Back
[00:17:45] <Casper> dropamp...
[06:18:08] <LeoNerd> RikusW: You around? I have some HVSP musings in my head. Namely: wondering what sort of chip I should put on my HVSP burner board. A tiny2313 has the advantage of a real hardware UART to talk STK500 proto. with, but at only 2Ki of flash and a massive 20 pins, I feel both too small and yet too big. A tiny84 has enough pins to talk to it, but no hardware UART. Still, in 8Ki I'm sure I can write up a software one
[06:18:15] <RikusW> yep
[06:18:49] <RikusW> t84 might work ok
[06:19:02] <RikusW> I used mega32u2
[06:19:21] <Lambda_Aurigae> does the stk500 protocol support hvsp?
[06:19:24] <LeoNerd> Yeah but that's like... hard. And also massive overkill
[06:19:26] <LeoNerd> Totally
[06:19:42] <RikusW> stk500 supports ISP HVPP HVSP
[06:19:45] <LeoNerd> I'm planning to make a little clone that -only- talks HVSP, for doing full GPIOs on tiny85 and 84 cips
[06:20:00] <LeoNerd> I think for my next project I might do a debugWire probe :)
[06:20:13] <malinus> LeoNerd: it's funny that people still use the STK500. They make us use it in school :V
[06:20:24] <LeoNerd> The hardware is /horrible/
[06:20:39] <malinus> and very outdated.
[06:20:42] <LeoNerd> But the protocol is damn-near universal; e.g. the only hvsp thing avrdude knows what to talk, is STK500
[06:21:12] <Lambda_Aurigae> I have an stk500 here...and an avr dragon.
[06:21:43] <Lambda_Aurigae> might have used either of them half a dozen times in the last 10 years.
[06:22:01] <RikusW> LeoNerd: does your PC even have a serial port ?
[06:22:13] <LeoNerd> RikusW: I have a metric tonne of FTDIs and clones :)
[06:22:20] <Lambda_Aurigae> bah.
[06:22:21] <malinus> RikusW: usb-serial devices are $1
[06:22:28] <LeoNerd> Also, you can get USB <-> TTL devices for, .. yah. what he said
[06:22:30] <Lambda_Aurigae> hardware parallel and serial on the motherboard here.
[06:23:14] <LeoNerd> http://www.amazon.co.uk/gp/product/B008AGDTA4 <== GBP1, for USB <-> {GND,Tx,Rx,Vcc}
[06:23:17] <Lambda_Aurigae> on a quad core 64bit amd FX-4150
[06:23:58] * RikusW also got lpt and com ports on a _new_ motherboard MSI Z77A-G43
[06:24:27] <Lambda_Aurigae> this is a gigabyte motherboard.
[06:24:48] <Lambda_Aurigae> dual bios is fun when working with linux kernel in bios.
[06:24:54] <RikusW> LeoNerd: tiny84 is 8 pin ? so _just_ enough pins....
[06:24:58] <Lambda_Aurigae> the ability to boot to the backup bios rocks.
[06:26:02] <LeoNerd> RikusW: Hehe... not quite. I need 5 pins to talk HVSP (SDI, SII, SDO, SCI, RESET control) and 2 to talk UART to the host.
[06:26:02] <malinus> but yeah, no point in getting horrible stk500 hardware, when you can get a atmega32u4 devBoard with all the pins flashed out etc. for 10% of the price
[06:26:45] <RikusW> if you use ST662 you need 1 more, so 6 + 2 + 2 power...
[06:27:39] <RikusW> and one more is you control Vtg
[06:27:52] <RikusW> (power to target chip)
[07:57:22] <MarkX> hey, weird question.
[07:59:42] <MarkX> i've always heard of 0.1 as 100 thou. (thousands of an inch). in altium it's being called a "mil", what is that :S
[07:59:58] <LeoNerd> A 'mil' is a 1/1000th of an inch
[08:00:29] <MarkX> so its another way of saying thou
[08:00:31] <MarkX> ?
[08:01:01] <LeoNerd> I expect it's French-inspired, as in 'mille'
[08:01:25] <MarkX> interesting
[08:01:47] <MarkX> its like the royale with cheese
[08:29:11] <MarkX> any good habits to follow when designing a schematic library that gets linked to a footprint?
[08:29:36] <MarkX> for example, in the case of an IC, the VCC at the top and GND at the bottom (as suggested by a video i'm watching)?
[08:31:26] <LeoNerd> I don't tend to put much relationship between the schematic and the footprint
[08:35:00] <MarkX> so your schematic designs for parts don't match the footprints (ie. pins that are on the left of the chip, are on the left in your part)?
[08:43:55] <LeoNerd> Oh, definiely not
[08:44:07] <Tom_itx> i put them where they're most convenient
[08:44:10] <LeoNerd> E.g. my ATtiny84 puts the two power pins on one side, then all of port A nicely sequential, then all of port B
[08:46:28] <MarkX> ah awesome!
[08:47:03] <MarkX> so theres no real standard for it as long as all the pins are there, its neat, and labelled properly?
[08:47:31] <LeoNerd> Indeed. Especially for things like complex logic chips, it's quite common to group logically related pins together in a similar place, regardless of their physical layout
[08:48:53] <MarkX> awesome
[08:49:45] <Tom_itx> the power pin MUST be on top!
[08:49:59] <LeoNerd> Liiies :P
[08:50:08] <Tom_itx> heh
[08:50:25] <LeoNerd> Oh, while we're on Eagle though, I made some symbols last night and I find that wires don't quite attach to the pins properly
[08:50:39] <Tom_itx> who said eagle?
[08:50:41] <LeoNerd> If I draw a wire just to the tip of a pin, it doesn't "stick" - if I move the symbol, the wire stays wher eit was, detached
[08:50:48] <Tom_itx> he's using altium...
[08:50:53] <LeoNerd> Whereas if I draw the wire properly down to the base of the pin, it does stick
[08:50:56] <LeoNerd> Ah
[08:51:06] <Tom_itx> use net not wire silly
[08:51:08] <LeoNerd> Well.. <.< -can- we talk Eagle instead?
[08:51:14] <LeoNerd> Yeah, net does the same thing I think
[08:51:18] <Tom_itx> net will snap to the pin
[08:51:20] <Tom_itx> wire won't
[08:51:27] <LeoNerd> Hmmm
[09:06:03] <MarkX> oh boy
[09:06:18] <MarkX> updating already placed components in altium is not as easy as pressing f5 :|
[09:11:13] <MarkX> lmao
https://twitter.com/WhyAltiumSucks
[09:12:50] <LeoNerd> How's it compare to Eagle?
[09:15:57] <MarkX> eagle is i think 2 buttons. you change a symbol. then pretty much click the refresh button
[09:16:02] <MarkX> and the changes are done
[09:16:06] <LeoNerd> I mean.. in general
[09:16:24] <LeoNerd> I know specifically in this case the 'update all' option in the library menu
[09:16:35] <MarkX> oh
[09:16:55] <MarkX> well you gotta click update from library, then you gotta choose what you want to update, then you have to choose what you want to do with it
[09:17:10] <MarkX> and maybe i did something wrong but my connections (led to pin 3 for example) was lost.
[09:20:49] <malinus> let's talk about free software instead - KiCad
[09:21:44] <rue_house> kicad, thats the name I was trying to thik of
[09:22:54] <malinus> I would feel filthy if I used eagle :)
[09:24:32] <twnqx> the one and only time i tried kicad
[09:24:41] <twnqx> it was an unusable abonimation
[09:25:15] <twnqx> then i just bought eagle on a whim and was happy, but the recent licensing policies make me rethink :X
[09:25:21] <rue_house> agreed, its changed a lot since eagle isn't free
[09:25:37] <kline> whats this about recent eagle licensing?
[09:25:41] <twnqx> i mean, 5 was cool
[09:25:52] <twnqx> paying the update to 6 after like 5-6 years was ok
[09:26:06] <twnqx> but i won't be another upgrade to 7 just a year later
[09:26:20] <twnqx> s/be/pay
[09:28:07] <twnqx> more the frequency than anything
[09:28:53] <malinus> twnqx: aww, was the gui not to your liking :)?
[09:31:37] <twnqx> the unusability? no, not really
[09:32:14] <MarkX> i dont understand, eagle has freeware option :S
[09:32:20] <MarkX> i used that and was happy with it
[09:33:07] <twnqx> i don't
[09:33:21] <twnqx> not enough layers, not enough schematic sheets, not enough size
[09:33:29] <MarkX> ah
[09:33:34] <MarkX> maybe i didn't hit those limits
[09:33:43] <MarkX> with my basic designs
[09:36:10] <Jartza> oh well
[09:36:16] <Jartza> http://tagsu.io/
[09:36:22] <Jartza> I guess that's cool enough now :)
[09:43:59] <MarkX> btw Jartza cool concept
[09:44:35] <MarkX> throw a thing on the back of it and make it into a belt buckle
[09:44:56] <Jartza> :)
[09:45:14] <Jartza> to be honest, that's very unconventional way to use web :D
[09:45:37] <MarkX> or better yet, give it some IR, get a remote for it. program it with the ability to have multiple phrases. then have it able to select phrase based on remote
[09:45:49] <MarkX> then put it on the back of cars so you can tell other drivers to go !@#$ themselves
[09:46:15] <autrilla__> How does CAN behave with interference? It only has a 1V differential between high and low, whereas RS485 has higher ones
[09:46:28] <Jartza> it already has ability to show multiple screens of text
[09:46:31] <Jartza> 42 to be exact
[09:47:13] <Jartza> MarkX: that small device is hard to see from the backwindow of a car
[09:47:19] <Jartza> that would need much bigger display
[09:47:34] <MarkX> make my back windshield into a massive !@#$ YOU sign
[09:47:38] <MarkX> XD
[09:48:18] <LeoNerd> autrilla__: if it's a differential current loop, the voltage isn't so important
[09:48:27] <Jartza> needs some kind of led matrix maybe
[09:49:06] <autrilla__> LeoNerd, well, noise will be more or less the same whether I use 500V or 1V
[09:49:16] <autrilla__> But it's more noticeable on 1V
[09:50:04] <LeoNerd> See again under "differential omde"
[09:50:06] <LeoNerd> mode
[09:51:27] <autrilla__> yes, using a differential loop noise _should_ be the same on both wires
[09:51:40] <autrilla__> Well, it should be opposite actually
[11:04:43] <rebrec> hello
[11:15:06] <MarkX> rebrec: hi
[12:21:52] <rebrec> i am trying to play with a "hackme challenge" based on an atmega32 that i try to port to arduino uno. The aim of this challenge is to light a led using (i guess) a buffer overflow present in the code. The firmware ask for a password over serial line. It store each input char inside a fixed size array and only stop adding char when it reads a carriage return. (==> buffer overflow). I am new to MCU, specifically
[12:21:55] <rebrec> on this kind of games. I don't relly know how the stack is handled, and if i overwrite the EIP (or Program Counter i suppose) value of the return function address on the stack, might i point it to some data in SDRAM ? or can it only point to Flash data please ?
[12:28:59] <LeoNerd> AVRs are Harvard architecture
[12:30:00] <Casper> rebrec: avr can not execute code in ram, only in flash
[12:31:46] <LeoNerd> (-basically- what I said, only reworded :) )
[12:42:08] <rebrec> ok LeoNerd and Casper thanks both :)
[12:43:11] <N1njaneer> rebrec: Yes, this would probably be a bit challenging on an AVR unless you can buffer overflow in to a section of RAM which can then in turn affect the state of the LED. That would require looking at the code and the memory map.
[12:43:57] <N1njaneer> If there is some variable called like "ledState" which falls in to SRAM somewhere after the unprotected buffer, then it should be absolutely trivial.
[12:45:38] <rebrec> N1njaneer: i guess i find the way to do it : the firmware ask the user to authenticate and after being authenticated (default password=admin) i have access to a menu where i can store a 10 char password saved in flash... i guess i might change the pass to some ASM value setting the pin to output and then setting pin high and then i would have to know the address in flash of this data so that i can overwrite eip
[12:45:41] <rebrec> to this value ?
[12:48:39] <rebrec> there is nothing except : an authentication message asking for a valid password to access to an "admin screen" where i can 1-change password and save it to flash (10 char with ending \0) 2-go back to "user mode" to authenticate again very basic
[12:51:14] <rebrec> are there ways for me to debug/detect/calculate where my "password" is stored on the stack and where is IP stored on the stack ? under a computer i usually use a debugger but as saide before, MCU are new to me (and buying hardware debugger is not an option) i can change the firmware at will for testing purpose if it can help but the resulting "exploit" might have to work with the original firmware
[12:52:27] <rebrec> sorry for questions mens but even if this "challenge" is not realistic at all, it raises questions to me that i never had before on MCU architecture and compilation issues
[12:56:26] <EEight> i have a basic question regarding bit masking and combining to a single byte:
https://gist.github.com/patricksebastien/db1d59bcafaf3a300c18
[12:56:32] <EEight> if anyone can spare 2 min
[12:56:38] <N1njaneer> rebrec: FYI -- the hardware debugger for doing full in-circuit stepping and debug is only $35 USD, so not terribly expensive if you want to consider it. It used to be several hundred dollars, but not any more. It is invaluable to have for working with AVR's
[12:56:45] <N1njaneer> EEight: Sure, ask away.
[12:57:00] <LeoNerd> A debugWire interface?
[12:57:23] <jeremyabel> N1njaneer: the dragon?
[12:57:34] <EEight> N1njaneer: i am masking 2 port(B&D) for 7 switches (3 on b, 4 on d). now i need to pack this to a single bytes
[12:58:23] <N1njaneer> Jeremy/Leo: No, the ATMEL-ICE. Does ISP/JTAG/SWD/DW for AVR and full SW/JTAG for ARM. Swiss-army-knife programmer that does everything for AVR and SAM all in one device :)
[12:58:36] <N1njaneer> EEight: In what arrangement do you need the packing?
[12:58:37] <LeoNerd> Ah.. hmmm
[12:58:50] <EEight> N1njaneer: doesn
[12:58:51] <LeoNerd> But dW is still an "undocumented" format, asides RikusW's work
[12:58:52] <N1njaneer> EEight: Generally it's just a combination of shifts and OR's
[12:58:57] <LeoNerd> So.. no free tools for me :(
[12:59:07] <jeremyabel> N1njaneer: ah the new thing
[12:59:20] <jeremyabel> the replacement for the mkII yeah?
[12:59:33] <N1njaneer> jeremyabel: Yes, they are very nice! That and I just had another ISP-MKII go belly-up on me, which they tend to do after a few years.
[12:59:38] <N1njaneer> jeremyabel: Yep!
[12:59:42] <EEight> N1njaneer: oups... it doesn't matter, in the end i will simply "unbyte" it
[12:59:49] <jeremyabel> wow I've had my mkII for 4 years and it's steady as ever
[12:59:57] * jeremyabel buys ICE though
[12:59:59] <jeremyabel> cause cheap
[13:00:36] <N1njaneer> EEight: Just remember that AND will mask bits, OR will combine bits, XOR will flip bits, and then << and >> shifts are your friends.
[13:02:15] <N1njaneer> jeremyabel: Yes, I've had several mkII's last for a few years, but then will just randomly stop working and will give an orange flashing light no matter what you do. If lucky it's just the ribbon cable that needs replacing, but generally once they start doing that it's just pitch-out and get a new one. Atmel seems peripherally aware of this problem but the official answer is that "The ISP-MKII is
[13:02:16] <N1njaneer> not considered to be a production-environment device, only for doing development work"
[13:02:23] <N1njaneer> but +they will
[13:02:30] <jeremyabel> weird
[13:02:41] <N1njaneer> And of course now the MKII's is officially EoL'd
[13:02:44] <jeremyabel> yeah
[13:02:51] <jeremyabel> I like the look of the ICE better too
[13:02:58] <jeremyabel> looks like it might sit on my desk a bit better than the MKII did
[13:03:28] <jeremyabel> fit in my pocket better too
[13:03:35] <N1njaneer> Yeah, it just looks a bit funky on the weird adapter board. Though I've started using the miniature 10-pin connectors on my SAM designs now.
[13:03:47] <jeremyabel> wtf do you use to program multiple devices anyway?
[13:03:53] <jeremyabel> like multiple at once
[13:04:29] <jeremyabel> I've never looked into that, and the project I was doing just finished a few days ago, so I don't really need it, but I am curious
[13:04:50] <jeremyabel> in a factory-level deal, would you just buy the chips pre-programmed from atmel?
[13:04:58] <N1njaneer> jeremyabel: You can actually do this with any Atmel programmer
[13:05:08] <jeremyabel> oh?
[13:05:13] <jeremyabel> is it as simple as daisychaining?
[13:05:15] <N1njaneer> You can gang-address them on the command line, so like if you want to run 10 of them at once it's trivial to do.
[13:05:20] <jeremyabel> dope
[13:05:27] <N1njaneer> You just need 10 units.
[13:05:35] <jeremyabel> ahhhh
[13:05:46] <jeremyabel> yeah I doubt I'll need that for a while :P
[13:06:01] <N1njaneer> There ARE production programmers out there, but the cost for a produciton programmer is astronomical compared to just buying cheap Atmel programmers and ganging them together and/or replacing when they stop working.
[13:06:07] <jeremyabel> yeah, I bet
[13:06:28] <N1njaneer> We'll easily do many thousands of production programmings with an MKII with no issue.
[13:06:45] <jeremyabel> yeah I probably did the same in total for all my cup boards just over time
[13:06:49] <jeremyabel> they're out in the world now!
[13:06:55] <jeremyabel> they've flown away
[13:07:11] <jeremyabel> good luck, little cups
[13:07:41] <N1njaneer> I suspect it may just be a case of the MKII's getting ESD'd or such over time - might not have had transient protection built in to them. It's always been the ones on my desk that get used on many different boards that eventualy seem to go!
[13:07:52] <jeremyabel> huh
[13:07:57] <N1njaneer> jeremyabel: Yeah, scary when you realize you have micros out in the world running 24/7!
[13:08:02] <jeremyabel> yeah...
[13:08:16] <jeremyabel> they'll be journying across canada for a while
[13:09:32] <jeremyabel> happy to move on to my fpga synth project now
[13:09:49] <N1njaneer> Doooo eeet!
[13:10:00] <N1njaneer> Go Xilinx!
[13:10:06] <jeremyabel> yep yep!
[13:10:09] <N1njaneer> Spartan3 or Spartan6
[13:10:13] <jeremyabel> yeah, that's the plan
[13:10:22] <jeremyabel> probably the spartan3, the $80 one
[13:11:10] <N1njaneer> I'm not an expert on them, but I've done a dozen or so designs with Spartan2/3 over the past decade and they've all worked great. Most still in service. Though just EoL'd the one design here that uses a Spartan2 since like 3-4 pieces of silicon for it are no longer avaliable. :(
[13:12:05] <learath> man. spartan2 is *ancient*
[13:12:45] <jeremyabel> N1njaneer: looking at the basys2
[13:12:48] <jeremyabel> http://www.digilentinc.com/Products/Detail.cfm?NavPath=2,400,790&Prod=BASYS2
[13:14:01] <N1njaneer> Seems like a resonable starting point, thouh not a whole lot of breakout.
[13:14:47] <jeremyabel> yeah, only 24 pin breakout
[13:15:02] <jeremyabel> http://www.digilentinc.com/Products/Detail.cfm?NavPath=2,400,1288&Prod=BASYS3 the basys3 has 4x12
[13:15:07] <N1njaneer> Spartan-3 is avaliable in TQFP to up around the 200,000 gate level, so pretty easy to lay out your own board.
[13:15:18] <jeremyabel> oh nice
[13:15:25] <jeremyabel> was hoping I could get a non-bga one
[13:15:48] * MarkX highfives N1njaneer
[13:15:51] <N1njaneer> Really just need a few resistors, decoupling caps, and a <$1 SPI flash memory for the non-volatile FPGAs
[13:16:21] * N1njaneer highfives back! "Why are we high-fiving?" :D
[13:17:08] <jeremyabel> word
[13:17:19] <jeremyabel> does it take just 3.3v or what?
[13:17:22] <N1njaneer> The other nice thing is that with the SPI flash, you can reprogram the SPI via the FPGA's JTAG port via the Xilinx utilities. iMPACT will automatically upload a little image in to the FPGA that lets JTAG reprogram the SPI, so it's very simple to do a bring-up
[13:17:23] <jeremyabel> I hear some of them take weirdo voltages
[13:18:07] <N1njaneer> jeremyabel: Spartan3 is generally 3.3V or 2.5V (for the IO) and then a seperate 1.2V for the core, which you can just stick on a simple SOT-223 regulator for.
[13:18:14] <jeremyabel> word
[13:19:11] <N1njaneer> The IO is actually settable to like a dozen different standards from LVCMOS (ala 3.3V TTL) to LVDS, etc and different IO banks on the device can even run on different voltages.
[13:19:30] <jeremyabel> weird
[13:19:42] <jeremyabel> yeah I have a feeling once I like jump into this it'll be a big "culture shock"
[13:19:48] <jeremyabel> like working in a different universe
[13:20:01] <jeremyabel> hell I'll be happy one I get a clock pulse out ;)
[13:20:09] <N1njaneer> Only pain is that AFAIK the Spartan3 series IS NOT 5V IO tolerant, so occasionally have to add either resistors or (better case) level shifters if you need to interface with 5V logic. Though it's generally just easier to try to run everything as 3.3V these days if possible :)
[13:20:26] <jeremyabel> yeah, I've mentally switched myself over to 3.3V these days
[13:20:26] <learath> N1njaneer: 200k gates in a tqfp? I only see the slx4 and the slx9 offered in tqg144
[13:20:30] <MarkX> high five for the sake of it
[13:20:32] <MarkX> XD
[13:20:37] <jeremyabel> total fives of high!
[13:20:44] <MarkX> aww yisss
[13:21:16] <N1njaneer> The FPGA stuff is probably the most challenging in the way that you have to think about things in a completely different way from an implementation standpoint. You have to remember that you are fundamentally defining things in terms of register-transfer logic (RTL) and that unless you build state machines to sequence things, EVERYTHING inherently runs all at once, in parallel
[13:21:38] <jeremyabel> yeah I feel like seeing some examples would help
[13:21:45] <jeremyabel> I assume the dev board will come with some
[13:21:48] <N1njaneer> learath: Spartan3 -- xc3s200a and friends. Pretty sure Spartan6 has a few as well.
[13:21:53] <jeremyabel> wtf is the hello-world for fpgas?
[13:22:32] <learath> N1njaneer: ahh ok I'm looking at the spartan6
[13:22:37] <N1njaneer> jeremyabel: You can synthesize a UART and print "Hello World!" to a UART! But in simpler cases something like a blinking LED based on a high-speed clock signal coming in would probably be more practical :)
[13:22:57] <jeremyabel> yeah that last bit sounds good to me
[13:23:08] <jeremyabel> I don't want to start cramming modules in, I want to get the fundamental small things first
[13:23:17] <N1njaneer> jeremyabel: I prefer VHDL, but Verilog is pretty much equivalent - just a different language syntax. VHDL just looks and feels more comfortable to me, os I use that instead. :)
[13:23:19] <jeremyabel> can you do an ADC with an FPGA?
[13:23:29] <N1njaneer> jeremyabel: If you connect one, sure :)
[13:23:39] <jeremyabel> oh so you can't build one in?
[13:23:51] <learath> there are FPGAs with analog built in
[13:24:03] <N1njaneer> Nope, unless you have an FPGA that has it.
[13:24:06] <jeremyabel> ah ok
[13:24:24] <jeremyabel> does the same go for analog output?
[13:25:41] <N1njaneer> The rub is (from my understanding) that generally digital and analog silicon are usually fabbed with very differing processes, so it's very challenging to get high performance digital AND analog on the same device. There are exceptions to this and you are starting to see more on-board analog elements with digital devices, but usually they aren't as high a performance. So it really depends on your
[13:25:41] <N1njaneer> specific needs.
[13:26:28] <N1njaneer> Personally I'd rather just not have the analog elements on the digital stuff anyhow, since there's far more flexibility in externally specing the right part for the job, etc.
[13:26:42] <N1njaneer> ALSO --
[13:27:05] <N1njaneer> It's important to understand the strengths AND weaknesses of a microcontroller, microprocessor, CPLD, and FPGA
[13:27:50] <jeremyabel> of course
[13:28:10] <jeremyabel> trying to ponder the best way to do this synth
[13:28:19] <jeremyabel> like do I output just waveform junk to a dac
[13:28:28] <jeremyabel> that's what most digisynths do at least
[13:28:46] <jeremyabel> to start I'll just focus on making pwm output :P
[13:28:53] <jeremyabel> after getting a clock output
[13:29:13] <jeremyabel> need to study how the SID works
[13:29:38] <[F_F]> is there a way to manipulate a ring oscillator frequency?
[13:29:47] <[F_F]> I am doing my ring oscillator using NOT gates
[13:30:39] <N1njaneer> SID is an analog/digital hybrid chip -- you won't be able to truely do it entirely in analog or digital exclusively, though my understanding is that some people have done some decent digital emulation of it :)
[13:30:47] <jeremyabel> indeed
[13:30:53] <jeremyabel> I think there's already an FPGA sim out there too
[13:31:03] <jeremyabel> mostly just the architecture that interests me
[13:31:13] <N1njaneer> F_F: Probably by adding some frequency limiting with some R/C values between ring gate elements.
[13:31:14] <jeremyabel> like how do you communicate with it, specifying waveforms, pitch, etc
[13:31:40] <jeremyabel> wonder if it just uses SPI or something
[13:31:54] <[F_F]> N1njaneer, I want to be able to modulate the frequency, using voltage or some sort of PWM
[13:32:18] <MarkX> VHSIC Hardware description language - VHDL
[13:32:32] <MarkX> Verilog Hardware description language ...
[13:32:33] <N1njaneer> jeremyabel: You may be interested in some of the work I've done a while back -- real-time digital waveform synthesis via graphical nodal-based elements. Runs out to ASIO at 48Khz, fully real-time, anything-patch-to-anything architecture.
https://www.dropbox.com/s/2s14plidxb6czcz/DSPatchBig.png?dl=0
[13:32:37] <specing> ve he de el
[13:32:51] <specing> also
[13:32:56] <jeremyabel> N1njaneer: yes, quite interested!
[13:33:11] <jeremyabel> N1njaneer: could you send me some more info on that? sourcecode?
[13:33:15] <N1njaneer> F_F: I would just manually code function generators that you can alter the parameters at will
[13:33:16] <specing> can anyone eli5 what '64 macrocells' means?
[13:33:35] <specing> also
[13:33:44] <specing> Im enrolled in a FPGA course @ uni \o/!
[13:34:11] <N1njaneer> jeremyabel: It's in a commercialized product right now, but I am happy to discuss any of the concepts. It's also several hundred source files since the entire GUI is also writting from scratch. :)
[13:34:41] <jeremyabel> ooh fun
[13:34:43] <N1njaneer> It makes pretty vector images out to laser:
https://www.dropbox.com/s/x11rmf4txe1kah0/DS3LiveExample.png?dl=0 :)
[13:34:47] <jeremyabel> I'm working on a node-based gui too
[13:34:58] <jeremyabel> well, porting a friend's to OSX, at least
[13:35:07] <jeremyabel> N1njaneer: oh dope
[13:35:10] <jeremyabel> what commercial product is it?
[13:35:27] <jeremyabel> but yeah I'll def bug you about it when I get some ideas down on paper
[13:35:44] <N1njaneer> We call it DigiSynth. It's also been wrapped in to an OEM version for the lighting market for controlling lasers. Been on a bunch of big name tours as of late :)
[13:36:11] <rebrec> N1njaneer and EEight thanks for your help, i might buy a debugger later but for now, this "challenge" should not require buying tools i guess. Might be more reading of the datasheet or of the way avr-gcc build the code maybe to deduce some memory address... thanks again for your help
[13:37:41] <jeremyabel> N1njaneer: way cool!
[13:37:49] <N1njaneer> Thanks!
[13:38:08] <N1njaneer> rebrec: No problem! The hardware debugger is money well spent if you expect to do more with AVRs
[13:55:29] <MarkX> http://i.imgur.com/F3Jz1GV.gif omg i howled
[14:01:19] <jeremyabel> hahaha
[14:15:42] <rebrec> N1njaneer: any link to a debugger you would recommend please ?
[14:16:44] <N1njaneer> Search "ATMEL-SAM"
[14:17:06] <N1njaneer> DigiKey and Mouser have them in stock, but not sure where in the world you may be.
[14:32:41] <TechChristoph> hi everyone
[14:33:12] <MarkX> N1njaneer: quick question. im designing my first altium pcb.... all the components and traces are grey
[14:33:16] <TechChristoph> do you know a good avr-site for beginners
[14:33:24] <TechChristoph> there are many
[14:33:34] <TechChristoph> but they confusing me a little
[14:33:38] <TechChristoph> :-)
[14:33:43] <Casper> TechChristoph: start by the big datasheet
[14:33:46] <MarkX> the traces are on the bottom layer, i'm assuming i'm just looking at the top of the board?
[14:34:02] <TechChristoph> ok
[14:34:23] <TechChristoph> maybe i should learn c first right ?
[15:10:50] <MarkX> TechChristoph: why don't you start off with arduino?
[15:21:08] <Tom_itx> NOOOO!!!
[15:21:50] <Tom_itx> learn how to blink an led and work your way up. then you will have an understanding of the chips
[15:23:09] <Tom_itx> TechChristoph,
http://tom-itx.ddns01.com:81/~webpage/how_to/atmega168/mega168_howto_main_index.php
[15:26:57] <MarkX> LMAO
[15:27:08] * MarkX knew that comment wouldn't be accepted
[15:27:13] <MarkX> muhuhahahaha
[15:28:15] <[F_F]> what's the smallest AVR chip out there?
[15:29:03] <[F_F]> 1.5mm x 1.4mm?
[15:37:07] <timemage> [F_F],
http://www.atmel.com/Images/Atmel-8127-AVR-8-bit-Microcontroller-ATtiny4-ATtiny5-ATtiny9-ATtiny10_Datasheet-Summary.pdf see page 2. smallest avr package i know of.
[15:41:01] <jeremyabel> [F_F]: if you want SMALL, I'd go with a PIC
[15:43:00] <jeremyabel> they're well-known for having lots of features on low-pincount devices
[15:43:26] <jeremyabel> http://www.microchip.com/wwwproducts/Devices.aspx?product=PIC12F1571
[15:50:25] <autrilla__> You should add "We hate Arduino" to the topic :3
[15:51:54] <rebrec> thanks N1njaneer from France here
[15:51:55] <rebrec> SAM-ICE seems to be used for ARM MCU N1njaneer in fact i am more about ATMEGA 8/16/32 but thanks anyways
[15:57:33] <N1njaneer> rebrec: I assue you it works fine for AVR :)
[15:58:45] <N1njaneer> It is Atmel's new official programer that covers all AVR and ARM parts, and all programming interfaces for both device series.
[15:58:47] <rebrec> good to know N1njaneer from ATMEL it cost 100$ but why not, one of theses days... as you said everything is so easier with a debugger
[15:59:09] <N1njaneer> rebrec: $35 USD if you just want the PCB and no case.
[15:59:17] <rebrec> does it works under linux ?
[15:59:40] <N1njaneer> I believe that programmer is only supported under Atmel Studio right now, so probably just Windows for the moment.
[16:00:09] <rebrec> argh... i was almost ganno click on the buy button ... :/
[16:04:43] <TechChristoph> MarkX: Well i started off with arduino
[16:04:52] <TechChristoph> but now it gets boring in a way
[16:05:10] <TechChristoph> for now i soldered my own developer board
[16:05:33] <TechChristoph> not something great but it works with avrdude so far
[16:06:51] <TechChristoph> thanks for all your help
[16:31:41] <avri> http://www.instructables.com/id/Stupid-Simple-Arduino-LF-RFID-Tag-Spoofer - I have a 100nf capacitor available, is it possible to construct this circuit with a different value inductor? Is it possible to use an inductor instead of a coil?
[16:46:37] <LeoNerd> Well, boo. Having spent 4 hours last night designing the perrrrrrfect (well almost) stripboard layout, I discover I don't have any pieces big enough to fit it. Looks like I shall have to order some more
[17:15:27] <malinus> rebrec: are you talking about the
[17:41:33] <Lambda_Aurigae> avri, a coil IS an inductor..
[17:41:45] <Lambda_Aurigae> and an inductor is a coil.
[17:42:26] <avri> I know... but I am wondering if I can use a component inductor, instead of an actual coil as an anthena - i.e. instead of looping a wire around something for that many turns
[17:42:50] <Lambda_Aurigae> smaller tighter inductor will have a shorter range I do believe.
[17:42:56] <avri> or why would they sell these coils instead of just use an inductor
[17:43:26] <Lambda_Aurigae> and the 10nF / 160uH is not much changeable...it's a tuned circuit...
[17:44:03] <Lambda_Aurigae> if you used a 100nF cap you would have to majorly change the coil/inductor.
[17:44:22] <avri> http://www.instructables.com/id/Stupid-Simple-Arduino-LF-RFID-Tag-Spoofer/step4/Building-the-circuit/
[17:44:54] <N1njaneer> Yes, you want larger surface area with the coil for better range.
[17:44:56] <avri> everything I have is correct (found the right cap)
[17:45:13] <Lambda_Aurigae> http://wiki.smallroom.net/doku.php?id=terd:projects:rfidspoofer
[17:45:19] <avri> the coil, I have 150uH available
[17:45:29] <avri> I am touching the rfid reader's coil, and no reading
[17:45:36] <avri> wondering if my circuit is correct
[17:45:38] <Lambda_Aurigae> that one does it with a 9mH coil and 220pF cap.
[17:45:47] <avri> not that its a complicated circuit or anything
[17:46:01] <avri> but using a digispark instead of an arduino
[17:46:10] <avri> (shouldn't matter really afaik)
[17:46:15] <Lambda_Aurigae> no clue what a digispark is.
[17:46:17] <avri> using pin 5
[17:46:31] <avri> http://digistump.com/wiki/digispark/tutorials/connecting
[17:46:58] <avri> http://digistump.com/products/1
[17:48:20] <Lambda_Aurigae> http://pdf1.alldatasheet.com/datasheet-pdf/view/548198/COILCRAFT/4312RV-905XGLB.html this is what the smallroom.net version uses...
[17:50:33] <avri> yeah, but that's a coil
[17:50:37] <avri> I have a regular inductor
[17:50:53] <avri> that's why I was asking if I should be able to swap these
[17:51:18] <avri> I was assuming that _touching_ the inductor to the reader' coil should compensate for the distance problem
[17:51:20] <avri> but no reading
[17:51:39] <avri> I once ago was able to use an inductor connected to an attiny85 as a tag
[17:51:55] <avri> passive tag - i.e. no power to the attiny
[17:52:08] <avri> but now I want it programmable in the field
[17:53:18] <avri> anyway... thanks for the help time to get some rest :)
[17:53:29] <avri> gn
[17:53:32] <Lambda_Aurigae> what I posted is an inductor...coil...same thing...just very compact.
[17:54:50] <avri> what I have:
http://www.ebay.com/itm/190863432067?_trksid=p2060778.m2749.l2649&ssPageName=STRK%3AMEBIDX%3AIT
[17:56:14] <Lambda_Aurigae> I've never done rfid stuff myself...I got away from analog/RF about 20 years ago.
[17:57:01] <avri> thanks Lambda_Aurigae
[17:57:21] <avri> I will go to sleep now... I'll play more with this tomorrow I guess... :) good night
[17:57:55] <Lambda_Aurigae> nighty night..wet dreams.
[20:23:46] <N1njaneer> So people still hanging about?
[20:24:03] <Tom_L> that's just what we do
[20:24:45] <N1njaneer> Do it more loudly!
[21:02:56] * N1njaneer orders more custom aluminum extrusion.
[21:03:30] <Tom_L> what you gonna make?
[21:03:48] <N1njaneer> We make LED luminary fixtures, among other things. This is a new profile. :)
[21:04:18] <Tom_L> you should come visit the local surplus sometime:
http://tom-itx.ddns01.com:81/~webpage/cnc/local_stock/stock_index.php
[21:04:21] <N1njaneer> Not too bad - only about $880 USD for a new die, then the cost of pushing the aluminum.
[21:04:32] <Tom_L> yeah
[21:04:59] <N1njaneer> Nice, whereabout is said surplus place?
[21:05:07] <Tom_L> central US
[21:05:21] <N1njaneer> More specifically?
[21:05:25] <Tom_L> Ks
[21:05:48] <N1njaneer> Oh right, forgot you are in that area. Good stuff.
[21:05:48] <Tom_L> http://www.yardstore.com/?gclid=CJK0zffPvMECFQqTaQodor8AhQ
[21:06:03] <Tom_L> their metal isn't on there i don't think
[21:06:19] <N1njaneer> We have an aluminum extruder in Canada who hasn't come up with a problem yet pushing some tricky items for us that others simply didn't want to touch.
[21:06:53] <Tom_L> i'd like to see that process sometime
[21:07:05] <N1njaneer> There are some good YouTube videos of aluminum extrusion.
[21:07:26] <N1njaneer> At some point I'd like to drive up to shoot video of our actual stuff being pressed.
[21:07:26] <Tom_L> it's always better when you're there though
[23:34:14] <Casper> has anyone here used the qtouch lib from atmel?
