#avr Logs
Jun 14 2017
#avr Calendar
12:12 AM wirts-leg: i just saw a drawing of a 1-wire programming interface that i can't make sense of
12:15 AM wirts-leg: seems to tie TX and RX from an FTDI together and feed them to a single pin.
12:18 AM wirts-leg: but wouldn't that cause problems?
12:28 AM Casper: no
12:28 AM Casper: err
12:28 AM Casper: maybe
12:28 AM Casper: do they use a resistor on the tx pin?
12:32 AM wirts-leg: i can link you a drawing
12:32 AM wirts-leg: https://github.com/4712/BLHeliSuite/blob/master/Manuals/How%20to%20Build%20a%20BLHeli%20bootloader%20interface%20with%20USB-UART%20board.pdf
12:33 AM wirts-leg: at the bottom of that page
12:34 AM Casper: actually, let me check something, it might be totally right
12:34 AM Casper: yes it is
12:34 AM Casper: the diode is the magic element
12:34 AM wirts-leg: what do you mean casper?
12:34 AM Casper: iirc, serial port idle high
12:35 AM Casper: the diode there prevent the 5V from flowing 'out'
12:35 AM Casper: instead they use a resistor to provide the 5V, which can safelly be grounded
12:35 AM Casper: so basically, they use it as an open collector
12:37 AM wirts-leg: ah that idle high bit makes it all so clear
12:37 AM wirts-leg: thanks
12:49 AM wirts-leg: i have 16 of the devices in question, and i've been soldering parallel port cable wires to the SPI pads in order to configure them.
12:51 AM wirts-leg: it's starting to feel like using their 1-wire solution is worth doing. since it means just sticking a plug over a pin instead of all that soldering and then desoldering.
01:05 AM c4017w is now known as c4017w_
01:05 AM c4017w_ is now known as c4017w__
01:07 AM rue_bed: pinpad
01:07 AM rue_bed: or chipclip
03:43 AM day is now known as daey
03:51 AM day is now known as daey
05:31 AM Tachawai is now known as Tachyon`
09:07 AM Emil: My hot air blower arrived today
09:07 AM Emil: Customs value was declared as 6.27$ :D
09:07 AM Emil: But it works great
09:07 AM bss36504: like for rework?
09:07 AM Emil: It's quite big
09:07 AM Emil: and the controls are manual
09:07 AM bss36504: thats what she said
09:07 AM Emil: but for the price it's very good
09:07 AM Emil: hehe
09:07 AM Emil: bss36504: for rework and reflow, yeah
09:08 AM bss36504: cool
09:08 AM Emil: I'll probably take it apart and try to make it a) digital control and b) a bit smaller
09:08 AM bss36504: I bought one from sparkfun a few years ago. Still going strong
09:08 AM bss36504: what is it?
09:08 AM Emil: http://www.ebay.com/itm/450W-220V-LCD-Display-Electronic-Hot-Air-Heat-Gun-Soldering-Station-Nozzle/371785689421
09:08 AM Emil: This is what I bought
09:08 AM bss36504: Oh interesting. I havent seen one fully integrated into the wand like that
09:09 AM Emil: Will be a great little project making it digital
09:09 AM bss36504: https://www.sparkfun.com/products/10706
09:09 AM bss36504: That's the one I have
09:09 AM bss36504: Works great
09:09 AM Emil: Lookin through the hot end I think it has the thermistor at the tip
09:09 AM Emil: Where it should be
09:09 AM Emil: I bet I could easily add some pid control
09:10 AM Emil: bss36504: heh, that looks a bit like metcal
09:10 AM bss36504: Might be
09:10 AM bss36504: I think it's just re-branded china import though
09:10 AM Emil: But yeah, this is a lot larger compared to others
09:11 AM Emil: Makes it harder to use
09:11 AM bss36504: I
09:11 AM bss36504: Yeah I'd expect that
09:11 AM bss36504: With the blower and everything integrated like that
09:11 AM bss36504: I like the bench-box ones personally
09:11 AM Emil: But I could try to add a longer metal pipe so that working under the microscope works
09:11 AM Emil: Yeah, bench ones are more pro
09:11 AM Emil: but this one is just cute :3
09:12 AM Emil: Also my pcbs from itead arrived
09:12 AM Emil: Obviously not as great quality as oshpark
09:12 AM Emil: but nice all the same
09:12 AM Emil: and the price is unbeatable :D
09:13 AM Emil: and damn
09:13 AM Emil: there's actually one way better thing that the itead ones have
09:14 AM Emil: there's no connecting fingers
09:14 AM Emil: the edges are all smooth
09:15 AM bss36504: Are the connecting fingers that inconvenient for you? A few passes on some 220 grit sandpaper and you're set
09:15 AM Emil: They are
09:15 AM bss36504: :P
09:15 AM Emil: of course it's only a small amount of work
09:15 AM Emil: but annoying work all the same
09:17 AM bss36504: I guess If I were doing production I would care but for three small boards, it's like 30 seconds of work
09:19 AM Emil: 2017-06-14 16:48:16 +0300 < Elepaja> [Emil Fihlman] <photo> https://elepaja.fi/telegram/media/AgADBAAD0KgxG1NVCFLDpOOdU657N5z_nBkABLXzhVH98zSXkJsAAgI.jpg
09:19 AM Emil: 2017-06-14 16:48:17 +0300 < Elepaja> [Emil Fihlman] <photo> https://elepaja.fi/telegram/media/AgADBAAD0agxG1NVCFIkXAyzHe_mLRc-uxkABDWdumYnFK8oafQBAAEC.jpg
09:19 AM Emil: 2017-06-14 16:48:17 +0300 < Elepaja> [Emil Fihlman] <photo> https://elepaja.fi/telegram/media/AgADBAAD0qgxG1NVCFIh72eIMyRVWL9mqRkABP3QerY1nd2YmnICAAEC.jpg
09:20 AM bss36504: snazzy
09:20 AM mark4: so is the LPM opcode on an xmega vastly different from LPM on other AVRs?
09:20 AM LeoNerd: xmega generally is quite different from ATtiny/ATmega
09:21 AM mark4: my memory says you just set an index register to the address and lpm from that address but the appnote im reading says that to read flash you store the address in an SFR and execute the LPM and then read the data out of another SFR
09:22 AM LeoNerd: The instruction set summary I have for XMEGA family A, says that LPM takes address from (Z) and loads into R0 or optionally some other register
09:22 AM mark4: i see only "avr instruction set" pdf, not a "xmega instruction set" pdf which would be absolutely required if the lpm opcode behaved differently
09:23 AM mark4: leonerd im reading AVR1316 and its saying something different
09:23 AM LeoNerd: I'm just looking at the XMEGA family A overview
09:23 AM mark4: or.. implying
09:23 AM mark4: aha. you have to do a NO_OPERATION maybe?
09:24 AM mark4: yea
09:25 AM mark4: i understand now. there are numerous LPM based commands. like READ_USER_SIG_ROW etc
09:25 AM mark4: the LPM opcode is the same same
09:25 AM mark4: ty
09:27 AM LeoNerd: Oh, yes.. assemblers are... "helpful" like that
09:42 AM NoHitWonder^: nice pcbs Emil
09:44 AM LeoNerd: Hrm.. odd... The XMEGA A family overview document mentions in the PORT section that digital peripheral pin functions can be remapped
09:44 AM LeoNerd: This page is the only page on which the word "remap" appears. It doesn't seem to say anywhere *how* they can be remapped
09:50 AM polprog: Emil: ive seen that board before :P
09:52 AM mark4: yes virtual ports allow you to move any non bit addressable port into the bit addressable region of the io space
09:53 AM LeoNerd: mark4: not talking about that bit
09:53 AM LeoNerd: "Remapping of digital peripheral pin functions
09:53 AM LeoNerd: – Selectable USART, SPI, and timer/counter input/output pin locations"
09:54 AM LeoNerd: ^-- that bit. The only other occurance of the word "remap" in the entire data sheet is: Other digital peripherals, such as
09:54 AM LeoNerd: USART, SPI, and timer/counters, can be remapped to selectable pin locations in order to optimize pin-out versus
09:54 AM LeoNerd: application needs.
09:54 AM mark4: oh yea with the USART you can swap the IN and OUT pins
09:54 AM mark4: THATs what you mean :)
09:55 AM * LeoNerd was hoping for something more akin to PIC24's IO crossbar
09:55 AM mark4: ugh i hate pic24's every time ive developed for them and the code space has exceeded 80, any time i make a change in module X, something in module Y breaks
09:56 AM mark4: ALWAYS happens
09:56 AM mark4: its like tuning a guitar. any time you tune string A up a bit all the other strings detune slightly
09:56 AM Emil: polprog: prolly ;)
09:56 AM Emil: polprog: actually I don't know if you have
09:56 AM Emil: but prolly you have
09:57 AM Emil: NoHitWonder^: they are okay, learnt that I should put all the passives on one side
09:57 AM Emil: I have to hot air the other side even though I can reflow the other T.T
09:57 AM polprog: Emil: i like the difference between green and violet mask
09:57 AM Emil: polprog: the green mask is super clear, I kind of like it
09:57 AM polprog: yeah. green ones are the prettiest ones
09:58 AM Emil: colour wise I don't personally prefer them the best
09:58 AM Emil: but they are the cheapest
09:58 AM polprog: which colors do you prefer then>
09:58 AM polprog: ?
09:58 AM Emil: black and blue
09:59 AM polprog: blue would definitely make your board stand out
09:59 AM Emil: black is horrible for checking for errors
09:59 AM Emil: Oh
09:59 AM Emil: Matte black would be absolutely kawaii
09:59 AM polprog: black wnd white is nice if you wanna hide the traces, and you dont really see the chips on black well, so that's also nice if it shouldnt be exposed
10:00 AM polprog: matte any color would be nice
10:00 AM mark4: im not a fan of any microchip pic devices, the skid bucket makes me want to puke
10:00 AM LeoNerd: They're generally a mixed bag, but the IO crossbar is one feature I'd really like to see on some AVR chips
10:01 AM LeoNerd: A few of the latest ATtiny chips are inching towards it, but still some way to go yet
10:01 AM Lambda_Aurigae: skid bucket?
10:01 AM mark4: yes. the only reliable way to put a breakpoint on any opcode is to either put two nops in front of the opcode you want to stop on and put the breakpoint on the first nop
10:02 AM mark4: or to put two nops after that opcode and a breakpoint on the opcode BEFORE the one you want to stop on
10:02 AM mark4: you cannot simply put a breakpoint on opcode 1 because you cant guarantee its going to stop on opcode 1
10:02 AM mark4: it could stop on opcode 1, 2 or 3
10:03 AM mark4: so you need to inject nops so you can come to a screeching halt
10:03 AM mark4: they call that the skid bucket
10:03 AM polprog: why is that?
10:04 AM polprog: it definitely looks like some design error for me
10:05 AM mark4: its not an error. its an accepted flaw
10:05 AM mark4: makes the devices cheaper
10:05 AM polprog: huh, interesting
10:05 AM mark4: any opcode already prefetched is going to be excecuted. PERIOD
10:05 AM mark4: well im assuming thats what the cause is
10:05 AM bss36504: I was just about to ask why this is...
10:06 AM bss36504: I guess that makes sense; if it's already in the pipeline you can't really preempt it.
10:06 AM polprog: you could always try to stop the clock
10:06 AM mark4: you can on every other processor ive ever worked with
10:06 AM mark4: no other line of devices requires a skid bucket. pic's ALL do
10:06 AM mark4: every pic.
10:06 AM mark4: avr's dont
10:07 AM polprog: i was thinking about changing avr clocks on the fly like a real cpu does
10:07 AM polprog: but in the datasheet there's a maximum delta f specified
10:08 AM mark4: you can change clocks on the fly
10:08 AM cehteh: polprog: you can change the prescaler and you can change OSCCAL (in slow increments)
10:08 AM bss36504: why not just use an xmega at that point?
10:08 AM cehteh: but you cant save much power from that
10:08 AM mark4: or you can simply select an alternative clock source for sytem clock
10:08 AM mark4: you can flip clock sources.
10:08 AM cehteh: you know the 'run-to-idle' theorem
10:08 AM polprog: hmm, i never thought about osccal
10:09 AM polprog: i dont really know much about cpu theory
10:09 AM mark4: using clock A... turn on clock B. wait for clock B to stablize... select clock B. turn off clock A
10:09 AM mark4: oscal is a tweak
10:09 AM cehteh: at some point you rather pick a higher clock and get something faster done, to extend the ultra low power sleep phases
10:09 AM mark4: and if you have auto calibration enabled tweaking oscal by hand isnt going to work :)
10:10 AM cehteh: mark4: osccal is pretty coars
10:10 AM Emil: osccal is absolutely awesome
10:10 AM Emil: only works for internal rc, though
10:10 AM cehteh: but i have some code which adjust the internal osc clock to some external signal which works well
10:10 AM polprog: what i had in mind was an experiment like this on an avr running some program
10:10 AM polprog: https://www.youtube.com/watch?v=nwJQxD8LLNY
10:10 AM mark4: i usually just set the maximum osc setting i can get away with (usually dependent on baud rate requirements etc)
10:11 AM mark4: having two clock soruces running at the same time and flipping between them isnt that difficult
10:11 AM mark4: but might not be efficient with battery
10:11 AM mark4: if battery is critical using the ultra low osc is the way to go
10:11 AM cehteh: polprog: you may have 2 prescaler settings, one for when busy and one for slow background jobs .. and finally use sleep as much you can
10:12 AM LeoNerd: Yah; bumping the prescaler is simple enough
10:12 AM cehteh: ultra low osc isnt the way to go
10:12 AM mark4: it is if timing is not critical
10:12 AM cehteh: see above run to idle
10:12 AM polprog: what would happen if the clock suddenly stopped?
10:12 AM mark4: youn cant stop your active clock
10:12 AM cehteh: you want the clock as fast enough to get things fast done without wasting power
10:12 AM mark4: hardware wont let you
10:12 AM polprog: say en external clock pulse
10:13 AM polprog: like from a signal generator
10:13 AM mark4: cehteh, faster clocks use more power
10:13 AM cehteh: thats not necessary the highest clock, could be 128khz
10:13 AM LeoNerd: Then the CPU stops
10:13 AM LeoNerd: Some of the later *mega chips have clockfail detection
10:13 AM cehteh: but sleeping save a magnitude more power than running
10:13 AM mark4: you cant stop the CPU by turning off its system clock
10:13 AM cehteh: you can
10:13 AM Emil: avrs have no minimum clock
10:13 AM LeoNerd: But yah overall, I've never found a need to adjust clocks to save power
10:13 AM LeoNerd: That's what SLEEP is for
10:13 AM mark4: cehteh, what if your system needs to use the absolute minimum power even when running
10:13 AM polprog: it's not about saving power
10:14 AM Emil: if you give an external clock
10:14 AM polprog: it's about messing around
10:14 AM Emil: you can clock it at whatever you want
10:14 AM mark4: using a fast clock to get the work done fast might use more power than using a slow clock and just plodding along till done
10:14 AM Emil: and there's no mimimumn delta f
10:14 AM Emil: osccal has a minimumn delta f because if you change it quickly enough it will destabilise
10:14 AM cehteh: mark4: if it needs that then it needs it, no way around, then its good to caclulate the slowest clock to do the job yes
10:14 AM Emil: because the phase shifts are fucking all over the place
10:14 AM Emil: sorry
10:14 AM Emil: maximum dleta f*
10:15 AM Emil: of course
10:15 AM mark4: cehteh, ive seen systems that need to run for years on the factory installed battery. sleep, wake up, do job, sleep. wake up and do job have to concerve as much power as they can
10:15 AM cehteh: yes
10:15 AM LeoNerd: Usually I find the way to do that is to run as fast as possible so it's awake for the shortest time
10:15 AM mark4: so yes: get the job completed fast. but a faster clock rate uses more power to do a job than the slowest clock rate does
10:16 AM LeoNerd: AVR chip power consumption scales sublinearly with respect to clock frequency
10:16 AM cehteh: so you batch these jobs, wake as short as possible, use some approbiate clock
10:16 AM polprog: Emil: http://www.atmel.com/images/doc2466.pdf p293
10:16 AM polprog: delta Tclcl
10:16 AM mark4: LeoNerd, meaning what?
10:16 AM LeoNerd: I.e. running the CPU twice as fast uses *less* than twice the amount of power
10:16 AM cehteh: also wake as least as possible that what i meant with batching
10:16 AM LeoNerd: But if you're running twice as fast and are able to complete the job in half the time, you've now *saved* power
10:16 AM cehteh: instead doing one job every 100ms consider if you can do 10 jobs each second
10:16 AM mark4: so using ultra low osc IS the best way to go if power consumption is critical and its still fast enough to do the job fast enough
10:17 AM mark4: cehteh, or 10 sub jobs
10:17 AM LeoNerd: Not if doing so means you're running out of SLEEP for longer
10:17 AM cehteh: as long it can sleep most of the time thats good
10:17 AM Emil: polprog: only for osccal
10:17 AM cehteh: aka if these jobs are really small
10:17 AM polprog: hmm
10:17 AM polprog: thanks for the info then
10:17 AM Emil: polprog: also, ctrl+f "delta" finds nothing
10:18 AM cehteh: its not the highest clock whats optimal you have to figure out what clock is the best to do your workload
10:18 AM polprog: delta meaning the symbol not phrase
10:18 AM Emil: ah
10:18 AM cehteh: but sleeping is still magnitudes better
10:18 AM cehteh: and lower voltage saves a lot power
10:18 AM polprog: anyway, ill just mess around if i get my hands on some signal gen
10:18 AM Emil: polprog: oh
10:18 AM Emil: it is actually for external clock drive
10:18 AM Emil: interesting
10:19 AM cehteh: so you run at the lowest voltage you can, use about the highest frequency with the lowest power OSC which runs stable on that voltage
10:19 AM mark4: what include file defines the NVM commands?
10:19 AM polprog: gotta go eat something, a lunch
10:19 AM cehteh: you may have other constraints like clocks for serial or ADC
10:19 AM mark4: NO_OPERATION REA_USER_SIG_ROW etce tc
10:19 AM cehteh: but just aiming for the lowest possible clock alone will get you noqhere
10:19 AM mark4: cehteh, yes you cant run 115200 on the ultra low osc for sure :)
10:20 AM mark4: lowest one that still lets you do the job... but thats in a power critical application
10:20 AM cehteh: and some problems mean you cant really use deep sleep
10:20 AM mark4: any other application i go with fastest clock that lets me get the job done
10:20 AM cehteh: because you have some interrupt sources which are not available thne
10:21 AM LeoNerd: Having done actual battery low-power stuff, I can say I did actually find that "fastest clock" was the best option
10:21 AM cehteh: you have to put a lot things in the bowl and rate/consider then
10:21 AM LeoNerd: I have battery discharge graphs to demonstrate
10:21 AM cehteh: yeah
10:21 AM LeoNerd: Yah; these things depend a lot on the task at hand. It's hard to say a general rule. Many factors acting contrary to each other
10:21 AM cehteh: thats likely for most problems
10:21 AM LeoNerd: It depends on the use-case. Only way to know for your problem is to measure it
10:22 AM LeoNerd: I also pulled quite a few other tricks in mine; like using the ADC to read multiple buttons from a single pin, but using the AC to detect if any button was pressed, so Idon't have to poll it
10:22 AM LeoNerd: That was fun :)
10:23 AM cehteh: just saying lowest clock is most likely suboptimal, highest clock for the given voltage might almost always be the best choice and there are reasons/exceptions to these rules
10:23 AM mark4: LeoNerd, you could tell WHICH buttons were pressed by the adc result?
10:23 AM LeoNerd: Yes
10:24 AM LeoNerd: Each button bridged a different resistor onto a divider on the ADC/AC input
10:24 AM cehteh: thats common
10:24 AM LeoNerd: So tuned that with no button, the input was one side of the bandgap, but any combination of buttons put the result on the other side
10:24 AM LeoNerd: So going between none and atleast-one button pressed would flip the AC and I'd know to start polling
10:24 AM mark4: oh. yea thas not the keypad i was thinking of tho lol
10:24 AM cehteh: more often external components waste power
10:24 AM cehteh: leds, voltage dividers, LDO's
10:24 AM LeoNerd: This was zero-power when idle
10:25 AM LeoNerd: No buttons -> no low-side of the divider -> input at VCC
10:25 AM cehteh: needs some careful desin (not such simple things like this buttons, but generally)
10:25 AM LeoNerd: Powered directly by a LiFePO4 cell, so no regulator :)
10:26 AM cehteh: thats always a good idea
10:27 AM LeoNerd: Overall I'm rather happy with that design. After attaching the radio and sensor modules, I only had 4 spare pins on the ATtiny for all my buttons and LEDs. Charliplex up to 6 LEDs on three pins, and 3 buttons on the remaining analog pin.
10:29 AM LeoNerd: I have a related task sometime to try to share SPI (for radio) and I²C (for some sensors) on the ATtiny's USI module
10:29 AM LeoNerd: I think it can be done with no other IO pins and a single externally-mounted 1G66 analog switch
10:31 AM cehteh: sidenote: internal osc is not the most powersaving
11:43 AM szwetsloot: Hey guys, would this be the right location to ask about the ATSAM4N16C from Atmel?
01:12 PM Emil: szwetsloot: no and yes
01:14 PM Lambda_Aurigae: you can ask. there MIGHT be someone here who has used that chip.
01:14 PM Lambda_Aurigae: used to be just an atmel AVR channel but we are branching out and we answer questions on anything from apple repair to zebra hunting so,,go for it.
01:15 PM cehteh: how do you hunt zebras with AVR's?
01:15 PM Emil: cehteh: easy
01:16 PM Emil: cehteh: you become frustrated enough at all the people on this channel that it sets up a butterfly chain reaction
01:16 PM LeoNerd: I imagine some sort of pattern discriminator on a camera
01:16 PM * cehteh just thought about an attiny in a railgun
01:16 PM Emil: Alternatively C-x M-c M-zebra
01:18 PM cehteh: C-x M-c is undefined
01:18 PM cehteh: :/
01:36 PM Lambda_Aurigae: you have to use dip package chips. Attach them to the end of blowgun darts and tip the pins with poison.
01:36 PM polprog: stepping on an upside down dip is 100x worse than stepping on a lego
01:36 PM polprog: +5% lead poisoning too
01:37 PM Lambda_Aurigae: only thing more painful is jamming a piece of copper wire under your fingernail.
01:37 PM Lambda_Aurigae: I have a whole bag of short copper wire bits marked, "For torture purposes only."
01:40 PM polprog: under your fingernail or one of those whiskers inside your finger :(
01:41 PM polprog: btw, should we add zebra hunting to the topic?
01:41 PM Lambda_Aurigae: of course.
01:41 PM Lambda_Aurigae: from AVR to Zebra!
01:41 PM cehteh: first step: zebra detector
01:42 PM cehteh: with avr and laser scanner
01:42 PM LeoNerd: Step 1: place a known zebra in Cape Town
01:47 PM polprog: random number generaton with barcode scanners and zebras
01:50 PM Lambda_Aurigae: could use barcode scanner to do unique identifiers for each zebra as their stripe patterns are unique like fingerprints.
01:51 PM cehteh: thats actually done by some zoologists
01:52 PM polprog: i hope not with barcode guns
01:52 PM Lambda_Aurigae: was just handed my next project.
01:52 PM Lambda_Aurigae: a weeding bot.
01:52 PM polprog: weeding or wedding
01:52 PM Lambda_Aurigae: wifey wants one but doesn't want to pay 250 dollars for it.
01:52 PM Lambda_Aurigae: weeding
01:52 PM Lambda_Aurigae: as in, keeping weeds down in the garden.
01:53 PM polprog: proof of concept: https://www.youtube.com/watch?v=6OnOvrTugiA
01:54 PM polprog: that will definitely keep weeds down
01:55 PM Lambda_Aurigae: does bad things to the taste of the tomatoes and cukes too.
01:57 PM cehteh: oh wrong bot .. i thought about a weed-growing bot :D
01:57 PM cehteh: would be vastly simpler, just regulate lights and water, maybe harvest, and keep the police out
01:59 PM polprog: every time i see those smart plant growers made out of an arduino/ti launchpad and a plastic box, i have a feeling that somewhere there is another version that grows weed...
02:00 PM cehteh: for sure
02:00 PM polprog: i just dont believe someone would make something like that just to grow lettuce
02:00 PM polprog: not worth it
02:10 PM pwillard: for a geek factor... it would be
02:15 PM Lambda_Aurigae: the ones I am seeing(one from the guy who designed the roomba) they have a little string trimmer underneath and just go around your garden wacking weeds...anything taller than a certain height gets to keep growing but if it is under 2 inches or so it gets wacked.
02:23 PM pwillard: i smell disaster with that approach
02:24 PM polprog: i guess you will have a month of totally wrecked garden while you fine tune the blades/whatever
03:05 PM day__ is now known as daey
03:13 PM day__ is now known as daey
05:44 PM kre10s: ok. So I have this telit modem connected to the pi. I don't have a sim card in the card slot, but I should still be able to communicate with the modem. Problem is I don't know if the rx/tx cts/rts lines are hooked up in the correct order.
05:46 PM kre10s: It should be fine, since they are connected over lvl shifters, so they won't burn any ios. Is there a way to read/write directly with the uart? just to test/see if the modem is alive?
05:46 PM Lambda_Aurigae: probably.
05:46 PM Lambda_Aurigae: what com port is the uart on?
05:47 PM Lambda_Aurigae: you might need to use setserial or similar to get it setup.
06:04 PM kre10s: setserial has so many options... How do i know what to set? I have the uart on gpio14&15
06:04 PM Lambda_Aurigae: sorry...I know nothing about rPI uart.
06:04 PM Lambda_Aurigae: and I didn't ask what pins.
06:04 PM Lambda_Aurigae: I asked what com port.
06:05 PM Lambda_Aurigae: in other words, on linux, what /dev/* is it?
06:05 PM kre10s: np. I should probably not be asking this on #avr anyway.
06:06 PM kre10s: I don't yet know what device it is.
06:13 PM Lambda_Aurigae: well, I would recommend learning to crawl before trying to run.
06:14 PM Lambda_Aurigae: learn your OS, your interface to the serial device known as a uart, and how to use said serial interface.
06:14 PM Lambda_Aurigae: google is your friend.
06:14 PM Lambda_Aurigae: lots of information out there on rPI serial port/uart interfacing.
06:50 PM Emil: kre10s: you want to use the raspberry pi serial port?
06:50 PM Emil: It is as easy as disabling the os on that with raspi-config
06:50 PM Emil: and then it appears like a normal serial port
06:56 PM kre10s: right I used the config tool to enable the uart hardware but disable the console. but I only get an /dev/ttyAMA0 not the /dev/ttyS0
06:57 PM kre10s: I want to use the second port for something other than a console.
11:55 PM day__ is now known as daey