Back
[00:53:37] <inflex> Jordan_U: datasheet gives full details
[10:49:50] <rue_mohr> Jordan_U, I'd say maybe 100ma on the crazy high end (overclocking and all)
[10:50:11] <rue_mohr> Jordan_U, and few uA in proper sleep
[10:53:30] <specing> 100ma if you have a lot of leds attached
[10:53:59] <monobit_> i wont to give 100A to the face of the guy that invented electronics
[10:54:03] <specing> In practice the chip itself should be way below 10ma
[10:54:19] <theBear> well, as long as you wont
[10:54:20] <rue_mohr> I actually had to downclock an m32 once to make a battery last a descent amount of time
[11:10:48] <Tom_itx> fri the 13
[11:11:37] <theBear> until a minute ago :)
[11:12:02] <Tom_itx> you're safe then
[11:12:32] <theBear> thank god
[11:15:19] <elektrinis> hi
[11:15:19] <tobbor> elektrinis! like, totally tell us about the project!
[11:15:44] <elektrinis> this bot makes me smile
[11:15:46] <elektrinis> each time :)
[11:17:08] <elektrinis> so listne... i have this AVR project... micro fith 8k flash
[11:17:20] <elektrinis> looking for smallest possible serial bootloader
[11:17:55] <elektrinis> any suggestions?
[11:37:08] <Steffanx> Get some uart bootlaoder written in assembly elektrinis
[11:37:22] <elektrinis> m
[11:37:42] <elektrinis> it will be a bit hard to modify it
[11:37:50] <Steffanx> Noo
[11:38:05] <Steffanx> AVR assembly isn't THAT hard
[11:38:11] <elektrinis> or get suitable one, as I am using a micro that is not that common
[11:38:21] <elektrinis> I do know avr asm
[11:38:27] <elektrinis> coded for few years
[11:38:30] <Steffanx> not very common = ?
[11:38:36] <elektrinis> AT90PWM3B
[11:38:47] <elektrinis> specific for motor drive
[11:44:34] <theBear> mmm.. i always found the AT90PWM series very interesting
[11:55:27] <elektrinis> this one has a powerful power stage controller
[11:55:36] <elektrinis> but very little pins and memory
[11:55:39] <elektrinis> unfortunately..
[11:57:58] <theBear> from memory i liked the 2... back then at90s series were err, VERY limited in some regards, including pwm outputs
[12:00:05] <elektrinis> m, other question..
[12:00:05] <elektrinis> http://www.asix.net/tools/prg_presto.htm
[12:00:09] <elektrinis> I have this programmer
[12:00:25] <elektrinis> it has a command line support, similar to avrdude
[12:00:40] <nusse> at least it was expensive
[12:00:41] <elektrinis> can I use it with eclipse AVR environment?
[12:01:10] <nusse> i guess eclipse is using avrdude for flashing?
[12:01:13] <elektrinis> yes
[12:02:22] <nusse> check the preferences if you find the option where you can set how to call avrdude and replace it with your call
[12:03:02] <theBear> back when i was really into code highlighting i just setup a macro or two in a given env for a given programming util/cmdline
[12:04:18] <elektrinis> cant wind the way to modify avr dude
[12:04:28] <elektrinis> it only lets me to choose some parameters
[12:04:32] <elektrinis> in style on/off
[12:04:38] <nusse> http://avr-eclipse.sourceforge.net/user%20manual/other/whatsnew.html if this is the eclipse integration, its kinda advanced and just replacing the call wont do anything :)
[12:05:07] <elektrinis> yes thats exactly what I see
[12:05:33] <nusse> maybe theres somewhere support for your programmer in avrdude?
[12:06:00] <elektrinis> unfortunately its not in the list
[12:07:00] <nusse> whats the reason for using the presto programmer anyway?
[12:09:10] <Tom_itx> mine works with avrdude or studio
[12:13:26] <elektrinis> Tom_itx, could you please repeat the link to your programmer?
[12:13:39] <elektrinis> nusse, because its the only hardware I have
[12:14:26] <Tom_itx> http://tom-itx.dyndns.org:81/~webpage/boards/USBTiny_Mkii/USBTiny_Mkii_index.php
[12:14:49] <Tom_itx> oh wait, were you the one that needed HVPP?
[12:14:55] <Kevin`> the programmers from atmel aren't priced too bad, have you looked at them elektrinis ?
[12:14:58] <Tom_itx> kuz mine don't do that
[12:15:04] <Kevin`> elektrinis: avr dragon specifically is a useful device
[12:15:11] <nusse> there are lots of cheap programmers
[12:15:30] <elektrinis> Tom_itx, yeah, same one
[12:15:31] <Tom_itx> some work better than others too
[12:15:39] <elektrinis> I guess I could do without HV for now
[12:15:45] <Tom_itx> why do you need hvpp?
[12:16:10] <elektrinis> there is a produjt in my company that uses tiny45
[12:16:26] <elektrinis> and HVPP is used to get one extra pin
[12:16:37] <Tom_itx> ahh
[12:16:44] <elektrinis> we need a programmer for production line
[12:16:58] <elektrinis> we currently use dragons, but there are an absolute crap
[12:17:32] <elektrinis> hanging all the time, dying once in a while...
[12:17:36] <elektrinis> and no enclosure..
[12:18:14] <Kevin`> the protocols for programming aren't secret, if you are doing high volume stuff you can just do it directly. the development tools are kind of designed for development, usb reset and all..
[12:19:10] <elektrinis> Kevin`, to "just do it directly" requires some additional work and human resources
[12:19:22] <elektrinis> as well as money to feed them
[12:19:31] <elektrinis> to buy a tool is cheaper.
[12:19:32] <Tom_itx> they don't work free?
[12:19:41] <nusse> dont feed them
[12:19:43] <Tom_itx> wtf
[12:19:52] <elektrinis> crazy, I agree. but they just dont work for free
[12:20:42] <Kevin`> elektrinis: it sounded like you had enough problems to benefit from it, I would expect it to only be a few hours work
[12:20:59] <elektrinis> :)
[12:21:19] <elektrinis> to me it looks like a week of work at least
[12:21:38] <elektrinis> designing schematics, pcb, enclosure, firmware, pc software. ...
[12:21:51] <elektrinis> a week if your'e lucky
[12:22:12] <Tom_itx> and the first prototype seldom is what you really want
[12:22:30] <elektrinis> Tom_itx, is your programmer built from open source schamatic and firmware?
[12:22:37] <Tom_itx> firmware is
[12:22:53] <elektrinis> I'd like to look into it
[12:23:02] <CapnKernel> tom's programmers == good
[12:23:14] <Tom_itx> i posted the schematic briefly and someone stole the design. not that it was special but it was a pisser to say the least
[12:23:35] <Tom_itx> right down to the description on my webpage :)
[12:23:39] <elektrinis> yeah, I know the feeling
[12:23:42] <CapnKernel> tom hates indonesians...
[12:23:54] <elektrinis> is your v2 programmer available to purchase?
[12:23:54] <Tom_itx> i've had no real complaints with it
[12:24:00] <Tom_itx> yes
[12:24:07] <Kevin`> Tom_itx's programmer has minimal external components, the design is pretty obvious. you can actually use the firmware with just the bare chip connected to usb
[12:24:07] <elektrinis> huw much is it? :)
[12:24:08] <Tom_itx> i'm running low on those right now
[12:24:18] <Kevin`> i'm suprised someone would steal it :/
[12:24:29] <Tom_itx> well they did
[12:24:42] <Tom_itx> dean was pissed at the time as well
[12:24:57] <elektrinis> Kevin`, where can I read on about it?
[12:25:03] <Tom_itx> elektrinis click on the 'purchase items' menu link
[12:25:13] <Tom_itx> if you're interested
[12:25:25] <elektrinis> m, $30
[12:25:37] <elektrinis> it would be shipped from US?
[12:25:44] <Tom_itx> yes
[12:25:49] <Tom_itx> standard post
[12:25:52] <elektrinis> m, plus tax 25%
[12:26:08] <elektrinis> shipping is like $3, right?
[12:26:10] <Kevin`> elektrinis: the programmer is included as an application demo in LUFA
[12:26:13] <Tom_itx> 5
[12:26:15] <elektrinis> ok
[12:45:18] <elektrinis> m
[12:45:27] <elektrinis> noticed my micro has the debugwire
[12:45:54] <elektrinis> anyone stumbled across a serial to debugwire adapter?
[12:48:47] <elektrinis> Tom_itx, your programmer support dW?
[12:49:13] <Steffanx> No his programmer doesn't support dW
[12:50:13] <Steffanx> but hey, go for it elektrinis :
http://ruemohr.org/docs/debugwire.html :P
[12:50:47] <elektrinis> huh
[13:01:07] <Tom_itx> elektrinis no
[13:01:37] <elektrinis> ok
[13:02:03] <Steffanx> but hey, go for it elektrinis :
http://ruemohr.org/docs/debugwire.html :P
[13:02:28] <elektrinis> not in the mood to go that deep
[13:02:47] <Tom_itx> you got dragons to do debugwire
[13:03:02] <Tom_itx> and hvpp if needed
[13:14:49] <elektrinis> http://www.keeelectronics.com/AVR_debugWire_ISP.htm
[13:14:55] <elektrinis> interesting
[13:16:13] <elektrinis> $39
[14:10:55] <elektrinis> is atmel studio 6 eclipse-based?
[14:11:58] <Corwin> no, MS Visual studio based
[14:12:54] <elektrinis> oh boy
[14:12:58] <elektrinis> 600MB
[14:13:01] <Corwin> :D
[14:13:01] <voodoofish430> youch.
[14:13:02] <elektrinis> the hell
[14:13:36] <elektrinis> a bit heavy for fancy notepad
[14:14:21] <specing> I've seen 6GB IDEs...
[14:16:44] <elektrinis> android for example? :)
[14:16:58] <elektrinis> ok, so its done downloading..
[14:17:47] <specing> No, FPGA
[14:18:05] <specing> there is no android ide
[14:18:16] <specing> only a 100KB or so eclipse plugin
[14:18:44] <voodoofish430> Yeah, it's just a big sdk download.
[14:19:16] <elektrinis> :) ok
[14:23:23] <elektrinis> http://translate.googleusercontent.com/translate_c?hl=en&ie=UTF8&prev=_t&rurl=translate.google.com&sl=zh-CN&tl=en&twu=1&u=http://www.mcuzone.com/shop/%3Fproduct-58.html&usg=ALkJrhiK8puUffp8Bc8tB3OWVmUA-nB8-w
[14:23:32] <elektrinis> Y280 = $5??
[14:24:03] <elektrinis> 3.5*
[14:24:27] <Corwin> quite cheap... even for fake clone
[14:24:49] <Corwin> uart interface :D
[14:25:13] <Corwin> ah, usb too
[14:25:23] <Corwin> i want that
[14:25:32] <Corwin> ... so i can take it apart
[14:26:46] <Martyn> When dealing with Yen, just divide by 120
[14:26:56] <Martyn> If you divide by 100, you're still close
[14:28:56] <elektrinis> mm
[14:29:19] <Martyn> But that's not YEN elektrinis
[14:29:20] <elektrinis> I want RS232 device so I can connect it to BT module in SPP mode
[14:29:21] <Martyn> that's YUAN
[14:29:26] <Martyn> it's about $43
[14:29:29] <elektrinis> oh.
[14:29:35] <Martyn> actually, $48
[14:29:42] <elektrinis> thats a bit more than 3.5$
[14:29:45] <Martyn> 280*0.17-ish
[14:29:51] <Martyn> yep ...
[14:29:57] <Martyn> and don't forget, there's a pretty hefty tax too
[14:30:08] <Martyn> (they never, ever print the chinese import taxes)
[14:30:12] <elektrinis> yeah, 25% on import
[14:30:20] <Martyn> So, closer to $55
[14:30:30] <Martyn> which is exactly what I'd pay for that
[14:30:40] <elektrinis> :)
[14:30:55] <elektrinis> i'm still looking for a hacked schematic/hex
[14:31:03] <elektrinis> so I could build like I like it
[14:31:09] <elektrinis> with embedded BT module
[14:31:33] <elektrinis> so you just stick it to your project board... pair your laptop with it... and your'e on
[14:31:53] <Martyn> Do you need a JTAGICE2? Or can you just live with a jtagice?
[14:32:01] <elektrinis> and like "look maa, no wires!!"
[14:32:04] <Martyn> If you can, there are plenty of MK1 clones out there in the market
[14:32:09] <elektrinis> I need debugwire
[14:32:21] <elektrinis> MK1 does not support it, AFAIK
[14:32:29] <Martyn> nope, sure doesn't
[14:32:38] <Martyn> one of the reasons that mk2's are $150 or so
[14:32:45] <elektrinis> yes.
[14:33:14] <elektrinis> the mk2 looks pretty old, like 4 years or so
[14:33:18] <elektrinis> and still no clones? :(
[14:33:25] <elektrinis> I mean open clones
[14:34:05] <elektrinis> google for like 4 hours..
[14:34:08] <elektrinis> nothing.
[14:34:23] * elektrinis sad
[14:35:00] <elektrinis> ok, leaving this at that and going to meet the gf. looks like no more electronics tonight.
[15:23:30] <Essobi> RikusW: what'sup
[15:28:26] <RikusW> Hi Essobi
[15:29:14] * RikusW have been walking around outside until about 21:30... (now 22:20)
[16:06:58] <OndraSter_> this looks like a waste of a traffic
[16:07:02] <OndraSter_> 2kB/s is my current speed
[16:07:09] <OndraSter_> RikusW, I am slowly reaching your internet! :D
[16:08:14] <RikusW> heh
[16:08:43] <RikusW> slow server ? or slow connection ?
[16:12:25] <OndraSter_> well
[16:12:32] <OndraSter_> my ISP's router kept rebooting every 5 minutes
[16:12:38] <OndraSter_> while after the reboot the internet worked about 10 seconds
[16:12:54] <OndraSter_> then nothing until next reboot
[16:12:56] <OndraSter_> I plugged my laptop into my home network, connected it to my phone and re-routed all traffic through that
[16:13:03] <OndraSter_> well, my monthly FUP was depleted rather quickly
[16:13:22] <OndraSter_> I went to the center for few hours
[16:13:28] <OndraSter_> when I came back I couldn't even ping the ISP's router anymore
[16:13:32] <OndraSter_> so. yeah
[16:13:40] <OndraSter_> I am stuck on depleted FUP on phone
[16:13:41] <OndraSter_> = 2kB/s :(
[16:15:18] <RikusW> now I don't have the slowest internet anymore ;)
[16:16:33] <RikusW> My GPRS connection goes down occasionally, probably due to maintenance or overload...
[16:18:08] <OndraSter_> well I get 2 bars of HSDPA in this room
[16:18:12] <OndraSter_> but it is pointless when you have depleted FUP :D
[16:18:20] <OndraSter_> so I have switched the phone to use only GPRS/EDGE
[16:18:59] <OndraSter_> it actually consumes less battery while in standby :D
[16:19:02] <OndraSter_> by a lot
[16:19:49] <RikusW> my GPRS modem got a 1m long yagi
[16:20:01] <OndraSter_> lol
[16:20:04] <RikusW> so no problems with signal strenght
[16:20:06] <OndraSter_> imagine 1m long yagi on your phone
[16:21:21] <RikusW> you'll get noticed everywhere :-D
[16:24:04] <OndraSter_> :D
[16:33:20] <specing> RikusW: lol
[16:34:17] <specing> I remember the signal strength on my old SAGEM mobile with an antena
[16:34:46] <specing> You could went hiking to the middle of the alps and still get full reception
[16:34:56] <specing> and the 200mAh battery lasted for two weeks
[16:35:29] <specing> ...compared to my new android that has 1/2 signal on the open and drains its 1.3 Ah battery in 2 days
[16:35:43] <specing> Fuck I wish they kept making those pohnes
[16:36:24] <specing> 0.o that phone would last two months on that battery!
[16:36:39] <RikusW> Imagine how long the old phone could last on a new 1.3Ah battery =-O
[16:36:52] <specing> what I was saying
[16:37:03] <specing> 2 months man
[16:37:11] <specing> TWO WHOLE MONTHS
[16:37:29] <RikusW> seems like technology goes downhill sometimes....
[16:39:05] <RikusW> Imagine how slow windows 7 would be on a pentium 1
[16:39:41] <specing> heh
[16:40:09] <specing> Imegine how fast the PDP os would be on a modern xeon
[16:44:09] <OndraSter_> specing, imagine how many applications would exist
[16:44:21] <OndraSter_> let alone games
[16:44:24] <OndraSter_> with supadupa graphics
[16:44:27] <Steffanx> *wouldnt
[16:44:45] <OndraSter_> *wouldn't (if you want), but would fits too
[16:44:53] <OndraSter_> it is ironic though
[16:45:23] <specing> Imagine running 50 quakes simultaneusly on 512M ram
[16:45:35] <Steffanx> Go for it
[16:45:45] <OndraSter_> wait
[16:45:51] <OndraSter_> 50 quakes on 512MB RAM?
[16:46:03] <OndraSter_> *fap fap fap*
[16:46:05] <OndraSter_> lol
[16:46:09] <OndraSter_> or Wolf3Ds!
[16:46:28] <Steffanx> I prefer Doom, never liked wolfenstein 3d
[16:46:32] <OndraSter_> heh
[16:50:03] <tlvb> by sequencing 'DEC r16' and 'BRNE label' I should jump to label if r16 != 1 (i.e. --r16 == 0) and continue otherwise, correct?
[16:51:27] <age> hi. I'm trying to get an usb bootloader into atmega8535. what troubles me is bootloader address. makefile specifies it as 1800 for 8k devices, but in datasheet it's 0xc00, which is 3072. what's right?
[16:52:24] <RikusW> tlvb: yes
[16:52:40] <Kevin`> age: there's a difference from specifying it in words or bytes
[16:52:55] <RikusW> tlvb: when r16 is decremented to 0 the branch wont be taken
[16:53:01] <age> oh
[16:53:46] <RikusW> age: and mega8535 don't have usb...
[16:53:55] <age> RikusW, it's software usb
[16:54:04] <age> usbasploader
[16:54:43] <RikusW> compile it, and see what the smallest boot section is it will fit into
[16:54:53] <age> Kevin`, I presume ld takes it bytes?
[16:55:01] <RikusW> then calculate the address and set the fuses correctly
[16:55:15] <age> RikusW, just about 2k bytes
[16:55:22] <age> and I did set fuses for that.
[16:55:31] <RikusW> then you'd place it at 6kb
[16:55:41] <RikusW> 1024 words ?
[16:55:57] <age> yes
[16:56:51] <RikusW> the byte address is 0x1800 then
[16:57:13] <age> let me try that again.
[16:57:17] <RikusW> that should be the address in the hex file too (you can read .hex files right ?)
[16:57:38] <age> I can hexdump it, but not sure what to look for
[16:57:55] <RikusW> open the hexfile with a text editor
[16:59:37] * specing slaps age for not using $ file before talking
[16:59:38] <RikusW> :LLAAAATT.......CC
[16:59:50] <RikusW> age; AAAA is the address in the hexfile
[16:59:51] <age> and?
[17:00:14] <RikusW> so you can check it it is :10180000......
[17:00:29] <age> the top line is
[17:01:21] <age> I've also added a bit of code to see if it gets to bootloader, now let me flash it and see.
[17:02:40] <RikusW> age: did you set the BOOTRST fuse ?
[17:02:55] <age> yes.
[17:03:02] <tlvb> Anyone know the proper constraint to use if I want to write to UDR0 in inline assembler? STS is listed as needing a label and a register, but I seem unable to find a constraint that matches label
[17:03:03] <RikusW> good
[17:03:32] <RikusW> tlvb: on which avr would that be ?
[17:03:45] <age> apparently it doesn't get to bootloader
[17:04:13] <tlvb> RikusW: atmega164 ...of course I could get the hard-coded address from the register description, but it seems to be the wrong way
[17:04:24] <age> I've added code to set PD0 high on bootloader init, and have a LED on it, no light there.
[17:04:52] <tlvb> RikusW: I'm following
http://www.nongnu.org/avr-libc/user-manual/inline_asm.html
[17:05:19] <RikusW> tlvb: UDR0 = 0xC6, so you have to use STS in this case
[17:05:45] <RikusW> but beware of using STS with names defined for use with IN/OUT.....
[17:06:04] <RikusW> it can cause HARD to find bugs :-P
[17:06:41] <RikusW> tlvb: did you include avr/io.h ?
[17:06:43] <age> oh, maybe not
[17:06:46] <tlvb> RikusW: yes
[17:07:45] <tlvb> RikusW: (on 0xC6): yes I know, the problem is that I was trying to avoid hardcoding 0xC6, but using the gcc/avr-libc way of putting a placeholder in the asm, and then specifying the standard UDR0 used in normal c at the end, as per the link above
[17:09:29] <age> no, definitely not in bootloader
[17:09:36] <RikusW> see if it is defined in the m164 header
[17:09:52] <RikusW> or try UDR instead of UDR0 ?
[17:14:31] <age> funny. it was plain 1800 in the makefile, not 0x1800
[17:14:43] <tlvb> that would not work, as it has two usarts... perhaps I'm thinking about it the wrong way, I want to write to UDR0, but UDR0 is not a register in the normal sense I'm writing to the address UDR0 in the sram, maybe UDR0—the address should be in the input list
[17:15:42] <tlvb> it compiles at least...
[17:16:48] <RikusW> age: everythng is hex in a .hex file
[17:17:44] <OndraSter_> do you have the code for the inline assembly?
[17:17:47] <age> RikusW, the makefile
[17:17:50] <OndraSter_> pls put it here the line or two
[17:17:55] <OndraSter_> I can'T open anything on the internet :)
[17:18:13] <age> but that changed nothing.
[17:19:37] <tlvb> it puts in the right 0xc6 in the disasm
[17:19:59] <tlvb> OndraSter_: was that directed at me?
[17:20:05] <tlvb> is pastebin ok?
[17:21:02] <OndraSter_> no
[17:21:05] <OndraSter_> my internet goes less than 2kB/s :)
[17:21:11] <OndraSter_> it is like 2 or 3 lines of code, right?
[17:21:44] <RikusW> OndraSter_: now go surfing on facebook :-P
[17:22:08] <RikusW> (Then you'll know why I just about never go there anymore)
[17:22:19] <tlvb> kind of, but it needs context, and it's not the really the actual asm that gives/gave me problems, but here goes:
[17:22:22] <tlvb> sts %[udr0], r17 ; 2l 9l \n\t\
[17:22:28] <tlvb> ...tabs are bad I see...
[17:23:16] <OndraSter_> and the error?
[17:23:31] <tlvb> "sts %[udr0], r17" is one of the assembler lines, and : [pind] "I" (_SFR_IO_ADDR(PIND)), [udr0] "" (UDR0)
[17:24:35] <tlvb> is the input list, this compiles, and seems to put the right 0xC6 at the right place, I'm just unsure of if 1) UDR0 should be on the input list, as I'm trying to write to UDR0 and 2) if I really should leave the constraint string empty
[17:25:17] <tlvb> I do seem to have some bug somewhere though, but it seems that this is not it
[17:25:17] <OndraSter_> hmm
[17:25:34] <OndraSter_> if disassembly fits and no errors are shown - it is not there :)
[17:26:23] <tlvb> then I have to abandon this goose chase and see if the bug is somewhere else then...
[17:29:00] <tlvb> I'm trying to sample a pin at specific intervals, storing it at a bit in a byte, and each 8th sample push the byte onto UDR0, but it does not seem to work
[17:30:41] <RikusW> do you check that the uart has finished transmitting first ?
[17:30:49] <tlvb> but I do possibly have an unreliable signal source and my logic analyzer weighs 10kg and is some 350km away
[17:31:27] <tlvb> RikusW: not explicitly, but I have done calculations, so that it should not sample more often than it is able to handle, with regards to the set baud rate
[17:32:19] <RikusW> maybe check UDRE anyways
[17:33:30] <tlvb> Perhaps I should, if only to eliminate that possible source of error, I should probably check the flow control pins of the ft232 so I'm not overrunning it's buffer also
[17:33:58] <OndraSter_> believe me
[17:34:02] <OndraSter_> FT232 can do more than mega :)
[17:34:25] <tlvb> OndraSter_: yes, but it is not guaranteeing that the bytes are sent as they arrive
[17:35:18] <tlvb> so it is possible to fill the buffer (as in I have done it in a test application just to make sure that flow control works)
[17:35:27] <OndraSter_> I've been pushing 1Mbit through ft232 to mega128a just fine, no buffer overruns, no nothing
[17:35:36] <OndraSter_> hmm
[17:35:40] <OndraSter_> I haven't used flow control
[17:35:46] <OndraSter_> I just used "simple & it works" :)
[17:36:17] <tlvb> How about 1Mbit continuous from mega to computer?
[17:37:04] <tlvb> never mind, it is possible that my usb bus is just encumbered
[17:37:31] <OndraSter_> hmm only computer -> mega
[17:38:01] <specing> duh 1Mbit is 128000 bytes/s
[17:38:04] <tlvb> I think that there may be some more work behind the scenes with buffers and stuff in that direction
[17:38:10] <specing> slow as ....
[17:39:08] <tlvb> I think you forgot a 0
[17:39:13] <specing> you get roughly 125 cycles between recieving another byte
[17:39:14] <tlvb> oh
[17:39:15] <tlvb> no
[17:39:19] <age> hmm, if I bricked my atmega by setting wrong clock fuses, it would not program?
[17:39:21] <tlvb> I read bits, not bytes
[17:39:22] <tlvb> sorry
[17:39:49] <specing> Im talking about the best case scenario that never happens
[17:40:06] <specing> yes, rs232 is only 80% efficient
[17:40:48] <tlvb> age: yes, that is a symptom
[17:40:57] <specing> so you get 156 cycles in between
[17:41:21] <age> good
[17:42:37] <RikusW> age: you can supply an external clock to XT1
[17:43:18] <age> RikusW, I didn't brick it, just wanted to verify that bad clock is not the cause of non-working bootloader
[17:43:29] <tlvb> specing: the case were I had the ft232 throw flow control at me were when I sent from the atmega at 1.25Mbd continuously (using interrupts), the atmega cycles were not important, I were checking if I could fill the internal buffer of the ft232 faster than it could scoop the data onto the usb bus
[17:44:38] <specing> Isn't the FT232 480Mbps?
[17:44:42] <tlvb> age: unless you have disabled normal programming you should not be unable to do that just becauser you have downloaded a bootloader (or failed to do so) I think
[17:44:47] <specing> Or is that FT232H only?
[17:44:55] <OndraSter_> interesting
[17:44:57] <RikusW> age: software usb usually requires an exact clock (crystal)
[17:45:02] <OndraSter_> I saw post when people were pushing 2Mbaud in arduino
[17:45:05] <OndraSter_> which has no flow control
[17:45:38] <age> RikusW, there is a crystal. I've had different atmegas working with this usb stack, that's not the problem
[17:45:50] <age> it just never gets to the bootloader.
[17:46:00] <tlvb> specing: I don't know honestly, but there is a lot more overhead too (although it would hardly account for a drop from 480Mbit to 1.25Mbd)
[17:46:52] <specing> use the FPGA interface ;P 30MBytes / sec
[17:47:14] <tlvb> OndraSter_: at 20MHz 2.5Mbd should be possible
[17:49:11] <OndraSter_> FT232 is full speed only
[17:49:25] <specing> blegh
[17:49:41] <OndraSter_> FT232R
[17:50:04] <RikusW> FT22232H is hi speed
[17:50:11] <RikusW> *FT2232H
[17:50:22] <tlvb> I think I have the RL