Back
[03:09:11] <MarkX> hi, i have a question that might be a little off topic but i was hoping someone might be able to help me with it
[03:09:28] <MarkX> i have a custom usb device that is currently using usbccgp.sys parent driver. it has 2 interfaces, one for audio using the generic audio driver and one for serial using usbser. i would like to use the winusb driver for the main device but when i assign that driver i lose my interfaces in device manager
[03:09:54] <MarkX> do i have to make a custom INF file that states WinUSB for the main device, then generic audio for the audio interface and usbser for the serial interface? can a device even have winusb as the parent driver with 2 other drivers for its interfaces?
[03:11:15] <carabia> MarkX: are you the assembler guy?
[03:11:39] <MarkX> carabia, i can work with it, i wouldn't say i'm super proficient though
[03:12:20] <MarkX> oh wait sorry i misread
[03:12:27] <MarkX> "the" assembler guy?
[03:13:09] <carabia> You were writing assembler stuff for AVR, am I correct?
[03:13:17] <MarkX> nope
[03:13:20] <carabia> Oh.
[03:13:46] <carabia> So, you're not the guy who got busted driving with pot?
[03:13:55] <MarkX> nope :S
[03:14:05] <Jartza> wut
[03:14:05] <carabia> I am disappoint.
[03:14:20] <carabia> ...Totally my bad.
[03:14:27] <MarkX> hahah all good
[03:14:38] <MarkX> are you on the lookout for that guy for a specific reason?
[03:14:40] <Jartza> I do write avr asm, but I don't know anything about getting busted driving with pot
[03:14:46] <MarkX> or he's just your idol or something?
[03:15:01] <MarkX> evening Jartza
[03:15:04] <MarkX> :)
[03:15:09] <Jartza> morning :)
[03:17:42] <carabia> MarkX: it's just that he got busted by the cops, and the bust was "unconstitutional" according to him, so he was planning on a lawsuit against the state for damages to his... alleged career and ego I guess. The ask was in the order of USD tens of millions. I just wanted to check how that worked out for him... :)
[03:18:09] <MarkX> HA
[03:18:20] <MarkX> not me :)
[03:18:29] * MarkX eyes Jartza
[03:18:29] <carabia> I swear the nick was Mark something :)
[03:18:43] <carabia> mark4 I guess. I figured markx is close 'nuf
[03:19:01] <MarkX> heh
[03:19:20] <MarkX> carabia, do you by any chance know about windows usb driver development?
[03:19:40] <carabia> fortunately no
[03:19:44] <carabia> [sic]
[03:19:44] <MarkX> haha
[03:22:20] <Jartza> I've been lucky and avoided windows quite long :)
[03:23:04] <Jartza> and no, I don't do pot either
[03:25:06] <MarkX> haha "do pot"
[03:25:12] <MarkX> it checks out, Jartza doesn't partake
[03:28:22] <Jartza> asm is enough for me :)
[03:28:27] <Jartza> you can mess your head with it too
[03:41:37] <Jartza> ...or make the avr do VGA
[03:41:44] <Jartza> what could be next... HDMI? :D
[03:57:28] <Skippy_421> good mornin' fellaws
[04:54:50] <carabia> Skippy_421: I see you've added some more numbers.
[04:55:15] <carabia> Jartza: if you pull off a legit hdmi interface off of an avr I'll consider tipping my hat
[04:58:46] <cehteh> i think that should be really impossible, 8 pins at over 100mhz at least
[05:04:22] <cehteh> mhm that saied, with some external IC handling the hdmi stuff, external ram for framebuffer, that could be nice, bit hey that external IC can be called "Raspberry PI zero" :D
[05:05:33] <twnqx> reminds me of my quest for a "cheap and simple PCIe VGA chip"
[05:05:43] <twnqx> didn't find many.
[05:06:07] <cehteh> there are pcie->hdmi chips :D
[05:06:15] <twnqx> huh?
[05:06:20] <twnqx> how does that work?
[05:06:44] <twnqx> i mean, sure, you can simple expose some random arm's frame buffer
[05:06:49] <cehteh> dunno just google if there are any chips and found some hints, no details
[05:06:54] <twnqx> (if that chip can handle being a pcie slave)
[05:07:04] <carabia> cehteh: oh, you can actually get that cheaper than a lot of avrs, some arm soc at 700mhz :)
[05:07:11] <cehteh> https://www.maximintegrated.com/en/app-notes/index.mvp/id/4252
[05:08:58] <cehteh> ah needs more stuff
[05:09:30] <cehteh> well i doubt you can make it simpler and more economic than using a rpi0 .. but the zero is so hard to get
[05:10:07] <twnqx> there are basically 3 options
[05:10:10] <twnqx> http://www.siliconmotion.com/EW_Pages/Embedded%20Graphics_v1.html
[05:10:21] <twnqx> http://www.aspeedtech.com/products.php?fPath=20&rId=377 and friends
[05:10:34] <twnqx> and asking matrox
[05:11:07] <twnqx> aspeed is analog only, which is pretty yuck in these days
[05:11:50] <twnqx> the sm750 can at least do single display (or clone mode) dvi + vga at full hd
[05:13:25] <carabia> Even checked hotmcu, they only have a single ca-8 off the higher end arms :(
[05:13:57] <carabia> it has a lot of memory onboard though... but it comes bundled with a screen and costs $220 so wtf :(
[05:17:38] <Jartza> carabia: well yeah, so far I've only gotten 32x16 characters on VGA out from attiny85
[05:17:57] <Jartza> with UART@9600 and ANSI-escape (subset) support :)
[05:18:09] <Jartza> outputting 640x480@60Hz industru standard VGA
[05:23:23] <Skippy_421> carabia what do you mean? my nick?
[05:24:08] <Skippy_421> also th cehteh for the stackexchange post, the answer someone gave you was exactley what I was looking for
[05:24:37] <cehteh> wasnt my query
[05:25:05] <Skippy_421> oh ok, then I can't remember who it was, but someone here :D thx
[05:25:53] <Jartza> carabia: or well, don't know about that "only"-part, because 32x16 = 512 characters, and the chip has 512 bytes of ram, so... :)
[05:46:07] <LeoNerd> Huh. Something I didn't realise. It appears like the arduino bootloader at least /does/ have the chip signature bytes hardcoded into it, instead of looking them up from the chip itself using LPM
[05:46:21] <LeoNerd> https://github.com/arduino/Arduino/blob/master/hardware/arduino/avr/bootloaders/atmega/ATmegaBOOT_168.c#L743
[05:46:48] <LeoNerd> So while I can take, say, a 328P bootloader and burn it onto a 328PB chip, it will now claim to just be a 328P over the serial wire
[05:49:20] <cehteh> lol
[05:49:33] <cehteh> well should be fixable, fork it :D
[05:49:51] <cehteh> or use some other bootloader
[05:49:53] <LeoNerd> I may have to. It'll come out -slightly- bigger though.. I wonder how much by
[05:50:13] <LeoNerd> Well, if I'm selling these boards as being similar to a Nano, I figure I ought to put an arduino-shaped bootloader on it
[05:50:17] <cehteh> since it is arduino i bet you can optimize it by tons at other places easily :D
[05:50:21] <LeoNerd> Anyone with a real ISP will just splat over it /anyway/
[05:51:18] <cehteh> does arduino share bootloader code (clib) with arduino programy by a linker script? -- no
[05:51:35] <cehteh> i wanted to look into that for the micronucleus bootloader
[05:51:53] <cehteh> aka shared v-usb code on tiny85
[05:52:00] <cehteh> would save a lot flash
[05:53:06] <LeoNerd> I don't generally use a bootloader; I like having the real ISP accessible
[05:54:06] <cehteh> i like bootloaders on usb enabled chips
[05:54:27] <LeoNerd> Oh true, that's different I guess
[05:54:34] <cehteh> btw did you stop the nonsense to reset the chip on serial connect? make at least a solder bridge?
[05:54:47] <LeoNerd> 32Ki of flash, so less pressure to save space
[05:54:58] <cehteh> usb enabled .. includes nano or any other with usb brindge
[05:55:11] <cehteh> so 328p/b too
[05:55:16] <LeoNerd> The DTR# -> RESET# connection is done by a 10nF cap; you can easily desolder that if it's annoying
[05:55:38] <cehteh> solder bridge would be nicer
[05:55:47] <cehteh> but ok
[05:56:05] <LeoNerd> I considered some other solder bridge configurations for the other handshaking lines too
[05:56:16] <LeoNerd> I might do it at some point, in a redesign. Probably drop most of the LEDs too
[05:56:50] <carabia> Oh Skippy_421, reminds me, how's your project coming along?
[05:56:52] <cehteh> lots of #ifdefs
[05:57:19] <cehteh> maybe you just add the 328pb the simple way add another #define for the sigs
[05:57:32] <LeoNerd> Hrm?
[05:57:39] <cehteh> see line 200
[05:57:45] <LeoNerd> They use SIG1 / SIG2 / SIG3 constants. which are defined in the #include <avr/io.h>
[05:57:58] <LeoNerd> Oh..
[05:58:10] <LeoNerd> Bah why do they do that?? They /should/ be using the io.h file which knows them
[05:58:21] <cehteh> its arduino
[05:59:13] <cehteh> hey you'll get bonus points when your bootloader can use any of the 2 uarts .. or even both
[05:59:45] <cehteh> not necessary, but could be handy sometimes
[05:59:57] <LeoNerd> I don't really see a point there. I'm only building this loader so I can pre-burn it onto the breakout boards I sell
[06:00:13] <cehteh> yeah, likely not worth the efforts
[06:00:15] <LeoNerd> The only reason there is because the onboard USB-CDC bridge is already attached to USART0. There'd be no real gain in making it support 1 as well
[06:00:33] <cehteh> but some people my like to have one production and one maintenacne port
[06:01:18] <LeoNerd> Eh... I figure such people will be making their own boards
[06:01:25] <cehteh> prolly
[06:01:31] <LeoNerd> This is /mostly/ for Arduino-types, or other early experimenters on breadboard
[06:02:00] <cehteh> but i already saied, sometime for small production runs its completely viable to solder a arduino on your product
[06:02:03] <LeoNerd> Oh speaking of, I think my next project will be on a mega128A. Huge great 64 pin monster
[06:02:07] <LeoNerd> 53 GPIOs :)
[06:02:20] <LeoNerd> I need 24 of those to be driving a R/G 8x8 LED matrix
[06:02:23] <cehteh> huw much pwm/timers?
[06:02:37] <cehteh> i never needed that much gpio
[06:02:38] <LeoNerd> Dunno.. some. Not a lot. I don't need those though.
[06:03:06] <LeoNerd> I would be happy with a much smaller chip and three 595s to talk to the LEDs, or other such hacks. But turns out this chip is cheaper/smaller/less power than all of the other ideas
[06:03:14] <cehteh> but having much more 'equal' pwm would be handy .. like having 6 or more compare match units on one timer
[06:03:23] <LeoNerd> YES
[06:03:24] <LeoNerd> Manalive yes
[06:03:50] <LeoNerd> I would love to have "some", maybe one or two 8 and 16-bit counters, and *separate* compare units that can be switched between them
[06:04:10] <cehteh> initializing 3 timers (2 8 bit and one 16 bit) just to get 6 pwm's suck
[06:04:18] <cehteh> or that
[06:04:32] <LeoNerd> Yeah it's rare I need more than two timers
[06:04:40] <cehteh> ack
[06:04:45] <LeoNerd> A 16bit ticker to give me a system clock, and another (could be 8bit) to do PWMing
[06:04:59] <cehteh> or opposite if you need higher res PWM
[06:05:13] <LeoNerd> Quite often I run out of PWM IO pins but I don't think I've ever been short a timer/counter unit
[06:05:48] <LeoNerd> Eh, I've only ever done PWM for LEDs.. 8bit is fine
[06:06:05] <LeoNerd> If I needed higher res it's probably because I was driving servos or somesuch; I'd use a dedicated servo PWM chip for that
[06:06:15] <cehteh> 16 bit pwm is often nicer, and 16bit for timer isnt enough anyway you need to extend it in a overflow isr already
[06:06:16] <LeoNerd> I have a nice 16channel I²C one on a board I should play with
[06:06:42] <LeoNerd> For a system timer I like to have a fast-ish "decimal" rate, like close to 5msec, and then keep my own tickers in software
[06:06:46] <carabia> Serial. Yuck.
[06:06:53] <cehteh> esp for pwm LEDs 16bit would be nicer because you can de-logarithmize better
[06:07:02] <LeoNerd> Oh.. hm.. I suppose that's a point
[06:07:10] <LeoNerd> Low-levels look terrible otherwise
[06:07:13] <cehteh> then you get real 256 brightness levels
[06:07:30] <LeoNerd> mmm
[06:07:54] <LeoNerd> Also, I think "de-logarithmize" can be better expressed "exponentiate" ;)
[06:07:58] <cehteh> timers counting logarithmically would be fun :D
[06:08:16] <LeoNerd> Well, if that was in base2 that would be trivially easy in hardware
[06:08:17] <cehteh> whatever
[06:08:22] <cehteh> yes
[06:08:30] <LeoNerd> Tell it /which/ bit of the counter it should flip on :)
[06:08:30] <cehteh> but 16bit pwm makes sense
[06:09:15] <cehteh> but for 16bit you need fast counters .. PLL on atmega?
[06:09:21] <cehteh> hey the tinys have that
[06:10:34] <LeoNerd> The tiny85 has a fast PLL, but then doesn't have a 16bit timer to go with it
[06:10:44] <LeoNerd> A decision I utterly fail to comprehend
[06:10:49] <cehteh> ah moment mega is 16mhz not 8 .. if i recall correctly you get around 300hz pwm freq with prescale 1 on 16bit
[06:11:06] <cehteh> thats enough for leds, but not enough for other things
[06:11:28] <LeoNerd> I did a bunch of stage lighting on PWM'ed LEDs. I considered tuning the PWM frequency to 25.1Hz just to fuck with the video team ;)
[06:11:43] <cehteh> lool
[06:12:06] <Jartza> how about 27.5Hz
[06:12:20] <cehteh> got strobed :D
[06:12:31] <LeoNerd> In the end they had the last laugh.. my blue LEDs kicked out a lot more UV light than we were expecting. Looked fine to human eyes, but the cameras picked it all up
[06:12:51] <LeoNerd> Funnily eonugh, they weren't expecting to need sunny-outdoor-spec UV filters on an indoor nighttime shoot ;)
[06:13:04] <cehteh> :)
[06:13:41] <LeoNerd> Things you learn, eh? :)
[06:13:58] <cehteh> i have a computer case with 2 bright blue leds here ... couldnt stand it so i covered them woth neon postit notes, which gives really nice colors
[06:14:03] <cehteh> until it bleached away
[06:14:16] <LeoNerd> Mmm.. yes that UV in action
[06:18:40] <Skippy_421> thx carabia my project is going pretty well, project management ftw
[06:20:39] <Skippy_421> I am currently wondering why sometimes I get a 3 in my i2c frame after the read byte --> < start | addr | read | 3 | .... | stop >
[06:20:58] <Skippy_421> I either had 0 for no data or 1 for data, but I just had a 3 there, and after that it froze
[06:21:38] <Skippy_421> currently I don't ahve that error again, but if I get it again, I have to check if this contains some avr i2c codes or some error msg from the slave
[06:25:20] <Lambda_Aurigae> or improperly terminated i2c bus or noise on the line.
[06:34:33] <Jartza> i2c bus termination, aka black magic voodoo :)
[06:34:41] <Skippy_421> I was able to reproduce the error, it seems like this happens when I read too fast
[06:35:36] <Skippy_421> when I read every 1ms it crashes at some point, with 10ms everything is fine
[06:35:39] <Skippy_421> at least for the moment
[06:36:11] <Lambda_Aurigae> so, the device at the far end can't go as fast as you want.
[06:36:24] <Skippy_421> I guess so
[06:37:50] <Jartza> does it stretch the clock?
[06:38:18] <Jartza> I have i2c lcd displays that don't have any kind of busy-signal, but they do stretch the clock when going too fast :)
[06:44:24] <Skippy_421> no, the clock stays the same
[06:44:40] <Skippy_421> but it's ok for me, 10ms is ok
[07:05:27] <Skippy_421> I cannot break a while(Wire.available() && i < len){ ... } loop
[07:07:42] <daey> can you post your Wire struct
[07:07:49] <MarkX> anyone have any idea about the question i posted earlier?
[07:09:11] <Skippy_421> http://pastebin.com/bpRzwQmA
[07:09:14] <Damme> hey guys, I am having troube reading out a mega48pa .. tried different versions of avrdude and now from git (avrdude version 6.1-svn-20130917 <- is there newer version??, or alternative to avr-dude?) I get this error:
http://pastebin.com/FiE6gmGz
[07:09:30] <Damme> fuses sais downloading is allowed
[07:09:43] <Damme> avrdude crashes on both linux and windows
[07:11:19] <LeoNerd> avrdude crashing sounds like a bad sign
[07:11:38] <Damme> ops, paste is pasted now, I didn't press 'I'm no robot'
[07:16:24] <daey> Skippy_421: i mean the struct and the constructor. out of curiousity
[07:17:10] <Damme> I think my usbtiny might be to blame.. en Attiny44A .. wonder if there are different firmwares
[07:17:54] <zaninime> Hi everybody! I'm struggling with some strange behaviour of avr-gcc. As soon as I create a function in addition to main, the firmware hangs/doesn't boot
[07:18:12] <zaninime> Any advice? :)
[07:20:37] <zaninime> The problem is related to function which body is declared before main
[07:20:48] <daey> http://paste.debian.net/788633/ oh baby yeah
[07:21:23] <Skippy_421> I don't have a Wire struct
[07:21:41] <daey> how did you get the funktion names with a dot inbetween? :o
[07:21:54] <Skippy_421> I use the Arduino Wire.h and I just wrapped a simple class called I2C around it
[07:22:25] <Skippy_421> Arduino Wire is a static class
[07:22:31] <daey> c has classes o.O
[07:22:35] <Skippy_421> c++
[07:22:37] <daey> oh
[07:22:41] <Skippy_421> it's arduino ^^
[07:22:44] <daey> nvm i dont know c++
[07:22:47] <daey> nor arduino
[07:22:48] <Skippy_421> at least partially
[07:23:30] <carabia> I die a little inside when I see c++
[07:24:00] <daey> i die a little each time i read the word arduino :(
[07:24:05] <Skippy_421> lol ^^
[07:24:19] <LeoNerd> It's spelled "arduino", it's pronounced "duplo"
[07:24:36] <carabia> börnö stringsöup really has done a great job evolving c into clusterfuck
[07:24:40] <daey> dont get me wrong, i think the platform itself is nice.. just not the software debacle
[07:25:22] <carabia> daey: what makes it anything other than an avr breakout is the _software_
[07:25:33] <Skippy_421> my university teached me and made me realize nothing is good or bad. Everything has its advantages and disadvantages and it always depends on what the goal is
[07:25:34] <daey> y :p
[07:25:35] <LeoNerd> Yah; a lot of the hardware is fine enough... though I'd prefer if they labeled the pins properly
[07:26:05] <daey> Skippy_421: the main problem with arduino is if you google for a lib. you get arduino libs. tons of them. one shittier than the next
[07:26:13] <carabia> Skippy_421: your university is a bunch of retards then. Don't c++ on a god damn avr. End of discussion
[07:26:38] <Skippy_421> I never said my university told me to do that ^^
[07:26:40] <daey> Skippy_421: i already got punshed for my struct oop attempt here :P
[07:27:28] <carabia> what you just said there is they taught you that. One way or another it implies that is exactly what they "told you to do"
[07:27:43] <daey> that being said, one of our external devs uses c++ for PICs. He gets the job done as well...
[07:27:50] <carabia> is = it's
[07:28:19] <carabia> daey: if we're talking pic32s, yes, you can have more overhead
[07:28:26] <Skippy_421> it's ok, every computer science subject has its own melt-down topics ^^ same for linux, windows, java, etc.
[07:29:05] <carabia> avrs are dead slow and then some (with arduino) compared to pic32s
[07:29:43] <daey> carabia: some are pic18s
[07:29:48] <Skippy_421> I didn'T even use avrs in my university, I just learned different stuff and through that, that everything has its advantages and disadvantages and what you choose to use depends on a lot of factors, but nothing is bad or good, it always depends
[07:29:55] <daey> but it goes up to 32 and ARMs
[07:32:10] <daey> i wonder whats going to happen to atmel after the takeover
[07:33:16] <carabia> Skippy_421: well the distinction is clear between using c and arduino. It's not language vs use-case, it's ease-of-use vs speed, period
[07:33:19] <aczid> I think the atmel brand will remain
[07:33:54] <carabia> and the convenience factor largely goes away once you actually get yourself around to learn c
[07:34:37] <_ami_> in case of single relay module, I don't have to worry about current drawn from mcu to signal port of relay.
[07:34:44] <carabia> and for very timing-critical things you're even better off learning assembly
[07:34:52] <_ami_> Is that correct?
[07:34:54] <carabia> _ami_: really depends on the relay
[07:35:24] <carabia> datasheet?
[07:35:41] <_ami_> https://www.parallax.com/product/27115
[07:35:52] <_ami_> I have this one.
[07:36:16] <carabia> oh it's a board
[07:38:01] <_ami_> I checked the DS, it says 85 mA but it's relay supply current. so I think signal port only needs 3.3v to 5v input voltage.
[07:38:15] <carabia> _ami_: I'm quite sure you can drive that straight from an mcu pin, looks like transistor right there
[07:38:38] <_ami_> ah. transistor is needed then?
[07:38:43] <carabia> well yes
[07:39:17] <carabia> the supply current determined is most likely through it
[07:39:43] <carabia> you don't want to drive it straight with an mcu.
[07:40:17] <_ami_> Yeah. that's why I asked here.:)
[07:41:33] <carabia> Yeah. If you want to drive high loads you can always use a solid-state relay. That won't need a transistor to drive it
[07:41:43] <daey> aczid: im sure it will. i jsut hope the tool chains all grow together. hopefully pic will be integrated into avr and not the other way around *shudder*
[07:41:50] <carabia> YMMV, still check datasheets.
[07:42:00] <_ami_> carabia, so mcu pin attached with base of npn transitor and 1k ohms resistor along with 10k ohms pull down
[07:42:25] <_ami_> and collector is powered to 5 v and emitter is connected to signal port of relay?
[07:42:31] <_ami_> Is that right?
[07:42:36] <aczid> daey: yeah, there is still quite a bit of uncertainty
[07:43:18] <daey> tbh. i dont really understand why they bought atmel. it doesnt really seem to add anything to their portfolio
[07:43:53] <carabia> _ami_: no
[07:44:02] <_ami_> hrm
[07:44:07] <carabia> try again :)
[07:44:31] <_ami_> thinking.. 😇
[07:44:50] <carabia> also you want the flyback diode so you won't blow up the tranny
[07:44:56] <carabia> when the coil discharges
[07:45:29] <_ami_> Yeah diodes are required .. how can I forget it.
[07:45:37] <carabia> just one.
[07:45:49] <_ami_> yah 1 only.
[07:45:56] <_ami_> rest is okay?
[07:46:05] <carabia> no.
[07:46:57] <daey> carabia: theres a talk in which linus claims he assembler code that is generated by the c code when he looks at it :P
[07:47:31] <daey> he knows the assembler code*
[07:48:10] <carabia> daey: well that depends on the platform
[07:48:21] <carabia> and. umm. linus says a lot of other stupid shit too.
[07:49:12] <daey> x86 obviously
[07:50:04] <daey> meh havent seen him talk much shit. sure he behaves like a playa...but then again he kindaaa is one
[07:51:17] <_ami_> carabia, damn
[07:51:25] <_ami_> umm...
[07:51:27] <carabia> daey: have you seen how long the x86 instruction list is?
[07:51:30] <daey> Skippy_421: btw. theres #arduino with 500+ users
[07:51:57] <daey> carabia: im not the one claiming it :D
[07:53:24] <carabia> with all of the god damn extensions. And the code generated varies on the compiler. I guess here we'd be talking of gcc. Or clang. Dunno. With optimizations, ..., hurr durr
[07:54:03] <_ami_> carabia, hints please
[07:54:22] <carabia> _ami_: just low side switch it
[07:55:01] <daey> i.e. Power rail - coil - transistor - gnd
[07:55:50] <_ami_> carabia, I did n't get.it.
[07:56:04] <daey> then google low side switch >.>
[07:56:14] <_ami_> okay
[07:56:32] <carabia> ...and the diode parallel with the coil. And remember to flip it
[07:57:35] <carabia> i need me some grub.
[08:02:37] <_ami_> I will try this when I reach home. Thanks for the help.
[08:10:38] <carabia> _ami_: if you wanted to make things really hard, you COULD use it as a high-side switch as you suggested, but it would get really wonky cause npns aren't really meant to do that.
[08:11:57] <_ami_> Yeah. I googled a bit. For that pnps are right choice.
[08:12:16] <carabia> yes.
[08:13:21] <_ami_> or you can say danglilton pair.. A npn with a pnp actually.
[08:13:53] <carabia> that's not a darlington... it had a different name.
[08:14:08] <carabia> anyway. it's not necessary
[08:14:33] <carabia> if we want to go there you could use the mcu to make a charge pump to drive the transistor \:D/
[08:15:07] <_ami_> Oh okay
[08:15:42] <_ami_> I thought I knew transistors .. but it's not.
[08:15:59] <_ami_> Will experiment on it today.
[08:20:32] <carabia> Actually, you could use a transistor to switch off a say, max232 and use its charge pump, to drive the transistor for the coil. Now there's an idea.
[08:22:18] <carabia> rs232 chips are awesome for their charge pumps to generate the proper voltages. You get + and - 2 * rail voltage
[08:22:49] <carabia> obviously can't draw much current, the voltage drops off fairly fast
[08:48:30] <LeoNerd> MAX660
[08:48:46] <LeoNerd> (is just that part; the doubler and inverter)
[09:02:20] <carabia> LeoNerd: yeah but you're more likely to have 232s around
[09:03:49] <carabia> though i guess nowadays everything ships with a uart usb bridge
[09:05:18] <Lambda_Aurigae> bah...my copiers are shipping with a 3.3V TTL level uart port.
[09:06:42] <LeoNerd> carabia: I wouldn't consdier either kind of part to be "jellybean"
[09:06:45] <LeoNerd> (I love that word lately ;) )
[09:07:29] <carabia> well, not anymore at least...
[09:08:15] <LeoNerd> I got myself a MAX1044 yesterday. Have to play with that this weekend I think
[09:08:40] <LeoNerd> It's supposed to be a voltage inverter (e.g. make -5V out of +5V), but there's an interesting application note to create a doubler and inverter (of the input)
[09:08:52] <LeoNerd> So I want to use it to make +10V and -5V out of a single +5V supply, for an opamp
[09:09:23] <LeoNerd> I don't need symmetric rails, but I would like the +10V headroom over the top, and a -slightly- negative bottom rail to let the opamp output go to ground
[09:09:48] <LeoNerd> MAX1044 lets you do it with a smaller chip and fewer caps
[09:10:17] <carabia> and the output characteristics seem to be nice with a quick glance, it promises 10 mA with only .5 V drop
[09:10:35] <carabia> though there's probably some gotchas in the datasheet hidden somewhere :)
[09:10:53] <LeoNerd> Yah. I don't require huge output stability - I'm driving a single opamp, which wants maybe 2mA quiescent on top of my feedback network. tiny
[09:11:10] <LeoNerd> Plus the opamp has quite good power supply rejection ratio, so the output ripple doesn't bother me
[09:11:24] <LeoNerd> I'm hoping to make a much-improved version of my oscilloscope current probe
[09:19:42] <carabia> a jelly bean probe?
[09:19:46] <carabia> What flavor?
[09:21:27] <LeoNerd> Perhaps it's one of those trolling flavours, like earwax :)
[10:33:43] <eszett> hi
[10:36:59] <carabia> Days off are nice and all, but I was gonna be ultra-productive today and here I am, it's 6 PM and I haven't gotten around to get my pants on yet.
[10:37:16] <carabia> ultra-productive in terms of non-work related stuff. Oh well.
[12:02:06] <daey> looking at the buspirate...it seems to be the perfect uC programmer.. am i missing something?
[12:02:18] <daey> it seems to support avr, arm, pic ...
[12:04:48] <LeoNerd> It's a bit slow. Well, very slow
[12:05:02] <LeoNerd> I find it a little unreliable as well.. or maybe I just have a bad unit
[12:05:12] <daey> i see
[12:09:10] <_ami_> daey: how does it support all uCs? jumper settings?
[12:09:19] * _ami_ googles abt it.
[12:10:20] <Lambda_Aurigae> doesn't do hv programming for avr or pic.
[12:10:31] <Lambda_Aurigae> but if your pic can be programmed low voltage ICSP mode then go for it.
[12:10:52] <Lambda_Aurigae> it would be needing software on the computer to support it...it is basically a bitbanged interface for programming.
[12:11:09] <daey> Lambda_Aurigae: most native programmers dont do hv...
[12:11:15] <Lambda_Aurigae> because it's bitbanged, it is, as LeoNerd said, very slow...and not real reliable.
[12:11:55] <LeoNerd> I sell an adapter to let the bus pirate do HVSP ;)
[12:12:05] <LeoNerd> Only for the 8 or 14-pin ATtiny chips though
[13:58:36] <Damme> hey guys, I am having troube reading out a mega48pa .. tried different versions of avrdude and now from git (avrdude version 6.1-svn-20130917 <- is there newer version??, or alternative to avr-dude?) I get this error:
http://pastebin.com/FiE6gmGz
[14:01:12] <Lambda_Aurigae> looks like 6.3 is the newest version.
[14:01:22] <Lambda_Aurigae> 6.1 should work just fine I would think.
[14:01:30] <Lambda_Aurigae> what hardware programmer do you havE?
[14:01:49] <Lambda_Aurigae> do you have a clock source on the clock pin? or a crystal?
[14:02:06] <Lambda_Aurigae> depending on how the fuses are set, it might need either a crystal or external clock source to work.
[14:02:19] <Lambda_Aurigae> also, is this one you programmed yourself or from a commercial product?
[14:02:33] <Lambda_Aurigae> often times commercial products have lock bits set that will keep you from reading the chip.
[14:02:38] <Lambda_Aurigae> can you read just the fuses?
[14:03:31] <Damme> Lambda_Aurigae, No clock, usbtinyisp called ispmap or something too
[14:03:36] <Damme> Fuses are fine
[14:04:02] <Lambda_Aurigae> so, you can actually read the fuses with that programmer?
[14:04:09] <Damme> h01, e-df, le2
[14:04:45] <Lambda_Aurigae> that's what the safemode reports...
[14:04:54] <Damme> hmm
[14:04:57] <Lambda_Aurigae> have you tried actually reading them ?
[14:05:16] <Damme> what about atmega328 as isp programmer ?
[14:05:23] <Lambda_Aurigae> what about it?
[14:05:39] <Lambda_Aurigae> I'm sure there is code out there to do that.
[14:05:51] <Damme> if its a difference between programmers
[14:06:18] <Lambda_Aurigae> cheap usbtiny programmers are worth about half what you pay for them.
[14:06:19] <Damme> Lambda_Aurigae, yes, I have code for that too. I'm just curius why it would read fuses wrongly
[14:06:40] <Lambda_Aurigae> anything v-usb based I would suspect as a problem myself.
[14:06:49] <Lambda_Aurigae> unless verified that it works elsewise.
[14:07:13] <Damme> Lambda_Aurigae, aah ok, I know it works with mega328 .. I bricked one the other day xP
[14:07:24] <Damme> but this 48pa seams wierd..
[14:08:17] <Damme> Maybe there are software protect then.. Its the controller to a voltcraft power supply.. I was hoping I could modify the existing code instead of writing from scratch :P
[14:08:31] <Lambda_Aurigae> so it's a commercial product?
[14:08:48] <Lambda_Aurigae> they probably have the memory lockbits set so you can't read it.
[14:09:18] <bss36504> Does anyone know where you can buy single quantities of a pre-programmed EDBG chip? It would be cool to be able to build boards with an embedded programmer/debugger
[14:09:32] <bss36504> I found this on digikey:
http://www.digikey.com/product-search/en?keywords=atmel%20edbg
[14:09:42] <bss36504> Not about to buy 4000 of them though :)
[14:10:28] <Damme> Lambda_Aurigae, probobly rather easy to reverse though... I might just do that a rainy day ^^
[14:11:06] <Damme> What happens if you re-program fuses with hi-voltage programmer and clear read-protect ?
[14:11:38] <Damme> does the code dissapear
[14:12:00] <Lambda_Aurigae> yup.
[14:12:10] <Lambda_Aurigae> to erase the lock bits you have to erase the entire flash.
[14:12:16] <Damme> ok
[14:12:21] <Damme> Too bad xP
[14:12:32] <Lambda_Aurigae> if you had some NSA level gear to hork into it and dig deep you might be able to recover it.
[14:12:46] <Lambda_Aurigae> but if you have to ask that kind of question I'm betting you don't have gear of that level.
[14:12:48] <Damme> yeah, decap and read bits from internal flash ^^
[14:13:19] <bss36504> Oh yeah, just a simple decap and read of a flash, easy as pie
[14:13:23] <Damme> I just thought you always could read the fuses.. and they looked okey thats why I asked
[14:13:33] <Lambda_Aurigae> then, if you do get the data out, it's binary file basically...you have to convert back to assembly...
[14:13:39] <Lambda_Aurigae> then reverse engineer the assembly.
[14:13:49] <Lambda_Aurigae> which is not for the faint of heart...
[14:14:06] <Damme> Yeah I know .. :) I'm reversing a robotic mower, uses arm cpu... and had jtag so it was easy to read out code :)
[14:14:24] <Damme> (ida does most of the work there though)