Back
[03:20:04] <Eszett2> yea
[03:22:09] <Eszett2> i have a question
[03:22:58] <Eszett2> I try to flash my Atmega with Atmel FLIP, but the application gives me a red light (error?) for "program" (programming)
[03:23:50] <Eszett2> what could be wrong there?
[03:24:16] <Eszett2> With a ISP programmer I succesfully flashed the same Atmega with the same hex file.
[03:26:46] <Eszett2> "adress is out of range"?
[03:28:40] <Eszett2> oh wait..
[03:48:18] <Eszett2> hmm
[03:50:29] <Eszett2> jo
[04:15:19] <Eszett2> hi
[04:18:42] <lorenzo> hi
[04:18:50] <Eszett2> ich meine neophyte =)
[04:19:04] <Eszett2> I said hi to neophyte..
[04:19:59] <Eszett2> have to go, laters..
[07:53:32] <hetii> hi
[07:59:40] <hetii> Q: is it possible to test jtag interface without connected device (some loop or something?:)
[08:01:18] <WormFood> Can you test an engine without fuel?
[08:01:31] <hetii> I load ixo-usb-jtag into Saleaellogic analyzer (based on cypress) so should act as a usbblaster but in urjtag I just get : Warning: TDO seems to be stuck at 1
[08:01:48] <WormFood> I see no way, to properly test a jtag interface, without something to connect it to.
[08:01:54] <hetii> when I pull down it then its always 0
[08:03:05] <WormFood> Is this the channel you want to ask this question in? Perhaps #Electronics might be more suitable.
[08:03:51] <hetii> Its always worth to ask people smarted then me :)
[08:06:17] <hetii> ok I found some board with lpc11c124 maybe form that chip I get some response
[08:45:18] <hetii> still no luck
[08:45:57] <hetii> I tested two board (one buffered and one non buffered) and even with lpc11c24 don`t have response ;(
[09:02:29] <twnqx> so many potential problems
[09:02:49] <twnqx> jtag is "fun"
[09:06:58] <LeoNerd> I'm having a mental block moment. LM2597 is a nice 5pin buck-mode SMPS controller. What's the boost-mode version of this? I want to turn 3.3V into 5V
[09:07:53] <twnqx> just use TI's selector?
[09:08:02] <twnqx> 3.3vin, 5vout, it tells you :P
[09:14:26] <LeoNerd> LM2577
[09:14:32] <LeoNerd> The boost version of a 2597
[09:15:26] <LeoNerd> Oh.. boo. They start at 3.5V input. :/
[09:22:15] <WormFood> LeoNerd, I expect 3.3 volts, would be within their margin of error.
[09:22:34] <LeoNerd> Hm? Nah.. the datasheet looks like it trails off very rapidly after 3.5
[09:22:49] <WormFood> That could be. I didn't look at the datasheet.
[09:23:00] <LeoNerd> I want to attach it to a LiFePO4, which has a cell voltage that drops down to about 3.2V when still at 20% capacity
[09:23:03] <LeoNerd> So that's too low for me
[10:06:01] <liwakura> uhm. im compiling with avr-gcc
[10:06:16] <liwakura> but i want to run some assembler instructions before the stack initialisation
[10:11:43] <twnqx> replace __init() :>
[10:12:38] <twnqx> http://www.nongnu.org/avr-libc/user-manual/mem_sections.html
[10:18:48] <liwakura> thanks
[11:34:10] <LeoNerd> *sigh* Ahh.. Farnell... tut tut tut... I hit the "datasheet" link for a MAX481. I get "revision 08" of the sheet, published in 2008. Whereas in my datasheets folder on my laptop I have revision 10, published 2014.
[12:25:16] <lorenzo> LeoNerd: why do you keep a datasheet folder? :p
[12:33:59] <LeoNerd> lorenzo: Because I like to keep local caches of PDFs of chips I usually use
[12:37:14] <lorenzo> ahh
[13:26:41] <sebus> 19:08:14 <LeoNerd> lorenzo: Because I like to keep local caches of PDFs of chips I usually use.
[13:26:42] <sebus> Same here :-)
[13:29:25] <lorenzo> btw, does anyone have or know where to find 8 bit sprites / bitmaps
[13:29:27] <lorenzo> for the alphabet?
[13:30:09] <eszett> like animals and such?
[13:30:21] <lorenzo> letters mostly
[13:30:25] <eszett> ah you mean fonts, yes
[13:30:27] <eszett> there is a website
[13:30:38] <eszett> one moment i look for the link
[13:33:09] <eszett> http://int10h.org/oldschool-pc-fonts/
[13:33:34] <eszett> a collection of all the old PC-BIOS-Fonts, as Bitmap or Truetype.
[13:34:44] <lorenzo> thanks, let's see if I can turn those into some sort of binary representation for the 8x8 display
[13:35:52] <eszett> np. well , you might look out for fontexport.exe i use this tool to export to BMP
[13:37:19] <eszett> http://www.alternate-tools.com/pages/c_fontexport.php?lang=ENG
[13:37:23] <eszett> this tool i mean
[13:42:28] <eszett> lorenz?
[13:44:45] <lorenzo> eszett: yes?
[13:45:42] <eszett> just wanted to know if you read my answer
[13:54:22] <lorenzo> starting to make my owns glyphs.h :-)
[13:54:27] <lorenzo> with numbers first
[14:31:24] <twnqx> i think i used a tool once to generate C-files from bitmaps
[14:31:44] <twnqx> linux bitmap fonts, that is
[14:32:46] <twnqx> yeah, .psfu fonts to .h files
[14:33:28] <Lambda_Aurigae> I used to do it with a spreadsheet
[14:33:47] <Lambda_Aurigae> put 1 or 0 in the various boxes to draw my fonts
[14:34:03] <Lambda_Aurigae> then export as a csv file and process that into my header or font files.
[14:34:09] <twnqx> heh
[14:34:56] <twnqx> lorenzo: seems i can give you .h files from lat2-08.psfu and lat2-10.psfu (8 pixels/10pixels high, i think) if you care
[15:37:37] <eszett> i have a question. On my circuit (atmega), i have RST floating, and when i connect it to GND i trigger the Reset. It works, so why is it recommended to pull RST to VCC by default?
[15:38:04] <eszett> instead of floating?
[15:39:05] <lorenzo> eszett: I don't think you want the RST pin to stay in a high impedance state
[15:39:40] <lorenzo> I think it may cause unintended resets if something happens to pull it down below a certain threshold
[15:40:02] <eszett> you mean floating is risky, as it could trigger by any accidant by itself?
[15:40:16] <eszett> high impedance = floating?
[15:40:27] <Tom_itx> because it will randomly reset if you don't
[15:40:36] <eszett> hmm but it never did
[15:40:42] <Tom_itx> but it will
[15:40:47] <eszett> i could measure its floating voltage with my multimeter
[15:40:54] <eszett> i guess its 0.08v, can that be?
[15:41:29] <eszett> assumed its 0.08v, how can it be that it would drop down to 0.00v?
[15:42:29] <eszett> Tom?
[15:47:01] <Tom_itx> i'm not concerned with what you're measuring. you need a pullup on reset.
[15:49:04] <cehteh> reset is never floating, it has a internal pullup
[15:49:21] <eszett> ahh, good point cehteh!
[15:49:37] <cehteh> but that pullup is rather weak, unless you leave reset completely disconnected you should add some external pullup
[15:49:39] <eszett> so thats why i dont need an external pullup
[15:49:46] <eszett> ok
[15:50:32] <eszett> one moment im measuring..
[15:50:52] <cehteh> a very small cap on reset swallows noise and spikes too and helps against accidental resets, but it must be small enough not to trigger 'reset' on any startup
[15:51:31] <cehteh> cap to ground
[15:52:11] <twnqx> doesn't the reset pin have a pullup integrated into the chip?
[15:54:25] <cehteh> thats what i just saied
[15:55:40] <cehteh> when you have reset connected to some longer traces/wires and the environment is noisy then the internal pullup is not always enough to prevent accidental resets
[15:55:53] <cehteh> atmel has an appnote about that
[15:56:33] <eszett> cehteh: i tried to measure between RST and GND, but somehow this always triggered a RST, why?
[15:56:57] <cehteh> because you have a crappy DMM, long wires, low impedance
[15:57:17] <cehteh> i saied that pullup is very weak
[15:57:36] <eszett> ugh
[15:57:57] <cehteh> maybe not all avrs have this internal reset pullup? datasheet helps
[15:58:00] <eszett> but that cant be.. my multimeters probes dont create continuity
[15:58:02] <cehteh> the ones i use have it
[15:58:31] <cehteh> the probes have either stray/static charges or low impedance
[15:58:39] <eszett> oh... i see
[15:58:44] <cehteh> what does it show when you leave it connected?
[15:58:49] <eszett> cehteh: 0.00v
[15:58:57] <eszett> i would have expected 5.12v
[15:59:06] <cehteh> me too :D
[15:59:13] <eszett> jeez, what the heck!?
[15:59:17] <cehteh> you havent plugged the probes into the current sockets?
[15:59:31] <eszett> i have touched with the probes to solder joints
[15:59:43] <cehteh> what avr is that?
[15:59:49] <eszett> atmega32u4
[16:00:26] <cehteh> never used that, read datasheet about its reset circruit
[16:00:31] <eszett> ok
[16:01:36] <cehteh> otherwise just put a 10k pullup and done
[16:02:10] <eszett> the pcb is already manufactured ;-(
[16:03:26] <eszett> would be a 1k sufficient too, or does it have to be 10k?
[16:04:03] * eszett has an idea
[16:04:38] <cehteh> yes 1k should work too
[16:09:31] <Tom_itx> either one
[16:09:34] <eszett> I have to correct myself, when i measure between RST and GND it gives me 0.08v, and between RST and VCC 0.00v. which is somehow odd..
[16:10:12] <cehteh> R RST Reset Pull-up Resistor 30-60 kOhm ... datasheet says
[16:10:57] <cehteh> what gives your dmm when you connect it vcc to gnd?
[16:11:23] <eszett> second..
[16:11:31] <cehteh> (magic smoke)
[16:11:58] <eszett> 5.13v
[16:12:07] <eszett> circuit works, LED blinks, etc.
[16:12:15] <cehteh> what DMM is that?
[16:12:23] <cehteh> some <$15 takeaway?
[16:12:30] <eszett> oh a very old one, noname from the 80ies, yes cheap
[16:12:40] <cehteh> get a decent dmm
[16:12:44] <eszett> argh..
[16:13:02] <cehteh> anything in the range of $40 to $100 should be sufficient
[16:13:08] <eszett> I will borrow some from my neighbour tomorrow
[16:13:30] <eszett> but the mm isnt the troublemaker, i guess.
[16:13:39] <cehteh> i am quite sure it is
[16:13:56] <eszett> hmm ok
[16:14:04] <cehteh> this ultra cheap ones have really bad inputs
[16:14:20] <cehteh> they are ok when you want to measure a cars battery or so
[16:14:38] <cehteh> but you cant measure high impedance electronics with it
[16:15:19] <cehteh> you dont need a ultra expensive one, but there is some difference between these really cheap ones and serious ones
[16:16:54] <eszett> cehteh: I see.
[16:17:11] <eszett> I will cross-check with a second circuit now (a clone)
[16:17:20] <eszett> same circuit, but different device
[16:17:23] <cehteh> you are in germany?
[16:17:28] <eszett> of course i am
[16:17:37] <cehteh> me too :d
[16:18:06] <cehteh> just look at reichelt or so, some uni-t or peaktech are ok
[16:18:33] <eszett> also muss es kein fluke für 200€ sein
[16:18:58] <eszett> ich werd mir das auf die besorgliste schreiben
[16:20:13] <eszett> mal ne andre sache.. du hast oben gesdchrieben das datenblatt sagt 30-60k, ich meine das steht da ja, nur was ich nicht kapier ist, dass alle mir bekannten schaltungen wie Teensy2.0, .. Arduino und was nicht alles, da nur 1k oder 10k haben, aber nicht mehr?
[16:20:25] <cehteh> http://www.reichelt.de/Multimeter-digital/PEAKTECH-3410/3/index.html?&ACTION=3&LA=2&ARTICLE=107371&GROUPID=4058&artnr=PEAKTECH+3410
[16:20:43] <cehteh> i think thats the one i have, its good enouh for hobbyist use
[16:21:00] <cehteh> of course you can always put more money into it and get some more features
[16:21:13] <eszett> oh, gut, aber das kommt ohne probes?
[16:21:18] <cehteh> mit
[16:21:29] <cehteh> und tasche iirc
[16:21:30] <eszett> ich brauch die feinen, für mein steckbrett
[16:21:45] <cehteh> eh english? oder /query
[16:22:21] <eszett> i gonna need the fine probe tips for my breadboard
[16:28:36] <lorenzo> in lab we use Amprobe multimeters
[16:28:53] <lorenzo> it's an Agilent/Keysight sub-brand iirc
[16:29:04] <lorenzo> decent multimeters for 30-40 eur
[16:30:11] <cehteh> yes some decent ones are as cheap as 30Eur ...
[16:30:38] <cehteh> just check what features you like
[16:31:06] <cehteh> latched beeper, fast response, backlight, what ranges
[16:31:45] <cehteh> most pita thing with dmm's is imo when they are too slow
[16:31:55] <cehteh> also Hz/% range is nice
[16:32:25] <cehteh> LC meter too but i'd rater buy another device for that
[16:33:22] <cehteh> the one i have can measure down to few µA .. thats nice sometimes
[16:43:50] <LeoNerd> I quite like my Uni-T UT61-E
[16:44:00] <LeoNerd> (note the E - it's quite different from the A/B/C/D variants)
[16:46:49] <LeoNerd> It's not *the* cheapest, it's GBP50 or so.. but feature-wise it's excellent for the price. 4½ digit / 22000 count, .1% basic DCV accuracy.. does mili and microamps
[16:47:25] <LeoNerd> Also not overburdened with silly features like temperature/light/randomcrap... It does have a capacitance meter but I can ignore that as I have a proper LCR anyway
[17:01:20] <twnqx> eszett: don't believe the reset nonsense :P
[17:01:29] <twnqx> lower values just mean higher current
[17:01:45] <twnqx> ehrn you force it to low in order to really reset
[17:01:49] <twnqx> when*
[17:20:51] <eszett> I have a question. A minute ago my ISP programmer went smoking.. the elemente called F1 is destroyed so it seems, what is that? a resistor or capacitor?
[17:24:15] <eszett> A diode, right?
[17:25:36] <twnqx> f = fuse
[17:25:47] <twnqx> sicherung.
[17:27:37] <eszett> twnqx: ahh, und worauf muss ich achten, wenn ich mir die kaufe, 500mA und Bauteilgrösse 1206, daswars oder?
[17:28:05] <twnqx> naja, und warum sie durchgebrannt ist natürlicj
[17:28:56] <eszett> ich habe einen verdacht.. ich hatte die schaltung mit einem usb kabel am pc angeschlpssen, und der ISP programmer auch noch angeschlossen, das heisst die schaltung hatte zweimal strom
[17:29:12] <eszett> ich dachte das wäre nicht so schlimme, weil 5v und 5v, macht ja immer noch 5v..?
[17:29:24] <twnqx> die sind nicht sooo identisch
[17:30:24] <eszett> twnqx: ich kann zur not doch einfach die fuse rauslöten, und als notbehelf ne drahtbrücke einsetzen oder
[17:30:34] <eszett> dann hab ich zwar keine sicherung mehr aber was solls
[17:30:54] <twnqx> der usb port sollte eh bei mehr als 0.5A abschalten *g*
[17:31:23] <eszett> also der isp programmer wird noch erkannt, aber ich glaub er tut nicht mehr so richtig..
[17:32:55] <eszett> boah das ganze zimmer stinkt!
[17:37:48] <twnqx> also *eigentlich* sollten sicherungen nicht so riechen...
[17:38:07] <twnqx> is das ding flach und grün? das sind eigentlich selbstheilende polyfuses...
[17:44:32] <cehteh> also wenn die fuse abraucht würde ich nicht probieren da ne brücke zu setzen
[17:44:54] <cehteh> als nächstes raucht dann das mainboard ab :D
[17:46:24] <eszett> also das war keine fuse im glasröhrchen, sondern so n 1206 bauteil irgendwie mit kohle oder so
[17:46:46] <eszett> jedenfalls hat das mit der brücke setzen auch nicht geklappt, weil die pads schwarz wie asche sind da ist überhaupt kein kuper mehr..
[17:47:03] <eszett> ich hab natürlich vorher mit alkohol gereinigt, aber die pads waren danach immer noch schwarz
[17:47:30] <twnqx> das klingt aber nicht nach 500mA ;)
[17:47:44] <eszett> steht aber so im datenblatt von dem 2€ china billig isp programmer
[17:48:04] <eszett> jetzt i st er kaputt, was für den preis nicht so ärgerlich ist, sondern eher weil ich jetzt keinen mehr hab...
[17:48:53] <twnqx> sowas:
http://www.mouser.de/images/littelfuse/lrg/1210l002.jpg ?
[17:48:57] <cehteh> da das sind selsbt zurücksetzende polyfuses .. wenn die durbrennt währ ich noch vorsichtiger
[17:49:09] <cehteh> da ist definitv was faul
[17:49:19] <cehteh> das muss schon nen heftiger kurzschluss sein
[17:50:05] <eszett> ja sowas in dr art
[17:50:17] <eszett> es hat auch eine leiterbahn angehoben
[17:50:27] <cehteh> http://www.mouser.com/images/microsites/Littelfuse%20Lo%20Rho%20PTC%20Resettable%20Fuses.jpg
[17:50:28] <twnqx> woah
[17:50:47] <eszett> das ist mein teil hier:
http://www.ebay.de/itm/380876438916?_trksid=p2057872.m2749.l2649&ssPageName=STRK%3AMEBIDX%3AIT
[17:50:53] <cehteh> eben .. check die schaltung, mess alles, mach auf keinen fall ne drahtbrücke
[17:50:55] <twnqx> naja, wer weiß was chinesen in nen 2€ programmer verbauen ;)
[17:51:10] <cehteh> weil als nächstes raucht der usb port
[17:51:14] <eszett> cehteh: jetzt ist es leider irreparabel
[17:51:30] <eszett> cehteh: wie jetzt wirklcich?
[17:51:39] <cehteh> ich hab auch solche programmer .. die funktioniueren bei mir
[17:52:11] <eszett> schick mir mal einen bitte =) weil wenn ich aus china bestelle muss ivch 6 wochen warten
[17:52:13] <cehteh> du hast nen heftigen kurzschluss oder ne ground loop wo strom drüber geht
[17:52:36] <cehteh> 2x 5v .. nicht aus dem gleichen netzteil?
[17:52:38] <eszett> cehteh: ja.. aber jetzt isses im müll
[17:53:07] <cehteh> ich sag ja nur, als nächstes ist der rechner mit dem usb port kaputt, das wird teuerer
[17:53:12] <eszett> cehteh: also beide kabel hatten die schaltung mit dem pc verbunden, also selbses (pc)netzteil
[17:53:22] <cehteh> und die programmer bekommste in 2 tagen von amazon oder ebay
[17:53:39] <cehteh> dann haste was verpolt oder sonstwie nen kurzschluss
[17:54:19] <twnqx> selnst dann muss man sich fragen, warum nicht wenigstens einer der ports abgeschaltet hat...
[17:54:33] <twnqx> oder waren das usb 3 anschlüsse?
[17:55:09] <cehteh> selbst dann
[17:55:10] <eszett> twnqx: weiss nicht... zusätztlich hatte ich mit dem Multimeter auch noch auf die Platine rumgestochert, ich glaub das war dann irgendwo zuviel des guten
[17:55:38] <eszett> also entweder warens die zwei stromquellen, oder das Multimeter-Gestocher, oder beides
[17:55:56] <cehteh> andere leute hier basteln nur hinterm nen billigen powered usb hub ... damit bei fehlern nicht das mainboard abraucht
[17:56:15] <twnqx> ich hab nen billigen unpowered :P
[17:56:26] <eszett> ach du jeh, wie kann denn von dem usb device ausgehend überspannung aufs pc mainboard kommen?
[17:56:29] <cehteh> hab ich auch hier aber ich weiss nicht ob das hilft
[17:56:39] <cehteh> ich rauch aber auch selten was ab :D
[17:56:39] <twnqx> da sind keine dioden oder sowas
[17:56:57] <twnqx> und die datenleitungen haben auch in der regel keinen schutz außer clamp dioden
[17:57:01] <twnqx> im chip.
[17:57:05] <cehteh> eszett: es reicht schon überstrom
[17:57:16] <cehteh> kurzschluss
[17:57:31] <cehteh> wenn da ne sicherung nicht schnell genug anspricht dann brennt ne leiterbahn
[17:57:37] <cehteh> oder gar die regler
[17:57:42] <eszett> sind teurere ISP programmer dagegen immun oder kann das bei jedem passieren?
[17:57:48] <cehteh> und das kann das komplette mainboard killen
[17:57:51] <twnqx> mh
[17:58:01] <twnqx> da musste schon SEHR teure haben
[17:58:07] <_26thmeusoc> aus dem grund des mainboard rauchens will ddas programmieren nach möglichkeit an einen rpi auslagern, damit das nicht mehr direkt am rechner hängt :/
[17:58:08] <cehteh> du kannst dir immer in den fuss schiessen wenn du was falsch machst
[17:58:28] <twnqx> eigentlich mal ne gute idee. isolierten programmierer bauen.
[17:58:30] <eszett> twqnx: ah, ne ich bin eher sparsam
[17:58:39] <cehteh> selbst/gerade der avrice war ja dafür bekannt das er schnell abraucht bei fehlern
[17:58:50] <twnqx> und die teuren stk600 auchj
[17:59:07] <cehteh> einfach nicht wild was verdrahten sondern mitdenlen
[17:59:27] <cehteh> und nur eine stromversorgung oder über dioden wenn es unbedingt sein muss
[18:00:00] <cehteh> das schützt aber nicht vor kurzschluss :D
[18:00:16] <eszett> ja denken ist manchmal nicht so meine stärke
[18:00:16] <twnqx> oder 5V auf usb data :3
[18:00:24] <cehteh> und wenns so heftig ist das ne polyfuse abraucht dann haste großen mist gebaut
[18:00:38] <twnqx> ja... ich frag mich echt wie nen usb port das durchlassen kann
[18:01:29] <cehteh> mainboard hersteller sparen gerne, vllt is auffm mainboard noch ne weitere polyfuse aber wenn du pech hast hängst das direkt am 5V .. und so nen modernes netzteil pumpft da auch mal nen paar dutzend ampere drauf
[18:01:49] <cehteh> das macht keine leiterbahn mit
[18:02:04] <_26thmeusoc> wobei der einfache usb hub ja auch nicht dazwischen reicht, wenn der backfeedet leitet der den überstrom meines wissens nach auch direkt an den rechner dahinter durch
[18:02:11] <twnqx> hab grad nen altes asus zerlegt. so viele ployfuses, und leiterbahnen im 5mm bereich für usb stromversorgungen
[18:02:40] <cehteh> haste aber nicht bei allem und vor allen dingen sind polyfuses sehr träge
[18:02:57] <twnqx> maybe
[18:03:09] <twnqx> wie gesagt, ich mag wenigstens nen unpowerd hub
[18:03:21] <cehteh> paar sekunden anprechzeit .. . wenn sie nicht vorher vaporsieren
[18:03:39] <twnqx> eigentlich... könnte man echt mal nen usb protektor bauen
[18:03:44] <twnqx> gehört ja nicht viel zu
[18:04:23] <cehteh> oder einfach schaltungen so aufbaun das nix passieren kann
[18:04:29] <twnqx> TI 0.5A USB current limiter, paar dioden für negative spannungen, z-dioden für positive, nur auf kapazität achten damti 480mhz noch gehen
[18:04:39] <twnqx> ach, da kann immer mal was schiefgehen
[18:05:14] <twnqx> ich hab schonmal ne komplette kette (2 reels) ws2812b aus nen usb port betrieben
[18:05:30] <twnqx> hat mich schon überrascht, dass 120 LEDs an einem port tun...
[18:06:28] <cehteh> :D
[18:06:32] <cehteh> und die ziehn
[18:06:38] <twnqx> jo, waren paar amps
[18:06:46] <twnqx> firmenlaptop halt :P
[18:07:10] <twnqx> war aber auch als hochstromport ausgewiesen.
[18:31:35] <eszett> I have a question, if a pin is floating what is its voltage?
[18:31:52] <twnqx> uh
[18:32:03] <twnqx> here's the problem
[18:32:07] <twnqx> cmos inputs, floating
[18:32:20] <twnqx> can turn on both the bottom and the top transistor
[18:32:25] <twnqx> causing a short
[18:32:36] <twnqx> kill chips due to overcurrent
[18:32:53] <twnqx> which is why the rules of cmos say "NEVER leave an input pin floating"
[18:33:08] <Lambda_Aurigae> never leave any digital input floating.
[18:33:12] <Lambda_Aurigae> never never never
[18:33:14] <Lambda_Aurigae> NEVER
[18:33:43] <twnqx> with TTL it's not that bad :P
[18:33:45] <eszett> and the input pins of my atmega, if i dont use them, then they are pulled up by default?
[18:33:51] <twnqx> no
[18:34:01] <twnqx> but you can turn on the pullups.
[18:34:09] <eszett> how do i do that?
[18:34:19] <Lambda_Aurigae> nobody reads datasheets....sheesh.
[18:34:38] <twnqx> set to input via DDRx, use PORTx to turn on pullup
[18:35:50] <eszett> ah, thx. And the rule is, when i set something to input i shall never forget to use or activate any pullup (internal, or external) as well right
[18:36:12] <twnqx> if something is connected, no need :P
[18:36:39] <eszett> oh, then i have misunderstood Lambdas sentence
[18:36:56] <twnqx> it doesn't float if some other device's output is cvonnected
[18:38:30] <eszett> and the 0.08v i measure at many pins on my Atmega, where ist that voltage coming from?
[18:38:49] <Lambda_Aurigae> leakage
[18:39:28] <eszett> Lambda: ah, and that value depends on my pcb, on anothers ones circuit it may be 0.09v, or 0.07v., right
[18:39:43] <twnqx> depends on the chip, supply, ...
[18:39:55] <Lambda_Aurigae> and various other parts of the circuit.
[18:39:56] <twnqx> and problem solar flares.
[18:40:01] <twnqx> probably*
[18:40:49] <eszett> and you say the atmega pulls up the RST pin by default, i dont have to write any code for that?
[18:41:05] <twnqx> well, you can't,
[18:41:17] <eszett> i cant?
[18:41:33] <twnqx> it's a dedicated pin, not a gpio pin...
[18:41:44] <twnqx> wenn ich fragen darf: wie neu bist du zu mikrocontrollern? :)
[18:41:55] <LeoNerd> It's both. See the RSTEN fuse. ;)
[18:42:10] <twnqx> well, it's one or the other...
[18:42:33] <eszett> oh, right. also ich bin sehr neu, es ist alles so furchtbar kompliziert
[18:42:38] <twnqx> but i think there are chips that do not have that fuse, and where the pin is ALWAYS input
[18:43:20] <twnqx> i know i "ruined" some chips by activating that fuse and not having a high voltage programmer :P
[18:43:21] <LeoNerd> Yah.. often the larger chips like the 64+ pin ones, usually have dedicated reset and xtal pins that aren't anything else as well
[18:43:29] <eszett> im asking, because my RST pin has 0.08v, and that value is not a wrong number by my DMM
[18:43:33] <LeoNerd> *tut* should always have an HVxP programmer handy
[18:44:05] <LeoNerd> My HVSP controller can fairly easily become an HVPP instead by the application of a couple of 595s
[18:45:27] <Casper> LeoNerd: is there a protocol difference between HVSP and normal serial programmer?
[18:45:49] <LeoNerd> Lots. totally different
[18:46:22] <LeoNerd> HVSP puts +12V on the reset line; uses one clock (SCI) and two data lines in (SDI, SII), one data out (SDO). HVSP uses an odd 11-bit clock cycle
[18:47:05] <LeoNerd> HVSP is just a serial version of the HVPP used on bigger chips. Internally, it's the HVPP memory controller hardware with two shift registers in front of it. SDI feeds an SR in front of the D0-D7 pins, and SII feeds an SR in front of the remaining handshaking/control signal lines
[18:47:12] <Casper> :(
[18:47:21] <Casper> stupidly made then :(
[18:47:28] <LeoNerd> Nop.. very sensibly made
[18:47:34] <twnqx> disagree
[18:47:47] <twnqx> could have just use "apply 12V to RST and use normal ISP"
[18:47:56] <Casper> I have this tiny85, and a normal programmer... could have been great to be able to reuse the reset pin
[18:48:05] <Casper> twnqx: that's what I was thinking...
[18:48:12] <Casper> use 12V instead of 0v...
[18:48:23] <LeoNerd> Casper: You could just get an HVSP programmer.. {disclaimer: I sell them :P }
[18:48:24] <Lambda_Aurigae> eszett, as to why those inputs are reading some output voltage...there is some gate leakage on mosfets which is what those inputs are connected to...
[18:48:45] <Casper> LeoNerd: I'm cheap :D
[18:48:53] <Lambda_Aurigae> Casper, or make a fuse doctor from another avr.
[18:49:12] <Casper> Lambda_Aurigae: that is what I actually think of doing, maybe
[18:49:19] <Lambda_Aurigae> http://mdiy.pl/atmega-fusebit-doctor-hvpp/?lang=en
[18:49:26] <Casper> I am in decision of how I do this project
[18:49:26] <Lambda_Aurigae> I have one I made for just that problem.
[18:49:28] <eszett> Lambda: ok, und die Mosfests sind im Atmega integriert?
[18:49:30] <LeoNerd> "Fuse doctor" can help... my programmer is a full stk500 clone though, so it can do normal flash/eeprom programming operations as well as fuses
[18:49:31] <twnqx> i just use my stk600 for that :X
[18:49:47] <Lambda_Aurigae> eszett, no clue what you said. I don't read german so well.
[18:49:53] <twnqx> eszett: i doubt he speaks german, and with that question, you should be happy about it...
[18:50:12] <Lambda_Aurigae> something about atmega having integrated mosfets I think.
[18:50:19] <eszett> sorry i switches language
[18:50:37] <Lambda_Aurigae> yes...atmega is a cmos device...the transistors are mosfet not bjt for the most part.
[18:51:18] <Casper> 8 pins... 2 power pins and a reset... 5 pins left... 3 for RGB, so 2 left, 1 for serial... or do I go 2... if 1 I need to use a 1 wire derivative, 2 I could use 2 wire...
[18:51:43] <LeoNerd> 3 pins for RGB? Why not use an intelligent LED of some kind? WS2812s can be done over a single pin for example
[18:53:14] <Lambda_Aurigae> probably more expensive than a plain RGB LED
[18:53:45] <Casper> LeoNerd: because :D
[18:54:05] <Casper> but I'm wondering if I shouln't move ti it...
[18:54:32] <Casper> LeoNerd: thru-hole? linky?
[18:55:27] <LeoNerd> Adafruit (and likely elsewhere) have some 5mm packaged versions of the WS2812. Elsewise, they come in easy to hand-solder 5050 packages; you can get some breakout boards for that to DIP layout if required
[18:56:38] <LeoNerd> Ahhah.. APA-106 is the part number
[18:56:58] <twnqx> Casper: pl9823
[18:57:29] <Casper> do they do PWM dimming? or just on/off?
[18:57:39] <twnqx> pwm, all three numbers you got
[18:57:42] <LeoNerd> 8bit PWM on all three channels.
[18:57:52] <Casper> I need chip, not led
[18:58:23] <LeoNerd> The PWM rate isn't amazing. Fine for static use but I don't like to use them on anything moving (like costumes or animatronics)
[18:58:37] <LeoNerd> For moving stuff I use APA-102s. They're loooovely
[18:59:47] <LeoNerd> One small detail to be aware of though: often these 5mm packages have different colour component order than the 5050 SMT devices.. Usually you have to swap red and green channels.
[18:59:57] <LeoNerd> That caught me out first time ;)
[19:00:56] <twnqx> or even slightly different timing, like the pl9823
[19:01:41] <LeoNerd> Oh.. that's fine. That's just "800kHz WS2812"... that's fairly standard
[19:01:58] <LeoNerd> Though a right bitch to drive from an AVR. I'd probably use a dedicated offload chip for that
[19:02:17] <LeoNerd> I twisted my head up enough trying to do mixed 2812+DMX last year. It works but it's not pretty
[19:05:00] <twnqx> the pl9823 are... different
[19:05:09] <twnqx> i have separate code for both :P
[21:19:43] <phinxy> When was atmega 8bit family "created" or should i say put on the market?
[21:30:05] <Casper> both?
[21:30:35] <eszett> something round the 1980ies i guess
[21:33:05] <Casper> oh when... skipped that word .... woops
[21:33:17] <Casper> no wonder the question didn't made sense :D
[21:33:24] <eszett> oh no, that cant be atmel was founded 1984 hehe
[21:34:15] <Casper> from wiki: The AVR is a modified Harvard architecture 8-bit RISC single-chip microcontroller, which was developed by Atmel in 1996.
[21:35:00] <cehteh> yes for a µC architecture they are quite new
[21:35:14] <cehteh> but where developed esp for embedded stuff
[21:35:58] <eszett> 1996? interesting..
[21:36:16] <Casper> so I got the email for onedrive that drop in size
[21:36:32] <Casper> it would be a wize move from dropbox to raise the base size!
[21:37:02] * Casper needs to remember to bring his usb-serial ttl adapter to work
[21:37:13] <Casper> I may have a router to recover the firmware....
[21:40:09] <phinxy> Casper, i need one of those , naow
[21:40:37] <Casper> chinese order one now, so you'll get it in a month!
[21:40:44] <phinxy> eggsactly
[21:47:28] <Casper> hmmmm is that airplane really in difficulty? or is it someone that misset their transponder?
