Back
[00:55:14] <buhman> are 8051-clones generally considered "better" than any arbitraty atmega?
[00:55:37] <buhman> (more-featureful/more-well-documented/etc..)
[00:58:35] <LoRez> hah
[01:02:00] <ziph> Anyone know of a driver that will do 5v-12v output from a 5v ttl input at >= 30mA at 5V?
[01:02:16] <ziph> (Short of a 0.5A MOSFET driver :)
[01:04:19] <CapnKernel> ziph: I'm a little confused by the >= 30mA. What is the maximum drive current you need?
[01:04:42] <ziph> 30mA is the minimum.
[01:04:54] <CapnKernel> What is the maximum? 31mA? 30A?
[01:06:06] <ziph> If you go much above 50-100mA you're in MOSFET driver territory.
[01:06:45] <CapnKernel> tl;dr
[01:08:06] <ziph> I can give you some references if you need to get up to speed with basic electronics.
[01:10:04] <CapnKernel> Lovely, that'd be great. Can you sign inside the front cover, "to the newbie, from patronising git"?
[01:10:09] <CapnKernel> ziph: You need to do a better job of specifying your requirements. I don't mind playing 20 questions with you, but it's based on the premise that you'll actually answer said questions.
[01:13:25] <ziph> Don't give offense if you can't take it.
[01:14:18] <ziph> But to be clear, I don't mind how high the maximum current output is, but I don't want to use a MOSFET driver with current sense and the other complexities that they usually have.
[01:15:15] <CapnKernel> What is the problem you're trying to solve? What would be a typical load?
[01:16:02] <CapnKernel> Would a ULN2003 help?
[01:16:20] <ziph> It's so that the driver can handle both a current hungry opto-isolator (~20-30mA at 5V) or 12V TTL.
[01:16:23] <CapnKernel> (You can gang them up for higher current if you need it)
[01:16:57] <CapnKernel> I don't know what 12V TTL is.
[01:17:20] <ziph> Neither do I, but this silly RS Components board pretends it exists. :)
[01:17:35] <CapnKernel> Which board?
[01:19:48] <ziph> This thing:
http://int.rsdelivers.com/product/rs/unipolar-stepper-motor-drive-board-2a/2173611.aspx#header
[01:20:37] <ziph> Actually it's "Open collector TTL" or "CMOS operating at 12V".
[02:17:59] <dekroning> huy guys
[02:18:38] <ziph> Hey there.
[03:42:51] <Bushman> ave
[03:43:14] <OndraSter> mornin
[04:27:35] <pc_magas> good morning
[04:31:48] <OndraSter> mornin
[05:57:54] <Roklobsta_> buhman: depends what memory model you use. the 8051 has a teeny tiny stack which has bit me in the arse before.
[05:58:43] <Roklobsta_> and 8051 code isn't the friendliest for compiling from C unlike AVR.
[05:59:16] <Roklobsta_> some 8051 chips like the ones from SiLabs run at 100MHz and have a cache.
[05:59:43] <Roklobsta_> bit it seems a bit silly. if you need 100Mhz just find a nice 16 or 32bit cpu at a lower clock.
[06:01:01] <OndraSter> don't have 51 cores low IPC?
[06:01:36] <Roklobsta_> IPC?
[06:01:45] <OndraSter> instruction per cycle... not sure if this applies to RISC cores
[06:01:46] <amee2k> instructions per clock?
[06:01:46] <OndraSter> you know
[06:01:51] <OndraSter> yes
[06:01:55] <OndraSter> cycles/clock, whatever
[06:01:57] <Roklobsta_> duh of coutse.
[06:02:04] <amee2k> seriously, if AVR or PIC or MSP doesn't cut it anymore, i would rather move up to ARM instead of 8051
[06:02:11] <OndraSter> I remember seeing some 51 core with 12 clocks per instruction ?!
[06:02:16] <OndraSter> with memory access ofc
[06:02:30] <amee2k> entry level ARM stuff has become dirt cheap over the last 5 years or so
[06:02:37] <OndraSter> ARMs are cool nowadays
[06:02:54] <Roklobsta_> yes if you use the SIlabs part with the cache then you get pretty much full speed without much or any waiting of instruction loads from flash.
[06:03:02] <amee2k> a half decent 3$ ARM will beat the fuck out of an arduino any day
[06:03:08] <OndraSter> yeah
[06:03:14] <OndraSter> STM32 Discovery is super cheap
[06:03:54] <Roklobsta_> and silabs might have made some IPC improvements on the 8051 ISA anyway.
[06:03:55] <amee2k> a friend of mine is working on a 5-7$ ARM USB stick in the segment to compete with arduinos
[06:04:19] <OndraSter> ARMduino?
[06:04:21] <Roklobsta_> amee2k: sure but arduino has so much momentum
[06:04:23] <Valen> i hope he isnt still using a ftdi ;->
[06:04:34] <amee2k> first prototype used some noname MCU, now he is thinking entry level STM series
[06:04:52] <amee2k> lol no, STM32 L1 has usb built in
[06:05:19] <amee2k> Roklobsta_: 1/4th the price and several times the performance sounds like a selling point to him
[06:06:02] <amee2k> the board really is only the MCU, usb connector, LDO and xtal and a bunch of headers
[06:06:07] <amee2k> lemme dig up the link
[06:06:13] <Roklobsta_> sure. why are arduinos so expensive? the parts must be cheap
[06:06:29] <amee2k> brb phone
[06:06:44] <Roklobsta_> teh gumbies who hear of and use arduino don't appreciate the subtlety
[06:07:35] <Roklobsta_> i have an arduino mege 128, barfed when i tried the arduino ide and went straight to using Bertos in AVS4.
[06:08:15] <OndraSter> the arduino "IDE" is just terrible
[06:08:41] <Roklobsta_> Bertos is nice, especially seeing as there is a build package for the atmega1280.
[06:09:48] <OndraSter> oh it is BeRTOS... I thought it was some IDE again... because Bert = short for Albert (at least here) lol
[06:11:13] <Roklobsta_> if you have an arduino mega definitely try out the BeRTOS kit.
[06:11:40] <OndraSter> I have only mega168 in my arduino
[06:11:47] <OndraSter> and mega128a in my project :)
[06:13:25] <amee2k> https://github.com/corecode/mchck/wiki
[06:14:01] <OndraSter> not bad
[06:14:06] <amee2k> Roklobsta_: the arduinio parts are like 5$, at least in the quantities they seem to be making these things. what is not cheap is sparkfun
[06:15:57] <Valen> my mums tv computer has a usb tuner on it
[06:16:06] <Valen> problem is the usb tuner only works ok from a cold boot
[06:16:24] <amee2k> lolusbtuner
[06:16:45] <Valen> and her computer doesn't turn power off to the usb ports when it shuts down
[06:16:56] <Valen> (aside from this issue it works well)
[06:17:30] <Valen> so i was thinking
[06:17:47] <Valen> it wouldn't be that hard to kill the usb power when the computer is off
[06:17:53] <amee2k> cheap vagoo usb stuff from china generally works well unless you try to use the more obscure features (which are frequently the ones you bought the product for, but still)
[06:17:59] <Valen> but i worry about the d- and d+ lines
[06:18:13] <OndraSter> they just couldn't make it "bipolar"
[06:18:13] <OndraSter> as in
[06:18:17] <OndraSter> d+ and d- switchable
[06:18:17] <amee2k> when the board is powered off, the data lines shouldn't be doing anything anyway
[06:18:25] <SianaGearz> the FTDI chip on the arduino is a bit expensive (and rather pointless, since it's better to have it separate from the device board)
[06:18:34] <Valen> it doesn't turn the usb off
[06:18:56] <Valen> and my understanding is that the data lines sit at some voltage
[06:18:58] <amee2k> OndraSter: you could make a custom USB cable with the power wire pulled out and wired to the PSU's normal +5V (instead of 5Vsb on the mainboard)
[06:19:11] <OndraSter> how would that help?
[06:19:18] <OndraSter> to me ?!
[06:19:27] <amee2k> err, that was supposed to be directed at Valen >_>
[06:19:30] <Valen> not solving the d+ / d- voltage > vcc issue
[06:19:44] <amee2k> i don't get why mainboards keep powering USB devices from the puny 5V standby rail anyway
[06:19:55] <SianaGearz> Valen: are you sure there is no jumper on her mainboard, labelled "USB_5VSB" or so
[06:20:02] <amee2k> Valen: if the board is powered down, D+/- isn't going to do anything
[06:20:11] <Valen> which "board" do you speak of?
[06:20:11] <CapnKernel> So you can wake up your computer with a mouse click
[06:20:12] <OndraSter> uh, USBs are powred by +5VSB?
[06:20:16] <Valen> SianaGearz: yes
[06:20:19] <OndraSter> isn't that rail like 1A?
[06:20:30] <specing_> 0.5A
[06:20:46] <amee2k> usually rated for 2A or so, but rarely does more than 0.5A from my exp
[06:21:02] <OndraSter> and during normal operation it is switched to regular +5
[06:21:08] <amee2k> either way 5Vsb usually has tiny filter caps that are grossly overloaded at 500mA already
[06:21:27] <Sgt_Lemming> evening all
[06:21:29] <amee2k> OndraSter: on the mainboard maybe, but i have yet to see a PSU to do that on its own
[06:21:36] <OndraSter> oh
[06:21:42] <SianaGearz> OndraSter: it's not specified whether USB has to be powered from +5V or +5VSB. many board simply have a 3-pin jumper allowing you to chose USB power suplly.
[06:21:52] <amee2k> maybe the 300$ ones do that
[06:22:05] <OndraSter> mine was $110
[06:22:21] <OndraSter> :P
[06:22:45] <amee2k> CapnKernel: one or two designated standby ports would do for that, but keeping all 6+ USB ports on a modern box powered off standby is ridiculous
[06:23:04] <OndraSter> only 6? :P
[06:23:14] <amee2k> "6+"
[06:23:23] <OndraSter> oh
[06:23:50] <amee2k> counting onboard headers for front panels and slots i think figures up to 10-14 ports are common by now on large boards
[06:23:53] <CapnKernel> Why? USB devices must support a standby mode, and in standby mode are not allowed to draw more than 500uA.
[06:24:11] <amee2k> the key word being "allowed"
[06:24:28] <amee2k> i seriously doubt anyone is actually enforcing that
[06:24:40] <SianaGearz> electric specifications of USB infrastructure are routinely ignored -.-
[06:24:53] <amee2k> no shit
[06:25:10] <amee2k> i blame a significant portion of the reliability issues that cheap USB gear has on exactly that
[06:25:25] <SianaGearz> well there used to be some boards enforcing maximum currents strictly, but the last i have seen was 2002. since then, puh, burn baby burn.
[06:25:56] <amee2k> most USB stuff doesn't have more than a self-resetting fuse that trips somewhere north of 500mA
[06:26:37] <amee2k> that means unless it is a dead short, it'll happily flow amps on end for a few seconds before it trips
[06:26:39] <CapnKernel> http://www.beyondlogic.org/usbnutshell/usb2.shtml
[06:26:39] <SianaGearz> oh even that would be progress. -.-
[06:27:09] <Valen> i'm suprised usb3 isn't more like 5A or something
[06:27:11] <Valen> and 12v
[06:27:27] * amee2k has seen a USB hub with the power trace blown off the board from the upstream USB connector all the way to one of the output ports
[06:28:16] <amee2k> Valen: USB was designed like 10 years ago to take over the same job that RS232 usually had back then. it was never meant for high speed or high power applications
[06:28:27] <CapnKernel> amee2k: delightful!
[06:28:28] <Valen> usb3 is rather newer however
[06:28:32] <amee2k> which imo reflects in the not very effective phy layer protocol
[06:28:43] <amee2k> Valen: USB3 has higher current limits too
[06:28:49] <OndraSter> 900mA
[06:28:51] <OndraSter> I think
[06:28:57] <amee2k> yeah, something like that
[06:29:11] <CapnKernel> USB 3 devices can't pull more than 900mA, even after negotiation
[06:29:13] <SianaGearz> it was like "haaah, want sirious bus? why, that's none of our consern, there's firewire already >.<"
[06:29:31] <OndraSter> thunderbolt!
[06:29:37] <OndraSter> dunno its power maximums there
[06:29:45] <amee2k> < CapnKernel> amee2k: delightful! << i think i had a slightly less polite way to put it, but that was the general idea
[06:29:56] <CapnKernel> Quite
[06:30:53] <amee2k> firewire is kind of like an external DMA channel from what i know, and it was made with video stuff in mind. that means sustained transfer at high bandwidth and relatively precise timing requirements
[06:32:24] <CapnKernel> Firewire is great - you can rock up to someone's machine, jack into their firewire port, and slurp their kernel memory. There is no countermeasure.
[06:33:55] <amee2k> if USB hadn't become so popular and displaced firewire, we'd probably have ended up with a combo of firewire and some low-end bus like ADB or some I2C derivative or something like that
[06:34:25] <Tom_itx> good it came along to save the day huh
[06:34:32] * inflex does like Firewire
[06:34:37] <inflex> but I see now they have Thunderbolt
[06:34:42] <amee2k> Tom_itx: not really, no
[06:34:48] <OndraSter> once tb switches to optic wiring
[06:34:51] <OndraSter> there won't be any power
[06:34:58] <amee2k> usb is a crutch on steroids
[06:35:05] <inflex> OndraSter: perhaps they'll still run power in the loom
[06:35:09] <OndraSter> hmm
[06:35:14] <OndraSter> it is Apple!
[06:35:18] <OndraSter> (just as fw tho)
[06:35:18] <inflex> the thing that botheredme with USB was always the complexity of it
[06:35:19] <amee2k> its only advantage is that it is cheap
[06:35:39] <inflex> or rather, the driver/software complexity
[06:35:58] <amee2k> yeah, the hardware is pretty dead simple
[06:36:15] <inflex> I'm sure they had their reasons... but it was a massive slap in the face tring to go from //port or serial-port programming to USB
[06:36:43] <ziph> How would it be made simpler?
[06:36:55] <amee2k> taken a look at I2C yet?
[06:37:15] <ziph> Me?
[06:37:17] <amee2k> yes
[06:37:22] <OndraSter> I wish there was simple I2C builtin interface
[06:37:23] <OndraSter> or SPI
[06:37:25] <OndraSter> or anything!
[06:37:29] <amee2k> its more than enough to run a keyboard and a mouse
[06:37:51] <ziph> It doesn't support self describing devices and multiple channels.
[06:38:01] <ziph> Add those and you've got USB level complexity.
[06:38:09] <amee2k> thats because it is a physical layer
[06:38:31] <amee2k> i2c doesn't define a protocol beyond framing and some device addressing
[06:38:56] <amee2k> there was someone proposing an i2c based desktop bus for low-speed peripherals in the 90s but i don't remember the name
[06:39:26] <amee2k> or maybe it was what became i2c when it didn't take off on the computer market
[06:39:43] <OndraSter> SMBus is (was?) for sensors in your PC
[06:39:48] <OndraSter> and it is pretty much I2C
[06:39:50] <OndraSter> just lower I Think
[06:39:53] <OndraSter> slower I think
[06:39:56] <amee2k> smbus is still quite current
[06:39:59] <amee2k> i meant an external bus
[06:40:11] <amee2k> ooh, ACCESSbus it was
[06:40:16] <OndraSter> picking up SMBus and making it into external header would work out too
[06:40:26] <amee2k> proposed by philips in the 90s
[06:40:34] <ziph> So why did you tell me to look at I2C? :)
[06:40:55] <RikusW> the video port does have an i2c interface ;)
[06:41:16] <amee2k> ziph: because A.b is basically i2c
[06:41:42] <ziph> A.b?
[06:41:46] <amee2k> ACCESSbus
[06:43:22] <amee2k> OndraSter: SMbus has stronger timing requirements than i2c, most notably slave devices aren't allowed to stretch clock cycles indefinitely to prevent hang-ups. other than that smbus is pretty much i2c as phy with a higher level protocol
[06:43:38] <OndraSter> yeah
[06:43:45] <OndraSter> that's why taking it out in header would be enough
[06:43:46] <amee2k> mixing smbus and generic i2c devices seems to work very well in most cases too
[06:46:01] <amee2k> ziph: accessbus is about as simple as they come and fit the same profile that USB was conceived for. the DDC pins in VGA connectors are actually a vestigial accessbus and some comps have made peripherals like mice and keyboards for it in the 90s
[06:46:59] <amee2k> when manufacturers started to push for USB it displaced both firewire and low-end busses like accessbus and ADB
[06:47:20] <amee2k> or PS/2, which has survived usb the longest of all i suppose
[06:48:46] <amee2k> i mean, how cool is that. atmega + I2C keyboard
[06:52:24] <ziph> Adding HID to A.b would make it about as complex as USB, minus the difficultly of having higher speeds and no clock.
[06:52:56] <amee2k> a.b *had* HID
[06:58:52] <ziph> How do you know what kind of device is attaching to the bus?
[06:59:23] <amee2k> how about bus discovery?
[06:59:33] <ziph> Hmm?
[06:59:45] <CapnKernel> ziph: Are you asking about USB?
[06:59:54] <ziph> No, A.b.
[06:59:59] <ziph> I've written USB stacks.
[07:00:14] <ziph> So did some Uni student in here, so it can't be that complex. ;)
[07:09:49] <amee2k> ziph: from what i remember, after a power up a device sends a message to register with the bus. the bus master can then query an identification structure and device capabilities
[07:09:55] <amee2k> ziph:
http://www.mcc-us.com/abspec30.htm << here are the specs if you want to know more
[07:10:21] <amee2k> its been a while and since noone is making peripherals for it i didn't dive any deeper into it
[07:20:25] <ziph> Sounds like the drivers for it would just just as complex as USB.
[07:23:17] <amee2k> you could try implementing a stack and find out :)
[07:24:17] <amee2k> i'm not trying to sell you something that has been dead for 15 years. just saying that while usb did come out as the winner, there definitely was quite a bit of competition around at the time
[07:24:49] <Valen> I wonder how firewire would go for EMC
[07:25:28] <amee2k> European Muon Collaboration?
[07:25:43] <Valen> i spose i should say linuxcnc these days
[07:26:46] <ziph> I suspect having to license from the MPEG license administration company would put a crimp on it.
[07:27:04] <Valen> people always want to use USB but the latency is too high/unpredictable
[07:27:20] <Valen> you need a licence from mpeg to use it as a generic data device?
[07:27:37] <ziph> Why do people want the PC in the servo loop?
[07:28:21] <Valen> because doing inverse kinematics on 9 axies simultaniously is easy in a $100 atom PC and hard in a $eleventy billion PLC?
[07:29:14] <ziph> More than a good FP DSP can handle?
[07:29:42] <Valen> if you want to make one that will do it and interface with the gcode generator sure go ahead ;->
[07:29:43] <amee2k> i think the important point here is the use of a comodity pc, not the number of axis
[07:30:17] <ziph> PCI cards work too. :)
[07:30:25] <amee2k> noone uses PCI anymore
[07:30:29] <Valen> btw i do literally mean a $100 PC, the intel atom boards give the best latency figures
[07:30:41] <Valen> I use PCI for the driver board in our mill
[07:30:42] <amee2k> or thats not entirely correct
[07:30:55] <amee2k> noone wants pci anymore. doesn't mean people don't use it anymore
[07:31:34] <ziph> PCI being PCI/PCIe
[07:31:41] <amee2k> ah, okay
[07:31:48] <amee2k> i thought you meant normal PCI only
[07:32:15] <amee2k> speaking of which, i've got quite a stack of old PCI cards to get rid of, if anyone wants
[07:32:43] <ben1066> which AVRs support host?
[07:32:46] <ben1066> usb host
[07:33:38] <CapnKernel> EMC2 is an amazing machine, let's face it, of 100 representative samples, how many are doing inverse kinematics on 9 axies simultaniously?
[07:33:46] <amee2k> iirc some do USB-to-go or whatever its called
[07:33:53] <CapnKernel> Most people just use it to drive their 3-axis mill.
[07:34:00] <ben1066> i need something that can handle HID
[07:34:31] <Valen> CapnKernel: so rather than use an off the shelf $100 PC people should use something else with no gui?
[07:34:36] <Valen> or nfs
[07:34:45] <Valen> or any other nifty thing a general purpose PC can do?
[07:35:29] <amee2k> people should use what gets the job done :P
[07:35:44] <amee2k> and yes, i know i have an attitude problem.
[07:36:20] <Valen> i like emc because its compeditors are mach3 a windows based thing that doesn't work too well
[07:36:22] <Valen> and well nothing
[07:36:31] <Valen> everything else is built into $100k fanuc machines
[07:36:51] <CapnKernel> The problem with EMC2 is that it conflates a GUI with the motion control. It shouldn't be necessary to use a PC for the motion control, just because one wants a GUI.
[07:36:51] <Valen> put EMC cd into computer, install and done
[07:37:05] <Valen> CapnKernel: do you know anything about emc?
[07:37:20] <amee2k> CapnKernel: so what do people use who want motion control with a neat gui?
[07:37:30] <Valen> if you want to you can run it headless and speak network commands to HAL directly
[07:37:32] <CapnKernel> I have used it a few times, and had "fun" getting it working for the first time, that's all.
[07:37:37] <amee2k> huge pile of FPGAs that implement VGA in software?
[07:37:46] <ben1066> pah ill just use an ARM uC I guess
[07:38:10] <inflex> btw, E14 has Dragons at $56 -
http://au.element14.com/atmel/atavrdragon/emulator-programmer-for-avr/dp/145508801?Ntt=145508801
[07:38:17] <inflex> going to get myself two of them tomorrow
[07:39:14] <ben1066> inflex: dragons seem alright, except they appear limited by atmel
[07:39:18] <ben1066> especially in speed
[07:39:40] <OndraSter> yep
[07:39:50] <OndraSter> bigger firmware takes ages to upload
[07:40:00] <inflex> I'm only getting them for the HV programming
[07:40:13] <inflex> most of my stuff is 1K~4K size any how
[07:40:22] <inflex> biggest I've ever used is a 168
[07:40:30] <ben1066> i think im gonna get a jtagice if any
[07:40:44] <ben1066> although 8 bit is kinda limiting me, but avrs are so nice to use I dont wanna move to arm
[07:40:47] <ben1066> :P
[07:40:58] <inflex> xmega or avr32?
[07:41:13] <inflex> though I suppose xmega is really more of a 8bit still
[07:42:03] <inflex> I really only plan to use the dragon to handle situations where I have to work on situations where I'm using the RST as an I/O
[07:48:40] <ben1066> i wish old avrs were like the tinies
[07:48:44] <ben1066> so if you used reset as io
[07:48:50] <ben1066> raise it to 12V and use standard
[07:54:40] <ben1066> can you do bga breakout on a 4 layer board?
[07:55:16] <ben1066> since the new LPC4xxx are only avaliable in BGA atm
[07:55:39] <ziph> I doubt you could do them on 4 layer board.
[07:55:53] <ziph> Use the STM F4.
[07:55:53] <amee2k> probably easier to find a device in a sane package
[07:56:34] <ziph> Or the 17xx if you don't need the M4.
[07:56:57] <ben1066> ziph: I dont really like the STM micros so much
[07:57:08] <ben1066> they cost more at the higher end
[07:57:12] <ziph> Use the LPC17xx's then.
[07:57:18] <ben1066> amee2k: they will be avaliable in LQFP...eventually
[07:57:36] <ben1066> ziph: but the reason I wanted to try was the whole M4/M0 thing :p
[07:58:06] <ziph> Get a 6-10 layer board done then. :)
[07:59:37] <ben1066> They arent as accessable
[07:59:49] <ben1066> I can get 4 layer boards done by say dorkbot maybe seed
[07:59:58] <ben1066> I know they dont do any higher :P
[08:00:47] <ziph> As soon as you have different power planes on a 4 layer you're practically back to one routing layer anyhow.
[08:01:07] <ben1066> how do you mean?
[08:01:57] <ziph> If you have more than one voltage on the board you really can only route freely on the top layer.
[08:02:01] <RikusW> ben1066: at90usb647 / 1287 support host mode aka otg
[08:03:06] <ben1066> RikusW: :D
[08:04:08] <amee2k> ben1066: well, their loss if they're too late for the party, eh?
[08:05:18] <ben1066> amee2k: how do you mean? for the packages or the pcb manafacture?
[08:05:56] <amee2k> if they don't think QFP is worth the hassle anymore and they put it on the back burner
[08:06:14] <amee2k> only natural that people who want to get on with their projects pick something else instead, no?
[08:06:38] <ziph> Yeah, it's a bit silly to have Cortex M's in such huge packages.
[08:06:51] <ben1066> amee2k: true, however unless I order from digikey and have to pay rediculous shipping excess, I cant get the lqfps
[08:07:12] <ben1066> lqfp also costrs 3x as much
[08:07:18] <amee2k> *QFP isn't exactly huge
[08:07:36] <amee2k> but BGA is ridiculous for prototyping imo, even with fabricated boards
[08:08:08] <ben1066> hmm I guess
[08:08:45] <ziph> And by the time you're in a huge BGA package you're competing with every DSP and high end CPU around.
[08:09:25] <ben1066> 256 pin >_>
[08:09:31] <ziph> The memoryless M4's NXP are putting out are the same complexity as a TI or AD DSP but much slower.
[08:12:09] <ben1066> I love the ti swebsite
[08:12:12] <ben1066> Problem: and
[08:12:44] <amee2k> i've been sorting out a huge stack of older-but-usable hardware lately... lots of network cards, video cards, optical/floppy drives, you name it. i want to see if i can still get a few bucks for it because i'd feel kinda bad for throwing it all in the dumpster, really
[08:12:54] <OndraSter> atmega8a for 1.36€
[08:12:58] <OndraSter> in tqfp32
[08:13:00] <OndraSter> ripoff
[08:13:09] <OndraSter> (other atmegas are even way, way bigger ripoff)
[08:13:13] <ben1066> avrs are overpriced these days
[08:13:15] <OndraSter> yes
[08:13:18] <OndraSter> and it bugs me off
[08:13:27] <amee2k> should i put it all up for sale as one position? or sort by type or otherwise into a a small number of positions?
[08:13:31] <OndraSter> atmega128a? 6.415€
[08:13:34] <OndraSter> and that is some supercool price
[08:13:39] <OndraSter> because normally they were like 10€
[08:14:05] <amee2k> putting every part up individually probably isn't worth it anymore because any single part is only like <5$ worth anymore
[08:14:18] <amee2k> what do you guys think?
[08:14:20] <OndraSter> woot, they have now QFM/MLF-64 version
[08:14:23] <OndraSter> which looks cooler
[08:14:27] <OndraSter> and it is even cheaper
[08:14:31] <OndraSter> too bad my boards are made for TQFP one :(
[08:15:22] <amee2k> right now i intend to put some better items up individually, like a drive rack and some video capture cards. then sell the rest as one position
[08:22:24] <RikusW> ben1066: PIC are even more expensive for slower MCU's....
[08:22:38] <RikusW> with less flash
[08:24:53] <RikusW> seems mentioning PIC is worse then arduino ;)
[08:24:59] <OndraSter> haha
[08:25:06] <CapnKernel> By not playing hard-to-get, in price terms, ARMs are eating the competition for breakfast.
[08:25:29] <ben1066> CapnKernel: Im thinking about switching, but there is no nice IDE and tools like avrs have
[08:25:35] <ben1066> the one treason I still use them
[08:25:45] <CapnKernel> my programming's in Linux, so no IDE anyway.
[08:25:53] <CapnKernel> Well, not like AS4
[08:25:57] * amee2k points at what CapnKernel said
[08:25:59] <amee2k> ^^
[08:26:03] <ben1066> and most of the IDEs are eclipse
[08:26:07] <ben1066> and I loathe eclipse
[08:26:14] <CapnKernel> YESS
[08:26:26] <ben1066> yESS?
[08:26:29] <CapnKernel> It's the new Emacs.
[08:26:34] <CapnKernel> Executable religion.
[08:26:42] <ben1066> eclipse is horrible
[08:26:50] <ben1066> its clunky, feels slow, never does what you want
[08:26:57] <OndraSter> haha eclipse
[08:26:58] <ben1066> the whole workspace system is rediculous imo
[08:26:59] <OndraSter> = pure evil
[08:27:03] <OndraSter> just as netbeans
[08:27:11] <OndraSter> but hey, whole android runs on that crap
[08:27:19] <OndraSter> on java :(
[08:28:51] <RikusW> probably why it eats the battery for breakfast....
[08:29:06] <ben1066> android doesnt use java
[08:29:09] <ben1066> it uses dalvik
[08:29:12] <ben1066> there is a difference :P
[08:29:35] <OndraSter> but dalvik is java bytecode compiler (or just interpreter, who knows)
[08:29:41] <RikusW> dalvik ?
[08:29:43] <OndraSter> I am glad my phone uses .NET
[08:29:45] <OndraSter> yes, Dalvik
[08:29:51] <ben1066> dalvik compiles further than java
[08:30:06] <ben1066> the code it reads is more optimised than java bytecode
[08:30:13] <ben1066> only the language you write programs in is java
[08:30:17] <OndraSter> it doesn't look like that :P
[08:30:25] <ben1066> it then has its own VM
[08:30:31] <ben1066> thats totally different to JVM
[08:30:33] <OndraSter> it still runs like crap, only SGS2 with its superfast CPU manages to handle it reasonably
[08:31:00] <ben1066> try running cyanogenmod
[08:31:05] <OndraSter> I rather not
[08:31:10] <OndraSter> I like my phone running Mango :)
[08:31:15] <ben1066> often the versions released by vendors are slow since they add all their crap
[08:31:24] <ben1066> Windows Phone is meh, they allow no native access
[08:31:34] <ben1066> and have very few apps
[08:31:41] <OndraSter> very few = over 60k now
[08:32:29] <OndraSter> I am waiting for MWC
[08:32:35] <ben1066> android have 400,000+
[08:32:50] <OndraSter> where 399 900 are fart apps and 899 are other crappy apps
[08:32:53] <ben1066> app store has 550,000+
[08:32:55] <OndraSter> they will hopefuly announce as much as possible from Apollo
[08:32:59] <OndraSter> even moar fart appz!
[08:33:21] <ben1066> OndraSter: your just acting like a fanboy D:
[08:33:26] <OndraSter> yao
[08:33:29] <OndraSter> yap
[08:33:32] <OndraSter> I <3 WP7 lol
[08:33:33] <ben1066> Also android and ios has steam
[08:33:39] <OndraSter> can you play steam games?
[08:33:46] <OndraSter> earn steam achievements?
[08:33:57] <amee2k> hmm is it a good idea to send PCI cards in an antistatic baggy in a bubble-wrap envelop?
[08:34:14] <RikusW> sound ok
[08:34:24] <amee2k> i mean, by mail
[08:35:08] <ben1066> No you can access the store, talk to steam friends, use the community stuff
[08:35:23] <RikusW> it might get damaged...
[08:35:28] <ben1066> android also costs less to develop for
[08:35:36] <RikusW> free sdk ?
[08:35:39] <ben1066> its a more mature platform
[08:35:39] <OndraSter> ben1066, I like my XBL where I can actually earn achievements
[08:35:46] <OndraSter> talk to friends
[08:35:57] <ben1066> xbl costs money...no?
[08:36:00] <OndraSter> no
[08:36:05] <OndraSter> only premium
[08:36:22] <ben1066> what can you do without premium?
[08:36:29] <OndraSter> no idea
[08:36:40] <ben1066> precisely...
[08:36:41] <OndraSter> I don't have xbox :)
[08:36:52] <OndraSter> so my phone is the only thing that has XBL
[08:36:55] <ben1066> so why are you using it as an example if you dont use it :P
[08:37:10] <OndraSter> because you used steam
[08:37:11] <ben1066> android is also open source
[08:37:20] <OndraSter> what does that matter
[08:37:21] <ben1066> last I checked windows phone 7 isnt
[08:37:29] <ben1066> its nice, it allows you to make your own version
[08:37:36] <ben1066> ala cyanogenmod or miui
[08:37:37] <OndraSter> I prefer just picking up the phone and it works
[08:37:51] <ben1066> android phones just work
[08:37:56] <ben1066> but your allowed to change it too
[08:38:10] <ben1066> s/allowed/can
[08:38:40] <ben1066> HTC and Sony even allow you to unlock your phones
[08:38:42] <OndraSter> oh well
[08:38:44] <OndraSter> enjoy your android
[08:38:46] <ben1066> they provide utilities to do it :p
[08:38:49] <OndraSter> I will be enjoying my WP7
[09:03:03] <ziph> Please don't say Open Source doesn't matter in here.
[09:03:26] <ziph> It's dangerous.
[09:04:34] <OndraSter> lol ok
[09:05:14] <amee2k> i think what that line is supposed to say is whether it works for the task at hand has nothing to do with the source license
[09:05:27] <amee2k> but you're welcome to start a FOSS flame war anyway, if you desire
[09:16:12] <vectory> foss software never works, is always too old and makes life living dependency hell
[09:16:23] <vectory> only beardtastic zealots promote it
[09:16:34] <vectory> amee2k: like that?
[09:18:51] <amee2k> vectory: its a start :P
[09:19:52] <amee2k> i'll counter with debian license fapping giving linux a bad name and there are a lot of distros with more reasonable versions like mint or ubuntu
[09:22:14] <vectory> welp, linux is just a unix ripoff and you better had run a real unix like osx
[09:23:26] <vectory> my washingmachine seems broken :( the noise it makes sounds unhealthy
[09:24:27] <amee2k> OSX is just a cheap BSD rip-off and not a real unix
[09:25:01] <vectory> its a fork not a ripoff
[09:25:10] <impulze> ripofffork
[09:25:13] <impulze> forkoff
[09:25:15] <vectory> darwin that is
[09:25:56] <amee2k> ripfork
[09:25:59] <impulze> yep
[09:26:00] <amee2k> satisfied?
[09:26:00] <amee2k> :P
[09:26:03] <impulze> no
[09:26:07] <amee2k> good.
[09:26:31] <rue_mohr> but as I understand thats what the bsd liscence is about
[09:26:41] <impulze> liscence :P
[09:26:45] <impulze> i've seen license and licence
[09:26:47] <rue_mohr> microsoft could use it
[09:26:48] <impulze> but not liscence!
[09:26:49] <amee2k> fapscense
[09:26:58] <eatyourguitar> license? apple paid too much for open source
[09:27:08] <eatyourguitar> but they are also selling it
[09:27:18] <impulze> so the consumers paid for it
[09:27:25] <amee2k> no, /you/ paid too much for open source with your macbook
[09:27:30] <eatyourguitar> at least they paid for linux
[09:27:36] <eatyourguitar> oh wait damn it
[09:27:41] <amee2k> well, i didn't.
[09:27:53] <eatyourguitar> apple did something evil to even the most harmless
[09:27:54] <impulze> wait, someone is eating guitars here
[09:28:13] <eatyourguitar> its a long story
[09:28:18] <amee2k> eatyourguitar: it took you this long to realize that?
[09:28:50] <eatyourguitar> name comes from eatyourguitar.com which I no longer own, it was a blog about nice hand made guitars and the people that build them
[09:29:23] <rue_mohr> I wonder how apple will obsolete all their hardware with 'the new model' next
[09:29:25] <ziph> NXP is now marketing to the dead:
http://ics.nxp.com/welcome/images/fairchild.logic.eol.gif
[09:29:36] <eatyourguitar> I am soooooo pissed at apple right now
[09:29:49] <eatyourguitar> I'm om osx 10.6 and they killed xcode 4.0
[09:30:19] <eatyourguitar> so now I can be a free developer for mac and make free apps IF I PAY SATAN FOR OSX LION
[09:30:35] <rue_mohr> thye have a history of making sure users have no input on their stuff
[09:30:51] <eatyourguitar> but then the free apps will force everyone to upgrade to lion baaaaaaah!!!!!!!!
[09:30:57] * eatyourguitar sheds a tear
[09:31:17] <rue_mohr> linux rules
[09:31:40] <rue_mohr> gcc still works on apple no?
[09:32:24] <eatyourguitar> how the fuck can they say free free free free (hahah just kidding) UP-Up-up-UPPPPPGRADE COMBO BREAKER 2,000,000 points
[09:32:40] <eatyourguitar> I'm going to eclipse
[09:33:20] <eatyourguitar> if gcc works, I want it
[09:33:27] <impulze> can#t
[09:33:29] <impulze> can't
[09:33:31] <impulze> i already have it
[09:33:38] <impulze> though i could sell it to you
[09:33:47] <eatyourguitar> I need a small C compiler to test stdio to the screen and keyboard
[09:33:53] <eatyourguitar> just text
[09:34:14] <eatyourguitar> dos or unix would be perfect but I don't have VM or anything on my mac book pro
[09:34:42] <impulze> an assembler and linker would suffice to use printf/putchar
[09:35:13] <eatyourguitar> ok so GCC for osx snow leopard?
[09:42:59] <ziph> "make somefile" where somefile.c is in the directory usually works on a Mac.
[09:49:10] <vectory> http://narf-archive.com/pix/2c2a022eea7295883d84536a40f4633147d0b0cf.jpg haha
[09:50:01] <impulze> :)
[09:50:22] <impulze> wow, narf-archive in #avr
[09:50:31] <impulze> never would've expected that, or did i post a link recently?
[09:58:02] <vectory> no, i was there all along
[10:13:29] <chupas> anyone know of a LDO reg for @ ~20mA
[10:14:14] <chupas> it would be nice to find something ~0.8 @ 5mA
[10:14:45] <ziph> What voltages?
[10:15:45] <chupas> 11.5-15
[10:15:58] <chupas> ~11V vout
[10:16:31] <chupas> I have the LM1085 in my sights
[10:16:42] <chupas> just checking to see if anyone knows of something better
[10:17:11] <ziph> I was going to suggest the Microchips because they're cheap in quantity, but I don't know if they do 11V.
[10:17:36] <chupas> Ill see what they got
[10:30:51] <ElMonkey> hi, does anyone have an AT90CAN128? or can help me with the fuse bits?
[10:31:01] <ElMonkey> the engbedded fusecalc thing isnt working right for this part
[10:32:28] <ElMonkey> my main problem is that it seems to be running at a measly 2MHz instead of 16
[10:32:54] <ElMonkey> so i thought maybe the CKOPT fuse is wrongly set
[10:32:55] <amee2k> sounds like the CLKDIV8 fuse is set wrong
[10:33:01] <ElMonkey> as a 16MHz crystal is attached to it
[10:33:09] <ElMonkey> well, thats the thing
[10:33:14] <amee2k> if that fuse is set, the clock prescaler will start out as 1/8 instead of 1/1
[10:33:16] <ElMonkey> there's no CLKDIV8 in the datasheet
[10:33:20] <amee2k> o.O
[10:33:23] <amee2k> ds linky?
[10:33:24] <ElMonkey> and no mention of any prescaler
[10:33:40] <amee2k> also, how are you measuring the clock frequency?
[10:34:01] <ElMonkey> i just measured the uart
[10:34:18] <ElMonkey> which i thought to have set to 9600 baud
[10:34:21] <amee2k> sure you got the baudrate calculations right?
[10:34:27] <ElMonkey> nope :)
[10:34:31] <amee2k> hehe
[10:34:56] <ElMonkey> oh fuck me, i'm looking at the wrong datasheet :/
[10:35:04] <amee2k> my usual trick is to set a timer pwm channel to CTC, 0xFF and toggle pin
[10:35:57] <amee2k> if i want to be sure, pull the bypass caps off, set the scope to AC coupling, then hold the probe on one of the supply pins *right* where the pin comes out of the case
[10:36:06] <ElMonkey> just get a logic sniffer
[10:36:14] <ElMonkey> no operable scope at hand, unfortunately
[10:36:26] <ElMonkey> er, got, not get
[10:36:49] <ElMonkey> anyway, now that i am looking at the right datasheet, things make more sense :)
[10:36:56] <amee2k> hehe
[11:02:18] <ElMonkey> amee2k, you were right about the CLKDIV8
[11:02:29] <ElMonkey> i was just too confused :/
[11:07:55] <amee2k> you're welcome :)
[11:10:00] <OndraSter> http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=250990211623&fromMakeTrack=true&ssPageName=VIP:watchlink:top:en
[11:10:03] <OndraSter> $9.10
[11:10:06] <OndraSter> 16 hours before end
[11:10:14] <OndraSter> it is already full price of the parts on it :(
[11:11:02] <amee2k> well, you do pay for assembly and board fab too
[11:11:20] <amee2k> so you can let it up to 9.80 without feeling bad about it :P
[11:11:29] <OndraSter> :D
[11:11:46] <OndraSter> http://uk.farnell.com/atmel/atmega128a-mu/mcu-8bit-avr-128k-flash-64qfn/dp/1773398
[11:11:49] <OndraSter> 7.5 GBP ?!
[11:11:59] <OndraSter> it is locally for 5.1 GBP
[11:12:36] <OndraSter> atmega1280 --> double that price
[11:12:38] <amee2k> hrm
[11:12:39] <OndraSter> holy s*it
[11:12:50] <OndraSter> I don't dare to open 2560
[11:12:52] <amee2k> on that arduino, is that a switching reg there on the far left?
[11:13:00] <OndraSter> no
[11:13:03] <amee2k> the SO-8 right under the xtal
[11:13:04] <OndraSter> that is mega8u2 or something
[11:13:06] <OndraSter> oh
[11:13:07] <OndraSter> nope
[11:13:09] <OndraSter> that is opamp
[11:13:14] <OndraSter> to auto-switch input power
[11:13:21] <amee2k> oh, okay
[11:13:25] <OndraSter> cool, mega2560 is cheaper than 1280
[11:13:27] <OndraSter> lmao
[11:13:34] <amee2k> lol
[11:13:49] <OndraSter> no QFP package though
[11:14:00] <amee2k> dang
[11:14:28] <OndraSter> but why ALL of these megas have like 2 - 4 UARTs
[11:14:30] <OndraSter> but only ONE SPI
[11:14:32] <OndraSter> and ONE I2C?
[11:14:34] <amee2k> i seriously need to make some pics of that USB switcher that i made
[11:14:49] <OndraSter> hmm
[11:14:52] <OndraSter> USART in PSI mode
[11:14:54] <OndraSter> SPI*
[11:15:08] <amee2k> yeah, second SPI is usually combined with some uart
[11:15:31] <amee2k> and i do think i've seen some mega with two twis, but don't quote me on it
[11:15:45] <OndraSter> hmm
[11:16:08] <OndraSter> and what about integrated somehow bank switching.. or expanding memory address bus (although that is not possible without changing the assembly... thus bank switching)
[11:16:38] <vectory> usb switcher?
[11:16:44] <amee2k> bank switching?
[11:16:56] <OndraSter> to use multiple 64kB RAM banks
[11:17:05] <OndraSter> classic stuff from older days
[11:17:18] <amee2k> there are avrs with 64k ram?? 0.0
[11:17:21] <OndraSter> 16bit address bus, but another 2 - 4 bits or more to switch banks
[11:17:23] <OndraSter> external RAM
[11:17:26] <amee2k> or anything close to that
[11:17:34] <OndraSter> XMEM
[11:17:47] <OndraSter> my project is using external 128kB RAM with bank switching :)
[11:17:47] <amee2k> if you have external ram anyway can't you just use some 7400 to gate the control lines?
[11:17:52] <OndraSter> you can
[11:18:07] <OndraSter> but it is pain to switch it amongst regular GPIO lines somehow
[11:18:15] <OndraSter> instead just writing to extra register "use this bank"
[11:18:20] <OndraSter> don't do XMEGAs have that?
[11:18:35] <amee2k> PORTx = (1<<bank); ?
[11:18:43] <OndraSter> that's one option
[11:19:01] <amee2k> or use more fancy 74xx and do PORTx = bank;
[11:19:17] <OndraSter> 154 should be hex decoder
[11:19:20] <OndraSter> 74154
[11:19:27] <amee2k> 161/162/163 are the gigital muxers i think
[11:19:32] <amee2k> or was it 15x
[11:19:42] <Bushman> hmm... ram switching, heh. i remember what was it to code for gameboy where there was only 32kb address range and yet you had to switch a whole 1024kb
[11:19:44] <amee2k> digital* lol
[11:19:48] <OndraSter> 161 is counter
[11:20:00] <OndraSter> 154 it should be IIRC
[11:20:04] <ElMonkey> alas, it works:
http://monkeyinthejungle.net/stuff/upshot/upshot_u4nIp4OB.png
[11:20:05] <amee2k> ah, then it was 15x
[11:20:34] <amee2k> nice
[11:20:40] <amee2k> ElMonkey: whats that red board there?
[11:20:53] <OndraSter> that haxing stuff
[11:20:54] <amee2k> open logic sniffer thingy?
[11:20:55] <ElMonkey> thats the open logic sniffer
[11:20:58] <amee2k> ah
[11:21:17] <amee2k> what software are you using to use it btw?
[11:21:40] <amee2k> i've been considering getting one, but all the info i could find yet was kinda meh
[11:21:55] <ElMonkey> this software:
http://ols.lxtreme.nl/
[11:22:00] <ElMonkey> first time i used it, though
[11:22:04] <ElMonkey> it kinda sucks, but works
[11:22:15] <vectory> omg, alone the clips on that table are worth ~50 bucks
[11:22:17] <ElMonkey> the hardware itself is pretty neat, for all i know
[11:22:46] <amee2k> interresting
[11:23:15] <Bushman> ElMonkey: <
http://w.nycz.bytom.pl/photo/lcd02.jpg > you too
[11:23:35] <ElMonkey> hehe, nice :)
[11:24:04] <ElMonkey> i am porting some stuff from the butterfly to this AT90CAN
[11:24:13] <amee2k> ElMonkey: one rather important question that i have about it is, whats the fastest signals it can handle while continuously recording it on the PC
[11:24:17] <Bushman> good for you
[11:24:23] <ElMonkey> which i dont even need for the CAN part, but its the only big AVR i have left
[11:24:25] <amee2k> i.e. arbitrary length until the PC runs out of memory
[11:24:40] <ElMonkey> amee2k, pretty low, i think
[11:24:48] <ElMonkey> i dont know how fast the PIC is that works the USB
[11:25:24] <amee2k> well, thats the important question, really >_>
[11:25:26] <ElMonkey> but with this software its running at 57.6k only
[11:25:45] <amee2k> o.O
[11:25:59] <ElMonkey> thats the speed to the computer :/
[11:26:24] <amee2k> isn't it supposed to do 200Msps x 16ch ??
[11:26:32] <ElMonkey> oh, nevermind, its 900k now
[11:26:36] <ElMonkey> that didnt work before
[11:26:44] <amee2k> okay
[11:26:46] <vectory> 100 MHz at 32 channels afaict
[11:26:54] <ElMonkey> yes, 200Msps but its limited to 24k samples for 8 channels
[11:27:05] <ben1066> ElMonkey: AT90? old micro :P
[11:27:26] <ElMonkey> ben1066, yea, i only have it for putting it into a car
[11:27:31] <ElMonkey> a project i didnt quite get to yet
[11:27:41] <amee2k> 24kS is like what, 100us @ 200Msps??
[11:27:44] <ElMonkey> alas, normal DIP megas were on backorder
[11:27:57] <ben1066> ElMonkey: what distributor?
[11:28:05] <amee2k> that sounds mildly useless
[11:28:26] <ElMonkey> amee2k, 120 :)
[11:28:35] <amee2k> ...
[11:29:29] <amee2k> also, my cheap ass DSO has more memory depht than that
[11:30:11] <ElMonkey> yea, well, i didnt ask any questions for 50 bucks :)
[11:30:15] <ben1066> wry u make me jealous, I still need any DSO D:
[11:30:32] <OndraSter> haha
[11:30:34] <OndraSter> so do I
[11:30:43] <amee2k> well, i do
[11:30:43] <OndraSter> and my logic analysator blew up
[11:31:02] <amee2k> and i don't think i have anything slow enough to snif with it left :(
[11:31:52] <Steffanx> lol abcminiuser's famous domain, fourwalledcubicle.com expired :)
[11:31:58] <Steffanx> * i mean :(
[11:32:15] <amee2k> Steffanx: quick, buy it!
[11:32:20] <ElMonkey> the logic sniffer was interesting to me because of the fpga, too
[11:32:28] <amee2k> and then sell it back to him for an obscenely inflated price
[11:32:32] <ElMonkey> i imagined i can hack it up if i ever need anything it doesnt do
[11:32:34] <Steffanx> Yeah amee2k
[11:34:38] <amee2k> well, its usb so it should be able to do everything
[11:34:44] <amee2k> just in software on the host computer
[11:35:19] <amee2k> but if they're stuck in full speed usb, thats not going to happen lol
[11:36:39] <ben1066> brb buying fourwalledcubicle.com
[11:37:16] <ben1066> doesnt appear to have expired
[11:37:19] <ben1066> teh site is just down
[11:37:20] <amee2k> also, pretty huge fail for using a hosting service that doesn't automatically renew domains for you :P
[11:37:50] <amee2k> Expires on............: 7 Feb 2012 19:47:13 EST
[11:37:54] <amee2k> Status................: ACTIVE
[11:37:58] <amee2k> nope, its not
[11:38:05] <OndraSter> my hosting ends 2morrow
[11:38:07] <OndraSter> I am not renewing lol
[11:39:22] <amee2k> i'm paying 4.something for a domain email http with ftp/cgi/php/blah sql and more disk quota than i'll ever see myself using
[11:39:23] <Steffanx> amee2k " Status: clientHold"
[11:39:24] <Steffanx> :)
[11:39:53] <amee2k> Steffanx: sure you've whois'd the right domain?
[11:39:56] <Steffanx> So I guess it expired, but isn't free yet
[11:39:57] <Steffanx> Yep
[11:40:11] <OndraSter> yeah
[11:40:14] <OndraSter> it takes about a month or two
[11:40:15] <amee2k> Domain Name: FOURWALLEDCUBICLE.COM
[11:40:16] <ben1066> how realistic is a solar powered wireless sensor array?
[11:40:37] <amee2k> ben1066: depending on the sensors and the desired range, quite realistic i'd say
[11:40:51] <Steffanx> FOURWALLEDCUBICLE.COM looks the same to me amee2k
[11:41:11] <amee2k> well, and depending on the location i suppose
[11:41:25] <amee2k> o.O
[11:41:31] <Steffanx> Nah, it shows to different statusses
[11:41:44] <amee2k> sounds like something is wonky with the whois service
[11:42:12] <Steffanx> I see that active status too
[11:42:13] <amee2k> what i find quite funny is that the "last update" date is tomorrow noon from my timezone lol
[11:42:30] <amee2k> Record last updated on: 8 Feb 2011 11:25:09 EST
[11:42:33] <Steffanx> 2011 :P
[11:42:40] <ben1066> amee2k: some 3.3V sensors, couple hundred meters range through 1 wall
[11:42:51] <amee2k> oh, err.
[11:42:53] <amee2k> LOL
[11:42:55] <Steffanx> Hehe amee2k
[11:42:55] <amee2k> you're right
[11:43:20] <amee2k> what time is it in EST right now anyway?
[11:43:34] <Steffanx> wolframalpha knows it all
[11:43:38] <amee2k> ben1066: sounds quite doable to me
[11:43:56] <amee2k> sorry, my penis is not long enough
[11:44:03] <Steffanx> -5 from UCT amee2k
[11:44:11] <Steffanx> -6 from CET
[11:44:27] <Steffanx> So 12:35pm :)
[11:44:29] <ben1066> Steffanx: UTC not UCT?
[11:44:30] <amee2k> hmm that would put it at 12:35
[11:44:38] <ben1066> still not right :P
[11:44:45] <ben1066> its still last updated tommorow
[11:44:50] <amee2k> so the domain is gona go out in just over 6 hours
[11:45:03] <Steffanx> It's down though
[11:45:17] <amee2k> actually, just over 7 even
[11:45:50] <amee2k> yep, doesn't resolve
[11:46:21] <Steffanx> UTC ben1066 :)
[11:46:49] <Steffanx> The order of the T and C doesn't matter, you know what i mean
[11:48:26] <amee2k> only the order of the first and last letter doesn't matter
[11:48:28] * amee2k runs
[11:48:34] <ben1066> yet we cant steal it yet >_>
[11:48:43] <ben1066> TUC?
[11:48:46] <ben1066> thats not right
[11:48:47] <ben1066> :)
[11:49:05] <amee2k> oh, did i do it wrong again? 0.0
[11:49:19] <ben1066> the U and C are still the right way around
[11:50:24] * amee2k . o O ( the art of rearranging the letters "UTC" until it reads something obscene... )
[12:36:20] <SilicaGel> is there a zigbee stack for 8-bit avrs that you can just download and use for free?
[12:41:08] <SilicaGel> maybe not
[12:41:39] <SilicaGel> since this BitCloud Pro thing says 100K of flash and 6K of ram
[12:47:12] <Jan-> It's cold :(
[12:48:02] <Jan-> it's going to be -15c tonight
[12:48:06] <Jan-> can AVRs warm me up?!
[12:48:16] <mrfrenzy_> it was -25 this morning
[12:48:31] <mrfrenzy_> only two glow plugs working, had to recharge the battery and try again...
[12:48:32] <Jan-> shouldn't that be MrFreezy?
[12:48:48] <mrfrenzy_> yeah ;)
[12:49:10] <Jan-> anyway, where on earth are you
[12:49:14] <mrfrenzy_> sweden
[12:49:25] <Jan-> Feh! well, you're used to it
[12:49:42] <Jan-> you have triple glazed windows, wall insulation three feet thick, and ethylene glycol for blood.
[12:49:57] <Jan-> the walls of this house are made out of two layers of very thin bricks.
[12:50:01] <mrfrenzy_> correct on all points ;)
[12:50:36] * Jan- force-feeds mrfrenzy_ surstromming
[12:51:02] <mrfrenzy_> I don't touch that stuff
[12:51:18] <mrfrenzy_> this is how you build a wall in Sweden
http://www.paroc.se/resources/bise_x_vagg_passivhus.jpg
[12:51:32] <Jan-> YOU DO NOW!
[12:51:38] <Jan-> muahahahah *Cram* *stuff*
[12:52:27] <Jan-> sorry, I can't see pics
[12:52:28] <Jan-> what is that
[12:53:09] <mrfrenzy_> it's a 7 layer wall
[12:53:40] <Jan-> On the other hand, it is actually worth doing there
[12:53:43] <Jan-> which it isn't here
[12:53:50] <mrfrenzy_> yep
[12:53:57] <Jan-> because here, it rarely gets to the point where the free helium starts CONDENSING OUT OF THE AIR.
[12:53:59] <Jan-> :)
[12:54:12] <mrfrenzy_> hehe
[12:54:43] <Jan-> what did swedish people do hundreds of years ago when there was no central heating :/
[12:54:53] <mrfrenzy_> they had a small fire in each room
[12:54:53] <Jan-> "Battled frostbite" I suspect
[12:55:34] <mrfrenzy_> another picture for you ;)
http://t3.gstatic.com/images?q=tbn:ANd9GcSC0IVMiyYIk8ivCmOz6ugF9mAt0wwr-la9UkDGYhF3sWRw-59dDQ
[12:56:14] <Jan-> I really can't see pics, mrfrenzy_
[12:56:28] <mrfrenzy_> I know, but maybe later ;)
[12:56:32] <mrfrenzy_> what client are you using?
[12:56:36] <Jan-> I can guess what "passivhus" means though
[12:57:08] <Jan-> client? icechat
[12:57:26] <mrfrenzy_> on a phone, or what other system that can't render images?
[12:57:35] <Jan-> I'm using a screenreader
[12:58:49] <mrfrenzy_> ooh, there was no way of suspecting here
[12:58:56] <mrfrenzy_> clever technology
[13:00:04] <Jan-> well, I'd rather be able to read the text on the screen, but there's no dance like the Got No Choice Samba.
[13:00:35] <abcminiuser> Jan-, woah, what's the internet like through nothing but a screenreader?
[13:00:48] <mrfrenzy_> I have often imagine what it would be like not seeing and all the possible technology to work around it
[13:00:58] <Jan-> abcminiuser: not half as much fun as code.
[13:01:08] <Jan-> seirously, I have no idea what blind people did before the internet
[13:01:25] <mrfrenzy_> I guess not much more than listen to books and radio...
[13:01:46] <mrfrenzy_> atleast that's what my great grandma did
[13:01:52] <abcminiuser> Jan-, you're a blink hacker? How's that work for C code?
[13:01:53] <Jan-> I used to wonder what it'd be like too, then I found out :)
[13:02:02] <mrfrenzy_> she received a tape with todays newspaper each day
[13:02:04] <abcminiuser> Is Perl just read out as white noise?
[13:02:11] <Jan-> Perl is not my favorite :)
[13:02:12] <mrfrenzy_> listened to that, then it was pretty much just radio
[13:02:23] <Jan-> mrfrenzy_, I just do the normal TV news.
[13:02:39] <mrfrenzy_> python should be pretty easy to listen to
[13:02:54] * Jan- holds the suspicion that Perl was developed to event out the wear and tear between the alphanumeric keys, and all the OTHER keys
[13:03:11] <Jan-> fact: anything typed while holding down shift and avoiding the letters is valid perl.
[13:04:04] <mrfrenzy_> all those characters must be very slow to spell out
[13:04:39] <Jan-> well, you can increase the speech speed, and it gets faster with practise
[13:04:48] <Jan-> as I say I'd rather be able to read the text, but that's life.
[13:05:34] <mrfrenzy_> perl for ($i=0; $i<10; $i++) { print i; } versus python for i in range(10): print i
[13:05:52] <Jan-> I write mainly javascript and C#
[13:06:04] <Jan-> although it looks like for microcontrollers I may have to write...
[13:06:06] * Jan- spitspit
[13:06:08] <Jan-> ...C...
[13:06:27] <vectory> asm all the way, Jan-
[13:06:29] <Kevin`> abcminiuser: you added a function to lufa to call bootloader code from application code, right?
[13:06:45] <mrfrenzy_> C# is quite nice
[13:06:54] <Jan-> well
[13:07:01] <Jan-> you know there's lots of kinds of java
[13:07:25] <Jan-> J2EE, Standard Edition, Enterprise Edition, JavaFX, etc
[13:07:31] <Jan-> I view C as JavaSensible :)
[13:07:32] <abcminiuser> Kevin`, there's example code in the LUFA manual
[13:07:33] <Jan-> er, C#
[13:08:13] <mrfrenzy_> all those government websites have these buttons "speach this text". is that really any more useful than your existing screenreader?
[13:08:37] <Kevin`> abcminiuser: ok. I may want to do something similar a project of mine, although not with usb. just want it as an example (although writing to flash is certainly nice)
[13:08:37] <Jan-> nope
[13:09:03] <mrfrenzy_> I've always suspected it was just a way to waste more government money (I work in web design)
[13:09:13] <Jan-> web designers should just do what they do
[13:09:19] <Jan-> just don't use images of text
[13:09:39] <mrfrenzy_> there are worse things to do
[13:09:47] <mrfrenzy_> like use a flash file for each button ;)
[13:09:53] * Jan- shudders
[13:10:17] <Jan-> that's evil for more reason than inconveniencing the 0.1 blind people per day who will hit it
[13:10:28] <mrfrenzy_> then label the flashfiles like button1.fl, button2.fl etc, with no alt tag
[13:10:43] <mrfrenzy_> indeed
[13:10:49] <Jan-> Oh well that's always fun
[13:10:56] <Jan-> "ButtonCvRt.gif"
[13:11:09] <Kevin`> Jan-: I probably shouldn't show you this:
http://netduino.com/
[13:11:13] <abcminiuser> Kevin`, I'd point you to my online docs, but apparently my domain has expired
[13:11:19] <Jan-> Kevin` I'm sort of aware.
[13:11:20] <mrfrenzy_> yep, lots of failed web designers around
[13:11:26] <Jan-> It seemed a bit
[13:11:27] <Jan-> er
[13:11:31] <Jan-> inefficient
[13:11:42] <Kevin`> Jan-: well of course, it's .net running on a microcontroller
[13:11:55] <Jan-> you can stand the clr overhead when you have a 3ghz 64-bit cpu with an instruction set that's documented in a 90-volume encyclopedia.
[13:12:02] <Jan-> when you have a little avr, it's less OK
[13:12:25] <mrfrenzy_> I got disappointed :( why is there no ethernet on a NETduino?
[13:12:57] <Jan-> I uhoh
[13:12:57] <Tom_itx> abcminiuser, what's up?
[13:13:02] <Jan-> I guess they use the 32 bit AVR, right?
[13:13:04] <Kevin`> mrfrenzy_: dumb hardware design. I hope they have other models
[13:13:05] <abcminiuser> or something. new line mrfrenzy_ I am a meat popsicle new line Tom_itx So am I
[13:13:24] <abcminiuser> Hey Tom_itx
[13:13:29] * Jan- sticks her tongue out at abcminiuser
[13:13:33] <abcminiuser> Hehe
[13:13:42] <Jan-> You've been watching too many Bruce Willis Milla Jovovich collaborations.
[13:13:53] <abcminiuser> Tom_itx, my domain expired yesterday, my hosting co didn't renew :(
[13:13:58] <Jan-> abcminiuser: Are you classified as human?
[13:14:00] <abcminiuser> Also, I think I'm close with the clone code
[13:14:00] <RikusW> abcminiuser: cold over there ?
[13:14:14] <abcminiuser> Jan-, sweet monkey jesus, you've see that? MULTI-PASS
[13:14:28] <Tom_itx> good
[13:14:30] <abcminiuser> RikusW, not at the moment
[13:14:34] <Jan-> abcminiuser: supergreen
[13:14:41] <abcminiuser> Tom_itx, PM for details so we don't clutter it up here?
[13:15:05] <Jan-> It's one of my favorite movies.
[13:15:06] <Steffanx> So got your domain back yet abcminiuser ? :P
[13:15:45] <Jan-> RikusW: do I want to eat boerewors?
[13:15:48] <abcminiuser> Steffanx, once my friend who actually registered it wakes up hopefully he'll re-register
[13:15:56] <Tom_itx> abcminiuser, have you tested it?
[13:16:01] <abcminiuser> Jan-, does it make sense with only audio?
[13:16:07] <RikusW> Jan-: there is many different varieties
[13:16:12] <abcminiuser> Tom_itx, I was doing some off-the-books testing today
[13:16:14] <RikusW> Jan-: its a beef sausage
[13:16:15] <Jan-> abcminiuser: well, I've actually seen, seen Fifth Element
[13:16:17] <Jan-> but yeah, it does.
[13:16:34] <Jan-> Most things do if you're willing to interpolate.
[13:16:39] <RikusW> Jan-: usually quite tasty, contains coriander too
[13:16:46] <Jan-> Anyway you can get audio description.
[13:16:55] <abcminiuser> Jan-, have you got the reader set up to give us all different voices?
[13:16:57] <Tom_itx> jan's toungueing ppl now?
[13:17:08] <abcminiuser> Because I dibs The Terminator voice
[13:17:18] <Jan-> RikusW: Our local butcher has some he's trying to sell us. I thought it sounded south african, so I thought I'd ask.
[13:17:47] <Tom_itx> abcminiuser, anything for me to test yet?
[13:18:07] <abcminiuser> Tom_itx, PM
[13:18:10] <Tom_itx> i was looking at wireshark yesterday but the usb part doesn't work on wondows
[13:18:24] <Tom_itx> or windows for that matter
[13:18:28] <Jan-> abcminiuser: no you all sound like stephen hawking
[13:18:32] <Jan-> on acid
[13:18:42] <abcminiuser> Jan-, sweet, so I sound smarter then
[13:19:01] <Jan-> Well, you don't sound Australian :)
[13:19:04] * Jan- runs away very fast
[13:19:13] <Jan-> *scoot* *dust* *pebbles*
[13:19:33] <RikusW> Jan-: so do you type on the keyboard ?
[13:19:37] <Steffanx> Tom_itx .. who needs the usb part of wireshark when you have usbsnoop? :)
[13:19:41] <RikusW> or use voice recognition ?
[13:19:43] <Jan-> RikusW: sure, what else would I do :)
[13:19:48] <Jan-> oh
[13:19:52] <Jan-> well no I just type
[13:20:27] <Jan-> most people don't look at the keyboard when they're typing anyhow
[13:20:40] <RikusW> I'd say give the boerewors a try
[13:20:48] <Jan-> Roger dodge
[13:21:10] <RikusW> not all butchers make it quite the same way...
[13:21:57] <Jan-> well basically it sounds like spicy sausage
[13:22:00] <Jan-> and that's OK by me :
[13:22:08] * RikusW occasionally peeks at the keyboard
[13:22:17] <Jan-> I bet you can all type with your eyes closed
[13:22:19] <Steffanx> You do RikusW ?
[13:22:26] <Steffanx> It's not very clean is it?
[13:23:00] * RikusW should have said rarely ;)
[13:24:07] <Tom_L> abcminiuser you around?
[13:24:23] <RikusW> my brother have a keyboard where some of the letters disappeared, other people complained to him that its difficult to type on...
[13:24:30] * RikusW had no problems :)
[13:24:33] <Tom_L> i bet he times out again
[13:25:04] <Steffanx> ping him Tom_L ?
[13:25:11] <RikusW> Jan-: we make boerewors ourself, usually we only add salt and coriander
[13:25:18] <abcminiuser> Sorry
[13:25:21] <abcminiuser> Back
[13:25:22] <Tom_L> oh
[13:25:23] <abcminiuser> :P
[13:25:30] <Tom_L> well you've been timing out alot lately
[13:25:40] <abcminiuser> I know
[13:25:43] <abcminiuser> Crappy router
[13:25:51] <RikusW> abcminiuser: what was down with your connection yesterday ?
[13:25:53] <RikusW> ah
[13:26:01] <RikusW> got it replaced ?
[13:26:11] <Jan-> Mmm
[13:26:14] <Jan-> Pizza :)
[13:26:17] <RikusW> with cisco ?
[13:26:18] <Jan-> *cram*
[13:26:58] <RikusW> other people like to add some pork too
[13:27:31] <Jan-> thounth goodth
[13:27:35] * Jan- talks with mouth full
[13:27:41] <Jan-> y'want some?
[13:27:41] <abcminiuser> RikusW, it's a shared router in the building somewhere
[13:27:46] * Jan- offers pizza to all
[13:27:50] <abcminiuser> Internet access is free, but it dies a lot
[13:28:10] <RikusW> sound irritating
[13:30:00] * Jan- hugs her GodNet
[13:30:31] <Jan-> OK time to listen to Whitesnake
[13:31:37] * Jan- http://www.youtube.com/watch?v=Gin-l4LDdXQ
[13:31:51] <asteve> hilariousness
[13:31:54] <asteve> man i hate css
[13:32:29] <Jan-> Me too
[13:32:33] <Jan-> I thought I was the only one
[13:32:43] <Jan-> not that I'm really equipped to use it very well :/
[13:32:54] * RikusW hates bulky slow webpages
[13:34:18] <Landon> hey guys, anyone have recommendations for books/reading on control systems?
[13:34:26] <mrfrenzy_> css can be very powerful if used properly
[13:35:58] <mrfrenzy_> what kind of control systems Landon ?
[13:36:23] <Landon> well, a general introduction would be ideal
[13:36:26] <Landon> but more specifically
[13:36:43] <asteve> Signals and Systems, 2nd Edition by A. O. Oppenheim and A. S. Willsky
[13:36:43] <asteve> with S. H. Nawab, Prentice Hall, 1997
[13:36:48] <mrfrenzy_> this is a good introduction to industrial control
http://engineeronadisk.com/V2/book_PLC/engineeronadisk.html
[13:36:50] <asteve> the only book you will ever truly hate with your life
[13:37:15] <Landon> I was talking to an RC car company today at the career fair and realized I don't have any projects that have any sort of control systems in them
[13:37:35] <Landon> so I'm trying to find a good starting point
[13:37:53] <Jan-> surface-to-air guided missile?
[13:37:57] <Jan-> (prison escape device?)
[13:37:59] <asteve> Landon: college or high school career fair?
[13:38:01] <Landon> college
[13:38:06] <asteve> what's your major?
[13:38:10] * Landon has graduated, still looking for some embedded work
[13:38:13] <Landon> computer engineering
[13:38:19] <asteve> nice, same
[13:38:36] * Jan- winces
[13:38:38] <Jan-> you poor sods :(
[13:38:48] <ziph> asteve: I love that book.
[13:39:02] <asteve> i'm currently employed; but in a different field and i need to switch in as soon as i get my big payout
[13:39:17] <asteve> ziph: what about that book do you love?
[13:39:58] <Jan-> Oh well good for you
[13:40:01] * RikusW have signals systems and controls printed in 1974....
[13:40:05] <Steffanx> :
[13:40:21] <RikusW> sort of ancient
[13:41:30] <RikusW> those old books have a tendency of being hard to understand...
[13:42:17] * RikusW have read a book about transistors but couldn't understand it at all, confused me, even though I know how transistors work...
[13:42:33] <ziph> asteve: It's thorough and fairly rigorously mathematical.
[13:43:05] <asteve> i think it's the opposite of thorough
[13:43:19] <asteve> i took two classes which used the book, first analog signals then discrete
[13:43:41] <asteve> it wasn't until I got to discrete that I understood analog and I blame the book
[13:44:48] * Jan- looks a bit nonplussed
[13:44:58] <Jan-> Phil just asked me if I had any M5 threaded rod
[13:45:12] <Jan-> do I look like the kind of person who'd have a foot long piece of M5 threaded rod in my back pocket
[13:45:23] * Jan- mimes pulling some out
[13:45:31] <Jan-> "Ah, yes, here we are! I got some the other day, it was on discount."
[13:45:32] <ziph> I studied it outside of a course with only differential equations as a background.
[13:45:33] <Jan-> WTF?!
[13:46:51] <ziph> I can imagine it not being very helpful for exam study.
[13:46:52] <ziph> :)
[13:46:55] <asteve> ya
[13:47:05] <ziph> His discrete book would be even worse.
[13:47:19] <ziph> (Discrete Time Signal Processing)
[13:47:58] <ziph> I can also imagine an engineer being upset that it doesn't cover control theory lingo.
[13:49:36] <asteve> i think i need to go back and try to relearn much of what i didn't :/
[13:49:54] <ziph> For electronics?
[13:50:22] <ziph> Teach yourself filter theory.
[13:51:15] <ziph> That's a fairly brutal application of it that will find holes in your knowledge fairly quickly.
[13:52:14] <Jan-> what does ziph stand for
[13:52:26] <ziph> ziphkinzelzag
[13:52:29] <Jan-> (and what's filter theory)
[13:52:30] <asteve> i fully understand FIR and bandpass filters - now applying filter theory enough to actually filter a circuit...
[13:52:31] <asteve> :)
[13:52:48] <ziph> asteve: I mean designing analog filters from scratch.
[13:52:55] <asteve> outside of placing decoupling/bypass capacitors
[13:52:56] <Jan-> People tell me FIR filters are simple, but the only explanation of them I can find involves hugely complicated equations that I don't understand.
[13:53:28] <asteve> Jan-: that's all they are
[13:53:34] <Jan-> They usually use annoying mathematical sort of gambits, along the lines of "FIR filters are extremely simple. Thus: [extremely complicated equation]. See?"
[13:54:19] <ziph> Jan-: You take an impulsive signal and put it through a filter and see what comes out.
[13:54:40] <Jan-> I had to write a filter in C# once
[13:54:49] <Jan-> I just had it average all of the recent n samples and assume that was the real value.
[13:54:55] <Jan-> I guess that's a low-pass filter.
[13:54:55] <ziph> Jan-: e.g. you take the signal 1, 0, 0, 0, 0, 0, 0 and notice that putting it through the filter 0, 0, 0, 2, 5, 6, 5, 2, 0, 0, 0, 0, 0... comes out.
[13:55:15] <ziph> Jan-: The "taps" of your FIR filter are just those values that came out.
[13:55:29] * Jan- ducks as the concept flies high overhead
[13:56:43] <ziph> Jan-: Do you know what a weighted average is?
[13:56:51] <Jan-> um, er
[13:56:54] * Jan- looks nervous
[13:57:16] <Jan-> Yes!
[13:57:18] <Jan-> LOOK AT THAT!
[13:57:21] <ziph> It's where you might take the average of say 20% of one thing and 80% of another.
[13:57:24] * Jan- points at something beyond ziph and runs
[13:57:28] <ziph> Instead of 50/50.
[13:57:40] <Jan-> Well that's what it intuitively means, I guess...
[13:57:45] <ziph> Yeap.
[13:58:13] <ziph> A FIR filter is a weighted average of (for example) the input 1, 2, 3, 4 and 5 samples ago.
[13:58:34] <ziph> So instead of taking an average of the last 5 samples, you take a weighted average.
[13:58:44] <Jan-> I just had it add five samples, divide by five, and assume that was the value.
[13:59:02] <Steffanx> :D
[13:59:11] <RikusW> thats moving average ?
[13:59:13] <ziph> You did a FIR filter then.
[13:59:14] * Jan- is probably a non-1337 h4x0r
[13:59:24] <ziph> Moving average is a FIR filter too.
[13:59:29] <ziph> It's called a box car filter.
[13:59:54] <Jan-> oooh.
[13:59:58] <Jan-> I just made it up
[14:00:01] <ziph> And it isn't as good as other filters because you can have higher frequencies that are less attenuated than lower ones.
[14:00:02] <asteve> back to this whole graduate thing; I graduated last spring with a 3.07, many of the schools I'm looking at masters for require a 3.3
[14:00:19] <asteve> man do i wish I would have retaken 3 classes to bump my average up
[14:00:20] <asteve> :/
[14:00:32] <Jan-> Whoops.
[14:00:32] <Steffanx> :P
[14:00:44] <Jan-> ziph: I knew I was an idiot :/
[14:00:57] <ziph> Jan-: It isn't all that bad.
[14:01:21] <Jan-> it's prob'ly also doable in an AVR
[14:01:28] <Jan-> if I make it say 4 or 8 samples long :)
[14:01:57] <ziph> There are better filters for things that don't have good multipliers.
[14:02:08] <ziph> Although the most popular ones actually involve the use of box car filters. :)
[14:02:22] <Jan-> Oh well
[14:02:25] <ziph> (And other tricks to improve them)
[14:02:27] <Jan-> Clearly I'm an original genius.
[14:02:42] <Jan-> I mean it won't be very linear I guess.
[14:02:50] <Jan-> but it's enough to just take jitter out of samples.
[14:03:03] <ziph> It's a linear filter.
[14:03:28] <Jan-> what do I know
[14:03:32] <Jan-> anyway I have to go make PANCAKES
[14:03:46] <ziph> One sec
[14:04:01] <ziph> http://www.bitplantation.com/transfer/BoxcarVsSimpleIir.pdf
[14:04:09] <ziph> The lumpy one is a boxcar filter.
[14:04:10] <Jan-> Later guys :)
[14:48:41] <OndraSter> I think I have found perfect box for my project
[14:48:41] <OndraSter> http://www.ges.cz/images/pictures/k/kp15.jpg
[14:48:50] <OndraSter> but I don't know the size of the curved part
[14:48:56] <OndraSter> so I will have to go to the shop and measure it
[14:49:02] <OndraSter> but it would be the pwn
[14:50:27] <Tom_itx> must be a big project
[14:51:42] <OndraSter> 180x120 are the LEDs on the front
[14:51:45] <OndraSter> plus few boards
[14:52:00] <OndraSter> 4x 10x3 boards, 1x (maybe 2x) 10x6 boards, 1x 10x8 board :)
[14:55:28] <OndraSter> http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=250990211623&fromMakeTrack=true&ssPageName=VIP:watchlink:top:en
[14:55:29] <OndraSter> $21.5
[14:55:32] <OndraSter> still half a day left
[15:24:12] <somaen> So, I'm trying to get interrupts setup on a STK1000, but I can't wrap my head around the EVBA-setup, what do I need to pass to mtsr to get it setup (that is, how do I know the address of my vector-table)?
[15:31:40] <vectory> say what? all i know is that interrupt table is defined by fusesettings, on mega8 at least
[15:32:06] <vectory> dont knopw what evba nor mtsr means
[15:33:11] <somaen> And i dont know what fusesettings is.
[15:33:51] * somaen is trying to write assembly, and finds most documentation to use C.
[15:38:57] <OndraSter> stk1000 is avr32
[15:38:59] <OndraSter> hmm
[15:43:20] <somaen> wrong channel?
[15:44:28] <OndraSter> dunno if anybody here uses avr32? :)
[15:44:41] <somaen> my bad
[15:45:25] <OndraSter> myself I'd rather go for ARM
[15:45:39] <OndraSter> AVR32 uses Eclipse, doesn't it?
[15:45:46] <OndraSter> or is it now part of AVR Studio 5 too
[15:46:18] <somaen> Dunno, I'm using the terminal-stuff on linux.
[15:46:24] <OndraSter> oh
[16:50:30] <pc_magas> Hello disi somebody manage to use avr-gc and arduino libraries?
[16:51:03] <OndraSter> only in arduino IDE
[16:51:48] <bronson> pc_magas: I looked into it but the arduino stuff isn't modular... not easy to separate out just what I wanted.
[16:52:04] <bronson> So I cribbed off it as I rewrote some pwm code.
[16:52:19] <pc_magas> Why both are using the same tools
[16:52:56] <pc_magas> ho does IDE looks and compiles the correct libraries with the code
[16:52:56] <pc_magas> ?
[16:53:35] <bronson> IDE does some source transforms too apparently. More than just #include.
[16:53:49] <bronson> But I didn't look too closely, just went straight avr.
[16:54:27] <pc_magas> So ti would be easier to make my own library
[16:54:35] <pc_magas> I thing or use the IDE
[16:55:03] <pc_magas> But that many scetches use globar variables is not a good tactick
[16:55:05] <pc_magas> tactic
[16:55:31] <pc_magas> and many scetches avoid to seperate the problewm into classes
[16:55:44] <bronson> Yep. Welcome to embedded. :)
[16:56:27] <pc_magas> Eg if I want to control a motor I wouls prefer to use a MOTOR class in order to control a motor
[16:57:04] <pc_magas> I think it is ok to have many people accesible technology to learn about embeded etc etc
[16:57:11] <pc_magas> like arduino
[16:57:43] <pc_magas> but then I think it is a good idea to leaarn how to make a sofrware with good way
[17:12:38] <Roamin> I'm so confused, why would passing a local variable, or a constant to a function work fine, but passing a global variable yields unexpected results?
[17:16:02] <jadew> maybe you forgot to set it as volatile and you're modifying it from an ISR
[17:16:33] <Roamin> volatile uint8_t abcd = 22;
[17:16:34] <Roamin> int main(void)
[17:16:34] <Roamin> {
[17:17:15] <jadew> how are the results unexpected?
[17:17:28] <jadew> you should try to confirm the value of the variable
[17:17:37] <Roamin> i did, thats whats weird
[17:17:45] <Roamin> when i do confirm it via LEDs , its correct
[17:18:19] <Roamin> the variable is passed to a function, which uses it to select an array element
[17:18:37] <jadew> is the parameter defined as volatile?
[17:18:42] <jadew> wait, that makes no sense
[17:18:57] <Roamin> parameter is define a u8int_t also
[17:18:59] <jadew> unless you're passing a pointer
[17:19:03] <jadew> I see
[17:20:45] <Roamin> its incredibly basic stuff .. 3 functions, pass parameter to 1 function, which calls another function to setup portC and D , then calls the last function that sets portb to the array element
[17:20:55] <jadew> paste some code on pastebin
[17:21:08] <bronson> Roamin: have you looked at the assembler output?
[17:22:00] <Roamin> no, i dont know enough to do that :(
[17:22:03] <jadew> are there any free programs that can simulate a circuit containing an AVR?
[17:22:27] <Roamin> i use avr studio, but it doesnt support my mcu for emulation (atmega32u2
[17:22:30] <jadew> the proteus demo doesn't allow me to start simulations with an avr in the circuit
[17:22:33] <bronson> Roamin: it actually isn't too bad if you take an hour have the AVR ISA reference next to you.
[17:22:39] <bronson> I didn't know anything about AVR assembly last week.
[17:23:23] <jadew> Roamin, yeah, I know about avr studio, but it doesn't simulate a circuit tho
[17:23:46] <bronson> (kinda wish I still didn't know because now I can see firsthand just how crappy the code generation is...)
[17:25:20] <Roamin> http://pastebin.com/AGH6ALc7
[17:25:32] <Roamin> only 1 function seems big, but its the same thing 16 times
[17:27:48] <jadew> Roamin, so where's the problem?
[17:28:21] <Roamin> calling print_letter(abcd); and calling print_letter(22); yields different results?
[17:28:36] <jadew> interresting
[17:28:54] <bronson> yea, that's unexpected. what's cap_letter?
[17:29:28] <Roamin> an array of 36 elements of 16 bytes, that represent a letter over a 16x16 led matrix
[17:29:40] <Roamin> i'll paste both headers too
[17:30:00] <Kevin`> jadew: why not just use a real avr in a real circuit with the debugging connection active?
[17:30:36] <jadew> Kevin`, I don't have debuging capabilities on my programmer and that's not the issue really, the problem is that I want to see the entire circuit working
[17:30:45] <Roamin> http://pastebin.com/TGAhZxyC thats fonts.h , its only 1 array initialized using PROGMEM
[17:31:04] <jadew> Roamin, what MCU are you using?
[17:31:23] <Roamin> atmega32u2 , programming it usb with flip, and avr studio 5
[17:31:28] <Roamin> http://pastebin.com/Ann0A44K matrix.h
[17:32:48] <jadew> is it possible that you are using up the ram?
[17:33:10] <jadew> and therefore overwriting abcd with one of the functions your calling?
[17:33:17] <jadew> it happened to me before
[17:33:17] <Roamin> shoulnd be ? unless im not using progmem properly?
[17:33:45] <Roamin> i believe i was doing something similar earlier, where the wrong letter would print , but i was initializing the array in ram instead
[17:33:57] <Roamin> switched to progmem, and the letters were correct
[17:33:58] <jadew> I noticed you're storing that part in the flash, but I don't know, maybe you're loading it or allocating too much stuff on the stack?
[17:34:30] <Kevin`> Roamin: you have your code loading the data from flash right, not just trying to grab data from the array?
[17:34:47] <Kevin`> bbl
[17:35:15] <jadew> oh right, good catch
[17:35:20] <jadew> this is not correct: cap_letter[letter][8]
[17:35:30] <Roamin> mm must i reference the array differently if i stored it in flash?
[17:35:44] <jadew> you have to load the data
[17:35:48] <jadew> from the flash
[17:36:17] <Roamin> how?
[17:36:30] <jadew> pgm_read_byte
[17:36:57] <jadew> so to read cap_letter[letter][8]
[17:37:17] <jadew> you would do pgm_read_byte(cap_letter + letter * 16 + 8)
[17:37:48] <jadew> so start pointer + letter * 16 (the letter offset) + the collumn offset
[17:39:02] <bronson> I think everyone has made that mistake before... Either that or they will.
[17:39:17] <bronson> Would be nice if gcc would detect that but I guess it's not so easy.
[17:39:34] <jadew> why wouldn't it be easy?
[17:39:41] <jadew> gcc is the one that compiles PROGMEM
[17:40:28] <jadew> actually I think the problem resides with the pointers
[17:40:56] <bronson> I remember reading an explanation on a mailing list post... it made sense at the time.
[17:40:56] <jadew> yeah, it can't be done
[17:41:05] <bronson> apparently they tried and it caused more problems than it solved.
[17:41:09] <jadew> well, the thing is that one points to ram the other one points to flash
[17:41:12] <jadew> 2 different memory spaces
[17:41:18] <jadew> so they could overlap
[17:41:37] <jadew> I suppose one solution would be to use bigger memory pointers
[17:41:50] <jadew> but that's not a socultion in the avr world
[17:42:18] <jadew> *solution
[17:42:50] <Tom_itx> evening peeps
[17:46:22] <bronson> anyone know of a quick way of detecting whether an addition overflows? for an ISR.
[17:46:53] <Tom_itx> check the flag?
[17:46:58] <bronson> if(add_with_overflow(a,b)) { overflow(); } /* adds b to a */
[17:47:01] <bronson> something like that
[17:47:16] <bronson> Tom_itx: saw a mailing list post saying that's dangerous because the optimizer can reorder
[17:47:26] <Tom_itx> oh
[17:48:20] <bronson> found a lot of posts saying what not to do but not much suggesting what to do...
[17:48:53] <Tom_itx> well dean could tell you but he fell asleep
[17:50:05] <bronson> doh. owell, I can keep using my horrible if statements until he wakes up.
[17:51:03] <Tom_itx> norway time
[17:51:17] <Tom_itx> 12:42AM
[17:51:46] <bronson> It's OK, I can wait a week or two. :) Just trying to clean things up.
[17:52:19] <Tom_itx> well i bet he's not the only one that could help
[17:53:07] <bronson> At some point I might roll up my shirtsleeves and try writing a little inline asm... that would be a hell of a lesson.
[17:54:31] <bronson> just hoping someone can shoot me a working gist and save the couple hours that would take me.
[17:55:27] <Tom_itx> http://www.nongnu.org/avr-libc/user-manual/inline_asm.html
[17:55:44] <Tom_itx> http://www.ibiblio.org/gferg/ldp/GCC-Inline-Assembly-HOWTO.html#s5
[17:57:35] <bronson> oh don't tempt me! not until I get all the features in...
[18:28:39] <grummund> bronson:
http://rotateright.com/forum/index.php?topic=33.0
[18:29:50] <bronson> grummund: that's clever. I'm using unsigned tho. I'll have to browse hacker's delight, that's new to me.
[18:30:35] <grummund> for unsigned surely it's easy... ?
[18:32:04] <grummund> sum = a + b; if (sum < a) { /* handle overflow */ }
[18:33:12] <bronson> grummund: true, it's probably not worth optimizing past that.
[18:35:39] <bronson> was hoping to get carry as a side-effect of the addition to save a register but I'd have to rewrite my whole ISR in asm to get much benefit.
[18:36:12] <grummund> do you know that gcc doesn't do that anyway?
[18:36:53] <bronson> grummund: yep, the asm is pretty ugly
[18:37:12] <bronson> I haven't tried massaging it tho, I might be making gcc's job harder than it should be.
[18:45:46] <grummund> bronson: the add and compare resolves to 4 instructions for me... and no more registers than required for the add alone.
[18:46:20] <bronson> huh, I must be screwing something up then... I'll look.
[18:46:30] <bronson> probably an accidental int or signed var or something.
[18:50:12] <Roamin> all right! using pgm_read_byte solved my problem. Thanks all
[18:51:31] <bronson> right on
[18:51:44] <buhman> Roklobsta_: thanks
[18:52:09] <buhman> Roklobsta_: is there a gcc-51? how would cross-compiling work anyway?
[18:52:29] <Roklobsta_> yes sdcc
[18:52:37] <Roklobsta_> i had used the tasking compiler
[18:52:42] <Roklobsta_> but that was my old job
[18:52:49] <Roklobsta_> i haven't used sdcc
[18:53:36] <Roklobsta_> with the 8051 you can choose a memory model like small medium and large. they use overlays and a small stack but you can't used function pointers and recursion.
[18:53:49] <Roklobsta_> but they work well with the small 8051 stack model.
[18:54:27] <Roklobsta_> i found such a memory model to be a blessing and a cure
[18:54:58] <Roklobsta_> blassing because you didn't have function pointers, curse because i nested to many functions once and my stack overloaded.
[18:55:23] <Roklobsta_> and with overlays you have to seperate out the functions you don't want overlaid like interrupts.
[18:58:39] <Roklobsta_> http://sdcc.sourceforge.net/
[18:58:46] <Roklobsta_> but hey, there must be another irc channel for that
[18:59:18] <Roklobsta_> why are you using 8051? it's crapulent.
[19:45:09] <rue_mohr> buhman,
[19:45:27] <rue_mohr> I'v been ok with sdcc so far
[19:45:36] <rue_mohr> buhman, who's 8051 are you using?
[19:51:21] <Roklobsta_> rue: whose have you used?
[19:56:21] <rue_shop> nxp and they ROCK
[19:56:27] <rue_shop> but they aren't made anymore
[19:56:36] <rue_shop> you can user std serial to flash them
[19:56:39] * bronson used the Cypress fx2 with sdcc back in 2004... it was a pretty good little setup.
[19:56:43] <rue_shop> just send it your ihex file
[19:57:46] <rue_shop> I think I found some by atmal that do the same
[20:16:24] <Roklobsta_> oh that;s nice
[20:16:28] <Roklobsta_> but dangerous.,
[20:21:19] <buhman> Roklobsta_: no I was just wondering whether it would be a good idea to play with
[20:22:09] <Roklobsta_> nope
[20:22:09] <Roklobsta_> stay away
[20:22:27] <Roklobsta_> it's still around because I htink it's so old and well known and that I think the IP for the 8051 is public domain so it's cheap to use.
[20:23:09] <buhman> are there any microcontrollers similar to atmegas that instead use x86 instructions?
[20:23:21] <buhman> (or is that even desirable to begin with?)
[20:23:56] <nevyn> there are a few SOC's built around x86
[20:24:19] <CapnKernel> buhman: What's the problem you're trying to solve?
[20:24:25] <nevyn> but yeah.
[20:24:37] <rue_shop> thye have a million pins
[20:29:28] <buhman> CapnKernel: I guess because I can't answer that, I have no business asking such questions ;P
[20:30:44] <CapnKernel> Haha, you can ask whatever you like :-)
[20:31:10] <CapnKernel> But the more info you give us, the more we can help you.
[20:32:07] <CapnKernel> For example, "I'd like to run old-school EGA DOS games" or "I'd like to make my own Linux computer from scratch"
[20:32:18] <buhman> CapnKernel: all of the above
[20:32:43] <buhman> CapnKernel: I don't really care what I do; I just want to make fun an interesting projects
[20:33:13] <CapnKernel> What kind of things have you done in the past?
[20:34:34] <buhman> CapnKernel: well, nothing too non-trivial; the last thing I did was I used a parallel DAC and played 8-bit PCM (compiled-in the program).
[20:35:24] <CapnKernel> Niiice :-)
[20:35:52] <CapnKernel> Have you played around with AVR chips before?
[20:37:39] <buhman> CapnKernel: yes I have a mega1284p and a tiny48
[20:37:51] <buhman> that's all that I've ever played with
[20:38:14] <Roklobsta_> get an xmega. that will keep you busy
[20:38:19] <Roklobsta_> or avr32
[20:38:39] <Roklobsta_> or an arm
[20:38:40] <buhman> Roklobsta_: heh, do they sell breadboardable xmegas?
[20:38:47] <Roklobsta_> ST have some nifty low pin count cortexes
[20:39:01] <buhman> when I looked, all I found were surface-mount xmegas
[20:39:14] <buhman> but the integrated DACs and such looked like fun
[20:39:20] <CapnKernel> You can get them on breakout boards
[20:39:23] <Roklobsta_> http://www.avrfreaks.net/index.php?module=Freaks%20Tools&func=viewItem&item_id=1023
[20:39:38] <Roklobsta_> for example
[20:40:03] <Roklobsta_> er you can use google?
http://shop.chip45.com/AVR-Xmega-Microcontroller-Modules-Boards
[20:40:13] <buhman> I feel like that's doing too much work for me :P
[20:40:36] <Roklobsta_> Ugh
[20:40:43] <Roklobsta_> it's all surface mount now
[20:40:59] <Roklobsta_> DIP packahes are old school and for fat fingers.
[20:41:04] <CapnKernel> Yes it's all surface mount now.
[20:41:21] <Roklobsta_> qfp packages are easy to hand solder anyway
[20:41:45] <CapnKernel> Here's an ARM chip in PDIP though:
http://www.nxp.com/news/press-releases/2011/10/nxp-cortex-m0-microcontrollers-in-high-volume-tssop-and-so-packages-target-8-16-bit-applications.html
[20:41:51] <buhman> Roklobsta_: exactly :D
[20:41:57] <CapnKernel> Memory sizes aren't great, not much bigger than AVRs.
[20:42:38] <Roklobsta_> www.bertos.org is a nice free RTOS that I have working with atmega1280 and it now support cortex and xmega with a lot of drivers written. it sort of just goes.
[20:43:01] <CapnKernel> Is it event-driven? MCU sleeps when nothing's happening?
[20:43:09] <CapnKernel> Tasks can schedule a wakeup?
[20:44:06] <CapnKernel> Haha, I probably should have looked at the web page before typing :-)
[20:44:48] <buhman> Roklobsta_: that seems like it would take away half the fun
[20:45:56] <CapnKernel> The BeRTOS website sure has a lot of link spam...
[20:46:09] <buhman> link spam?
[20:46:54] <CapnKernel> Comment spam
[20:47:13] <buhman> oh
[20:48:12] <CapnKernel> I haven't yet found out how much a license is.
[20:49:12] <nevyn> buhman: buy a protyping board. build the thing you want
[20:49:28] <nevyn> prototype your project.
[20:50:20] <CapnKernel> Sounds like buhman is still looking into what "you want" means. Hacker in search of fun project!
[20:50:51] <nevyn> ah.
[20:51:02] <buhman> CapnKernel: exactly
[20:51:25] <nevyn> pretty, functional, arty, practical, ?
[20:52:18] <nevyn> what skills do you have? what sort of thing are you interested in?
[20:52:38] <CapnKernel> buhman: You can get this special kind of sand, and you can heat it up, and grow crystals...
[20:55:48] <nevyn> actually IC's from scratch would be really fun
[20:56:11] <nevyn> even 7000 series sort of levels would be really cool
[21:08:21] <nevyn> or a transistor
[21:23:59] <nevyn> hrm rockstars can do this now..
http://www.flickr.com/photos/jeriellsworth/2835459827/in/set-72157607161498665/
[21:25:20] <nevyn> jeri ellisworth is a rockstar.
[21:28:09] <buhman> nevyn: I'm not understanding what's going on there
[21:32:40] <nevyn> buhman: it's a homemade transistor
[21:33:17] <LoRez> it's also 3.5 years ago.
[21:33:22] <nevyn> yeah
[21:34:00] <nevyn> this is true.
[21:43:45] <Roklobsta_> bertos is nice. you can run it in a preemptive mode or not.
[21:44:43] <buhman> I've been wondering: where do I get a bunch of cheap wire for breadboarding?
[21:45:12] <buhman> unfortunately I don't have any solid cat5 (already destroyed one just to find out it was stranded)
[21:45:48] <Tom_itx> well you can tell by bending it before you destroy it
[21:45:53] <Roklobsta_> just how cheap are you trying to be?
[21:46:09] <Roklobsta_> a spool of wirewrap wire is good
[21:46:17] <inflex> I always found CAT5 to be too thin any how
[21:46:25] <inflex> preferred a slightly thicker wire for bboarding
[21:46:27] <buhman> Tom_itx: I didn't know there *was* stranded cat5 before then ;P
[21:46:27] <CapnKernel> Wirewrap wire is often too thin to hold in the breadboard properly
[21:46:27] <Roklobsta_> too thin?
[21:46:29] <Tom_itx> inflex, do you use avrdude for programming?
[21:46:34] <inflex> Tom_itx: yes
[21:46:48] <Roklobsta_> oh yes, you need thick wire for the holes to grab on
[21:47:24] <Roklobsta_> how about some mains wire for lighting, it shold be solid but lower guage than main power cabling.
[21:47:29] <buhman> and cat5 isn't thick enough?
[21:47:48] <Tom_itx> inflex, what version are you using?
[21:48:00] <Roklobsta_> 5.12 is due soon, 5.11 is latest
[21:48:15] <inflex> Tom_itx: looks like 5.10
[21:48:33] <Roklobsta_> the version that comes with even the newest atmel supplied gcc is still 5.10
[21:49:25] <Roklobsta_> i have a prebuilt win32 5.11 on helix.air.net.au
[21:49:26] <Roklobsta_> but you are all linux heads.
[21:56:24] <rue_mohr> nobody wants to help me make a digital scope for everyone
[22:05:33] <Roamin> what type of mcu would you use to make a scope?
[22:06:50] <Tom_itx> avr of course
[22:07:11] <Kevin`> Roamin: what bandwidth?
[22:08:38] <Roamin> mm, kinda what i meant, you'd need a pretty fast mcu to have a decent bandwidth scope no?
[22:08:45] <Roamin> like 100mhz scope ?
[22:09:46] <Kevin`> Roamin: at 1000msps you'll need external adcs, and it may be cheaper or necessary to use external capture storage too
[22:10:17] <Kevin`> and at that point the price of the mcu won't matter, so you'd use whatever is easiest to program the ui in
[22:11:13] <Roamin> here's an idea, build a scope around a raspberri pi
[22:12:49] <Tom_itx> Addy_
[22:13:05] <Tom_itx> i hope your connection is better than the other day
[22:14:04] <Kevin`> Roamin: what FAST io does a rasbperry pi have? does it have an external memory interface for directly attaching memory mapped devices? you wouldn't want to do this over usb or spi
[22:15:03] <Landon> rue_mohr: like the xprotolab? :P
[22:19:21] <Kevin`> pci-express would work, but you probably want a parallel interface, since it's not particularly easy to design pci-express controllers
[22:22:33] <Roamin> Kevin` details are still scarce about the GPIO of the raspberri pi , but it has 16 gpio + a few other buses like i2c , the ARM runs at 700 mhz, and the board is meant to be 25$ for 128mb of ram, and 35$ for 256mb
[22:23:18] <Kevin`> i2c is way too slow, you'd be better off using usb 2.0 than that (and that isn't a good choice either particuarly)
[22:25:36] <Roamin> im just saying it has i2c, im pretty sure a lot of cool projects will come out on raspberri. Are you familiar at all with the project?
[22:26:01] <Kevin`> yeah, i'll probably get one when it's available
[22:32:38] <jadew> any idea why a pin would go high even tho I can clearly see on the osciloscope that it doesn't?
[22:32:55] <jadew> a decoupling cap seems to get it working again, but for a brief period
[22:33:13] <jadew> (I'm doing ADC on this MCU, if it matters)
[22:33:22] <Tom_itx> maybe you need a pulldown on it
[22:33:52] <jadew> hmm
[22:34:03] <jadew> I'm currently using it with a pullup
[22:34:15] <jadew> let me try
[22:34:23] <Tom_itx> what's the off state?
[22:34:38] <jadew> 1
[22:35:08] <Tom_itx> you got ddrx set
[22:35:10] <Tom_itx> i'm sure
[22:35:25] <jadew> what's ddrx?
[22:35:31] <Tom_itx> the direction reg
[22:35:35] <Tom_itx> DDRx
[22:35:37] <jadew> oh, yeah
[22:35:47] <jadew> I'm doing input on another one
[22:35:52] <jadew> oh wait
[22:35:56] <jadew> sorry, output
[22:36:03] <jadew> (kinda tired, it's 6am)
[22:36:05] <Tom_itx> i don't use the internal pullups for much
[22:36:34] <Tom_itx> and if it's breadboarded check the connections good
[22:37:18] <jadew> it's a prototyping board, good call
[22:38:02] <bronson> it seems weird that this doesn't work:
[22:38:04] <bronson> uint8 save = TCNT0; while(TCNT0 == save) { idle(); }
[22:38:38] <bronson> isn't TCNT0 declared volatile? the compiler just optimizes it right away
[22:40:15] <jadew> weird, I disabled the pullup and it does the same thing
[22:40:39] <Tom_itx> the io pin could be bad but i doubt it
[22:40:52] <jadew> it's not bad, cuz it's working in the first step
[22:41:15] <jadew> but then I do ADC
[22:41:20] <Tom_itx> check your code over good then
[22:41:25] <jadew> and I somehow that must be influencing it
[22:43:20] <jadew> this is the only write I do on that port, it shouldn't touch that pin: SOFT_USART_PORT = (SOFT_USART_PORT & ~(1 << SOFT_USART_TX)) | ((sUSART_to_send & 1) << SOFT_USART_TX);
[22:45:04] <rue_mohr> Landon, prolly not, its just an avr digital sample catcher
[22:45:09] <bronson> wow, this works... and it even generates great code:
[22:45:34] <bronson> volatile uint8_t *port = &TCNT0; uint8 save = *port; while(*port == save) { idle(); }
[22:45:38] <rue_mohr> gcc?
[22:46:10] <bronson> yep, 4.6.1
[22:46:42] <rue_mohr> gcc can, if you write the code right, generate fine asm
[22:46:54] <bronson> seems strange that I should have to force the compiler to treat registers as volatile but owell.
[22:47:17] * inflex has found quite often that the fancier you try getting in your C code, the worse the ASM becomes
[22:47:24] <bronson> now that I know this I can recognize it in the future.
[22:48:22] <bronson> inflex: except, in this case, my earlier (broken) example is less fancy...
[23:08:01] <rue_mohr> hahah I should turn this javascript calculator into an avr calculator and use up some of my led displays
[23:13:08] <rue_mohr> ....finish my scope first so I can tune pid servos on my 12' mecha
[23:13:55] <CapnKernel> rue_mohr: If there's no pictures, it didn't happen.
[23:57:21] <jadew> ok... my ADC wasn't showing a constant value so I took a look on the line and found some ringing:
http://imagebin.org/197872 also found this spike
http://imagebin.org/197873
[23:58:14] <jadew> tried to take the ringing out with a low pass filter and got this:
http://imagebin.org/197874 http://imagebin.org/197875 http://imagebin.org/197876
[23:58:36] <jadew> any clue how to get rid of it? my filter should have taken out pretty much anything above 1mhz
[23:58:54] <jadew> and what's the deal with that spike? what might generate it?
[23:59:15] <jadew> the way I feed the ADC at this point is trough a voltage divider from it's own power source
[23:59:23] <jadew> (USB)
