#avr Logs
Mar 08 2017
#avr Calendar
12:21 AM rue_house: XOR IT WITH ITSELF!!!!
12:23 AM rue_house: oh :( its past bedtime
12:23 AM rue_house: what did I do?
12:23 AM rue_house: I was playing with the amplifier I pulled out of a scrap metal pile and replaced a diode to make work
12:23 AM rue_house: thats it?
12:23 AM rue_house: ... yea
12:23 AM rue_house: I got some rectangular RGB leds, the urge to make an rgb bar graph is almost overwelming
12:24 AM rue_house: it bedtime
12:24 AM rue_house: I have to sleep for 8 hours
12:24 AM rue_house: but, how do I make it productive?
12:24 AM rue_house: I wonder if I can master sleep-coding
12:24 AM rue_house: I mean, code cant be 100% syntax errors and typos can it?
12:25 AM rue_house: ECHO
12:25 AM rue_house: Echo
12:25 AM rue_house: echo
12:42 AM * Casper gives a mixture of high concentration of caffeine and sugar to rue_house
12:48 AM rue_house: nonon sleep sleep I need sleep _I_ need SLEEP
12:48 AM rue_house: sleep sleep sleep sleep sleep
12:48 AM rue_house: peels peels peels peels peels
12:48 AM rue_house: sleep sleeep sleep
12:52 AM Casper: caffeine
04:29 AM cousteau`: o/
04:31 AM cousteau`: does AtMega have bit access? can I access individual bits within a byte without reading the whole byte, replacing the bits, and writing the byte?
04:31 AM cousteau`: (asking because config bits)
04:32 AM cousteau`: ...then again, AVR doesn't even define bitfield structs like TI's Piccolo does, so I guess not
04:34 AM cousteau`: Google says that CodeVision does allow you to do REGISTER.0 = 1 to set bits (definitely not C!), but whether or not this translates directly as a "write bit" instruction is another thing
04:39 AM specing: s/AVR/avr-libc/
04:46 AM cousteau`: yeah that
05:14 AM eballetbo: by chance there is here any of the avrdude maintainers?
05:56 AM CORDIC: cousteau`: Yes, AVR does have bit-wise instructions. `and', `or', eor', `com
05:56 AM CORDIC: ' and few others.
05:57 AM CORDIC: C _programming language_ has bit fields.
06:37 AM cousteau`: CORDIC: but I mean, there's no way for the AVR to know that bit N was modified and the rest weren't
06:38 AM cousteau`: asking because the doc says that "writing to 1 on this bit blah blah blah", and I wanted to ensure that reading the register, writing a 1, and writing it back was OK
06:40 AM NoHitWonder: its ok
06:40 AM NoHitWonder: its how you set bits with avrs
06:40 AM NoHitWonder: or clear
06:40 AM cousteau`: ok then, wanted to make sure :)
06:43 AM cousteau`: How should I configure the PWMs? What I'm doing is: write 10000011 on GTCCR (General Timer/Counter Control Register), which iiuc stops the timers, then configure their registers (including a part where I write to their count value in order to set the phase between two PWMs), and then release the reset on the timers by writing 00000011 on GTCCR
06:49 AM Lambda_Aurigae: if you are working in assembly there is CBI and associated command SBI for clearing and setting bit in io register.
06:50 AM Lambda_Aurigae: and sometimes gcc is smart enough to use that and/or one of several other bit manipulation commands.
06:51 AM Lambda_Aurigae: CBR and SBR for clearning and setting bits in general purpose registers
06:55 AM Jartza: https://github.com/rakettitiede/octapentaveega/blob/master/vga.asm
06:55 AM Jartza: oops
06:56 AM cousteau`: Lambda_Aurigae: oh I see... damn, maybe that means this is wrong
06:56 AM Jartza: cousteau`: you should really learn to set and clear individual bits, writing that kind of constant values to registers is nasty.
06:57 AM Jartza: and not very readable code
06:57 AM Jartza: if you find handling bits in registers hard, then I have something for you
06:57 AM Jartza: https://gist.github.com/Jartza/29f70ad559a0f1e7b6ae
06:57 AM cousteau`: I could do it in a more bitwise way, but instead I'm just writing comments
06:57 AM Jartza: there are some macros
06:58 AM Jartza: with those you could write BITS_SET(GTCCR, TSM, PSRASY, PSRSYNC) for example
06:59 AM Jartza: which would then set bits TSM, PSRASY, PSRSYNC in GTCCR register
06:59 AM Jartza: or, you could make a "variable" out of single bit in register
07:00 AM Jartza: for example, if you want to just handle PB2 in PORTB for example
07:00 AM Jartza: let's say you have LED there
07:00 AM Jartza: #define LED BITP(PORTB, PB2)
07:00 AM Jartza: then you could just write LED = 0;
07:00 AM Jartza: or LED = 1;
07:01 AM Jartza: and that will compile into single SBI or CBI
07:03 AM Jartza: I use those macros all the time with AVR
07:03 AM * cousteau` . o O ( instead of so many BSO_x() macros I'd make a single variadic macro )
07:03 AM cousteau`: #define BSO_(a,b,c,d,e,f,g,h,...) (0xFF & (1<<(h) | 1<<(g) | 1<<(f) | 1<<(e) | 1<<(d) | 1<<(c) | 1<<(b) | 1<<(a))) #define BSO(...) BSO_(__VA_ARGS__, 8,8,8,8,8,8,8,8)
07:03 AM NoHitWonder: oh
07:04 AM NoHitWonder: i didnt even know about these macros
07:06 AM Jartza: no wonder you didn't know about them because I made them :P
07:06 AM NoHitWonder: PORTB |= (1 << 7); does this translate to SBI?
07:06 AM Jartza: haven't really advertised it anywhere
07:06 AM Jartza: NoHitWonder: yes
07:07 AM NoHitWonder: okay good
07:07 AM cousteau`: anyway, if I felt more elegant I'd do TCCR0A = 0<<COM0A0 | 2<<COM0B0 | 3<<WGM00; or something like that
07:08 AM cousteau`: and, in any case, I'm configuring all bits in the registers, not toggling a few; otherwise I guess |= << would be better indeed
07:10 AM Jartza: there's BITS_SET_EX also
07:10 AM cousteau`: also, &=~ and |= don't allow me to set a two-bit field to 2
07:10 AM Jartza: I also really, really hate those 2 << and 3 <<....
07:10 AM cousteau`: Jartza: then how do you set a 2-bit field?
07:10 AM Jartza: because every bit has name in AVR
07:10 AM Jartza: just set the bits
07:10 AM cousteau`: bit by bit?
07:10 AM Jartza: WGM00 and WGM01
07:11 AM Jartza: still makes it much more readable
07:11 AM cousteau`: that seems excessively verbose
07:11 AM Jartza: and same code size
07:11 AM Jartza: verbose is good
07:11 AM Jartza: verbose is friend
07:11 AM cousteau`: no, in this case it's less readable
07:11 AM Jartza: it's not
07:12 AM cousteau`: "set CS0 to 3" looks better than "set CS00 to 1, set CS01 to 1, and set CS02 to 0" IMO
07:12 AM Jartza: it does not
07:14 AM cousteau`: looks better to me
07:14 AM Jartza: with only that opinion you would fail many of the programming tests in interviews for embedded programming
07:14 AM cousteau`: anyway, let's go back to my original question
07:14 AM iwancoppa: CS0 to 3 is not immediately obvious which bits are set and which aren't imo
07:14 AM Jartza: indeed
07:15 AM Jartza: it's like...
07:15 AM cousteau`: ...and what if you want to set PSRASY and PSRSYNC simultaneously?
07:15 AM Jartza: which is better. setting "11" to CS10 of TCCR1, or setting CS10, CS12 and CS13 of TCCR1?
07:16 AM Jartza: especially when datasheets plenty of time have tables like this: http://i.imgur.com/nnPqqPh.png
07:16 AM Jartza: BITS_SET_EX(TCCR1, CS10, CS11, CS12);
07:16 AM Jartza: that's immediately clear what is set
07:16 AM cousteau`: yeah but that doesn't allow you to set it to 011
07:16 AM cousteau`: only 000 or 111
07:17 AM Jartza: ?
07:17 AM Jartza: yes that does
07:17 AM cousteau`: well, not in one step
07:17 AM NoHitWonder: i think this is the most readable way: UCSRC |= (1<<URSEL)|(1<<UCSZ1)|(1<<UCSZ0);
07:17 AM NoHitWonder: to me at least
07:17 AM Jartza: that BITS_SET_EX sets CS10, CS11 and CS12 to 1 and all the other bits to 0
07:18 AM Jartza: NoHitWonder: that would be with my macros BITS_SET(UCSRC, URSEL, UCSZ1, UCSZ0);
07:18 AM NoHitWonder: yeah
07:19 AM cousteau`: Jartza: on a completely tangent topic, have you considered the variadic macro suggestion I mentioned?
07:19 AM NoHitWonder: does that translate to read/or/write?
07:20 AM NoHitWonder: Jartza
07:20 AM Jartza: NoHitWonder: that translates to what you wrote. and then that translates to whatever compiler thinks is optimal.
07:20 AM Jartza: cousteau`: didn't really occur to me to try shifting bit 8 times :)
07:21 AM NoHitWonder: well that woulb be read->or->write
07:21 AM NoHitWonder: since it cant use sbi
07:21 AM Jartza: in case of multiple bits, usually yes
07:21 AM cousteau`: well I was going to use zeros, then I realized I was an idiot, then considered using -1, then realized I was an idiot again, and finally went with 8
07:22 AM Jartza: yeah, looks like gcc at least optimizes 1<<8 away if writing to 8bit register
07:22 AM Jartza: although that's still a bit hacky :)
07:27 AM cousteau`: https://gist.github.com/cousteaulecommandant/baf5732fb801d6a8adcea61392948ee5 if you want it made :) (can I make a pull request on a gist?)
07:30 AM NoHitWonder: oh man, there is simulator in ATMEL Studio 7 and Disassembly window
07:30 AM NoHitWonder: i thought these were avr studio 4 only features
07:31 AM Lambda_Aurigae: NoHitWonder, they have been there a long time.
07:31 AM Lambda_Aurigae: too bad it's windows only.
07:32 AM Lambda_Aurigae: hopefully microchip will fix that problem
07:33 AM Lambda_Aurigae: maybe they will integrate avr into mplabx!
07:33 AM Lambda_Aurigae: hehe
07:38 AM Jartza: NoHitWonder: atmel studio simulator is actually very good
07:38 AM Jartza: up to so good that if you accidentally define chip to use external clock, it'll just halt forever :D
07:44 AM NoHitWonder2: i love this dissambler window http://pastebin.com/7qbMmC2c
07:45 AM Lambda_Aurigae: I can get that right out of gcc.
08:12 AM rubend: hello everyone
08:51 AM cousteau: ok, so the timers are doing something weird
08:54 AM cousteau: I stop the timers by holding the corresponding reset on GTCCR, then set some parameters, write to OCRnA/OCRnB, then *write TCNTn in order to set the phase*, and finally release the reset
08:55 AM cousteau: apparently, if I write a non-0 value to TCNTn, *the counter does not stop at OCRnA*, but instead runs up to MAX (0xFF)
08:56 AM cehteh: is tcnt biggier than ocr?
08:57 AM cehteh: (withhin some marigin, i remember something in the datasheet)
08:57 AM cousteau: so for a pulse width of 100 (OCRnA=99), if I want TC2 to be delayed 50 wrt TC1, I have to write 256-50 on that register, not 100-50
08:58 AM cehteh: havent followed your discussion, but looks like the timer is somewhat wrong configured
09:35 AM yids_ is now known as yids
09:50 AM cousteau: cehteh: the code is http://ideone.com/acktSd
09:51 AM cousteau: I have to set the counter to (MAX+1)-delay instead of (COUNT+1)-delay. It seems that the first time after setting the value the compare is skipped.
09:52 AM cousteau: The doc says "Writing to the TCNT0 Register blocks (removes) the Compare Match on the following timer clock." but I understand that means "on the next clock cycle", not "on the next compare match"... right?
09:59 AM cehteh: i think that means a full timer spin
10:00 AM cehteh: order of initialization is important there
10:08 AM cousteau: do you mean, order in which I assign the registers?
10:12 AM cehteh: and starting the timer (how did you stop it? prescaler to 0?)
10:13 AM cehteh: i cant remember the exact procedure, would have to read the datasheet too
10:15 AM cousteau: I stop it by setting and holding the prescaler reset
10:15 AM cousteau: and then release the reset once I'm done
10:15 AM cousteau: It seems the OCRnx registers are double buffered! Brb, reading about that
10:21 AM cousteau: Fast PWM -- Update of OCR0x at: BOTTOM
10:21 AM cousteau: does that mean even if the prescaler clock is disabled, or it must be running?
11:41 AM cousteau: ...so writing a 0 to TCNTn doesn't seem to force an update of the OCRnx
11:46 AM cehteh: tried a -1?
11:48 AM * cehteh is busy with other things, i didnt read the datasheet
11:48 AM cehteh: iirc new comparator values are latched when the timer overflows
11:49 AM cehteh: (depending on timer mode)
11:51 AM skz81: <cehteh> tried a -1? >> will write 0xFF to the register but nor resolve the double buffering thing
11:52 AM cehteh: that should immediately overflow and latch the buffer, my guess .. i still didnt read the datasheet :D
11:57 AM cousteau: Seems that forcing the OCRnx to update the double buffer is impossible in FastPWM mode. That only happens immediately in CTC mode.
11:58 AM cousteau: So I'm setting it in CTC mode, updating the OCRnx registers, and then setting it back to FastPWM mode. :P
11:58 AM cousteau: And then enable the prescaler so that it starts running.
11:58 AM cousteau: This is hacky as hell, but it's working.
11:58 AM cehteh: do you use the GTCCR.TSM flag to sync / start the timers?
11:59 AM cehteh: and do you really need fast pwm? other modes might be more approbiate when you dont need that much speed
11:59 AM cousteau: yes, I set that to 1 (and the other 2 as well), do the changes, and then set that to 0 so that the other 2 are released by the hardware
11:59 AM cousteau: I need accurate synchronous PWMs
11:59 AM cehteh: what pwm freq do you need?
12:00 PM cousteau: well, the idea is that this can implement arbitrary periods/phases/duties, so I can't know
12:01 PM cehteh: in what ballpark?
12:03 PM cehteh: one timer or multiple timers in sync? (how much pwm's?)
12:05 PM cousteau: multiple timers in sync
12:05 PM cousteau: that's why I need to specify the delay between timers
12:06 PM cousteau: (with this AtMega, I can have up to 3 independent PWMs; the other 3 PWM outputs would be restricted)
12:09 PM cehteh: depends on the atmega, but i guess 328p
12:10 PM cehteh: so you dont want to start all timers at once but have some time delay between?
12:22 PM cehteh: cousteau: see table 17-8 in the datasheet
12:22 PM cehteh: only normal and CTC modes dont latch ocr registers
12:27 PM cousteau: cehteh: yep, that's what I ended up figuring out
12:27 PM cehteh: one look at the datasheet :D
12:28 PM cousteau: so what I do is set the mode to CTC, set the OCRnx registers (which get updated immediately), and then set FastPWM mode
12:28 PM cehteh: why?
12:28 PM cehteh: cant you stay in CTC?
12:28 PM cousteau: well, CTC does not generate PWMs directly, it's more for interrupts and the like
12:29 PM cehteh: mhm you cant enable the output pins in ctc?
12:30 PM cehteh: table 17-5 :D
12:30 PM cehteh: and 17.2 rather
12:30 PM cousteau: ...what datasheet are you reading? Mine is ordered differently. (It's the one that says only 328/P, not 48/88/168/328)
12:30 PM cehteh: could be
12:31 PM cehteh: you can connect the output pins in non pwm mode too
12:31 PM cehteh: just try
12:32 PM cousteau: ...I think in Normal and CTC modes it only sets the signal but doesn't clear it (or the other way around)
12:32 PM cousteau: wait, I'll check the datasheet
12:33 PM cousteau: see Table 19-3. Compare Output Mode, non-PWM
12:33 PM cousteau: (yours will be different though)
12:34 PM cousteau: As I suspected; if I leave it in CTC mode it won't set automatically
12:37 PM cousteau: in CTC mode, OCnx just gets set/cleared/toggled when TCNTn = OCRnx, depending on what you write on COMnx. You were right in that it *does* generate a signal... but it's not useful as PWM
12:37 PM cousteau: unless I want a PWM with a duty cycle of 50% and half the frequency
12:37 PM cousteau: Anyway; my ugly workaround works.
12:37 PM cehteh: i bet it works better
12:38 PM cousteau: what, the workaround, or the... other options? (I didn't find any; I can only think of using interrupts)
12:39 PM cehteh: i have to read it too
12:44 PM cousteau: http://codepad.org/ORZJdtfu in case you want to see it
12:44 PM cousteau: I was too lazy to write the register info in a pretty way
01:27 PM _ami_: linking static libraries to executable does not make code slower?
01:28 PM _ami_: i think .a inclusion is similar to compiling the library source files with executable (main.cpp) code.
01:29 PM cehteh: try -flto (and read about how it works)
01:29 PM cehteh: i have good success with that
01:31 PM cehteh: usually gcc optimizes per compilation unit, iirc it doesnt matter if you gcc main.c lib.a or gcc main.c lib.c
01:31 PM NoHitWonder2: i have atmega1284 and i want to debug it with atmel studio, i also have atmel ice. what interface would you recommend?
01:34 PM _ami_: thanks cehteh. gonna try this.
01:34 PM cehteh: note that you need to recompile your lib with -flto as well
01:39 PM _ami_: cehteh: i have a toolchain cmake file where i set the linker flags and that sets it for lib and executable..
01:39 PM _ami_: i hv already used -Os so adding -flto did not reduce the size.
01:39 PM _ami_: set(CMAKE_EXE_LINKER_FLAGS " -flto -T ${CMAKE_SOURCE_DIR}/libopencm3.ld -nostartfiles -lopencm3_stm32f1 -lc -lnosys --specs=rdimon.specs -Wl,--gc-sections" CACHE STRING "")
01:40 PM _ami_: Scanning dependencies of target proj1-size
01:40 PM _ami_: text data bss dec hex filename
01:40 PM _ami_: 1220 12 0 1232 4d0 proj1.elf
01:40 PM _ami_: same size for both cases. so you were right abt that.. -Os does the magic
01:44 PM _ami_: time to sleep
01:44 PM _ami_: nn!
02:59 PM hetii: Hello :)
03:01 PM NoHitWonder2: hello
03:12 PM hetii: hmm I wonder what is the best way to exchange data between avr like atmega8/attiny85 and my odroid pc.
03:13 PM hetii: Currently I use spi pooling in slave mode: http://paste.ubuntu.com/24141458/
03:14 PM hetii: but wonder if I will use some bit-banging and more bus wire get better performance
03:14 PM specing: uart?
03:14 PM hetii: well uart is lower then spi I guess
03:17 PM hetii: *slower
03:21 PM specing: there is no speed limit for uart
03:23 PM hetii: ?
03:24 PM hetii: ok so what will exchange data faster,when use uart set to 1Mbps or SPI ?
03:25 PM hetii: imho uart consume more bits to send the same amount of bytes
03:28 PM specing: hetii: yes, but uart transfers are asynch and use one wire less
03:30 PM specing: so instead of SPI clock you can run another uart channel there
03:30 PM hetii: as I sad, its not the point to use additional wire but with maximum speed that I can use to exchange data.
03:32 PM specing: and as I said, there is no inherent maximum speed
03:32 PM Rickta59: what is the max speed of spi on the odroid?
03:34 PM Rickta59: is dns involved?
03:34 PM Rickta59: * ignore errant history paste
03:35 PM Rickta59: what is the max speed of spi on the odroid? @ hetii
03:43 PM hetii: Rickta59: I found: "It can't be over 30Mbps with a single data line."
03:44 PM Rickta59: seems like spi would be the winner then
03:44 PM Rickta59: and it has the added advantage is that you can just pause without breaking the data stream
03:44 PM arij is now known as arij_work
03:47 PM hetii: yep, but as I read avr in spi slave mode is able to talk just with speed like fosc/4
03:47 PM hetii: and based on this: https://www.tablix.org/~avian/blog/archives/2012/06/spi_interrupts_versus_polling/
03:48 PM hetii: polling is the best choice if I don`t need to do anything else
03:48 PM Rickta59: how much data are you transferring? is it larger than ram?
03:51 PM hetii: I control 155 rgb leds
03:51 PM Rickta59: time sensitive leds or clocked?
03:52 PM hetii: so 155*3 give us 465 just for rgb data.
03:52 PM hetii: they are time sensitive: ws2812b
03:54 PM hetii: General what I have now is a bridge between ws2801 protocol and ws2812b
03:54 PM arij_work is now known as arij
04:04 PM Hfuy: Hello.
04:05 PM Lambda_Aurigae: olleH
04:08 PM JanC_ is now known as JanC
04:11 PM Hfuy: Does anyone make a module with an adjustable amplifier and DC offset for routing audio into an AVR?
04:11 PM Hfuy: I've looked around, but all I can find is mic amps.
04:11 PM Hfuy: Application is reading linear timecode, which is an audio signal containing serial data.
04:13 PM Hfuy: I could probably just DC offset it with a divider, but I'd like to have some sort of buffering to better deal with variable impedance inputs.
04:14 PM daey_ is now known as daey
04:14 PM NoHitWonder__: op amps
04:14 PM Hfuy: I suspected they might be involved.
04:15 PM Lambda_Aurigae: lm741 or lm386 for the win?
04:15 PM Hfuy: Well, yes.
04:15 PM Hfuy: I'd prefer to find something prebuilt. I guess I could use a mic amp board and alter the feedback resistor to reduce gain.
04:16 PM Lambda_Aurigae: lm386, 2 caps, 2 resistors...
04:16 PM NoHitWonder__: 4 resistor
04:16 PM NoHitWonder__: s
04:16 PM Lambda_Aurigae: hmmm...maybe yeah..
04:16 PM Lambda_Aurigae: it's been a while since I built such.
04:16 PM NoHitWonder__: http://www.radio-electronics.com/images/op-amp-inverting-circuit-single-rail.gif
04:17 PM Hfuy: It's not really a line amp. It's just a buffer with a bit of level control.
04:17 PM Hfuy: Nobody seems to make a board for it.
04:17 PM Lambda_Aurigae: NoHitWonder__, yeah..I always forget about that feedback resistor and never use the input resistor.
04:17 PM Lambda_Aurigae: Hfuy, because everybody just builds their own into the circuit.
04:17 PM Hfuy: I was getting that impression.
04:18 PM Hfuy: The reason for sticking on this is that I'm trying to create a design that non-experts can make by plugging together readily-available stuff.
04:18 PM Hfuy: This is also helpful due to my non-expert status.
04:19 PM Hfuy: I suppose in extremis I could just stick it through a diode. I don't think it would affect the readability.
04:19 PM Hfuy: Unfortunately, line level audio is only likely to be around a volt, and I don't know what the avr's input impedance is.
04:19 PM arij is now known as arij_work
04:20 PM Hfuy: I think a large part of why I wanted a module for this is to avoid requiring a dual rail PSU, as well.
04:20 PM Lambda_Aurigae: there should be a low voltage reference for the ADC that should read line level just fine.
04:21 PM Lambda_Aurigae: who needs dual rail?
04:21 PM Hfuy: Many op-amp designs.
04:22 PM Lambda_Aurigae: the one NoHitWonder__ posted is single rail
04:22 PM Hfuy: Despite the fact that I actually need the damn thing to offset it to 2.5v
04:22 PM Hfuy: Grr.
04:23 PM NoHitWonder__: that circuit does that
04:23 PM Hfuy: Are there any values for the Rs and Cs, and a type number for the op-amp?
04:23 PM NoHitWonder__: r3 = r4
04:24 PM Hfuy: One could practically build that on one of those IC breakout boards.
04:24 PM NoHitWonder__: r1 = r2 if you dont waint voltage gain
04:24 PM Lambda_Aurigae: if one bothered to do a google search I bet there's even a page or twelve that has calculators for all those values.
04:24 PM NoHitWonder__: just look around, you will find correct values
04:25 PM Hfuy: One of them should ideally be a pot
04:25 PM NoHitWonder__: what for? voltage gain?
04:26 PM Hfuy: Yes, probably
04:26 PM Hfuy: Just so it can deal with a bit of variability in the signal
04:26 PM NoHitWonder__: yeha
04:26 PM NoHitWonder__: *yeah
04:31 PM Hfuy: I'm seeing a lot of NE5532 headphone amp boards on ebay.
04:31 PM Hfuy: I wonder if they'd do it.
04:32 PM Hfuy: Problem is, it'd still be a negative-going signal
04:34 PM Lambda_Aurigae: https://hackaday.io/project/20161-breadboard-vga-from-ttl-and-eeprom for Jartza
04:37 PM NoHitWonder2: negative-going signal? what do you mean by that?
04:39 PM Hfuy: Audio signals tend to be centered around ground.
04:40 PM Hfuy: I'm not actually sure how much that matters for this specific situation, but it's a problem with audio input to microcontrollers.
04:43 PM NoHitWonder2: i would do it with that inverting op amp
04:43 PM NoHitWonder2: only few components
04:43 PM NoHitWonder2: its a trivial circuit
04:44 PM Hfuy: Yes, I know what you mean. I'm just trying not to make boards for this.
04:44 PM Hfuy: Or, well, I'm definitely not going to make boards for this.
04:44 PM NoHitWonder2: and put pot in place of r4 and you can adjust the offset
04:44 PM Hfuy: R3 and R4, I assume.
04:45 PM Hfuy: Although really if it's a 5V device, they should both be the same.
04:45 PM Hfuy: Not sure what C2 is for.
04:46 PM NoHitWonder2: for keeping the 2.5 V (which is "ground") steady
04:46 PM Hfuy: Just sort of surprised nobody's made this. Audio into AVR (or at least arduino) is not a rare ask.
04:47 PM NoHitWonder2: its a decoupling cap
04:50 PM Hfuy: I suppose the only remaining question is whether the capacitors are (for instance) 0.01pF or 5F, and whether the resistors are 0.1 ohm or 5 megs.
04:53 PM NoHitWonder2: c1 needs to be big (enough), c2 0.1uF maybe
04:53 PM NoHitWonder2: just find a schematic for a guitar pedal and look from there
04:54 PM Hfuy: It's quite a difficult problem, because most designs assume bipolar in and out
04:54 PM NoHitWonder2: r3 = r4 = 100K
04:54 PM NoHitWonder2: most battery power guitar pedal are single rail
04:56 PM Hfuy: R2 controls gain, I guess.
04:56 PM Hfuy: But I'm not sure what the value should be.
04:57 PM Hfuy: Does R1 just increase the input impedance?
04:58 PM NoHitWonder2: https://en.wikipedia.org/wiki/Operational_amplifier_applications#Inverting_amplifier
04:59 PM NoHitWonder2: i think they can be 100K
05:02 PM Hfuy: Perhaps this could be built with LM741 or equivalent.
05:03 PM Hfuy: At least, I'd like to use the commonest possible parts
05:03 PM NoHitWonder2: yes
05:04 PM Hfuy: I once had some very useful little protoboards which had a DIL layout in the middle and a scattering of connected pads.
05:04 PM Hfuy: they were very useful for this, but I can't think of how to google for them
05:06 PM NoHitWonder2: https://ae01.alicdn.com/kf/HTB1F6aDJVXXXXcWXFXXq6xXFXXXS/5pcs-lot-universal-Stripboard-font-b-Veroboard-b-font-vero-Board-Single-Side-4-8x13-3cm.jpg
05:06 PM NoHitWonder2: this?
05:06 PM Lambda_Aurigae: https://www.radioshack.com/products/multipurpose-pc-board?gclid=COaom5n8x9ICFVi4wAodH3gNIg
05:06 PM Lambda_Aurigae: just google for protoboard
05:07 PM Hfuy: NoHitWonder2: : Like that, but very small, probably 30mm-40mm square
05:07 PM Hfuy: they were very handy for this sort of thing
05:08 PM Lambda_Aurigae: https://www.radioshack.com/collections/breadboards-ic-sockets/products/general-purpose-ic-pc-board
05:08 PM Lambda_Aurigae: https://www.radioshack.com/collections/breadboards-ic-sockets/products/dual-mini-board
05:09 PM Hfuy: that's the one
05:09 PM Hfuy: the, er, "general purpose ic pc board"
05:09 PM Hfuy: Just right for this sort of thing
05:10 PM Hfuy: I once made about 25 4017 sequencer boards with 10-segment LED bargraphs with loads of those stacked on pillars, for sci fi props.
05:11 PM Lambda_Aurigae: I have all 3 of those in quantity here.
05:11 PM Hfuy: Actually a bit overkill for a DIL-8 based buffer amp, but it's about the minimum imaginable proto board.
05:11 PM Hfuy: Presumably you don't get them from radio shack.
05:11 PM Lambda_Aurigae: nope.
05:12 PM Lambda_Aurigae: other sources out there.
05:12 PM Lambda_Aurigae: those just happened to be the first links I found for them.
05:12 PM Hfuy: I think we did actually get them from (about four different branches of) radio shack but I wasn't paying :)
05:12 PM Lambda_Aurigae: I also have 8x10 inch boards with copper pads on every hole.
05:12 PM Lambda_Aurigae: that I can cut up to fit what I want.
05:12 PM Hfuy: Jesus, what are you prototyping, a Cray-1?
05:12 PM Lambda_Aurigae: I've had this stuff for YEARS
05:13 PM Hfuy: I first used those DIL breakouts in the 80s, when I was about 12.
05:13 PM Hfuy: Well, a theoretical maximum of nearly 12.
05:13 PM Lambda_Aurigae: but, back in the day,,,when I was 16,,,I used such to built a 4bit computer from 74xx series chips.
05:13 PM Hfuy: I was pleased to find they were still available in the late 90s. I haven't used them since.
05:13 PM Hfuy: Reencountering them now is like discovering an old friend.
05:13 PM * Hfuy wells up with nostalgia for LED sequencers
05:13 PM Lambda_Aurigae: hell, I still have an old radio shack etching kit...never opened.
05:14 PM Hfuy: Last time I went in a radio shack it was in LA.
05:14 PM Hfuy: And they didn't have what their website claimed they had.
05:14 PM Hfuy: Bitches.
05:14 PM Lambda_Aurigae: yeah.
05:14 PM Lambda_Aurigae: they turned into a cellphone store.
05:15 PM Lambda_Aurigae: they are filing bankruptcy again.
05:15 PM Lambda_Aurigae: probably gonna close more stores.
05:15 PM Hfuy: My impression was they were not sure what they were trying to be.
05:15 PM Lambda_Aurigae: think there is one store still open within 100 miles of me.
05:15 PM Hfuy: Either a hobbyist emporium or a... toyshop.
05:15 PM Lambda_Aurigae: cellphones and ardweenies and rc dinosaurs.
05:16 PM Lambda_Aurigae: that's what they had the last time I was in one.
05:16 PM Hfuy: I get the feeling their halfhearted attempt to be both is going to be doubleplusungood.
05:16 PM Lambda_Aurigae: they closed like 90% of their stores across the country a couple years back.
05:16 PM Hfuy: Somewhere around I have a 5mm flashing red LED from there.
05:16 PM Hfuy: In original packaging. It's probably worth $100 as 80s kitsch.
05:17 PM Hfuy: Here in the UK they were, for some reason, called Tandy, although the goods were all still branded Radio Shack. I guess it at least explains the name of the TRS-80.
05:17 PM Lambda_Aurigae: I also have every engineers mini notebook ever put out in the 70s and 80s
05:17 PM Hfuy: Forrest M. Mims, III!
05:18 PM Lambda_Aurigae: yeah...Tandy was a side company...they had tandy stores in aussieland too.
05:18 PM Hfuy: Suspect rights issues with the Radio Shack name.
05:18 PM Lambda_Aurigae: probably.
05:18 PM Hfuy: Also there's a company in the UK called RS, for Radio Spares, which might have been a concern.
05:19 PM Hfuy: When I was a kid, the fact that Tandy was really Radio Shack was a sort of open non-secret.
05:19 PM Lambda_Aurigae: yup.
05:19 PM daey_ is now known as daey
05:19 PM Lambda_Aurigae: I forget which was the parent company.
05:19 PM Lambda_Aurigae: for some reason I want to say Tandy was.
05:19 PM Hfuy: I didn't visit the US until much later, and when I first saw a real Radio Shack (same livery, same branding) it was something of a moment.
05:19 PM Lambda_Aurigae: TRS-80...Tandy Radio Shack
05:19 PM Hfuy: Yes, quite.
05:20 PM Lambda_Aurigae: first computer I ever worked with was a trs-80 model 16...
05:20 PM Hfuy: You're dating yourself.
05:20 PM Lambda_Aurigae: I had one of those nifty little tandy pocket computers in high school too.
05:20 PM Lambda_Aurigae: of course.
05:20 PM Lambda_Aurigae: I'm 49
05:21 PM Lambda_Aurigae: timex sinclair was the dream computer when I was a teenager.
05:21 PM Lambda_Aurigae: my first computer was a commodore vic-20.
05:21 PM Hfuy: I'm about ten years younger
05:21 PM Lambda_Aurigae: which,,,I still have,,,it still works...I still play games onit.
05:21 PM Lambda_Aurigae: connected to my projector with the 15 foot diagonal projection screen
05:21 PM Lambda_Aurigae: the pixels are an inch across.
05:21 PM Hfuy: I still want a three-gun CRT projector.
05:22 PM * Hfuy is a fan of obsolete yet amusing things
05:22 PM Lambda_Aurigae: I've torn a few of those apart for the optics.
05:23 PM Hfuy: and do what with?
05:23 PM Lambda_Aurigae: burn ants
05:23 PM Lambda_Aurigae: amongst other awesome sideline projects.
05:24 PM Hfuy: One of the things about those is that you can now get very good ones for very little money.
05:24 PM Lambda_Aurigae: I like magnifying glasses.
05:24 PM Hfuy: New tubes are an issue, but the images are very nice.
05:24 PM Lambda_Aurigae: yeah...I find mine by the roadside.
05:24 PM Lambda_Aurigae: or people give them to me if I haul them out of the basement.
05:24 PM Lambda_Aurigae: they usually work.
05:24 PM Hfuy: the main issue is that they weigh slightly more than a recently-collapsed star.
05:24 PM Lambda_Aurigae: or at least mostly.
05:24 PM Hfuy: I suspect the tubes are generally thrashed, no?
05:25 PM Lambda_Aurigae: not always.
05:25 PM Lambda_Aurigae: I've had 2 in the last 3 years that worked perfectly.
05:25 PM Hfuy: You have my envy, sir.
05:25 PM Lambda_Aurigae: even got the curved screens with them.
05:25 PM Hfuy: there's a chap in Canada who specialises in redoing them. I think he found a place that could recondition the tubes - snap the necks off, wash out the burned phosphor, and recoat.
05:26 PM Hfuy: Quite a process.
05:26 PM learath: that seems like a lot of work.
05:26 PM Hfuy: Some of them are in military flight simulators, so the cost of a system replacement is vast.
05:26 PM Lambda_Aurigae: for not a lot of return unless you can find someone who really really wants one.
05:26 PM Hfuy: I suspect that's what makes it worthwhile.
05:27 PM Hfuy: They do have some quite nice characteristics. Very good contrast, in some ways at least. Peak brightness tends to be a bit variable with picture content.
05:27 PM Lambda_Aurigae: you can play the old lightgun based games with them too!
05:27 PM Hfuy: yes!
05:28 PM Hfuy: Sadly none of them that I'm aware of will handle 4K.
05:28 PM Lambda_Aurigae: so?
05:28 PM Lambda_Aurigae: I don't have a single monitor, tv, or video card that handles 4k
05:28 PM Hfuy: We considered them for a grading suite (colour adjustment for film and TV work)
05:29 PM Hfuy: Lack of 4K was a downer. the variable peak brightness with different picture content was the killer. Otherwise they're better than most modern displays.
05:29 PM Hfuy: Just huge, heavy, power hungry, and a cow to set up :)
05:29 PM Lambda_Aurigae: that's why I keep that 21inch IBM monitor around.
05:29 PM Lambda_Aurigae: weighs more than a honda
05:29 PM Lambda_Aurigae: but has an absolutely awesome picture.
05:30 PM Hfuy: Well, it's really OLED world in postproduction now. Very hard to beat, but at professional levels, extremely expensive.
05:30 PM Lambda_Aurigae: https://www.cnet.com/products/ibm-p260-crt-monitor-21-series/specs/
05:31 PM Hfuy: I'm amused by the fact that the retro gaming community is picking up what were very high end Sony CRT video monitors to connect to 80s consoles.
05:31 PM Hfuy: Each pixel is being etched almost individually onto the screen. They seem to like that.
05:32 PM Lambda_Aurigae: 70 pounds!
05:32 PM Lambda_Aurigae: hehe
05:35 PM Hfuy: They were often 21" trinitron types
05:35 PM Hfuy: Suspect that IBM monitor you mention may be using the same 21" tube as their broadcast displays
05:36 PM Lambda_Aurigae: FD Trinitron
05:36 PM Lambda_Aurigae: that monitor can display basically anything up to SVGA.
05:36 PM Lambda_Aurigae: hercules mono, cga, ega, vga, etc...
05:36 PM Hfuy: I practically gave away a Sony BVM-20F1E a few years ago, before the retro gaming community went "Huh?"
05:37 PM Lambda_Aurigae: used something like the NEC multisync technology.
05:37 PM Lambda_Aurigae: feed it video and it'll display something.
05:37 PM Hfuy: Now they go for £1000 on ebay.
05:39 PM Lambda_Aurigae: used t have 3 of them that I had color matched
05:39 PM Lambda_Aurigae: but 2 of them died.
05:39 PM Lambda_Aurigae: drove them with a matrox parhelia triplehead video card.
05:46 PM Hfuy: Not sure what happened there. Damn webchat.
05:48 PM Lambda_Aurigae: I'm still here.
05:50 PM Lambda_Aurigae: some might consider that a negative,,,but,,
05:59 PM Hfuy: You're a CRT connoiseur. You'll do.
06:02 PM Hfuy: Can't find another source for those general-purpose DIL IC boards, though.
06:03 PM Hfuy: They appear variously but it seems to come back to Radio Shack.
06:03 PM Lambda_Aurigae: you looking for the one that has the dip in the middle, yes?
06:04 PM Hfuy: that's the one. I'll have a better look around once I'm not on my laptop, whch is so slow that opening another tab is time to get a coffee.
06:05 PM Lambda_Aurigae: yeah...those are likely a rat-shack only thing these days.
06:07 PM Lambda_Aurigae: https://www.circuitspecialists.com/64-8931.html
06:07 PM Lambda_Aurigae: just perfboard...no planned out dip socket
06:07 PM Lambda_Aurigae: but still usable.
06:07 PM Hfuy: I'm quite heartened by some of the good perf/vero style stuff that's available these days.
06:08 PM Hfuy: I got some very cheap stuff (some holes missed, not the ends of invention) which has several rows of adjacent DIL rows, with the power down the middle.
06:13 PM Lambda_Aurigae: http://www.digikey.com/product-detail/en/twin-industries/LED12-C2/438-1129-ND/3693591
06:13 PM Lambda_Aurigae: http://www.digikey.com/product-detail/en/twin-industries/8300SB2-LF/438-1138-ND/5119056
06:13 PM Lambda_Aurigae: http://www.digikey.com/product-detail/en/phoenix-contact/2947190/277-9475-ND/2528636 that's cute
06:14 PM Lambda_Aurigae: expensive though.
06:19 PM Hfuy: I liked the Rat Shack ones for the fact that there was just about room on them to knock up something like a 555 oscillator or preamp.
06:19 PM Lambda_Aurigae: yup.
06:20 PM Hfuy: Propmaking requires a lot of that sort of lashup, although I notice that a lot of modern throwaway things can be done with an arduino.
06:21 PM Hfuy: (AVRs will actually drive a lot more output current than the datasheet suggests they will. I'm not sure why.)
06:21 PM Lambda_Aurigae: ardweeny is the evil microsoft of the microcontroller world.
06:21 PM Lambda_Aurigae: of course they will.
06:21 PM Lambda_Aurigae: they will even survive dead shorts on the output for short periods.
06:22 PM Lambda_Aurigae: heck, they will survive having VCC and GND reversed.
06:23 PM Hfuy: I buy arduino clones by the handful simply because it's a useful AVR dev board with USB programming built in.
06:23 PM Hfuy: That's all.
06:23 PM Lambda_Aurigae: I'm just a bastard and won't buy anything with the arduino name on it.
06:23 PM Hfuy: I can't get that hardcore about it.
06:23 PM Lambda_Aurigae: I know it limits and things cost me more, but,,,
06:24 PM Hfuy: If I'm just flashing lights I'll use their libraries, because why not.
06:24 PM Lambda_Aurigae: I just do it as a hobby anyhow.
06:24 PM Lambda_Aurigae: and get lots of free samples.
06:24 PM Hfuy: Heh.
06:24 PM Lambda_Aurigae: fujitsu is sending me some fram chips even.
06:24 PM Hfuy: One thing I did discover is that modern LEDs are bright.
06:25 PM Lambda_Aurigae: of course they are.
06:25 PM Hfuy: I rebuilt a device which had used green LEDs by replacing them with very tiny white SMT ones.
06:25 PM Hfuy: Old LEDs: 20mA apiece. Modern LEDs: one milliamp each and still several times brighter.
06:25 PM Lambda_Aurigae: them blue ones are too flippin bright and annoying
06:26 PM Hfuy: You can run the damn things at one milliamp. One!
06:26 PM Lambda_Aurigae: I just bought a bunch of LED christmas lights that were on 90%+ clearance.
06:26 PM * Hfuy gesticulates wildly
06:26 PM Lambda_Aurigae: 150 LEDs for $0.75 each.
06:27 PM Hfuy: Um.
06:27 PM Lambda_Aurigae: bought 4 all white and 4 multicolor red/green/blue
06:27 PM Hfuy: 75 cents the set? Crikey.
06:27 PM Lambda_Aurigae: yeah.
06:27 PM Lambda_Aurigae: they were returns at a local store...none of the strings work.
06:27 PM Hfuy: Another approach is to strip them off sticky strip.
06:27 PM Lambda_Aurigae: but they all have good LEDs
06:27 PM Hfuy: Well. LEDs are hard to kill, unless you do something utterly inane.
06:27 PM Lambda_Aurigae: yup.
06:28 PM Lambda_Aurigae: the two I've futzed with had bad plugs.
06:28 PM Hfuy: What a superb technology, really.
06:28 PM Hfuy: I'm sitting in a room lit by LEDs looking at an LED-backlit display.
06:29 PM Hfuy: I even have some LED movie lights.
06:29 PM Lambda_Aurigae: one thing I'm told by firefighters,,,they still prefer incandescent flashlights for fires.
06:29 PM Lambda_Aurigae: LED light for some reason glares off the smoke.
06:30 PM Lambda_Aurigae: where those slightly yellow incandescent bulbs penetrate better.
06:30 PM Hfuy: hmm.
06:30 PM Hfuy: They could fix that I'm sure.
06:30 PM Hfuy: The tungsten-balanced film and TV lighting is very convincing.
06:31 PM Lambda_Aurigae: good friend of mine is a firefighter and they have tested many and still go back to the incandescents.
06:31 PM Hfuy: Fair play
06:31 PM Hfuy: Otherwise I suspect LED dominates utterly.
06:31 PM Lambda_Aurigae: getting there.
06:31 PM Hfuy: Also, you can get red, green and blue drivers for them, which I've used as accent lighting on camera.
06:32 PM Hfuy: You have to be a bit careful as they're very saturated and can cause odd gamut clipping problems.
06:34 PM Lambda_Aurigae: I have several flashlights that I use for viewing copier/printer output.
06:34 PM Lambda_Aurigae: different color temps.
06:34 PM Lambda_Aurigae: one is incandescent but the other three are LED.
06:34 PM Hfuy: Might want to be a bit careful with that, as LED output is very spiky and will mess you up.
06:34 PM Hfuy: At least unless it's very, very carefully engineered, which flashlights aren't.
06:34 PM Lambda_Aurigae: they are made specifically for this.
06:35 PM Lambda_Aurigae: I am a xerox field analyst...I deal with a lot of color printing issues.
06:35 PM Hfuy: Ah, right. I suspect they're expensive.
06:36 PM Lambda_Aurigae: yup.
06:36 PM Hfuy: Certainly the reasonable-quality movie stuff is alarmingly spendy.#
06:36 PM Lambda_Aurigae: I particularly like the blacklight one.
06:36 PM Hfuy: Heh.
06:37 PM Hfuy: If I want blacklight, I tend to get an HMI and put three layers of Lee's "congo blue" filter on it.
06:37 PM Lambda_Aurigae: great for finding the hidden code on the pages.
06:37 PM Lambda_Aurigae: the yellow shines out good under the blacklight LED
06:38 PM Hfuy: the challenge in movie production is to make them sufficiently powerful.
06:38 PM Lambda_Aurigae: as does the special coating on some paper which totally changes the colors on the prints.
06:38 PM Hfuy: Most of them are up to 100W, 300-400 is coming out now. We want 5kW.
06:39 PM Lambda_Aurigae: I want a 100W laser diode dangit!
06:39 PM Hfuy: Colour management is a fricken nightmare in movies.
06:39 PM Hfuy: It's a much older problem in print so you have more universal systems for it.#
06:39 PM Lambda_Aurigae: I want to burn ants in the front yard from my porch with an LED!
06:40 PM Lambda_Aurigae: problem in print is the manufacturers are always modifying the technology.
06:40 PM * Hfuy nods glumly
06:40 PM Hfuy: Same.
06:40 PM Hfuy: "Ooh, look, we have $BETTER this month."
06:40 PM Hfuy: "Stop screwing around with it!"
06:40 PM Lambda_Aurigae: new model of the same device might have the same exact print engine but they change the toner or the laser or drum surface or something and it totally throws things off.
06:40 PM Hfuy: "But it's $BETTER!"
06:40 PM Lambda_Aurigae: we have 29 new xerox models coming out this year.
06:41 PM Lambda_Aurigae: biggest rollout of any copier manufacturer ever in one year.
06:41 PM Hfuy: Is that...
06:41 PM Hfuy: ...necessary?
06:41 PM Lambda_Aurigae: total line refresh
06:41 PM Lambda_Aurigae: from b&w printers through production level digital press machines.
06:41 PM Hfuy: I I see
06:41 PM Hfuy: Can't say it's a name we see here much.
06:41 PM Lambda_Aurigae: and I gotta learn em all.
06:42 PM Lambda_Aurigae: cause I'm the guy they come to when things don't work right and nobody else can figure it out.
06:43 PM Hfuy: My girlfriend used to have a terrible office job and has a fund of stories with titles like New Intern and the Exploding Magenta Toner.
06:43 PM Lambda_Aurigae: I'm a specialist on 5 different model lines currently plus analyst specialist on all xerox copier/printer systems.
06:43 PM Lambda_Aurigae: hehe.
06:43 PM Hfuy: They had to replace several items of furniture and the carpet.
06:44 PM Lambda_Aurigae: I trained as a copier tech on the job with other techs for 3 weeks then they turned me loose on my own.
06:44 PM Lambda_Aurigae: my first day on my own I went out to empty a waste toner as my first ever alone call.
06:44 PM Lambda_Aurigae: they had a nice white carpet
06:44 PM Lambda_Aurigae: so, I took the waste toner bottle outside.
06:44 PM Hfuy: So you have picky clients who want to be able to pick a pantone colour in photoshop, press print, and have it come out and match their T-shirt.
06:44 PM Lambda_Aurigae: started to empty it and the wind came up
06:44 PM Lambda_Aurigae: I ended up with multicolored toner from head to toe.
06:45 PM Hfuy: Geenee
06:45 PM Lambda_Aurigae: at the end of the day I still had toner in my ears and other places.
06:45 PM Lambda_Aurigae: looked like I had been attacked by sesame street!
06:45 PM Hfuy: Well it's ball-milled down to a microfine size, isn't it.
06:45 PM Lambda_Aurigae: yeah....we do have those kinds of customers.
06:45 PM Hfuy: So it's outrageously difficult to clean up.
06:45 PM Lambda_Aurigae: actually, the new xerox toner isn't ball milled.
06:45 PM Lambda_Aurigae: it's grown.
06:45 PM Lambda_Aurigae: floating crystal growing operation.
06:46 PM Hfuy: Is that so they can charge champagne prices for it?
06:46 PM Lambda_Aurigae: it's even finer and more consistent size and shape than milled toner.
06:46 PM Chillum: woah
06:46 PM Lambda_Aurigae: it's how we can get 2400dpi resolution on a 500 dollar printer.
06:46 PM Hfuy: And quite purple, from what I've seen.
06:47 PM Hfuy: Also, is it really a $500 printer, or is it a $2500 printer which costs $500 to refill :)
06:47 PM Lambda_Aurigae: we can do purples and reds that nobody else can do.
06:47 PM Lambda_Aurigae: hell, xerox printers can hit the red in the sharp logo....sharp printers and copiers can't do that.
06:47 PM Hfuy: There has been much change in film and TV gamuts recently. We could never really do teal before. Now, with modern gear, we can.
06:48 PM Lambda_Aurigae: which we find to be humerous.
06:48 PM Hfuy: TV red was always a bit orange.
06:48 PM Hfuy: That's improved too.
06:48 PM Lambda_Aurigae: yup.
06:48 PM Hfuy: Now we can actually have teal and ruby, if everyone puts the switches in the right place.
06:48 PM Hfuy: They don't.
06:49 PM Lambda_Aurigae: the fun part is trying to explain color gamuts between that cheap LCD monitor and a laser printer/copier
06:49 PM Lambda_Aurigae: how they can show something on screen and no way the printer can reproduce it.
06:49 PM Chillum: hehe
06:49 PM * Hfuy facepalms with feeling
06:49 PM Hfuy: You're not kidding.
06:49 PM Lambda_Aurigae: the difference between RGB and CMYK processes.
06:49 PM Hfuy: "It doesn't look like that at work!"
06:49 PM Lambda_Aurigae: and
06:49 PM Hfuy: "What do they have at work?"
06:49 PM Hfuy: "A sony BVM-X300"
06:49 PM Lambda_Aurigae: the differences between monitors!
06:50 PM Hfuy: "And you have?"
06:50 PM Hfuy: "My ipad!"
06:50 PM Lambda_Aurigae: and how if their monitor isn't calibrated then there is no way I can attempt to match what they see there with what the copier can produce.
06:50 PM * Hfuy discreetly crushes a small object while grinding his teeth
06:50 PM Lambda_Aurigae: oh glub..ipad displays are HORRID!
06:50 PM Hfuy: I feel your pain.
06:50 PM Lambda_Aurigae: the color temp is soooo fucked on those things.
06:50 PM Hfuy: BVM-X300 is a 1500-nit OLED costing upward of USD30,000
06:51 PM Hfuy: Their ipad... isn't.
06:51 PM Lambda_Aurigae: had a customer with an ipad air and a macbook pro....and they would print from both...and get the same output from the printer even though the displays showed different colors.
06:51 PM Lambda_Aurigae: and they expected the document to print to match the display.
06:52 PM Hfuy: It's a huge problem in film and TV too
06:53 PM Hfuy: At least in print, once you print it, assuming reasonable longevity from the pigments, it is what it is.
06:53 PM Hfuy: We distribute something, everyone's TV is different.
06:53 PM Lambda_Aurigae: yeah.
06:53 PM Lambda_Aurigae: toner pigments tend to fade over time.
06:53 PM Lambda_Aurigae: even without light exposure.
06:53 PM Hfuy: I always assumed the colours were flaked plastic.
06:53 PM Lambda_Aurigae: xerox EA toner seems to not fade as bad...that's the grown crystal toner.
06:53 PM Lambda_Aurigae: they are.
06:53 PM Lambda_Aurigae: most use colored PVC
06:54 PM Hfuy: And the black is... soot.
06:54 PM Lambda_Aurigae: xerox uses colored polyester.
06:54 PM Lambda_Aurigae: PVC colored with carbon black.
06:54 PM Lambda_Aurigae: xerox EA toner doesn't have any carbon black in it.
06:54 PM Hfuy: I once made a set of dye cells for colour mixing light using CMY inkjet inks. Worked pretty well.
06:54 PM Lambda_Aurigae: it's just tinted polyester balls.
06:55 PM Hfuy: They don't do the half-intensity pigments in electrostatic imaging, do they?
06:55 PM Lambda_Aurigae: well,,,
06:56 PM Lambda_Aurigae: some inkjets do that.
06:56 PM Lambda_Aurigae: usually HP wide format units.
06:56 PM Hfuy: I had a very nice HP A3-width printer that did. Was superb. Only printer I ever wore out.
06:56 PM Lambda_Aurigae: I've only seen two laser devices that use more than the standard CMYK toners
06:57 PM Lambda_Aurigae: and those were 5 color process only with the 5th color changeable to fit specific color requirements.
06:57 PM Hfuy: My memory of colour lasers isn't that great, but it's probably from the mid-90s.
06:57 PM Lambda_Aurigae: gold, silver, bright green, bright blue, and clear.
06:57 PM Hfuy: I remember it being very... well. 300dpi and halftone-y.
06:57 PM Lambda_Aurigae: it's all halftones.
06:58 PM Hfuy: Sure, but at 300dpi, it doesn't actually look very good.
06:58 PM Lambda_Aurigae: lowest res device we sell these days is 600dpi and that's a black and white printer.
06:58 PM Hfuy: At 2200, I suspect things improve.
06:58 PM Lambda_Aurigae: most are 1200dpi
06:58 PM Lambda_Aurigae: new ones are 2400dpi.
06:58 PM Hfuy: I fear that inkjets still rule the roost for absolute image quality, though?
06:59 PM Lambda_Aurigae: well,,,
06:59 PM Lambda_Aurigae: depends on the image.
06:59 PM Lambda_Aurigae: really, ink press rules...but who can afford a 100K dollar press?
06:59 PM Lambda_Aurigae: and that's bottom of the line.
07:00 PM Lambda_Aurigae: the new digital presses are putting out some really amazing images these days.
07:02 PM Lambda_Aurigae: https://hackaday.com/2017/03/08/light-replaces-electrons-for-giant-vector-graphics-asteroids-game/
07:02 PM Hfuy: the only thing about lasers I notice is that the fused toner is often more reflective than the paper, so there's a difference in gloss.
07:02 PM Lambda_Aurigae: glub I miss vector graphics!
07:03 PM Lambda_Aurigae: that's why the new devices have adjustments for plain vs glossy paper...you can get major differences in image quality by setting the proper paper type before printing.
07:03 PM Hfuy: how does it adjust the gloss on the toner? Some sort of roughness control?
07:03 PM Lambda_Aurigae: and the Xerox Color 1000 has a 5 color process...you can set the 5th color as clear and do a full clearcoat on the entire page.
07:04 PM Lambda_Aurigae: not sure really how...I know it changes some temperature and pressure settings in the fuser.
07:04 PM Hfuy: I'd like to have a laser, but probably a simple mono one that I can tear apart and use to print PCBs.
07:04 PM Hfuy: Eventually someone will make one for that...
07:04 PM Lambda_Aurigae: hehe
07:04 PM Lambda_Aurigae: I've done that.
07:05 PM Lambda_Aurigae: hp laserjet 8100...based on the old 5si
07:05 PM Lambda_Aurigae: nice straight paper path with minimal modifications.
07:05 PM Lambda_Aurigae: I got it feeding thick cardstock the same thickness as the PCB material I was using
07:05 PM Lambda_Aurigae: then cut a hole in the middle of the cardstock and taped a 2x2 inch piece of PCB in the hole.
07:06 PM Lambda_Aurigae: fed that through and printed on it.
07:06 PM Lambda_Aurigae: biggest problem I had is arcing from the high voltage on the copper clad edges.
07:06 PM Lambda_Aurigae: which killed 3 HV power supplies in a week.
07:06 PM Lambda_Aurigae: luckily I have access to piles of those printers.
07:07 PM Lambda_Aurigae: had to mod the fuser a bit and the paper path a bit more to get it to feed
07:08 PM Lambda_Aurigae: and lots of adjustments to high voltage and fuser temp
07:09 PM Lambda_Aurigae: I've seen several other people do direct to copper clad printing with lasers but they always remove the fuser and do hand fusing with an iron.
07:09 PM Hfuy: The guide I saw omitted the fuser, and simply placed the pcb, with loose toner, in an acetone vapour atmosphere for a few minutes.
07:09 PM Hfuy: Worked nicely.
07:09 PM Hfuy: I was surprised the whole electrostatic thing worked on copper, but it seems to.
07:10 PM Lambda_Aurigae: my other direct to copperclad printing is using a zebra thermal transfer printer.
07:10 PM Hfuy: One day someone will do a device for this.
07:10 PM Hfuy: Although I guess it's something you can also do with a 3D printer and a sharpie.
07:10 PM Lambda_Aurigae: the current zebra printer I have can only get me down to 8mil traces and spacing and it's very blocky.
07:11 PM Lambda_Aurigae: heck, for many years I did it with just a sharpie and rat-shack tracing guide.
07:11 PM Hfuy: Me too.
07:11 PM Lambda_Aurigae: hell, still do on occasion.
07:11 PM Hfuy: If I need to do something now I tend to do UV
07:11 PM Hfuy: though frankly, with arduino boards effectively free, I rarely need to.
07:11 PM Lambda_Aurigae: I stll build things with discreet 74xx series chips.
07:12 PM Lambda_Aurigae: or 555 timer based toys.
07:13 PM Hfuy: I spent an unconscionable amount of time plotting out roughly how I'd do an 8x8 matrix LED driver using discretes, including discrete SRAM to hold the bitmap.
07:13 PM Hfuy: Grayscale, you understand. None of this single-bit stuff.
07:13 PM Lambda_Aurigae: hehe.
07:13 PM Lambda_Aurigae: back in my teens I built a 4-bit computer from discreets
07:14 PM Hfuy: OK, you win.
07:14 PM Lambda_Aurigae: including an 8x8 pixel display.
07:14 PM Lambda_Aurigae: wire wrapped boards and all.
07:14 PM Hfuy: I wonder why so many of the early logic ICs are four-bit devices.
07:14 PM Lambda_Aurigae: I've killed waaay too many brain cells since then.
07:14 PM Hfuy: I believe Cray made his machines out of that sort of stuff. I suppose it was fast
07:14 PM Lambda_Aurigae: back then 4 bits is all we needed!
07:15 PM Lambda_Aurigae: my mass storage and programming medium was a strip of graph paper and a pencil.
07:15 PM Hfuy: One, two, miss a few... sixteen?
07:15 PM Lambda_Aurigae: used two contacts to read the pencil on the paper in the square as it went through...4 bits wide plus timing squares.
07:16 PM Lambda_Aurigae: had a totally twisted idea while driving to a call today.
07:16 PM Lambda_Aurigae: kinda-almost-binary-coded-dice display.
07:17 PM Lambda_Aurigae: a 3x3 led digit.
07:17 PM Lambda_Aurigae: done kinda like dice.
07:17 PM Hfuy: Reminds me of the (very obscure) Marain language invented by Iain M. Banks for his fictional Culture.
07:17 PM Lambda_Aurigae: only you can display 0 to 9 with it with the standard dice patterns plus a few extras.
07:17 PM Hfuy: That was all 3x3 cells and nonary counting.
07:18 PM Lambda_Aurigae: I think it would make an awesome clock.
07:18 PM Lambda_Aurigae: might even get me on hackaday frontpage!
07:18 PM Lambda_Aurigae: hehe
07:18 PM Lambda_Aurigae: they seem to like clocks with non-standard displays.
07:19 PM Lambda_Aurigae: and if I built it around an ardweeny I'm sure to win something!
07:19 PM Hfuy: Recently I was asked to research a fun project - electronic props for a production set in a world best described as 80s-punk.
07:19 PM Hfuy: So everything's wood veneer and brushed metal and red LED displays, but it does modern things.
07:19 PM Lambda_Aurigae: hehe
07:19 PM Hfuy: I was trying to find a source of orange plasma displays, but the only way to get them was to rip them out of old Grid Compass laptops, which are 80s enough to begin with.
07:19 PM Lambda_Aurigae: that sounds like one I got to work on a couple years back.
07:20 PM Hfuy: Trying to find a pixel display that worked for it was hard.
07:20 PM Lambda_Aurigae: a computer point of sale system that didn't look like a computer....for an amish store.
07:20 PM Hfuy: oh for pete's sake
07:20 PM * Hfuy holds no truck with religion
07:20 PM Lambda_Aurigae: had to be battery powered.
07:21 PM Lambda_Aurigae: we ended up using an old mechanical adding machine and making flip segment displays.
07:21 PM Hfuy: Making? Good grief, how much money did they have
07:21 PM Lambda_Aurigae: I find religion to be nothing more than another form of government...a few people controlling how the masses think and act.
07:21 PM Lambda_Aurigae: a shitload.
07:21 PM Lambda_Aurigae: we made 4 of them.
07:22 PM Lambda_Aurigae: at 800 dollars each.
07:22 PM Hfuy: 4 displays
07:22 PM Hfuy: or 4 units?
07:22 PM Lambda_Aurigae: 4 units.
07:22 PM Lambda_Aurigae: cost of parts was under 200 dollars.
07:23 PM Lambda_Aurigae: the displays were 3d printed
07:23 PM Lambda_Aurigae: other guy did that part.
07:23 PM Lambda_Aurigae: I did the electronics and programming
07:23 PM Lambda_Aurigae: seems so long as it can run off grid they don't mind.
07:23 PM Lambda_Aurigae: I also did the wooden cases for them.
07:24 PM Lambda_Aurigae: they came out really neat.
07:24 PM Lambda_Aurigae: even had receipt printers.
07:24 PM Hfuy: Those people are crazy
07:24 PM Lambda_Aurigae: of course.
07:24 PM Hfuy: Unfortunately it's almost 1am here, so I'd better fall into bed.
07:24 PM Lambda_Aurigae: but they pay good money for good work.
07:25 PM Lambda_Aurigae: nearly 7pm here.
07:25 PM Hfuy: Pleasure talking
07:25 PM Lambda_Aurigae: getting close to bedtime for me too.
09:13 PM Rez is now known as L
10:13 PM L is now known as Lo
11:27 PM arij_work is now known as arij
11:50 PM daey_ is now known as daey