Back
[02:27:22] <inflex> lo folks
[02:27:31] * inflex finishes submitting yet another design for protoing
[02:32:18] <soul-d> im to lazy to even fan wire out 2 level converter chips to some pin headers maybe after grocery shopping :)
[02:52:11] <inflex> soul-d: don't feel bad - none of my stuff is actually in the workshop as such - it's just stuff I can do sitting here being a big fat f*ck
[02:54:54] <soul-d> yeah i could do it just to lazy still need to figure out what pcb layout tool to use ( linux ) but also waiting a bit need to reinstall os for change and new release is overdue
[02:59:10] <ziph> inflex: Want to see my latest board?
[03:03:19] <inflex> sure
[03:05:10] <ziph> Can I pm you?
[03:07:00] <inflex> oh sure
[03:12:35] <h4x0r`> aw
[03:12:42] <h4x0r`> now i wanna see it 2
[03:12:45] <h4x0r`> :/
[03:12:53] <inflex> heh
[03:21:06] <ziph> How do I know that you aren't my main competitor? :)
[03:24:54] <Valen> i wanna see it too
[03:25:15] <Valen> inflex: you should do gyprocking
[03:25:23] <Valen> or remove airconditioners from walls
[03:25:26] <inflex> oh god I hate that stuff
[03:25:30] <Valen> great way to fill in time ;-P
[03:25:49] <Valen> more or less than gluing fans into northbridges of your mothers computer
[03:25:54] <ziph> Or you could organise to install pink batts and not install them and get paid anyhow. ;)
[03:26:06] <inflex> haha
[03:26:20] <Valen> could do what the indians did at my dads place
[03:26:35] <Valen> rip the batts in half and only put them around the manhole
[03:26:47] <Valen> they turned up with 2 bags
[03:27:03] <Valen> he went up outside the roof and lifted some tiles off
[03:27:33] <Valen> they had to go and get another 10 bags before they had actually covered the roof to the right depth
[03:27:43] <inflex> lolz
[03:27:52] <Valen> also took them 8 hours rather than the 2 they had booked in
[03:30:11] <Valen> I think somebody worked out that if all the eligible places were insulated
[03:30:31] <Valen> it would equal about 4 days worth of emmisions from one power station in victoria
[03:30:42] <Valen> in terms of savings
[03:56:54] <lovre> hi all
[03:58:09] <lovre> i could use a bit of OT help... searching ebay for coppered and varhished PCB boards for creating my own avr dev board, but i dont know how to find those, if there even are some. All i can find is coppered ones, not varnished ones (if that is the correct term, english is not my native language)
[03:58:36] <Valen> what are you trying to do?
[03:59:04] <lovre> just get a pcb and light it, etch it, produce my board...
[03:59:26] <lovre> how can i search ebay for this, i am not familliar with english terminology for this
[03:59:36] <Valen> usually called resist
[03:59:47] <Valen> though most people do toner transfer now
[03:59:53] <Valen> or photo-resist
[04:02:39] <ziph> Most sane people to photo-resist.
[04:03:28] <ziph> lovre: Got $20-40 bucks?
[04:03:30] <mrfrenzy> you can get the boards from wherever you buy components. farnell, digikey etc
[04:04:11] <lovre> ziph: 20-40 bucks for one board, or a pack :)
[04:04:29] <inflex> no, sane people just get a PCB service to do the boards, and they wait 2 weeks.
[04:04:42] <ziph> lovre: The road to getting good results with photo-resist is long and expensive.
[04:04:47] <ziph> lovre: You're better doing this:
http://dorkbotpdx.org/wiki/pcb_order
[04:05:25] <inflex> hrmm... itead or seeed is a bit cheaper
[04:05:42] <Steffanx> ziph unless they fuckup the pcb
[04:06:06] <ziph> Steffanx: Oh?
[04:06:09] <mrfrenzy> if they fuckup they will just have to redo it ofcourse
[04:09:16] <lovre> so whats the best way to make my pcb at home?
[04:10:05] <Valen> we do toner transfer and milling
[04:10:11] <Valen> milling is better
[04:10:18] <ziph> Seriously?
[04:10:24] <mrfrenzy> photo-etch is best
[04:10:27] <Valen> toner transfer has higher detail, but its fussy
[04:10:37] <Valen> milling is easy
[04:10:44] <mrfrenzy> milling is only for low detail, and toner transfer is fussy as you said
[04:10:53] <lovre> i was thinking more in lines of hobbyst-affordable-low-cost :D
[04:10:54] <mrfrenzy> photo-etch makes good boards once you learned it
[04:11:02] <ziph> lovre: You can buy about 600 commercially made boards for the cost of one home set up. :)
[04:11:15] <Valen> i don't know about "low detail"
[04:11:29] <Valen> tqfp is pretty easy
[04:11:38] <Valen> its only those 3mm dfn's that really are hard
[04:11:48] <Valen> well, impossible lol
[04:12:02] <Valen> http://www.vapourforge.com/jake/temp/IMAG0117.jpg
[04:12:13] <Valen> pcb for atmega32u4
[04:12:20] <mrfrenzy> can you mill a board for an xmega128?
[04:12:29] <Valen> we also did all the drilling and a double sided board
[04:12:49] <Valen> what package are they?
[04:12:55] <mrfrenzy> 100TQFP
[04:13:04] <Valen> probably
[04:13:14] <mrfrenzy> ooh, I didn't think that was possible
[04:13:19] <Valen> even made a solder mask
http://www.vapourforge.com/jake/temp/IMAG0135.jpg
[04:13:28] <mrfrenzy> then it is a very good solution (if you have a milling machine) :)
[04:13:47] <Valen> well we did 44tqfp which is the 32u4
[04:14:02] <Valen> so if the pin spacing is the same it should be doable
[04:14:09] <Steffanx> No we didn't Valen :P
[04:14:14] <Steffanx> You did
[04:14:28] <Valen> we is my company
[04:15:20] <Valen> not doing the bga package though
[04:16:29] <mrfrenzy> do you have specs of your machine somewhere?
[04:16:49] <Valen> http://www.cnczone.com/forums/general_metal_working_machines/118358-phenolic_basalt_head_our_hm45.html is the head we made for it
[04:17:06] <mrfrenzy> and yes, for soldering tqfp you pretty much want solder ask
[04:17:12] <mrfrenzy> cool, will checkout later
[04:17:14] <mrfrenzy> now work
[04:17:24] <Valen> sorry not mask, a stencil
[04:17:32] <Valen> solder mask would be really handy
[04:29:54] <ziph> Valen: Any backlash in that thing? :)
[04:48:25] <karlp> mrfrenzy: we do milling here, and it's fabulous, we use it _for_ the fine detail.
[04:48:30] <karlp> but we do have a fairly nice mill :)
[04:52:42] <jacekowski> i went to flour mill on monday
[04:52:50] <AnalogSound> hello all
[04:52:52] <jacekowski> they do milling there as well
[04:52:58] * karlp grins
[04:53:05] <jacekowski> they use it to make fine flour
[04:53:37] <AnalogSound> :))
[05:01:41] <karlp> Valen: that's a much more serious mill you have there!
[05:02:52] <Valen> ziph: none that we can measure
[05:03:08] <Valen> we have linear glass scales on the axies and use that for the feedback
[05:03:22] <Valen> so if there is any backlash it makes the control loop go mental
[05:03:33] <ziph> Ah.
[05:04:23] <Valen> so there is less than .001mm (resolution of the scales) lash in it lol
[05:04:40] <ziph> You should do online panel manufacturing with it. ;)
[05:04:55] <Valen> probably not enough $ in it really
[05:05:10] <Valen> and there is no soldermask or plated through holes
[05:05:20] <Valen> but if you want a board made urgently ;->
[05:05:23] <ziph> Oh, I mean enclosure panels.
[05:05:30] <Valen> oh that, sure
[05:05:38] <Valen> we can make the enclosure too if you want
[05:15:37] <ziph> Mmm, 50MHz SPI.
[05:16:16] <inflex> O_o
[05:16:30] <inflex> well, I took a big fat gamble and submitted my experimental board for production
[05:16:49] <ziph> That's what the poker players do when they're low on chips. :)
[05:17:06] <inflex> experimental because I'm not used to dealing with negative regulators, let alone ones being 'misused' for odd purposes... and it's a board that has to handle 50A and 40V
[05:18:29] <inflex> damn, just realised, that's 2kW
[05:18:50] <ziph> Planning on baking cookies with it? :)
[05:19:11] <specing> inflex: I can smell it burning already...
[05:21:29] <inflex> ziph: mmm.. yah, might have to think carefully about it
[05:21:39] <h4x0r`> gimme a yell when those cookies are ready lol
[05:21:45] <inflex> ziph: esp since the shunt resistor I'm planning on using will have to dump ~3W
[05:21:54] <inflex> but I'm a tad worried about the copper
[05:21:57] <Valen> heh we have done 600A stuff before
[05:22:07] <Valen> but with bus bars not PCB
[05:23:05] <inflex> http://dxp.me/i/baker.png that's it there
[05:23:19] <inflex> the +ve side goes through the top-left corner
[05:24:10] <inflex> the board is only 25 x 50mm
[05:24:38] <Valen> whats the minimum amps inflex?
[05:24:45] <inflex> I did at least mask off the large area for the wires, so at least I can flood it
[05:24:48] <inflex> Valen: zero
[05:24:49] <inflex> :)
[05:24:53] <Valen> whats it do?
[05:24:59] <inflex> counts mAh consumed
[05:25:16] <Valen> have you looked at the alegro parts?
[05:25:18] <inflex> tracks voltage, peak current... etc - it's a "big" version of the little mAh counter I did before
[05:25:27] <inflex> the hall-sense onces?
[05:25:28] <inflex> ones
[05:25:30] <Valen> yeah
[05:25:35] <inflex> yeah, I've looked at them
[05:25:37] <Valen> up to 100A and no resistance
[05:25:48] <inflex> *nod*
[05:26:06] <Valen> well not much anyway
[05:26:06] <inflex> push come to shove I might limit this to 30A, which still covers a fair bit of the market
[05:26:18] <inflex> I mean, the shunt I'm using is only 1milliohm
[05:26:43] <inflex> so, that's 2.5W dissipated at max power... and the shunt is 3W rated
[05:27:41] <inflex> I might have to actually make people do a 'tap' on the -ve side, since I don't need a lot of current for that
[05:27:59] <Valen> only concern is its 2.5W disipated into a padded foam wad stuffed ontop of the batteries ;->
[05:28:07] <inflex> else I fear that there just won't be enough PCB area for anything more than about 14~16AWG wire
[05:28:18] <Valen> stick a bunch of vias through that pad too I'd suggest
[05:28:35] <Valen> ahh single sided, nvm
[05:28:45] <inflex> well, there's nothing much on the other side
[05:29:04] <inflex> there's a flood plane but that's for the virtual-ground (Vcc -3.3V)
[05:29:22] <inflex> which is why there's a busshell of vias under the LM337 reg
[05:29:24] <Valen> why do you have a negative rail thing going on anyway?
[05:29:46] <inflex> because for what ever reason, the coulomb counting chip prefers to be on the "high side"
[05:29:59] <inflex> imo technically I don't see why... but I can't get any additional spec out of the datasheet
[05:30:13] <h4x0r`> it perfers floating?
[05:30:17] <inflex> maybe it's something to do with improving the CMMR
[05:30:19] <h4x0r`> prefers*
[05:30:39] <h4x0r`> floating signals are unreliable
[05:30:48] <h4x0r`> but not sure if thats what u meant
[05:31:03] <inflex> Anyhow, I hope it works, else that's a whole $14 I wasted!
[05:31:07] <Valen> lols
[05:31:09] <inflex> that's 5 PCBs this week... damn
[05:31:16] <h4x0r`> heh
[05:31:31] <Valen> if your using a sense resistor anyway just hook it up to the 200x amp in the AVR ;-> made of win
[05:31:42] <h4x0r`> lol
[05:32:01] <inflex> cept there's none in the mega?8 series
[05:32:23] <Valen> lesson is to use a real mans AVR ;-P
[05:32:35] <inflex> the charge-counter chip is a proper integrating type, as oppsed to a sample-and-assume type
[05:32:37] <Valen> stick a USB port on it and have datalogging
[05:32:37] <inflex> that's why I'm using it
[05:32:51] <inflex> that's not going to be much use in-flight ;)
[05:32:59] <inflex> besides, I'll leave that for the more advanced ones
[05:33:06] <Valen> fairynuff
[05:33:37] <inflex> I put in a servo logger this week too
[05:33:48] <inflex> bigger than I wanted but at least it might do a nice job
[05:37:26] <inflex> Anyhow, I've get ~10 boards in each batch, so I'll try at 50A and see what happens
[05:37:33] <inflex> ... probably from behind a blast barrier
[05:37:55] <inflex> can just imagine there being a fault in the solder reflow and being hit with 50A at high resistance... melt and boom
[05:38:31] <grummund> inflex: out of interest what IC is that?
[05:38:54] <inflex> grummund: LTC4150
[05:39:27] <grummund> ta
[05:41:11] <Valen> i have found shorts in PCB's that way
[05:41:11] <Valen> step 1, remove $100 ADC chip
[05:41:11] <Valen> step 2, apply car battery to output of voltage regulator
[05:41:11] <Valen> step 3 extinguish smoking holes in desk
[05:41:12] <Valen> step 4, extinguish tantalum cap on board
[05:41:13] <Valen> step 5 replace cap and voltage regulator
[05:41:13] <Valen> step 7 profit
[05:41:50] <inflex> damned tantalum caps
[05:41:54] <inflex> they're always naughty
[05:49:54] <h4x0r`> how so lol
[05:55:52] <inflex> they're so sensitive
[05:55:59] <inflex> "Boo hoo, I don't like that ripple... *BANG* "
[05:56:19] <inflex> or "WAAAAAH! you dare put even the slightest wrong polarity on me ... BIG BADDA BOOM"
[06:09:46] <h4x0r`> anyone know if avr's support ts-ms or etsi protocols?
[06:09:50] <h4x0r`> cnat find in datasheet
[06:10:11] <karlp> support what?
[06:10:37] <h4x0r`> gsm mutiplexer protocols
[06:10:44] <h4x0r`> its a form of logic
[06:11:00] <grummund> ho ho you're havin a laugh
[06:11:19] <h4x0r`> atmega32 apprently does
[06:11:30] <h4x0r`> whos laughin?
[06:11:34] <grummund> you
[06:11:40] <h4x0r`> nup
[06:11:57] <grummund> why not?
[06:12:12] <h4x0r`> coz starin like dis
[06:12:14] <h4x0r`> :|
[06:12:19] <h4x0r`> +readin
[06:20:09] <karlp> h4x0r`: where are you finding that in the atmega32 datasheets?
[06:25:53] <h4x0r`> well a little misleading, i found an atmel application note with reference to avrs..
[06:26:16] <karlp> I'm sure you can do anything in software if you try really hard :)
[06:26:18] <h4x0r`> turns out, the appnote is suggesting developers use an ata IC
[06:26:19] <soul-d> http://www.olimex.com/dev/pdf/AVR/AVR-GSM/AVR-GSM.pdf
[06:26:28] <soul-d> they use an atmega 32?
[06:26:57] <h4x0r`> note sure still reading
[06:27:15] <h4x0r`> soul-d, you're interested etsi comms ?
[06:27:20] <h4x0r`> avr-based* ?!
[06:27:26] <soul-d> no i just googled :P
[06:27:58] <soul-d> cause i thought you might be mistaken with maybe avr32
[06:28:10] <h4x0r`> hrm, lol
[06:28:32] <h4x0r`> that devboard u linked, its just an avr talking to a gsm module..
[06:28:34] <h4x0r`> um
[06:28:35] <h4x0r`> and
[06:29:12] <h4x0r`> yea the gsm mux protocol is different, it allows for asyncronous network operations over virtual ports similar to uarts
[06:29:13] <h4x0r`> um
[06:29:24] <h4x0r`> like faxing while talking on fone and stuff
[06:29:51] <h4x0r`> no avr atm i know of supports it (is capable of it)
[06:30:00] <Steffanx> Just curious where are you from h4x0r` ? The US?
[06:30:35] <h4x0r`> AU
[06:30:42] <Steffanx> Hmm, ok :)
[06:31:13] <h4x0r`> and yourself Steffanx?
[06:31:19] <Steffanx> NL
[06:31:28] <h4x0r`> hehe :)
[06:32:21] <Tom_itx> crazy foreigners
[06:32:32] <Steffanx> There is only one crazy man here
[06:32:53] <h4x0r`> hehehe :D
[06:33:17] <h4x0r`> crazy alright..
[06:33:28] <h4x0r`> crazy.. like a FOX ...
[06:33:31] <h4x0r`> :D
[06:34:32] <h4x0r`> anyway heres an example ic that'll boost an avr's ability to multiplex compatible cmux gsm units.. have fun with it
http://www.atmel.com/dyn/resources/prod_documents/doc4841.pdf
[06:34:39] <h4x0r`> coz i aint touchin it :D
[06:35:08] <h4x0r`> shiz, wronf pdf..
[06:35:38] <h4x0r`> lol, nvm..
[06:37:44] <Tom_itx> too much grog?
[06:38:04] <Steffanx> Hehe
[06:38:12] <Steffanx> Not enough sleep probably
[06:38:36] <h4x0r`> hey not funny
[06:38:38] <h4x0r`> hehe
[06:38:51] <Steffanx> Not? It isn't true?
[06:38:58] <h4x0r`> nah just posted the wrong pdf
[06:39:03] <h4x0r`> how isnt it?
[06:52:16] <h4x0r`> just to prove im not drunk/crazy
[06:52:27] <h4x0r`> in my gsm module ds we have this,
[06:52:32] <h4x0r`> To access the three virtual interfaces, both the GSM engine and the customer
[06:52:33] <h4x0r`> application must contain Mux components that communicate over the multiplexer
[06:52:33] <h4x0r`> protocol.
[06:53:10] <h4x0r`> all you need is a multiplexer onboard, as the gsm module already supports it
[06:54:21] <h4x0r`> the mux allows you to use three internet virtual interfaces, like fax/voice/data all at the same time, so theres no need for 4 seperate uarts to control them
[06:54:40] <h4x0r`> s/internet/internal
[06:55:09] <Tom_itx> what freq is the mux?
[06:55:16] <h4x0r`> which one?
[06:56:37] <Tom_itx> how fast does it have to multiplex the 3?
[06:59:37] <h4x0r`> I guess at whatever the spec is for gsm 7.1.0 which is what the multiplexer on the gsm ic uses
[06:59:55] <h4x0r`> but digging around for a spec
[07:01:17] <h4x0r`> these documents really do suck
[07:07:47] <h4x0r`> seems theres a way to handle it over software.
[07:08:10] <ziph> Heheh, welcome to a world of pain.
[07:08:41] <h4x0r`> hehe, lucky i read a lot
[07:08:42] <h4x0r`> hahaha
[07:09:10] <h4x0r`> i always thought that was a software thing, those damn italian documents arent written very well.
[07:09:14] <ziph> h4x0r`: Me too.
[07:09:27] <ziph> h4x0r`: (Read datasheets that is, instead of trying to get it working with headbutting)
[07:09:51] <h4x0r`> lol. its the datasheet that isnt written well lol.. ie:
[07:10:17] <ziph> h4x0r`: Are you using the GSM spec or a vendor datasheet?
[07:10:21] <h4x0r`> To access the three virtual interfaces, both the GSM engine and the customer application must contain Mux components that communicate over the multiplexer
[07:10:22] <h4x0r`> protocol.
[07:10:27] <h4x0r`> vendor ds
[07:10:39] <ziph> Grab the GSM specification for it, and this one:
[07:10:40] <ziph> http://www.google.com.au/url?sa=t&rct=j&q=telit%20cmux%20user%20guide&source=web&cd=1&ved=0CB4QFjAA&url=http%3A%2F%2Fwww.telit.com%2Fmodule%2Finfopool%2Fdownload.php%3Fid%3D616&ei=poKyTo6vIa6UiQfD8s2NAg&usg=AFQjCNH867sa9_JL7kDIoanFLfYSE1_i5A&sig2=JAcKvLqtDUhqzrujh8aGcQ
[07:10:56] <h4x0r`> ten bux says its a split virtual uart lol
[07:11:01] <ziph> It's for the Telit devices, but it is general enough to cover any device.
[07:11:11] <ziph> And it has handy test vectors that the GSM spec doesn't.
[07:11:22] <h4x0r`> kewl
[07:11:33] <h4x0r`> thats where the quote comes from lol
[07:11:37] <ziph> h4x0r`: Also, if your CPU resets and the modem doesn't you may be screwed.
[07:12:16] <ziph> h4x0r`: And in the case of one modem, the fact that you were in CMUX mode was actually saved to flash.
[07:12:31] <ziph> h4x0r`: So you had to probe it on power on to see if you were looking at an AT interface or a CMUX interface.
[07:12:45] <h4x0r`> :)
[07:12:48] <h4x0r`> hmm nice thinking
[07:13:03] <h4x0r`> which chip are you using lol
[07:13:17] <ziph> h4x0r`: And it is handy to have a stream abstraction for your modem routines so that you can use them both on raw serial and on a MUX port.
[07:13:28] <ziph> h4x0r`: This was on an MSP430 and the modem I don't recall.
[07:13:53] <h4x0r`> sweet, what do play with these days ?
[07:14:17] <h4x0r`> "Every CMUX instance has its own user profile storage in NVM"
[07:14:22] <h4x0r`> omg thats insane.
[07:14:29] <ziph> Whatever the customer uses, I only do hardware design for my own projects. ;)
[07:15:06] <h4x0r`> omg .. cmux really is a wolrd of pain :/
[07:15:13] <ziph> h4x0r`: You can get modems with multiple interfaces (e.g. 3 RS232 ports that you can use at the same time.
[07:15:34] <ziph> Yeah, that driver probably cost $10,000 or something to write and make reliable. ;)
[07:15:52] <h4x0r`> ziph, yup the one i use has 2
[07:15:55] <h4x0r`> hardware
[07:15:59] <ziph> Plus you have to adapt anything that uses the modem to sit on top of it.
[07:16:06] <h4x0r`> yea
[07:17:08] <ziph> And at high RF power levels the modem would sometimes screw up the CMUX state and do things that were very wrong. :)
[07:17:25] <ziph> (Before they released a firmware update :)
[07:17:55] <h4x0r`> lol no doubt whatsoever ahaha
[07:18:11] <ziph> Why do you need the three ports?
[07:18:56] <ziph> Oh, and with the simple CMUX (not full HDLC) you can lose frame sync and therefore need to have timeouts on frame splitting.
[07:19:19] <h4x0r`> really.
[07:21:14] <h4x0r`> man those telit documents can be very slippery.. i interpreted the cmux component statement into the need for a rf/gsm transceiver..
[07:21:32] <h4x0r`> almost..
[07:21:39] <ziph> Yeah, you need to go off the GSM spec and use the Telit document as supplemental detail.
[07:21:51] <ziph> You're using a Telit?
[07:22:29] <h4x0r`> yup
[07:22:51] <h4x0r`> good advice there man
[07:23:39] <h4x0r`> well i always presumed the mux feature was a coded solution..
[07:23:52] <ziph> As in one you could get a library for?
[07:24:00] <h4x0r`> i was hoping to be able to use several ports back to the avr usart tho..
[07:24:29] <h4x0r`> that may not be required though.. its been a long time since I started the research for this project..
[07:25:00] <h4x0r`> i realise its going to be an ongoing project though, so plenty of time for reading :)
[07:26:27] <h4x0r`> well sweet, I can say it looks promising
[07:28:03] <ziph> It was good when it worked.
[07:28:56] <ziph> I had the PPP TCP/IP stack running on one connection while the other was used to pull cell data out every few seconds.
[07:29:25] <h4x0r`> SER?
[07:29:35] <h4x0r`> hardware uarts?
[07:29:37] <ziph> (Once it worked, that is. Once it had a bit of testing it worked reliably).
[07:29:45] <ziph> No, single UART with CMUX running on it.
[07:29:47] <h4x0r`> excellent
[07:29:55] <h4x0r`> wow sweet .
[07:30:02] <ziph> I think this particular modem had 1 RS232 and 1 USB.
[07:30:09] <inflex> mm... now I'm kinda bored
[07:30:29] <ziph> inflex: Go look for errors in the expensive board you just sent off? ;)
[07:33:21] <inflex> no, that's just depressing, I prefer to leave those discoveries to right after the elation of receiving them.
[07:33:31] <ziph> :)
[07:33:45] <ziph> I haven't found any on this one yet.
[07:33:54] <inflex> I remember once, for my first commercial board actually, I had a 2nd batch done and then I realised, "These boards don't actually even work!"
[07:34:03] <inflex> it was hilarious
[07:34:08] <ziph> Hah.
[07:34:11] <ziph> How did that happen? :)
[07:34:13] <inflex> yet people had bought them, and they actually did sortof work
[07:34:28] <inflex> needless to say, I ditched the new batch, respun the boards with the fix
[07:35:35] <inflex> the fault was that there was no positive-feedback going on with my lipo discharger, so when it approached the 'cutoff' point, it just slowwwwwwly turned off depending on the transistor, temperature and everything else
[07:35:48] <inflex> once the feedback was fixed it SNAPPED off like it was supposed to
[07:35:54] <ziph> One I had recently had VCC/GND switched on an OPAMP because I rotated the supply schematic block 180 degrees instead of flipping it and wired the power to the top and ground to the bottom, and then failed to notice it while checking the netlist.
[07:36:04] <inflex> oooh, I did that recently
[07:36:20] <inflex> ended up bypassing the opamp (was a 2x gain stage) and just had to live with half the resolution
[07:36:26] <ziph> I'm thinking of writing some better ERC software..
[07:37:31] <ziph> inflex: Funny thing is that it killed the regulator which presumably shorted internally and killed a pin on the MCU. :)
[07:37:44] <inflex> my original one would have shown up in a DRC
[07:37:55] <inflex> but I didn't use it at that time... was still "young"
[07:40:51] <karlp> ziph: what package are you using that didn't catch that in ERC?
[07:41:23] <karlp> it's likely to more a problem in the part library, not defining what pins were power in and ground
[07:41:24] <ziph> karlp: What package supports nominating supply voltages?
[07:41:39] <karlp> eagle does.
[07:41:48] <karlp> you can say that a given pin is power in.
[07:42:11] <karlp> or a given pin is ground,
[07:42:29] <karlp> and it will fail ERC. I've had it flag exactly the problem you describe
[07:44:24] <ziph> The best way is to actually print out a netlist and tick everything off.
[07:47:48] <karlp> or, get better ERC
[07:47:59] <karlp> the hell I'm going to print a netlist and verify net by net
[07:48:22] <ziph> To save $300 and two weeks? Why not? :)
[09:07:09] <spow> Hi. I have a course about AVR realtime. The teacher said it's ok for me to borrow a microcontroller given the circumstances I joined the uni, but that's it up to me to "get the C library working on WinARV or AVR-Studio.". Do I have to copy files from a computer at the uni ? Or just download the environments he named ?
[09:07:29] <jacekowski> download it
[09:08:20] <spow> Can I test without the µC that everyhting will work ? I'm getting it tomorrow
[09:08:46] <ziph> AVR Studio has an emulator that you can try it in.
[09:12:00] <spow> thanks, will try that from the uni wifi :)
[09:13:00] <h4x0r`> does that emulator just run code, or can it do circuits?
[09:13:24] <h4x0r`> disregard, presuming just code.
[09:13:32] <ziph> It runs code, the one that ErrorDevelope*r does lets you write peripherals though.
[09:14:02] <ziph> You can get integrated environment that do both, but they seem a bit pointless.
[09:15:59] <h4x0r`> yea
[09:17:35] <ziph> Simulating complex systems with models is useful though (e.g. OpenModelica).
[09:24:55] <h4x0r`> periphery thing looks ok actually
[09:25:06] <h4x0r`> hmm a usart transceiver.. niice :p
[09:34:19] <_Shurik_> hello hello
[09:53:56] <ziph> Woho, another part of the board confirmed working.
[10:13:34] <_Shurik_> yay!
[10:57:41] <inflex> ziph: nice going
[10:57:46] * inflex should go to bed... nearly 2am
[10:59:27] <ziph> Is it late enough for it to be < 35degree or 99.9% humidity yet?
[11:04:24] <inflex> ugh yeah
[11:04:33] <ziph> :)
[11:04:38] <inflex> right, to bed... else I'll have no zest to work tomorrow (two orders to build and ship)
[11:04:51] <ziph> Sleep well.
[11:05:04] <ziph> I hope the mozzies leave a little blood for your heart. :)
[13:39:12] <hotch> hey Tom_itx it came, thx :P
[13:52:10] <RikusW> anyone know of something that will dissolve the black ic casing ?
[13:52:39] <RikusW> then my dead m32 may be useful as a display item ;)
[13:53:18] <jacekowski> HF
[13:54:03] <scuzzy> anyone here ever had ACL reconstruction?
[13:54:24] <RikusW> jacekowski: that will eat the ic legs too ?...
[13:55:47] <RikusW> seems acetone don't work...
[13:56:16] <jacekowski> it won't eat legs as fast as casing
[13:56:24] <jacekowski> but it does eat them
[13:57:28] <Kevin`> only apply it to the case. if you applied it AROUND the legs anyway they would come free of the case and be useless
[13:57:52] * RikusW don't have any HF
[13:58:03] <RikusW> not sure I want it either...
[13:58:09] <karlp> what's the project?
[13:58:18] <RikusW> I do have HNO3
[13:58:38] <RikusW> stripping a dead m32's casing away
[14:03:43] <karlp> just for fun?
[14:10:09] <_Shurik_> RikusW: thanks for the AVR Dragon cririque yesterday. Got stuff soldered and now works like a charm
[14:14:01] <RikusW> _Shurik_: you soldered the pins on dragon ?
[14:14:18] <RikusW> I use a jumper wire with connectors
[14:15:06] <_Shurik_> RikusW: What I did is I found some IC over-the-top sockets (with long pins) and soldered them into HV connector
[14:15:16] <_Shurik_> this way I can push regular wires into them
[14:15:53] <RikusW> I got some connectors like the jtag one, and solders 2 of those into my hv ports
[14:16:20] <RikusW> just take note you must still connect vcc to vtg on isp for hvpp to work properly... sort of dumb
[14:17:03] <_Shurik_> http://www.asia.ru/images/target/photo/51625710/IC_Sockets_and_SIP_Sockets_with_Long_Pin.jpg
[14:17:19] <_Shurik_> One of those on the left
[14:17:33] <_Shurik_> cut the plastic in the middle and soldered two halves in
[14:17:44] <_Shurik_> trimming pins of course
[14:17:58] <_Shurik_> now I can run wires from HV into breadboard
[14:18:14] <_Shurik_> yeah, and I did the VCC to ISP connection as specified in PDF
[14:18:32] <RikusW> using jumper wire ?
[14:18:42] <_Shurik_> those pins separately run about 30 cents each! One of those sockets is only 99c in the store
[14:18:48] <_Shurik_> yes
[14:18:54] <_Shurik_> jumper wires
[14:19:27] <RikusW> I got yy09 connectors, 1100 of them...
[14:19:36] <RikusW> for about 11 euro
[14:19:51] <RikusW> they push onto the normal pins
[14:20:11] <RikusW> then got 2x5 plastic casing for those
[14:20:31] <RikusW> use 15cm wire and made a few of those connectors
[14:20:44] <RikusW> my board also use 2x5 so...
[14:21:09] <_Shurik_> nice
[14:22:17] <RikusW> YY092C10 and YY09CT at www.mantech.co.za
[14:22:47] <RikusW> used with YWN10
[14:23:04] <_Shurik_> Hmm, I see
[14:23:19] <RikusW> also got some YRF10
[14:23:34] <RikusW> for putting 2 on a piece of ribbon cable
[14:23:54] <RikusW> and YRF06 for ISP cable
[14:23:54] <_Shurik_> All the fancy stuff eh? :)
[14:24:03] <RikusW> and quite cheap too
[14:25:37] <RikusW> used some 1.5mm heatshrink tubing to cover the YY09CT connectors to have loose wires too
[14:26:14] <_Shurik_> RikusW: If I would run AtTiny13a from 2AA batteries at 125Khz, I shouldn't need any caps on VCC?
[14:26:55] <RikusW> try adding 100nF ?
[14:27:02] <RikusW> just to be save
[14:27:06] <RikusW> safe
[14:27:14] <_Shurik_> yeah? Okay
[15:55:58] <amee2k> i've got a bunch of module boards that i want to repurpose, but they have no mounting holes whatsoever. they probably came in form-fitting cases with molded mounts. i can't see useful places to drill my own holes. any ideas how and from what to make mounting brackets to hold them in my own case?
[16:16:52] <karlp> hot glue
[16:26:33] <vectory> hi, i heard that with gcc, array[] will be created on the stack and malloc* will be on the heap, does that apply to avr-gcc, too?
[16:27:20] <Kevin`> vectory: yes
[16:27:27] <vectory> ok, next question
[16:27:31] <Kevin`> unless you malloc an array
[16:27:52] <vectory> like int a[]=malloc(x)?
[16:28:07] <vectory> or int* a= malloc(x)?
[16:28:11] <Kevin`> yeah, the data there would end up in the heap of course
[16:28:41] <vectory> didnt think the first was even possible, lol, never saw that
[16:29:50] <Kevin`> it would be kind of strange as written since you are declaring it right there. normally just do int a[x]. unless x isn't constant.
[16:31:07] <vectory> so, avr-gcc isnt c99 compliant, since dynamic array size initialization is allowd in the 99 std iirc
[16:32:55] <vectory> anyway
[16:33:00] <vectory> next question: if data, say a lookuptable, is on the stack and accessed, will everything above a value in the stack need to be popped to get said value?
[16:33:49] <vectory> at least thats how i understand a stack, so i thought it would be unwise to have a LUT on the stack
[16:34:06] <h4x0r`> does anyone here use the enable pin of a regulator to start their project? Id like an example using
http://www.st.com/internet/com/TECHNICAL_RESOURCES/TECHNICAL_LITERATURE/DATASHEET/CD00002172.pdf
[16:34:15] <Kevin`> vectory: don't put a lookup table in ram, put it only in program memory
[16:34:41] <vectory> hm, good point
[16:34:51] <Kevin`> vectory: memory locations on the stack are still memory locations. ignoring the LUT part.
[16:35:03] <vectory> k
[16:35:57] <Kevin`> the stack part of the stack is more about returning to the previous context than the variables there
[16:37:06] <vectory> silly me :S
[16:38:52] <vectory> its still faster to write a loop to fill the ram, then creating a file with a lut and integrating it in the makefile -_-
[16:39:11] <vectory> lazyness wins
[16:40:32] <edboogie2011> hi guys. I am not good at calculating wire gauges. If I need to power 24 VDC 2.0 AMP brushless motors, what gauge or type of wire should I use?
[16:42:21] <Tom_itx> http://www.mogami.com/e/cad/wire-gauge.html
[16:42:55] <Tom_itx> meh, that one's no good
[16:43:04] <edboogie2011> yeah,,,
[16:43:06] <edboogie2011> ?
[16:43:10] <edboogie2011> i don't understand how to use it
[16:43:44] <vectory> it only converts the units
[16:43:50] <vectory> from AWG to mm
[16:45:15] <Tom_itx> http://circuitcalculator.com/wordpress/2007/09/20/wire-parameter-calculator/
[16:45:17] <Tom_itx> closer
[16:46:43] <krphop> edboogie2011: depends on how long your wires are going to be
[16:47:28] <Tom_itx> http://www.powerstream.com/Wire_Size.htm
[16:53:25] <amee2k> karlp: well, i thought about it but just squirting a ton of hot glue and sticking it on sounds like the easy way out... and a pretty poor one too IMO
[16:53:34] <amee2k> (sorry for the slow reply :)
[17:26:26] <aar> anyone active?
[17:26:50] <aar> got some questions about mc in general
[17:27:13] <aar> ? nm then
[17:27:47] <aar> if anyone happens to pop in and see this please post, I'll be checking it for a little while
[17:27:58] <aar> need some advice on jury rigging :D
[17:32:01] <karlp> aar, always best to actually ask the question, then wait for a reply
[19:31:04] <inflex> lo all
[19:41:20] <Tom_itx> hey
[19:41:34] <inflex> hiya Tom_itx - how's matters?
[19:42:03] <Tom_itx> same as usual
[19:42:26] <inflex> shame - would hope that something would improve.
[20:23:06] <edboogie2011> Hey guys.. this is waaay off topic.. but i don't know who to ask... i bought fifty 512MB microSD cards for $4 each from china through ebay. and they came in loose packaging with Sandisk silkscreen, how do i tell if they are counterfeit.. i am worried...
[20:23:12] <edboogie2011> do these look real?
http://www.ebay.com/itm/50-x-512MB-San-Disk-MicroSD-TF-Memory-Card-GENUINE-NEW-/230447563753?pt=Digital_Camera_Memory_Cards&hash=item35a7bec3e9#ht_2556wt_1112
[20:23:31] <karlp> you should ask sandisk
[20:23:33] <edboogie2011> I need them to give one for each of my customers for an embedded device I am releasing
[20:23:41] <edboogie2011> yeah good idea!
[20:23:41] <edboogie2011> nice
[20:23:42] <inflex> heh
[20:23:50] <inflex> eBay was likely a bad place to pick up that sort of thing though
[20:24:16] <karlp> I'd say they're probably fake,
[20:24:22] <karlp> but whether that's a problem or not is a different matter
[20:24:28] <inflex> still, they are only 512MB
[20:24:42] <inflex> so it's not likely that they'd have bothered doing the space faking
[20:26:41] <edboogie2011> what is space faking?
[20:27:02] <edboogie2011> I feel like counterfeiting these things seem almost not worth it
[20:27:04] <edboogie2011> maybe I am wrong
[20:27:17] <inflex> that's what I'm thinking at this end - but they might just be 'generic' units rebadged
[20:27:28] <inflex> space-faking is when they say 4GB, but really it's 1GB
[20:27:54] <inflex> once you write beyond 1GB it just overwrites or dumps the data
[20:28:10] <edboogie2011> inflex: wow that is horrible
[20:28:14] <edboogie2011> i will test that now
[20:28:48] <inflex> easy way to test, yes... just create a random file that's about the size of the memory card/chip... md5 it... then copy it over and copy it back.. MD5 should be the same
[20:28:59] <inflex> if not, then tadaaaah, you have a faker
[20:32:03] <edboogie2011> great idea
[20:32:15] <edboogie2011> is there an easy way to get MD5 checksum in windows XP?
[20:34:09] <inflex> oh, no idea
[20:50:51] <Casper> edboogie2011: of files? there is md5sum.exe
[20:50:56] <Casper> and tons of progs too
