Back
[01:25:24] <inflex> uuuugh... hate waiting for PCBs to arrive
[01:25:29] <inflex> but less so than making them myself :D
[01:43:52] <hypermagic> hey
[02:06:21] <hypermagic> hey
[02:11:08] <DO9XE> has someone an idea, how I can check weather my DFFL in my xmega runs on the right frequency? :p
[02:12:52] <DO9XE> after some research i am thinking that it might be, that my USB has the wrong clocking.
[02:15:19] <hypermagic> possible
[02:15:46] <hypermagic> you can't reliably use an rc internal oscillator for usb, you must use crystaal to be accurate
[02:15:59] <hypermagic> (or you can calibrate your rc osc using any xtal)
[02:16:52] <DO9XE> but if I sync it with the SOF Event?
[02:17:12] <hypermagic> hmm using xmega ? ;)
[02:17:36] <DO9XE> abcminiuser, I wrote you two mails, ignore the first one ;) Its solved. I'm Hendrik
[02:17:47] <DO9XE> jipp, still xmega :)
[02:18:08] <abcminiuser> Ahoyhoy
[02:18:22] <abcminiuser> Saw them last night, haven't had a chance to respond yet
[02:18:28] <abcminiuser> Oh, neato, that takes care of that then :P
[02:19:20] <DO9XE> abcminiuser, how bad is the GenericHID example working on a xMega?
[02:21:02] <abcminiuser> Should work correctly I believe, unless I forked something up
[02:22:12] <hypermagic> i havent used xmega yet ;< i know it has pll :)
[02:22:13] <DO9XE> well, i wrote the makefile to xMega and nothing works :/ it is not recognized neither by linux, nor by windows, and wireshark shows only malformed packages :/
[02:22:27] <hypermagic> you just adjust your clock using a watch crystal, add it ;)
[02:22:44] <hypermagic> (or sync and adjust from your recved data)
[02:24:28] <DO9XE> This line should do the syncing:
https://github.com/abcminiuser/lufa/blob/master/Demos/Device/ClassDriver/GenericHID/GenericHID.c#L97
[02:25:10] <abcminiuser> Dang. I'll dig out an XMEGA later tonight and give it a burl and see what's up.
[02:25:36] <DO9XE> that would be super awesome :)
[02:25:45] <abcminiuser> Does the mouse demo work?
[02:26:15] <DO9XE> didnt test it yet. I am also using a selfmade board. just bought an xmega :S
[02:27:46] <abcminiuser> Ah, see if the other demos work first
[02:27:54] <abcminiuser> Did you set F_USB to 48000000?
[02:28:29] <DO9XE> jipp, i did
[02:31:25] <DO9XE> Do i need to set a board? i didnt saw that i need to.
[02:37:20] <DO9XE> nope, mouse Demo doesnt work either. Might there be something wrong with my hardware? I didn't found anything about anything I would need to add to the Hardware, like pullup/down resistors on D+/-
[02:37:48] <abcminiuser> The demos should work, so if all of them are horked it's likely the compile settings or hardware
[02:37:56] <abcminiuser> IIRC no external hardware required
[02:38:02] <abcminiuser> Do you have AS6 installed?
[02:38:30] <DO9XE> no, im on linux. i am using just gcc and the cloned git repo.
[02:39:16] <abcminiuser> Okiedokie, pastebin your makefile perhaps
[02:42:11] <DO9XE> http://pastebin.com/tMNQLtW7 here you are :)
[02:43:27] <abcminiuser> Yeah, that's supposed to work. Hrm.
[02:44:32] <abcminiuser> Ok I'll have a gid around for a XMEGAA3BU Xplained in a mo
[02:45:30] <DO9XE> okay :)
[03:46:30] <DO9XE> any new findings Dean? :)
[04:03:36] <Jartza> yay
[04:03:55] <Jartza> got alpha version of the audio bootloader to work
[04:04:03] <Jartza> (for attiny)
[04:05:20] <abcminiuser> Just found the board
[04:05:25] <abcminiuser> I have too much junk in storage
[04:05:48] <DO9XE> i know that :D I have like 2k of CMOS chips, just because they were cheap :D
[04:06:09] <Jartza> junk, yes.
[04:06:22] <Jartza> all kind of old crap :)
[04:07:21] <Jartza> like SN74xx and 74LSxx chips
[04:07:58] <twnqx> DO9XE: why does your nick kind of look like an amateur radio handle?
[04:08:08] <DO9XE> I think i got over 1k of 6in1 inverter CMOS chips :D
[04:08:15] <twnqx> lol
[04:08:21] <DO9XE> twnqx, because it is one ;)
[04:08:24] <twnqx> oh.
[04:08:46] <DO9XE> hihi :)
[04:09:05] * twnqx continues writing nasty macros
[04:09:10] <twnqx> __ ## #name = (value & 0x0014fff) | (__ ## #name & 0x00002000); \
[04:09:12] <twnqx> yey
[04:10:30] <twnqx> it's kinda bad that you can't put comments inside macros :(
[04:11:20] * abcminiuser is not a smart man
[04:11:36] <abcminiuser> "WHY ISN'T THE *&^(ING TOOL SHOWING UP?" "Oh, it's plugged into the wall charger..."
[04:12:04] <abcminiuser> Ok latest HEAD with the same makefile (but with xmega256a3bu) shows up fine
[04:12:26] <abcminiuser> I'll rebuild it for your device, so you can try with a known good binary
[04:12:43] <twnqx> just blame DO9XE's patched avrdude
[04:13:53] <DO9XE> twnqx, you cant :D my usart and blinking-led code works fine on the same setup :D
[04:14:01] <abcminiuser> Wait is that even building? AVR-GCC shouldn't accept "xmega64a3u", needs an "at" prefix
[04:14:24] <twnqx> DO9XE: when i uploaded code with my broken avrdude yesterday it got some data right
[04:14:36] <twnqx> enough to output from about 80% of progmem text to serial
[04:14:42] <DO9XE> well, i tryed it more than one time :D
[04:14:44] <twnqx> first 80% were garbage though
[04:15:26] <abcminiuser> http://www.fourwalledcubicle.com/files/temp/GenericHID.hex
[04:16:32] <twnqx> abcminiuser: how much would you charge for the sourcecode for a LUFA based usb-to-serial converter?
[04:16:40] <twnqx> i only found arduino code online, and i am lazy...
[04:19:21] <abcminiuser> Like the one already in the LUFA projects/USBtoSerial folder?
[04:19:29] <DO9XE> :D
[04:20:37] <DO9XE> nope, your compiled Hex is not working either :/ I have my 3V operatingvoltage. booting my oszilloscope to measure something on D+&D-
[04:20:54] <twnqx> abcminiuser: ummm... is my lufa too old then...
[04:21:12] <twnqx> sorry.
[04:21:56] <abcminiuser> Must be super mightily old then (>3 years or so)
[04:23:27] <twnqx> even for usbkey, hmmm
[04:23:39] * twnqx eyes usbkey board left of the keyboar
[04:23:40] <twnqx> d
[04:24:29] <twnqx> just sad that i never found useful connectors for these weird 1.27/2.54 combo pins
[04:24:36] <twnqx> (well, never really looked either)
[04:24:42] <DO9XE> abcminiuser, there aren't even signals on the bus :/
[04:24:58] <twnqx> you're doing something wrong
[04:26:46] <abcminiuser> Yeah I have some female breakout boards in a box somewhere, but it was a silly design choice
[04:26:53] <abcminiuser> Made the board smaller, but also unusable
[04:27:06] <DO9XE> i fixed some points in LUFA, which occured odd to me. this made the computer at leas recognize that there is something arround. For example i change a calculation to a fixed value, because the xmega datasheet said so.
[04:28:20] <DO9XE> and i took out the device specific definitions, e.g. fot the clock settings.
[04:30:27] <abcminiuser> Might be that I've missed some family specifc config code; IIRC there were a few community patches around the XMEGA clock setup code as they claimed it wasn't working right on some of them
[04:30:45] <abcminiuser> I only have an A3BU for easy testing, didn'r want to loot Atmel when I left :P
[04:31:26] <twnqx> i should just change the usb serial to my leonardo clone hardware
[04:31:35] <DO9XE> hmm, then I have to look for some other clock setting patches.
[04:31:37] <twnqx> (which is full of bad design choices as well :P)
[04:34:39] <twnqx> the questions when writing a cpu emulator... "does a software generated irq trigger if, in a single transaction, irq generation is disabled AND triggered"
[04:35:38] * twnqx puts a // TODO
[04:35:53] <DO9XE> abcminiuser, do you have any patch in mind of which you can tell me the link to? :)
[04:37:27] <abcminiuser> https://github.com/abcminiuser/lufa/commit/c4e85215bff8363ae9a9a79581cc3e7427b597c7 was one
[04:37:57] <abcminiuser> And
https://github.com/abcminiuser/lufa/commit/c499a0b7550ba26e4b26e11ae13935f094f22ddc
[04:41:48] <DO9XE> the first commit was the point were i put in a fixed value. i am looking for the second one, weather it is in the code
[04:43:49] <twnqx> the weather? too cold :(
[04:47:37] <Tom_itx> hey dean...
[04:47:49] <Tom_itx> you shouldn't be a stranger!
[04:48:06] <DO9XE> still this
http://pastebin.com/PFx5xL1U :(((
[04:48:15] <DO9XE> he istn't ;)
[04:49:53] <Tom_itx> i gotta remind him before the Mrs starts taking _all_ his time
[04:57:34] <abcminiuser> Ahoyhoy
[04:57:38] <abcminiuser> Yeah it's been a while
[05:04:00] <DO9XE> hmm, lunchbreak. im not allowed to stay in the office, safety terms and stuff :P see you in 45 minutes max. ;)
[05:38:12] * DO9XE is back from lunch :)
[05:43:34] <LeoNerd> The AVR Studio part info XML files don't seem to give the default values of the fuses
[05:43:39] <LeoNerd> This is upsetting.. :/
[05:45:51] <DO9XE> abcminiuser, i think i will now calculate the clock registers and set them directly via hex values :)
[06:03:50] <RikusW> LeoNerd: just mailed you a grep from my parts files with default fuse values
[06:04:25] <RikusW> indeed the xml files seems not to contain it, I had to add it manually
[06:04:47] <LeoNerd> Ahhh... tricky
[06:04:50] <LeoNerd> OK
[06:35:41] <DO9XE> is the 2MHz RC Oscillator switched on by default on xMega devices? seems so...
[08:09:02] <DO9XE> hmm... still nothing -.-#
[08:09:47] <DO9XE> I've set the clocksettings like they should be in the datasheet, but, theres nothing coming from the device :P
[08:10:18] <Jartza> what device
[08:16:40] <vsync_> Has anyone ever used murata's bluetooth chips?
[08:23:47] <vsync_> or does anyone know of a chip with a smaller footprint than the RN42 which is 25.6 x 13.4 mm
[08:24:40] <vsync_> murata has one that is 7 x 7.4 mm, with a built-in antenna, however the datasheet is cryptic at best
[10:41:44] <Jartza> https://pbs.twimg.com/media/B6rL0ZMCEAAvb9n.jpg:large
[13:57:15] <vsync_> Mehh.
[14:01:14] <twnqx> indeed.
[14:01:21] <Tom_itx> poo
[14:02:11] <Thrashbarg> snuh
[14:03:51] <vsync_> the electronics channel proved its usefulness once more
[14:04:15] <Thrashbarg> is that a hint of sarcasm I detect?
[14:04:20] <vsync_> hey guys, wanna talk about linux?
[14:04:34] <vsync_> Thrashbarg: there's quite ways more than a hint
[14:04:45] <Thrashbarg> :)
[14:30:58] <vsync_> bluetooth seems to be a pain in the ass
[14:31:32] <vsync_> i am really tempted to go for the RN42 but the footprint is quite big
[14:31:43] <vsync_> i am dealing with minimum amount of space :(
[14:32:23] <vsync_> panasonic and murata offer a really tiny ble chips but afaik you need to write your own programs and drop them into them via jtag
[14:32:29] <vsync_> -a
[14:33:20] <twnqx> checked the brits... crs or something?
[14:35:24] <vsync_> huh?
[14:35:32] <twnqx> for bluetooth chips
[14:35:37] <twnqx> or csr
[14:35:41] <twnqx> cambridge something
[14:37:18] <vsync_> Oh, let's see. Thanks!
[14:39:08] <vsync_> Oh. They seem to have small solutions aswell. They also require you to write your own stack I think
[14:39:35] <twnqx> not sure which parts of that they have out of the box
[14:39:41] <twnqx> but i think they have some
[14:39:52] <vsync_> all i need is a simple uart off the chip. The rfduino actually seems pretty good, but the upshot is that it lacks IO for my application
[14:40:23] * twnqx checks chinese serial over bluetooth board
[14:40:26] <twnqx> check, CSR chip :P
[14:43:57] <vsync_> hmm
[14:47:04] <vsync_> i checked the CSR's, they require an external flash. The chinese ones just have a UART implementation I suppose
[14:47:29] <vsync_> it means the footprint is roundabout the same as the RN42, which is well documented
[14:47:49] <vsync_> thanks anyway, much obliged :)
[14:48:40] <vsync_> the rfduino looks very, very sexy. Let me see just how many IO i would need
[15:55:47] <vsync_> nrf8001 seems perfect
[15:56:21] <vsync_> getting nitty gritty with tracing a copper antenna will be a pain in the arse though
[19:25:45] * LeoNerd hunts SMT multi-channel optoisolators so he can hack his dishwasher
