Back
[00:13:33] <xrlk> congrats
[00:20:00] <Casper> and the pl2303 seems to be too slow
[00:29:05] <obnauticus> Hello. I have a custom avr32u4 board that I am trying to get working with the Caterina-Leonardo.hex bootloader. After uploading the bootloader over ICSP, I get an error that my programmer is not responding. When I try a different bootloader (LUFA / BootloaderCDC.hex) I am able to communicate and program it via USB. Any reasons why?
[00:30:58] * Casper wonders if bootloaders work fine with usb-serial adapters
[00:31:13] <Casper> pl2303 cause bitbang to be WAY too slow
[00:32:16] <obnauticus> http://pastebin.com/5yGXaNTt console output
[00:41:31] <xrlk> if a programmer says it will work with the ATMEGA328P, will it also work for the ATMEGA328
[00:42:48] <Casper> xrlk: who are you talking to?
[00:43:04] <xrlk> any1
[00:44:38] <Casper> oh you didn't put the ? I read it as an affirmation..
[00:44:41] <Casper> and yes it will
[00:44:54] <Casper> but the software may need to be told that you have a different part...
[00:45:11] <Casper> the P and non P version use a different signature
[00:45:21] <Casper> but afaik, they can use the same binary
[00:45:47] <xrlk> ok
[00:45:52] <xrlk> thanks bro
[00:46:15] <Casper> in fact, the same programmer will work on most avr
[01:02:18] <jadew> Casper, maybe they're already programmed and are expecting an external clock?
[01:02:55] <Casper> jadew: no...
[01:03:04] <Casper> the pl2303 seems to be the issue
[01:03:14] <Casper> it's not an hardware serial port
[01:03:21] <Casper> beside, it sit on a hub too
[01:03:50] <Casper> heck... it's actually 2x 6ft extension with a 10ft a to b, going into those port extender thing for laptop...
[01:04:22] <Casper> that have audio, network, serial, parallel, few usb and ps2
[01:11:02] <Casper> jadew: I just need to flash a serial bootloader and I guess it will be fine after...
[01:11:16] <Casper> but for now, time to close the stuff... and go to bed
[01:12:37] <jadew> night
[01:14:16] <Casper> and... the internal RC of that 328p... with the rounding error... cause the 9600 baud serial to be really.... 9000
[01:21:15] <xrlk> Casper, do you know a good site budgetwise for purchasing parts like header pins and small pcbs
[01:21:22] <xrlk> I was looking @ sparkfun
[01:56:33] <dunz0r> xrlk: Ebay :)
[01:56:50] <xrlk> ok
[04:50:02] <mux_> hey abcminiuser
[04:50:19] <abcminiuser> Ahoyhoy
[04:50:25] <abcminiuser> On strong cold meds, so be gentle
[04:50:34] <mux_> don't worry, no hard questions today
[04:50:37] <mux_> just a dreaded question
[04:51:05] <mux_> the lufa virtual serial driver isn't signed, so it's an inconvenience to install under windows 8
[04:51:42] <mux_> as far as I understand, this should be entirely fixed with a code signing certificate, which costs something like a couple hundred dollars, right?
[04:53:47] <abcminiuser> Yeah $90 if you go for a cheap one
[04:53:57] <abcminiuser> (per year, non extended verification)
[04:54:27] <mux_> alright, and what does a 'permanent' one cost? is there such a thing? will it work indefinitely for all identical pid/vid LUFA devices?
[04:55:25] <mux_> what I'm getting at mostly is that it's a pain and I'd like to buy one, preferably one that you can give away with LUFA CDC-ACM to the community so that nobody else needs to deal with it anymore
[04:58:31] <abcminiuser> I'm actually not sure you can buy one - pretty sure you "rent" one, as the issuer is then responsible for validating and revoking it
[04:58:41] <abcminiuser> $99/year for a basic one is the cheapest I saw
[04:58:53] <mux_> so it's still a pain then
[04:59:32] <abcminiuser> Whole thing grinds my gears since it offers no protection (malware guys would just keep buying them as they get revoked) and for CDC-ACM the INF just says to use Microsoft's own *(&^ing driver
[04:59:43] <mux_> I know
[05:00:02] <mux_> I'm sure that if I contact microsoft directly they can offer some kind of deal
[05:00:20] <abcminiuser> See if you can, I'd be very interested if there's a non-terrible solution
[05:00:21] <mux_> because I only see advantages for them if people are able to just build virtual serial devices
[05:01:03] <mux_> alright, where to start....
[05:02:34] <mux_> hah, this is rich, their 'contact us' page is a 404
[05:07:06] <mux_> ugh
[05:07:25] <mux_> so, eh, immediate answer is no
[05:08:19] <mux_> next question: as a workaround for the coming months that it's probably going to take to get a certificate going, is it possible to be an impostor and use the Atmel ASF vid/pid?
[05:09:35] <abcminiuser> Sure just change the VID and PID to match their signed one
[05:09:46] <abcminiuser> (don't you feel secure already?)
[05:10:43] <mux_> yeah I know it's ridiculous, but they don't really make it easy to do it the 'proper' way
[05:11:39] <mux_> as far as I can see the simple act of signing a driver for very malleable microcontroller requires me to register as a company with a CA, all of the top 5 trusted ones are in the US, which requires a ton of paperwork to go through
[05:11:57] <mux_> then I need to do all kinds of crazy shit
[05:12:07] <mux_> with eventually 65 bytes being added to the inf
[05:13:21] <abcminiuser> TADA
[05:13:24] <mux_> I'm sure I can convince somebody at microsoft that it's way more useful to the hacker/maker community just to have a 'generic CDC-ACM device' vid/pid like vid_FFFF is used right now
[05:14:09] <mux_> anyway, I'm willing to spend some time and effort on this to spare other people the trouble
[05:14:27] <mux_> because windows 9/10/whatever it's called also requires signing
[05:14:53] <mux_> maybe I'll just post to the forums for now
[05:15:02] <mux_> if I can do that with a fake account
[05:34:07] <Valen> abcminiuser: did you see those GE led bulbs?
[05:34:56] <abcminiuser> Yeah we had one in the office today
[05:35:03] <abcminiuser> Three hours in, we still can't control it :S
[05:35:09] <Valen> lol
[05:35:29] <Valen> in a legit fashion with their app or trying to break in? ;->
[05:35:36] <abcminiuser> The Wink hub was angry that day my friend, like {something from Seindfeld I forget}
[05:35:48] <abcminiuser> Literally "let's set it up and take a look
[05:35:57] <abcminiuser> "..."it's still flashing pink, is pink good?"
[05:36:11] <Valen> they 240v or still 110 only?
[05:36:25] <abcminiuser> Ah "was angry that day, my friends, like an old man trying to send back soup"
[05:36:35] <Valen> lol
[05:36:37] <abcminiuser> 110V for the one we had at least
[05:36:51] <abcminiuser> Granted we have a lot of wireless at our work, but hey, our stuff copes (mostly)
[05:36:51] <Valen> why would you make a 110v anything these days, i mean really
[05:37:21] <Valen> i mean I bet step 1 in their thing is bridge rectifier then stepdown converter
[05:37:36] <abcminiuser> No space I guess
[05:37:49] <abcminiuser> A19 is pretty teeny tiny
[05:37:50] <Valen> probably saves them .001 cents on the stepdown
[05:42:59] <_abc_> Hi guys. Are there figures somewhere as to atmega avr interrupt latency, effective, when compiled with avr-gcc? I mean, do I have to read the listing files and figure it out (tedius), or has someone written it up a bit already? Thanks.
[05:43:35] <twnqx> the latency itself is low, but the extra code generated by register saving is not
[05:43:41] <abcminiuser> 7 cycles HW
[05:43:54] <twnqx> but it seems to me that how many registers gcc actually saves depends on later usage
[05:44:05] <abcminiuser> But GCC will add more to that, register prologue/epilogue and potential multiple-cycle jumpts
[05:44:10] <abcminiuser> *jumps
[06:51:03] <_abc_> Ah. Thanks. And I could probably save some cycles using volatile global regs instead of locals so less saving is needed?
[06:51:46] <_abc_> I have to add an ISR to a project which already uses two ISRs in a TMR1 context and I need to figure out how fast that needs to be to not nuke the TMR2 based timing
[06:51:57] <_abc_> It will probably be okay but I would like to *know*.
[06:52:27] <_abc_> Also, is there a way to tell gcc to make the listing file nice, in the sense that gcc -O3 etc tends to make a mess of source code and asm correspondence?
[06:53:18] <_abc_> Nitpick: should avr-libc docs mention that something like PORTA &= ~ONE_BIT_MASK; is always (?) compiled as cbi 0x...
[06:53:31] <_abc_> And is thus interrupt safe without cli/sei guarding it?
[06:53:39] <_abc_> Or am I wrong about that too.
[06:54:07] <Lambda_Aurigae> http://www.motorauthority.com/news/1094786_production-ready-aeromobil-flying-car-to-debut-this-month-video
[06:54:36] <_abc_> I looked at the generated code .l file (tediously, see above about -O3 and code reordering), and it seems to use only sbi/cbi if only one bit is set/reset at a time. Does it use an atomic xor or and when more bits are affected?
[07:11:03] <_abc_> Iow, are io port modify instructions usually atomic?
[07:11:22] <_abc_> (when that is possible, iow or, and, xor, not at byte level)?
[07:13:24] <LeoNerd> Unless it's an Xmega, there are no atomic multi-bit RMW insructions
[07:15:19] <_abc_> Ah you can't directly or xor etc with an IO right?
[07:15:29] <_abc_> Okay, noted.
[07:15:49] <_abc_> cli/ioop/sei should be only as long as the clocks assumed for each right?
[07:16:07] <_abc_> Assuming I guard the io ops with critical or cli/sei
[07:20:23] <LeoNerd> Yah; though I tend to do { uint8_t SREG_save = SREG; cli(); .... SREG = SREG_save; }
[07:20:49] <_abc_> Hmm?
[07:20:52] <_abc_> Why is that?
[07:20:55] <_abc_> @ LeoNerd
[07:22:33] <_abc_> ?
[07:26:19] <_abc_> Ah I see, you do not assume I is set.
[07:26:41] <_abc_> I can assume it is set at all times since it is never cleared in my code after setup (critical for it to run, or there will be smoke).
[07:27:11] <Tom_itx> who's sending smoke signals!
[07:27:30] * _abc_ tries to send them SELDOMLY
[07:27:51] <_abc_> Multiplexed displays do not like to freeze on one digit.
[07:28:08] <_abc_> Okay, I got the point about SREG saving, thanks LeoNerd
[07:28:10] <Tom_itx> don't overdrive em :)
[07:28:11] <_abc_> and all
[07:28:38] <_abc_> Tom_itx: They are not overdriven when *multiplexed. Stop that and they will be OLEDs for a few seconds (Orange LEDs)
[11:30:10] <megal0maniac> twnqx: I think I'm making progress :D
http://i.imgur.com/tOLVaAR.png
[11:30:49] <twnqx> i'd find it interesting to see what happens if you disable learning
[11:30:54] <twnqx> 'cause then it should flood...
[11:32:32] <megal0maniac> Apparently I was supposed to patch the kernel driver for the switch as well, but I did a diff against the source I have and the file I downloaded and it spat out nothing. So all I need to do now is try it :D
[11:33:57] <megal0maniac> twnqx: This is the full switch config
http://i.imgur.com/CCT3wOh.png
[11:34:45] <megal0maniac> Port 0 is WAN (connected to modem, pppoe) and the rest are LAN. If I were to make port 4 my listening port, would I just remove it from the VLAN to isolate it?
[11:35:33] <twnqx> uh
[11:35:45] <twnqx> i suppose it would flood across all ports that are part of the same vlan
[11:35:48] <megal0maniac> Obviously I don't want my "listening device" to be present on the network. Not quite sure how to configure that
[11:37:41] <megal0maniac> I guess I could untag the CPU from any VLAN associated with the switch and make it my listening port. Then the switch becomes dumb and I can monitor on eth0
[11:38:14] <megal0maniac> Heck, I could plug in a flash drive and tcpdump straight to it. Don't even need a PC
[11:50:33] <Getty> someone here a bit experience with usage of FatFS? I am just in the planning phase (and we dont have a testenv to test it yet) but i wanna know how i should think about the optimization of directories for speeding up access
[11:51:02] <Getty> will it be like wise to say if i have a list of items to say every 1000 items to make a subdir based on the number? like 1000/ 2000/ subdirs?
[11:51:30] <Getty> or is that in the end not that relevant as its all anyway "guessing" in entries around to find the right one?
[12:09:08] <Casper> how well does an attiny work as an usb programmer?
[12:14:18] <LeoNerd> Given you can buy those things for like, $5, I doubt many people have bothered making one
[12:15:58] <Jartza> or for $2
[12:16:10] <Jartza> I just bought 10 avr programmers made out of atmega8
[12:16:24] <Jartza> they did cost 17€ total with postages to finland
[12:16:42] <Jartza> seem to work just fine
[12:16:55] <Jartza> although to be fair, I also soldered one programmer myself out of attiny85 :D
[12:17:25] <Casper> I have the parts at home
[12:17:45] <Casper> and the tiny85 won't do the job I tought it would do
[12:17:50] <Casper> so I could use them
[12:18:13] <LeoNerd> The tiny85 has too few pins for this
[12:18:55] <LeoNerd> You need 4 to talk to the ISP port of the target and 2 to talk USB. Hrm. well, I suppose you cooould if you had an HVSP
[12:19:59] <Casper> oh right can't disable reset without hvsp
[12:20:06] <Casper> chicken and egg again
[12:20:33] <Casper> this "new" setup is without a serial port... I mean none hardware
[12:20:44] <Casper> and the pl2303 is sloooow as hell
[12:20:58] <twnqx> like every serial port
[12:21:01] <Casper> I mean 1.4k take like 3 minutes to flash
[12:21:08] <twnqx> though they work perfectly fine at 2mbit/s
[12:26:48] <LeoNerd> I'm working on a simple/cheap/lowcost HVSP burner, incidentally
[13:23:10] <twnqx> LeoNerd: connect car battery to reset, done
[13:28:48] <obnauticus> Where does the arduino store the header that defines the pin alieses for the Arduino Leonardo/m32u4 ?
[13:30:44] <megal0maniac> twnqx: lololol
[13:33:56] <theBear> obnauticus, err, it doesn't, it's umm, in a header file somewhere maybe, i dunno, arduino do lots of stuff in unusually stupid ways
[13:47:33] <kdehl> Which 8-bit AVR has the most SRAM?
[13:48:42] <jadew> kdehl, check the parametric search on atmel's site
[13:48:49] <kdehl> I have the 1284p, which only has 16kB, and I recently purchased the 1280, but even though it's TQFP-100, and more flash it has less RAM, only 8kB
[13:48:58] <kdehl> Oh. Didn't know you could do that.
[13:49:35] <twnqx> just add external ram (for those that support it)
[13:49:54] <twnqx> hm. with bank switching you could make an arduino with 1MB ram
[13:50:03] <twnqx> would probably be fun
[13:50:14] <twnqx> i mean, sell it to people who have no clue how to use it
[13:50:52] <kdehl> Heh.
[13:50:58] <kdehl> No, I need linear memory.
[13:51:11] <twnqx> not a problem
[13:51:13] <kdehl> Seems like the most you can get internally on an 8-bit controller is 16 kB.
[13:51:14] <twnqx> with some chips :P
[13:51:27] * twnqx has an external 32kB attached
[13:51:35] <kdehl> Does the 1284 support external RAM?
[13:51:42] <kdehl> Too few mins, I would guess.
[13:51:44] <kdehl> *pins
[13:51:52] <twnqx> i have it on an at90can
[13:51:58] <twnqx> the 1280 might
[13:52:10] <twnqx> also it's multiplexed address/data
[13:52:17] <twnqx> 20 pins or so for the interface
[13:52:19] <kdehl> Oh!
[13:52:31] <kdehl> That reduces the pin requirement a lot
[13:54:16] <jadew> kdehl, looks like xmega384c3 and d3 have 32kb of ram
[13:55:34] <kdehl> ah, yes. Heh. I only looked at the first page of my search.
[13:55:36] <kdehl> :)
[13:56:28] <kdehl> I need at least 64 kB. Which will be hard, since the address registers are 16 bits. Or did I make that up myself?
[13:56:51] <kdehl> Even with external RAM.
[13:57:10] <twnqx> at least 64 kb?
[13:57:11] <twnqx> lol
[13:57:23] <twnqx> yes, exactly 64 is possible
[13:57:30] <twnqx> more only with bank switching
[13:57:33] <kdehl> Yeah.
[13:57:34] <kdehl> Hm.
[13:57:45] <kdehl> Hm, hm, hm.
[13:57:52] <kdehl> No.
[13:58:05] <twnqx> yes.
[13:58:17] <kdehl> Yes.
[13:58:27] <twnqx> but you can't put a single 64kB object into ram.
[13:58:59] <kdehl> Yeah. No, I'm stupid. I need 32 kB for my data, and some more for code. So 64 is sufficient.
[13:59:07] <kdehl> I was mixing some things up here.
[13:59:23] <kdehl> Alright then. I need a controller with exernal RAM, it seems.
[14:00:16] <kdehl> Hm. The 90can you use states in the introduction that it supports "external memory storage," something the datasheet of the 1284p does not mention.
[14:00:25] <kdehl> Which means it probably doesn't support external RAM.
[14:00:35] <twnqx> can't you find that in the parametric search as well?
[14:00:54] <kdehl> No. I just checked the datasheets.
[14:01:29] <kdehl> When you use it, you can tell avr-gcc to use it, no?
[14:02:30] <twnqx> yes
[14:02:41] <twnqx> i wuold possibly not even bother
[14:02:55] <kdehl> What do you mean?
[14:02:56] <twnqx> and just use a static pointer for your 32kB object in external ram
[14:03:01] <kdehl> Yeah.
[14:03:03] <kdehl> True.
[14:03:25] <kdehl> Or.
[14:03:26] <kdehl> Hm.
[14:03:28] <twnqx> but yeah, the "generic Makefile" has some linker parameters that move the heap there
[14:03:36] <kdehl> Alright.
[14:03:56] <kdehl> So now I just need to choose controller. The simplest one that support external RAM, I suppose.
[14:04:11] <kdehl> Preferable DIL. But something tells me that ain't going to happen.
[14:04:24] <twnqx> http://hackaday.com/2010/12/16/external-ram-for-an-atmega128/ ?
[14:04:52] <kdehl> Yeah. I saw that. But it uses a surface mounted controller, I believe.
[14:05:11] <twnqx> well.....
[14:05:17] <twnqx> you will likely use surface mounted ram, too
[14:05:33] <kdehl> I have 32 kB SRAM in DIL.
[14:05:36] <twnqx> oh.
[14:05:56] <twnqx> and the 74hc should be uch of a problem either
[14:06:01] <N1njaneer> Or just bit-bang-bus the RAM connections.
[14:06:18] <twnqx> no, he wants it mapped
[14:06:34] <kdehl> Yeah. I need a linear memory buffer.
[14:06:39] <kdehl> It needs to be fast.
[14:06:51] <twnqx> it's not that fast
[14:06:57] <twnqx> at least an extra cyle for multiplexing
[14:07:04] <N1njaneer> kdehl: You might want to look in to the SAM series, then. There are parts that will do exactly what you are looking for a LOT easier and more efficiently.
[14:07:18] <twnqx> in dil? :>
[14:08:03] <kdehl> Heh. Nah, I'm afraid I will have to give up DIL.
[14:08:09] <kdehl> N1njaneer: Interesting. Thank you.
[14:08:38] <kdehl> Wait.
[14:08:53] <kdehl> Actually. I could use an EEPROM. I don't even need it in SRAM.
[14:09:29] <N1njaneer> kdehl: You can also use SPI-bus for external flash/ram
[14:09:36] <kdehl> Too slow.
[14:09:48] <kdehl> One cycle is fine, but to have to do it all serially will be too slow.
[14:12:08] <kdehl> N1njaneer: SAM is ARM?
[14:12:20] <kdehl> That's a 32-bit controller, isn't it?
[14:12:20] <twnqx> yes
[14:12:26] <kdehl> I'm not that modern. :)
[14:12:55] <twnqx> given they cost the same as my at9c0can
[14:13:00] <twnqx> i should have used them
[14:13:16] <kdehl> Heh.
[14:13:45] <kdehl> No, this is something that I want to try out on an 8-bit controller, actually.
[14:16:08] <kdehl> Huh. The 8515 has only 512 bytes of SRAM, but supports up to 64 kB external memory.
[14:16:55] <kdehl> And it's DIL.
[14:19:26] <twnqx> uhhh
[14:19:35] <twnqx> that's a completely different arch though, isn't it
[14:21:23] <N1njaneer> kdehl: Yep, ARM.
[14:24:05] <N1njaneer> I've recently started moving in to Atmel SAM stuff after 12+ years in AVR. A bit of a transition, but not horribly so.
[14:25:48] <kdehl> twnqx: Not 8051.
[14:30:35] <tunixbsd> I'm trying to simulate a rjmp instruction by save the stack in a point and load in another , but it's not working as the expected.
[14:30:39] <tunixbsd> http://pastebin.com/6MCmje9J
[14:31:15] <tunixbsd> it's just for learning purposes
[14:44:44] <megal0maniac> twnqx: Set up a monitor port, it works flawlessly :D And because the router runs Linux and has USB, I can add a flashdrive and run tcpdump on the router itself. Magic!
[14:45:10] <twnqx> i probably misunderstood what you wanted
[14:45:21] <megal0maniac> Me too :P
[14:56:31] <kdehl> Hehe.
[15:03:09] <Shavik> Pad positions of my power supply IC are off by 0.55mm and are as a result, unsolderable. yay, $1000 in pcb's down the drain
[15:03:48] <N1njaneer> D'oh.
[15:03:55] <Shavik> Yea, Pretty bummed
[15:04:02] <N1njaneer> If that's the only thing wrong, consider making a little castellated riser-board adapter.
[15:04:13] <Shavik> Interesting thought
[15:04:15] <Shavik> Thanks
[15:04:26] <Shavik> IFX90121 SMPS
[15:04:39] <Shavik> 0.525mm off I think heh
[15:05:09] <N1njaneer> SSOP?
[15:05:37] <N1njaneer> This looks very fixable if you have some thin adapters made. Not idea, but saves money.
[15:07:43] <Shavik> Thought about just trying it. heh The inside edge of the leg may rest on the outside area of the pad
[15:07:45] <N1njaneer> +ideal
[15:07:51] <Shavik> wonder of the solder would still connect them
[15:08:00] <Shavik> with this being the PSU, i'm hesitant heh
[15:08:05] <Shavik> if*
[15:10:54] <jadew> LeoNerd, are you available to do that test on the bus pirate for me?
[15:12:19] <twnqx> Shavik: i have to ask you a question
[15:12:32] <twnqx> that i ask a friend of mine who constantly has similar problems
[15:12:43] <twnqx> why don't you order single quantity first to verify your design?
[15:13:12] <twnqx> and: don't you do a printout of the pads only first and put the parts on to check this part?
[15:13:52] <Shavik> twnqx, this is a prime example of not always doing that because everyone before it was right and management was in a hurry
[15:13:53] <Shavik> heh
[15:13:53] <N1njaneer> All good techniques. Or build the footprints yourself for each part off the datasheet.
[15:13:57] <Shavik> ultimately my fault
[15:14:01] <Shavik> I did
[15:14:04] <Shavik> That was half the problem
[15:14:10] <Shavik> This was my first project doing that myself
[15:15:19] <Shavik> valuable lesson learned
[15:15:42] <jadew> last board I did I used a tqfp package instead of a lqfp one
[15:16:38] <twnqx> last board i did
[15:16:43] <twnqx> i ordered the wrong chips
[15:16:50] <twnqx> tssop instead of soic.
[15:17:00] <jadew> did they still fit?
[15:17:04] <twnqx> luckily i always do the paper test
[15:17:09] <twnqx> lol, no
[15:17:15] <twnqx> the chip was a quarter of the size :P
[15:17:20] <jadew> ah :P
[15:17:54] <jadew> I didn't do the paper test because I didn't have that chip yet
[15:18:16] <twnqx> yeah, i can always wait the extra few days
[15:18:17] <jadew> so I simply put the paper next to the dev board and thought "ok, looks like it's the right footprint"
[15:18:42] <twnqx> (now i have three of the CAN boards and no idea what to do with them)
[15:18:47] <twnqx> well, i need one
[15:19:10] <kdehl> 10 ns SRAM ought to be safe for a 20 MHz AVR, right?
[15:19:26] <jadew> oh, speaking of which, I received some documentation about can in the mail today
[15:19:34] <jadew> forgot because I was sleeping when that happened
[15:20:15] <twnqx> 10ns is enough for 100mhz
[15:20:15] <twnqx> >_>
[15:20:41] <kdehl> I thought you needed margins?
[15:20:43] <xrlk> my resistors and sram arrived
[15:20:56] <twnqx> sure, but ot that much :D
[15:21:01] <kdehl> Heh. Alright.
[15:21:19] <twnqx> though
[15:21:21] <xrlk> FEDEX won't leave the MCUs at the front door though
[15:21:31] <twnqx> i have a -12 on my 16mhz board
[15:21:37] <xrlk> did they expect anybody to pick it up at 10am on a weekday
[15:22:16] <twnqx> and the worst about this CAN board
[15:22:21] <twnqx> i'd love to redo it with a sam
[15:23:40] <kdehl> Ah, it's 20 ns. Nvm. But that ought to be low enough too, right?
[15:25:46] <kdehl> How come I can find tons of 32 kB SRAMs on ebay, but no 64 kB?
[15:25:59] <kdehl> Doesn't feel like an odd size to me.
[15:26:33] <twnqx> beats me, i order at digikey :P
[15:27:11] <kdehl> I'm stupid. There are many 64 kB too.
[15:28:51] <kdehl> But you said that the address and data bus are multiplexed. How do I then interface with a non-multiplexed SRAM chip?
[15:29:06] <twnqx> like in the link
[15:29:45] <twnqx> http://i2.wp.com/www.scienceprog.com/wp-content/uploads/2010/12/XMEM_schematics.png
[15:30:44] <kdehl> Ah. Thanks!
[15:31:22] <jadew> twnqx, look what I received today:
http://dumb.ro/files/IMG_0195.png :P
[15:31:51] <twnqx> can fd i haven't heard of :p
[15:32:01] <twnqx> that's not even in my book about car bus systems!
[15:32:13] <jadew> flexible data rate
[15:32:32] <twnqx> yeah, i can read that
[15:32:44] <jadew> neat chart
[15:33:03] <twnqx> crc is also new
[15:33:12] <twnqx> neither in 2.0a nor 2.0b
[15:33:41] <jadew> that's good to know
[15:33:52] <LeoNerd> jadew: Hi. Mmmmm. was just swimming, also I have no machine tonight. PM me some instructions if you want and I'll take a look tomorrow
[15:34:29] <twnqx> jadewif you want some of the standards as pdf... tell me
[15:34:30] <jadew> LeoNerd, cool, I'll do that if I can't find someone else with a bus pirate
[15:34:39] <twnqx> some were harder to find on the internets than others.
[15:34:56] <twnqx> same for higher layer protocols
[15:35:00] <jadew> twnqx, I do, I'm interrested in the CAN at the moment
[15:35:09] <jadew> I got Lin down I believe
[15:35:26] <twnqx> i never dived down into that (yet)
[15:35:35] <jadew> it's super simple
[15:36:19] <jadew> if you need any quick pointers let me know
[15:36:41] <twnqx> for car purposes?
[15:37:17] <jadew> yeah, I never tried it on a car tbh, but I wanted to implement it in my project so I bought some transcievers and had a play with it
[15:37:28] <jadew> I also have a chart for LIN if you're interested :)
[15:37:38] <twnqx> i needed it... to get my car through inspection...
[15:37:45] * twnqx won't go into details
[15:41:32] <twnqx> no idea if this works or not behind weird NAT systems
[15:41:44] <twnqx> oh... and with ipv6...
[15:41:46] <twnqx> lol
[15:43:55] <kdehl> Hm. I have a bunch of 74373s already. According to the datasheet of the '573 they are identical but have different pin arrangement.
[15:50:03] <jadew> twnqx, what are you talking about?
[15:50:44] <twnqx> the dcc send request i sent you...
[15:51:07] <jadew> ah, yeah, it didn't get through
[15:51:30] <jadew> sorry I was afk for a bit, my wife had me watch the wedding video, couldn't stay through it all
[15:53:24] <twnqx> oh, romania. cool
[16:00:39] <jadew> twnqx, when I'll get to CAN I may have some questions if you don't mind
[16:01:11] <twnqx> sigh, i fail to upload this mallish 21MB file to mega...
[16:01:11] <jadew> actually I have one right now
[16:01:16] <twnqx> how much can a single website suck
[16:01:34] <jadew> twnqx, I can give you an upload link in a moment
[16:06:31] <jadew> did you get it?
[16:07:21] <twnqx> i am 3% from completing on mega :P
[16:09:09] <jadew> what do you mean?
[16:09:22] <jadew> I'm not sure if I should think about atmega or mega bytes
[16:09:32] <jadew> and 3% as in project progress or upload :P
[16:10:50] <obnauticus> Does anyone here have disconnect problems on the arduino leonardo? I ran a custom board, uploaded the bootloader and have it all working but it will intermittently disconnect randomly
[16:11:11] <jadew> I also got some video tutorial CDs today (related to CAN & all that), they're new but they look like they've been used for a year
[16:11:15] <Tom_itx> i'm way disconnected from it
[16:11:23] <Tom_itx> so far i may never see one
[16:12:09] <jadew> twnqx, thanks, looks like it arrived
[16:14:02] <kdehl> Do you guys have a simple VGA signal chip to recommend?
[16:14:16] <twnqx> intel haswel
[16:14:18] <twnqx> l
[16:14:35] <twnqx> jadew: you said you had a question right now?
[16:14:38] * twnqx wants to head to bed
[16:15:46] <jadew> yes, I'm working on a device that talks various protocols and I want to add support to CAN as well
[16:16:08] <jadew> do you think I'll have to support the higher layers too, or just the plain CAN protocol?
[16:16:39] <jadew> *for CAN
[16:16:49] * jadew is very tired lately, didn't drink enough coke
[16:17:11] <twnqx> no idea what you want to do, i would not implement higher layers unless i have to
[16:17:15] <twnqx> and there are so many...
[16:17:22] <jadew> twnqx, that's what I thought
[16:17:28] <twnqx> KWP2000, UDS
[16:17:39] <jadew> you have the link in the PM
[16:18:28] <jadew> it might fail, given that it's not getting many visitors, but I still want to finish this for myself
[16:18:38] <jadew> it's almost done anyway
[16:32:31] <kdehl> Shit, are there any TQFP AVRs that can handle 5 volt?
[16:32:40] <kdehl> No.
[16:32:43] <kdehl> Sorry, I mean 20 MHz.
[16:32:44] <kdehl> Heh.
[16:32:59] <kdehl> Most of them seem to max out at 16.
[16:35:28] <jadew> kdehl, parametric search
[16:35:33] <jadew> http://www.atmel.com/v2pfresults.aspx#%28actives:!%288238,8394,8362,8282,8431,8300,8358,8392,8378,8445,8236,8449,8474,8248,8264,8447,8256,8254,8286,8462,8429,8458,8466,8400,8302,8278,9999%29,data:%28area:%27%27,category:%2734864%27,pm:!%28%28i:8238,v:!%280,18%29%29,%28i:8394,v:!%280,20%29%29,%28i:8362,v:!%2812,42%29%29,%28i:8282,v:!%284%29%29,%28i:8431,v:!%281,36%29%29,%28i:8300,v:!%281,9%29%29,%28i:8358,v:!%281,68%29%29,%28i:8392,v:!%280,1%29%29,%28i:8378,v:!n%29
[16:36:08] <Tom_itx> those never paste too well
[16:36:27] <kdehl> jadew: Ah, yes.
[16:37:06] <jadew> Tom_itx, did they paste right now?
[16:37:28] <Tom_itx> URL not found Please close this window and try your search again
[16:37:45] <jadew> awesome
[16:37:58] <Tefad_> use a URL shortener.
[16:37:59] <Tom_itx> :)
[16:38:12] <Tom_itx> i don't think that would help
[16:40:58] <jadew> http://tinyurl.com/superbigb00bs
[16:42:30] * kdehl won't click that
[16:42:36] <kdehl> I said NO!
[16:42:42] <kdehl> Not clicking it.
[16:42:47] <kdehl> Nope. Uh-uh.
[17:26:28] <kdehl> Alright. I've settled on ATmega 8515. Does anyone have experiences with those chips?
[17:26:30] <xrlk> I'm spending with a budget of $75 Will I need this
https://www.sparkfun.com/products/12878 If I buy this
https://www.sparkfun.com/products/9325 solder
[17:27:38] <jadew> you need solder paste only if you're planning to do many boards
[17:27:48] <jadew> in which case you also need a stencil
[17:27:48] <xrlk> ok
[17:28:06] <jadew> otherwise just solder, if it's for home use, go for leaded, not lead-free
[17:28:22] <jadew> 0.6mm or 0.5
[17:28:28] <jadew> and liquid flux
[17:28:30] <jadew> and you're set
[17:29:14] <jadew> I have a big bottle of liquid flux which I'm using to refil a nail polish bottle
[17:29:32] <jadew> and when I need flux, I just apply it with the brush
[17:30:02] <jadew> seems like the least wasteful solution so far
[17:34:36] <jadew> anyone with a bus pirate willing to test a few baudrates for me?
[17:35:43] <jadew> bus pirate and a scope :P
[17:36:07] <jadew> unless the bp reports the real baudrate, then we don't need the scope
[17:38:52] <theBear> does it help if i'm an ass pirate ?
[17:38:55] <theBear> matey
[17:39:52] <jadew> not really, this has nothing to do with audio stuff
[17:40:28] <jadew> 10 fps (farts per second)
[17:44:00] <theBear> wtf does ass pirate have to do with audio stuff ?
[17:44:16] <theBear> i was reporting real baudrates when you were in primary school man
[17:44:25] <jadew> I supposed you take over fart based protocols, no?
[17:45:29] <jadew> or do you mean "ass pirate" as in surprise buttsex?
[17:46:01] <theBear> i didn't say fart smuggler, or fart wrangler
[17:46:24] <theBear> and not surprise so much, just ya know, general marauding
[17:46:47] <theBear> getting drunk, talking parrots, that kinda stuff, only with asses involved
[17:49:47] <jadew> theBear, looks like my project won't get funded
[17:50:01] <jadew> and I don't think it's because of the price like you said
[17:50:20] <jadew> it looks like the conversion rate is at about 10% which is huge
[17:50:33] <theBear> i also said i'm the furthest thing from a consumer there is anywhere, ever
[17:50:40] <theBear> what's a conversion rate ?
[17:50:55] <jadew> leads / unique visitors
[17:51:00] <theBear> i also said you should ask in the big channel, but do you listen to me ?
[17:51:10] <theBear> what are leads
[17:51:12] <jadew> as in how many people bought it vs how many unique visiotrs it had
[17:51:35] <jadew> I didn't want to spam
[17:51:42] <theBear> is that a bad thing
[17:52:02] <theBear> 1 in 10 ain't bad in my formerly professional opinion
[17:52:15] <jadew> it's great actually
[17:52:25] <jadew> the problem is it doesn't get enough visitors
[17:53:07] <jadew> there's obviously a market for it, but it's hard to get to it
[17:54:02] <theBear> yeah well, only asshats visit those kickstarter style sites anyway :)
[17:54:36] <Tom_itx> hehe
[17:54:45] <theBear> :)
[17:55:25] <jadew> yeah, the people that actually commited to it didn't come from the site, they came from eevblog (I have a thread there)
[17:55:45] <jadew> 2 from there and 2 from a guy who bought it and tweeted about it
[17:56:27] <jadew> the stuff that seems to get funded a lot is more in the line of funky impossible watches
[17:56:29] <jadew> rings
[17:56:47] <jadew> or *duino stuff
[17:57:08] <theBear> you have a WHAT !~?!?!?! and to think i kinda respected you
[17:57:34] <jadew> what did I say I have?
[17:57:48] <jadew> and you "kinda" respect me? :)
[17:58:27] <jadew> ah, the thread
[17:58:45] <jadew> yeah, there's a crowdsourcing section there
[17:58:58] <jadew> I take it you don't like eevblog?
[18:02:11] * jadew pokes theBear, you awake?
[18:03:11] <Tom_itx> never poke a sleeping bear!
[18:04:08] <jadew> it's alright, he's domesticated
[18:06:49] <theBear> no i ain't, not toilet trained either
[18:06:56] <theBear> and this time of day i'm relatively sober usually
[18:07:52] <theBear> and he's a complete asshat, i've got tired of listing the reasons for sceptical persons like yourself, but look up your ##electronics log, it'll be there a lotta times, keywords include "asshat dave" and "fuck that dave guy" and "claims i don't exist"
[18:08:58] <jadew> heh, I find the video blog entertaining and the forum is really great
[18:10:09] <jadew> I actually found someone there to open and sniff his data adapter (from a brymen multimeter) and was able to make my own
[18:10:31] <jadew> I really like that crowd
[18:12:50] <Valen> jadew: I don't know why but i imagined somebody actually sniffing an adaptors insides
[18:12:52] <Valen> with their nose
[18:13:09] <theBear> entertaining shouldn't make up for technically flawed and/or incorrect, almost constantly, and using made up or guessed/assumed facts constantly in the non-technical realm
[18:13:09] <jadew> Valen, haha
[18:13:32] <Casper> is HV programming plain SPI but with 12V reset? or the protocol is different?
[18:13:54] <theBear> Casper, on avr i THINK it's parallel programming, at least on the older chips
[18:15:09] <jadew> Casper, I think so
[18:15:39] <jadew> http://www.avrfreaks.net/forum/tut-hard-avr-programming-methods
[19:44:36] <gee__> hey, dumb/quick question - I've got a uint8_t number (a counter, really) which I want to use as an address to eeprom_read_byte(), which expects (uint8_t *) as an argument type... casting my variable to a (uint8_t *) throws a "cast to pointer from integer of different size" warning in the compiler.
[19:46:06] <gee__> the function is actually pretty simple - for(i=0;i<64;i++) { if (eeprom_read_byte(i) != 0xFF) break; }
[19:46:34] <gee__> err, s/function/application
[19:48:19] <gee__> so yeah, what's the best way to get that small snip of code to compile cleanly and do what I want? thanks!
[19:49:19] <jadew> cast it
[19:49:39] <jadew> are you using c++ or c?
[19:50:05] <gee__> eeprom_read_byte((uint8_t *) i) throws a warning
[19:50:07] <gee__> c
[19:50:57] <jadew> well, you could declare i as uint8_t *
[19:51:41] <jadew> or you could do (char *)0 + a
[19:51:55] <jadew> (should generate the same code)
[19:53:33] <gee__> eeprom_read_byte((uint8_t *) 0 + i) compiles cleanly
[19:53:50] <gee__> doesn't seem right adding a uint8_t directly to a pointer though *shrug*
[19:54:02] <jadew> it's actually fine
[19:54:07] <jadew> and commmon
[19:55:37] <gee__> good to know! thanks.
[19:55:43] <jadew> np
[20:00:06] <gee__> one more Q... just to fix the last nagging thing in something I'm coding up. I've got a couple of arrays in progmem, and I want to make an array of pointers to those arrays, with that sitting in progmem also. example
[20:00:22] <gee__> const uint8_t array_0[] PROGMEM = {1, 2, 3};
[20:00:31] <gee__> const uint8_t array_1[] PROGMEM = {4, 5, 6};
[20:00:45] <gee__> const uint8_t arrays[] = {array_0, array_1};
[20:01:03] <gee__> err, *arrays[] = ...
[20:01:28] <gee__> this right now compiles fine, but to put arrays[] in progmem
[20:01:58] <gee__> if I do const uint8_t *arrays[] PROGMEM = {...}; I get "variale arrays must be const in order to be put into progmem"
[20:02:26] <gee__> am I missing a second const?
[20:05:04] <gee__> cancel that! got it figured out
[20:05:13] <gee__> const uint8_t * const arrays[] = { ... };
[20:05:23] <gee__> thanks for the help!
[20:09:45] <jadew> why did he leave...
[20:13:47] <Tom_itx> you were too slow
[20:53:31] <Casper> does anyone have a good bootloader suggestion? for serial port, compatible with usb-serial
[20:54:12] <Tom_itx> is app109? about bootloaders?
[20:54:18] <Tom_itx> isn't it serial?
[20:54:39] <Tom_itx> i forget the app # but i think that's it
[20:55:52] <Casper> with the usb-serial and my bit banged serial programmer... it's unacceptably slow
[20:56:38] <Tom_itx> you just want a faster programmer?
[20:57:20] <Casper> actually yeah
[20:57:32] <Casper> because there is a limit of slowness...
[20:57:40] <Tom_itx> mine are pretty quick
[20:58:04] <Casper> usb-spi?
[20:58:08] <Tom_itx> yes
[20:58:14] <Tom_itx> hardware
[20:58:36] <Tom_itx> i gotta run for a bit... back in 30-45
[20:58:44] <Casper> do you know what is the difference between standard and high voltage?
[20:58:50] <Casper> is it only and only the reset voltage?
[20:58:51] <Tom_itx> 12v
[20:58:53] <Tom_itx> on reset
[20:58:56] <Tom_itx> yes
[20:59:07] <Tom_itx> unless it's HVPP
[20:59:15] <Tom_itx> then i'm not sure
[20:59:17] <Casper> so basically you take a standard one and wire it so it send 12V instead of 0V and that's it?
[20:59:20] <Tom_itx> bak in a few...
[20:59:28] <Tom_itx> different protocol
[20:59:30] <Tom_itx> slightly
[20:59:37] <Casper> ah
[20:59:50] <Tom_itx> it does work that way on TPI though
[20:59:56] <Casper> Reading | ################################################## | 100% 82.45s
[20:59:57] <Tom_itx> not ISP
[21:00:03] <Casper> that's 1290 bytes
[21:00:13] <Tom_itx> brb
[21:00:16] <Casper> ok
[21:31:56] <Tom_itx> Casper that's kinda slow
[21:34:09] <Casper> yeah
[21:34:27] <Casper> that's due to the usb-serial dongle...
[21:34:46] <Casper> I used to do that in like 2-3 seconds I think with the hardware port
[21:35:51] <Tom_itx> i don't have a large program to test with...
[21:36:14] <Casper> that ain't a big program
[21:36:20] <Casper> it's only 1290 bytes..
[21:37:24] <Tom_itx> my blink is 713
[21:37:52] <Tom_itx> takes about half sec
[21:38:41] <Casper> that's more than acceptable
[21:39:29] <Casper> how much do you sell those with shipping to canada and how long it should take?
[21:40:19] <Tom_itx> just a sec
[21:40:50] <atom1> Reading | ################################################## | 100% 0.07s
[21:40:50] <atom1> avrdude: verifying ...
[21:40:50] <atom1> avrdude: 246 bytes of flash verified
[21:41:03] <Tom_itx> tested on the linux box
[21:41:31] <atom1> Writing | ################################################## | 100% 0.08s
[21:44:48] <Tom_itx> Casper it takes as long as your customs want's it to :D
[21:45:10] <Tom_itx> pm
[21:45:11] <Casper> yeah.... stupid customs...
[21:45:35] <Casper> I think the delays is volountary
[21:45:42] <N1njaneer> Fill out your commercial invoice correctly and it generally sails right through :)
[21:45:44] <Tom_itx> so your file might take a sec to program
[21:45:48] <Casper> so people don't send food/animals throught the customs
[21:46:50] <Casper> I once made one with an atmega88pa-pu... but I'm not sure if that chip is junk or if I got a bad batch... but it never worked proprelly
[21:54:05] <Tom_itx> should program yours in .4 sec if my calcs are right
[21:54:27] <Tom_itx> and i'm still at the default spi rate
[21:54:42] <Tom_itx> you can bump that up if you're feelin froggy
[21:55:31] <Tom_itx> it just has to be at least 1/4 the clock rate
[22:08:31] <Tom_itx> one took 8 days but another took 16 to canada
[22:08:32] <anton02> how does one import the gcc version of millis()?
[22:09:48] <anton02> Tom_itx: how come you dont hang around ##electronics
[22:10:03] <Tom_itx> because it got to be a stupid channel
[22:10:11] <Tom_itx> maybe it's better now but i left
[22:10:18] <anton02> oh okay
[22:10:32] <Tom_itx> more social than electronics
[22:10:37] <Casper> brb
[22:26:14] <Tom_L> ping
[22:26:39] <jadew> PONG < > << <> > < < > > <
[22:26:44] <Tom_itx> what a delay...
[22:26:58] <Tom_L> finally caught up with the rest of the world
[22:27:06] <Tom_L> http://tom-itx.ddns01.com:81/~webpage/boards/USBTiny_Mkii/new_kits/USBTinyMkII_adapter_kit_desc.jpg
[22:27:18] <Tom_L> Casper
[22:27:23] <Casper> o/
[22:27:30] <Tom_itx> the bb adapter
[22:27:56] <Casper> ok
[22:27:57] <Tom_itx> i don't think i have any of the 10 pin now
[22:28:05] <Tom_itx> nobody really used them except rue_house
[22:29:01] <rue_house> 10 PIN RULEZ!
[22:29:10] <rue_house> YYYYYYYYYYYYYYYYYYAAAAAAAAAAAAAAAAAAAAAAAAAAAAAHHHHHHHHHHHHHHHHHHHHHH!!!!!!!!!!!!!!!!!!!!!!!
[22:29:58] <Tom_L> hah that pic shows one of the earlier programmers
[22:30:45] <Tom_L> Casper, you were asking about the 12v on reset...
[22:30:54] <Tom_L> http://tom-itx.ddns01.com:81/~webpage/boards/USBTiny_Mkii/new_kits/Accessory1_desc.jpg
[22:31:13] <Tom_L> that works on the TPI (attiny 4,5,9,10,20 etc)
[22:31:38] <rue_house> ok I just spent $8 on ebay, callin it quits for the night
[22:31:40] <Tom_L> since the first 4 there only have 4 io
[22:31:53] <Tom_L> rue_house, you're $7 over your budget!
[22:32:18] <rue_house> but I'm only buying like 4 or 5 times a month
[22:32:32] <rue_house> definitly not more than 15
[22:40:05] <Casper> but rue.... why 10?!?
[22:41:10] <rue_house> 10?
[22:41:14] <rue_house> oh
[22:41:28] <rue_house> there has NEVER been a dispute about the pinout of a 10 pin isp
[22:41:39] <rue_house> and you get a programming led
[22:42:04] <Tom_itx> pfft
[22:42:08] <Tom_itx> so does the USB one
[22:42:23] <rue_house> ?
[22:42:34] <Tom_itx> it's got a programming led on it
[22:42:37] <rue_house> oh
[22:42:48] <rue_house> but 10 pin has a rock solid pinout
[22:43:01] <rue_house> there has never been a programmer out there with mixed up pins
[22:43:02] <Tom_itx> 6 is less board space
[22:43:13] <rue_house> really? .2" of space?
[22:43:18] <Tom_itx> uh huh
[22:43:29] <Tom_itx> you'd be surprised how many smt parts will fit there
[22:43:46] <rue_house> and tell me, when your making baords, do you also lose the mounting screws to save space?
[22:43:51] <Tom_itx> take a phone apart if you have doubts
[22:44:09] <Tom_itx> hotglue!!
[22:44:15] <rue_house> yea...
[22:44:35] <rue_house> its not cause the board costs too much eather, its cause your obsessed about saving space
[22:44:57] <rue_house> iphone pico on the head of a pin
[22:45:13] <rue_house> $4000 and it borks on the first firmware upgrade you give it
[22:45:33] <Tom_itx> i don't doubt it
[22:45:42] <rue_house> THE DAY WILL COME!
[22:46:03] <rue_house> and, as usual, everyone will just shake their head and keep asking why people buy apple
[22:59:40] <anton02> how essential do you think it is to super sample with photodiodes?
[23:00:14] <anton02> do you get random fluctuations very often for no reason?
[23:01:43] <N1njaneer> Dig deeper to understand what is happening, rather than "for no reason" :)
[23:02:23] <N1njaneer> Optical noise, circuit noise, trigger noise, etc. Characterize your input first. Then it will become obvious if you need to condition it in analog, digital, etc :)
[23:03:01] <N1njaneer> That question is too application-specific to give a broad answer to.
[23:03:10] <anton02> alright, thanks
[23:03:36] <N1njaneer> What are you trying to do?
[23:06:10] <anton02> match the speed of a toy car to another vehicle, and the other vehicle as an emitter, and the toy car has a photodiode. By differntiating the change of signal with respect to time i can derive the required velocity
[23:07:10] <anton02> just wondering how long i should wait between taking the two samples to derive velocity
[23:07:25] <anton02> and also whether super sampling would be needed as well
[23:07:59] <anton02> other vehicle has an emitter*
[23:19:33] <N1njaneer> Have you calculated the pulse ranges, shifts, and temporal resolutions necessary to match the kinds of velocity accuracy you are looking for?
[23:20:01] <N1njaneer> I assume you are trying to measure Doppler shift of the pulse trains
[23:20:45] <anton02> i thought doppler was sound waves
[23:21:16] <anton02> im talking about light
[23:21:31] <N1njaneer> Doppler applies to any shift of preceived wavelength based on relative velocities of source and observer. It's what causes red and blue-shift of wavelengths in cosmology, among other things.
[23:22:04] <anton02> isnt light way too fast to measure the doppler shift with a simple MCU
[23:22:24] <N1njaneer> You would have to take it in to the analog realm first, then measure.
[23:22:53] <N1njaneer> I am missing how you are intending to do this with photodiodes, if you aren't measure doppler shift of the pulse-trains.
[23:22:54] <anton02> 2 metres apart they are
[23:23:27] <N1njaneer> Are you just trying to convey data, or to actually measure the velocity between the two vehicles using light?
[23:23:39] <anton02> 2nd one
[23:24:09] <anton02> this will be done by measure the rate of changing of intensity
[23:24:10] <N1njaneer> What is your intention with the photodiodes, then? Explain the principle you are trying to use in broad-stroke to do what you are proposing?
[23:26:05] <anton02> with knowledge of the toy cars speed by having control over its motors, this information can be used in conjunction with the rate of change of intensity from the IR beacon to the sensor
[23:26:25] <anton02> to determine the velocity difference between the two vehicles
[23:26:32] <N1njaneer> Oh, so you're not proposing to do a time-of-flight measurement, then.
[23:26:46] <anton02> nope
[23:26:53] <N1njaneer> Okay, that helps clarify
[23:27:10] <N1njaneer> Problem is you will not get a linear relationship of intensity from the beacon to the sensor.
[23:28:56] <anton02> yeah i looked at the angle intensity graph and its exponential, but my estimation will be quite rough any way and no greater than the error induced by there not being a linear relationship
[23:28:58] <N1njaneer> There are far too many things which will affect the fall-off, as well as introduction of noice.
[23:29:25] <N1njaneer> One thing you might want to consider, though, still using the optical transmission method
[23:30:08] <anton02> there wont be any other sources of infrared light in the room
[23:30:08] <N1njaneer> Put an encoder on a wheel of the car, and pulse the beacon to the encoder. That will give you an absolute locked relationship of velocity between the vehicles with error only down to uncertainty in the encoder and the wheel diameter.
[23:30:55] <N1njaneer> Any incandescent, flourescent, sunlight, or phosphor source will contribute IR
[23:31:16] <N1njaneer> (such as white LEDs)
[23:31:17] <anton02> are you sure about incandescent
[23:31:29] <N1njaneer> Incandescent is the worst possible of all
[23:31:46] <anton02> sorry i meant fluourescent
[23:32:03] <N1njaneer> Yes, flourescents emit IR up and past 940nm
[23:33:05] <N1njaneer> Trust me on this, I've done extensive work with 940nm IR data transmissions systems for the past three years :)
[23:33:15] <N1njaneer> There's a lot of little gotchas
[23:33:52] <anton02> okay well my beacons emit 940nm DC
[23:34:06] <N1njaneer> Yes, which means it's impossible to pick out of background noise.
[23:34:44] <anton02> i already know I can, I've tested it. The beacon's intensity is much greater than the background noise
[23:35:07] <N1njaneer> Then go make it happen :)
[23:35:28] <N1njaneer> But there is a reason why IR transmission systems pretty much always modulate at ~40Khz or so
[23:35:37] <anton02> im just trying to take things a step further with speed matching now. Before I was just making my car stop when it reached the other vehicle
[23:36:23] <N1njaneer> Then I would use the encoder on the wheels and relay exact velocity and match in a deterministic, linear fashion :)
[23:37:20] <anton02> do you mean relay as in send data from the other vehicles beacon?
[23:37:38] <N1njaneer> Yes
[23:37:54] <anton02> i guess that would sort of take the challenge out of it is all
[23:38:30] <N1njaneer> Implement your idea, then use that as a fallback if it doesn't work as you expect it to.
[23:39:07] <N1njaneer> But most importantly, work the problem well enough that you can absolutely say why or why something is or is not sufficient to do what you are trying to do.
[23:39:19] <N1njaneer> why or why +not
[23:39:46] <anton02> alright, thanks
[23:39:57] <N1njaneer> Hopefully that was useful
[23:40:17] <anton02> yep it was
[23:47:39] <N1njaneer> Do let us know how it works for you! I'll be curious to see your findings :)
[23:47:54] <N1njaneer> I'm always for trying new things, but I know how fickle IR can be.
[23:48:16] <N1njaneer> If you are going to go that route, I'd suggest trying to get as best an isotropic emission you can from your beacon.
[23:48:29] <N1njaneer> The detector may be the harder part.
