Back
[07:28:43] <bluebie> heyas!
[07:30:39] <bluebie> I come with a question! attiny's have a neat chart in the datasheet for minimum voltage for various system frequencies. On a chip which can do 16mhz @ 5v, and 8mhz @ 3.3v, would it be within spec do you think to run the chip @ 3.3v on a 16mhz crystal, but booting with ckdiv8 on and then switching clkpr down to /2 so the 16 gets scaled to 8mhz, or is 16mhz even behind the prescaler no good?
[07:32:36] <Tom_itx> it would be running out of spec now wouldn't it?
[07:33:08] <Tom_itx> i presume ALL the chip is spec'd that way, not just part of it
[07:33:21] <bluebie> well.. most of the chip wouldn't be.. but yeah, that's pretty fair
[07:33:33] <specing> I think the OSC is at 1V anyway so you should be safe
[07:33:42] <specing> check with a scope
[07:35:04] <bluebie> I don't know much about EE - I pick up what I need to along the way to make little art projects and things happen, but my semi-guess understanding is the clock speed issue is one of power supply, where lower voltages have a harder time supplying the whole chip through it's own internal resistances each time the CPU steps forward and needs a surge of current from the nearby capacitors to transition correctly
[07:35:36] <bluebie> so if it's just a few transistors off in one corner running overclocked at 16mhz it'd probably not be a problem?
[07:35:50] <bluebie> but maybe I'm totally wrong about how that all works, hence asking you all!
[07:37:18] <RikusW> it should work at 3v3 and ckdiv8
[07:37:58] <specing> what attiny is this bluebie ?
[07:39:22] <bluebie> tiny167 I think
[07:43:29] <RikusW> I've even overclocked at 3v3 and 16MHz
[07:43:40] <RikusW> seems to work ok
[07:44:12] <RikusW> the whole idea behind CKDIV8 is to keep the clock in a safe zone for low V
[07:44:14] <bluebie> for production stuff tho
[07:44:19] <bluebie> mm yeah
[08:35:47] <gagan662> hello, any good simulator for avr 8bit for linux?
[09:08:38] <rue_house> dont simulate
[09:08:59] <rue_house> really, it wont give you an accurate representation of your programs behavoir int eh real world
[09:09:11] <rue_house> UNLESS you never plan to use it in the real world
[09:19:51] <RikusW> simulation can help you verify your code does what you think it does
[09:20:06] <RikusW> but its useless for actual realtime work
[09:23:35] <theBear> also useless for a lot of things 'cos a lot of what gets (by most people i know including me at least,) done on micros is kinda dependant on 'custom' hw or inputs/outputs that are at best 'tricky and involving much messing about' to simulate
[09:24:40] <theBear> and even if a sim has port status or fancy stuff like 'virtual leds' for visualising/showing outputs/function, most times things are so fast that at best you'll see rapid wiggling of pins, or more often just a blur
[09:30:28] <bss36504> My tow cents: Simulators are good at simulating internal operations, but not anything that requires external stimulus.
[09:30:34] <bss36504> two* yeesh.
[09:32:45] <twnqx> cow centes! moooooo
[09:33:00] <rue_house> simulators verify what your system does IN A SIMULTOR , but dont expect it to reflect the real world
[09:33:39] <Tom_itx> UPS plane down in Bama today
[09:33:40] <rue_house> that makes it good for university projects that dont actually get put togethor
[09:33:47] <Tom_itx> hope you didn't have any parts ordered
[09:33:59] <rue_house> not by ups
[09:35:15] <Tom_itx> ur gonna be late
[09:35:27] <mindw0rk_office> what's the cheapest, well documented qvga display module?
[09:57:02] <bss36504> abcminiuser: In my mass storage implementation (which built off of the dataflash/USBKEY demo), sometimes during a read or a write, the program gets stuck for a while somewhere inside MS_Device_USBTask(). I poked through the functions it calls and the only place I noticed that it might loop and get stuck is in MS_Device_ReadInCommandBlock() where it waits for Endpoint_Read_Stream_LE() to return. Any idea why this happens or how I c
[09:57:03] <bss36504> it? Other than this, it works nicely. AT90USB1287, 8Mhz internally clocked.
[09:57:51] <abcminiuser> Hrm. I guess it's waiting for the read to time out if the command block takes a long time to be sent
[09:58:08] <abcminiuser> You could lower the timeout, but that might have implications elsewhere
[09:58:28] <bss36504> yeah, this is many, many seconds.
[09:58:39] <bss36504> of "stuckness"
[10:00:33] <bss36504> also, due to my handy indicator LEDs, there is no memory activity which implies to me that it's not really even reaching the SCSI layer. It's like the host issues the command and the device misses it, then after the host notices nothing happening, it maybe sends it again?
[10:01:24] <bss36504> I'll move around my LED toggle to see if I can find exactly where it stalls.
[10:02:19] <abcminiuser> Strange, havn't seen that myself (as far as I remember)
[10:02:28] <theBear> hmmm, how exactly are you working your handy little leds ? not possibly stepping on something timing sensitive are you ?
[10:02:32] <abcminiuser> If you use Linux, see if dmesg shows anything interesting
[10:03:25] <bss36504> theBear: It takes like one clock cycle to toggle the LEDs. abcminiuser: I'll try it on my ubuntu VM.
[10:03:37] <theBear> mmm, i was just gonna say, timing OR noise/signal kinda issues might show up if you can look closer at the pc side usb details
[10:04:03] <bss36504> theBear: also, no change when I DONT toggle the LEDs.
[10:04:17] <theBear> toggling an led takes a single cycle, and i'm not abusing you, just suggesting you think over if you did something 'funny' without thinking about it, we all do 'funny' things now and again
[10:05:03] <bss36504> No worries, I know what you mean :)
[10:05:15] <bss36504> Thanks for suggestions, anything is welcome
[10:05:27] <theBear> :)
[10:08:35] <bss36504> another wonderful part of this issue is how intermittent it is.
[10:13:25] <bss36504> abcminiuser: Alright, dmesg has some interesting things. shall I paste the whole thing?
[10:13:49] <bss36504> actually, its a little bit long. here are some excerpts.
[10:14:00] <bss36504> disk3s1: media is not present.
[10:14:01] <bss36504> 0 [Level 3] [ReadUID 0] [Facility com.apple.system.fs] [ErrType IO] [ErrNo 6] [IOType Read] [PBlkNum 59] [LBlkNum 0] [FSLogMsgID 406322873] [FSLogMsgOrder First]
[10:14:14] <bss36504> USBMSC Identifier (non-unique): A4133353239351517041 0x3eb 0x2045 0x1
[10:14:15] <bss36504> USBF: 155710.312 AppleUSBEHCI[0xd6f4000]::Found a transaction past the completion deadline on bus 0xfd, timing out! (Addr: 12, EP: 3)
[10:14:15] <bss36504> USBF: 155790.313 AppleUSBEHCI[0xd6f4000]::Found a transaction past the completion deadline on bus 0xfd, timing out! (Addr: 12, EP: 3)
[10:18:54] <abcminiuser> Hrm, so it looks like it gives up on the transfer early then gets cranky when the response comes
[10:19:05] <abcminiuser> I guess the media is really slow on your board or something
[10:19:15] * abcminiuser is cooking dinner, so here on and off
[10:21:33] <bss36504> dinner eh? It's 11am where I am. The media is no slower than the atmel dataflash. It's a serial phase change memory chip and the SPI clock is running at 4Mhz. should take 1ms/512bytes if my math is correct.
[10:21:54] <bss36504> 1.024ms*
[10:25:55] <bss36504> plus, during these little timeouts, the board does nothing with the memory.
[10:26:59] <abcminiuser> Amy chance you don't return the correct amount of data?
[10:27:12] <abcminiuser> Requested Blocks * 512 bytes
[10:28:07] <Tom_itx> how's things in norway?
[10:29:25] <bss36504> it does the right amount. I send the address to the memory (block address * 512) then I read 64 bytes at a time (clear endpoint IN each 64) looped 8 times.
[10:30:08] <abcminiuser> Can you show the media access code?
[10:30:29] <abcminiuser> Tom_itx, crappy, I'm on support and India's on holidays tomorrow/Friday
[10:30:55] <Tom_itx> so we call for support and get an aussie accent?
[10:31:07] <Tom_itx> at least we could understand that
[10:31:27] <abcminiuser> No phone support thank christ
[10:31:32] <Tom_itx> haha
[10:31:40] <abcminiuser> You can get a "G'day" via email support tho
[10:31:50] <bss36504> noob moment: media access code?
[10:33:28] <abcminiuser> The code that reads and writes
[10:33:46] <bss36504> like my actual code? yes I can, if thats what you want.
[10:37:00] <bss36504> http://pastebin.com/07yn3D9L
[11:16:44] <bss36504> abcminiuser: I'm assuming you haven't seen my code yet, since you left:
http://pastebin.com/07yn3D9L
[11:18:43] <abcminiuser> Sorry, had a visitor
[11:18:45] <abcminiuser> Looking now
[11:18:53] <bss36504> not a problem at all.
[11:18:58] <bss36504> I appreciate the help
[11:19:23] <abcminiuser> Debugging code via IM, reading backlog, emails, a forum thread and trying to digest a ton of pasta...
[11:20:54] <abcminiuser> Add at the start of the read/write functions: if (Endpoint_WaitUntilReady())
[11:20:54] <abcminiuser> return;
[11:21:07] <abcminiuser> Add at the end of them: if (!(Endpoint_IsReadWriteAllowed()))
[11:21:07] <abcminiuser> Endpoint_ClearIN();
[11:21:39] <megal0maniac_afk> Tom_itx: I had a very pleasant chat with a gentleman named Brijesh recently. I don't know what your issue is :)
[11:27:38] <bss36504> abcminiuser: Well, it didn't break anything, but it still takes a little bit to start going and dmesg is still indicating a transaction timeout on addr 12, EP 3. What is endpoint 3?
[11:28:41] <abcminiuser> Check your Descriptors.c for what endpoint is what
[11:30:15] <bss36504> endpoint 3 is the data IN endpoint.
[11:30:16] <abcminiuser> Balls, it should be Endpoint_ClearOUT() for the read function
[11:30:27] <bss36504> alright, no problem
[11:30:38] * abcminiuser is not thinking well after a day on support....
[11:31:28] <bss36504> soooo....I'm clearing the recieve buffer at the end of a read?
[11:31:38] <abcminiuser> You're flusing it to the host
[11:31:48] <abcminiuser> Otherwise the last chunk won't be sent
[11:32:01] <bss36504> I thought the OUT was from Host->device. but i'll take your word for ot.
[11:32:03] <bss36504> it*
[11:32:40] <abcminiuser> Goddamit, you're right
[11:32:46] <bss36504> haha alright.
[11:32:52] <abcminiuser> I really can't think today :S
[11:33:08] <bss36504> I know the wonderful 6:18PM feeling.
[11:33:36] <abcminiuser> When the last 6 support tickets came in this afternoon in the same minute I just gave up and went home :S
[11:34:08] <bss36504> Yeah, that's called "fuck this shit O'clock"
[11:34:58] <abcminiuser> Roughly what I said when I left....
[11:35:13] <abcminiuser> "Baesj, hjemme tid"
[11:35:55] <bss36504> haha (lol@google translate)
[11:38:06] <bss36504> another detail that I forgot here: I added an eject handler similarly to how you did it in the Mass Storage Bootloader demo (the newest one from git) and sometimes even that gets hung up. By similarly, I basically look in the SCSI decoder for the Star/stop command and then set a flag to false that controls the main loop in MassStorage.c. Then I just call USB_Disable() and sit there.
[11:38:24] <bss36504> This is why i'm concerned that it is not due to my media drivers.
[11:38:40] <bss36504> some sort of sticking in a lower level USB related thing.
[11:39:01] <abcminiuser> Yeah, can you use Wireshark to trace the comms?
[11:40:30] <bss36504> well yes I have Wireshark, but I'm not very familiar. So we'll see how it goes. If you feel like bailing for the night, please do.
[11:40:44] <bss36504> (This PCB layout isn't going to do itself)
[11:40:59] * abcminiuser is trying to come up with a way to add distributables to the LUFA AS6 plugin via scripts
[11:41:10] <abcminiuser> I'm making yet another horror methinks
[11:41:50] <bss36504> you mean you don't just magically get your code to work on the first try?
[11:42:03] <abcminiuser> Heh
[11:42:12] <abcminiuser> Some of the LUFA projects have accompanying host code
[11:42:17] <abcminiuser> Like Python or C# APPS
[11:42:36] <abcminiuser> But the Studio extension system can't cope with marking a directory as being a support distributable
[11:43:01] <abcminiuser> So I need to add marker text files, then search through when I generate the extension for them and archive the directories into a single zip
[11:44:12] <bss36504> That sounds like a classic pain in the ass.
[11:45:56] <bss36504> I'm hoping my poor mac can support 2 VMs at once....
[11:52:49] <abcminiuser> Bugger me, this works
[11:52:51] <abcminiuser> Hrm
[12:09:14] <abcminiuser> Hrm, seems the tools team added what I wanted anyway
[12:09:18] <abcminiuser> Yay, I guess
[12:10:01] <bss36504> well thats helpful I suppose. I
[12:10:13] * bss36504 would really like it if Wireshark would open...
[12:24:57] <bss36504> Alright, well I
[12:25:25] <bss36504> I'm beginning to think that these issues may just be related to OSX. My Ubuntu VM shows no errors in dmesg.
[12:25:31] <theBear> maybe call it ethereal, it might like that
[12:25:37] <theBear> heh, make it feel 10 years younger
[12:27:34] <abcminiuser> Oh OSX
[12:27:40] <abcminiuser> Screw that
[12:27:53] <abcminiuser> I tried to set up a VM of it for testing several times, no go
[12:32:18] <RikusW> I got it working
[12:32:42] <RikusW> abcminiuser: I used VirtualBox and iAtkos ML2
[12:33:31] <RikusW> abcminiuser:
http://iatkos.me/2012/12/30/iatkos-ml2/
[12:33:37] <abcminiuser> Ah, haven't tried that one
[12:33:41] <abcminiuser> I'll give it a go now
[12:33:59] <RikusW> there are newer versions too
[12:34:15] <RikusW> ML3 can boot from a stick
[12:35:37] <RikusW> ML3U is newest
[12:36:25] <bss36504> I like my mac, ok you haters? Anyway, I run windows in a VM for AS6, and i have an Ubuntu VM as well.
[12:36:39] <abcminiuser> Use whatever you want, doesn't affect me
[12:36:49] <bss36504> haha i know, I'm not upset.
[12:36:51] <abcminiuser> I just chose not to test or support something that actively fought me
[12:37:06] <abcminiuser> If it works great, if not, I don't know so I sleep happy
[12:37:55] <inkjetunito> windows 8 fixed osx
[12:38:02] <bss36504> Oh i get it, the apple fanboyism and closed source nature of the company is annoying, but I do like their products so it's a tradeoff. But maybe in the future I can return some of the support by helping port LUFA to OSX.
[12:38:08] <bss36504> inkjetunito: wut?
[12:38:41] <inkjetunito> bss36504: yes, after having to use win8, the small disturbances of osx don't bother at all anymore %)
[12:38:55] <bss36504> haha I see what you mean now.
[12:39:10] <bss36504> You mean to tell me you didn't want to turn your PC into a tablet?
[12:39:38] <inkjetunito> bss36504: no. it does fail on tablets aswell. anyway: offtopic
[12:39:53] <bss36504> good point, and good point.
[12:43:12] <inkjetunito> wtf. a distro provides avrdude and binutils for avr in its stock repos, but the cross compiler isn't there
[12:45:33] <theBear> abcminiuser, you know, while searching for perfectly reasonable things like distro livecd downloads <grin> i may have spied the odd "osx vm - installed and ready to run on XYZ vm software" here and there
[12:46:16] <abcminiuser> theBear, "comes with free rootkit"
[12:46:23] <abcminiuser> I'm a bit wary of preconfigured OSes like that
[12:46:37] <abcminiuser> Although granted reinstalling from a modified binary doesn't guarantee anything either
[12:47:23] <OndraSter> it is not like osx will run in the VM out of the box
[12:47:23] <OndraSter> let alone since there is no actual installation DVD, is there?
[12:47:24] <OndraSter> isn't*
[12:47:24] <OndraSter> any*
[12:47:25] <theBear> abcminiuser, that's a point... i never really wanted to try one, so i never really considered it
[12:47:48] <abcminiuser> OndraSter, yeah they're modified
[12:48:07] <abcminiuser> But at least it prevents script kiddies from installing it, installing a rootkit and then packaging it up
[12:48:17] <abcminiuser> I want a test environement, not a botnet client
[12:48:22] <theBear> tho you'd think something 'decent' like vmware or vbox would be able to pretend to be apple hw... at the same time i haven't kept up, i got some distant memory about a tiny 'apple prom' or something that might be needed to REALLY look like an apple x86(/64)
[12:49:06] <theBear> at the same time, being vm software isolating from the network and doing some basic and honeypot/general-faked-network tests is relatively safe and easy
[12:59:05] <abcminiuser> Bahaha, oh god
https://www.youtube.com/watch?v=xgAXCAWQUic&feature=youtu.be&t=4m33s
[12:59:08] <abcminiuser> Watch the birds
[13:15:35] <bss36504> Ha!
[13:15:50] <bss36504> Bet it smelled like chicken.
[13:17:39] <N1njaneer> "Flllammmmeee broooilllled!"
[13:17:54] <N1njaneer> Next year they just made it squirt BBQ sauce.
[13:17:59] <N1njaneer> Before ignition
[13:35:53] <RikusW> pest control 2000L of petrol kills 5 million birds in seconds
http://www.farmersweekly.co.za/article.aspx?id=6385&h=Where-do-you-fill-your-tank
[13:37:53] <masterq> Hi
[13:39:15] <N1njaneer> RikusW: Wow, there's efficiency and actual environmental forward-thinking we'll never have in the USA.
[13:40:34] <RikusW> There is explosives in the petrol drums to vaporize it first
[13:40:52] <RikusW> Unfortunately I cannot find any YT vids
[13:42:02] <bss36504> very interesting
[13:42:38] <RikusW> http://c1planetsavecom.wpengine.netdna-cdn.com/files/2012/07/Red-billed_quelea_flocking_at_waterhole-e1341814405220.jpg
[13:42:48] <RikusW> they destroy crops in a hurry
[13:43:05] <bss36504> Interesting. I'm glad we dont have that problem where I live.
[13:45:21] <RikusW> the actual article www.total.co.za/os/content/NT0004B556.pdf
[13:46:10] <RikusW> 4ha ~~ 8 acre fireball :-D
[13:46:14] <theBear> heh, and they say they go all out for olympic ceremonies :)
[14:01:52] <abcminiuser> Sciencing ML2 into a VM now
[14:02:00] <abcminiuser> Got 3, didn't work due to UEFI screwiness
[14:07:45] <RikusW> So are torrents _that_ fast over there ? :)
[14:08:06] <RikusW> I got xcode and xquartz too
[14:08:29] <RikusW> you actually only need the compiler not the whole studio
[14:10:19] <abcminiuser> 3MB/s if I'm lucky
[14:10:25] <RikusW> nice
[14:10:30] <abcminiuser> 1MB/s is more usual
[14:10:38] <RikusW> 400k max here
[14:10:43] <RikusW> more like 100k
[14:11:00] <RikusW> actually 470k
[14:11:09] <bss36504> I can get 5MB/s at school. more like 1.5 at home.
[14:11:38] <avrdood> N1njaneer: rebuilt the circuit and re-solder a few pins, now it works like a charm !
[14:11:45] <avrdood> N1njaneer: except for a little offset, its all good
[14:12:28] <bss36504> R0b0t1: How goes the PWM?
[14:32:56] <RikusW> http://imgur.com/gallery/o7F3Hbl
[14:33:02] <RikusW> windows compared to houses
[14:33:11] <RikusW> WinME
http://i.imgur.com/9AVULgD.jpg
[14:33:11] <abcminiuser> Baahaha
[14:33:18] <abcminiuser> BAHAHAHAA
[14:33:34] <abcminiuser> Mass storage demo works...
[14:33:46] <Joggl> abcminiuser, grats
[14:33:53] <R0b0t1> bss36504: Eh, I'mma leave it until it becomes a problem
[14:34:06] <R0b0t1> probably have to cut a trace and swap the pinout with hookup wire
[14:34:46] <abcminiuser> LUFA on an XMEGA running on OSX, I never thougth I'd see the day...
[14:34:46] <bss36504> Well that timer 1 lib I sent you to might work as well if the system is Arduino. But cutting a trace isn't so bad either.
[14:35:14] <bss36504> abcminiuser: holy shit, congrats. What do you use for a demo platform so you can test all these different chips?
[14:35:19] <Joggl> abcminiuser, do you have suicide feelings now?
[14:35:25] <R0b0t1> I do the same thing as I would with the timer1 lib
[14:35:36] <abcminiuser> I have a...selection of AVR products
[14:35:39] <R0b0t1> like, I don't think I can remove the overhead
[14:35:42] <abcminiuser> And several VMs
[14:36:05] <bss36504> I figured you had a nice selection. All I have is perf board and a JTAGICE III lol
[14:36:12] <Joggl> and if the selection is not big enough he let build him new xplain-boards ;)
[14:36:13] <bss36504> then again, i dont work for atmel....
[14:36:24] <abcminiuser> I got all my loot before I started
[14:36:31] <abcminiuser> Atmel and other engineers just sent me free stuff
[14:36:41] <bss36504> well thats nice of them.
[14:36:54] <abcminiuser> Yeah, except I feel guilty since I only use a portion of it reguarly
[14:37:01] <bss36504> and they say it's not what you know but who you know.
[14:37:26] <bss36504> well if you are feeling generous, feel free to send some loot to a poor college student/application engineering intern.
[14:37:36] <abcminiuser> Hah
[14:37:44] <abcminiuser> What do you have your eyes on?
[14:37:55] <N1njaneer> bss: Whereabouts are you?
[14:38:54] <bss36504> Nothing in particular. I'd like to play with some Xmega stuff, but I dont have the time. I am THE one application engineer at the company I intern with, so I make little demo widgets and such to show off our memory (such as a mass storage device).
[14:39:09] <bss36504> N1njaneer: I'm in Burlington VT, USA
[14:39:19] <N1njaneer> bss: Cleveland here. :)
[14:39:58] <N1njaneer> bss: Find out who your local Atmel FAE is and contact them. They are VERY generous with demo kits if you get to talking to them, especially if you have a commercial application potential :)
[14:40:05] <Joggl> if you would live in germany i could have send you some of my eval-kits for playing ;)
[14:40:24] <RikusW> Win 3.1
http://t1.gstatic.com/images?q=tbn:ANd9GcT5EJ4ixkaF8mkM6nZrSqSJkripAiJq0Vw0KSraaFl0T67pXlIF
[14:40:58] <bss36504> Plus in addition to application stuff, I make PCB layouts to use on our tester, write test code etc. basically whatever they need. My cube has all sorts of neat toys. Thanks for the tip N1njaneer :0
[14:41:06] <bss36504> also, thanks for the though Joggl
[14:41:11] <bss36504> thought*
[14:41:41] <Joggl> ask dean, my kit is great (even if i think its still unassembled....)
[14:42:07] <abcminiuser> Half assembled
[14:42:29] <abcminiuser> I'll finish it tomorrow methinks
[14:42:33] <abcminiuser> Support is kinda crappy
[14:42:39] <abcminiuser> So it'll take my mind off it
[14:42:40] <bss36504> which kit is it? actually, heres a good question for the group: what kit would be good to get?
[14:42:40] <Joggl> ;)
[14:42:58] <bss36504> "which kit is it" was directed to Joggl
[14:43:04] <abcminiuser> What the christ
[14:43:12] <Joggl> http://www.ehajo.de/Bausaetze/aTeVaL
[14:43:15] <abcminiuser> Why the hell does scrolling down in OSX scroll upwards?
[14:43:28] <Joggl> omg, english translation is not available ;)
[14:43:32] <RikusW> the menus are ubuntu style too
[14:43:43] <RikusW> abcminiuser: iirc there is a setting to change that
[14:43:45] <bss36504> abcminiuser: It's more natural on a trackpad. just change it in settings
[14:43:51] <abcminiuser> Jesus
[14:44:09] * RikusW is definitely keeping his PC
[14:44:34] <bss36504> But but....it's a setting. meaning you can change it....
[14:44:38] <abcminiuser> If Hitler made an OS, it would definetely be like this
[14:44:54] <abcminiuser> LOOK AT ZE SHINY BEACHBALLS
[14:44:55] <bss36504> that is harsh. his would be way less inviting.
[14:45:12] <Joggl> "you are jewish, you cant scroll."
[14:45:17] <abcminiuser> So far it quacks like a Unix duck and walks like a one legged Kangaroo
[14:45:27] <RikusW> lol
[14:45:50] <bss36504> It's really a nice marriage once you get used to it. I bought mine before I was really into computers and programming, and it works just swell for me.
[14:46:04] <bss36504> But I bet you know more about Kangaroos than me.
[14:46:09] <abcminiuser> Nope, I'm out
[14:46:17] <abcminiuser> I'll just pretend I tested with it
[14:46:20] <abcminiuser> Jesus
[14:46:28] <abcminiuser> Give me Windows or Linux any day
[14:46:34] <RikusW> About how I felt...
[14:46:45] <RikusW> Defaulting to Mint with XP in VBox
[14:47:00] <abcminiuser> Win7 native here, with XP and Mint VMs
[14:47:10] <bss36504> It's basically linux wrapped in a nice UI with some less threatening apps, and more stable than windows.
[14:47:26] <RikusW> abcminiuser: somehow your AVRISP doesn't want work in the VBox, the dragon and my board works just fine
[14:47:44] <abcminiuser> Not super suprised given what I do with the USB controller to make it workj
[14:47:52] <RikusW> heh
[14:48:02] <RikusW> it does work in plain XP though
[14:48:20] <abcminiuser> Anyone feel brave to try a new LUFA AS6 extension build?
[14:48:30] <RikusW> link ?
[14:48:36] <abcminiuser> I'll upload it now
[14:48:44] <N1njaneer> I have one machine in the office here that still runs Win 3.1 :)
[14:49:10] <RikusW> what for ?!
[14:49:12] <bss36504> RikusW: How is the dragon? I have one, and i had issues with it. The JTAGICE III is way better for only a little more IMHO.
[14:49:25] <N1njaneer> RikuxW: One of our BTU ovens
[14:49:48] <N1njaneer> It's funny to remember back to exiting to a DOS prompt before you shut down :)
[14:49:50] <abcminiuser> Yes, please go ICE3 :S
[14:53:04] <abcminiuser> Balls, have to regen it
[14:53:13] <abcminiuser> Jesus the documentation transform takes forever
[14:53:17] <abcminiuser> And yet no one uses it :S
[14:53:34] <N1njaneer> abc: Doxygen?
[14:53:39] <abcminiuser> Yeah
[14:53:55] <abcminiuser> But I did a BUNCH of work to make it transform into the format needed for Atmel Studio's help system
[14:54:02] <abcminiuser> So you get native help inbuilt
[14:54:21] <abcminiuser> It uses a bunch of XML transforms, Docbook, and a few other bits and pieces
[14:54:28] <soul-d> works in eclipse to sorta like popup with the function
[14:54:37] <N1njaneer> I thought Doxygen was a cool concept, but frankly a zero-sum game. It helps to generate documentation at help, but often at the expense of starting to make the code unweildly to write.
[14:54:51] <N1njaneer> +and help
[14:54:58] <abcminiuser> It's not to bad if you keep writing in a consistent style
[14:55:05] <abcminiuser> My gripe is the Doxygen parser bugs
[14:55:09] <abcminiuser> Of which, there are many
[14:55:22] <bss36504> abcminiuser: I skimmed your documentation :p
[14:55:38] <abcminiuser> Did you at least skim it inside Atmel Studio? :P
[14:55:56] <abcminiuser> http://fourwalledcubicle.com/files/temp/LUFA-TESTING-13.08.14-21.42.00.vsix
[14:56:15] <bss36504> and here WE GOOOOOOO!
[14:58:18] <abcminiuser> Remove any previously installed releases
[14:58:20] <abcminiuser> Including testing
[14:58:31] <RikusW> bss36504: My dragon works okay for me
[14:58:52] <RikusW> had some buggyness with breakpoints in dW though
[14:59:45] <N1njaneer> avrdood: Sorry, from earlier -- great to hear you got it working! Get the temps you are expecting now, then?
[14:59:54] <bss36504> I really hated dW. I bricked more than one 328 trying to get that to work. (that is the one where you have to enable debugging but disable SPI or some crap?) I like JTAG the best
[15:00:46] <abcminiuser> Yeah, dW is a bit unstable at best
[15:01:58] <RikusW> bss36504: the dragon can recover that
[15:02:08] <RikusW> or use HVPP
[15:02:11] <RikusW> that WILL work
[15:02:18] <bss36504> yeah, but with the parallel junk. It's just such a pain is all.
[15:02:29] <RikusW> a bit cumbersome, 21 connections to be made
[15:02:34] <avrdood> N1njaneer: yes! it turns out my DMM is off so the reading is just fine!
[15:02:35] <bss36504> yeah, that part.
[15:03:00] <RikusW> the dragon can disable dW without HVPP
[15:03:10] <bss36504> how?
[15:03:24] <RikusW> either commandline or in the debugger itself
[15:03:39] <RikusW> AS4 had a commandline tool
[15:03:44] <RikusW> not sure about AS6
[15:03:59] <bss36504> I think so. I just thought dW couldn't set any fuses.
[15:04:04] <RikusW> while debugging in debugger options there is a disable dW button
[15:04:31] <RikusW> well technically dW is temporarily disabled and ISP used to change the fuses
[15:04:43] <RikusW> my programmer can do that automatically
https://sites.google.com/site/megau2s/
[15:04:59] <bss36504> http://www.joeydevilla.com/wordpress/wp-content/uploads/2010/10/the_more_you_know2.jpg
[15:05:07] <RikusW> bss36504:
http://www.ruemohr.org/docs/debugwire.html
[15:05:09] <bss36504> thanks
[15:05:45] <RikusW> dW is just one wire UART
[15:06:23] <RikusW> studio docs should explain how to do it using the dragon
[15:06:28] <bss36504> huh. well at any rate, I have an ICE3 now. I might try and get it working though
[15:07:13] <RikusW> that might work to disable dW too
[15:07:27] <RikusW> you'll need a clock on the dW chip
[15:07:38] <RikusW> so if you set it to crystal connect one
[15:07:56] <bss36504> Well yeah, if you access to jtag, then you can do anything. but for non-jtag chips like the 328, it's dW and ISP
[15:09:01] <RikusW> abcminiuser: how do I make the vsix work with AS6.1 ?
[15:09:09] <RikusW> Tools -> Addin ?
[15:09:15] <RikusW> or extension ?
[15:09:20] <abcminiuser> "Run" it and it should prompt to install
[15:09:28] <abcminiuser> Double-click
[15:10:30] <RikusW> do I need to be online ?
[15:10:42] <abcminiuser> No
[15:10:53] <abcminiuser> VSIX is just a fancy ZIP that Visual Studio understands
[15:11:32] <RikusW> helplibmanager crashed... ?
[15:11:55] <abcminiuser> Balls, that shouldn't happen
[15:12:07] <abcminiuser> Not to worry, there's an option to force re-install the help inside Studio for it
[15:12:13] <RikusW> is the LUFA source part of the vsix ?
[15:12:25] <abcminiuser> Yes
[15:12:28] <RikusW> nice ! :)
[15:12:30] <abcminiuser> If you rename it to zip
[15:12:40] <abcminiuser> You can pull it out of the "contents.zip" inner file
[15:13:08] <abcminiuser> But if it's installed in AS6.1 you get all the example magic
[15:14:40] <RikusW> created sample CDC project :)
[15:14:54] <RikusW> so now its _that_ easy ? ! :)
[15:15:03] <abcminiuser> Has been for 8 months now :P
[15:15:25] <RikusW> build done :)
[15:16:32] * RikusW hit the wrong close button....
[15:16:43] <N1njaneer> avr: Great to hear!
[15:16:56] <abcminiuser> Now marvel at that beautiful XAML Getting Started page
[15:17:22] <RikusW> nice indeed
[15:17:48] <RikusW> Is there any way to add a custom programmer / debugger to AS6 ?
[15:18:01] <abcminiuser> External tools would allow you to call AVRDude
[15:18:05] <abcminiuser> But no debuggers
[15:18:18] <RikusW> No public Atmel docs on the interfaces ?
[15:18:27] <abcminiuser> No public ones :)
[15:21:53] <RikusW> Linux house
https://encrypted-tbn0.gstatic.com/images?q=tbn:ANd9GcRwGcOr8dOK7BNpjhJ-7JIiVgSlbopKDS89Qjc3Lp9_ZpuLYNy2
[15:25:40] <RikusW> zlog
[15:27:37] <RikusW> OSX
http://elitedaily.com/wp-content/uploads/2013/01/luxury-prison-cell-elite-daily.jpg
[15:51:28] <megal0maniac> Quick! I need the wizards!
[15:52:39] <megal0maniac> C makes no sense
[15:53:21] <megal0maniac> http://pastebin.com/0P0LuD0A
[15:54:21] <megal0maniac> Lines 25 and 35
[15:54:30] <RikusW> megal0maniac: have you seen
http://imgur.com/gallery/o7F3Hbl ?
[15:54:46] <megal0maniac> I should be getting floating point numbers formatted to two decimal places
[15:54:51] <megal0maniac> Instead I'm getting 908692438223645460000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000.0000
[15:55:01] <megal0maniac> Why? This never happens in Java :/
[15:55:10] <abcminiuser> Kaching
[15:55:19] <abcminiuser> Who's Michael here?
[15:55:43] <megal0maniac> ...Me?
[15:55:45] <megal0maniac> Why?
[15:55:45] <RikusW> megal0maniac: try just %f
[15:55:52] <abcminiuser> Just PayPal'd me?
[15:55:52] <megal0maniac> I did. Same crap
[15:55:58] <megal0maniac> Not that I know of
[15:56:03] <abcminiuser> Hrm
[15:56:09] <megal0maniac> I guess I could?
[15:56:33] <megal0maniac> abcminiuser: Why is my program shit?
[15:56:53] <N1njaneer> Mega: Do %02f
[15:57:02] <abcminiuser> No, just assumed it was someone here
[15:57:15] <abcminiuser> When I get donations it's usually from someone I'm talking to
[15:57:22] <abcminiuser> Guess I just have secret admirers
[15:57:37] <megal0maniac> N1njaneer: Same thing :/
[15:58:50] <N1njaneer> mega: It sounds like perhaps you don't have an extended formatting library included.
[15:59:03] <megal0maniac> What is conio.h?
[15:59:09] <N1njaneer> I BELIVE there is a lighter-weight sprintf() function. Trying to remember since it's been a while.
[15:59:28] <N1njaneer> Wow haven't seen conio.h in forever - that's an old one for CONsoleIO
[15:59:42] <N1njaneer> What is the architecture/compiler/target?
[16:00:01] <N1njaneer> Mega: Sorry reading backscroll :)
[16:00:14] * abcminiuser Emergency Power Save Mode Activated
[16:00:18] <abcminiuser> Talk to y'all tomorrow
[16:00:24] <N1njaneer> See'ya Dean!
[16:00:29] <abcminiuser> Night all!
[16:00:30] <megal0maniac> Night!
[16:00:39] <megal0maniac> Sleep with one eye open
[16:00:44] <megal0maniac> You never know with these admirers
[16:00:49] <N1njaneer> mega: This going out to an AVR?
[16:00:52] <megal0maniac> N1njaneer: Windows CLI
[16:00:54] <megal0maniac> x86
[16:00:56] <megal0maniac> MinGW
[16:01:00] <N1njaneer> Gotcha
[16:01:06] <megal0maniac> Programming 1
[16:01:07] <N1njaneer> I was gonna say with conio.h :)
[16:01:42] <megal0maniac> Tried. No success
[16:02:58] <N1njaneer> I see %.2f, have you tried %0.2f ?
[16:04:17] <megal0maniac> I have
[16:04:27] <megal0maniac> Just tried on linux, same thing :/
[16:04:31] <megal0maniac> wth?
[16:05:32] <N1njaneer> Hmm, hold on a sec. Playing with i.
[16:05:33] <N1njaneer> it
[16:05:47] <N1njaneer> er
[16:05:54] <N1njaneer> Your temps are ints :)
[16:06:01] <megal0maniac> Of ffs
[16:06:03] <megal0maniac> Oh
[16:06:05] <megal0maniac> ;lkjfdsa;lkj
[16:06:05] <N1njaneer> temp_f and temp_c need to be floats
[16:06:15] <N1njaneer> You're seeing it do a weird cast of the datatyp
[16:06:16] <megal0maniac> Fell asleep in maths test today. So there
[16:06:18] <megal0maniac> Thatnks :)
[16:06:22] <megal0maniac> !!!
[16:06:29] <megal0maniac> s/Thatnks/thanks
[16:07:06] <N1njaneer> No prob. I like easy ones :)
[16:08:46] <megal0maniac> Oh, well that's good to know ;) C is a funny thing.
[16:09:05] <N1njaneer> It's a great language but can take some getting used to the nuances :)
[16:09:18] <megal0maniac> Cool. Working now. %02f doesn't work, needed %0.2f
[16:09:49] <megal0maniac> I learnt in Java, and this is making it feel like Arduino in comparison
[16:10:10] <N1njaneer> %02d will give you leading zeros of two places for integers. So "05" instead of "5" if you need to maintain fixed widths
[16:10:43] <N1njaneer> Though to be fair those are format specifiers for the standard function calls, so TECHNICALLY not part o the language. :)
[16:11:04] <N1njaneer> And you can always rewrite them or write new functions when the existing ones don't do what you need :)
[16:11:17] <masterq> has somebody used a avr32 uc3 before?
[16:12:09] <megal0maniac> masterq: Yes, but I had no idea what I was doing. As usual
[16:12:14] <megal0maniac> N1njaneer: Gotcha :)
[16:12:30] <masterq> i have just finish my first board and first avr32 application, and now i like to program it, but i got an error/warning: Bootloader and code overlap.
[16:12:42] <masterq> i use dfu-programmer and the bootloader
[16:12:57] <masterq> i have no avr32 jtag programmer.
[16:14:22] <N1njaneer> master: Not worked with AVR32 specifically, but are you sure the application isn't too large?
[16:14:57] <masterq> im not sure, what is gonna happend if i use --suppress-bootloader-mem, am i overwriting the bootloader? and if why? my firmware is soo small. Or am i just overwriting the jump command at the beginning of the flash space? which makes the device jump into bootloader space?
[16:15:09] <N1njaneer> At least on AVR the bootloader goes to the top of program flash, so if your application is too large to fit it will run in to the bootloader. Check the final compiled size of your application against the total that should be avaliable in the device minus the reserved size for bootloader.
[16:15:12] <masterq> avr32-size -B main.hex
[16:15:16] <masterq> text data bss dec hex filename
[16:15:16] <masterq> 0 6002 0 6002 1772 main.hex
[16:15:53] <N1njaneer> How big is the device flash?
[16:16:00] <masterq> 512 kbyte
[16:16:19] <N1njaneer> Yeah, doesn't seem like that would be the issue, then. :)
[16:16:33] <masterq> hehe no...
[16:16:38] <masterq> su
[16:16:42] <masterq> sry^^
[16:16:55] <masterq> good is was not the pwd^^
[16:17:09] <masterq> i use this command: dfu-programmer at32uc3b0512 flash main.hex
[16:17:30] <masterq> maybe there is something wrong
[16:17:56] <N1njaneer> I'm not sure on the commands for that, and Dean just logged out for the night :(
[16:18:25] <masterq> there is an other command called: flash-user
[16:19:51] <masterq> maybe i have to use this... i'm just careful, because if i destroy the bootloader i have to build a new device or by a jtag programmer
[16:20:07] <masterq> *buy
[16:20:35] <megal0maniac> masterq: I'd suggest you ask at #aery32, but they're very quiet.
[16:20:57] <megal0maniac> If anyone is there, though, then they'll probably know
[16:21:32] <N1njaneer> I would imagine the DFU bootloader SHOULD be protected
[16:21:41] <megal0maniac> N1njaneer: It is by default
[16:21:58] <megal0maniac> And it's a pesky business putting it back when JTAG kills it
[16:22:10] <masterq> thanks, but first i will try to google a while and then ask there
[16:22:15] <megal0maniac> This is the full extent of my knowledge
[16:22:36] <N1njaneer> I like having JTAG since you can fix all problems with it.
[16:22:47] <N1njaneer> Save for setting a clock fuse wrong, or, well... disabling JTAG :)
[16:23:05] <Roklobsta> shaddap
[16:23:46] <Roklobsta> don't remind me
[16:24:11] <masterq> uff, is it possible to disable jtag?
[16:24:17] <twnqx> of course
[16:24:21] <Roklobsta> yes AND SPI
[16:24:27] <Roklobsta> %&#^&^@$^%$@
[16:24:27] <twnqx> both by fuse and i think even by software
[16:24:30] <megal0maniac> HVPP!
[16:24:44] <twnqx> HVSP/HVPP are the only ones you can't disable
[16:24:44] <Roklobsta> 240VAC on reset
[16:24:44] <megal0maniac> Connect ALL of the wires
[16:24:54] <megal0maniac> Roklobsta: Yes, I think it tolerates that
[16:24:56] <megal0maniac> Oh wait
[16:25:09] <megal0maniac> Sorry, no. Only DC. 240VDC
[16:25:12] <masterq> hehe i tried to disable spi on mega, but there is a hardware programmer you can build for reparing it...
[16:25:29] <megal0maniac> masterq: Were you trying to ruin someone's day?
[16:29:09] <masterq> Roklobsta: i thing i have some where the eagle files for a stk500 clone, if you have disabled you spi interface... if you have builded it, there is no need to be afraid anymore
[16:29:16] <masterq> *think
[16:37:57] <Roklobsta> masterq: i have a dragon - and it's a bloody QFP100. I got a SMD rework station the other day, time to use it....
[16:38:17] <Roklobsta> i pissed and moaned about it last week
[16:38:38] * Roklobsta pisses and moans some more
[16:39:04] <masterq> ok, thats nice
[16:40:50] <Roklobsta> masterq: thanks for the offer, but it's way easier for me to glare at the chip and hope my disdain will flip the SPIEN fuse back.
[16:41:39] <Roklobsta> quantum mechanics says if i glare for long enough it will happen.
[16:43:49] <masterq> ^^ good luck
[16:46:44] <megal0maniac> Quantum mechanics blah blah
[16:46:50] <megal0maniac> Goodnight all :)
[16:47:10] <megal0maniac> N1njaneer: Thanks again. I'll include nap time in tomorrow's schedule :P
[16:47:39] <Roklobsta> i think the "if you wait long enough it will happen" excuse is a copout for physicists with regards to glaring holes in their knowledge,.
[16:52:00] <Roklobsta> speaking of which, i put on a lecture by lenny susskind on quantum mechanics last night to help me to go to sleep. worked like a charm.
[17:21:04] <avrdood> lol
[17:21:18] <avrdood> einstein got nicer hair
[17:27:44] <masterq> ok, i think i understood the warning now, bootloader on the avr32 is located @ 0x80000000 so i need to start programming @ 80002000h. Or i have to fill up the first 2000h bytes with something i don't need, the bootloader can not be overwritten by the bootloader
[17:28:55] <masterq> but the best think is to tell the linker the program start is @0x80002000
[17:29:09] <masterq> does somebody know how to do that?
[17:31:06] <masterq> *best thing
[17:35:03] <avrdood> aww
[17:35:10] <avrdood> you are programming in assembler?
[17:35:18] <masterq> ne c
[17:35:21] <masterq> *no c
[17:35:34] <avrdood> what are you doing
[17:35:35] <avrdood> ?
[17:36:30] <masterq> i used this code, gcc and objcopy command:
http://devzone.aery32.com/2012/07/06/how-to-install-avr-32-bit-toolchain-on-linux/ for avoiding other problems
[17:36:40] <masterq> its my first project on avr32
[17:38:08] <avrdood> right
[17:54:33] <N1njaneer> maste: You should be able to tell it where to start the section, hold a sec
[17:56:33] <N1njaneer> master: Assuming you are using gcc, try using ".txt=0x80002000" in your linker command in the makefile. That should place the code starting at 0x80002000
[17:57:13] <N1njaneer> If you are building this in AVRStudio, you can add that command in the AVR/GNU Linker Memory Settings tab under project properties
[18:00:59] <masterq> N1njaneer: Im using linux, i will try it!
[18:50:29] <ColdKeyboard> Are there some ICs that are 16-bit or something like that shift registers with serial-in parallel-out ?
[18:51:09] <ColdKeyboard> I want to drive 20 LEDs and I'm trying to avoid using 3x74xx595...
[18:57:49] <ColdKeyboard> ANy suggestions?
[19:03:10] <avrdood> digi-key.com
[19:03:26] <LoRez> minus the dash
[19:03:28] <pWNAGE> you could try an i2c gpio expander if you want to direct drive
[19:03:56] <pWNAGE> I'm sure there are 16 bit shift registers as well
[19:11:48] <masterq> i found something here:
http://www.ultimaserial.com/avr_lwip_tips7.html
[19:12:50] <masterq> now i'm able to program the device without warning, but the program is not being startet until
[19:13:00] <masterq> but i have to sleep a while
[19:13:24] <masterq> maybe see you tomorrow!
[19:27:52] <ColdKeyboard> Is it "ok" to chain 2 74HC164 by connecting Q7 of 1st IC to Input of 2nd IC if they have same clock or is better to use 74HC595?
[19:28:28] <ColdKeyboard> I'm trying to drive 20 LEDs and I want my board to be as small as possible and since 164 has less pins and it's smaller than 595
[19:28:57] <ColdKeyboard> Or there is better way to drive 20 LEDs? (I want to make 2 led bar graphs)
[20:08:17] <MrTrick> ColdKeyboard: if they're not common-anode/catchode, charlieplex!
[20:10:36] <N1njaneer> ColdKeyboard: TLC5951 -- 24 outputs, constant-current, built-in PWM
[20:11:08] <N1njaneer> I've used a single Atmega168 to run 576 individual LEDs with no multiplexing :)
[20:12:05] <MrTrick> http://i.imgur.com/uYMO8Qg.jpg?1 <-- This runs 20 LEDs, and 4 buttons, off 5 GPIO pins and nothing else except resistors.
[20:12:47] <N1njaneer> TLC5951 is also avaliable in a tiny 6x6mm QFN package
[20:13:14] <N1njaneer> MrTrick: Yes as long as you don't need smooth fading or 100% duty-cycle on any of the parts.
[20:14:39] <MrTrick> That's true N1njaneer. It's possible to run the charlieplexed array at 1/5 duty cycle though, so while there's a limit to fading, it's certainly implementable.
[20:15:05] <N1njaneer> Cuts your brightness out from TDM, but if your application can accept that then you're good to go.
[20:56:39] <MrTrick> Hmm... I have a PCB-mounted motor with a grounded case, but the case is right up against the positive part of the coin cell battery.
[20:56:58] <MrTrick> Is there any reason I shouldn't "ground" the motor to VCC instead of 0V?
[20:58:53] <theBear> erm, so long as it's isolated from everything else (like its own coils etc) and not making a whole lot of noise, it's not exactly good practice, but can't see why not... just gotta be careful it doesn't hit a grounded casing or something during use
[20:59:52] <MrTrick> well that's just it - all the stuff it might hit is at VCC.
[21:00:11] <MrTrick> The terminals are isolated from the case, and connected to an H-bridge driver.
[21:11:09] <theBear> and the metal shaft/gears aren't gonna cause any issues ?
[21:13:06] <N1njaneer> Apply kapton tape :)
[21:16:28] <theBear> is he friends with kapton kirk ?
[21:23:02] <MrTrick> theBear: they don't connect to anything, it's a pager motor.
[21:23:38] <theBear> vibrator you mena ?
[21:23:39] <theBear> mean
[21:23:47] <MrTrick> yep.
[21:24:04] <MrTrick> N1njaneer: I did consider the "duct tape" solution, but I thought that might not be so good for manufacturing.
[22:07:21] <Casper> MrTrick: as N1njaneer said, kapton tape may be a good solution, but you may want to consider 3M double sided tape too
