#avr Logs
Apr 03 2018
#avr Calendar
12:16 AM _ami_: https://www.reddit.com/r/korea/comments/88y5cy/3_americans_assult_koreans_and_make_racist/
12:19 AM _ami_: its in news now
12:19 AM _ami_: https://www.youtube.com/watch?v=X-YQ3Ea-PvQ&feature=youtu.be
01:52 AM _ami_: https://time4ee.com/news.php?readmore=342 Saleae Logic Analyzer 1.2.22 beta - real-time feature
01:52 AM _ami_: this is nice.
01:52 AM _ami_: and its working for clones too.
01:52 AM _ami_: :D
03:48 AM Haohmaru: how do i find my strings in the .lss file?
03:48 AM Haohmaru: it seems they are not written as plain text, cuz i can't ctrl+F them
03:49 AM Haohmaru: i see the printf() calls, but i don't understand teh asm voodoo magic shizzle
04:44 AM polprog: have you tried the strings command
04:45 AM polprog: or they just arent there
04:48 AM Haohmaru: strings command?
05:22 AM polprog: `strings file.bin`
05:22 AM Haohmaru: maybe that's some linuxy thang
05:22 AM polprog: ah yeah
05:32 AM MrFahrenheit: try binwalk too
05:32 AM MrFahrenheit: :P
05:33 AM Haohmaru: why bin :/
05:33 AM Haohmaru: i want to know if they go in RAM or flash or what
05:34 AM MrFahrenheit: binwalk analyses binary files
05:34 AM Haohmaru: cuz i have further questions that i hope will get automatically answered from that.. instead of having to ask here or there
05:35 AM Haohmaru: hm, are they in the .lss at all? they surely must be inside the .elf
05:36 AM MrFahrenheit: they must be there, yes
05:36 AM MrFahrenheit: you have to copy them into the hex file by selecting the correct data segment
05:36 AM MrFahrenheit: (if you're generating the hex file yourself)
05:37 AM MrFahrenheit: strings should go to the .data segment if I'm not mistaken
05:37 AM Haohmaru: avr-objcopy -R .eeprom -R .fuse -R .lock -R .signature -O ihex $(TARGET_OUTPUT_FILE) $(TARGET_OUTPUT_DIR)$(TARGET_OUTPUT_BASENAME).hex
05:38 AM MrFahrenheit: yes, that should copy data to the hex
05:39 AM polprog: odd, my CD4066 takes 60mA when just connected to power. could that be because of floating inputs?
05:39 AM MrFahrenheit: Haohmaru, try objdump -s -j .data file.elf
05:40 AM MrFahrenheit: or avr-objdump
05:41 AM Haohmaru: i'll see
05:41 AM MrFahrenheit: polprog, datasheet says you can't inputs them floating
05:41 AM Haohmaru: polprog what does the datasheet say about unused inputs? ;P~
05:41 AM MrFahrenheit: *can't leave inputs floating
05:41 AM polprog: Haohmaru: tie them to gnd
05:42 AM Haohmaru: ;P~
05:42 AM polprog: but, when i give it 0-5 V it takes nothing
05:42 AM Haohmaru: takes?
05:42 AM polprog: let me get some resistors
05:42 AM MrFahrenheit: "Generally they are tied to GND or VCC, whichever makes more sense or is more convenient."
05:42 AM polprog: when vss is at 0 and vdd at 5 it takes 1ma
05:46 AM polprog: okay. looking good here
05:46 AM polprog: i dont like how they call the pin Vss and then talk about pulling the inputs to GND
05:46 AM polprog: this causes some confusion
05:47 AM polprog: ESPECIALLY in analogue where GND ins't necesarily VSS
06:06 AM polprog: https://puu.sh/zVjyQ/eb18af474f.png
06:07 AM polprog: :/ doesnt look good
06:11 AM polprog: i need tho think of something better
08:28 AM rue_bed: polprog, what did you build?
08:34 AM polprog: rue_bed: i tried to make that "voltage controlled resistor" with the cd 4066.
08:34 AM polprog: but it didnt work out well
08:34 AM polprog: :/
08:34 AM polprog: see pic
08:34 AM Haohmaru: did it result in a resistance-controlled voltage? ;P~
08:35 AM polprog: i connected an RC circuit to the control pin of the analog switch but it makes a waveform that's unusable
08:35 AM polprog: a -4V negative peak
08:36 AM polprog: i need to find a way to slowly change the gain from -inf to 0dB
08:49 AM MrFahrenheit: why not just use some fets
08:49 AM polprog: i am
08:49 AM polprog: just a sec, let me show you what i came up with
08:51 AM MrFahrenheit: or if you don't need precision or more than 8 bits, maybe a digital pot?
08:51 AM polprog: no, digital pot not really
08:52 AM polprog: id have to add a whole MCU there
08:52 AM LeoNerd: Uhm, analog switch is usually digitally *controlled*; it's either off or on
08:52 AM LeoNerd: It isn't continuously variable
08:52 AM polprog: LeoNerd: i just found out about that expermientally :D
08:57 AM polprog: seems like id either need to find amosfet that doesnt have the reverse biased diode
08:57 AM polprog: or i have no idea
08:57 AM LeoNerd: What are you attempting to do?
08:59 AM polprog: a circuit that mutes an input signal (clamps it to ground) with the restriction that when its switched on/off it will gradually clamp/unclamp it to avoid a steep edge on the output
08:59 AM polprog: short, a mute button on my mixer box
09:02 AM LeoNerd: Ah, you possibly then want a FET with an RC sloped gate
09:02 AM polprog: yes
09:02 AM polprog: hmm now that i think aboutit
09:02 AM polprog: ill post a falstad link in a sec
09:03 AM polprog: http://tinyurl.com/y82bgjbx
09:04 AM polprog: falstad fet doesnt have the reverse biased diode (from S to D) like for example 2n7000 has
09:04 AM polprog: but luckily it seems that it doesnt matter
09:04 AM polprog: as ive just added such a diode and it just doesnt change anything
09:05 AM LeoNerd: Yeah, that sort of thing should suffice
09:05 AM polprog: great, let me now test it in reality
09:05 AM polprog: well, nope
09:06 AM polprog: the diode clamps the signal at above -0.6V
09:06 AM polprog: so thats wrong
09:06 AM LeoNerd: You'll need the source to be at least 0.6V lower than the entire range of your signal
09:07 AM LeoNerd: To ensure it's properly off
09:08 AM polprog: but then it clamps my signal at the source voltage (in my case -8.5V - the negative analog supply of my whole design)
09:08 AM LeoNerd: Your processing circuit has to be power by rails in excess of the signal voltage; that's just generally true
09:08 AM LeoNerd: *powered
09:09 AM polprog: yeah i know
09:09 AM polprog: i have the preamp ready and tested,
09:09 AM polprog: the only thing left is the mute button and the summing circuit
09:09 AM polprog: with the latterone being simple
09:10 AM rue_mohr: the 4066 is a a switch, not var. resistor...
09:10 AM LeoNerd: Alternatively, if this is all audio, you could get away with some DC-blocking caps and shifting a section of the circuit with a bias
09:10 AM polprog: that could work
09:10 AM polprog: let me think
09:10 AM rue_mohr: polprog, did you look up voltage multiplier circuits?
09:10 AM LeoNerd: In fact, maybe that plus the 4066 would be sufficient for your soft-on anyway
09:10 AM LeoNerd: The nonzero resistance of the analog switch (plus maybe add more resistance) to slowly bring a capacitor up from discharge
09:11 AM polprog: i dont think so, 4066 doesnt do soft on from what youve told me and what i found out
09:11 AM rue_mohr: polprog, the linear range of a fet is reaaaly small
09:11 AM polprog: rue_mohr: yes, but it doesnt matter that much here
09:11 AM polprog: it doesnt have to be linear - it just needs to be smooth
09:11 AM rue_mohr: but, in that model, it looks like your trying to SWITCH the audio on and off, not set its volume
09:11 AM LeoNerd: Yes; that's the usecase
09:12 AM polprog: thats the aim of this circuit
09:12 AM LeoNerd: A simple on/off mute, with soft-start
09:12 AM rue_mohr: if you want to fade, you need to have a volume control and fade it in
09:12 AM rue_mohr: VGA Variable Gain Amplifier, good luck googling that :)
09:13 AM polprog: yes i have a fader.
09:13 AM rue_mohr: https://i.stack.imgur.com/ZKnN8.png
09:13 AM rue_mohr: hah, binary gain
09:13 AM rue_mohr: polprog, if you have a fader, you need to just bring it in
09:14 AM rue_mohr: thats your soft mute
09:14 AM polprog: the fader is a knob/ slider
09:14 AM polprog: apart from that i want to have a button that mutes it immediately
09:14 AM rue_mohr: you need an electroic one then ! :)
09:14 AM rue_mohr: https://encrypted-tbn0.gstatic.com/images?q=tbn:ANd9GcT9cwGGZGuZbtZmwrj3EHZlKdktVcRqxH--eYE2hQzsAdRssasY
09:14 AM LeoNerd: Honestly; I think the circuit you sent the link above for was fine
09:14 AM LeoNerd: Just bias it appropriately outside of the signal range
09:16 AM rue_mohr: https://i.stack.imgur.com/yvqRr.png
09:16 AM rue_mohr: huh
09:25 AM rue_mohr: in either case, variable gain amplifiers are one of the more challanging things
09:25 AM polprog: thats why im trying to avoid making one
09:25 AM rue_mohr: wanna try one of those schematics?
09:26 AM polprog: maybe later
09:26 AM rue_mohr: you dont have to use the LM101
09:26 AM polprog: im thinking now - im chanrging the capacitor from a voltage source. what if i used a current source
09:26 AM polprog: http://tinyurl.com/ydy6em65
09:26 AM rue_mohr: thats the joke in the old databooks, the schematics were not speicifically for the chips they were listed with
09:28 AM polprog: https://puu.sh/zVoIQ/57b9a419ab.png this nasty edge bothers me
09:28 AM rue_mohr: do you know how big the linear region of that fet IS!?
09:28 AM polprog: i guess the only way to find out is to build it
09:28 AM rue_mohr: yea I pushed the buttons
09:28 AM MrFahrenheit: python lib done
09:29 AM MrFahrenheit: well... working, not done
09:29 AM rue_mohr: (as in its small)
09:29 AM rue_mohr: eek, work, bye!
09:29 AM polprog: i know its small
09:29 AM polprog: cya
09:30 AM polprog: i guess if i sapplied the RC circuits output in the linear region then it could work out nicely
09:30 AM polprog: but thats not reall possible
09:32 AM polprog: i guess ill build the last one and see for real if it works or not, measure it, decide if i wanna go with it or keep digging
09:32 AM rue_mohr: depends if you use another fet in an op-amp circuit to do a feedback loop that also generates a linear control signal of the secondary audio fet
09:36 AM polprog: which secondary fet?
09:36 AM polprog: that would be the only one discrete fet in the whole signal chain, or whatever you call that
09:39 AM MrFahrenheit: doesn't have to be linear though
09:39 AM MrFahrenheit: in fact, it's better if it's logarithmic
09:39 AM polprog: ill see
09:39 AM MrFahrenheit: assuming it's audio
09:39 AM polprog: it is
09:39 AM polprog: if it doesnt pluck, im happy
09:39 AM polprog: now let me do some maths and test it
10:12 AM polprog: it works!
10:16 AM MrFahrenheit: neat
10:28 AM polprog: im so happy
10:38 AM polprog: https://youtu.be/38lTSFt5QgU
10:47 AM MrFahrenheit: cool
11:18 AM aczid: nice :D
11:31 AM polprog: testing the peak detector now
11:32 AM polprog: https://puu.sh/zVszq/185b1cd9d5.png
12:08 PM * Ameisen waits for LLVM to rebuild after updating after a week
12:24 PM Emil: uh oh
12:25 PM Emil: I tried to take apart a Casio CFX-9800G
12:25 PM Emil: For the LCD
12:25 PM Emil: The fucking lines are individually driven
12:25 PM polprog: well, if they have an asic for that
12:25 PM Casper: matrix like most likelly?
12:25 PM polprog: yeah sounds like a graphing one
12:26 PM Ameisen: Trying to figure out how to 'convince' LLVM-Clang to not bother with a true call stack until it hits the 'last' function in the NORETURN chain
12:26 PM Ameisen: your 'true' entry point rather than all the random init functions
12:26 PM Ameisen: no reason to waste stack on return addresses and such to functions that will never be returned to
12:27 PM Emil: 2018-04-03 20:25:00 +0300 < Elepaja> [emilfihlman] <photo> https://elepaja.fi/telegram/media/AgADBAADVK0xG0tSIVL9wg1IRr7DUjxpmhoABHDHB1tFCvLvk18AAgI.jpg
12:27 PM Emil: 2018-04-03 20:25:16 +0300 < Elepaja> [emilfihlman] <photo> https://elepaja.fi/telegram/media/AgADBAADVa0xG0tSIVIrhb19MBNHTIQ_nRoABBamlUtrg3qnuGAAAgI.jpg
12:27 PM Emil: 2018-04-03 20:25:28 +0300 < Elepaja> [emilfihlman] <photo> https://elepaja.fi/telegram/media/AgADBAADVq0xG0tSIVKheeHe4TjZN4hpiRoABKYKbl2qCz1uMz0EAAEC.jpg
12:27 PM Emil: 2018-04-03 20:25:38 +0300 < Elepaja> [emilfihlman] <photo> https://elepaja.fi/telegram/media/AgADBAADV60xG0tSIVL1CPueUk4V8WbmkBkABP0Y1RQdG8L2GOwAAgI.jpg
12:27 PM Emil: 2018-04-03 20:25:44 +0300 < Elepaja> [emilfihlman] Luulin et saisin kivan näytön
12:27 PM Emil: 2018-04-03 20:25:50 +0300 < Elepaja> [emilfihlman] Eisit.jpg :D
12:28 PM Ameisen: also trying to, but failing at, add an __attribute__((address=blah))
12:28 PM Ameisen: so a symbol has a fixed address
12:29 PM polprog: nice chip lol
12:29 PM Ameisen: adding an address attribute is suprisingly difficult
12:29 PM Ameisen: doesn't seem that the compiler keeps address information with symbols
12:30 PM polprog: why did you take apart that magnificent device
12:30 PM Emil: polprog: ikr
12:30 PM polprog: and does it still work
12:30 PM Emil: I couldn't get it on
12:31 PM polprog: ah
12:31 PM polprog: https://www.youtube.com/watch?v=rqVDx2zRSyg
12:32 PM polprog: peak detector tester
12:33 PM Emil: polprog: wrong link?
12:33 PM polprog: no thats the track i used
12:33 PM Emil: Ah
12:33 PM polprog: u can upload the video of the actual peak det working
12:33 PM polprog: :D
12:33 PM polprog: i can*
12:34 PM Emil: doit
12:39 PM polprog: https://youtu.be/zIndG0bsLgQ
12:40 PM Emil: polprog: is that purely analog?
12:40 PM polprog: yes
12:40 PM polprog: now im adding the RS latch so you have a led that you have to reset (like a real peak detector)
12:41 PM Emil: Nooo
12:41 PM Emil: just add a low pass
12:41 PM Emil: that you can adjust
12:41 PM Emil: so you can adjust the period
12:41 PM polprog: hmm
12:41 PM Emil: where it calculates the peak
12:41 PM Emil: that's just a leaky integrator
12:41 PM polprog: the led you are watching is the "PEAK NOW" https://puu.sh/zVszq/185b1cd9d5.png
12:41 PM Emil: (through a rectifier of course)
12:42 PM polprog: it is ran trough a rectifier
12:42 PM polprog: :)
12:42 PM Emil: https://en.wikipedia.org/wiki/Leaky_integrator
12:42 PM polprog: hmm
12:43 PM polprog: i was thinking about adding an integrator with a certain time constant
12:43 PM polprog: that will be added as the input feed for the level meter in the future
12:43 PM Emil: https://www.youtube.com/watch?v=2cOY7mDv_pA
12:43 PM Emil: https://www.youtube.com/watch?v=a_8Dgcp-xfw
12:43 PM Emil: Much easier digitally
12:44 PM polprog: yeah i know
12:44 PM Emil: Though pure analog is fiiiiiineeeeeeee
12:44 PM Emil: daaamn guurl fine
12:44 PM polprog: saw that vid ! cool ;)
12:44 PM polprog: good speakers in the lab is a must
12:45 PM Emil: 100%
12:45 PM polprog: whats the track in the log one
12:45 PM Emil: Actually some fucker has stolen our speakers
12:45 PM Emil: I need to get them back
12:45 PM polprog: lol
12:45 PM polprog: ive got two pairs, one on each desk
12:46 PM polprog: and i took out an old pair of creative speakers as the output for my breadboard as youve seen
12:46 PM Emil: https://www.youtube.com/watch?v=Iu_B35mUWH8
12:46 PM Emil: I think it's this one
12:46 PM Emil: Thank god for good quality video :D
12:46 PM polprog: ty
12:46 PM Emil: and
12:46 PM Emil: time
12:46 PM Emil: is
12:47 PM Emil: 33:27
12:47 PM Emil: Droid Bishop - In Your Love
12:47 PM polprog: https://www.youtube.com/watch?v=oOMBp81NvLY i like this track
12:48 PM polprog: haha that video
12:48 PM polprog: VHS intensifies
12:48 PM polprog: but waut
12:48 PM polprog: wait!!!
12:48 PM Emil: wot
12:48 PM polprog: those are digital video distortions
12:48 PM polprog: :D
12:48 PM Emil: hahah
12:48 PM Emil: heresy!
12:49 PM polprog: mpeg errors
12:49 PM polprog: :D
12:50 PM polprog: cool trax overall
12:53 PM polprog: also got those nice momentary switches today
12:53 PM polprog: https://puu.sh/zVvbC/e3fb59ba64.jpg
01:12 PM polprog: time to make supper
01:12 PM Tom_L: what we havin?
01:13 PM Tom_L: tops tend to break off those
01:13 PM Tom_L: but i still use em
01:37 PM Ameisen: Hmm... ideally, I want to autogenerate system-specific tables for registers/etc based upon some kind of configuration file...
01:53 PM MrFahrenheit: polprog, I have a couple of those, bought them thinking they weren't momentary
01:54 PM polprog: ideally id have some that are like those but stable as well
01:54 PM polprog: i could unify the fron panel design then
01:54 PM polprog: front*
02:03 PM MrFahrenheit: I wish I could find relegendable keycaps for a low price, cause I have a whole bunch of keyboard switches that would be great for front panels
02:03 PM MrFahrenheit: I guess I could buy blank caps and use a label maker...
02:03 PM MrFahrenheit: oh shit, or I could 3d print some now \o/
02:09 PM Ameisen: I could release the ARM toolchain now. It's incomplete though.
02:09 PM Ameisen: It is functional though.
02:10 PM Ameisen: the AVR side is very incomplete, so not at all ready.
02:11 PM MrFahrenheit: for what
02:11 PM Ameisen: LLVM-based ARM Embedded toolchain
02:12 PM Ameisen: as an alternative to the GCC-based one
02:12 PM Ameisen: mainly so the entire toolchain, including libraries, can all by LTO
02:12 PM Ameisen: be*
02:12 PM Ameisen: so your final binary is optimized as one compilation unit merged together
02:13 PM MrFahrenheit: with clang?
02:14 PM Ameisen: yes
02:14 PM Ameisen: I tried to get a full LTO toolchain working with GCC
02:14 PM Ameisen: either libgcc or newlib via GCC were causing ICEs in GCC when built with LTO
02:14 PM Ameisen: the LTO-compilation worked fine, but linking failed miserably.
02:14 PM Ameisen: clang+newlib+compiler-rt, entirely LTO, work fine
02:14 PM Ameisen: probably will replace newlib though
02:15 PM Ameisen: it's too 'heavy' for my taste for embedded
02:15 PM Ameisen: it still strictly conforms to the spec, which is overkill for embedded. It also still presumes side-effects, like setting errno
02:16 PM Ameisen: I'd like the toolchain to handle AVR as well, but LLVM's AVR support is still very primitive.
02:16 PM Ameisen: GCC's AVR support is as well, but GCC is more like in the mesolithic for AVR support, whereas LLVM is still in the paleolithic.
02:24 PM MrFahrenheit: interesting
02:29 PM Ameisen: There are a few features I'm trying to add. One that I had to give up on was getting Clang's __builtin_constant_p to the point of GCC's
02:29 PM Ameisen: their internal tables function differently. GCC can determine that an argument is potentially constant. Clang cannot.
02:29 PM Ameisen: _really_ useful on AVR
02:29 PM Ameisen: where you could do branching for progmem behavior
02:30 PM Ameisen: the goal in both AVR and ARM is getting the compiler to help with compiling C++ to minimum-size/maximum-speed
02:30 PM MrFahrenheit: reminded me of that lkml post from last week
02:30 PM Ameisen: it's usually the compiler getting in the way
02:30 PM Ameisen: yeah. Slightly different purpose.
02:30 PM Ameisen: Tohugh technically __builtin_constant_p should work for them
02:30 PM Ameisen: wasn't sure why they weren't just using that
02:30 PM Ameisen: they already only support gcc
02:34 PM MrFahrenheit: wasn't there experimental clang support
02:34 PM Ameisen: for AVR? Yes.
02:34 PM Ameisen: It's... not good
02:34 PM Ameisen: I improved it a little
02:34 PM Ameisen: it's missing 99% of the code reducers
02:35 PM Ameisen: it just requires a lot of work to get the compiler generating good AVR code
02:35 PM Ameisen: GCC has had twenty years to do it
02:35 PM Ameisen: however, the GCC AVR devs have sort of abandoned ship due to the GCC maintainers not being friendly towards 8-bit ISAs
02:35 PM Ameisen: so... you are likely to see LLVM catch up to speed
02:36 PM Ameisen: hopefully, with __flash in C++, giving C++ in AVR the same features as C in AVR
02:36 PM Ameisen: meaning you won't need a 2000-line header file to give type-safe program memory variables :|
02:38 PM MrFahrenheit: no, I mean for the kernel
02:46 PM MrFahrenheit: is there even a point in making a python wrapper library available on windows
02:54 PM Ameisen: I'm not sure what a python wrapper library is useful for ever.
02:55 PM Ameisen: but I don't write python
02:59 PM MrFahrenheit: I use it for what you use ruby for
02:59 PM Ameisen: My build scripts are in Ruby. That's about it.
02:59 PM Ameisen: If I really need a quick script for something random, I'll prolly use ruby.
02:59 PM Ameisen: I like my ruby buildscripts more than makefiles.
04:27 PM twnqx: does any of you have experience with https://jlcpcb.com/ ?
04:28 PM Emil: twnqx: yeah
04:28 PM Emil: twnqx: they are good
04:28 PM Emil: (though I ordered through easyeda but it's the same thing)
04:29 PM twnqx: cool, thanks
04:29 PM twnqx: i might use them for my next project, just to test
04:31 PM Emil: well I mean, they are okay
04:32 PM Emil: I'm not qualified enough to say that any of them are better than others
04:32 PM Emil: Except that OSHParkis _fine_
04:32 PM Emil: but of course they cost more
04:34 PM MrFahrenheit: wait, $2 for 2 layer ones? let me guess, shipping is ~$12?
04:35 PM Emil: MrFahrenheit: no
04:35 PM Emil: MrFahrenheit: it's free
04:35 PM Emil: but only for first order
04:35 PM Ameisen: typical drug dealer strategy.
04:38 PM polprog: Heh
04:38 PM polprog: Nites
04:38 PM polprog: Niters*
04:39 PM polprog: Im thinking about adding a peak meter to my mixer, but thats tomorrow
08:47 PM _ami_: Almost finished with my diy bench power supply: https://pbs.twimg.com/media/DZ5hklFVMAETDcc.jpg
09:48 PM rue_mohr: polprog, did the volume get under control?
09:48 PM rue_mohr: fer mute?
09:49 PM rue_mohr: did you find a way to use an amp to control the resistance responce of a fet?
10:56 PM _ami_: rue_mohr, asking me or polprog ?
10:57 PM rue_mohr: polprog,
11:02 PM polprog: rue_mohr: no, im using a log pot as a voltage divider and then theres a 5+dB amp
11:03 PM polprog: First one did have variable gain amp
11:04 PM polprog: Just a sec let me find this
11:06 PM rue_mohr: but for your mute
11:07 PM polprog: Old design http://puu.sh/zTrIP/980aeeff26.jpg if i was to do it this way now i would remove biasing resistors (they add a dc component), get rid of that 36R on thr input, and use biger pot value
11:07 PM rue_mohr: you need a variable gain amp for your mute to get soft off
11:07 PM polprog: Ah
11:07 PM polprog: No, I managed with a mosfet and a dc biased signal
11:08 PM polprog: http://puu.sh/zVzn7/920ba4804d.jpg
11:08 PM rue_mohr: it works, not like this morning with the pop?
11:08 PM polprog: Works pretty well
11:09 PM rue_mohr: k
11:09 PM rue_mohr: nuffing it out should work, yea
11:10 PM polprog: Had a little chat in music-electronics about lack of headroom but it turns out it can go up to +4dBu, or +12 if i used +V at the top of the divider.
11:10 PM polprog: Now i just need to get the flip flop working and i can start putting this all together
11:11 PM polprog: If you scroll up i uploaded a video of it working
11:12 PM polprog: And a video of the peak detector comparator working
11:12 PM polprog: :)
11:22 PM polprog: I feel a bit sick
11:34 PM day__ is now known as day