Back
[03:03:42] <Jordan_U> Is an external crystal required for accurate time keeping with an Atmega64A1? Specifically, I'd like to generate an interrupt at least 10 times per second (100 ideally) that I can use to measure durations (stop watch rather than a clock) with no more than 10 seconds of drift from real time per hour.
[03:10:23] <malinus> Jordan_U, look for the ppm+/- of the internal oscillator
[03:13:27] <malinus> Jordan_U, iirc it's 10% accuracy for non-calibrated, and down to 1% for callibrated. So a callibrated internal oscillator will have a drift larger than 10sec/hour, yes.
[03:16:33] <Jordan_U> malinus: I can't find "ppm" anywhere in the datasheet:
http://www.atmel.com/Images/Atmel-7647-Automotive-Microcontrollers-ATmega16M1-32M1-64M1-32C1-64C1_datasheet.pdf
[03:27:20] <Thrashbarg> http://www.maxwell.com/products/microelectronics/docs/hsn1000_rev3.pdf <-- I suppose they have radiation hardened power regulators too?
[03:29:28] <malinus> Jordan_U, I remembered correctly. Callibrated, it can be 1% accurate @ 8MHz (page 29).
[03:32:45] <Jordan_U> malinus: Thanks.
[03:36:23] <Jordan_U> malinus: Does that 1% incude variation due to heat and voltage?
[03:36:58] <malinus> no :D
[03:37:32] <malinus> so yeah, don't use the iternal osci. for anything that requires precise timing
[03:37:51] <malinus> especially not precise timing over time
[03:39:46] <Jordan_U> malinus: Not the answer I was hoping to hear, but thanks again for giving it :)
[03:40:28] <malinus> Jordan_U, with external crystal, you get down to ~4sec/day drift
[03:42:13] <Jordan_U> malinus: And does that 4 sec/day include variation due to temperature and voltage?
[03:42:56] <malinus> I don't think so, but it shouldn't be that big
[03:43:05] <malinus> again, would need to check in the datasheet
[04:50:40] <mog> i have a 32khz crystal connected to my xtal pins, is there anyway to output that frequency out another set of pins to test the oscillator?
[04:51:00] <mog> if i use the interupts even with scaler disabled i still end up with 64 hertz
[05:09:09] <malinus> mog, something is definetly wrong
[09:05:47] <rue_more> mog, thats a divider of 512
[09:05:55] <rue_more> because you have a 32.768khz crystal
[09:06:13] <rue_more> opposed to 32.000khz
[10:39:46] <mog> i want to measure the frequency of my 32khz crystal for calibration.
[10:40:07] <mog> from what ive read online best way to do this is to output the crystal frequncy out another pin
[10:40:54] <mog> i cant get the full speed out the pin? but i should be able to i would think as 32khz is slow compared to the 8mhz internal clock its running at?
[10:41:23] <malinus> mog, yes
[10:41:51] <malinus> mog, how are you measuring?
[10:42:17] <mog> i was going to connect a frequency counter to the pin im bouncing the output to
[10:42:42] <mog> i tried just doing toggle in the timer2 interupt and setting the timer count to 254
[10:42:55] <mog> but what i get is nothing close to 32khz
[10:44:46] <malinus> mog, what did you get?
[10:45:21] <mog> 64 hz, 128 hz, and 4khz . with different things ive tried
[10:50:27] <malinus> mog, I'm not sure you understand how the timer works. If it's set to 254, it will tick 254 times before triggering the interrupt/pin.
[10:53:05] <mog> tcnt2 is the counter timer2 uses to overflow at 255
[10:53:15] <mog> is how i understand it
[10:53:31] <mog> so setting at 255 will cause the timer interupt to overflow everytime i would think
[10:57:00] <malinus> mog, no, it would count 256 times
[10:58:34] <malinus> 0-1-2-3-4-5-6-[...]->255->OVERFLOW->0
[10:59:13] <mog> umm reading datasheet seems otherwise?
[11:00:16] <mog> i could be wrong, are you saying i should set it to 1 then?
[11:06:38] <aandrew> it interrupts when it overflows
[11:06:45] <aandrew> so the 255 -> 0 transition
[11:07:27] <mog> thats what i thought / expected
[11:07:42] <mog> so setting it at 255 means it should overflow every time the crystal ticks
[11:09:09] <malinus> mog, I got disconnected. Did you respond
[11:09:10] <malinus> ?
[11:09:48] <mog> malinus, so you are saying i should set it to 1 then?
[11:09:59] <mog> my goal is to have it trigger every time it ticks
[11:10:54] <malinus> mog, hmm. or 0. I think 0. Try both.
[11:11:30] <malinus> read the datasheet. I always get confused with the number of counts vs value ;P
[11:36:10] <malinus> mog, have you figured it out? Also don't use the overflow, that will obviously add additional delay, make it switch the pin on match
[11:36:59] <mog> wwhats best way to do it then?
[11:37:13] <mog> im not at my board this second cant test till after lunch
[11:38:49] <N1njaneer> Hi all!
[11:39:12] <malinus> hi!
[11:39:35] <N1njaneer> Sushi and AVR code this afternoon - a good combination!
[11:41:03] <malinus> mog, ask N1njaneer he will explain it much better. but you want to use it in the switch on compare match mode
[11:41:05] <malinus> or whatever it's called
[11:41:47] <malinus> I gotta go, I'm sure N1njaneer will take it from here ;)
[11:41:55] <N1njaneer> Mog: Which micro, and what's the goal?
[11:42:14] * N1njaneer makes note to self -- when starting to write AVR articles, do a bunch on setting up timers and stuff :D
[11:42:32] <mog> im trying to measure my 32khz crystal to get better time from it
[11:42:36] <mog> its a 328p
[11:42:50] <mog> i want to toggle a pin every tick of the crystal
[11:42:58] <mog> so i can hook a frequency counter to that pin
[11:42:59] <N1njaneer> Are you doing a low-power application?
[11:43:07] <mog> yes
[11:43:33] <mog> application works great im trying to minimize drift now
[11:43:48] <N1njaneer> What's the ppm spec on your crystal?
[11:43:54] <mog> 10ppm
[11:44:14] <mog> im probably getting closer to 30-40 ppm now
[11:44:36] <N1njaneer> So seeing more drift than expecting?
[11:44:58] <mog> yes so i want to verify the frequency
[11:45:46] <N1njaneer> Unless the clock crystal is out of spec (unlikely) the only other hardware item I would verify is to ensure your resonance caps for the crystal are the correct value as per the spec of the crystal.
[11:46:29] <N1njaneer> Other than that, can you describe what you are doing on the software side and how you are counting/dividing time down for your measurement? It's likely you have an off-by-one error if you are seeing drift.
[11:46:59] <mog> im just using the scaler to give me a 1 second interval and doing the counter ++
[11:47:03] <mog> nothing complicated
[11:47:27] <N1njaneer> How long until you are seeing drift occur?
[11:47:54] <mog> its hard to say
[11:48:04] <mog> within a day i noticed greater than normal drift
[11:48:30] <N1njaneer> A +/-10ppm crystal will give you at most a +/- 0.864 second drift.
[11:48:44] <mog> ya in a day im seeing more than a second of drift
[11:49:02] <N1njaneer> What counter are you using for the divide-down?
[11:49:28] <mog> TCCR2B = (1<<CS22)|(1<<CS20); //Set CLK/128 or overflow interrupt every 1s
[11:49:48] <mog> you cant really mess that up
[11:50:00] <mog> any other divider would have much different drift
[11:50:41] <N1njaneer> Also, can you tell me what the overall application is that you're doing? Sometimes another strategy to consider is using a dedicated RTC crystal, let that keep time for the Atmel, and then sleep the Atmel until it needs to wake up and do something. You can get exceedingly low power consumption with that method, and not lose time/date when the Atmel resets :)
[11:51:06] <mog> that is what im doing
[11:51:21] <mog> brb
[11:51:41] <N1njaneer> Sorry, RTC clock and crystal
[11:55:51] <Kev> maybe it's extremely hot/cold where you're testing the xtal
[11:57:25] <N1njaneer> mog: Are you using WGM mode 0, then? Are you having to reset the count back to zero or is that occuring automatically? I usually always use CTC mode for my applications.
[11:59:20] <N1njaneer> Kev: If it's a +/10ppm crystal, that means that over the operating range of the crystal it shouldn't drift more than that. If it's cheap crystal that's like -10-+70C, a good crystal is more like -55C to +125C
[11:59:41] <N1njaneer> If he's outside of those ranges in his testing, I'd be concerned about his personal health more than the clock running at the wrong speed :D
[12:01:30] <N1njaneer> Hmm okay, in Normal Mode (WGM 0) it doesn't need resetting, so that eliminates the possibility of not accounting for interrupt overhead before resetting the counter and getting compounding error.
[12:03:23] <N1njaneer> mog: Dumb question, but you ARE sure the Atmega is set to EXTLOFXTAL mode for the input clock, yes? And you have verified your load capacitance cap values on the quartz are correct?
[12:03:33] <Kev> yeah
[12:04:01] <Kev> also try soldering the xtal case to GND, it might or might not help a lot
[12:04:43] <Kev> may or may not *
[12:05:01] <N1njaneer> We've actually run AVR boards immersed in LN2 here to see how they work. They actually function just fine for the most part, and only stopped when the ceramic resonators (at least on the ones we were testing) hit the LN2. Gotta retry this with quartz instead.
[12:05:15] <Kev> whichever is gramatticaly the most corect
[12:05:16] <N1njaneer> And the boards restarted fine as soon as the resonator was back up to a resonable temp :D
[12:05:17] <Kev> whichever is grammatically the most correct *
[12:05:30] <mog> sorry im back. im not 100% sure part is even in error, why i want to measure it
[12:05:50] <mog> no caps
[12:06:00] <N1njaneer> The quartz crystal has no caps??
[12:06:09] <N1njaneer> What's the part #?
[12:06:26] <N1njaneer> Generally you'll need a pair of 18-22pf caps on each leg of the quartz to ground.
[12:06:31] <mog> correct everything ive seen about using 32khz means no cap
[12:06:53] <mog> its an abs07 cant see it this second
[12:07:13] <Kev> xtal = caps
[12:07:40] <Kev> (for the most part)
[12:08:17] <mog> i think the wgm pwm stuff is what i need to do
[12:09:26] <Kev> +/- 20ppm says the datasheet
[12:10:34] <mog> i must have been thinking of tolerance not stability
[12:12:05] <N1njaneer> Mog: Did you run the calculations for external capacitance as per the section of the datasheet regarding low frequency crystal oscillators?
[12:13:02] <N1njaneer> If those caps aren't present there's no guarantee your crystal will resonate at the correct rate, or at all :)
[12:13:43] <mog> like i said
[12:13:44] <mog> it resonates
[12:13:51] <mog> it resonates very close to accurate but a bit fast
[12:13:54] <mog> several of them do that
[12:13:59] <mog> i just want to try and clock the rate
[12:14:40] <mog> i appreciate all the tangential advice though
[12:14:44] <N1njaneer> http://www.atmel.com/images/doc8333.pdf -- please read, especially the top paragraph on page 8 :)
[12:15:36] <N1njaneer> "Ultra low power 32kHz oscillators typically dissipate significantly below 1μW, and the current flowing in the circuit is, therefore,
[12:15:36] <N1njaneer> extremely small. In addition, the crystal frequency is highly dependent on the capacitive load."
[12:15:54] <mog> im aware of this
[12:17:02] <N1njaneer> Is this on a PCB, or a perf-board?
[12:17:39] <Kev> it's an smd xtal so i'm going to guess that it is a pcb
[12:17:54] <N1njaneer> Kev: I've see stranger things :D
[12:18:22] <N1njaneer> Oh well, my overall reccomendation is to work through all of the aspects of that appnote thoroughly in order to double-check what is going on.
[12:18:59] <mog> that is what im trying to do.... im trying to see what the crystal is outputing.
[12:19:01] <mog> its on a pcb
[12:19:53] <N1njaneer> Do you have a precision frequency counter or digital scope?
[12:20:34] <mog> i do have a precision frequency counter and a good scope
[12:22:07] <N1njaneer> What are they measuring the crystal as?
[12:22:37] <mog> i havent tried measuring it directly, as everything ive read says thats invalid way to test it
[12:24:29] <N1njaneer> "everything" being what, if you don't mind me asking? You have to be careful not to introduce stray parasitic capacitance from the test leads, but I've measured AVR quartz crystals many times without difficulties here.
[12:24:56] <mog> ive read thats exactly what happens even with a 10 probe
[12:25:18] <N1njaneer> Use a near-field probe, then :)
[12:25:53] <mog> i guess ill do more reading to figure out how to pwm the pin at same rate as crystal
[12:26:48] <N1njaneer> Or write an AVR program that is simply two lines - GPIO on, GPIO off - nothing else. That will run at either a half or quarter-frequency of the system clock depending on what instruction you are using to toggle the GPIO pins - some are 1 cycle, some are 2 :)
[12:27:14] <N1njaneer> Then probe the TTL output level and you'll be buffered from the clock.
[12:33:14] <N1njaneer> Dunno, that's about all I've got. This sounds like a hardware problem to me.
[12:34:26] <Kev> at some point if he doesnt want to try the actual correct design with caps, there's not much we can do
[12:35:41] <N1njaneer> The caps may or may not be required, but there's a need to run the calculations and closely examine the datasheets for calculating load capacitance and ESR to make sure of that. :)
[12:36:06] <N1njaneer> Atmel provides the handy appnotes to assist with issues exactly like this.
[12:37:20] <N1njaneer> With that simple of a software division of the frequency (unless there's something else in software we're not aware of, since we've not seen a complete code-dump), if the clock is drifting that much the clock isn't running at the correct frequency, and that would be a hardware thing. :)
[12:40:06] <N1njaneer> In any case, hopefully he solves it.
[12:40:21] <N1njaneer> mog: Definately let us know what you find out!
[13:04:37] <umquant> anyone have experience implementing websockets on an avr?
[13:18:39] <umquant> One other non related question. Know of any good online code resources that show good use of C++/OOP on an embedded platform
[13:21:31] <jadew> you want to stay away from OOP on AVRs
[13:21:36] <jadew> unless you really know what you're doing
[13:22:06] <jadew> also, wth is a websocket?
[13:22:39] <umquant> jadew, It is an interface that is used in COMET (basically long polling)
[13:23:14] <jadew> blocking HTTP call?
[13:23:29] <umquant> jadew, also, the reason I asked about cpp is many companies want applicants to have experience with embedded C++ I am really just trying to get my feet wet with the methodology and learn some of the common embedded C++ situations
[13:23:40] <umquant> http://en.wikipedia.org/wiki/WebSocket
[13:24:07] <umquant> For fun I wanted to try to implement it on an AVR
[13:25:02] <jadew> well, programming for these devices is the same as regular programming, except you have tighter limits
[13:25:33] <aandrew> not sure why everyone wants to try to do OO on tiny devices
[13:25:48] <jadew> ok so websockets are blocking HTTP calls
[13:25:50] <umquant> aandrew, I don't nessarily want to, but employers ask about the experience
[13:26:06] <jadew> kinda hard to get that from the wikipedia article
[13:26:12] <jadew> sounds like another web 2.0 bullshit
[13:26:19] <aandrew> umquant: I've been doing this for 20 years. I can count on one hand the number of times any company has asked me for "embedded OO" experience and who was serious about hiring
[13:26:48] <umquant> aandrew, Well that is good to know! I am going off of employment descriptions and openings that are sent to me
[13:26:57] <aandrew> yeah don't trust those :-)
[13:27:08] <umquant> I haven't had an embedded job interview yet since I have a job currently just not doing the super fun stuff :)
[13:27:20] <umquant> haha yea! I know the job descrip for what I do now was WAAAAY off
[13:27:40] <umquant> Just trying to be the best canidate possible. I am coming from a completely different field so I need all the help I can get
[13:27:53] <aandrew> yep no worries
[13:28:09] <aandrew> it'd be good if you had a bit of a portfolio of stuff you've done (esp. on your own)
[13:28:11] <umquant> aandrew, do you work for yourself?
[13:28:28] <aandrew> the HR weenie won't be impressed but your eventual manager and peers may well be
[13:28:30] <aandrew> umquant: yes
[13:28:33] <umquant> aandrew, Yep! One of my big things I am doing while in school / working is working on and documenting my projects
[13:28:44] <umquant> no big embedded projects yet but I working towards some
[13:30:15] <umquant> jadew,
http://tech.wordnik.com/websockets-non-blocking-by-nature/
[13:30:16] <jadew> http://emptybottle.org/bullshit/
[13:30:35] <umquant> I don't know much about the spec, just that it is insanely popular atm
[13:30:43] <jadew> umquant, no, they are blocking in the sense that they keep the HTTP connection open
[13:30:45] <jadew> that's how they work
[13:30:54] <umquant> gtocah
[13:31:17] <jadew> I implemented something like that many years ago, way before this thing became a trend
[13:31:28] <jadew> it's a shitty approach
[13:32:56] <umquant> it seems like it is good for web stuff
[13:33:15] <jadew> I didn't know it has a name now, but I've seen it before and always thought it can be done better
[13:33:22] <jadew> I used something similar for an online chess game
[13:34:17] <jadew> I think it's a waste of resources
[13:35:16] <umquant> sounds like I need to do more research on it
[13:35:52] <jadew> I'm turning into a bitter old man really, the world is turning into a direction I don't like :P
[13:36:41] <N1njaneer> aandrew: I do OOP on AVRs because it allows much cleaner code in complex systems, as well as allows for easy compositing of functionality between dissimilar layers and peripherals.
[13:36:44] <jadew> there's a bunch of people generating "cool trends" and the rest of the world following them, with out questioning or understanding them
[13:36:53] <jadew> like that XHTML garbage
[13:37:30] <jadew> which eventually got trashed
[13:38:41] <aandrew> ninyou have to really know what you're doing to keep the code size down
[13:38:43] <jadew> N1njaneer, there's not enough space to allow for such complexity that you NEED OOP
[13:38:57] <jadew> you can make clean code with out OOP
[13:39:35] <N1njaneer> Entirely depends on what you are doing, and how you do it.
[13:39:44] <jadew> one thing I do, whenever I think I may need OOP is to consider macros
[13:40:07] <umquant> Hi N1njaneer :)
[13:50:45] <t4nk746> hello there!!
[13:51:52] <umquant> hi
[14:00:36] <t4nk746> umquant:im having a little trouble with avr's adc can u help
[14:00:42] <t4nk746> ?
[14:00:59] <umquant> sure I can try. Post your code / problems and others can help too
[14:02:09] <t4nk746> while(1) { ADCSRA |= (1 << ADSC); while((ADCSRA & (1 << ADSC)) != 0); LEDvalue = ADC;
[14:02:23] <umquant> t4nk746, post your code to pastebin.com please
[14:02:57] <t4nk746> ok
[14:03:53] <umquant> Thanks!
[14:27:57] <Jartza> hmmh
[14:28:07] <Jartza> https://www.dropbox.com/s/5dan6djv8ucdxwe/IMG_20140613_213756.jpg
[14:28:08] <Jartza> whee
[16:42:08] <Jartza> some strange stuff
[17:01:41] <Trackk> hello
[17:02:07] <Trackk> is there any way I can make an xmega program another xmega through USART/I2C/SPI interface?
[17:02:43] <Lambda_Aurigae> sure.
[17:03:21] <Lambda_Aurigae> put a bootloader on the device to be programmed...program the other device to send the code to the bootloader via the chosen interface.
[17:44:25] <Trackk> Lambda_Aurigae: any bootloader in particular? I mean it can be done through ISP, PDI, JTAG... i'm confused
[17:44:46] <Lambda_Aurigae> well, the bootloader has to support communication through said port.
[17:45:13] <Lambda_Aurigae> most likely you will need to write one...or modify an existing one to your purposes.
[17:46:23] <Lambda_Aurigae> a bootloader is a small program that lives on the microcontroller that allows you to upload a program through a method other than the standard ISP/PDI/JTAG interface.
[17:48:19] <Trackk> i see
[17:48:30] <Trackk> i read about XBoot, seems like a nice alternative
[17:48:42] <Trackk> thanks Lambda_Aurigae :)
[18:12:18] <Jartza> hmmh
[18:12:37] <Jartza> I guess avr doesn't like negative voltages that much? :(
[18:16:11] <Lambda_Aurigae> Jartza, yeah, not so much.
[18:16:29] <Jartza> 400mV bad? :)
[18:16:49] <Jartza> (momentarily)
[18:19:18] <Lambda_Aurigae> did the magic blue smoke escape?
[18:19:24] <Jartza> no
[18:19:32] <Lambda_Aurigae> and did you check the datasheet?
[18:19:57] <Jartza> not for negative voltane no
[18:20:54] <Lambda_Aurigae> Voltage on any pin except RESET
[18:20:55] <Lambda_Aurigae> with respect to ground ............................... -0.5V to VCC + 0.5V
[18:20:55] <Lambda_Aurigae> Voltage on RESET with respect to ground ......-0.5V to +13.0V
[18:21:01] <Lambda_Aurigae> that's the for the atmega1284p
[18:21:34] <Jartza> oh, ok
[18:21:38] <Jartza> just didn't find that
[18:21:40] <Lambda_Aurigae> it's in the datasheet.
[18:21:48] <Lambda_Aurigae> electrical characteristics.
[18:22:49] <Jartza> doh
[18:22:51] <Jartza> yeah
[18:22:55] <Jartza> so it is
[18:22:59] <Jartza> an seems to be same for attiny85
[18:23:03] <Jartza> I was blind (again)
[18:23:14] <Jartza> but then that fits inside the specs
[18:23:18] <Jartza> so I don't worry
[18:24:20] <Jartza> actually fits to specs barely
[18:24:25] <Jartza> some undershooting in the signal
[18:24:41] <Jartza> now it's been running 1 hour and now the minimum seems to be -480mV
[18:24:42] <Lambda_Aurigae> diodes are a wonderful thing.
[18:25:03] <Jartza> sure they are :)
[18:25:15] <Jartza> but everything is working, I was just wondering will I smoke the chip with this
[18:25:32] <Lambda_Aurigae> in my experience the specs are on the safe side.
[18:25:46] <Lambda_Aurigae> about 80% of actual maybe.
[18:25:54] <Jartza> yeah they usually are
[18:26:19] <Jartza> but now that I have the specs and it hasn't even gone to -0.5V yet, I dare to let this test run over night :)
[18:27:07] <Jartza> I've had my first oscope for few days now
[18:27:08] <Lambda_Aurigae> zener diodes are fun too.
[18:27:11] <Jartza> learning how to use it
[18:27:19] <Jartza> (and learning electronics, still)
[18:27:34] <Lambda_Aurigae> I've been learning electronics for 38 years.
[18:27:35] <Jartza> wonderful thing, even this cheap one
[18:27:49] <Jartza> well, I've been learning for like 3.8 months :)
[18:28:33] <Jartza> mostly learning by doing
[18:28:42] <Lambda_Aurigae> same here.
[18:28:46] <Jartza> I can only read theory so much, but after a while I need to see what happens
[18:29:19] <Lambda_Aurigae> I started out by fixing radios as a kid then got into the forrest m. mimms III books, specially Getting Started In Electronics and all the mini-notebooks.
[18:29:32] <Lambda_Aurigae> I had practicals long before any theory.
[18:29:43] <Jartza> yeah
[18:30:34] <Lambda_Aurigae> went through several of those radio shack electronics kits..50-in-1, 100-in-1, 300-in-1...the ones with the springs for contacts to connect wires between components.
[18:30:52] <Lambda_Aurigae> build circuit by the book then modify and see what happens.
[18:30:59] <Jartza> cool
[18:31:17] <Lambda_Aurigae> that was when I was 8 to 10 years old.
[18:31:20] <Jartza> I remember by big brother had some electronics sets when I was a kid
[18:31:36] <Jartza> blue board with those spring-things where you connect stuff
[18:31:57] <N1njaneer> Ahh yes, the old RadioShack specials!
[18:32:03] * N1njaneer heads home - see you all later on!
[18:32:06] <Lambda_Aurigae> they were awesome back in the day.
[18:32:19] <Jartza> yeah, I wasn't allowed to touch those
[18:32:25] <Jartza> as I was "too young"
[18:32:53] <Jartza> big brother is 10 years older than me
[18:33:08] <Jartza> (or, one of them :))
[18:33:09] <Lambda_Aurigae> if you are old enough to read resistor values you are old enough to learn electronics.
[18:33:30] <Jartza> :)
[18:33:35] <Jartza> I guess I wasn't
[18:33:52] <Jartza> but I still remember the blue boards and the components
[18:33:59] <Jartza> they looked magical to little kid
[18:34:15] <Lambda_Aurigae> blue ones and red ones and white ones..
[18:34:52] <Lambda_Aurigae> one of them I had was little square and rectangular plastic boxes so you could arrange them and connect them with these little metal tabs between the springs.
[18:37:55] <Jartza> I never got to play with those
[18:38:10] <Jartza> I guess my bro lost them... or burned the parts, for all I know :D
[18:40:13] <Lambda_Aurigae> the new ones are kinda chincy.
[18:42:20] <Jartza> and just as I was cursing, I got rid of the negative voltage :)
[19:23:47] <Jartza> well
[19:24:05] <Jartza> looks like my signal stays within 480mV and 3.68V
[19:24:42] <Jartza> and as my VCC is 3.3V, it goes a bit high, but still within specs
[20:16:52] <Jartza> well
[20:16:58] <Jartza> from 40mV to 3.34V
[20:17:09] <Jartza> I guess I'll leave it there :)
[20:34:59] <Klapo> is connecting TFT LCD from eg xperia x10i to avr possible?
[21:01:44] <learath> possible? Yes!
[21:01:48] <learath> sane? not so sure.
