Back
[00:37:26] <veek> the whole harvard architecture of the avr - how does that work? one bus for operands(bid), one for opcodes (bidirectional). What about the other two?
[00:43:06] <veek> http://storage8.static.itmages.com/i/16/0610/h_1465535771_9345089_c9fbeb90a9.png
[01:54:49] <veek> anyone awake? in an avr, when i do LDS R20, 0x90; this will get converted into byte code which is: E 0x9 R20-int and 0x0; however how is this stored - the whole code/data split???? does LDS/E goto code segment and the 0x90 go to data? what about Rwhatever?
[02:07:19] <Casper> ??
[02:08:23] <veek> Casper, any idea how arm stores code
[02:08:42] <veek> that effing book makes me want to scream in frustration
[02:08:57] <Casper> nope, your pic show an illogical way, so weird
[02:09:10] <veek> hang on let me upload
[02:09:57] <veek> Casper,
http://storage1.static.itmages.com/i/16/0610/h_1465541002_5619535_533ba09e9c.png
[02:10:52] <veek> LDS R20, 0x90
[02:11:27] <veek> that gets converted to bytes and gets stored. But in the ARM there are separate areas for code and data
[02:11:34] <veek> err AVR
[02:12:00] <veek> since LDS is code and 0x90 is data - they should be split?
[02:14:25] <Casper> that's what I understand
[02:15:08] <veek> okay so how would the data be loaded PC latches address of LDS
[02:15:44] <veek> and loads that opcode from code bus
[02:16:00] <veek> but how does it locate 0x90 from the data area?
[02:17:18] <Casper> it take the word, split it, the "E" cause the cpu to put 0x90 in the address bus and cause a read, that is stored at the R20 ...
[02:18:06] <veek> yes but how does the cpu find 0x90.. that's stored somewhere else..
[02:18:07] <Casper> I understand how it work, but not in detail enought to describe?
[02:18:18] <Casper> 2 bytes from ram
[02:18:21] <Casper> err
[02:18:23] <Casper> flash
[02:18:43] <Casper> split into nibles, reassemble the data, and do magic
[02:19:12] <veek> PC ONLY points to the LDS opcode..
[06:28:25] <Jartza> https://drive.google.com/file/d/0B2dTzW9TMeBxcW5lTVpvcHVXenM/view
[06:44:02] <LeoNerd> Does anyone know of a handy parametric comparison chart of current ATmega chips? Iwant a list of all the chips with >= 2 UARTs and >= 1KiB RAM
[06:47:22] <LeoNerd> Hrm.. though I think the smallest will still be a 28/32 pin thing, which is still bigger than the 20pin ATtiny1634
[06:54:00] <Jartza> LeoNerd: how about atmel MCU selector?
[06:54:11] <Jartza> http://www.atmel.com/selector.html
[06:54:23] <Jartza> oops
[06:54:27] <Jartza> http://www.atmel.com/selector.html#(actives:!(),data:(area:'',category:'34864',pm:!(),view:list),sc:1)
[06:55:19] <LeoNerd> Aha.. that looks to be just the thing. :) Thanks
[06:55:41] <Jartza> http://bit.ly/22XKpQB
[06:55:44] <LeoNerd> .. iohfuck this is stupid
[06:55:47] <Jartza> that might work petter
[06:55:53] <LeoNerd> It spins local javascript for about 8 seconds, every time I toggle a tickybox
[06:55:56] <Jartza> uh. better.
[06:56:02] <LeoNerd> Ihave to turn *off* all the CPU types to turn on just 8bit AVR
[06:56:05] <Jartza> LeoNerd: indeed it does :)
[06:56:10] <Jartza> there's "clear all" button
[06:56:24] <Jartza> and for me it only spins like 0.1s
[06:56:47] <Jartza> but anyway, that's pretty handy still
[06:58:08] <LeoNerd> Ahyes, as suspected: the ATtiny1634 is *the only* tiny chip with >= 2 UARTs and >= 1KiB RAM, and thus the smallest at only 20 pins. Even the smallest ATmega is 32
[06:58:22] <LeoNerd> Annoyingly, the max speed of the ATtiny1634 is only 12MHz. Hrm.
[06:58:43] <LeoNerd> Still good enough for DMX though
[06:59:48] <LeoNerd> Hrm.. or maybe I bump up to the mega328PB
[06:59:57] <LeoNerd> Could be a fun excuse to try out that new shiny chip
[07:00:29] <Jartza> yeah. I've been planning to test it too
[07:00:53] <LeoNerd> It's annoying that you can't quite just take a regular Arduino Nano/etc.. board and rechip it
[07:01:13] <LeoNerd> Otherwise you risk destroying the PE0/PE1 pins, which sit where there used to be a second copy of VCC/GND
[07:02:05] <LeoNerd> Hrm.. I don't see any mega328PB stuff on tindie... I feel a gap in the market here ;)
[07:02:09] * LeoNerd opens up KiCad
[07:04:31] <LeoNerd> I reckon on doing it to a similar form factor as the Arduino Nano.. probably still with a USB-CDC bridge + USB socket
[07:04:34] <LeoNerd> Because everyone likes those
[07:05:49] <Jartza> they're handy with breadboard :)
[07:06:26] <LeoNerd> Mhm
[07:06:56] <LeoNerd> https://twitter.com/cpan_pevans/status/724252582877777920 <== my various breadboarded AVR chips
[07:07:20] <LeoNerd> Actually I've got two more since that pic - A tiny85 and a tiny841
[07:08:12] <LeoNerd> One of those 328s has a modification to it; I replaced the 16MHz resonator with a 14.7456, for doing common UART baud rates
[07:20:01] <Jartza> LeoNerd: ahh. I've been using 18.432MHz oscillators :)
[07:21:15] <LeoNerd> I don't think the 328P(B) is technically rated up to 20MHz, is it?
[07:21:37] <LeoNerd> Oh.. huh. it is
[07:21:56] <LeoNerd> Why do I do these at 14 then? I should totally use that 18
[07:22:37] <Jartza> I have no idea why you ues those :)
[07:22:52] <Jartza> maybe because the "standard" is 16MHz in arduinos (I never understood why)
[07:23:16] <Jartza> they could use 20MHz instead
[07:59:28] <tavish> hi, I have an atmega1280 and the watchdog doesn't seem to be working as expected. it does 'reset' the device, but it seem to execute the program, only a power-cycle makes it work again
[08:00:26] <tavish> all I'm doing to test it is wdt_enable(WDTO_8S); and then not pet the watchdog
[08:03:54] <Jartza> do you disable the watchdog on boot?
[08:04:17] <Jartza> maybe it goes into watchdog-boot-loop?
[08:04:38] <tavish> Jartza: i did do that, also it's not enabled in the fuses
[08:04:58] <tavish> so it shouldn't be setting the watchdog again on reboot
[08:05:44] <Jartza> watchdog reset doesn't disable watchdog automatically
[08:05:52] <Jartza> so, there's also nothing to disable it
[08:06:56] <Jartza> there is no need to "set it again" as it's never disabled.
[08:20:00] <skz81> ami, why testing the flag before setting here ?
https://github.com/amitesh-singh/amiduino/blob/master/avr/avr_programmming/pwm/powersleep/sleep.c#L17
[08:20:43] <skz81> hu, he's not here... Was scrolled up, please forgot my noises :)
[08:20:50] <skz81> and forgive
[08:22:26] <veek> skz81, bow low, only then shall you be forgiven
[08:23:15] <veek> skz81, not the volatile - bow even lower
[08:24:39] <veek> err s/not/note
[08:24:53] <skz81> veek, what ????
[08:25:17] <veek> dude volatile.. it's being set elsewhere through hardware
[08:25:29] <veek> that's an ISR so it makes sense
[08:25:49] <veek> volatile uint8_t f_timer = 0;
[08:26:29] <veek> not bad :p i'm of some use in here :)
[08:26:30] <skz81> okay, but, it's set in ISR, reset in main's while loop....
[08:27:16] <veek> and..
[08:27:30] <skz81> Don't see the point to make the test before setting the flag. In worst case, you set it twice in row
[08:28:24] <skz81> There is no exception handling mecanism anyway, so it just "optimize" the worst case (but doesn't lead to actual optim, actually !)
[08:28:54] <skz81> normal case : test/branch, set, worst case, just test/branch
[08:29:09] <veek> dude, eg: ./a.out; main() invoked and initializes flag then goes to sleep; hw int occurs; Ooo! ISR called; ISR does something else with flag
[08:29:09] <skz81> so just set this flag and exit ISR, non ?
[08:30:43] <veek> skz81, i can't tell you exactly what that's doing because i'm still learning avr :( like started daybefore so..
[08:31:17] <veek> not yet read timers
[08:32:05] <LeoNerd> Mmmmmm timers
[08:33:14] <skz81> veek, hum, good point, after all this flag is completly useless. I beleived _ami_ was driving timer2 from external (async) pins
[08:34:03] <skz81> Only ADC and IDLE sleep mode are working fine. Rest are not. After reading the DS, it seems async timer2 can bring back MCU from dead in PWR_SAVE mode.
https://github.com/amitesh-singh/amiduino/blob/master/avr/avr_programmming/pwm/powersleep/sleep.c
[08:34:15] <skz81> (quote from _ami_)
[08:41:20] <veek> grr disconected
[09:28:16] <rue_house> if you dont want the timer to wake you up, turn it off
[09:29:02] <_ami_> rue_house: pwr_down does that exactly. it shuts off timers.
[09:29:45] <rue_house> not if its waking it up
[09:32:10] <darsie> hey
[09:35:06] <darsie> Is an ATMega328P likely to work as a replacement for an ATMega168?
[09:35:27] <LeoNerd> They're prettymuch the same, yup. Just bigger flash/RAM. You might need software recompilation
[09:35:38] <LeoNerd> But pinwise and featurewise it'll be fine
[09:35:45] <darsie> k
[09:35:45] <darsie> thx
[09:43:57] <skz81> <Jartza> maybe because the "standard" is 16MHz in arduinos (I never understood why)
[09:44:22] <rue_house> hmm, not sure if you do have to recompile it...
[09:44:31] <skz81> <Jartza> they could use 20MHz instead >>> But maybe older models did have chip that had lower max freq
[09:45:31] <skz81> by a time it was atmega168, don't know it's max freq
[09:45:34] <skz81> its*
[09:46:44] * skz81 answers to himself : "hu, no... 20MHz, loser !"
[09:48:12] <skz81> "It's historical, the ATmega8 used in the original Arduino had a maximum clock speed of 16MHz." [
http://forum.arduino.cc/index.php?topic=27599.0 ]
[09:48:31] * skz81 gives himself a good point after all
[09:52:00] <rue_house> iirc all avrs overclock pretty good
[10:24:17] <tavish> Jartza: okay, I think i solved the issue. I had to disable watchdog before c/c++ runtime initialization, because I think I have too many things needing to be initialized in RAM.
[10:24:33] <tavish> so i am now doing this in the reset vector
[11:02:26] <LeoNerd> What's anyone's favourite USB-CDC bridge of choice these days? FTDI generally annoys me and they're superexpensive
[11:02:39] <LeoNerd> Prolific PL2303?
[11:03:34] <aandrew> I'm still a fan of FTDI. Nothing but trouble with Prolific IME. There's a chinese one that seems to be getting a lot of attention (can't remember the name)
[11:04:13] <LeoNerd> Silabs CP2102?
[11:04:43] <LeoNerd> CH340?
[11:05:10] <aandrew> CH340 that was it
[11:05:32] <LeoNerd> Olimex is the manuf.
[11:05:39] <aandrew> olimex?
[11:05:45] <aandrew> they don't make ASICs do they?
[11:06:23] <LeoNerd> Hrm.. doesn't look too bad. It needs an xtal though.
[11:07:53] <LeoNerd> Meh.. and not in KiCad's library
[11:07:59] <LeoNerd> Seems lately hardly anything I want is :/
[11:42:30] <LeoNerd> Ooooh oops. The CH340 doesn't have TX/RX LED support
[11:43:52] <LeoNerd> Hmm.. nor the PL2303
[11:44:19] <cehteh> most use leds directly on the TX/RX io port ... nice for optical snooping :D
[11:44:37] <LeoNerd> Yah, but... pulse stretching :(
[11:45:11] <LeoNerd> Youknow.. I might prefer the CP2102 for this - no need for an xtal
[11:45:12] <cehteh> ugly hack, but leds are fast
[11:45:39] <skz81> nice week-end folks !
[11:45:59] <cehteh> i remember back in time with 10mbit ethernet and cards which had a led directly connected to the dataline it was possible to snoop traffic on the leds too, with a bit signal processing
[11:46:11] <LeoNerd> Mm.. .and the 2102 says it natively supports 250kbaud, which is a thing I want
[11:46:39] <Jartza> LeoNerd: CH340 (y)
[11:47:59] <LeoNerd> Hrm... and what for LEDs? Just put them direct on the lines?
[11:53:15] <Jartza> why not
[11:54:36] <LeoNerd> TTL UART idles high, right? So LED should be VCC to signal
[11:59:54] <Jartza> yes, TTL UART does
[12:01:05] <LeoNerd> What should I stick unused modem handshaking lines at? (DSR/CTS)
[12:02:17] <Jartza> depends what you're doing :)
[12:15:40] <Jartza> I actually have no idea where to get ch340 chips
[12:17:14] <Jartza> maybe aliexpress or something
[13:28:49] <_abc_> Hello. I saw there are several versions of "commercial" usb-asp programmers, some with mega88, some with other things. Is there some difference? There are stories of "timeout, update firmware" and such too.
[13:37:35] <lowin> I always get that update firmware crap on mine!
[14:06:46] <_abc_> And did you do something about it?
[14:06:56] <_abc_> There appears to be an update somewhere.
[14:08:07] <_abc_> Does anyone know by <3 what the clear watchdog instruction is in avr atmega8 asm? I think it's just 2 bytes, i.e. a word? Correct?
[14:08:38] <_abc_> http://www.atmel.com/webdoc/avrassembler/avrassembler.wb_WDR.html right?
[14:08:54] <_abc_> 0x95A8 ?
[14:10:06] <_abc_> Next question: in avr hex files bytes are stored lsb at lower addr, msb at higher, or vice versa?
[14:12:21] <Lambda_Aurigae> _abc_, usb-asp was one of the early v-usb projects and has been co-opted by many people around the world to make their own avr programmers.
[14:13:00] <Lambda_Aurigae> there are dozens of different usbasp firmwares out there.
[14:13:19] <_abc_> I noticed.
[14:13:36] <_abc_> There's one on sale around here made by atnel.pl (ha)
[14:13:36] <Lambda_Aurigae> it is, at best, a fun toy.
[14:13:46] <_abc_> It seems to be liked by a lot of people too.
[14:14:01] <Lambda_Aurigae> I still have troubles with v-usb implementations on several machines here.
[14:14:13] <Lambda_Aurigae> yes, it is cheap and seems to work for the most part.
[14:14:36] <_abc_> Lambda_Aurigae: I know. I read the fine print and did one smart thing: Bought the cheapest hub I could find ($1.5), and used that to connect it. Never saw a problem.
[14:14:45] <Lambda_Aurigae> hehe..yeah.
[14:14:56] <Lambda_Aurigae> externally powered hub is best
[14:15:03] <_abc_> Laptop usb port protected, and no timing to f* the host stack.
[14:15:21] <_abc_> power is optional, the little things do not draw anywhere close to 50mA.
[14:15:41] <Lambda_Aurigae> a lot of people try to power their circuits through the usbasp and usb port.
[14:15:58] <Lambda_Aurigae> I found that a charger hub made a lot of difference in some projects.
[14:16:07] <Lambda_Aurigae> specially one that can provide 1A to 2A per port.
[14:16:16] <_abc_> So I am seriously considering embedding a hub into what I make now... v-usb on one output, a usb stick on another (inernal, so I never lose settings and the firmware again). 2 free ports left after that. Sounds nice, eh?
[14:16:24] <Lambda_Aurigae> much more expensive, but solves a lot of power issues.
[14:16:32] <_abc_> Lambda_Aurigae: I try and succeed to not use power when not needed ;)
[14:16:39] <_abc_> The art of miserable power supplies.
[14:16:44] <Lambda_Aurigae> hehe.
[14:17:02] <Lambda_Aurigae> a lot of newbies don't know that they can't get more than 100mA from a normal usb port without requesting it in software.
[14:17:21] <Lambda_Aurigae> then complain when their device horks hairballs when they try to drive a big LED display or motor or something from usb.
[14:17:29] <_abc_> Oh you can get 0.5A per port group, the mobo makers skimp on fuses, only the port group is fused with 2A ptc
[14:17:43] <_abc_> Yeah, okay, arduino people should be excused.
[14:17:45] <Lambda_Aurigae> you can, yes.
[14:18:01] <Lambda_Aurigae> but, by default, by usb spec, 100mA is the max you can get without a power request.
[14:18:20] <_abc_> The fuse skimping has become a norm because people buy external hdds with 2 usb power plugs to make up 1A and if the ports are individually fused the hdd does not work.
[14:18:22] <Lambda_Aurigae> some ports will just give you their max output without the request though...those are relatively new.
[14:18:27] <_abc_> I know what the spec says.
[14:18:40] <_abc_> Rely on 6 billion monkeys to respect that... brbr
[14:18:47] <Lambda_Aurigae> yeah.
[14:18:47] <_abc_> *brb
[14:19:10] <Lambda_Aurigae> and in theory they will create something worthwhile....the internet has disproven that theory.
[14:19:13] <Lambda_Aurigae> :}
[14:19:54] <_abc_> re
[14:20:54] <_abc_> The internet has taught us that the horrible statistics which apply to 100 average people gathered in one place, apply with a vengeance to 500 million congregating online. Statistically, in 100 people, you have 5 dying of cancer, 2 pedophiles, 5 gays, about 10 thieves, etc.
[14:21:16] <Lambda_Aurigae> I used to play around with v-usb quite a bit until I discovered certain PIC (yes, wash my fingers with lye soap) chips that can do usb without any external components.
[14:21:45] * _abc_ notices his blackened toenail is about to turn yellow. Should not kick any more steel chassis when not wearing hardened toe boots.
[14:22:00] <_abc_> Lambda_Aurigae: I use pics too, no need to be shy.
[14:22:21] <_abc_> Bit banging is usually pic in asm, C usually atmel, larger things are C and MIPS
[14:22:26] <Lambda_Aurigae> my little fingernail has finally grown back in from january when I crushed it between a copier and a wooden box hanging on the wall.
[14:22:47] <Lambda_Aurigae> I've grown fond of MIPS over the last year.
[14:22:57] <Lambda_Aurigae> with the pic32mx270f256b chip.
[14:23:16] <_abc_> Yes the larger pic32s can run retrobsd. Did you see that?
[14:23:27] <Lambda_Aurigae> yup.
[14:23:39] <_abc_> I always use outdated machines as hosts so mplab x has the speed of a drugged slug.
[14:23:42] <Lambda_Aurigae> this is one of the smaller ones....256K flash, 64K sram.
[14:23:50] <_abc_> I should really upgrade. Some money coming my way soon and I'll do it.
[14:24:12] <_abc_> iirc the smaller pdps which ran the earlyest nixes had 64k ram to
[14:24:32] <_abc_> xenix could run in iirc 480k ram or less on x86 without a mmu
[14:25:09] <Lambda_Aurigae> my current machine here is a 2.5ghz i5 with 8GB ram 2 core hyperthreaded laptop
[14:25:11] <Lambda_Aurigae> works well.
[14:25:17] <_abc_> I wonder if something, anything, could make Mr Ellison's open source sloth (Java) run faster in the future.
[14:25:20] <Lambda_Aurigae> I scraped it out of a junk pile.
[14:25:27] <Lambda_Aurigae> replaced the ram and harddrive and away it flies.
[14:25:43] <_abc_> droid devel env, eclipse, and mplab x are by far the slowest apps I have to use.
[14:25:57] <Lambda_Aurigae> droid dev environment sucks.
[14:26:03] <_abc_> There is no choice.
[14:26:04] <Lambda_Aurigae> never gotten one to work right here.
[14:26:11] <_abc_> I have to make little apps to support my hw
[14:26:20] <Lambda_Aurigae> I guess I might have to break down and put a win7 vm on this box for that.
[14:26:27] <_abc_> It takes exactly twice as long to load its modules than netbeans
[14:26:31] <Lambda_Aurigae> hehe.
[14:26:37] <Lambda_Aurigae> I like netbeans, mostly.
[14:26:37] <_abc_> I use the linux version of both. Worky.
[14:27:02] <Lambda_Aurigae> if I could figure out how to get a droid dev env working on linux I would love it.
[14:27:22] <Lambda_Aurigae> I have a piece of hardware I would love to tie to android.
[14:27:23] <_abc_> Just make SURE your graphics card has full hw accel under linux because the device emulator on non accel kvm refuses to run and non kvm makes the afore mentioned slow startup of the devel env look fast.
[14:27:39] <Lambda_Aurigae> the emulator is the one piece I got working.
[14:27:40] <_abc_> I am serious. A droid image takes over 7 minutes to boot on a dual core 3GHz cpu.
[14:27:50] <_abc_> Without hw accel.
[14:27:56] <Lambda_Aurigae> but, yeah, I have ok graphic accel...I run kerbal space program on it.
[14:28:24] <_abc_> I don't know what that is. I have a nvidia and another things I forget what, and both have horrible linux accel driver problems so they run in VESA mode.
[14:28:47] <Lambda_Aurigae> intel chipsets seem to work well in linux.
[14:29:13] <Lambda_Aurigae> but I have 8 android devices I can test on without having to run an emulator.
[14:29:27] <_abc_> Been shopping around the other day and the situation is still bad. The local pc shops have displays with many laptops and half of them run mint because they come without windows (low cost), and that is unaccelerated on all I tried.
[14:29:30] <Lambda_Aurigae> I collect phones from the phone company.
[14:29:41] <_abc_> i915 unaccelerated was what I saw.
[14:29:51] <Lambda_Aurigae> they have a phone discard drop box just inside the front door.
[14:29:52] <_abc_> How good is i915 accel support now?
[14:30:00] <Lambda_Aurigae> they are one of my clients and let me dig through the box.
[14:30:03] <_abc_> Lambda_Aurigae: I would like to move in next to where you live ;)
[14:30:12] <_abc_> Not really but okay I am a little envious.
[14:31:18] <_abc_> Looking through v-usb code, it seems not to be using the watchdog at all, yet it is enabled in fuses?!
[14:31:21] <Lambda_Aurigae> I'm running kubuntu on this thing and it has an i915 chipset and I have no problem running ksp at full screen full res
[14:31:26] <Lambda_Aurigae> works very well.
[14:31:34] <_abc_> Lambda_Aurigae: So accel?
[14:31:41] <Lambda_Aurigae> yup.
[14:32:01] <_abc_> Hmm okay. What glxgears speed do you get [non authoritative]?
[14:32:13] <Lambda_Aurigae> standard glxgears pushes 70 to 90 FPS
[14:32:23] <Lambda_Aurigae> running it right now.
[14:32:24] <_abc_> F*** soccer match this country against France. Everyone is on beer and foot.
[14:32:29] <Lambda_Aurigae> I'm also running 2 screens on this thing.
[14:32:30] <_abc_> Lambda_Aurigae: with what cpu load?
[14:32:40] <Lambda_Aurigae> the internal laptop screen and an external hdmi
[14:32:56] <_abc_> glxgears has 70-90fps with what cpu load%?
[14:33:28] <Lambda_Aurigae> under 20% on all 4 cores...
[14:33:39] <Lambda_Aurigae> and firefox has 28 open tabs on 4 windows.
[14:33:43] <_abc_> glxgears -speed 20.0 -gears 20 -fps should be a benchmark [non authoritative]
[14:33:45] <Lambda_Aurigae> also have chrome loaded
[14:33:51] <_abc_> Okay, you win.
[14:33:56] * _abc_ pretends to drool
[14:34:12] <Lambda_Aurigae> and remoted in with teamviewer at the moment.
[14:34:29] <Lambda_Aurigae> I'm sitting at a campground with another laptop and a 4G wifi jetpack.
[14:34:31] <_abc_> Well the droid emu will be molasses slow if not accelerated and also complain about missing kvm
[14:34:51] <_abc_> Lambda_Aurigae: hehe. Weekend eh.
[14:34:55] <Lambda_Aurigae> yup.
[14:35:02] <_abc_> Nice.
[14:35:19] <Lambda_Aurigae> I was out at 5am on the lake in my kayak
[14:35:32] * _abc_ sneezes due to the grill smoke from the neighbors camping outside with the tv and howling as the ball goes around on the green on the telly
[14:35:36] <Lambda_Aurigae> took friday off for a long weekend.
[14:36:00] <_abc_> I have not been in a kayak since I was like 15
[14:36:03] <Lambda_Aurigae> it's damned hot out today so sitting here in the camper with the AC running while the wifey naps.
[14:36:27] <Lambda_Aurigae> I go out once a week at least. Every day when camping if we are near deep enough water.
[14:36:30] <_abc_> And it would be unsafe to get into a kayak now. I have my built in flotation device ;)
[14:36:43] <Lambda_Aurigae> I weigh 220 pounds right now..
[14:36:53] <Lambda_Aurigae> down from 300 about 8 years ago.
[14:37:43] <Lambda_Aurigae> my kayak is rated for 250 pounds.
[14:37:58] <_abc_> You log on with teamviewer into linux? That is a little strange. No remote X, you prefer tw?
[14:38:11] <Lambda_Aurigae> teamviewer works simply enough.
[14:38:31] <Lambda_Aurigae> I've done vnc in the past but if my ip at home changes then I gotta figure it out again.
[14:38:34] <_abc_> I didn't know the linux server is okay
[14:38:45] <_abc_> So tw uses the tw server as middle man?
[14:38:52] <Lambda_Aurigae> yes.
[14:38:52] <_abc_> You don't have dyndns or such?
[14:39:12] <Lambda_Aurigae> you can setup a computer to announce itself to teamviewer servers then connect remotely eaily enough.
[14:39:21] <Lambda_Aurigae> dyndns started charging.
[14:39:37] <_abc_> Meh just get a d-link or tp-link router, the router maker provides free dyndns service.
[14:39:44] <Lambda_Aurigae> I can also connect to my office computer at work with the same teamviewer....behind corporate firewall.
[14:39:56] <Lambda_Aurigae> I won't buy d-link again.
[14:40:00] <_abc_> Problems?
[14:40:05] <Lambda_Aurigae> they last 6 months tops for me.
[14:40:14] <Lambda_Aurigae> horrid power supplies.
[14:40:14] <_abc_> How do they die?
[14:40:27] <Lambda_Aurigae> the regulator circuit on the board goes to hell
[14:40:33] <_abc_> Oh those suck always, I just replace it when new with a good make, and I don't see problems.
[14:40:40] <Lambda_Aurigae> usually a bad cap but those tend to take out something else.
[14:40:46] <_abc_> You have to make sure you do not exceed 9V supply input.
[14:41:03] <Lambda_Aurigae> I use ubiquiti gear now.
[14:41:06] <_abc_> They skimped on the inductor. If you put in a larger uH inductor you can up the volts to 16V
[14:41:11] <_abc_> It's classic.
[14:41:14] <Lambda_Aurigae> more expensive..the one on my home network has been good for 3 years.
[14:41:29] <_abc_> I heard about ubiquity. What do you think about ruckus?
[14:41:31] <Lambda_Aurigae> it also runs linux.
[14:41:37] <Lambda_Aurigae> there was a rukus?
[14:41:45] <_abc_> It's a router maker.
[14:41:48] <_abc_> Nice boxes.
[14:41:53] <Lambda_Aurigae> oh.
[14:41:55] <_abc_> Wifi ap's mostly.
[14:41:58] <Lambda_Aurigae> never used them.
[14:41:59] <_abc_> American
[14:42:05] <Lambda_Aurigae> ubiquiti makes some nice gear.
[14:42:15] <_abc_> https://www.ruckuswireless.com/
[14:42:36] <Lambda_Aurigae> I have a pair of nanostation-loco boxes that I can get 3 miles range at 150Mb/s
[14:42:44] <Lambda_Aurigae> they cost me about 70 dollars each.
[14:42:45] <_abc_> I can't really recommend them but I configure/upgrade some and they seem okay.
[14:43:01] <_abc_> That's a lot of money for 3 miles imho.
[14:43:27] <Lambda_Aurigae> not for good quality gear.
[14:43:50] <_abc_> If you say so.
[14:43:59] <Lambda_Aurigae> outside rated, come with their own POE injector.
[14:44:05] <_abc_> I don't know what to do with 150MBps.
[14:44:08] <Lambda_Aurigae> run 1 ethernet cable out to the pole.
[14:44:39] <Lambda_Aurigae> I was using them to feed internet to my inlaws who lived just under half a mile from me.
[14:44:55] <Lambda_Aurigae> wife and mother-n-law could do video conferencing between the two houses with no problem.
[14:45:57] <Lambda_Aurigae> and both devices were inside...one was inside a mobile home with metal sides.
[14:46:27] <Lambda_Aurigae> set it in front of a window pointing at our house and no problems getting a stable link, even in heavy rain and thunderstorms.
[14:46:32] <_abc_> That is like, testing mr. Maxwell.
[14:46:46] <_abc_> Putting one end in a metal sided caravan.
[14:46:55] <Lambda_Aurigae> I did have the transmitters cranked to 110% rated max too.
[14:47:18] <Lambda_Aurigae> they will go to 125%
[14:47:27] <_abc_> No trees in between? Those are what? 5GHz?
[14:47:35] <Lambda_Aurigae> 2.5ghz.
[14:47:38] <Lambda_Aurigae> and 5ghz
[14:47:45] <_abc_> Hey 125% is cheating. You're not supposed to be able to take them to more than 11...
[14:47:48] <_abc_> ;)
[14:47:49] <Lambda_Aurigae> a,b,g,n
[14:48:26] <Lambda_Aurigae> that's the stereo that goes to 11....aka Spinal Tap
[14:48:26] <_abc_> http://xkcd.org/ someone is making fun of unix man pages.
[14:48:32] <_abc_> Lambda_Aurigae: yep
[14:48:50] <Jartza> evening
[14:49:05] <Lambda_Aurigae> morning Jartza
[14:49:09] <Lambda_Aurigae> how goes the good fight?
[14:49:24] <Lambda_Aurigae> I'm thinking of going outside and building something.
[14:49:42] <Lambda_Aurigae> I brought a bunch of pallet wood and some of my woodworking tools with me this weekend.
[14:50:00] <Lambda_Aurigae> might make a bunch of mini-crates and see if anybody around will buy them.
[14:50:05] <Lambda_Aurigae> they usually sell pretty good.
[14:50:16] <Jartza> cool
[14:50:21] <Jartza> I'm just happy it's friday :)
[14:50:31] <Jartza> workweek done
[14:51:22] <Jartza> time to dig out my xmega32a4u
[14:52:03] <_abc_> How can you tell the work week is done when you pull out a mcu to replace the mcu you just used at work (I assume)?
[14:52:06] <_abc_> ;)
[14:54:47] <Lambda_Aurigae> I fix copiers for a living.
[14:54:58] <Lambda_Aurigae> for fun I tear old ones apart and make robots out of them.
[14:55:02] <Lambda_Aurigae> or make sawdust.
[14:55:21] <Lambda_Aurigae> and what parts of the wood that don't become sawdust I stick together to make other things.
[14:55:25] <Lambda_Aurigae> which I'm gonna go do now..
[14:55:27] <Lambda_Aurigae> laters.
[15:07:12] <Jartza> _abc_: my current assignment is with embedded linux, so getting back to 8bit AVR is.... well, like vacation :D
[15:08:45] <_abc_> Embedded linux is everywhere. So are copyright (copyleft) violations. I have 5 wlan routers, all run linux, and I have the source for only 2... of course those 2 already run openwrt. The others... not so much.
[15:09:26] <_abc_> This is really an openwrt question but does anyone here have a nice design of 8 pin spi flash readers/writer for avr/whatever to read/write firmware chips on stubborn routers and laptops and tablets with?
[15:10:43] <Jartza> well, this device will run openwrt
[15:11:48] <_abc_> http://openprog.altervista.org/OP_eng.html hm interesting beasti
[15:11:48] <_abc_> e
[15:20:06] <Jartza> so is this:
https://speakerdeck.com/asb/lowrisc-plans-for-risc-v-in-2016
[15:20:39] <Jartza> ->
http://www.lowrisc.org/
[15:21:05] <_abc_> The world is awash in SoCs. Another SoC? Uhh.
[15:22:12] <Jartza> but fully open soc
[15:32:07] <_abc_> opencores.org will have half a dozen already. I am not dissing it or anything but, seriously.
[15:56:19] <LeoNerd> Wow... the 8/16/32U2 is a weird chip. It has more interrupts than the 32U4, but fewer IOs, no I²C, no ADC, only 2 timers
[15:56:33] <LeoNerd> or maybe the 32U4 is weird in having so few interrupt pins and they're all in stupid places
[15:56:39] * LeoNerd still bitter about that
[15:56:49] <RikusW> indeed
[15:57:26] <RikusW> I suspect the 32u4 is a at90128 6/7 repacked in 44pin
[15:57:44] <RikusW> at90usb1286/7
[15:57:46] <LeoNerd> Also, 32U2 has hardware flow control, which is new...
[15:58:27] <LeoNerd> Basically, I'm just running out of IO pins on a tiny841, so I added a 74'595 for slow output pins, but then also considering a USB-CDC bridge for UART, .. at which point a 16U2 is smaller/cheaper than that prior combination of 3 chips
[15:58:38] <LeoNerd> But the 16/32U4 is crazy big/expensive
[16:04:21] <LeoNerd> Jartza: So, turns out the CH340 is a family of chips. There's a couple of 20pin ones and a small 16pin one.
[16:15:06] <LeoNerd> Oh.. but the 16pin is a SOIC, which makes it physically larger than the 20pin which comes in TSSOP
[16:49:46] <darsie> Is it bad to have data wires from a uC to an RFID module on a breadboard going near the RFID antenna?
[17:16:23] <Lambda_Aurigae> darsie, depends.
[17:16:51] <Lambda_Aurigae> if those data lines are cycling fast they might cause some noise interference with the antenna.
[17:17:46] <darsie> ok. So the antenna won't cause trouble on the data lines?
[17:18:34] <Lambda_Aurigae> again, depends.
[17:19:09] <darsie> ok. I separated them now.
[17:19:34] <Lambda_Aurigae> what rfid frequency are you using? how much power? is it transmitting or receiving? your data, is it going fast, as in, short pulses that could be interfered with?
[17:20:30] <darsie> I'll try it.
[17:21:22] <Lambda_Aurigae> lots of variables.
[17:21:32] <Lambda_Aurigae> there are all kinds of books on circuit design.
[17:23:32] <Lambda_Aurigae> when getting into mixed signal with RF, analog, and digital, it is more like black magic than science. Don't forget to sacrifice several small furry mammals during the design phase.
[17:24:32] <LeoNerd> Mm.. so I think I've just finished designing myself an Arduino Nano-sized mega328PB breakout board
[17:24:37] <LeoNerd> Different pinouts, but same formfactor
[17:24:52] <LeoNerd> Also USB bridge and ISP header, as before
[17:24:59] <Lambda_Aurigae> is that chip available in dip or just smt packages?
[17:25:08] <LeoNerd> SMT. I'm using the TQFP size
[17:32:39] * LeoNerd contemplates using a 4layer stackup
[17:32:57] <LeoNerd> I mean.. I have managed it with 2 layers, but I wonder if it would behave nicer to have a cleaner power plane
[17:33:02] <Lambda_Aurigae> I've made my own 4 layer boards before.
[17:34:15] <Lambda_Aurigae> made a 2-layer and 2 1-layer then used long leads to solder things into them. Solder the parts to the 2-layer board first then slide the other 1-layer boards onto the leads that poke through and solder them down as well.
[17:34:25] <Lambda_Aurigae> wire wrap sockets were a must.
[17:34:38] <Lambda_Aurigae> aligning the drill holes was a pain in the rumpus too.
[17:51:39] <phinxy> have anyone experience with ssd1307 display who could teach me how to work out column/rows/addressing modes?
[17:52:50] <phinxy> There are 8 PAGES and 128 COLUMNS. The screen itself is 128x64 pixels. What is a page? on the ssd1327 There arent any "PAGES" but "ROWS" instead
[17:54:14] <Lambda_Aurigae> do you have the datasheet?
[17:57:34] <phinxy> So ive misunderstood how much display ram it was in it, someone said it was possible to do buffering but there is only 128x64 bits of ram
[17:57:35] <Lambda_Aurigae> seems the ssd1307 is 128x39 pixels...from the ones I'm finding online.
[17:57:56] <phinxy> there might be a 128x32 pixel version where buffering is doable
[17:59:17] <Lambda_Aurigae> O
[17:59:32] <Lambda_Aurigae> I'm just seeing 128x39
[18:04:41] <Lambda_Aurigae> this a monochrome display, yes?
[18:05:21] <Lambda_Aurigae> a page is 8 rows. lsb being top, msb being bottom.
[18:05:36] <Lambda_Aurigae> so one byte defines 8 pixels in a vertical column.
[18:07:05] <Lambda_Aurigae> so, a page is 128 bytes....128 columns by 8 rows.
[18:09:13] <Lambda_Aurigae> but that is for the ssd1307
[18:09:21] <Lambda_Aurigae> the ssd1327 is laid out different.
[18:11:02] <Lambda_Aurigae> on the 1327 it is laid out as 128x128x4bits...with the display being 4 bit greyscale
[18:11:54] <Lambda_Aurigae> http://www.seeedstudio.com/wiki/images/8/82/SSD1327_datasheet.pdf
[18:11:59] <Lambda_Aurigae> section 8.7, page 30
[18:16:01] <phinxy> Lambda, working things out. had a typo, meant the 1306 not the 1307.
[18:18:03] <Lambda_Aurigae> 1307 is like the 1306
[18:18:12] <Lambda_Aurigae> only more pages and more rows.
[18:18:23] <Lambda_Aurigae> again, each page is 128 pixels wide by 8 pixels high.
[18:18:34] <Lambda_Aurigae> each byte is 8 pixels in one column.
[18:19:43] <Lambda_Aurigae> https://cdn-shop.adafruit.com/datasheets/SSD1306.pdf
[18:19:46] <Lambda_Aurigae> page 25
[18:19:58] <Lambda_Aurigae> section 8.7
[18:41:56] <phinxy> in horizontal model it draws boxes instead of going from start to end. in vertical mode i get the desired scan-line thing
[18:43:56] <phinxy> i mean, in horizontal mode it draws left to right but after 8 pixels it jumps down a row
[18:49:38] <Lambda_Aurigae> there appears to be multiple modes.
[18:49:44] <Lambda_Aurigae> it's all in the datasheet.
[19:57:49] <LeoNerd> Hrm.. so.. discovery of the evening: Linux w/ FTDI USB chip can detect a break condition just fine; Linux w/ CP210x cannot.
[19:57:57] <LeoNerd> Tomorrow I shall try out some other chips
[19:58:28] <LeoNerd> I don't know if that's a fundamental failing of the CP210x chip, or just the Linux driver for it
[20:03:04] <Lambda_Aurigae> dunno.
[20:03:11] <Lambda_Aurigae> never used the cp210x series
