Back
[00:00:28] <N1njaneer> Wow $24.95 for that? Clearly I need to start building more consumer stuff :)
[00:01:15] <learath> I suspect most of that is the support and the low volume
[00:01:25] <N1njaneer> Indeed.
[00:01:26] <learath> also its' *way* overkill
[00:01:50] <N1njaneer> We could spit out thousands of those per day here. :)
[00:01:59] <learath> Yeah :)
[00:02:17] <learath> it's almost exactly thumb sized
[00:02:21] <N1njaneer> But then the support is a killer, since a single person will soak up $24.95 worth of your time not reading the manual. :)
[00:02:34] <learath> particularly with *really* shitty firmware
[00:03:01] <N1njaneer> Yeah no doubt
[00:03:13] <N1njaneer> All this FTDI USB connectivity stuff also kills me.
[00:03:31] <N1njaneer> I mean, it's easy, but... It's still a serial-port.
[00:03:44] <N1njaneer> And has serial-port limitations and abnormalities.
[00:04:30] <N1njaneer> But that's just me. And most of us around here love LUFA for AVRs :)
[00:05:24] <learath> LUFA?
[00:06:15] <Valen> I made soap with lufa engraved on it ;->
[00:08:11] <learath> man. This makes things *so much* nicer
[00:08:17] <learath> and it's trivial
[00:09:22] <N1njaneer> learth:
http://fourwalledcubicle.com/LUFA.php -- and Dean is on here regularly as abcuser
[00:14:59] <Roklobsta> abuser?
[00:15:52] <Roklobsta> N1njaneer: the nice thing about ftdi is the Win/Lin/wahtever serial drivers are robust
[00:20:04] <N1njaneer> True, though most times I seem to still need to load a driver for it when a product uses one. Versus, say, doing it as a HID-device if you don't need to push a gargantuan amount of data. I've done all of my USB designs with HID thus far for the ease of different OS reason :)
[00:20:04] <learath> crap. uint8_t EEMEM send_string_crc_1_address = ""; // need to compute this correctly does not actually appear to work
[00:20:35] <N1njaneer> Needs to be a pointer.
[00:21:54] <N1njaneer> And I'm not sure if "" is considered valid notation. It needs to have a string literal of some sort in there, even something like "\0"
[00:22:11] <learath> hm. I'll try that
[00:23:27] <N1njaneer> But if you are doing it as the address of the string, you may need to define it as a pointer. Either uint8_t * and cast the string (which will inherently be of 'char') to it, or do it as a "char *"
[00:23:46] <N1njaneer> I've not used the EEMEM notation, though, so I cannot say for certain what the compiler will be happy with :)
[00:23:52] <learath> looks like I was just being a dumbass
[00:24:23] <N1njaneer> Alternatively, do variable[] = "put stuff here" and it will autosize it, however in your application you probably want to use specific size allocations :)
[00:25:01] <learath> Yeah.
[00:25:11] <learath> Otherwise the saved strings will be less useful than they could be
[00:25:15] <learath> or I'd have to manually pad them out
[00:28:16] <N1njaneer> Just remember to provision for the null-terminator if necessary.
[00:40:20] <learath> built clean. Testing: tomorrow
[00:40:25] <learath> and thank
[00:40:26] <learath> s
[05:31:31] <megal0maniac> Not necessarily avr related, but I know there might be people here who can help. I'm looking for a PSU for a Cisco Catalyst 2960-S
[05:32:17] <megal0maniac> It's a Liteon PCB inside the switch. 60W and molex connectors with what I assume to be a Cisco specific pin-out
[05:37:50] <megal0maniac> It's much like this, except the connectors are different
http://www.ebay.com/itm/151088596482
[05:38:09] <megal0maniac> And it's LITEON PA-1600-4-LF, no A
[05:38:37] <megal0maniac> And Cisco isn't keen on helping, since we didn't spend $9999999999999999 on a higher end switch
[05:38:57] <megal0maniac> RikusW: Perhaps you'd consider fixing it?
[05:39:07] <megal0maniac> Or having a look at least
[05:40:57] <megal0maniac> Paid job, obviously. I know you have experience with SMPS and you're close
[05:44:56] <RikusW> hi megal0maniac
[05:45:37] * RikusW can get xmega32a4u for R21 (ex vat)
[05:45:42] * megal0maniac waves
[05:45:44] <megal0maniac> Nice!
[05:45:54] <RikusW> arrow.altech.co.za
[05:46:12] <RikusW> and that is for small quantities
[05:46:12] <megal0maniac> They're the ones who don't like small quantities
[05:46:20] <megal0maniac> Oh lol
[05:46:30] <RikusW> seems they do supply if you plan to order bigger
[05:46:42] <megal0maniac> Have you seen my previous msg?
[05:46:50] <RikusW> seems they specialise in supplying an entire BOM
[05:46:58] <RikusW> which one ?
[05:47:10] <megal0maniac> Dead SMPS in a Cisco switch
[05:47:16] <RikusW> zlog
[05:47:37] <RikusW> what happened ? surge ?
[05:48:00] <megal0maniac> Probably. Caps are all fine, fuse is intact, but no secondary power
[05:48:36] <RikusW> you could measure to see if there is 350V on the HV cap
[05:48:43] <RikusW> (be careful) :-P
[05:49:09] <RikusW> and check the transistor on the HV side
[05:51:08] * RikusW is back on GPRS :(
[05:51:08] <megal0maniac> Well I'm checking whether they'd be keen to send it to someone, I don't want to get involved because I don't know what I'm doing and they don't mind paying. The switch is worth a few thousand ZAR, and you can't get new PSUs. 1 month out of warranty :/
[05:51:28] <RikusW> ouch
[05:51:40] <megal0maniac> So basically, are you willing to have a look? Charge whatever you charge, make a few bucks
[05:52:16] <megal0maniac> And possibly fix the switch :P
[05:52:43] <RikusW> possibly
[05:52:50] <RikusW> what output voltages are there ?
[05:54:01] <RikusW> megal0maniac: so you opened it switch to have a look inside ?
[05:54:04] <megal0maniac> They have two of them, they've swapped the PSUs and confirmed that the switch still works, and the PSU is in fact dead
[05:54:13] <RikusW> ah
[05:54:47] <megal0maniac> I'm not sure what voltages. I did take a look, but I just checked the caps, fuse, looked for burn, dry joints etc
[05:54:53] <RikusW> is there any visible damage ? blackened parted ?
[05:54:58] <megal0maniac> pm
[05:55:04] <RikusW> or cracked ICs / transistors ?
[05:55:06] <RikusW> ok
[06:25:27] <ambro718> I'm seeing that multiplying a float by two makes gcc produce a call to __addsf3, that is it adds the value to itself.... is that optimal?
[06:25:56] <ambro718> I think it's just the matter of incrementing the exponent, with a small complication for subnormal values.
[06:35:27] <ambro718> I'll try to benchmark 2*x against ldexp(x, 1)
[06:51:47] <ambro718> indeed, 2*x takes about 112 cycles, ldexp(x, 1) takes about 73
[09:37:01] <AllinYourhead> Hello. I need some help with hardware design on an ATmega32C1. My problem is that the UART bus is using the same pins as the ISP (PD3 = MOSI/TX and PD4 = SCK/RX). I've found a few search hits regarding peripheral SPI devices sharing bus for ISP that should be blocked off with resistors but is it the same thing when SPI and UART is combined? I don't want my UART device to interfere during ISP
[09:37:01] <AllinYourhead> programming.
[09:49:58] <megal0maniac_afk> AllinYourhead: First thing that comes to mind is some kind of buffer between your UART device and the AVR which you can tristate when the reset pin is low. But there is probably a simpler solution
[09:52:48] <AllinYourhead> might have a lazy man's solution. My UART peripheral device got a reset pin working just like the ATmega so if i connect those two the ISP-programmer will pull the peripheral into reset as well and thereby force it to be quiet
[09:54:38] <megal0maniac_afk> That is indeed the lazy-man's solution, but it would work
[09:54:58] <megal0maniac_afk> This is what I had in mind
http://www.nxp.com/documents/data_sheet/74LVC1G125.pdf
[09:56:57] <megal0maniac_afk> Pity about the polarity, but there it is
[09:58:37] <megal0maniac_afk> The RX line on your peripheral will be high Z anyway, so one of these is the "nice" way to do it. A simple resistor might work, but now I'm speculating
[10:02:29] <AllinYourhead> recently discovered that running an 8-bit ATmega @ 3.3V and 16MHz is considered unsafe zone according to datasheet - but havnt noticed any problems at all with that setup. Any experiances on this?
[10:04:42] <moemoe> can work, can create heisenbugs
[10:07:04] <vsync_> there is an avr in a box
[10:30:10] <RikusW> it worked for me so far
[10:30:19] <RikusW> using mega32u2
[10:31:09] <RikusW> If I use a p-channel fet to switch on/off a 30mA load, will I need a resistor between GPIO and gate ?
[10:31:41] <RikusW> it will be a tiny fet so gate capacitance should be minimal
[12:13:08] <julius_> hi
[12:13:37] <julius_> im looking for a cheap way to experiment with ethernet/avr. what board/software would you use?
[12:19:01] <Joggl> wait until i finished my netEval-Board ;)
[12:37:54] <julius_> Joggl, what can it do besides ethernet?
[12:38:56] <Joggl> julius_, i plan a evaluationboard with lufa-clone of avrispmkii so you can program it via atmel studio without problems
[12:39:06] <Joggl> and i plan to add the wiznet-chip as ethernet-solution
[12:39:16] <Joggl> so you do not need to program the tcp/ip stack on the avr
[12:40:31] <julius_> oh you do plan to create your own board?
[12:43:07] <learath> can you get a proper tcp stack in an AVR?
[12:58:32] <ambro718> learath: look up uIP and lwIP
[12:58:53] <ambro718> though I'm not sure whether lwIP fits in any AVR
[14:29:08] <Joggl> learath, yes you can, but its pain in the ass
[14:29:15] <Joggl> too less ram and so on
[14:29:36] <Joggl> julius_, yes, improvement of my actual board:
http://www.ehajo.de/Bausaetze/aTeVaL
[15:38:36] -hitchcock.freenode.net:#avr- [freenode-info] why register and identify? your IRC nick is how people know you.
http://freenode.net/faq.shtml#nicksetup
[20:28:05] <jadew> http://www.eevblog.com/forum/buysellwanted/texas-instruments-25$-coupon-code!/
[20:28:21] <jadew> $25 coupon code on any order from the ti store :D
[20:29:03] <w|zzy> hmm
[20:29:07] <jadew> I just ordered two cortex-m4 launchpads for $1 :D
[20:29:12] <w|zzy> nice
[20:29:32] <jadew> if the price is lower than $25, it's free
[20:29:40] <jadew> for me it was 98 cents higher tho
[20:29:57] <jadew> free shipping as well
[20:31:00] <w|zzy> got enough toys at the moment
[20:31:08] <jadew> me too
[20:31:14] <jadew> but I can't stop shopping
[20:31:35] <jadew> lots of toys, no time to play with them, but that's probably gonna change after the wedding
[20:32:03] <Fornaxian> yeah...no more toys, and still no time to play with the ones you have.
[20:32:13] <jadew> hehe
[20:32:13] <w|zzy> Lol. because the sex stops after marriage?
[20:32:19] <Fornaxian> well, no more toys unless you like rattles.
[20:32:33] <w|zzy> Lol
[20:32:35] <Fornaxian> w|zzy, yeah....the best thing to kill a woman's sex drive is the first bite of a wedding cake.
[20:32:48] <jadew> haha
[20:33:20] <w|zzy> Lol
[20:33:41] <w|zzy> and I still don't have time
[21:09:47] <ambro718> hm, I calculate I probably didn't destroy my AVR by over-flashing it. That would have taken about 7 days of continuous flashing.
[21:11:00] <ambro718> it's broken in a weird way. It seems that some particular sequence of instructions resets it, with some probability. Or it crashes in a strange way without resetting.
[21:11:48] <Casper> ambro718: verify the code, if it still verify then you did not worn out the flash
[21:12:11] <ambro718> Casper: it's all written correctly, avrdude verifies it automatically
[21:12:26] <Casper> then the flash is ok
[21:12:38] <ambro718> this all started when avrdude reported on time that some fuse changed. I restored the fuses but from that time it's broken.
[21:12:40] <Fornaxian> would take quite a bit to wear it out....think someone even did an experiment on it to see how long it would take by continuous rewriting.
[21:13:58] <ambro718> I've previously destroyed a chip by programming it while unpowered, but I'm pretty sure this wasn't the case here
[21:14:29] <Casper> ambro718: that could damage the protection diodes, but not the rest of the chip
[21:14:44] <Casper> however, in some case it could cause it to erase the signature
[21:14:48] <Fornaxian> that's kind of odd....I've tried programming them without power before and never had a problem....I suppose it could do strange things though.
[21:17:26] <Casper> jadew: Estimated Ship Date: 09/17/2013 Note: In Warehouse, Pulled for Shipping
[21:18:03] <Casper> I wonder if they sent it and didn't sent me the tracking, or if they didn'T got time to ship today
[21:35:57] <ambro718> why do AVRs have relatively many timer output compare units (high-end AVRs have 6 or more, and xmega seems to have 22), but AVR32 and AT91SAM has few (3 it seems, that is the RC register for each of the 3 timers)?
[21:36:47] <Fornaxian> different function targets?
[21:37:06] <Fornaxian> avr32 and arm are more processors and less controllers....
[21:38:58] <ambro718> it's weird, why can't I have a fast microcontrollers that also has lots of peripherals like timers
[21:39:36] <ambro718> by fast I minimally mean 32-bit
[21:39:38] <Fornaxian> dunno.
[21:39:54] <Fornaxian> real world interface makes the necessity to slow things down perhaps?
[21:43:11] <Casper> ambro718: usually, when you need more power, you ain't controlling stuff directly
[21:45:28] <Enma_Hinobara> I love it when my disk drive says things like *WHIRRRRR SNAP beep beep beep*
[21:45:52] <Fornaxian> better than Whirrrr GRRRIIIIND....
[21:46:57] <Casper> click click beeeeep click click beeeeep click click beeeeep click click beeeeep click click beeeeep