_bed> progmem?
[05:38:00] <less> rue_bed: I have led connnected
[05:38:03] <_abc_> As in, SPM based internal flash memory writing, for atmega specifically?
[05:38:17] <_abc_> I would like to to not reinvent hot water. Been googling for a while now.
[05:39:30] <less> rue_bed: This lib is for 18mhz
[05:39:40] <less> 16mhz*
[05:41:15] <_abc_> Anyone? Is there boilerplate code to write SPM based flash arrays in C?
[05:41:36] <_abc_> I do not see PROGMEM segment marked read only so maybe it is implicit?
[05:42:03] <rue_bed> you want to load a table into progmem to read out runtime?
[05:42:19] <_abc_> I want to modify some tables at runtime yes.
[05:42:36] <_abc_> Largeish tables (think fonts and the like)
[05:42:36] <rue_bed> I do not suggest doing that in progmem then
[05:42:48] <_abc_> Well, can it be done? It will be done rarely
[05:43:03] <rue_bed> self flashing
[05:43:16] <_abc_> Yes, but that is usually banished into the boot loader.
[05:43:40] <_abc_> Could I simply borrow the code from there and take care I don't try to write to the segment page where the code resides?
[05:43:59] <_abc_> It's ok to turn off interrupts and the like when doing this so that is not a problem
[05:44:45] <_abc_> http://www.nongnu.org/avr-libc/user-manual/FAQ.html been googling for some time now. Nothing. Weird.
[05:44:54] <_abc_> Surely I am not the only one who came up with this need
[05:45:17] <rue_bed> its out there, I dont know it tho
[05:45:30] <_abc_> Also how does one align a vector at an address or page boundary? Can one force its address in flash after making sure some larger vector declared on purpose there overlaps it?
[05:45:43] <_abc_> rue_bed: Okay, thanks. Any clues on what names to search for?
[05:46:34] <_abc_> Meanwhile, found yesterday, interesting and useful: wikidevi.com
[05:48:07] <_abc_> Also, I found an avr-gcc bug which may be older: uint8_t x; uint16_t y; x = 0; y = (x << 8); /* -> y = 0x0100; */ y = (x << 7); y <<= 1; /* or any other combination: ok */
[05:48:26] <_abc_> So, careful about left shifting zero arguments by 8 bits.
[05:48:43] <_abc_> I tried all the usual forced casts, same outcome.
[05:49:18] <_abc_> I assume someone missed the non integer (i.e. unsigned) sign extension and a sign bit or carry got shifted there.
[05:49:57] <_abc_> Right shifts are always okay, the left shift seems to only be wrong with 0 argument.
[05:59:39] <_abc_> No comment? ;)
[06:05:08] <_abc_> http://www.avrfreaks.net/forum/write-flash-memory-atmega128 I think I found what I was looking for wrt SPM
[06:05:13] <_abc_> Will need to experiment a bit.
[06:05:35] <_abc_> Technically there is no reason not to be able to call boot loader segment located routines from normal code, right?
[06:08:05] <Xark> _abc_: I think it is required to "self-flash".
[06:09:07] <_abc_> http://avr.2057.n7.nabble.com/Sharing-code-between-a-bootloader-and-an-application-td9381.html self-flash is obvious, the questions are when, how, how much ;)
[06:54:20] <_abc_> http://oneweekwonder.blogspot.ro/2014/07/bootjacker-amazing-avr-bootloader-hack.html clever hack.
[06:56:38] <_abc_> I have to admit the guy did a clever thing.
[06:56:40] <_abc_> +++
[06:58:24] <Xark> _abc_: Yeah, neat. I remember hearing about this (but not in detail),.
[07:01:57] <_abc_> Ugh. How many devices did you people OMIT to set up to summer time today. I am still finding some...
[07:02:19] <_abc_> Fucking hell someone should remove this obnoxious idiocy from the world.
[07:02:33] <_abc_> Maybe it made sense in Ben Franklin's time, does not make so much sense now
[07:07:02] <_abc_> ##electronics has become a cesspool of #arduino idiots lately. I wonder how come they did not find THIS channel...
[07:13:02] <Mr_Sheesh> _abc_ - I'm thinking "Why not have all your house clocks talk either using IR or using something like a cheap nRF24L01+ so you have one master clock and the other clocks are auto set by the (GPS or 'Net connected, time set via USB?) master clock? But agreed, we don't really gain anything but pain ATM from DST
[07:13:34] <_abc_> I think it was a no brainer when the only clock to set was the town's city hall tower clock.
[07:13:45] <_abc_> NOW you have clocks where you don't even think about them.
[07:14:10] <_abc_> And NO I DO NOT WANT THEM NETWORK AUTOMATED, SOME NEED TO RUN ON DIFFERENT TIME, LIKE SIDEREAL OR DIFFERENT TIME ZONE.
[07:16:26] * Xark still likes "leap seconds" though - don't want Earth time to be totally make believe (vs solar system "clock")...
[07:17:07] <_abc_> I would prefer the leap seconds to be dithered in subtly, perhaps into the last second of the year every year. Leap anything makes me cringe.
[07:17:17] <_abc_> C code to handle leap years in microcontrollers is a PAIN.
[07:17:34] <_abc_> And that's just part of it.
[08:27:17] <_abc_> http://www.dailymail.co.uk/news/article-3016648/Big-brand-theory-professors-Brian-Cox-Stephen-Hawking-trademark-names-turn-brands.html good move. Idiotic companies always try to capitalize on person's names and it takes ages to ban that in court. Should be a granted right, the right to one's own name, I think.
[08:27:34] <_abc_> And no you can't trademark Peter. I did it already!
[08:27:37] <_abc_> <smirk>
[08:31:06] <Lambda_Aurigae> I want to copyright my likeness and all derivatives.
[08:31:20] <Lambda_Aurigae> then do something newsworthy.
[08:31:34] <Lambda_Aurigae> so the news has to pay me to show my likeness on tv.
[08:36:27] <_abc_> Do you resemble a pink unicorn?
[08:36:38] <Lambda_Aurigae> no
[08:37:28] <_abc_> In Europe people have a right to their own image, and, in particular, children have it automatically whereas adults may be asked for permission. This results in people's faces being blotted out and children's faces always being so in news articles.
[08:37:39] <_abc_> Makes perfect sense in a way.
[08:37:42] <_abc_> I'm in Europe.
[08:38:03] <Lambda_Aurigae> children are usually blotted out here if they were involved in something
[08:38:16] <_abc_> The most recent example is that of the Germanwings killer pilot, whose face was blotted out in news pics in German media.
[08:38:18] <Lambda_Aurigae> but the news can film randomly downtown and show everybody, adult and child.
[08:38:30] <_abc_> Don't know about that.
[08:38:54] <_abc_> I assume with HD filming the law will catch up as one can extract stills of very high resolution from a HD material.
[08:39:04] <_abc_> Lambda_Aurigae: Where is here in your case? USA=
[08:39:05] <_abc_> ?
[08:39:15] <Lambda_Aurigae> yeah.
[08:39:28] <Lambda_Aurigae> pretty close to the geographic center of the USA in fact.
[08:40:37] <_abc_> I have no idea where that is. I think someone from USA once said he's in a "flyover state". Cough.
[08:41:16] <_abc_> Kansas-Nebraska says Wikipedia.
[08:42:28] <Lambda_Aurigae> I am 400 miles from the geographic center of the USA.
[08:42:47] <_abc_> http://www.nytimes.com/2008/06/02/us/02land.html?pagewanted=all&_r=0 nice one
[08:43:30] <Lambda_Aurigae> yeah, but not the geographic center of the lower 48 states.
[08:43:54] <Lambda_Aurigae> https://www.google.com/search?client=ubuntu&channel=fs&q=geographic+center+of+the+USA&ie=utf-8&oe=utf-8
[08:44:50] <Lambda_Aurigae> I am at Latitude 42.173310 and longitude -93.093568
[08:44:59] <_abc_> Did they eventually find out what caused the crash of the Air Indonesia Surabaya-Singapore airbus? Was reading news on the Air Canada plane which "arrived" in Halifax today. 24 in hospital. Seems the Indonesia one was lost when a flight computer was rebooted in error (rebooting the one which worked instead of the busted one). Just another rumor. It's annoying how long it takes to find out these things.
[08:45:08] <_abc_> Computers kill! Fact!
[08:45:22] <Lambda_Aurigae> http://www.latlong.net/c/?lat=42.173310&long=-93.093568
[08:45:38] <_abc_> Some airbus people seem to refer to their killer as HAL. Should be Clementine I think (Stephen King's murderous car)
[08:46:03] <Lambda_Aurigae> HAL was only defending himself!
[08:46:18] <Lambda_Aurigae> poor misunderstood computer..
[08:47:02] <_abc_> Looks like you're in the boonies to me. Well just from pop. density.
[08:47:04] <Lambda_Aurigae> and the car is Christine
[08:47:15] <_abc_> Ah yes.
[08:47:15] <Lambda_Aurigae> yeah...I am in the middle of cornfields.
[08:47:50] * _abc_ keeps zooming out trying to find a town name he can relate to ...
[08:47:59] <Lambda_Aurigae> Des Moines is southwest of us.
[08:48:17] <_abc_> Iowa?
[08:48:21] <Lambda_Aurigae> yeah.
[08:48:38] <Lambda_Aurigae> nearest "town" is Union...about 500 people I think total.
[08:48:52] <Lambda_Aurigae> nearest "city" is marshalltown, about 35000 people..and half an hour away.
[08:50:03] <_abc_> Lambda_Aurigae: I wonder how you do your shopping? Weekly? Big fridge etc and shop at some large grocery store or do you go for self sufficient style, with farm products etc.
[08:50:15] <Lambda_Aurigae> big fridge and freezer..
[08:50:19] <Lambda_Aurigae> once a week groceries usually.
[08:50:29] <Lambda_Aurigae> but we do have a pretty good size garden.
[08:50:53] <_abc_> Interesting. I seem to see very few "rural" people from USA online.
[08:51:09] <_abc_> What's the pop. percentile living in rural areas in USA now? 20%?
[08:51:22] <Lambda_Aurigae> no clue.
[08:51:27] <Lambda_Aurigae> I doubt it's that high.
[08:51:46] <Lambda_Aurigae> https://www.google.com/maps/place/1729+Marble+Rd,+Union,+IA+50258/@42.1732963,-93.0935678,820m/data=!3m2!1e3!4b1!4m2!3m1!1s0x87efb535b9207049:0xa224e959427fd913
[08:52:08] <Fleck> chill _abc_ :D
[08:52:10] <Lambda_Aurigae> google maps of my place...I'm right at that marker...trees above and left of the marker.
[08:52:44] <_abc_> http://www.nationmaster.com/country-info/stats/People/Percentage-living-in-urban-areas Lambda_Aurigae
[08:52:47] <Lambda_Aurigae> my garden is north on 3rd stree there..up at the farm at the end of 3rd street it used to be...now it's just at the trees just north and east of my house.
[08:52:55] <_abc_> Fleck: ?
[08:53:12] <_abc_> It is 20% rural 80% urban for USA as I vaguely remembered it
[08:53:19] <Fleck> about arduinos, idiots and channels! :D
[08:53:44] <Lambda_Aurigae> in fact, you can see our old garden in the satellite view up there.
[08:53:53] <_abc_> Fleck: Hey, I could invite every arduino question asker from ##electronics to #avr. After all, they use avr's no?
[08:54:10] <Fleck> thats not the point _abc_
[08:54:23] <Lambda_Aurigae> _abc_, not necessarily anymore...
[08:54:32] <Lambda_Aurigae> a lot of new ardweenies are not avr anymore.
[08:54:43] <Lambda_Aurigae> heck, they have even ported ardweeny to the esp8266 module!
[08:54:52] * _abc_ wants to start a new channel for electronics called ##oniu with topic "##oniu is not oniudra"
[08:55:53] <_abc_> I have predicted that the ages of robots will come when the imlementation of artificial stupidity will become possible. And now we have arduino. The 8086 of embedded. The worst parts of programming and architecture rolled into a never-get-rid of grandfathered product.
[08:56:03] <_abc_> +p
[08:57:47] <_abc_> I have 2 major gripes against A*: 1) complete lack of port protection on hardware, even a 220R resistor on each pin would have saved a lot of duinos from smoky death - good for sales though - and shameless lies in the spec sheet such as 'each duino io pin can drive 40mA' 2) software which a) uses CPP in embedded b) abstracts hardware to the point where point and click "programming" interfaces are simpler.
[08:58:19] <_abc_> So who's for ##oniu ? ;)
[08:58:22] <Lambda_Aurigae> _abc_, there is now a "visual" programming interface for them.
[08:58:35] <Lambda_Aurigae> I'm there..
[08:58:36] <Lambda_Aurigae> hehe.
[08:58:41] <_abc_> I know. There are various other things which are much better for visual programming, including a nice UML based one.
[08:59:10] <_abc_> The latter called microuml. My link does not open now. F*
[08:59:17] <Lambda_Aurigae> hmm.
[08:59:23] <Lambda_Aurigae> I kinda learned uml at one job I had.
[08:59:28] <_abc_> microuml.net
[08:59:29] <Lambda_Aurigae> never got into it deep enough.
[08:59:43] <_abc_> I played with Petri nets and the like.
[08:59:59] <Lambda_Aurigae> company I worked for was doing software development for the telecom industry.
[09:00:02] <_abc_> It's cute but after you lay out the framework you are back to editing source as usual. Can't abstract real hardware enough for this.
[09:00:08] <Lambda_Aurigae> we had UML diagrams everywhere on the walls..
[09:00:24] <Lambda_Aurigae> we would do some round-robin work turning UML into code and versa-visa.
[09:00:42] <_abc_> Well UML is about the only nice formal way to do embedded. There should also be a way to compile behavioral and structural Verilog/VHDL into working code and that may yet come up.
[09:01:09] <Lambda_Aurigae> I don't see any microuml or micro-uml anywhere.
[09:01:11] <_abc_> Petri nets are closely related to UML and structured logic diagrams
[09:01:47] <_abc_> https://duckduckgo.com/q=microuml
[09:02:03] <_abc_> 2nd link is arduino modelling in microuml. The website is down probably for maintenance.
[09:02:22] <_abc_> Last seen last week. Try archive.org ?
[09:02:33] <Lambda_Aurigae> hehe.
[09:02:38] <Lambda_Aurigae> I'm kind of old fashioned.
[09:02:50] <_abc_> Looks like a data center related problem, dns microuml.net does not resolve
[09:02:57] <Lambda_Aurigae> heck, I still use full 8bit binary representation instead of the << operator usually.
[09:03:59] <_abc_> https://web.archive.org/web/*/http://www.microuml.net here
[09:04:34] <Lambda_Aurigae> like,, #define PB1_ON = 0b00000001 #define PB1_OFF = 0b11111110
[09:04:36] <_abc_> Well I'm an older guy but I'm the hex kind of old guy not the kind of octal old guy. Does that date me? ;)
[09:04:48] <Lambda_Aurigae> I do hex and octal too.
[09:04:57] <_abc_> Then you're older than I am...
[09:05:04] <Lambda_Aurigae> 47.
[09:05:20] <_abc_> Oh. Pretty close. Different backgrounds then.
[09:05:25] <Tom_itx> ole farts
[09:05:32] <Lambda_Aurigae> I learned 6502 assembly on a vic-20 with vicmon cartridge
[09:05:38] <Lambda_Aurigae> mmmm...oles
[09:05:50] <Lambda_Aurigae> like tater tots only 1/3 height.
[09:05:52] <_abc_> https://web.archive.org/web/20141220035714/http://microuml.net/microuml.html here. Looks nice and seems to work too.
[09:06:10] <_abc_> I hope it's a network/dns/domain renewal snafu and not some lawyer scum takedown
[09:06:25] <Lambda_Aurigae> uml is just digital flowcharts.
[09:06:28] <_abc_> I learned 8085 and Z80 asm 1st and some Basic.
[09:07:03] <Lambda_Aurigae> I started with basic on trs-80 then apple then vic...then 6502 assembly on vic and apple..then pascal on apple...then went on from there.
[09:07:12] <_abc_> Also their microuml editor Papyrus is an Eclipse app.
[09:07:29] <_abc_> I skipped Pascal and went on from asm to C.
[09:07:39] <Lambda_Aurigae> pascal was taught in our highschool...
[09:07:50] <Lambda_Aurigae> half a year of basic..half a year of pascal...in my junior year.
[09:07:59] <Lambda_Aurigae> I had the entire class finished in the first quarter.
[09:08:09] <Lambda_Aurigae> knew more than the teacher did.
[09:08:11] <_abc_> While spending several years in a shithole country, duped by my relatives to go there and take a degree. Which I did not. So I did a lot of things backwards, working through a career in between.
[09:08:49] <_abc_> Personally I do not like Pascal. And VHDL, for the same reason. Too academic.
[09:08:58] <twnqx> verilog > vhdl :P
[09:09:07] <twnqx> and well, pascal is a modula-2 clone
[09:09:13] <Lambda_Aurigae> have yet to learn verilog or vhdl.
[09:09:16] <twnqx> which is academic by background
[09:09:20] <_abc_> Weirdly, I once chewed myself through a Cobol (!) course book ~500 pages, which left me with a taste for clearly labelled sections in C...
[09:09:25] <Lambda_Aurigae> we learned fortran and ada in the military.
[09:09:38] <Lambda_Aurigae> I learned enough cobol to rework some mainframe code once.
[09:09:42] <malinus> I never got all that arduino hate. Isn't it just some elitist bullshit. "It shouldn't be that easy"? Why not let people/kids play with their toys?
[09:10:17] <Lambda_Aurigae> malinus, let them play...enjoy, have fun....but don't ask me questions about the massive bloatware that is the arduino libs.
[09:10:24] <_abc_> I learned very little Fortran, just enough to be dangerous and to hack about 20 lines of Spice3f4 or such to compile with gnu f2c
[09:10:47] <Lambda_Aurigae> we used fortran for doing satellite tracking math stuff in the military.
[09:10:52] <malinus> Lambda_Aurigae: but they never do, they go to #arduino for that :P
[09:11:01] <twnqx> malinus: it's like wasting three quarters of the whimsy performance the avr has for abstraction layers
[09:11:07] <Lambda_Aurigae> malinus, ummm...no,,,they don't...at least, not all of them.
[09:11:22] <_abc_> malinus: To me, having learned how to do things more or less right, it's the equivalent of a bunch of punks breaking into a concert hall in mid concert, pull out some makeshift drums and pipes, and joining in the music while screaming and shouting.
[09:12:08] <Lambda_Aurigae> _abc_, I compare it more to someone playing a CD of a fiddle vs playing the fiddle.
[09:12:24] <Lambda_Aurigae> yeah, it's easier, but you don't learn how the fiddle actually works.
[09:12:31] <_abc_> They do not have 16 bit brains. More like playing a CD through a GSM phone...
[09:12:40] <malinus> lol'd at that comparison _abc_
[09:13:34] <Lambda_Aurigae> question on the news..."What makes you get up and go to the office each Monday?"
[09:13:44] <_abc_> Lack of moneyß
[09:13:46] <_abc_> ?
[09:13:47] <malinus> twnqx: you must really hate people who buy expensive new notebooks to browse facebook and check mails :D
[09:14:00] <twnqx> i do
[09:14:01] <Lambda_Aurigae> my answer, "My job allows me to afford a few luxuries that I've grown accustomed to. Like, sleeping indoors and eating."
[09:14:04] <malinus> twnqx: :P
[09:14:06] <_abc_> malinus: Worse, some buy iShiny products for that. Eww,
[09:14:17] <Lambda_Aurigae> _abc_, iCrap
[09:14:23] <twnqx> i hate stupidity, i feel physical pain observing stupid people
[09:14:26] <_abc_> Lambda_Aurigae: The sleeping indoors one is good, I'll remember that
[09:14:29] <_abc_> brb
[09:14:32] <malinus> those are mostly accessories though (like expensive jeans or a watch)
[09:16:59] <_abc_> Lambda_Aurigae: One has to admit that from the seller's point of view, well broken in fanboi buyers who pay what one asks for a rounded cornered object are a bounty.
[09:17:15] <Lambda_Aurigae> yes.
[09:17:16] <_abc_> I mean crack dealers would wish to be able to keep their clients over several years.
[09:17:27] <Lambda_Aurigae> apple is living proof that idiots will buy anything shiny.
[09:17:43] <_abc_> Yeah the iShiny is more suggestive that iFruit.
[09:18:32] <_abc_> Latest iBfanboi test I saw in a paper was about recognizing the Apple logo from about 9 versions, apple hole facing left or right, leaf facing left or right. I'm proud to say I failed it.
[09:18:35] <Lambda_Aurigae> tried to install mac os on vmware last weekend here to test something..
[09:18:45] <Lambda_Aurigae> it won't install...because I have an AMD processor.
[09:19:38] <malinus> greaHardwareSupportOS.png
[09:19:51] <_abc_> Last time I touched iShiny stuff was on a colorful cube which I fixed electronically, it had a busted transistor in the monitor unit. I tried to put Yellowdog Linux on it. Gave up since the monitor died again. All Apple hw since Apple ][ shares one feature: It's pretty much impossible to take apart or work on.
[09:20:11] <Lambda_Aurigae> take apart...easy
[09:20:17] <Lambda_Aurigae> get back together working,,,more difficult.
[09:20:26] <Lambda_Aurigae> I can take ANYTHING apart.
[09:20:37] <Lambda_Aurigae> just can't always get it back together in working order.
[09:20:39] <_abc_> Oh don't say that. Did you see the latest laptops and power supplies? All glued shut and gooped full?
[09:20:59] <Lambda_Aurigae> just need the right hammer and/or solvent
[09:21:09] <_abc_> I worked for >12 years on camcorders and digital cameras, cramped assemblies and hairy 30 page disassemblies do not scare me.
[09:21:20] <_abc_> Ah yes, the solvent part sucks
[09:21:43] <Lambda_Aurigae> but with enough fuming nitric acid, nothing apple has can stand up before me!
[09:21:56] <_abc_> I once changed the CFL lamps in a laptop for someone I knew. It was glued shut. I had to force it open. When done, the LCD bezel looked like it had been run over by a tank but it worked.
[09:22:10] <Lambda_Aurigae> yup..been there
[09:22:27] <_abc_> They deliberately make it so you have to replace exterior cosmetic parts after the repair.
[09:22:41] <_abc_> This ensures proper repairs can only be done by an authorized place which gets the parts
[09:22:55] <_abc_> Camera makers do it exactly like that, I know all about that.
[09:23:10] <Lambda_Aurigae> these days most of that gear is designed to be disposable.
[09:23:18] <Lambda_Aurigae> not made to be worked on.
[09:23:26] <Lambda_Aurigae> hell, cars are going that route too it seems.
[09:23:40] <_abc_> It's deliberately done so, Lambda_Aurigae
[09:23:45] <Lambda_Aurigae> yup.
[09:24:14] <Lambda_Aurigae> along with designed obsolescence
[09:24:19] <_abc_> Worse, I live in a one diskette locust country and I have to make my projects copy proof. I've gotten to use microprocessors for everything for that reason alone. Blink led? mcu!
[09:24:28] <Lambda_Aurigae> build hardware designed to fail after a certain amount of time.
[09:24:35] <_abc_> One diskette country is one where you need only one copy of any software for the entire country.
[09:25:05] <_abc_> Ah yes you mean like inkjet printers which won't work after printing 10,000 pages because they "know" the waste ink sponge is "full"?
[09:25:13] <Lambda_Aurigae> yup.
[09:25:31] <Lambda_Aurigae> or, car engine mounts designed to fail after so many hours of operation.
[09:25:42] <_abc_> That I have not seen.
[09:25:49] <Lambda_Aurigae> I worked with a guy who was on a design team in the 70s who did just that.
[09:25:58] <Lambda_Aurigae> the entire team got fired for a slip of the slide rule.
[09:26:10] <_abc_> The finished part worked too well?
[09:26:12] <Lambda_Aurigae> they were told to make a certain engine mount that would fail after 100K miles.
[09:26:24] <Lambda_Aurigae> GM make many many thousands of the mounts.
[09:26:33] <Lambda_Aurigae> expecting them to fail just outside warranty.
[09:26:36] <Lambda_Aurigae> they didn't fail.
[09:26:50] <Lambda_Aurigae> this one mount was used on 8 or 10 different vehicles.
[09:27:08] <_abc_> Well there were those US buses which had engines dropping out of them. Was that erring in the opposite direction?...
[09:27:26] <Lambda_Aurigae> pretty much every big manufacturing company on the planet does this.
[09:27:34] <Lambda_Aurigae> probably was.
[09:27:45] <_abc_> Just as a curiosity, are military vehicles also affected?
[09:27:54] <Lambda_Aurigae> hell, copiers could be built to run a million copies between maintenance.
[09:28:04] <Lambda_Aurigae> but they are built with substandard parts that wear out and fail.
[09:28:12] <_abc_> Wait, they WERE built like that. Laserjet 4?
[09:28:28] <Lambda_Aurigae> _abc_, of course they are...they are just milspec which means they pass higher standards of testing.
[09:28:39] <_abc_> The fun part is LJ4s used plastic gears etc. Which proves long lived devices can be made with such parts!
[09:28:42] <Lambda_Aurigae> laserjet 5si...best laser printer ever.
[09:28:49] <_abc_> Yeah or 5
[09:28:52] <Lambda_Aurigae> everything since then has been crap.
[09:29:03] <Lambda_Aurigae> even the 8100 and 8150 which were the same frame as the 5si.
[09:29:11] <_abc_> Don't know those.
[09:29:13] <Lambda_Aurigae> quality of parts went downhill.
[09:29:37] <Lambda_Aurigae> they look exactly like the 5si...same exact frame...different controller and a few different sensors.
[09:30:25] <_abc_> 5si was post script without a cartridge?
[09:30:30] <Lambda_Aurigae> we have a xerox high end copier we sell that is designed to run a million copies between maintenance cycles.
[09:30:38] <_abc_> Oh that's a big one. 4 is small.
[09:30:41] <Lambda_Aurigae> why can't those same rollers and parts be used in the smaller ones?
[09:31:03] <Lambda_Aurigae> 5si has postscript without cartridge...I think it is actually a simm stick in the machine.
[09:31:24] <Lambda_Aurigae> yeah...we have LJ 4 machines in the field still..
[09:31:40] <Lambda_Aurigae> but fewer and fewer every month.
[09:32:18] <Lambda_Aurigae> operation cost of HP laserjets is horrid.
[09:32:22] <Lambda_Aurigae> not as bad as inkjets
[09:32:39] <Lambda_Aurigae> but still much higher than most xerox devices.
[09:32:58] <_abc_> Xerox no longer makes anything low end in USA right?
[09:33:04] <Lambda_Aurigae> cost per page can be a killer.
[09:33:08] <_abc_> Xerox uses Samsung engines for that
[09:33:10] <Lambda_Aurigae> xerox no longer makes anything in the USA.
[09:33:25] <Lambda_Aurigae> even though the company is based in the USA
[09:33:35] <Lambda_Aurigae> yeah....samsung for small desktop copiers.
[09:33:47] <_abc_> I know, same for everything else. I would have expected some high end and development labs to remain in USA,
[09:33:54] <_abc_> I guess the bean counters won that one.
[09:33:55] <Lambda_Aurigae> development, yes.
[09:33:58] <Lambda_Aurigae> manufacturing, no.
[09:34:35] <_abc_> Imnsho, companies which have more than 50% of manufacture or engineering offshore should be forced to a status where they are taxed as foreign companies, and do not enjoy local company privileges. All over the world.
[09:35:16] <Lambda_Aurigae> dang....LJ 4 toner cartridge works out to just over 0.03 USD per page.
[09:35:28] <_abc_> Is that a lot?
[09:35:44] <Lambda_Aurigae> we charge 0.008 USD per page under maintenance for most xerox copiers.
[09:35:59] <Lambda_Aurigae> that is horridly expensive for a laser device.
[09:36:19] <Lambda_Aurigae> inkjets usually work out to 0.20 USD for a black and 0.50 USD for a color page.
[09:36:54] <Lambda_Aurigae> xerox devices under service contract,,parts, labor, toner,,generally are 0.008 for black and 0.10 or slightly less for color.
[09:37:10] <Lambda_Aurigae> gotta go snuggle da wifey.
[09:37:12] <Lambda_Aurigae> bbiaf.
[09:58:00] <Lambda_Aurigae> now I gotta go to town
[09:58:06] <Lambda_Aurigae> weekly shopping trip.
[09:59:01] <_abc_> http://reprap.com/ wtf
[09:59:14] <_abc_> I assume maintenance?!
[09:59:18] <Lambda_Aurigae> barebones website?
[09:59:18] <Lambda_Aurigae> hehe
[10:15:14] <_abc_> http://gtalug.org/pipermail/talk/2015-March/001130.html fun
[10:15:20] <_abc_> The right site is reprap.org
[10:15:26] <_abc_> the other guys are probably squatters
[10:17:26] <_abc_> Lambda_Aurigae:
http://ineteconomics.org/institute-blog/william-lazonick-how-superstar-companies-apple-are-killing-america-s-high-tech-future iShiny related
[10:21:35] <_abc_> l8
[12:01:36] <LeoNerd> I'm sure I asked before but I'm not sure I ever got an answer I thought I was happy with: I'm looking into possible arrangements for reading a few (3 or 4?) pushbutton switches with an analog chain and a single ADC input pin
[12:02:07] <LeoNerd> Currently I have buttons on PCINT pins, but that uses up too many pins. I'm wondering if there's something fun I can do to still make it interrupt-based, so I can avoid regular polling of the ADC input
[12:02:30] <LeoNerd> Also, wondering what resistances are best to use; ideally some arrangement that I can tell if >1 button is pressed
[12:05:19] <LeoNerd> Ohoho! I have an idea
[12:06:43] <LeoNerd> Since the quiescent state is no button pressed, I could arrange that the input has a pulldown resistor, and all the buttons are some amount of resistor pullup. Then with no button pressed, the input is ~0V, and any button raises it. If I choose values sufficient that even the lowest "pressed" voltage is well above the logic HIGH level, then I can use a PCINT to tell me -that- a button is pressed, then switch the >>
[12:06:44] <LeoNerd> << pin into analog mode to read what value it was, then drop it back to PCINT to know when it was released
[12:07:59] <LeoNerd> What's the level required for that?
[12:09:09] <LeoNerd> Ooooh or I could use the anacomp
[12:39:48] <Lambda_Aurigae> back from shopping.
[12:39:51] <Lambda_Aurigae> spent way too much.
[12:44:29] <Lambda_Aurigae> but I got some really yummers extra sharp aged cheddar cheese.
[12:58:58] <LeoNerd> So.. at 3.3V input using the 1.1V bandgap, that feels like 10k pulldown, then a 10k and 4.7k pullup on the buttons; leading to ~0V off, or ~1.6 or around 2, when pressed
[13:00:03] <LeoNerd> Ooh I can probably detect >1 button at once like this, if I pick the pullups right
[13:02:22] <Lambda_Aurigae> with an r-2r ladder you could do several buttons easily.
[13:05:54] <Lambda_Aurigae> and with the ADC you could do multiple buttons at once.
[13:19:51] <Dominykas> Hey guys, i have a common anode 8x8 led dot matrix display, but only 8 transistors. Can i put them at the rows, where the + is? Or will i have to buy 24 transistors for every collum?
[13:20:12] <LeoNerd> What kind of transistors?
[13:20:27] <Dominykas> bc337
[13:20:32] <LeoNerd> PNP, NPN?
[13:20:35] <Dominykas> npn
[13:20:38] <Lambda_Aurigae> how do you come up with 24?
[13:20:45] <Lambda_Aurigae> 8x8 has 8 rows and 8 columns.
[13:20:48] <Lambda_Aurigae> that's 16 max.
[13:20:50] <Dominykas> oh i mean
[13:21:01] <Dominykas> if i add more led dot matrix displays (3)
[13:21:12] <Dominykas> so in total 3 displays
[13:21:38] <Lambda_Aurigae> generally, one transistor per column.
[13:22:07] <Lambda_Aurigae> then you can cycle through columns and cycle through rows.
[13:22:20] <Dominykas> just because its common anode, i have to use transistors on the collums?
[13:24:32] <Lambda_Aurigae> could go on either.
[13:24:55] <Lambda_Aurigae> I prefer to use the uln2803 driver chip over individual transistors.
[13:25:13] <Dominykas> so i can just place 8 on the row of the first display, and connect the second two?
[13:25:31] <Lambda_Aurigae> you will need 8 for each display most likely
[13:25:49] <Dominykas> yeah, the problem is when i bought everything i didnt know about the driver chips
[13:25:52] <Dominykas> because of the current?
[13:25:59] <Lambda_Aurigae> you are going to need 16 i/o pins to drive it unless you use some kind of shift registers.
[13:26:16] <Dominykas> yeah i have shift registers and one cd4017 for rows
[13:30:11] <Dominykas> So why do you think, i will need (most likely) 8 for each display?
[13:30:12] <Lambda_Aurigae> generally you use one driver/transistor per 8 LEDs
[13:31:34] <Lambda_Aurigae> often you will want to watch the max current draw at any one time.
[13:31:46] <Lambda_Aurigae> so, at most, 8 LEDs on at one time.
[13:32:24] <Dominykas> okay, i wont use 1 transistor for more then 8 leds
[13:32:51] <Dominykas> but i could then put the transistor at the anode side?
[13:33:26] <Lambda_Aurigae> you can put it at either anode or cathode.
[13:34:01] <Dominykas> alright, thank you once again!
[13:35:20] <Lambda_Aurigae> http://forum.allaboutcircuits.com/threads/8x32-dot-matrix-display-with-at89s52-in-c.83996/
[13:35:45] <Lambda_Aurigae> they use one uln2803 for all 4 displays.
[13:35:50] <Lambda_Aurigae> interestingly enough.
[13:36:23] <Lambda_Aurigae> but I'm betting they only turn on one column at a time...I'm betting not more than 8 LEDs at once.
[13:36:46] <Lambda_Aurigae> so, you COULD get away with 8 transistors.
[13:49:05] <LeoNerd> Woo. Well, I have 2 buttons working, anyway...
[13:49:51] <LeoNerd> (with multitouch ;) )
[13:50:16] <Lambda_Aurigae> LeoNerd, with an r-2r ladder you could easily do 8.
[13:50:29] <LeoNerd> Mm.. well, I only need 3 in reality I think.
[13:50:54] <LeoNerd> But 2 is enough to test. I like to have a fairly loose margin for the ADC comparisons, because otherwise it'll get quite tight tolerance by the end
[13:51:22] <Lambda_Aurigae> r-2r will still give you a decent differentiation.
[13:51:54] <Lambda_Aurigae> and you just need one value of resistor....or 2 if you go that route but can do it with 1.
[13:54:42] <Lambda_Aurigae> http://en.wikipedia.org/wiki/Resistor_ladder#R-2R_resistor_ladder_network_.28digital_to_analog_conversion.2C_or_DAC.29
[13:59:15] <LeoNerd> Well, I also need that the "off" value is the wrong side of the bandgap from any of the "on" values... so that I can use the anacomp
[13:59:23] <LeoNerd> This doesn't look like it'll do that
[14:10:40] <LeoNerd> Oh this is lovely :)
[14:11:03] <LeoNerd> Using the anacomp to tell me _if_ any buttons are held; if they are then poll the ADC every 50msec, report changes of state. If not, then no ADC polling, so shutdown.
[15:46:20] <Flipp_> anyone used a L3GD20H gyro before?
[15:52:21] <Flipp_> mainly just noticed that the chip has a few offsets and needs to be calibrated. I can turn on a high-pass to get rid of the DC offset, but I'd rather not use filtering if I can just add an offset to calibrate
[21:31:31] <Horuzar> Anyone familiar with the H44780 driver and the 16x2 LCD?
[21:31:39] <Horuzar> I have a simple question wondering if someone can help
[21:31:45] <Horuzar> Getting unexpected output
[21:36:33] <Horuzar> http://pastie.org/10061802 "You win." is showing up as soon as the program is launched on the FIRST line of output... idk why.. can
[23:43:54] <Flipp_> is it normal for gyros to have an offset while at rest? it's pretty constant, but all three dims are definitely not zero
[23:57:56] <Casper> did you paid a few thousands dollars for your gyros? if not then it won't be perfect