#avr Logs
Feb 14 2018
#avr Calendar
12:11 AM absynth is now known as dan2wik
12:34 AM Jartza: nuxil: I have 3 Tevo Tarantulas and Tevo Black Widow
12:35 AM Jartza: but they are DIY-kits, so "some" (a lot) assembly required
12:35 AM nuxil: i dont mind assembly :) i rather like doing builds
12:37 AM nuxil: where did you buy them from?
01:22 AM polprog: morning
01:25 AM Haohmaru: polprog ur LATE!
01:25 AM nohit: morning
01:41 AM Jartza: nuxil: https://www.aliexpress.com/store/product/2016-Newest-TEVO-Tarantula-I3Aluminium-Extrusion-3D-Printer-kit-printer-3d-printing-2-Rolls-Filament-8GB/2010004_32596996503.html?spm=2114.12010612.0.0.de1f40bb6gjiC9
01:43 AM Jartza: nuxil: there's also very active FB-group with >40k users for help :)
04:01 AM nohit: polprog you have that rigol 4 channel scope ?
04:02 AM nohit: have you managed to get that logic analyzer part working ?
04:03 AM nohit: or the decoder part
04:03 AM nohit: SA-DS1000Z Serial Bus Analysis Option
04:04 AM nohit: wouldnt want to pay 200bucks extra for feature like that
04:06 AM nohit: that seems like a dream scope but it takes a little off of its glory when they're trying to rip people off with these software upgrades
04:13 AM polprog: nohit: my version, 1054z doesnt have a la builtin
04:14 AM polprog: decoders are nice though
04:14 AM nohit: so there are a decoder out of the box ?
04:14 AM nohit: *is
04:14 AM polprog: not really
04:14 AM polprog: its an option , you get a 30 day demo
04:14 AM nohit: https://www.ebay.com/itm/Rigol-DS1054Z-Digital-Oscilloscopes-Bandwidth-50-Mhz-Channels-4/331941742120?epid=2254681137&hash=item4d49450628:g:EHYAAOSwUoNaA4By
04:15 AM Haohmaru: it sounds too similar to the DS1052E digital oscillotroll i have here at the job
04:15 AM polprog: 1054z, this one i have
04:16 AM Haohmaru: do all the rotary encoders respond "backwards" on it? ;P~
04:17 AM nohit: on the other hand i shouldnt really worry about that, i have logic analyzer that's way better
04:19 AM polprog: me too
04:19 AM polprog: Haohmaru: what do yo mean bavkwards?
04:19 AM polprog: backwards*
04:19 AM Haohmaru: like, you probe some signal, and say it's either too small or out of range, and you reach for the scale knob on the scope to adjust it.. just like you would reach for a volume knob on a hifi or pretty much anything..
04:19 AM Haohmaru: AND it responds in the opposite direction
04:19 AM nohit: i remember we had an old dedicated logic analyzer in our school's lab, it must have cost a fortune back when it was new. these days you can get one for 20bucks and it fits in your pocket :)
04:19 AM Haohmaru: whoever wrote the firmware on that thing is either an idiot or a troll
04:20 AM polprog: Haohmaru: mine is CW=+, CCW=-
04:20 AM polprog: nohit: one of those big CRT boxes?
04:20 AM Haohmaru: so, it's a troll thing then
04:21 AM Haohmaru: i suspected that, because why would you buy the more expensive model otherwise
04:22 AM nohit: polprog yes
04:22 AM polprog: heh, cool
04:22 AM polprog: of only i had a big enough lab
04:23 AM nohit: cant really remember the details, maybe something like this https://2.bp.blogspot.com/-Z7FJqRgULfI/VE-yuP__48I/AAAAAAAABEM/D-Ko3kKTJ5Q/s1600/IMG_0289.JPG
04:23 AM polprog: nice
04:23 AM polprog: *fallout intesifies*
04:24 AM Haohmaru: wow, a "DON'T CARE" button \o/
04:25 AM Haohmaru: and a hexadecimal keypad
04:26 AM polprog: is that the any key
04:26 AM Haohmaru: nah, this must be better
04:26 AM Haohmaru: i would press it SO MUCH..
04:26 AM polprog: lol. me too
04:27 AM polprog: tape it down pressed
04:27 AM nohit: damn me for being lazy last spring
04:27 AM polprog: gtg
04:27 AM nohit: see ya
04:27 AM polprog: its middle of maths right now lol
04:39 AM polprog: 2+2=5 for sufficiently big values of 2
05:04 AM nohit: Haohmaru
05:05 AM Haohmaru: yiz
05:05 AM Haohmaru: le moi
05:05 AM nohit: why dont you do DSP hardware synths ?
05:05 AM Haohmaru: i wish
05:05 AM nohit: you do them on PC so why not with hw
05:06 AM Haohmaru: i know on theory there is an ARM-based chip with floating point and sh*t which should be capable of decent amount of DSP
05:07 AM Haohmaru: but then that needs to be glued with <something> to even have a chance to get audio
05:08 AM Haohmaru: and, getting the basic "boring" code that runs everything in the background till i can have my audio process() function..
05:08 AM Haohmaru: no idea about those things
05:08 AM nohit: timer interrupt is the process() function
05:08 AM Haohmaru: probably
05:08 AM nohit: not probably
05:08 AM Haohmaru: i've not went higher than 8bit
05:10 AM nohit: you set up the timer to a certain frequecy, and that frequency is your sample rate
05:11 AM Haohmaru: everything is simple in theory
05:11 AM nohit: then do your magic in the ISR and after that put the resulting number to your external or internal DAC
05:11 AM Haohmaru: ;]
05:11 AM Haohmaru: internal DAC?
05:12 AM nohit: yes STM32's have internal DAC, not all but some of them
05:12 AM Haohmaru: is that gud enough?
05:12 AM nohit: yes
05:12 AM Haohmaru: how much bits
05:12 AM nohit: 12
05:12 AM Haohmaru: o_O
05:12 AM Haohmaru: that's kinda lofi like tape
05:13 AM nohit: no its isnt
05:13 AM nohit: https://www.thomann.de/fi/waldorf_streichfett.htm
05:13 AM nohit: this uses a single stm32
05:13 AM Haohmaru: maybe if you apply a logarithmic hack..
05:13 AM nohit: and its internal 12 bit DAC
05:13 AM Haohmaru: then 12bits should be decent
05:14 AM nohit: there's also some cheap external 16-bit DACs that rue_ linked some time ago
05:15 AM Haohmaru: in any case, it's above my head
05:15 AM nohit: but i would use the internal because its simple as writing a value into a register
05:16 AM Haohmaru: not sure about running a ~200kHz timer ;]
05:17 AM nohit: this must be somehow related to oversampling
05:17 AM nohit: because you only need 44100Hz
05:17 AM Haohmaru: if the DAC setup time is slow, it might be simpler to sacrifice 16 pins and do a R2R network ;P~
05:17 AM nohit: well that's what nuxil is doing
05:18 AM Haohmaru: and i don't have much spare time anymore, so if i'd be doing something, then i'd better have a very good idea exactly what
05:18 AM Haohmaru: iykwim
05:19 AM nohit: yeah
05:20 AM Haohmaru: i know exactly what i'd want to make
05:20 AM nohit: https://www.futurlec.com/Datasheet/Others/PT8211.pdf
05:21 AM Haohmaru: two things, the smaller one first
05:21 AM nohit: cheap as hell https://www.aliexpress.com/item/20pcs-bag-PT8211-S-SOP8-two-16-bit-analog-conversion/32354544565.html?spm=2114.search0104.3.15.6a011c32jzHirs&ws_ab_test=searchweb0_0,searchweb201602_3_10152_10151_10065_10344_10068_10342_10343_10340_10341_10084_10083_10618_10304_10307_10302_5711211_5722315_10313_10059_10534_100031_10629_10103_10626_10625_10624_10623_10622_10621_10620_10142,searchweb201603_2,ppcSwitch_5&algo_expid=f02dd998-e990-4afe
05:21 AM nohit: ffc102-2&algo_pvid=f02dd998-e990-4afe-9e6a-a6b87cffc102&transAbTest=ae803_5&priceBeautifyAB=0
05:22 AM nohit: damn
05:22 AM nohit: https://www.aliexpress.com/item/20pcs-bag-PT8211-S-SOP8-two-16-bit-analog-conversion/32354544565.html
05:22 AM nohit: ok tell me
05:22 AM twnqx: Sorry, this item is no longer available!
05:23 AM nohit: https://www.aliexpress.com/item/10PCS-The-new-8211-PT8211-PT8211-S-SOP8-16-bit-digital-analog-converter-IC/32839393249.html
05:23 AM twnqx: Estimated Delivery Time:28-56days lulz
05:24 AM nohit: damn :D
05:24 AM nohit: well they are cheap if you can get them
05:24 AM Haohmaru: nohit a 303, and a polysynth
05:25 AM nohit: https://www.aliexpress.com/item/Free-Shipping-PT8211-SOP-8-100PCS-LOT/32466443459.html
05:25 AM twnqx: also serial, 32bit per sample, no fun with avr
05:25 AM nohit: well you wouldnt use avr for DSP
05:26 AM twnqx: reminds me i wanted to make three PCBs for about 1.5 years now ;_;
05:27 AM nohit: Haohmaru waht kind of polysynth ?
05:27 AM nohit: analog?
05:27 AM Haohmaru: "SR-8050"
05:27 AM Haohmaru: i have it as vst plugin
05:27 AM nohit: you can get these analog oscillator chips these ddays https://electricdruid.net/analog-renaissance/
05:28 AM nohit: ok
05:28 AM Haohmaru: it's strongly inspired by roland's jp8000 and D-50
05:28 AM Haohmaru: with some twists ;P~
05:29 AM Haohmaru: i think i know a guy who knows that guy ;P~
05:31 AM Haohmaru: while the 303 works on any sampling rate since it's supposed to be "analog", my polysynth is supposed to be like a digital hardware synth, so running on a fixed sampling rate is a very important aspect IMO
05:31 AM Haohmaru: in the VST i do that by "resampling" (crudely) in realtime
05:33 AM Haohmaru: and in any case, knowing the sampling rate in advance is a very precious advantage
05:33 AM Haohmaru: no such luxury in vst ;]
05:34 AM nohit: yeah
05:35 AM Haohmaru: i'm doing a re-design of the x0xb0x IOboard pcb
05:35 AM Haohmaru: i want to finish that, and see if it works
05:35 AM Haohmaru: and if it fixes the things i wanted to fix
05:38 AM nohit: it would be cool to make a 3-voice analog polysynth with those chips. signal path all analog, but digitally controlled so you could edit and save presets
05:38 AM nohit: cool
05:38 AM Haohmaru: 3 voice is too few
05:38 AM nohit: nah
05:38 AM Haohmaru: i'm not really a fan of "analog" just because it's analog(ue)
05:38 AM Haohmaru: good sounds are good, no matter how they were made
05:38 AM nohit: yes
05:40 AM Haohmaru: basically, i wanted to do that, but i couldn't (that was before i knew how to kicad), then one day Roland came out with the TB-3
05:40 AM Haohmaru: it was a digital 303 in a box
05:40 AM nohit: yes i have it
05:41 AM Haohmaru: except it had a few major flaws that completely ruin it
05:41 AM Haohmaru: then they came out with the TB-03, which fixed all of the flaws........... except one
05:41 AM Haohmaru: x_x
05:41 AM Haohmaru: pls roland
05:42 AM nohit: whats the one
05:42 AM Haohmaru: and there's a guy who is doing a digital 303 in a box from before roland's attempts, but on some 16bit (i think) processor(s)
05:43 AM Haohmaru: he contacted me, but i've not heard from him for a long time, so no idea how's his progress
05:43 AM Haohmaru: nohit the accent cutoff modulation, but i suspect their whole cutoff modulation coupling is faked too
05:44 AM Haohmaru: the other stuff are pretty okay
05:44 AM Haohmaru: osc, filter, etc..
05:46 AM Haohmaru: technically that last flaw is purely software, so they can fix it with an update
05:47 AM Haohmaru: but they have that flaw from the TB-3 even, so it smells like they don't see it as a "flaw"
05:49 AM nohit: yeah
05:51 AM nohit: i actually have 2 old DSP audio dev kits from TI, sadly i dont have the software for them
05:51 AM nohit: they dont make those things anymore
05:51 AM nohit: well they do but they cost a fortune, nothing for the hobbyists
05:54 AM nohit: this was a cool product, i suggested to our teahcer that we would get it but the problem was that it needed a credit card and our school didnt have one https://line6.com/tcddk/images/devbackground.jpg
05:55 AM nohit: guitar pedal that you customize with your own code
05:57 AM Haohmaru: that reminds me of that other digital effects box, something-something-owl
05:58 AM polprog: now that i knoe how to pwm, as soon as i sort my lowpass out i will make a bitcrusher
05:58 AM Haohmaru: for the 303 i think i'd need two audio outs and one audio in, at the same time
05:59 AM Haohmaru: with low latency that is
06:04 AM nohit: yeah
06:04 AM nohit: polprog how are you gonna implement it ?
06:08 AM nohit: one way to do it(if your signal is +1.0...-1.0) is to multiply it with some integer and then round that result, and then go back to 1.0...-1.0
06:09 AM Haohmaru: if you got floating point, you can do it with floor() ;P~
06:09 AM Haohmaru: but bitcrushers are boring
06:10 AM Haohmaru: decimators ftw
06:11 AM nohit: my signal is 0...65535
06:12 AM nohit: i shift it with >> 4 so its fits into the 12bit DAC
06:13 AM nohit: here's the DAC output https://imgur.com/a/Eex2q
06:20 AM Haohmaru: getto DAC ;P~
06:20 AM Haohmaru: * ghetto
06:23 AM Jartza: hmmh
06:23 AM Jartza: strange
06:23 AM polprog: nohit: most likely left adjust the adc regsiter and keep copying it into output compare register
06:23 AM Jartza: looks like _delay_ms doesn't work in attiny5
06:24 AM Jartza: _or_ I have totally diffent clock than I think
06:24 AM polprog: i wanted to make a decimator as well but again i cant calculate it
06:24 AM polprog: but i simulated it in falstad
06:28 AM danci: here's my "process" function https://pastebin.com/yeENMa47
06:29 AM Haohmaru: eh?
06:29 AM Haohmaru: u run teh dsp inside the isr?
06:30 AM nohit: where else ?
06:30 AM Haohmaru: main thread?
06:30 AM Haohmaru: "thread" ;P~
06:30 AM nohit: no no
06:32 AM nohit: main thread is just for UI
06:35 AM nohit: i have to go to lunch, bbl
06:35 AM nohit: ->
06:35 AM Jartza: lol
06:36 AM Jartza: delay works
06:36 AM Jartza: but it seems attiny5 has irreversible fuses
06:38 AM Emil: Jartza: eh?
06:43 AM Jartza: yep
06:43 AM Jartza: I accidentally burned fuse WDTON
06:44 AM Jartza: avrdude: verifying ...
06:44 AM Jartza: avrdude: verification error, first mismatch at byte 0x0000 0xfc != 0xfe
06:44 AM Jartza: avrdude: verification error; content mismatch
06:44 AM Jartza: then tried burning 0xfe instead 0xfc
06:45 AM Jartza: tried also 0xff... no success
06:45 AM Jartza: tried erasing the whole chip... did check the lock bits, no lock bits set...
06:46 AM Jartza: so the led blinked too fast because watchdog kept resetting it
06:53 AM Emil: Jartza: you can disable the watchdog fuse no porblem
06:53 AM Emil: problem*
06:58 AM Jartza: avrdude don't want to co-operate with that :(
07:05 AM Haohmaru: Jartza you didn't forget the "safemode" thing, did you?
07:05 AM Jartza: nope
07:07 AM Jartza: avrdude -c usbasp -p attiny5 -u -U fuse:w:0xFF:m
07:07 AM Jartza: avrdude: verifying ...
07:07 AM Jartza: avrdude: verification error, first mismatch at byte 0x0000 0xfc != 0xff
07:07 AM Jartza: avrdude: verification error; content mismatch
07:07 AM Jartza: fuse burning to 0xfc works nicely
07:08 AM Haohmaru: "fuse" ?
07:08 AM Haohmaru: which fuse is that? ;P~
07:08 AM Jartza: it's the only fuse in attiny5
07:08 AM Haohmaru: could it be fuse0 then?
07:08 AM Jartza: nope
07:09 AM Jartza: it's "fuse"
07:09 AM Jartza: and like I said, burning fuse 0xFF -> 0xFC did work fine, or 0xFF -> 0xFE
07:10 AM Haohmaru: *shrug*
07:13 AM Jartza: and I'm using "high voltage TPI"
07:13 AM Jartza: so reset pin gets 12V
07:13 AM Jartza: and otherwise programming works just fine, plus zeroing fuse bits. just setting them back to 1 fails
07:17 AM Haohmaru: could it be internal hardware fault?
07:19 AM Jartza: tried 3 chips so far :D
07:19 AM Jartza: now I have to disable watchdog at boot in them, d'oh
07:19 AM Jartza: http://savannah.nongnu.org/bugs/?52318
07:19 AM Jartza: but seems I'm not the only one
07:21 AM Haohmaru: usbasp...
07:24 AM Jartza: tried with atmel ice too, fails also
07:43 AM emil is now known as Emil
07:45 AM Jartza: http://i.imgur.com/ms4jyO6.png
07:45 AM Jartza: this with atmel ice and atmel studio
07:46 AM Jartza: so it looks like attiny4/5/9/10 fuses are write-once
07:47 AM Emil: Allrighty
07:47 AM Emil: there we go
07:47 AM Emil: Updated my vm
07:48 AM Jartza: writing flash works fine, programming fuses *once* works fine, but resetting back to defaults does not
07:50 AM Haohmaru: could it be fake chips?
07:51 AM Jartza: I would be surprised if they are, as thees are from Farnell
07:52 AM Haohmaru: could it be a hardware bug?
07:53 AM Jartza: wouldn't be surprised
07:54 AM Haohmaru: i think almost all of the xmegas i programmed had this thing where one of the fuses doesn't get written the first time
07:54 AM Haohmaru: attempting again after that results in success
07:55 AM Haohmaru: but running the command twice one after another on a new chip doesn't do it
07:55 AM Haohmaru: for some reason, it seems there must be a pause between the first and the second attempt *shrug*
07:55 AM Haohmaru: that is, with avrdude5.x and olimex avrisp2 clone
07:55 AM Emil: Jartza: did you read that fuses are supposed to be writable multiple times?
07:56 AM Jartza: yes
07:56 AM Jartza: Configuration bits are not affected by a chip erase but they can be cleared using the configuration section erase command (see Erasing the Configuration Section in this chapter). Note that configuration bits are locked if Non- Volatile Lock Bit 1 (NVLB1) is programmed.
07:56 AM Jartza: but NVLB1 is not programmed
07:57 AM Jartza: oh well. I just throw these two chips into bin "use maybe later" :)
07:57 AM Jartza: and take third chip and *don't* program the WDTON fuse :D
07:57 AM Emil: haha :D
07:57 AM Emil: Hey hmm
07:58 AM Emil: Jartza: what you could try is program a small program that disables the watchdog
07:58 AM Emil: and then try reprogramming the fuse
07:59 AM Haohmaru: avrdude -c avrispmkii -P usb -p x128a3 -u -U fuse4:w:0xF3:m
07:59 AM Haohmaru: avrdude -c avrispmkii -P usb -p x128a3 -u -U fuse5:w:0xE5:m
07:59 AM Haohmaru: avrdude -c avrispmkii -P usb -p x128a3 -U flash:w:whatever.hex:i
08:00 AM Haohmaru: that's in my .bat script, and fuse4 fails to change from its default value on a new chip
08:00 AM Haohmaru: the rest of the commands work
08:00 AM Haohmaru: this happens also on xmega32a4u
08:01 AM Jartza: Emil: what good would that do, because giving 12V to rest pin switches to programming mode anyway
08:01 AM Jartza: and if WD would be running there, then flashing would fail too
08:01 AM Jartza: but flashing the chip works just fine
08:04 AM Emil: Jartza: you are already using high voltage?
08:04 AM Emil: I see
08:06 AM Jartza: yes
08:06 AM Jartza: have to, because I've disabled reset
08:06 AM Jartza: it gives me 33% more pins :D
08:07 AM Jartza: 3 -> 4 :D
08:07 AM Haohmaru: wow, so.. many.. pins..
08:08 AM cehteh: someday microchip will release an ATtiny in to92 case, just for Jartza
08:18 AM Jartza: I would make it do VGA
08:19 AM Jartza: I know how to get VGA out from single pin :D
08:20 AM twnqx: in stunning truecolor!
08:20 AM Jartza: that might prove to be a bit harder
08:21 AM Emil: Hmm
08:21 AM Emil: I should try generating vga also
08:21 AM Emil: Except without those silly restrictions you place on yourself ;)
08:22 AM Haohmaru: i have VGA on my comput0r ;P~
08:22 AM * Haohmaru hides
08:22 AM twnqx: do i still have vga on my computer.. obsolete tech...
08:28 AM Jartza: Emil: silly restrictions are purely for fun :)
08:36 AM cehteh: vga is so old skool
08:36 AM cehteh: dvi at least, better DP or HDMI
08:43 AM Jartza: those would require either dedicated transmitter chip or very powerful mcu :)
08:43 AM Jartza: or fpga
08:43 AM Emil: I don't think there's a general purpose mcu or cpu that can generate DVI, DP or HDMI
08:44 AM Emil: The bandwidth required is just too great
08:44 AM rue_: unless you reduce it to 320x200 at 4 bit colour
08:44 AM Haohmaru: you need only two colors, black and green
08:45 AM rue_: I disagree, black and orange are the colours you need
08:46 AM * Haohmaru vomits from his eyes
08:47 AM cehteh: low res / colors would be fine, just the lowest end of what HDMI or whatever allows and maybe some 'light' external hardware (pwm, shift registers?)
08:48 AM Jartza: rue_: IIRC lowest supported hdmi resolution is 640x480
08:48 AM cehteh: but hook up a microcontroller to some 'actual' lcd monitor with digital inputs would be nice and useful
08:48 AM Emil: okay I take my words back
08:48 AM Emil: that's doable
08:49 AM Jartza: what would be 251.25Mb/s bitrate in all three TMDS data channels
08:49 AM cehteh: well it has some protocol overhead you need to bang buffers quasi parallel
08:49 AM Emil: eh
08:50 AM Emil: 640x480x3x8x30/10^6=221Mbps
08:51 AM Emil: how'd you get 251.25Mbps?
08:51 AM cehteh: prolly the framing overhead?
08:51 AM Emil: hmm true
08:51 AM cehteh: how about monochrome?
08:51 AM Emil: no grayscale?
08:51 AM Emil: just divide by 24
08:52 AM rue_: hdmi in, who said anything about hdmi out?
08:52 AM Emil: 9.216Mbps
08:52 AM Emil: An AVR could generate that :D
08:52 AM rue_: cmon people 9600 baud
08:53 AM Emil: >using 9600
08:53 AM Emil: >not using 250k/500k/1Mbaud masterrace
08:53 AM Emil: pls
08:53 AM cehteh: lol
08:53 AM cehteh: ok get going
08:53 AM Emil: seriously
08:53 AM Emil: 9600 is cancer
08:53 AM Emil: I hate it got to be a standard
08:53 AM Jartza: Emil: an eight-bit color signal rides in a ten-bit word
08:53 AM cehteh: 9600 is excellent
08:53 AM Emil: Jartza: ah true
08:53 AM Emil: cehteh: wash your mouth
08:53 AM Emil: it's dirty
08:53 AM rue_: Emil, at 16Mhz, the avr can do 9600 properly, maybe 19200, but no higher
08:54 AM Jartza: eh
08:54 AM Emil: rue_: with 2x it can do 19200 also iirc
08:54 AM Haohmaru: 19230 ;P~
08:54 AM cehteh: Emil: its barely fast enough for interactive use, easy on the interrupts/mpu and you notice any problems visually still
08:54 AM Jartza: what does "properly" mean?
08:54 AM Emil: no
08:54 AM cehteh: i like 9600
08:54 AM Emil: 19200 and 9600 were both fine
08:54 AM Emil: with or without 2x
08:55 AM Haohmaru: xmega has fractional baudrate gen
08:55 AM Haohmaru: ;P~
08:55 AM rue_: http://wormfood.net/avrbaudcalc.php
08:55 AM Jartza: octapentaveega also gets data in at 9600 bps :)
08:55 AM rue_: well I thought I meant with 0% error
08:55 AM Jartza: while bitbanging VGA
08:55 AM cehteh: of course its not for high performance stuff, but for a serial CLI 9600 is gold
08:56 AM Haohmaru: some dudes type on the keyboard faster than that bitrate can handle ;P~
08:56 AM rue_: ah, I'm prolly thinking of the 8Mhz chart
08:56 AM cehteh: yes 9600 is on the edge there agreed, but eventually you pause typing
08:57 AM Haohmaru: no, you don't
08:57 AM Haohmaru: fish don't stop swimmin'
08:57 AM rue_: your saying someo can type at almost 1000cps?
08:57 AM Emil: rue_: bro
08:57 AM Emil: rue_: you can't generate 9600 with 0% error
08:57 AM Emil: on 16MHz
08:57 AM cehteh: i like the short 9600 baud lag, you see whats going on and if there any irregularites
08:58 AM rue_: Emil, I may have also been thinking with acceptable error, aka green on the chart
08:58 AM Emil: rue_: both 19200 and 9600 show green on the chart at 16MHz and with or without 2x bit ;)
08:58 AM _ami_: add CRC check?
08:58 AM Emil: anycase
08:59 AM rue_: tcp over serial
08:59 AM cehteh: i made a CLI with line editing on 9600 baud and full unicode (including double width characters) support, whcih needs to redraw the line frequently, 9600 is really slow, but it helped to optimize / debug the thing while still usable
08:59 AM Emil: I pretty much only use 1Mbaud anymore
08:59 AM Emil: but it requires 2x bit most of the time
08:59 AM cehteh: iff you want to get more data down the line, of course you need to up the datarate
08:59 AM cehteh: but for a simple CLI, its perfect
08:59 AM Emil: but yeah
09:00 AM Emil: 9600 is super supported everywhere
09:00 AM Emil: as is 115200
09:00 AM Haohmaru: my tcp-to-usart pipe runs on 230400
09:00 AM Emil: Haohmaru: dear god why
09:00 AM Haohmaru: i tried it on twice faster too, it worked
09:00 AM Emil: why not just 250k, or 500k, or 1Mbaud :D
09:00 AM rue_: no, I'm talking error retransmission
09:00 AM _ami_: Haohmaru: wow!
09:00 AM Emil: https://emil.fi/d/2018-02-14_17-00-02_SCZbCem7.png
09:01 AM rue_: Emil, dodgy odds of a random usb converter being able to do those rates tho
09:01 AM Haohmaru: Emil cuz i want to sniff it with the computer when sh*t happens
09:01 AM cehteh: Emil: note that with 9600 you also have more time handling things and possibly do other things (the mpu's main task)
09:01 AM Haohmaru: that sounded like i like to sniff sh*t O_o
09:01 AM rue_: serial programming in linux is hard enough without trying to cough up custom baud rates
09:01 AM cehteh: 115200 or faster put a lot load on a AVR when you need to parse data
09:01 AM Emil: rue_: pretty much all usb serial converters work up to 1M
09:02 AM cehteh: esp when runnin only at 8mhz
09:02 AM Emil: Haohmaru: yeah but
09:02 AM Emil: computers are fast bro
09:02 AM Emil: cehteh: yeah that's true, and it's the only issue with faster speeds
09:02 AM rue_: Emil, I'm not sure if BAUD1000000 is defined in termio.h
09:02 AM Emil: it is
09:02 AM Emil: termios.h
09:02 AM Haohmaru: Emil can't open the serial port at a very fast bitrate
09:02 AM rue_: k
09:02 AM Emil: Haohmaru: eh?
09:03 AM Haohmaru: 460800 (or whatever it is) certainly doesn't work on my computer
09:03 AM rue_: what we really need back is the 1990 automatic baud rated detection code
09:04 AM Emil: Haohmaru: god awful
09:04 AM rue_: at some point, people forgot how to do automatic baud detection
09:04 AM Emil: Haohmaru: just use 1M
09:05 AM Haohmaru: it should probably work, did i mention i use synchronous? ;P~
09:05 AM rue_: :) the tea machine put the tea in the pot this morning, not all over the counter :)
09:26 AM Jartza: hmmh
09:26 AM Jartza: wtf
09:27 AM Jartza: took the two chips that had those wdton fuses burned to other machine, but same atmel ice programmer
09:27 AM Jartza: this machine burns the fuses back to 0xFF just fine
09:27 AM Jartza: wtf
09:27 AM Jartza: same atmel studio version too
09:28 AM Jartza: the only difference I can think of is that the other computer had like atmel studio 5.x installed and then updated to 6 and 7 and the other one had 7 installed straight away
09:28 AM Jartza: could it just be atmel ice driver?
09:28 AM Jartza: those seem to be different version
09:30 AM Jartza: d'oh. it is the driver. updated driver on this machine too and now it can burn & "unburn" fuses too
09:30 AM Jartza: shit
09:30 AM Jartza: 5 hours well spent :D
09:31 AM Haohmaru: that sux
09:39 AM Jartza: also, usbasp can't seem to "unburn" fuses either
09:44 AM Haohmaru: another "great" thing about the usbasp
09:44 AM polprog: hehehe
09:44 AM * polprog flies away on a dragon
10:00 AM Jartza: I mostly use atmel ice, but avrdude doesn't support it's TPI mode
10:00 AM polprog: TPI?
10:01 AM Jartza: the programming mode attiny4/5/9/10 uses
10:01 AM polprog: ah the new one
10:01 AM Jartza: no it's not that new
10:01 AM _ami_: its old. usbasp support TPI
10:01 AM polprog: recently i was trying to program atmega8 with debugwire but it didnt work out. nether avrdude could not avarice
10:01 AM polprog: nor avarice*
10:02 AM Jartza: polprog: the UPDI is the new one, in recent attinies
10:02 AM Jartza: which is basically half-duplex uart
10:03 AM polprog: wonder if dragon supports that
10:03 AM polprog: and if not then probably could be hacked with a serial port from what you say
10:03 AM Jartza: TPI is only for those very tiny tinies :)
10:03 AM polprog: th only tini i worked with was quite big :P
10:03 AM polprog: 4313
10:03 AM Jartza: yeah, I programmed the mfg samples from atmel with just usb ttl-uart adapter and python script
10:04 AM Jartza: attiny817 samples
10:04 AM polprog: nice
10:04 AM Jartza: that was before atmel ice got support for it
10:04 AM polprog: i would think it needs some more hackery
10:04 AM _ami_: i worked with t13a
11:21 AM nohit: antto i have actually no idea how the pros do it since i have very little experience with this. i guess you could make some buffer based system, where the timer ISR would just feed the DAC with values from a buffer(that is filled elesewhere). but i have no need for that in my project, i have 25 s between the interrupts and that's easily enough, i have very fast MCU.
11:25 AM nohit: at some point i might try FreeRTOS and see how well it would fit for this
11:34 AM antto: nohit if you run sample by sample, then you have a tight limit, your process must be very consistent, time-wise
11:34 AM antto: if you process in blocks of samples, then some of these samples can take more time to process than others, and that won't cause glitches
11:38 AM antto: even with a DAC ISR like that, i would try to use a circular buffer so i can process() in the main.. "thread"
11:54 AM nohit: yeah
11:55 AM Jartza: lol
11:55 AM Jartza: my yuge firmware
11:55 AM Jartza: Program: 70 bytes (13.7% Full)
11:55 AM Jartza: Data: 0 bytes (0.0% Full)
11:55 AM * polprog claps
11:55 AM polprog: is that in C?
11:55 AM Jartza: yea
11:56 AM antto: my x0xb0x firmware is almost 100KB
11:56 AM Jartza: although ripped out all unneeded stuff
11:56 AM polprog: one of my projects takes up 6k if you enable debug commands
11:56 AM Jartza: like interrupt vectors as the fw doesn't use interrupts, and stack initialization, because I don't use variables :P
11:57 AM polprog: well, thats pretty close to writing assembly :P
11:57 AM antto: just one of my buffers is 4KB ;P~
11:57 AM Jartza: it just polls few inputs and sets outputs accordingly
11:57 AM Jartza: but it looks like I could tighten the logic in so small package that I can actually enable the interrupts also
11:58 AM Jartza: so I can even use MCU idle mode
11:58 AM Jartza: I was afraid that I fill the attiny5's 512 bytes flash
11:58 AM antto: 512 program flash?!
11:58 AM antto: and an instruction is 2 bytes?
12:03 PM nohit: antto are you a patreon for oli larkin?
12:03 PM antto: a wat
12:03 PM nohit: you posted that kvr link
12:03 PM nohit: iirc
12:04 PM nohit: he just released this https://olilarkin.github.io/wdl-ol/annotated.html , documentation for iplug2
12:04 PM antto: i tried to "contribute" to wdl-ol, but i guess i did it wrong
12:04 PM antto: then i kinda got dissapointed, and left it
12:05 PM nohit: ok
12:11 PM antto: nohit https://forums.cockos.com/showthread.php?p=1399294#post1399294
12:11 PM antto: 2014 x_x
12:30 PM nohit: hehe nice
01:14 PM polprog: jesus
01:15 PM polprog: i think ill install windows 7 over this 10 BS that is on the workshop computer
01:15 PM polprog: it has a 7 pro key
01:15 PM polprog: ill check if i have a 7 pro disk
01:15 PM polprog: for now i have a working windows here, i think
01:25 PM Emil: polprog: stop
01:25 PM Emil: you are making a grave mistake
01:29 PM polprog: what
01:30 PM polprog: it was here before
01:31 PM polprog: i put in a second disk that has debian on it
01:31 PM polprog: this windows here is just in case i need some obscure software
01:44 PM nohit: antto we actually use some heavy duty dsp shit in our product https://www.youtube.com/watch?v=Zm4ttYiuAmE
01:44 PM nohit: but i know very little about that part since i work with the wrist device/mobile software
01:49 PM nohit: this is how the wrist unit actually looks, the one on the video is prototype http://ariadna.tech/application/files/3014/8476/1004/Wristunit_withtext.png
02:28 PM antto: funky
03:28 PM nuxil: nohit, http://ariadna.tech/application/files/3014/8476/1004/Wristunit_withtext.png
03:28 PM nuxil: thats just a render right ?
03:33 PM nuxil: anywho. 1st suggestion is to it. remove the "ARIADNA, TECH" in the bottom on the lcd and replace it with the current depth, put the logo on the frame instead. or just have it as a spash screen when booting up.
03:33 PM polprog: discrete branding
03:33 PM polprog: vide garmin
03:33 PM polprog: just a small logo on the top of the bezel
03:34 PM nuxil: the other corner should include. descend ascend info and possible current
03:34 PM polprog: diving sounds like fun
03:34 PM nuxil: it is
03:34 PM polprog: ive only done snorkelling and i enjoyed it
03:34 PM polprog: :)
03:34 PM nuxil: i only tested a few times. but it was fun
03:37 PM nuxil: but diving never became my thing. hurts my ears to much.
03:38 PM polprog: could be a problem for me too
03:38 PM nuxil: and i dont like sniffing water to get the pressure equilized. :p just blowing with my nose closed only helps a few feet.
03:39 PM polprog: i dont know how deep you can go without any gear
03:39 PM polprog: i think i did 3 meters and at that point it hurt my sinuses
03:43 PM Emil: God damn I love undocumented features
03:43 PM polprog: hm?
03:45 PM Emil: cflow
03:45 PM Emil: the man page does not list all the features
03:45 PM polprog: looks like a cool tool
03:46 PM polprog: i need to try it
03:46 PM polprog: odd, i have the manpage for it but i dont have it
03:47 PM Emil: polprog: cflow --level 'begin=┃' --level '0= ' --level '1=┃ ' --level 'end0=┣━' --level 'end1=┗━' --cpp -n file.c 2>/dev/null
03:47 PM Emil: to make it prettier
03:48 PM polprog: i dont have the command
03:48 PM Emil: sudo apt-get install cflow
03:49 PM polprog: just a sec
03:49 PM polprog: huh rxvt supports *blinking* text
03:49 PM polprog: awesome
03:49 PM Emil: ohshit
03:50 PM Emil: _nice_
03:52 PM polprog: wow
03:52 PM polprog: niice
03:56 PM polprog: echo -e "\033[1;5;32m $ $ $ \033[33mJACKPOT\033[32m $ $ $\033[0m"
03:56 PM Emil: don't tease us
03:56 PM Emil: gib video
03:59 PM Emil: Ohshit
03:59 PM Emil: it works in putty too :D
04:00 PM polprog: https://puu.sh/znQLo/67ad80ac7c.mp4
04:01 PM polprog: putty? wow
04:01 PM polprog: xfce4-terminal doesnt support it though
04:01 PM Emil: polprog: yeah it's in the settings
04:01 PM Emil: https://emil.fi/d/2018-02-15_00-01-06_3JPZGHyg.png
04:02 PM Emil: polprog: is that which program btw?
04:02 PM polprog: urxvt
04:02 PM Emil: is it hte linux sharex one I linked you?
04:02 PM Emil: or some other?
04:02 PM Emil: I mean what did you use for capture
04:02 PM polprog: no, i didnt install it yet, used simplescreenrecorder
04:03 PM Emil: ah cool
04:10 PM polprog: damn the instructions on the github page for sharenix are wrong
05:56 PM polprog: night
09:30 PM unregisterednick: hi
09:30 PM unregisterednick: are tritium batteries cool for powering mcu? https://www.youtube.com/watch?v=8fOG5IZLECk
09:35 PM nuxil: cool, yes. practical. no.
09:37 PM nuxil: you be better off with some cr2032
09:52 PM unregisterednick: nuxil: ye a ;<
09:53 PM unregisterednick: nuxil: there are radioactve batts though, even in light houses in russia
10:01 PM nuxil: lolz.
10:02 PM nuxil: what isnt radioactive in russia :p
10:04 PM nuxil: but plutonium/uranium is a hole different ballpark than tririum.
11:37 PM day__ is now known as day