#avr Logs
Jan 29 2018
#avr Calendar
12:06 AM nuxil: https://www.youtube.com/watch?v=J187Pw6UkCM lol..
12:15 AM polymorph: nuxil: this is more meaningful https://www.youtube.com/watch?v=J9b0J29OzAU
12:15 AM polymorph: https://www.youtube.com/watch?v=Y1MDOerruDU
12:17 AM nuxil: no magic there.
12:18 AM nuxil: inductance :)
12:19 AM nuxil: its basically the same prinsible as when drop a magnet in a cobber pipe.
12:21 AM polymorph: i would do it without magnet
12:21 AM polymorph: and friction
12:24 AM nuxil: just dont dig too deep into these types of experiments. you will endup with the free-energy nuts :p
12:30 AM polymorph: nuxil: relax
12:32 AM polymorph: nuxil: started electronics in 98
12:34 AM nuxil: :)
12:38 AM nuxil: so i have like atm 15 packages on tranveling on the 7th sea's bought on ebay. just now i got sms from DHL that i can expect delivery from singaport the 2nd feb. wtf. when did they start using DHL. 1st time ever i have DHL as maildelivery from Ebay with "Freeshipping"
12:42 AM _ami_: nuxil, your ebay stuff must be pricy? over 100$?
12:43 AM nuxil: yes 1 item i ordered costed like around 150 bucks
12:43 AM _ami_: yeah, that explains
12:44 AM nuxil: i hope its my new soldering station :)
12:45 AM nuxil: but from my experiance with chinees quality, it will probably blow up on me :)
12:46 AM _ami_: nuxil, nice, which one did u buy?
12:46 AM nuxil: wep something. 2in1
12:46 AM nuxil: hotair + soldering station
12:46 AM nuxil: reflow
12:46 AM _ami_: nice, 150$ is a nice deal.
12:47 AM nuxil: that remains to be seen ;)
12:47 AM _ami_: which brand?
12:48 AM _ami_: https://www.aliexpress.com/item/YIHUA-995D-soldering-station-used-for-motherboard-repair-tool/1794837776.html -> i bought this in july 2017
12:48 AM _ami_: works well till now
12:48 AM _ami_: it was 150$ then.
12:52 AM nuxil: brand is Saike. had to go look it up :p
12:52 AM nuxil: but its not 2in1. its 4in1
12:53 AM nuxil: https://www.ebay.com/itm/202008666472
12:54 AM nuxil: if its shit i'll sell it to my friend :p
12:56 AM nuxil: he asked me if he could buy old one. i just said hell no. you cna buy this new off me if im not happy with it :p
12:56 AM _ami_: :)
12:57 AM _ami_: there are lots of free stuffs with this.
12:57 AM nuxil: yea thats why i bought it.
01:21 AM nuxil: _ami_, hows your current protection thingy going?
01:24 AM _ami_: nuxil, it seems that LM358 is not good option for this purpose.
01:24 AM _ami_: i need MCP602
01:24 AM nuxil: because?
01:26 AM _ami_: LM358 has high offset ~7mV?
01:26 AM _ami_: and it does not work well on 5v supply if i want high gain.
01:26 AM _ami_: i think i could use two LM358
01:27 AM nuxil: − Input Offset Voltage . . . 3 mV Typ
01:27 AM _ami_: one to achieve 100 gain in two steps
01:27 AM _ami_: and other as a comparator.
01:28 AM _ami_: i shall give a try again.
01:32 AM nuxil: how about just using a transistor instead ?
01:33 AM nuxil: http://www.electronicdesign.com/sites/electronicdesign.com/files/uploads/2015/02/IFD2634_F1.gif simple as that.
01:38 AM polymorph: _ami_: it has less than 10mV offset, usually ~1mV, it will work down to 3V
01:42 AM _ami_: nuxil, polymorph https://cdn.instructables.com/FES/MSKK/J7MFS5G8/FESMSKKJ7MFS5G8.LARGE.jpg
01:42 AM polymorph: cant open it now, what are you building?
01:42 AM _ami_: i was trying to modify this ckt since i have 0.01 ohms shunt and LM358 op-amp
01:43 AM polymorph: will not work well
01:43 AM _ami_: polymorph, current control on those XL6009 boost converter
01:43 AM _ami_: injecting voltage on feedback pin
01:43 AM _ami_: based on http://www.instructables.com/id/Adding-a-Current-Limit-Feature-to-a-BuckBoost-Conv/
01:43 AM _ami_: polymorph, yes, it did not work well.
01:44 AM _ami_: i could only able to control load current from 0.31A to 0.35A ( 22 ohms(2W) load)
01:44 AM nuxil: you only have a gain of around 30 didnt you want 100 ?
01:44 AM polymorph: _ami_: cant open it, but if you want to current limit then you need to reconfigure your boost's ground to V+ so it becomes a voltage mirror
01:45 AM polymorph: a flyback
01:45 AM _ami_: nuxil, i did try with 100 but ckt was unstable.
01:45 AM nuxil: did you add in a cap ?
01:45 AM _ami_: No
01:45 AM nuxil: over R5
01:46 AM nuxil: you should when doing high gain. you dont want any oscilation or stuff to happen.
01:46 AM polymorph: _ami_: lmc6462, op07 or other low offset thing can do things like that
01:47 AM polymorph: op07 is cheap but needs +-5V
01:49 AM nuxil: but 100 in gain isnt much, 100 of hfe is like whats normal for a jellybean transistor
01:53 AM _ami_: nuxil, 100 gain is feasible at 5v supply to LM358
01:53 AM _ami_: ?
01:55 AM nuxil: why not? havent tried tho.
01:55 AM polymorph: http://www.ti.com/lit/ds/symlink/lm158-n.pdf
01:56 AM polymorph: lm358 can do 25V/mV gain at 5V minimum
01:58 AM _ami_: hmm. then it should have worked.. i might be doing something wrong.
01:59 AM _ami_: i shall share the ckt, 1 min
01:59 AM _ami_: schema*
02:00 AM polymorph: _ami_: it amplifies its offsut voltage too.
02:01 AM nuxil: it amplifies everythin :p
02:01 AM nuxil: noise too
02:01 AM polymorph: if it has ±10mV offset you will have ±1V out voltage with zero input at 100 gain
02:01 AM nuxil: if your offset is 1v and your gain is 10. well then thats a offset of 10V and if your supply is 5v your will clamp it to the rail.
02:02 AM polymorph: nuxil: lm358 is not r2r out though
02:02 AM polymorph: LMV358 can do that
02:02 AM nuxil: ye its single rail
02:03 AM polymorph: nuxil: check four facts
02:03 AM _ami_: https://www.enlightenment.org/ss/e-5a6ed4f7155bf9.42094871.png
02:03 AM polymorph: your
02:03 AM polymorph: The connection was interrupted - lel
02:03 AM nuxil: ?
02:03 AM nuxil: what fact?
02:04 AM polymorph: [09:00] <nuxil> ye its single rail
02:04 AM polymorph: [09:00] <polymorph> nuxil: lm358 is not r2r out though
02:04 AM nuxil: − Single Supply . . . 3 V to 32 V
02:04 AM nuxil: quote from datasheet
02:04 AM nuxil: <nuxil> ye its single rail
02:04 AM polymorph: and i was referring to output stage topology what you said
02:05 AM nuxil: yeh ?
02:06 AM nuxil: i basically said the same as you. but used 1v instead of 10mv
02:07 AM _ami_: schema: https://www.enlightenment.org/ss/e-5a6ed4f7155bf9.42094871.png
02:08 AM _ami_: i did not get much control over current.
02:09 AM _ami_: i was able to control from 0.31A to 0.35A on a 22 ohms load (6.8 to 7 V - Vout)
02:10 AM nuxil: are you a doctor ?
02:10 AM nuxil: you surly write like one :P
02:11 AM _ami_: :)
02:12 AM nuxil: your shunt is not in series with the load ?
02:12 AM nuxil: oh never mind.
02:12 AM nuxil: i see now.
02:12 AM _ami_: it is in series with load
02:13 AM _ami_: also in real circuit, i did not put any diode
02:25 AM nuxil: hmm that 10k and 100K dividor for the + input.. should it be taped off the negative side of the load ? should it not be before the load. at + rail.
02:30 AM Trangar_ is now known as Trangar
02:37 AM nuxil: i see on your kicad you got it like that.
03:03 AM _ami_: nuxil, i shall give it a try again.
03:03 AM _ami_: today when i reach home.
05:12 AM nohitzwork: what are you doing nuxil?
06:20 AM nohitzwork: im reading a book about amiga game programming
06:20 AM nohitzwork: nice find, its in finnish
06:20 AM nohitzwork: http://amiga.unikko.org/BITTI-kirjat/
07:38 AM cehteh: mhm .. running up to 3A through a 200mA through a blow fuse for few milliseconds (less than 3ms) certainly wont blow it, but will it detoriate over time? my guess is no, any opinions?
08:23 AM Emil: what type of fuse
08:28 AM Emil: if 200mA is the maximum current rating
08:28 AM Emil: then 3A for <3ms might blow it
09:56 AM polymorph: soldered wires from an ata-133 cable in a gps connecting the friggin microusb connector socket and board ;/
09:56 AM polprog: still waiting for the stms
09:57 AM polprog: :(
09:57 AM polymorph: ;<
09:57 AM polymorph: Emil: try.... 250mA F ?
09:57 AM Emil: polymorph: hm?
09:57 AM polymorph: F is fast fuse
09:58 AM polprog: is that a traditional fuse or a polyfuse?
09:58 AM polymorph: i²t rating
09:59 AM polymorph: cehteh: it first need to heat up
10:12 AM polymorph: it works, now it won't break out of the panel and still charge, fixed the soldered wires with composite concrete
10:12 AM polprog: tracking shows that its in a post office 15 minutes by tram from here.
10:19 AM cehteh: traditional fuse, dunno if fast or slow blowing kind, but even fast are speced to need 20ms, 3ms and 3A where worst case, real case is maybe half of that or less
10:20 AM cehteh: i know that it needs to heat up to blow, but was wondering about side effect which lets it detoriate (oxidizing when getting warm, phase changes etc)
10:21 AM polymorph: polprog: finally ?
10:21 AM polprog: polymorph: from what i know now it got mis-routed (i called customer support)
10:21 AM polymorph: lel
10:21 AM polprog: i think ill take a walk
10:22 AM polprog: to that post office and ask there :P
10:22 AM polprog: or i just wait a bit more, this should sort out itself
10:22 AM polymorph: cehteh: the difference between a slow and a fast fuse is thermal mass, slow fuse has a solder ball in the wire's center for example
10:22 AM cehteh: jup
10:22 AM polymorph: and it has a little thicker wire
10:23 AM polymorph: polprog: i should order my 373s too ;>
10:24 AM polymorph: ill take some 48 and some 100 pin ones
10:24 AM polprog: polymorph: 373s?
10:24 AM polymorph: y
10:24 AM polymorph: http://www.st.com/content/ccc/resource/technical/document/reference_manual/fa/06/c7/15/3b/2a/4c/f1/DM00041563.pdf/files/DM00041563.pdf/jcr:content/translations/en.DM00041563.pdf
10:24 AM polprog: whats that? i dont reacall any 373
10:25 AM polymorph: this with the superadcs
10:25 AM polprog: aah
10:25 AM polymorph: i have found that not only it has 16 bit delta sigma adcs, they also have programmable gain from 0.5-8x at input
10:26 AM polymorph: and differential input capabilty
10:28 AM polymorph: so setting the internal 1V2 bandgap as reference and gainx8 gives 1V2/(2^19) bit steps
10:29 AM polymorph: 2.29 uV
10:30 AM polymorph: @150mV range
12:12 PM polprog: ill take a walk there tomorrow after school
12:15 PM McDonaldsWiFi: I'm having a hard time find some resources to help me init the adc on my 328p with assembly
12:15 PM McDonaldsWiFi: its all C :( anyone know if any resources?
12:16 PM LeoNerd: That's a Lego problem
12:16 PM LeoNerd: I.e. you have a brick ("How to initialise the ADC from C"), and a brick ("How to use AVR peripheral registers from assembly"), so just stick the two bricks together
12:17 PM McDonaldsWiFi: and hope I don't step on them :P
12:17 PM McDonaldsWiFi: okay, ill give it another go ^^
12:23 PM polprog: you do the same as in C. look up the register adresses in the datasheet
12:23 PM polprog: unless you use arduino hal to initialize the adc, then it will be an interesting ride for ya :P
12:23 PM McDonaldsWiFi: Yeah now that I think about it I'm not sure why I was so confused
12:24 PM polprog: you can use sbi/cbi for single bytes
12:24 PM McDonaldsWiFi: well I am using an arduino as a breakout but that's it, rest is in atmel studio
12:24 PM LeoNerd: Why assembly though? If I need to use assembly in little bits of code, I write in mostly-C with small chunks of asm where required
12:24 PM McDonaldsWiFi: essentially trying to make my own analogRead() stype loop in asm :D
12:24 PM LeoNerd: Usually rare
12:24 PM McDonaldsWiFi: to learn mostly tbh
12:24 PM polprog: you already know some assembly or not?
12:25 PM McDonaldsWiFi: i'm working on making a little self driving car so I want to make the calculations tight
12:25 PM LeoNerd: In particular, things like peripheral initialisation are a one-time setup thing, so you can do those in C on startup and not worry about performance
12:25 PM polprog: ^
12:25 PM McDonaldsWiFi: I know a bit of z80 asm... so I'm not totally lost
12:25 PM polprog: thats a good start
12:25 PM McDonaldsWiFi: but it only has like... 8 registers you use
12:25 PM McDonaldsWiFi: not 30+ xD
12:26 PM McDonaldsWiFi: its been fun so far though, the adc is the only real snag I've had. I forget how much the higher level stuff hides
12:26 PM McDonaldsWiFi: I really should learn to do this in C... i mean most C compilers can beat you at ASM anyways
12:26 PM McDonaldsWiFi: at least gcc usually can
12:27 PM McDonaldsWiFi: damn guys, why you gotta talk sense into me
12:27 PM McDonaldsWiFi: xD
12:27 PM polprog: just a sec
12:28 PM McDonaldsWiFi: oh wait, i forgot
12:28 PM polprog: in C you would do (amoung other assembly init stuff) for example ADMUX = somethign something
12:28 PM McDonaldsWiFi: I'm also trying to use this as an output for my z80 build
12:28 PM McDonaldsWiFi: so.. asm will be essential to keep the wait down xD
12:29 PM McDonaldsWiFi: yeah I've found that, thats where you set the refernce right?
12:29 PM McDonaldsWiFi: so you'd set that flag to 1.. whichever you want
12:32 PM polprog: so you just take the adress of admux, IN it to, say r16, sbi/cbi/ldi r16 soemthing and finally OUT admux, r16
12:32 PM polprog: rest is language-agnostic, ie. the actual bits and values
12:32 PM McDonaldsWiFi: right
12:32 PM polprog: s/agnostic/independent/
12:32 PM McDonaldsWiFi: okay, I'm gonna try this again
12:33 PM polprog: this is my old 595 code that uses assembly routine called from C. apart from the assembly file being a standalone program
12:33 PM polprog: http://polprog.net/papiery/avr/74hc595/
12:34 PM polprog: not that i look at it, it probably wont compile since it has two main labels defined
12:34 PM polprog: but stuff shown in the C file applies, extern in C code, .global in asm
12:37 PM McDonaldsWiFi: I've never actually called asm from C before
12:37 PM McDonaldsWiFi: This sounds like a great way to learn.
12:38 PM polprog: if you go up a directory i put there some pdf i found on the net that neatly explains calling convention on AVR
12:39 PM polprog: i put it there since it took me a long time to find that
12:45 PM McDonaldsWiFi: got it
12:45 PM polymorph: [16:55] <Emil> polymorph: hm? - because it was dropped and connector broke out of board...
12:45 PM McDonaldsWiFi: thank so much!
12:46 PM polprog: np
12:46 PM polprog: theres a whole into tutorial to assembly on the index page as well
12:46 PM polprog: :)
12:49 PM McDonaldsWiFi: thanks!!
12:50 PM polprog: always happy to help
12:50 PM polprog: now im off to maths homework lol
12:51 PM nuxil: what kind of math ? :)
12:51 PM polprog: nuxil: the title of the chapter is "quadratic function with parameter". Vietas formulas, etc.
12:52 PM nuxil: :D
12:52 PM nuxil: make batman
12:52 PM nuxil: batsign
12:52 PM polprog: for example this excercise im doing right now is to find such m that 5x^2-mx+1=0 will have such roots that abs(x1-x2) <= 1
12:52 PM polprog: yeah the batman curve is cool :P
12:57 PM nohit: have you guys ever used "register" keyword?
12:58 PM nohit: i havent seen anyone using it
12:58 PM polprog: i didnt even know something like this exists
12:58 PM nohit: until today
12:58 PM polprog: lol
12:58 PM nuxil: ?
12:58 PM polprog: what does it do?
12:58 PM nohit: https://en.wikipedia.org/wiki/Register_(keyword)
12:59 PM nohit: it was used in this DSP code http://www.wolinlabs.com/blog/stm32f4.adc.dsp.dac.html
01:00 PM polprog: "When used, register is typically for loop counters, or possibly for other very frequently used variables in the code." imagine a compiler that puts your loop counter in ram everytime :P
01:00 PM nuxil: does register work on avr's ?
01:00 PM polprog: for(volatile int i =0 ...) :P
01:01 PM nohit: idk
01:01 PM nuxil: how does it know what "register" to use if you use register.. im not sure this will work well on avr.
01:03 PM polprog: oh arm m3 has only 12 general purpose register. i wonder how much pic has
01:05 PM nuxil: i think register is more for x86 stuff where you have registers such as eax, ebx, ecx so on.
01:06 PM McDonaldsWiFi: Is there a best practice use of registers you guys have seen before? Kind like a standardized usse?
01:06 PM McDonaldsWiFi: Like "r5 is usually used as a counter"
01:06 PM McDonaldsWiFi: etc etc
01:07 PM nuxil: read the datasheet :p
01:09 PM polprog: i think the only limitation is to make sure you restore call saved registers so the compiler can be happy
01:09 PM polprog: and iirc LDI can load only to upper 16 registers
01:13 PM nohit: if it works with stm32 so why wouldnt it work with avr
01:13 PM nohit: both have general purpose registers
01:21 PM polprog: on the wikipedia page on that its said that even though mplab compiler supports that, it should not be used
01:21 PM polprog: you either trust the c compiler or write the critical code in assembly :P
01:21 PM nuxil: oh. the register will not actually put something on the register. it can. but it will only suggest to do so. and yea i read too its not reccomendedt o use it
01:39 PM Jartza: in avr-gcc it should be typically safe to use r2..r15 for register keyword
01:39 PM Jartza: but it might make compiler optimize other parts of the code like crap
01:39 PM Jartza: and also it's really just a suggestion, not real "instruction"
01:40 PM Jartza: so compiler might not allocate register for your variable
01:40 PM Jartza: and vice versa, compiler might allocate register for variable even if you don't use "register" keyword
01:42 PM Jartza: I've found generally for simple code the r12..r15 being relatively safe with avr-gcc
01:46 PM polprog: wow dev-c++ on windows gave me executable file size in MiB up to 14 decimal places haha
01:49 PM Neomex is now known as Pie
01:50 PM Pie is now known as what
01:51 PM what is now known as Guest79324
01:51 PM Guest79324 is now known as who
01:52 PM who is now known as fail-o-tron
01:52 PM fail-o-tron is now known as barkos
01:52 PM barkos is now known as Escobark
01:57 PM Escobark is now known as LouisGuddog
01:59 PM LouisGuddog is now known as Pawdilla
01:59 PM Pawdilla is now known as okitsboringnow
02:00 PM okitsboringnow is now known as Neomex
02:00 PM nuxil: cant deside on a name?
02:00 PM nuxil: Neomex,
02:01 PM Neomex: just irc stuff happening somewhere on irc
02:35 PM * polprog [make] *** building target sandwich
02:35 PM * polprog Warning: ISO forbids making sandwiches out of buns intead of bread slices [-Wimplicit-sandwich-declaration]
02:39 PM Thrashbarg: man overboard
02:40 PM Thrashbarg: ~$ make love
02:40 PM Thrashbarg: make: *** No rule to make target 'love'. Stop.
02:41 PM polprog: i learnt to type man.he.net or linux.die.net instead of googling 'man date'
02:41 PM Thrashbarg: hehehe
02:42 PM polprog: plus, hurricane electric has way better man page browser than die.net
02:42 PM Thrashbarg: why not just use the command line?
02:42 PM polprog: sometimes i have 5 or 6 terminals open
02:43 PM polprog: i dont wanna open another one
02:43 PM polprog: :x
02:43 PM Thrashbarg: that's a poor excuse
02:44 PM Thrashbarg: install gman
02:45 PM polprog: hmm
02:45 PM polprog: good idea
02:45 PM polprog: oh, i forgot. sometimes i have this single putty window on windows and i just dont wanna log in again
02:45 PM polprog: so its easier to bring up firefox, which usually is running anyway
02:57 PM polprog: http://i.lvme.me/dk3sz1d.jpg
03:21 PM polprog: this is very interesting https://www.youtube.com/watch?v=a73ZXDJtU48 and contrary to some breadboard videos it doesnt take 45 minutes
03:24 PM Thrashbarg: awesome
03:25 PM polprog: my first avr project was a prime number cycler
03:26 PM polprog: it would count upwards and pause at every prime number
03:26 PM Thrashbarg: nice
03:26 PM polprog: gone up to either 1024 or 9999
03:26 PM polprog: that display board then became a lm335 based thermometer, as soon as i learnt how to adc
03:28 PM Thrashbarg: I remember my first AVR thing was a programmable timer for a friend's project. I hadn't done AVR before but I had made a homebrew 8080 computer and was proficient in assembly on that, so transferring to the AVR wasn't too painful
03:29 PM polprog: i learnt assembly on avrs. before that i had a book on z80 in polish
03:29 PM Thrashbarg: nice
03:29 PM Thrashbarg: the Z80 is a superset of the 8080
03:31 PM polprog: i think i have a clone of 8080, UCY7880. needs some sick double clock. normal z80 i think could run off square wave
03:31 PM polprog: i guess i could make an analog generator with phase shifted clocks or something like that but this chip has more of a historic value now
03:33 PM polprog: https://puu.sh/ybLrA/74185fea6d.jpg
03:33 PM polprog: long one on the right
03:33 PM Thrashbarg: nice!
03:34 PM Thrashbarg: you'll find the 8080 uses a really old MOS process and needs +5V, +12V and -5V. The clock signals need to be 12V
03:34 PM polprog: it was used in Elwro 600 and Elwro 500 computers
03:34 PM polprog: this one runs off 5V
03:34 PM nuxil: god damn.. i was cleaning my bench with a vacum clean. then i hear ding ding ding.. and there goes 3 atiny85 .. now i need to cut this crap open to find them .lol think i need to go outside and do that.
03:34 PM Thrashbarg: wow okay
03:34 PM Thrashbarg: :/
03:35 PM polprog: lol
03:35 PM polprog: i once sucked a potentiometer head from my synth :( it took me 15 mins to find it
03:35 PM nuxil: lesson. clean you bench for chips before starting the vacume cleaner :p
03:37 PM Thrashbarg: polprog: https://pl.wikipedia.org/wiki/MCY7880 <-- says it needs the three voltages :/
03:37 PM Thrashbarg: unless the UCY is different
03:37 PM polprog: oh missed that :(
03:38 PM polprog: next chapter says "Characteristics of MCY7880" and under there "supply voltage 5V +- 0.25
03:38 PM polprog: "
03:38 PM Thrashbarg: yea
03:38 PM polprog: skimmed this article quickyl
03:39 PM Thrashbarg: there's a site around where they're collecting a database of characteristics of various 8080 processors
03:39 PM Thrashbarg: by running a rigorous test program on them you can submit the results
03:41 PM polprog: hmm
03:41 PM polprog: well, if i ever got a working system out of those
03:41 PM polprog: i think i would like an 8086 though
03:42 PM Thrashbarg: they're a little more cumbersome to work with. Multiplexed data/address bus, and you need a clock generator (again)
03:42 PM polprog: damn
03:42 PM polprog: well, stm32 is is then lol
03:42 PM Thrashbarg: the 8085 is good in that it has the clock generator built in
03:42 PM polprog: o
03:42 PM Thrashbarg: it's not difficult to find the clock generator, they're about
03:42 PM Thrashbarg: The Z80 is similar, just put a TTL clock on its CLK pin
03:43 PM polprog: i think thats all you need from the clock side
03:43 PM Thrashbarg: yeah
03:43 PM polprog: if i get my hands on any z80 ill surely take a loot at the book
03:44 PM Thrashbarg: cool
05:44 PM green_snow: hello, anyone has some experience with connecting a MIDI Input to an AVR? I looked through articles and most of people used the 6N138 optocoupler, but some suggest that it's too slow for MIDI... what is the truth?
05:45 PM LeoNerd: MIDI is 31.25kBaud
05:45 PM LeoNerd: 6N138 datasheet I have here says it's good up to 100k
05:45 PM LeoNerd: https://www.vishay.com/docs/83605/6n138.pdf
05:46 PM LeoNerd: ... wow, I've never seen that package style before. A surface-mount DIL-8
05:47 PM polprog: yeah, optos have some funny packages
05:48 PM green_snow: thank you :)
05:48 PM polprog: like LTV 847 has this funny dip variant
05:48 PM LeoNerd: I've also used some 6N variant (125? I forget) on DMX-512, and that runs at 250k
05:48 PM LeoNerd: Though these days I'd likely use one of the electromagnetic isolators instead
05:49 PM polprog: i should take a look at one of thosr ethernet isolators. they must be fast
05:49 PM LeoNerd: Trickier driving circuitry
05:49 PM polprog: i have a dead nic that i can use as a donor
05:49 PM LeoNerd: I meant the all-in-one SOIC16-W chips that Analog Devices etc.. all make
05:50 PM LeoNerd: There's some nice bidirectional ones if you need I²C and so on, too
05:51 PM polprog: one would think those transformer-like ones are bi directional
05:51 PM LeoNerd: At the component level yes they are, but the driver circuitry around them makes them not
05:51 PM LeoNerd: You put a differential driver on one end and a receiver on the other
05:51 PM polprog: hmm
05:52 PM polprog: ill poke them around with rs485 driver then
05:52 PM LeoNerd: Ah, Isolated RS-485 has its own solutions. There's lots of (almost)all-in-one chips for those
05:52 PM LeoNerd: I sell some :) - https://www.tindie.com/products/leonerd/rs-485-isolated-application-board/ :)
05:53 PM polprog: the aim is to get a general feeling of how those electromagnetic isolators work. 485 drivers are just one way to get them
05:53 PM polprog: i like to sometimes poke around without a particular aim
05:53 PM LeoNerd: The all-in-one chips are so self-contained you don't really need to know how they work
05:53 PM LeoNerd: They act just like a 74'240 or similar.. I.e. just a logic buffer
05:54 PM polprog: im too young to know what a 74'240 is off my head
05:54 PM polprog: i know its 7400 series
05:54 PM LeoNerd: 8 channels logic buffer
05:54 PM LeoNerd: But yah; I suppose none of these chips do 8 channels. Usually 2 or 4 is more common
05:55 PM LeoNerd: But really, you just give them power + ground on both sides, then the rest of the pins are logic in or out pins
05:55 PM polprog: ok, now i know
05:55 PM LeoNerd: Suuuper-easy :)
05:56 PM polprog: kinda like the ..3002 chip china uses in their stepper drivers
05:57 PM polprog: ?
05:57 PM LeoNerd: Not sure which chip that is
05:58 PM polprog: umm
06:00 PM polprog: ULN2003
06:00 PM polprog: my bad
06:01 PM LeoNerd: Oh them... kinda similar perhaps in idea
06:03 PM polprog: very popular here cos you can connect a motor to an arduino and it doesnt blow up... at least long enough that you get bored and take it apart lel
06:04 PM LeoNerd: Yeah; it's not -really- a logic chip though. It's really just 7 channels of darlington transistor array in a convenient package
06:05 PM polprog: yeah
06:06 PM polprog: doesnt replace a proper motor driver
06:06 PM polprog: it works reasonably well for bitbanging steppers
06:06 PM LeoNerd: It's also not an H-bridge, like the L293
06:06 PM polprog: i was about to say this
06:06 PM LeoNerd: Don't steppers still need an H-bridge?
06:08 PM polprog: i think i have one that doesnt
06:08 PM polprog: iirc each coil has one side common and one connevted to that chip
06:09 PM LeoNerd: Hmm.. ohright, that arrangement
06:09 PM polprog: its a chineese stepper motor though. i have some real steppers that i could try and drive with some beefy mosfets i have at hand
06:10 PM polprog: in a h-bridge arrangement
06:11 PM LeoNerd: Yah.. you're not going to manage any of the larger types of stepper with an L293
06:11 PM polprog: or.. whatever i come up with
06:12 PM polprog: i have some MSO033something crazy high current mosfets and some beefy BUZ.. TO92 mosfets. plus i just scavenged some medium sized heatsinks from a dead psu
09:40 PM sabor_ is now known as sabor
11:38 PM day__ is now known as day