Back
[00:41:31] <phinxy> Have anyone successfully cancelled a PCB order from Seeedstudio Fusion where the order was "processing" ?
[00:42:02] <phinxy> such a bummer, noticed a big mistake on my design
[00:50:28] <phinxy> designing PCBs was so much easier than what i imagined. It's almost scary how easy it is because if you come up with a good board you could easily start assembling it and make some nice income
[00:50:50] <x29a> did you contact seeed?
[00:51:14] <phinxy> I sent a message using the contact form a couple hours ago. maybe they are sleeping over there
[00:52:15] <Casper> I'ld expect it to be too late
[00:53:10] <Casper> usually if a compagny say it is processing, it mean it has been sent to be produced, they can't cancel at that stage
[00:53:28] <phinxy> Anyone of you who have a PCB layout you sell to people? Like a semi-business
[00:54:03] <phinxy> Casper, Ok. i guess i could build a litter box for the cat with them or something
[00:54:21] <phinxy> \s
[00:55:40] * Casper wonders about the safety of that
[01:06:18] <flyback> wow inflex ltns
[01:06:25] <flyback> how are days since you retired your shop
[01:06:29] <flyback> did you ever find anythign else?
[01:07:41] * Casper reverses the wire of flyback
[01:07:46] <flyback> BMCC
[01:10:42] <Casper> how is you?
[01:15:03] <Casper> bed time, nite
[01:23:54] <inflex> hi flyback
[01:24:17] <inflex> flyback: these days I repair computers and phones... doing a lot of data-recovery too.
[01:24:41] <inflex> I still build various AVR projects to help me with certain parts of the job, particularly with automating data recovery processes
[01:25:11] <flyback> cool
[01:25:32] <inflex> What have you been up to?
[01:25:38] <flyback> vegitating mostly
[01:25:44] <flyback> cleaning out basement getting rid of stuff
[01:25:54] <flyback> getting my data recovery project going
[01:26:32] <flyback> still have same main job
[01:29:13] <flyback> just got a new iron aoyue 2930
[01:29:24] <flyback> but still having troble on rohs stuff like some routers oh well
[01:29:56] <flyback> on high blood pressure meds now and I need bifocals
[01:30:02] <flyback> right in time for 42th bday
[01:51:13] <inflex> What's your data recovery thing?
[01:51:31] <inflex> Oh joy... yes, I've got a few months of 42 left... and yes... damned high BP
[03:49:25] <julius> hi
[06:34:38] <Lambda_Aurigae> non-avr question.... 1280 logic cells in an FPGA is teeny tiny, yes?
[06:36:42] <Lambda_Aurigae> and 7860 isn't exactly large I'm guessing.
[06:54:40] <Tom_itx> very teeny tiny
[06:55:01] <Tom_itx> 7860 would be considered tiny in today's standards too
[07:17:51] <theBear> kinda like how in the 80s a few mhz was a fast processor, and now a few ghz is average
[07:18:01] <theBear> that kinda teeny tiny
[07:35:08] <Tom_itx> and we're not getting any more done now than we did then
[07:45:38] <julius> well, theres more porn to download. takes more time
[07:47:47] <julius> flyback, the flyback diode in my circuit greats you
[07:47:53] <julius> for whatever reason it is still alive
[08:01:56] <vaskozl> Hey can anyone help me with some code:
https://skozl.com/s/main.c ?
[08:02:04] <vaskozl> Basically it's a line following robot.
[08:02:21] <vaskozl> When both sensors read white currently the motor turns off. What I need to do instead is make it do a spiral.
[08:02:31] <vaskozl> I'm using interrupts but I don't really understand them.
[08:02:51] <vaskozl> When I uncomment the code in the for(;;){} in main it doesn't do anything.
[08:04:35] <vaskozl> When I'm expecting it atleast perform a spiral.
[08:05:02] <theBear> first up interrupt handler routines should always be as very minimal as they possibly can to still achieve something useful, otherwise the concept doesn't really work
[08:07:34] <theBear> hmm, i guess you using servos then
[08:07:36] <theBear> ?
[08:07:41] <vaskozl> yup
[08:09:15] <theBear> hmm, is that tune really a tune or more of a clunking ?
[08:09:43] <apo_> vaskozl: if (OCR0B = OCR0A = 0) {
[08:09:47] <apo_> that assigns
[08:09:51] <vaskozl> oh
[08:09:52] <vaskozl> fuck
[08:12:10] <apo_> (And even if you used comparisons it probably wouldn't do what you expected)
[08:12:39] <theBear> oh yeah, if i hadn't been making coffee, my next question was wtf two equals in a row means
[08:12:51] <apo_> :)
[08:13:05] <apo_> theBear: it checks if OCR0B equals (OCR0A == 0), of course
[08:13:41] <theBear> i'm outta practice see, and while i pretty sure it should work that way even if it doesn't please my eye, i always preferred something like while (1) for an infinite loop
[08:13:50] <theBear> heh of course
[08:14:25] <apo_> for (;;) // for cthulhu
[08:15:28] <theBear> but you can'
[08:15:37] <theBear> t do nothing to nothing until it looks like nothing
[08:15:57] <theBear> and you can't deny that 1 is always not 0
[08:16:37] <apo_> hm?
[08:33:42] <mohsen_> Which gyroscope module do you suggest to use?
[08:39:34] <LeoNerd> The MPU6050 seems to be popular... at least, most of the usual suspects (Adafruit, SparkFun,...) make breakout boards for it
[08:42:22] <mohsen_> Will you explain me what is exactly a breakout board?
[08:43:41] <LeoNerd> Typically, some sort of sensor or whatever chip, mounted on a board with convenient access pins for all the IO lines you'll need
[08:43:48] <LeoNerd> usually useful for breadboarding and the like
[08:44:06] <LeoNerd> Often people put "essential" external components like decoupling caps and similar on that board, to make it easier to use
[08:46:36] <Jartza> I just made my own breakouts for xmega
[08:46:46] <Jartza> with pins for crystal and caps
[08:47:05] <mohsen_> Oh, a breakout board is what I call a module.
[08:48:58] <mohsen_> Does the chip specify the data IO interface(I2C,UART..) or the breakout board?
[08:48:58] <mohsen_> Jartza: Isn't any for xmega in the market?
[08:49:37] <Jartza> sure there are
[08:49:50] <Jartza> but most of them are just for plain chip
[08:49:58] <Jartza> I wanted crystal and it's caps there too
[08:50:03] <Jartza> makes it easier to breadboard
[08:50:49] <Jartza> I've been lately playing with LSM6DS3 accelerometer/gyroscope
[08:50:52] <Jartza> it's nice chip :)
[08:51:01] <Jartza> it can also act as autonomous i2c hub
[08:51:14] <Jartza> so you can collect data from other sensors too while MCU is in sleep
[08:51:22] <Jartza> and just periodically flush the data to MCU
[08:51:31] <Jartza> it has internal 8kB fifo buffer
[08:53:16] <mohsen_> I see
[08:53:16] <mohsen_> Hmmm
[08:54:50] <mohsen_> Do all gyros need to be recalibrated after sometime?
[08:59:22] <cehteh> mohsen_: look at mpu9150 thats gyro, acc and mag in one
[09:00:25] <cehteh> not as good as others and you need to be careful with mounting (away from current which produces magnet fields) ... and usually you want a own vreg just for the sensors
[09:01:24] <mohsen_> cehteh: No I want only a gyro..
[09:01:32] <cehteh> and you need to recalibrate a gyro, constantly by other sensors or on demand. some software also stores complex config data about temperature calibration curves (3 degree polynom)
[09:01:42] <cehteh> for what applications btw?
[09:03:47] <mohsen_> I want to calculate the drawn angle based on the duration of movement
[09:03:47] <mohsen_> in degrees
[09:04:06] <cehteh> drawn?
[09:05:07] <cehteh> for short timespans you can only zero the gyro when you know its not moving (accelerometer comes handy)
[09:06:13] <cehteh> but when you want to pick up longer slower things, it gets more complicated, not to mention that earth rotates once every 24hours, maybe to slow for measuring it, but enough to add an error to long time measurements
[09:07:10] <mohsen_> I see
[11:19:27] * flyback bites julius
[11:19:47] <flyback> so
[11:20:07] <flyback> has avr caught up to ti msp430 yet in inchworm running speed low power modes
[11:21:01] <LeoNerd> I'm not sure I can parse that sentence
[11:21:08] <LeoNerd> In fact I'm reasonably sure I can't. :)
[11:21:15] <flyback> msp430 has some super low power modes
[11:21:19] <flyback> where it barely ticks over
[11:21:53] <flyback> and some coma modes
[11:21:58] <flyback> where it just retains
[11:22:03] <flyback> state and ram
[11:22:06] <LeoNerd> Well... all AVR chips have SLEEP
[11:22:30] <LeoNerd> Some of the newer chips (e.g. tiny1634) let you change the main CPU clock on the fly, too
[11:22:38] <flyback> nice
[11:22:42] <LeoNerd> So maybe you could swap that to the 32kHz ULP if you needed low power but sleep wasn't enough
[11:23:37] <flyback> I love power scaling :)
[11:25:11] <flyback> I do realize avr is more of the hot rod class but still scaling is damn useful
[11:25:35] <flyback> if not for power then for survivng a bad situation, over/under temp
[11:27:15] <flyback> espically the wireless the wireless stuff
[11:27:19] <flyback> err
[11:27:25] <flyback> isorry it's early for me
[11:27:30] <flyback> brain not awake yet
[11:28:40] <flyback> low power wireless stuff
[11:30:01] <flyback> I have to laugh when you get into situations where the led is the power hog
[11:31:15] <cehteh> happens easily ..
[11:33:35] <cehteh> runnin attiny at 128khz, active power is then less than 100µA at 5V ... red led with resistor draws 200x that much
[11:33:44] <flyback> how low do you guys think they can go voltage wise before they hit a physics limit
[11:33:52] <flyback> cehteh, pretty cool
[11:34:19] <flyback> I seen 0.6v logic
[11:35:00] <LeoNerd> ECL
[11:35:46] <cehteh> current would be limited by transfering single electrons ... voltage dunno, thats more the 'urge' the electrons want to go from one to another potential
[11:37:13] <flyback> I was reading up on some company doing some weird type of trasistor process claiming they could get a arm cortex m0 down to msp430, etc low power levels
[11:37:43] <cehteh> you can do pretty amazing stuff under lab conditions
[11:38:02] <flyback> no this is actually a released product
[11:38:11] <aandrew> jesus fuck man you're everywhere, flyback
[11:38:15] <flyback> probably a good bit expensive though
[11:38:21] <flyback> aandrew, thx
[11:38:39] <flyback> CANUCK
[11:38:51] <aandrew> don't make break out that pic, man... don't do it
[11:38:57] <flyback> what pic?
[11:39:11] <flyback> and I was on this network yrs before I met you merry lot on efnet
[11:39:22] <flyback> invited here in the early days by my friend lilo
[11:39:34] <flyback> who sadly we had a fight and he temp klined me then died :(
[11:39:43] <aandrew> *nods* I'm as hold as the hills too, just don't remeber ever seeing your name in here
[11:39:56] <aandrew> yeah it was kind of shitty when that happened
[11:41:16] <flyback> have you canucked with dash7 yet?
[11:41:26] <flyback> looks neat considering it goes thru water and concrete
[11:41:30] <aandrew> nope
[11:41:46] <flyback> milspec but they released to public also
[11:41:52] <flyback> 4xxmhz band
[11:41:56] <aandrew> interesting
[11:42:12] <aandrew> so what're you doing that's uberlowpower on AVR?
[11:42:28] <flyback> oh I haven't even learned to code yet mabye later on in the year :)
[11:42:41] <flyback> came here to ask some soldering iron advice :)
[11:42:50] <aandrew> I'd built an infloor heating pump controller that used a capacitive reactance power supply and fired triacs to drive the relays between sleeping
[11:42:58] <flyback> oh nice
[11:44:24] <flyback> capacitive reactance not heard of that but by the name, does that have to do with improving your pf
[11:44:37] <aandrew> no
[11:44:44] <aandrew> it's just a dirt cheap power supply but has shitty current
[11:44:49] <flyback> ah
[11:45:05] <aandrew> you have an RC in series as your current limiting device
[11:45:18] <aandrew> 120V -- RC -- zener -- 3V out
[11:45:29] <flyback> bonus is you can use it to heat the water
[11:45:31] * flyback ducks
[11:45:32] <aandrew> er rather 120VAC -- RC -- diode/zener rectifier -- 3V out
[11:45:48] <aandrew> well no that's the thing... superlow current
[11:46:07] <aandrew> if I used a regular bridge and rectifier you'd end up with a lot more space being used and a lot of power dissipated across the regulator
[11:46:31] <flyback> I hope you got a couple crowbar diodes in there though
[11:46:47] <aandrew> you mean TVS? yeah
[11:46:51] <flyback> yeah
[11:47:21] <flyback> doesn't sound like a bad solution for low current but I woudl think it would have 0 resistance to a spike coming in
[11:47:43] <flyback> tvs are cheap ;)
[11:47:58] <aandrew> yes. you have to manage the momentary current spike in the zener but the tvs clips it to something reasonable, at least until it blows :-)
[11:48:16] <aandrew> but the spikes are usually high voltage but really low energy which semiconductors are decently good at surviving
[11:48:21] <aandrew> it's all about heat
[11:48:22] <flyback> cool
[11:48:43] <flyback> oh i thought with fets any voltage over punches a hole in the gate even low current
[11:49:20] <aandrew> with a fet, yes, because it's got a super thin insulator which has a (pretty low) voltage withstand rating
[11:49:24] <flyback> I work in IT and man have I seen some carnage from shitty psu's with 0 over voltage protection
[11:49:26] <aandrew> puncture the insulator and it's game over
[11:50:02] <flyback> client came in with some white box pc someone sold them, shitty powmax psu, king of firestarters BLEW THE CHIPS OFF THE HARD DRIVE
[11:50:05] <flyback> I kid you not
[11:50:36] <x29a> thanks for bringing "firestarter" back to my mind
[11:50:57] <flyback> one good thing about all this green stuff
[11:51:04] <flyback> and the 80 plus psu spec
[11:51:12] <flyback> they put in a crowbar mandate in the spec :)
[11:51:34] <flyback> crowbar is a zener and scr
[11:51:54] <flyback> if the psu loses it, it shoots the psu in the head saving the rest of the box
[11:53:40] <flyback> speaking of over voltage death, appears my ftdi2232 fell out of it's static bag and rolled around in my work bag too many times, it's gone
[11:53:42] <flyback> oh well
[11:54:36] <flyback> nice chatting with you, gotta head to work
[11:55:52] <flyback> aandrew, btw some of the newer server psu's in dells, etc are digital controlled and have a data bus
[11:56:07] <flyback> so the server can work with the psu when changing load
[11:56:14] <flyback> also monitor wattage used etc
[11:56:39] <flyback> I laughed when I saw firmware updates for a psu of a client's server
[11:56:43] * flyback bites HaibaneNero
[11:56:50] * flyback catches fire and dies
[11:57:20] <HaibaneNero> uhm, thats only how "Nero" works, flyback
[11:57:35] <HaibaneNero> in the other case, you'd just get slapped
[11:57:59] <HaibaneNero> i have no idea why i just woke up. Is 6:30 PM here
[11:58:08] <HaibaneNero> and i don't even know why i was even sleeping
[11:58:24] <flyback> did you see the 1000 ways to die episode about the peppers at indian reservation
[11:58:38] <HaibaneNero> No
[11:58:47] <flyback> some big shot texan decides he wants to try one since he's been in pepper eating contests
[11:58:57] <flyback> it starts burning the canuck out of his mouth
[11:59:11] <flyback> he panics and goes into the guys fridge and gulps down the first liquid he sees
[11:59:20] <flyback> which happens to be snake venum
[11:59:23] <HaibaneNero> -.-
[11:59:27] <flyback> normally harmless when drank
[11:59:34] <flyback> but all the sores in his mouth exposed blood
[12:00:03] <flyback> he died, true story
[12:00:06] <HaibaneNero> Was he nominated for the Darwin Awards?
[12:00:12] <flyback> probably
[12:00:43] <flyback> I want to use the reenactment blubbering convulsions as a ringtone
[12:01:55] <flyback> some of them I am glad the person died (not that one)
[12:02:26] <flyback> like the doctor who was trying to murder people on bicycles for getting in his way and he backended a truck and got a rebar between the eyes
[12:04:02] <flyback> HaibaneNero,
http://www.spike.com/video-clips/c9z5ey/1000-ways-to-die-snake-chug
[12:04:41] <flyback> 1:52-55 wanna make a ringtoine
[12:06:01] <flyback> bbl
[12:22:38] <julius> my timer0 currently fires its ISR every ms, do you guys have some code that updates variables showing that 1s, 10s, 20s went by?
[12:22:58] <LeoNerd> Didn't we do this yesterday?
[12:23:14] <julius> not with me
[12:23:15] <LeoNerd> static uint16_t counter = 0; counter++; if(counter >= 1000) { printf("That was a second\n"); counter -= 1000; }
[12:23:16] <LeoNerd> and so on
[12:23:25] <LeoNerd> I'm pretty sure I mentioned exactly that, to you
[12:23:32] <julius> oh
[12:23:40] <julius> im sorry, must have missed that
[12:23:49] <LeoNerd> Obviously just count to 10000 or 20000 for longer durations
[12:23:50] <julius> was more concerned about pwm
[12:24:56] <julius> sure...but then i set a variable like 10s_have_passed. but when do i reset it? in your code the counter would reset immediately, would the "if" statement in main still pick up on it?
[12:25:36] <liwakura> depends on what you want to do
[12:26:03] <LeoNerd> See what I just did :)
[12:26:27] <julius> oh, you reset in the main
[12:26:29] <julius> as you said
[12:28:13] <julius> you forgot to make the variable volatile
[12:28:34] <LeoNerd> No Ididn't
[12:28:48] <LeoNerd> I didn't need to. This variable is only accessed by the ISR
[12:28:57] <LeoNerd> It's static /inside/ the ISR
[12:29:51] <julius> but then you set it and immedialtely, after 1ms reset it
[12:30:04] <liwakura> i see this is is quite more of an C issue
[12:30:05] <julius> what if the main code needs more than 1ms to execute?
[12:30:37] <LeoNerd> Yeah; don't run it within the ISR :)
[12:30:43] <liwakura> i thought you can't do interrupts while in an ISR
[12:31:13] <LeoNerd> Personally what I'd do in that situation would be simply ISR(TIMER0_OVF_vect) { static uint16_t counter = 0; counter++; if(counter >= 1000) { task_schedule(TASK_1HZ); counter -= 1000; } }
[12:31:26] <LeoNerd> Because task_schedule just touches some volatile memory flags
[12:32:38] <julius> yes, that makes sense
[12:32:44] <liwakura> has someone of you an uart ISP which fills a ringbuffer? as code?
[12:32:46] <LeoNerd> Then the main body of the program is running the task function and if it happens to take more than 1msec that doesn't matter.. the ISR will briefly interrupt it, count that tick, then let it resume
[12:32:49] <liwakura> *ISR
[12:33:45] <liwakura> otherwise im just going to code it myself, but its like my 4th program
[12:34:04] <LeoNerd> If only there was some way to share C code between multiple applications...
[12:34:29] <liwakura> ah...
[12:34:36] <liwakura> is such a thing in the stdlib?
[12:34:43] <LeoNerd> I doubt it
[12:40:07] <aandrew> flyback: I could see current monitoring but I would not want the PSU to be controlled by the MB. the PSU should be entirely hardware controlled. SW has an ugly habit of having bugs.
[12:56:48] <DonaldDuck> Anyone have try the thread lib for small IC like Attiny. I have several trigger to look, doing it in a single thread. But was wondering if multi-thread would work or worth it.
[12:57:24] <julius> LeoNerd, thank you. that solved my problem
[13:36:13] <cehteh> DonaldDuck: i am working on a scheduler, not threads, but functioncalls scheduled
[13:37:00] <cehteh> multithreading is a lot waste of memory, and tinys have little
[13:37:21] <LeoNerd> My scheduler is tasklet based
[13:38:01] <cehteh> tasklets as in?
[13:38:13] <LeoNerd> task functions that run and then return status codes
[13:38:38] <cehteh> ah
[13:39:06] <DonaldDuck> cehteh, Ah ok thanks, anyways to sneak a peak on your scheduler?
[13:39:08] <cehteh> so far i have no returns, maybe i consider it (but mhm what? :D)
[13:40:15] <cehteh> git://git.pipapo.org/muos ... very work in progress
[13:41:00] <cehteh> actually i am developing on a arduino nano, have some attiny85 here but havent ported it yet
[13:41:15] <cehteh> would need only minimal changes, but still
[13:41:39] <cehteh> ah yes .. and no much docs yet :D
[13:42:12] <DonaldDuck> Well the code is already a good step ;)
[13:42:14] <cehteh> http://git.pipapo.org/?p=muos;a=blob;f=src/example.c
[13:42:42] <cehteh> it runs :) time based scheduler and high priority and low priority queues
[13:43:04] <cehteh> and quite advanced UART lib with lineeditor, completed yesterday
[13:43:46] <DonaldDuck> Ok. I was doing a small power manager for a project. And had a blinking LED until the attiny received confirmation that the system shutoff properly. But blinking was delay and not consistent... I've found a way around and it made me wondering about this.
[13:44:35] <cehteh> http://public.pipapo.org/pulseview1.png da blink :D
[13:45:47] <beanbag-> aandrew,
https://www.arm.com/products/system-ip/cordio-radio-cores/index.php
[13:45:58] <cehteh> channel 5 is a led blinking at 2hz
[13:45:58] <DonaldDuck> cehteh, thanks. Will look more into it ;)
[13:46:00] <beanbag-> I think this is the company I mentioned earlier with the subvolt logic arm
[13:46:14] <beanbag-> claim they can get a cortex m0 down to msp430 power levels
[13:46:25] <cehteh> i am planning to port it to tinys .. but patches are welcome too
[13:46:50] <beanbag-> oh wait that's arm itself my bad
[13:46:52] * beanbag- facepalms
[13:51:00] <DonaldDuck> cehteh, I've pull the project thank. Looking at it ;)
[13:56:24] <beanbag-> WTF
[13:56:28] <beanbag-> A T-STORM IN FEB
[13:56:36] <beanbag-> MOTHERCANUCKING moosesucker
[14:00:23] <julius> i got this function: motor_init(); first line in the function is: PORTB ^= (1 << LED1); which works nicely, led goes on. after that comes my pwm initialization code TCCR1A and company. any idea what could prevent the avr from running after the led?
[14:02:36] <julius> oh wait, im actually setting a variable to 1: pump_is_running = 1; commenting that out seems to make the code go on...but why? theres just the PWM initialization, OCR1B, TCCR1A, TCCR1A...
[14:03:35] <DonaldDuck> julius, without seeing your code. We are no psychic. Use paste bin or a website like that to share it
[14:10:53] <beanbag-> oh this is cool
[14:10:54] <beanbag-> http://www.networkworld.com/article/3037088/researchers-make-low-power-wi-fi-breakthrough.html
[14:11:15] <beanbag-> only up to 11mbit but 10,000 times less power usage on a wifi link
[14:11:24] <beanbag-> yeah i'll take the speed hit for a ton of apps
[14:21:11] <beanbag-> am I stupid for trying to use braided wick on via's and other thru holes
[14:21:22] <beanbag-> seems like I lost too many trying to do this on routers etc
[14:21:32] <beanbag-> or are routers just made from shit material that dies easily
[14:57:50] <Jartza> evening
[15:48:59] <beanbag-> http://realterm.sourceforge.net/
[15:52:28] * beanbag- gotta go not feeling well
[18:45:53] <julius> DonaldDuck, solved it....would be to much work in shortening the code. thanks
[18:52:10] <julius> on the last page of this humidity sensor datasheet they write: avoid polarizatio, use alternating current/voltage. does this mean that the resistance measurement should switch + and - at every measurement?
http://www.elecrow.com/download/HR202%20Humidity%20Sensor.pdf
[18:53:50] <Lambda_Aurigae> if that is what it says.
[19:09:53] <lorenzo> julius: it's AC powered, yes
[19:10:48] <lorenzo> you also probably want some signal conditioning in between, such as an opamp
[19:11:47] <lorenzo> horrible sensor anyway :-) I wouldn't even bother, grab a dht22 or mcp9808 or similar
[19:53:29] <Evidlo> So I've been struggling for a while with this problem. I'm trying to create 4 short pulses and 4 long pulses on, but I'm getting some weird output:
http://i.imgur.com/KA6DcET.png
[19:53:35] <Evidlo> my code:
https://dpaste.de/DxU4
[19:54:33] <Evidlo> My idea was to use PWM mode and change OCR1A and OCR1C to get 50% duty cycle pulses of varying length
[19:55:10] <Casper> there is a few way to trigger the interrupt
[19:55:40] <Casper> one is at the compare, one is at overflow... then more than one way to update the value... immediate, top, bottom..
[19:55:48] <Casper> I forgot which one is the best...
[19:56:12] <Evidlo> I'm using overflow, which interrupts at OCR1C
[19:57:54] <Evidlo> Is that considered an overflow?
[19:59:22] <Casper> I'ld have to analyse the code and read documentation, but it's too late for me for that
[19:59:54] <Casper> or too early... because, now I'm somewhat tired, but soon I'll be wide awake... as always...
[19:59:59] <Casper> ... damn sleep issue...
[20:00:03] <julius> lorenzo, ok thx. will take a look tomorrow
[20:00:08] <Lambda_Aurigae> sleep is for the weak.
[20:00:20] <Lambda_Aurigae> chug another gallon of jolt cola.
[20:01:54] <Casper> hmmm
[20:02:03] <Evidlo> Is it possible to interrupt on OCR1C?
[20:02:05] <Casper> good idea! I'll make hot chocolate!
[20:06:54] <Casper> on compare match?
[20:08:23] <Evidlo> yes
[20:09:17] <Evidlo> to get variable pulses, I need to be able to set OC1A at one compare match, then clear it and reset TCNT1 on another compare match
[20:13:45] <Evidlo> I think I've almost got it figured out
