#avr Logs
Jan 25 2018
#avr Calendar
12:03 AM day__ is now known as day
02:33 AM polygon: http://joshuawise.com/projects/ndfslave
02:34 AM _abc_: polygon: interesting, thanks for posting it
02:35 AM polygon: _abc_: hi dude whats up?
02:36 AM _abc_: Downing coffee #2 before chores. Morning here.
04:19 AM polygon: so a DVD disc has areal data density of ~416.260598146 KiB/mm²
04:26 AM polprog: hmm
04:26 AM polprog: polygon: pretty neat
04:27 AM polprog: damnit, sill waiting for the stms to arrive. its bloody priority mail, it should not take more than 3 days
04:30 AM polygon: if i were to create an ecc-encap-decode stream filter i would need to fix the blocksize to say 1MiB or try to brute-force the block-size? ;/
04:30 AM polygon: polprog: which one you like? :)
04:30 AM polprog: id say use FAT blocksize
04:30 AM polprog: or wahtever the crappy camera uses
04:31 AM polygon: what are you talking about? ;>
04:32 AM polprog: what are YOU talking about :P
04:32 AM polygon: your stms
04:32 AM polprog: aaaa
04:32 AM polprog: okay
04:32 AM polygon: [11:25] <polprog> damnit, sill waiting for the stms to arrive. its bloody priority mail, it should not take more than 3 days - [11:28] <polygon> polprog: which one you like? :)
04:33 AM polprog: 2x blue pill aka stm32f103c8 plus 4x some veroboard just for general use, they were cheap
04:36 AM polygon: did you know you can print your pcb to an office paper with a laser printer then iron it on a pcb?
04:36 AM polygon: you can ther scrub off the paper with a wet sponge and etch the copper with acid
04:38 AM polprog: you dont say
04:39 AM polprog: its not that i spent whole days trying to either iron them so they transfer whole to find that half of the print doesnt even stick
04:39 AM polprog: ;)
04:39 AM polygon: hm, i usually get it right ;l
04:40 AM polygon: the trick is put it on a towel or some insulater stuff/and/or preheat the board to >200°C then put the paper on and roll it on with a ruber roller
04:41 AM polprog: i still have some sulphate left. i could try
04:41 AM polprog: i have some clad left as well
04:41 AM polygon: toner melts at ~200°C
04:41 AM polprog: i set the iron to "3 dots" and i think i checked with a pyrometer that its >200C
04:41 AM polygon: and first you should degrease, even fine sand the copper surface
04:42 AM polygon: when i did photolithography my board looked like a mirror
04:42 AM polprog: i did apply every possible degreaser, and abrasive agent on the clad
04:42 AM polygon: then it was not hot enough, and you probably did not roll it down evenly
04:42 AM polygon: i found an awesome tool in a printer, the rubber roller rod
04:47 AM polprog: possible
04:47 AM polprog: i just pressed the clother iron
04:48 AM polygon: you know, the board must be >190°C to stick to toner
04:48 AM polprog: that could be it
04:49 AM polygon: i did weld foil with an iron at max setting too
04:49 AM polygon: nylon/pp foil
06:33 AM day: does a simple PID controller lib for avr's exist?
07:04 AM cehteh: pid is simple even when you write it on your own
07:39 AM Emil: day: current=(1-a)*current+a*input; integral+=current; derivative=current-previous;
07:40 AM Emil: and so on
07:41 AM Emil: obviously with caveats
08:00 AM day: cehteh: i guess i give it a try then. I think i know the theory behind it anyways.
08:30 AM polygon: emlinear interpolate? :)
08:31 AM polygon: Emil: linear interpolate? :)
08:32 AM polprog: if only i knew what those variables reflect :P
11:16 AM jragon: Here's the final(ish) use of the graph I've been making: https://streamable.com/9ufe3 It's a PID controlled boost converter
12:03 PM mudkip908_ is now known as mudkip908
12:20 PM polprog: atmega8l, uart dies when vcc goes below 3.9
12:21 PM polprog: but thats in its operational range ...
12:21 PM polprog: why does it do it
12:23 PM LeoNerd: BOD settings?
12:24 PM polprog: might need to take a look
12:24 PM polprog: but its not that it cuts out at 3.8
12:24 PM polprog: *9
12:24 PM polprog: at 3.85 it receives corrupted data
12:25 PM polprog: but ill take a look at bod
12:25 PM nuxil: arg. damn. i got outbid on the fluke 287.
12:26 PM nuxil: still one day left. its at $235
12:27 PM nuxil: i was also looking at this one. https://www.ebay.com/itm/Brymen-BM869s-BM867s-BM257s-BM235-BM859-BM857-Professional-Multimeters-Series-/302094989685?
12:27 PM polprog: i could use a bench meter
12:27 PM nuxil: BM869S
12:27 PM nuxil: but that dude wants 50€ in shipping
12:27 PM polprog: i have bm 867s. its pretty good
12:27 PM nuxil: thats like 1/4 of what the meter is worth :p
12:28 PM polprog: yeah... that sucks... i got mine from a local retailer
12:28 PM nuxil: i had to contact the seller since no shipping details listed.
12:29 PM nuxil: oh well. the DMM hunt continues.
12:31 PM nuxil: i'll wait until tomorrow. if the bid on that flue is less than $250 i'll make a bid again.
12:31 PM nuxil: *fluke
12:31 PM polprog: ive got kind of a space problem on this bench - ALL the cables are too long
12:31 PM polprog: i think the scope should be at eye height
12:32 PM nuxil: get your scissor
12:32 PM nuxil: :p
12:32 PM polprog: putting it higher could help with excess cable lenght making a total mess on the bench
12:33 PM nuxil: build shelfs
12:33 PM polprog: yeah
12:34 PM polprog: i was thinking of making this bench deeper like 40 cm more, and hanging shelves on the wall
12:34 PM polprog: also the cork has some quirks
12:34 PM nuxil: a deep bench is good
12:34 PM nuxil: cork?
12:35 PM polprog: i have cork laid on 12mm plywood as the bench top
12:35 PM polprog: i was watching mike harmans video on his shop
12:35 PM nuxil: get his https://www.ebay.com/itm/Green-Anti-Static-ESD-Desktop-Grounding-Mat-Wrist-Strap-Ground-Wire-50x60cm/323002670951?ssPageName=STRK%3AMEBIDX%3AIT&_trksid=p2057872.m2749.l2649
12:35 PM polprog: he has kitchen workop as a bench
12:36 PM nuxil: cork? wtf :p
12:36 PM polprog: look, i was short on money
12:36 PM polprog: :P
12:36 PM nuxil: :)
12:37 PM polprog: i have this green rubber cutting mat on it, or rather its minified version...
12:37 PM polprog: i think i have some pics
12:37 PM polprog: https://puu.sh/yvflA/7f7597849b.jpg
12:38 PM polprog: its a very shallow bench
12:38 PM polprog: the psu depth is half the depth
12:38 PM nuxil: how can you work on that. not "fat enuf" :)
12:39 PM nuxil: make it 40 cm deeper or so yea
12:39 PM polprog: that would mean rebuilding it entirely...
12:40 PM polprog: should measure the kitchen worktop we have, see if there are any shelves that dont look like warehouse furniture, and convince mom its time to renovate the kitchen
12:40 PM polprog: then i could have bench as the whole worktop, and all the instruments would be on that shelf
12:42 PM nuxil: wall mount that monitor :p
12:43 PM polprog: i was thinking about that kind of steel wire rack they use for stacking PC cases
12:44 PM polprog: http://ec-i21.geccdn.net/site/images/n-picgroup/30012684.jpg this could save drilling in that wall
12:44 PM polprog: thats a good idea
12:45 PM nuxil: meh. overkill.
12:46 PM nuxil: just go to your local hardware store and get some plywood, make them cut in wanted size, and get some decent strong shelf brackets
12:46 PM polprog: hmm
12:46 PM polprog: good idea
12:47 PM polprog: hahahaha ready made workshop bench/table costs 500 eur :P
12:48 PM nuxil: damn. i spend 100 on plywood and parts. then get a scope or a dmm for the 400 remaining :p
12:49 PM polprog: i think i spent less than 150pln (50 eur) on the whole bench (minus work)
12:50 PM polygon: i have never killed statically any device btw, i don't have a sparky t-shirt.
12:50 PM polprog: i just dont wear polar
12:50 PM polprog: and try not to spark
12:50 PM gjm: >pln
12:50 PM gjm: pierogi
12:50 PM nuxil: you know.
12:50 PM nuxil: you dont need sparks to kill a chip.
12:51 PM polprog: i know esd precautions
12:51 PM nuxil: esd can be sparkless
12:51 PM polygon: well i don't even comb my hair with chips
12:51 PM polprog: but sparks dont help :P
12:51 PM nuxil: why not be safe than sorry?
12:51 PM polprog: i am
12:51 PM polygon: i have killed 60$ laser diodes, but not with ESD
12:51 PM polprog: all chips are in static foam or static bags
12:52 PM nuxil: imagin your mother walks in and stars to poke at a chip and ask what is that and zaps it :p
12:52 PM polygon: noise in trimpots :) that is a killer
12:53 PM nuxil: old carbon tim pots are a night mare.
12:54 PM nuxil: *trim
12:54 PM polygon: oh btw how many of you predry their components prior soldering?
12:54 PM polprog: predry? in what? toaster oven :P?
12:55 PM polprog: im still hunting for a en electric grill oven to use as reflow oven
12:55 PM polygon: whatever, they have moisture trapped inside and it will damage the chip when heated
12:55 PM polprog: meh
12:55 PM polygon: polprog: lol i got one from tesco and made a pt100 termostat for it
12:55 PM nuxil: ffs. i why both of you have the same color on your nick.
12:55 PM nuxil: so confusing
12:55 PM polygon: the toaster is like 20 eur
12:55 PM polprog: nuxil: using irsii?
12:56 PM nuxil: hexchat
12:56 PM polprog: ah
12:56 PM polprog: well, i had often situations when people had the same nick length and color
12:56 PM polprog: thats confusing
12:56 PM polygon: hair-dryer works for drying too...
12:56 PM polygon: it should be temp controlled though
12:57 PM polprog: i think i only used a hairdryer on chips when i was testing some basic zener voltage regulator i built after reading a chapterin arts of electronics
12:57 PM polprog: polygon: i even had a working thermocouple thermometer proto for the oven
12:58 PM polygon: good
12:58 PM polprog: couple lnes of code and parts and it would make a nice controller
12:58 PM polygon: atmega?
12:58 PM polprog: yeah
12:58 PM polprog: i mean, i dont know any other microcontrollers than avr right now :P
12:58 PM polygon: i made an analog one, potmeter died and it caught on fire
12:58 PM polprog: 10/10 would buy
12:58 PM polygon: it was a tesco hotplate lel
12:58 PM polygon: 10 eur
12:59 PM polygon: i made a pt100 regulator for it, it was +-2°C accurate
12:59 PM polygon: 0 to 270°C
01:00 PM nuxil: nice
01:00 PM polprog: mine was using max31855 with K type thermocouple
01:00 PM polygon: now i refurbished it but my mcu tempcontroller is not ready
01:00 PM polygon: the hotplate's heating element survived glowing red-hot lel
01:01 PM polygon: so im gonna hack it, make it proper
01:12 PM Casper: one day I might need to do a controlled temperature box, able to both do heating and cooling
01:12 PM Casper: will be annoying :D
01:12 PM Casper: (really, just need an h-bridge)
01:20 PM polprog: looks like making a shelf for that bench would help reclaim most of the space
01:21 PM polygon: Casper: a 250W (or 2) peltier TEG elements and a H-bridge, and heatsinks, fans
01:22 PM polygon: and an ntc or pt100 or 2
01:22 PM polygon: TEG is good upto 125°C
01:23 PM polygon: maybe 30°C cooling / stage, if you want -30°C you might need multiple stages
01:23 PM polygon: they are kind of low efficiency
01:33 PM polprog: tegs seem like a cool thing
01:33 PM polprog: ive seen some in cooled ccd cameras
02:50 PM Casper: polygon: was thinking more of a 25W peltier
02:50 PM Casper: it will be only for a usb hub and some rtl sdr dongles
03:38 PM polygon: Casper: nah wont work
03:38 PM Casper: why not?
03:38 PM polygon: tiny
03:39 PM Casper: so?
03:39 PM polygon: unless you only want 1cm³ sample cooler
03:40 PM Casper: insulated box, 30°C delta... 10W... ok, let's go for 50W peltier then?
03:40 PM polygon: let's see, the peltier has losses, and box insulation is not perfect
03:41 PM polygon: was it 30°C or 60°C 0 td?
03:42 PM Casper: 50°C outside, 20°C inside
03:42 PM polygon: hot side warms up, cold side cools down, difference in temperature created by power, zero power is 30-60°C it is in ds
03:43 PM Casper: I think it will actually only cool..
03:43 PM polygon: let's say half power = half of td_zero
03:43 PM polygon: no.
03:43 PM polygon: the hot side will get hot
03:44 PM Casper: no care
03:44 PM polygon: you need to cool that good
03:44 PM Casper: 2 good fans
03:45 PM polygon: i have tried a 75W peltier in a tiny box, got some ice, that's all
03:48 PM polygon: Casper: math will work out approximately
03:49 PM Casper: still think a 25-30W peltier would be fine
03:53 PM polygon: It all depends on your requirements
03:53 PM polygon: but peltiers are disappointing
03:54 PM polygon: TEG has like 5% efficiency, and TEC is like less than 80%
03:54 PM polygon: (TEG is the generator type)
03:56 PM polygon: another problem you might run in is if you run the hot side hot' and you turn it off, the cold side will spike high temps
03:57 PM polygon: i would gues it destroys laser diodes in this occasion
04:11 PM polygon: i *think* i optimized some bit operations ;> test will show
04:12 PM polygon: 10MB/s was poor
04:50 PM Casper: in my case it is not direct contact, so not important
04:50 PM Casper: the box will be in te attic or in an antenna mast or alike
04:50 PM Casper: just need to keep the inside at 10-25°C, ideally closer to 10
04:50 PM Casper: that will help to keep the noise low
04:51 PM Casper: well, "low"
04:51 PM Casper: let's just say barelly noticably lower :D
04:52 PM polygon: http://peltiermodules.com/peltier.datasheet/TEC1-12706.pdf
04:53 PM polygon: aham, so 66°C max diff at 0 power
04:55 PM polygon: this thing could do -8°C at 25°C hot side temp, 25W, (33°C diff)
04:56 PM Casper: https://i.imgur.com/jWqgjiY.gifv overdue?
04:59 PM polygon: it has some dust
05:09 PM Casper: a weightable amount
05:42 PM polprog: wonder if my friend took my advice not to smoke next to his PC
05:54 PM nuxil: any alternative to kicad ? the more i use it the more pissed off i get. i mean i hate this hole "begine/end" tool thingy. and draggin components that has wires attached to it just SUCKS ASS compared to ltspice.
05:54 PM nuxil: maybe i should just stick to good old ltspice.
05:56 PM polprog: isnt ltspice a sim framework
05:56 PM nuxil: yea. but you can use it draw schematic in it.
05:56 PM polprog: well yeah
05:56 PM polprog: i agree, eeschema has its bad sides
05:56 PM polprog: i think one of the bigger flaws is that i uses bitmask renderer
05:57 PM polprog: which for me glitches out when i rotate something...
05:57 PM polprog: i have to rezoom to force redraw
05:57 PM polprog: but pcbnew has improved much
05:58 PM nuxil: ltspice is a bit old. and it feels so on it gui. also it lacks alot of packages/componetns. but its general inerface is good. as in how keyboard/mouse works.
05:58 PM nuxil: kicad had lots of packages and looks nice. thats it :p
05:59 PM polprog: i mean, eagle is old as hell too
05:59 PM polprog: i think it was glitcing out the same way when i was using it :P
06:00 PM polprog: Altium looks like its gui was untouched for the last 15 years !
06:01 PM nuxil: well. Altium is out of the question.
06:03 PM nuxil: fritzing?
06:04 PM polprog: lol
06:04 PM nuxil: :p
06:04 PM polprog: whats that :P EDA for schoolchildren ;)
06:05 PM nuxil: oh online crap.
06:05 PM nuxil: no thanks.
06:07 PM nuxil: librepcb ?
06:07 PM polprog: gEDA
06:07 PM polprog: i spent a day clicking around and never got anything working :P
06:08 PM nuxil: https://en.wikipedia.org/wiki/Schematic_capture#/media/File:Kicad_Eeschema_screenshot.jpg
06:08 PM nuxil: thats another 80's looking app :p
06:08 PM nuxil: http://librepcb.org/
06:08 PM nuxil: have you tested ?
06:08 PM polprog: never heard
06:09 PM nuxil: i will try it.
06:09 PM polprog: doesnt look bad
06:09 PM polprog: at least its not SERIF FONTS ON TECHNICAL DRAWINGS like Altium does, jesus
06:10 PM nuxil: ok. downloading hte nightly build to test with
06:10 PM polprog: what distro are you on?
06:10 PM nuxil: windows
06:11 PM polprog: okay
06:11 PM polprog: X windows i hope ;)
06:11 PM nuxil: lo-. no winshit 10
06:12 PM polprog: i was about to ask why dont you download it from the repo :P
06:12 PM nuxil: oki. so the the nighly link is broken. :(
06:13 PM nuxil: https://ci.appveyor.com/api/projects/librepcb/librepcb/artifacts/build/librepcb-nightly.zip
06:13 PM nuxil: downloads 1MB then fails
06:20 PM nuxil: idk. so much crap out ther aswell
09:16 PM * nuxil crys
09:46 PM day__ is now known as day
11:36 PM day__ is now known as day