#avr Logs

Oct 03 2017

12:16 AM day__ is now known as daey
01:42 AM rue_bed: yea
05:33 AM eth3: could anyone please help me with reading ADC12 on atmega32u4?
05:33 AM eth3: based on this picture: https://cdn-learn.adafruit.com/assets/assets/000/030/930/original/adafruit_products_2889_pinout_v1_0.png?1457306365
05:34 AM eth3: i want to read analog value from the pin labelled PB5 there
05:34 AM eth3: confused with different pin naming notations
05:48 AM Emil: eth3: configure the pin as input and then configure the adc
05:49 AM Emil: admux adcsra adcsrb
05:49 AM Emil: iirc
05:49 AM Emil: then just read the ADC register
05:52 AM Cracki: get the atmega32u4 data sheet and make sure you have handled setting all the registers (if needed) for that device
05:52 AM eth3: it's in front of me, and it's even more confusing
05:52 AM eth3: it says the chip has 12 adc channels
05:53 AM eth3: but labels are going from adc0 to adc13, with adc3 being skipped
05:54 AM eth3: so i'm not even sure which number do i need to use to access that pin
05:55 AM MrCircuitMatt: which adc# pin is the one that says PB5?
05:55 AM eth3: could you please have a look at the image here: https://cdn-learn.adafruit.com/assets/assets/000/030/930/original/adafruit_products_2889_pinout_v1_0.png?1457306365
05:55 AM eth3: each pin has like 4-5 different labels
05:56 AM eth3: the only one that rings a bell is pb5
05:56 AM MrCircuitMatt: says ADC12. I would confirm that with the datasheet and go from there
05:56 AM eth3: does that mean it's 11th channel?
05:58 AM MrCircuitMatt: eth3: check "10.3.1 Alternate Functions of Port B" on page 74 of the datasheet for an answer :-)
05:58 AM MrCircuitMatt: in particular table 10-3
05:59 AM eth3: this is insane. at the bottom of that page it mentions adc channel 13
05:59 AM eth3: whereas in the features on one of the first pages it says the chip has 12 adc channels
05:59 AM eth3: page 2, line 6
05:59 AM eth3: i'm definitely missing something here
06:00 AM MrCircuitMatt: yeah but one of them is not available... it's like taking 3 pigs and writing 1, 2 and 4 on them
06:01 AM eth3: so i guess i need to prepare admux like it's 13
06:02 AM MrCircuitMatt: I have to run, but I can check later. My gut feeling is that 13 is correct
06:02 AM eth3: which is "100101" according to page 314
06:02 AM eth3: thanks
06:02 AM eth3: i'll try with 13
06:03 AM MrCircuitMatt: you're looking at the wrong page
06:03 AM MrCircuitMatt: in the table it says 12, next page 12 as well
06:04 AM MrCircuitMatt: you were looking at PB6 I think
06:06 AM eth3: ooops, yes, picked up the wrong one
06:07 AM eth3: thanks for spotting
06:14 AM Cracki: if it says adc12, it's channel 12. don't overthink it.
06:17 AM MrCircuitMatt: OK back
06:17 AM MrCircuitMatt: eth3: generally the way I do this is first to locate the chip you're interested in. You need the pinout of the eval board for that. Turns out to be PB5 in this case. But from there I'd avoid whatever weird nomenclature the eval board manufacturer decided to adopt
06:17 AM MrCircuitMatt: eth3: datasheets are usually structured in sort of a similar manner. Usually quite in the beginning you get a pinout diagram. There already it says that PB5 is ADC12. Then you get either a pinout table below, or some more prose on the pins. Yours is the latter case which is the page that I quote. There again it says ADC12 and they even provide a table for your convenience that also says ADC12
06:17 AM MrCircuitMatt: and even the eval board diagram says ADC12 so I'd be fairly confident on ADC12... :-)
06:19 AM Cracki: http://www.atmel.com/Images/Atmel-7766-8-bit-AVR-ATmega16U4-32U4_Datasheet.pdf
06:19 AM Cracki: 24.9 starts the registers
06:19 AM Cracki: just go through them, check the tables, set the bits according to what you want
06:20 AM Cracki: admux lists the bits needed to get adc12
06:22 AM Cracki: note: mux4:0 bits are in admux reg, mux5 bit is in adcsrb reg
06:25 AM eth3: thanks i got it working
06:26 AM eth3: could you please enlighten me on what obscure numbering is "arduino-style"?
06:26 AM eth3: is it just some kind of "simplification" in their libs?
06:26 AM Lambda_Aurigae: it's arduidiot numbering
06:27 AM Cracki: they just number the pins on the PCB
06:27 AM Lambda_Aurigae: probably to make something easier to lay out on some board done originally for ardweeny and it stick.
06:27 AM Lambda_Aurigae: stick/stuck
06:27 AM Cracki: instead of labeling them according to what their names on the chip are
06:27 AM Cracki: also to make them consistent across avr models
06:28 AM eth3: so their numbers don't have any meaning?
06:28 AM Cracki: not really
06:28 AM Lambda_Aurigae: only to arduidiots.
06:28 AM Cracki: diagrams such as the one you have there are good. they tell you the names (and functions!) of the pins of chip
06:28 AM Cracki: the
06:29 AM eth3: so arduino-specific libraries just translate those numbers to normal port/pin conbination behind the scenes?
06:29 AM Cracki: yes
06:29 AM Cracki: things like "PB5" typically refer to gpio-capable pins, which are grouped into ports (port B here)
06:30 AM Cracki: there are pins that aren't gpio-capable, and just have special functions.
06:30 AM Cracki: adc pins can be such.
06:30 AM Cracki: you'd see that on the diagram though... any "modules" tend to only route gpio-capable pins
06:32 AM eth3: looking at that diagram i just got a suspicion that arduino uses another numbering for analog pins as well. is that so?
06:32 AM Lambda_Aurigae: very likely.
06:33 AM Lambda_Aurigae: the ardweeny abstraction gets just silly sometimes in my opinion.
06:33 AM Cracki: you'll note that the BLE control uses pins that aren't routed out to pin headers, so they're in the extra box on the top right
06:33 AM Cracki: arduino tends to label analog pins A<number>
06:34 AM eth3: yeah, for example there are both ADC10 and A10, which are different pins
06:34 AM Cracki: correct.
06:34 AM Cracki: adc10 is the avr name for the special function.
06:34 AM Cracki: A10 is arduino
06:34 AM Cracki: and they have no meaningful mapping
06:35 AM eth3: is it a common practice to mix arduino and normal avr libs together in one project?
06:35 AM MrCircuitMatt: eth3: correct on the numbering for analog pins. I think it's like A0 to A#. Totally unnecessary IMO
06:35 AM MrCircuitMatt: I would not use arduino libraries at all
06:36 AM eth3: i personally never did. but i think i stumbled upon such a mixed project
06:36 AM eth3: with the pin naming confusion significantly amplified
06:36 AM MrCircuitMatt: yeah imagine the mess if half of your codebase uses pin PB5 and the other half uses pin 9.
06:37 AM Lambda_Aurigae: specially don't use the wiring library
06:37 AM Lambda_Aurigae: that is soooo soooo horridly bloated.
06:37 AM eth3: what's that library for?
06:37 AM Cracki: I don't mix arduino libs into plain avr code. arduino libs tend to use arduino apis. they're user-friendly, but cancer.
06:37 AM Lambda_Aurigae: digitalwrite() and such.
06:38 AM Cracki: if I want to use an "arduino" lib, I'll butcher it, taking the functions and leaving the arduino crud
06:38 AM Lambda_Aurigae: https://pastebin.com/dSRSxgax
06:38 AM Cracki: wiring is what arduino started from
06:38 AM MrCircuitMatt: I see how they can abstract off the actual pin naming with their alternative nomenclature, making stuff "compatible" across platforms, but you'd really want to avoid reading the datasheet super hard to use that to your advantage. And then their documentation is kinda weird so you end up reading their source and then you read the datasheet to comprehend the source... there's potential for a shortcut right there
06:38 AM Lambda_Aurigae: shows how overly bloated the digitalwrite() command is.
06:38 AM Cracki: it has to "just work" for many cases.
06:39 AM eth3: yeah, that's an impressive difference
06:39 AM Cracki: it disables any devices that could be using the pin
06:39 AM Cracki: (timers, uart/spi/...)
06:45 AM eth3: thanks guys, you are very helpful. i'll do a bit of coding armed with the new knowledge, and might come back for help
06:46 AM * eth3 feels the power
07:10 AM ayjay_t: anyone here use olimex and linux?
07:10 AM ayjay_t: or use dfu_programmer?
07:12 AM Lambda_Aurigae: olimex?
07:12 AM Lambda_Aurigae: olimex is a manufacturer.
07:12 AM Lambda_Aurigae: they make dozens of boards.
07:13 AM ayjay_t: yeah true
07:14 AM ayjay_t: i'm using the mk2
07:14 AM ayjay_t: i was throwing off because you can run the dfu_programmer locally, you have to make install
07:14 AM ayjay_t: thrown off*
07:14 AM ayjay_t: i'm trying to upgrade th firmware on their avr-is-mk2 to pdi-upgrade a xmega128a4u
07:14 AM ayjay_t: it's been a ride.
07:14 AM Lambda_Aurigae: guessing you mean the avr-isp-mk2
07:15 AM Lambda_Aurigae: I have a programmer from Tom_itx here that does pdi...but never had any pdi devices to test it on.
07:15 AM Lambda_Aurigae: dfu is something totally different.
07:16 AM ayjay_t: yeah i have the avr-isp-mk2, thats right, it was a typo
07:16 AM ayjay_t: and yeah i'm going to test the pdi soon
07:16 AM ayjay_t: but apparently theres lots of issues with versions etc and the firmware needs to be upgraded
07:16 AM Lambda_Aurigae: I do have a device with a dfu bootloader on it and have the dfu uploader program that works just fine.
07:16 AM ayjay_t: and dfu_programmer is used to upgrade the avr-isp-mk2
07:16 AM Lambda_Aurigae: and I use linux.
07:17 AM ayjay_t: well i ahve three versions of avr_dude installed on my computer right now, 6.3, 6.3+olimex batch, 5.9
07:17 AM Lambda_Aurigae: never had any problems with dfu and linux.
07:17 AM ayjay_t: like i said, the issue with dfu_programmer is that it needs to be installed, you can't run it locally
07:18 AM Lambda_Aurigae: ummm.
07:18 AM Lambda_Aurigae: not sure what you mean by that...
07:19 AM Lambda_Aurigae: flip for linux doesn't need any install...just needs java installed.
07:21 AM Lambda_Aurigae: dfu-programmer, the open source version, doesn't need to be installed.
07:21 AM Lambda_Aurigae: you just build it
07:21 AM Lambda_Aurigae: then go to the src folder and run it.
07:21 AM MrCircuitMatt: ayjay_t: I was confused for a second as well, because I took "locally" as the opposite of "remotely". May I ask why you can't run it from the directory you compiled in?
07:21 AM Lambda_Aurigae: or copy dfu_programmer to wherever you want.
07:22 AM Lambda_Aurigae: I just did that.
07:22 AM Lambda_Aurigae: it runs from the compiled directory just fine.
07:22 AM MrCircuitMatt: I just tried as well, and the binary runs for sure, but I have no device to test. Hence my curiosity
07:23 AM ayjay_t: but don't you guys have it installed as well?
07:23 AM Lambda_Aurigae: if you are in the directory where duf_programmer binary is located, just type ./dfu_programmer
07:23 AM Lambda_Aurigae: no.
07:23 AM ayjay_t: yeah it didn't work for me
07:23 AM ayjay_t: i'm not a total n00b
07:23 AM Lambda_Aurigae: now, to get it to work you might have to run it as root.
07:23 AM ayjay_t: and the batch file that is dfu_programmer was obviously looking for other programs that were being installed by `make install` in PATH
07:23 AM ayjay_t: yeah i sudo'd it
07:23 AM ayjay_t: it works now that i've installed it
07:24 AM Lambda_Aurigae: btch file?
07:24 AM * ayjay_t shrugs
07:24 AM Lambda_Aurigae: batch?
07:24 AM ayjay_t: yeah shell script
07:24 AM ayjay_t: sorry it's because i've been reading the atmel docs
07:24 AM MrCircuitMatt: that's the autocomplete script :-)
07:24 AM ayjay_t: they keep calling it a "batch file"
07:24 AM Lambda_Aurigae: dfu-programmer is a binary, not a batch.
07:24 AM ayjay_t: well cat ./dfu-programmer was bash
07:24 AM MrCircuitMatt: it's the autocomplete script. You can copy it to /etc/bash_completion.d/
07:25 AM Lambda_Aurigae: mine is a binary file.
07:25 AM MrCircuitMatt: then your bash can autocomplete funky stuff with tab and all that, but it's got nothing to do with the actual binary. The binary is in src/
07:25 AM ayjay_t: oh the binaries in src?
07:26 AM Lambda_Aurigae: didn't watch the compile?
07:26 AM MrCircuitMatt: the binary is in src/ and seems to run just fine
07:26 AM ayjay_t: lol Lambda_Auri you are kinda mean
07:26 AM ayjay_t: shoot, i should have cd src
07:26 AM ayjay_t: we're about to test the final thing tho
07:26 AM Lambda_Aurigae: of course.
07:26 AM Lambda_Aurigae: people need to read
07:27 AM Lambda_Aurigae: I watched the compiler output to see where it put the binary.
07:27 AM MrCircuitMatt: I'm away for a bit. Be nice to each other please kthx
07:27 AM Lambda_Aurigae: no.
07:27 AM ayjay_t: lol
07:27 AM Lambda_Aurigae: [:
07:36 AM dan3wik is now known as dan2wik
07:46 AM ayjay_t: i wonder if i'd get anywhere by just trying to run this from windows instead
08:03 AM ayjay_t: invalid device sig...
08:03 AM ayjay_t: that's the first check it probably does to make sure it has the appropriate programmer?
08:07 AM _ami_: ayjay_t: your avr programmer problem did not solve yet?
08:08 AM ayjay_t: no im still working on it
08:08 AM ayjay_t: i might solder up a custom board i think this forum says that apull up resistor on the reset line is a no go
08:09 AM ayjay_t: thats exactly what i'm going to do
08:11 AM _ami_: ayjay_t: rst line of mcu programmer?
08:11 AM ayjay_t: no of the xmega128a4u
08:11 AM ayjay_t: one of the pdi lines
08:11 AM _ami_: generally programmer connects one of gpio to RST line of other mcu
08:12 AM _ami_: oh xmega stuffs. did not work on xmega yet.
08:12 AM ayjay_t: i'm more experienced with arm, so i'm a little bit new to how the clock domains and fuses work on avr
08:13 AM _ami_: ayjay_t: plz do write a blog abt how you create this programmer. it would be good information for others who likes to do what you are doing.
08:13 AM _ami_: ayjay_t: i feel you. i can understand how setting fuses can be annoying for you. :)
08:13 AM ayjay_t: _ami_ i usually do document my travels on various support forums so people can search
08:13 AM ayjay_t: but in this case i'm just moving towards more and more commercial solutions
08:13 AM _ami_: nice.
08:31 AM henkannie: Hi all
08:32 AM henkannie: I've got an Mega 2560 board here and when I try to transmit data over UART, the RX led lights up, not the Tx led.
08:33 AM henkannie: Is this some sort of reverse=psychology because people always forget to put Tx in Rx and vice-versa?
08:38 AM Haohmaru: depends how you wire the LED up
08:39 AM Haohmaru: how do you know which LED is "rx" ?
08:39 AM Haohmaru: does it go to the RX pin or does it just have some label on the silkscreen saying "arr eks"
08:41 AM ayjay_t: should i try programming it before i solder on the crystal?
08:42 AM Emil: ayjay_t: watch dem fuses
08:43 AM ayjay_t: uhh but
08:43 AM ayjay_t: yeah i mean, iw as hoping just to get a *connect* and read the correct err device number
08:44 AM ayjay_t: that would be father than i've got so far, but i knwo atmels are kinda weird with their fuses and oscillators and i just don't know if i should solder on the crystal before i try programming it with pdi
08:44 AM ayjay_t: i mean, it uses a clk signal right?
08:48 AM ayjay_t: i guess no harm in trying
08:48 AM Emil: If the crystal is on pins you don't use for programming it doesn't affect anything
08:51 AM ayjay_t: wow okay thats really convenient
08:51 AM ayjay_t: so i need four pins, pdi, power, and ground
08:57 AM henkannie: Haohmaru: It has a label on the silkscreen saying Rx
08:57 AM henkannie: And it lights up when I do Tx in my sketch
09:00 AM ayjay_t: https://twitter.com/misterpikul/status/915207115257638913
09:00 AM ayjay_t: it's growing
09:01 AM rue_bed: you need some smaller wire
09:02 AM rue_bed: scsi cable
09:03 AM ayjay_t: howcome? Just to make it more practical/easier to work with?
09:04 AM Haohmaru: henkannie are you sure the LED is actually wired to the RX pin?
09:18 AM ayjay_t: rofl i might try soldering the programmer directly to an xmega
09:24 AM polprog: ayjay_t: some proto?
09:26 AM ayjay_t: i'm trying to build a prototype at the moment
09:27 AM ayjay_t: i have a ancestor board but i think there are too many external components to test the programmer on it
09:27 AM ayjay_t: i might not have considered signal integrity of the programming lines when i built this prototype because i usually use nxp arm chips and they just use a slow uart connection so its nbd
09:28 AM ayjay_t: but it seems like atmels programming is more sensitive
09:28 AM ayjay_t: nxp kinetis has a bootloader kinda like lufe but it's rom based
09:28 AM polprog: you can always slow it down if using avrdude
09:31 AM ayjay_t: i just added a ton of bypass capacitors to the power lines and that manages to reduce a ton of noise
09:31 AM ayjay_t: i had literally just put the avr on the board and solder connections on for the programmer without anything else
09:37 AM ayjay_t: its all XPRG_CMD_PROGMODE failed htough. and i'm trying 5.11.1, 5.9, and 6.3 patched
10:08 AM ayjay_t: actually though when i lower all the bitrates, -B10, -b 9600, and i scope it, i don't notice a different :-|
10:08 AM Haohmaru: what is your programmer?
10:09 AM Haohmaru: if you're with the avrisp2 clone, then use avrdude 5.1 iirc
10:09 AM Haohmaru: cuz the programmer itself has USB-related issues in newer versions of avrdude
10:10 AM Haohmaru: i use the olimex avrisp2 clone and PDI works fine with avrdude 5.10
10:10 AM ayjay_t: i tried to compile avrdude 5.1 but it wants a lesser libtools
10:10 AM ayjay_t: i'm using an olimex avr isp mk2
10:11 AM polprog: oh yeah
10:11 AM ayjay_t: i updated the firmware on it too
10:11 AM polprog: I just found out Dragon supports PDI :>
10:11 AM polprog: http://www.atmel.com/webdoc/avrdragon/avrdragon.pdi_description.html
10:12 AM Haohmaru: there's a patch for the first problematic version of avrdude after 5.10, which supposedly also works on newer versions too
10:12 AM ayjay_t: okay wait
10:13 AM ayjay_t: so, i applied that patch to 6.3, but i didn't apply it to 5.11
10:13 AM ayjay_t: i'll try applying it to 5.11
10:14 AM ayjay_t: it fails on 5.11
10:14 AM ayjay_t: the patch, i think its 6.x.x
10:15 AM Haohmaru: i think the patch should definayely work on avrdude 6.0.. you should find the bug report discussion where the patch comes from, there are some better details
10:15 AM Haohmaru: gotta split
10:16 AM ayjay_t: i'm going to try windows /shrug
10:50 AM ayjay_t: meh commercial tools shouldn't be this difficult
10:50 AM ayjay_t: tough market, i guess
10:50 AM ayjay_t: brb
11:14 AM ayjay_t: alright well at least atmel studio recognized the olimex!
11:14 AM ayjay_t: although that wasn't exactly straight forward either *concern*
11:15 AM Lambda_Aurigae: olimex avr-isp-mk2 is a clone..not exactly OEM commercial device.
11:15 AM ayjay_t: OMG
11:15 AM ayjay_t: you have to switch the jumper to power the device to get it to wokr
11:15 AM learath: heh, just make sure you don't do that with the wrong device
11:15 AM learath: or you'll fry your programmer.
11:16 AM learath: (speaking from experience)
11:16 AM ayjay_t: wait how do you fry the programmer trying to power a target?
11:16 AM learath: ....
11:16 AM ayjay_t: no i'm serious
11:16 AM learath: target tries to draw too much power
11:16 AM learath: magic smoke goes -> that way
11:16 AM Lambda_Aurigae: looping back vcc to gnd.
11:16 AM ayjay_t: oh yeah if it was shorted
11:17 AM ayjay_t: it's hard to imagine someone would do that!
11:17 AM learath: I think mine was trying to drive a motor?
11:17 AM learath: (and yes, it was dumb)
11:17 AM ayjay_t: so should i go back to linux and make sure it all works?
11:17 AM polprog: ...
11:17 AM polprog: rip usb
11:17 AM ayjay_t: RIP
11:17 AM learath: nah the usb has a polyfuse on it
11:17 AM ayjay_t: i might just have to make a video or something of this, it took me 5 hours to set up this freaking programmer
11:17 AM learath: but the transistor blew to save the polyfuse!
11:18 AM ayjay_t: and it all came down to a freaking jumper
11:18 AM ayjay_t: i remember once i took apart my whole office to find a bug and i just had a usb cable hanging loose from the back of the computer
11:28 AM antto: yes, the olimex avrisp2 clone is a clone.. sort of.. made by the community.. it has jumper for voltage selection (5 or 3.3V) and whether you want to power the target from the programmer
11:28 AM antto: so, be careful
11:28 AM antto: keep that jumper to the OFF position unless you feel confident ;P~
11:30 AM ayjay_t: i already messed it up
11:30 AM ayjay_t: and powered the xmega with a abs max v @ 5v
11:30 AM ayjay_t: didn't break the chip tho
11:30 AM ayjay_t: i mean, the xmega has an abs max @ 3.3v, i gave it 5v
11:31 AM ayjay_t: i did have some nor flash on board that i'm worried about tho
11:33 AM Cracki: it might have a protection diode
11:33 AM antto: oh, xmegas are 3.3V
11:33 AM ayjay_t: i mean, its my board so, unless the chip itself has a protection diode then
11:34 AM antto: i thought you knew that
11:34 AM ayjay_t: pray for the nor flash.
11:34 AM Cracki: hehe
11:34 AM antto: xmegas are great
11:35 AM antto: pls microshit, don't f*ck em up, pls
11:35 AM ayjay_t: eh
11:35 AM Cracki: relax, they keep the avr cores and add more powerful peripherals
11:36 AM ayjay_t: okay back to linux and to test the nor flash and then avr dude
11:36 AM Cracki: they put an event system in the new tinys, which I believe is something that has only been done in xmegas before
11:36 AM antto: yes, and i'm pretty sure atmel came up with that, not microshit
11:46 AM ayjay_t: the nor survived the five volt overvoltage!
11:49 AM ayjay_t: avr dude works!
12:13 PM Cracki: treefiddy volts
12:21 PM thardin: what is wrong when avr-gcc doesn't understand inline asm?
12:22 PM polprog: thardin: how are you writing it
12:22 PM polprog: ?
12:23 PM thardin: asm volatile("jmp 0");
12:23 PM thardin: same as in my bootloader
12:23 PM thardin: -std=gnu99 did not help
12:23 PM thardin: I intend to use a different target address of course
12:23 PM polprog: you forgot to define the register you use and clobbers. even when empty gcc want them
12:23 PM mmfood: don't know if this is the place to ask, but what the heck: Would two of these in series work as inductors for a buck converter? https://www.farnell.com/datasheets/2339117.pdf
12:23 PM mmfood: I need to fit it between two boards with a distance of 8.5 mm so the bigger ones won't fit. It will drive an atmega328 and some multiplexed LEDs
12:24 PM polprog: so it's asm volatile("jmp 0"::);
12:24 PM polprog: also im not sure bout that jmp parameter, but try it
12:25 PM thardin: herp, now -std=gnu99 fixed it
12:25 PM thardin: still messing around currently. might need to do an ijmp with z holding the target
12:26 PM thardin: plus something to unwind the stack
12:26 PM thardin: what I'm trying to do is cleanly enter the bootloader from the application without losing the fact that it saw an 'S' coming in on USART
12:26 PM polprog: ah
12:27 PM thardin: have to do echo -n "S" > /dev/ttyUSB && sleep 1 && avrdude [...] right now
12:29 PM thardin: mmfood: assuming they meet whatever specs you need for the converter, sure. might need to orient them in a way such that you don't accidentally create a transformer
12:29 PM polprog: yeah that could be fun :P
12:29 PM polprog: to trouble shoot
12:30 PM thardin: accidental common mode choke :]
12:30 PM mmfood: thardin: how would I make sure to avoid that :p
12:30 PM polprog: i accidentally a transformer
12:30 PM thardin: think about how the B fields would orient themselves
12:30 PM thardin: are you standing these up or?
12:30 PM LeoNerd: I saw a bee in a field once
12:31 PM thardin: you could just put them physically in series. should be fine
12:31 PM thardin: if horizontal
12:32 PM thardin: but I doubt it's a huge issue
12:32 PM mmfood: oh ok
12:32 PM mmfood: B-field = some type of magnetic field type or what?
12:32 PM thardin: magnetic field yes
12:33 PM thardin: assuming any leaks out. depends on construction
12:41 PM thardin: personally I'm lazy so I'd just use a ready-made buck conveter module :]
12:44 PM ayjay_t: thardin: honestly, they'd be way better than anything you'd make
12:44 PM ayjay_t: i mean, unless you can lay down < half micron circuits and synchronous switching algorithms
12:44 PM ayjay_t: then hats off
12:45 PM ayjay_t: so i guess my point is, it should be "personally, i'm smart so I'd just use a ready made buck convert module :]"
12:45 PM * ayjay_t came off wayyy to negative in tone there
12:46 PM thardin: well I don't think myself cleverer than the engineer that spent who knows how many months full time optimizing a 7805 drop-in buck
12:47 PM thardin: but as always, It Depends. 4€ or whatever is low enough that I can't justify not spending it
01:09 PM JanC is now known as Guest6853
01:09 PM JanC_ is now known as JanC
01:49 PM carabia: https://en.wikipedia.org/wiki/NodeMCU
01:49 PM carabia: every day we stray further away from god
01:52 PM polprog: we killed god a long time ago: https://en.wikipedia.org/wiki/NanoVM
01:53 PM polprog: ;) may try that once
02:13 PM thardin: there's a jvm for amiga even
02:19 PM polprog: "jamiga"
02:19 PM polprog: is it because it jams all the time?
02:19 PM thardin: jamigaaaah
02:20 PM thardin: perhaps. I imagine it can run quite decent on a fully kitted out machine
02:20 PM polprog: yeah, probably
02:20 PM thardin: i should add that I'm not really an amiga guy. but I know some. they're.. interesting
02:21 PM polprog: me neither, sa one once on some retro gaming con
02:21 PM polprog: of which retro gaming was a single room, the rest was some indie game stands
02:21 PM polprog: becuase VR is a perfect thing to go along with a retro gaming con
02:22 PM thardin: VR on amiga
02:22 PM thardin: used to be a thing
02:22 PM polprog: VR is older than most people think
02:24 PM NoHitWonder2: was jurassic park made with amiga?
02:24 PM thardin: stereoscopic films used to be a thing back around the early 1900's if I'm not mistaken
02:24 PM NoHitWonder2: i heard that somewhere
02:24 PM thardin: NoHitWonder2: wouldn't surprise me. or maybe SGI machines
02:26 PM polprog: Voice commands were experimented with at mit in the 80's and then they were with us all through till today, i remember some old nokias (but new enough to run symbian i think?) had a feature that allowed you to dictate the contact name or number to be dialled
02:26 PM polprog: https://www.youtube.com/watch?v=RyBEUyEtxQo
02:26 PM polprog: now everybody is hyped for alexas and stuff
02:27 PM polprog: this is the bees knees. https://www.youtube.com/watch?v=SP8wSw4bBuA
02:28 PM thardin: AI becomes hip about once per generation
02:28 PM polprog: let's hope so
02:28 PM thardin: I'm still perplexed why NNs are so huge suddenly. used to poke with that 2002
02:29 PM polprog: as you can see on the attached video material, you dont need to have super-duper graphic effects like hair animation to have lots of fun
02:29 PM polprog: that low poly graphics could be probably done on a bigger avr :P
02:30 PM polprog: i wonder if they had to clean the puke from the flight sim in 2:40
02:31 PM polprog: maybe the differencenow is that more people can afford a vr kit than back then
02:31 PM polprog: not sure how much was an arcade machine like that worth
02:32 PM polprog: ca
02:32 PM polprog: oops missed a slash
02:32 PM polprog: ;)
02:35 PM thardin: flat shading was fairly standard at the time. the amiga had some nifty hardware acceleration for it even
02:37 PM thardin: crazy people (like britelite) have managed to do fully textured 3d on the
02:37 PM thardin: m
05:49 PM Emil: thardin: because back in 2002 you couldn't spread information like you can today
05:49 PM Emil: And the "tech is hip" hadn't reach the point it is at now
05:49 PM Emil: Neural Networks, machines learning, AI
05:49 PM Emil: Old ideas and subjects
05:50 PM Emil: but with the availability of todays processing power
08:46 PM rue_bed: c2=c1+- (d*sqrt(a^2+b^2))