irclog2html for #edev on 20070321

00:54.04*** join/#edev CosmicPenguin (n=nobody@www.crouse-house.com)
00:54.04*** mode/#edev [+o CosmicPenguin] by ChanServ
02:03.44*** join/#edev lgn2 (i=lgn@gateway/tor/x-f35ca2a30f215d6b)
02:25.27*** join/#edev catastr0p (n=catastr0@d14-69-178-65.try.wideopenwest.com)
02:54.51*** join/#edev Crofton|laptop (n=balister@66-207-66-26.black.dmt.ntelos.net)
03:07.49*** join/#edev diamat (n=cerne@2001:5c0:8fff:fffe:0:0:0:5b4f)
03:08.42*** part/#edev diamat (n=cerne@2001:5c0:8fff:fffe:0:0:0:5b4f)
04:03.07*** join/#edev redblue (i=star@ppp210.108-253-207.mtl.mt.videotron.ca)
04:36.27*** join/#edev lgn (i=lgn@gateway/tor/x-975d90c878dd12bc)
04:38.50flybackhttp://www.youtube.com/watch?v=9hCed_TT_KQ
04:48.53*** join/#edev DrZimmerman (n=theo@84-74-134-243.dclient.hispeed.ch)
05:22.53*** join/#edev lgn (i=lgn@gateway/tor/x-ef2024a7c7e39cf6)
06:02.15*** join/#edev lgn2 (i=lgn@gateway/tor/x-988c575570bae53c)
06:42.00*** join/#edev jbevren (n=jbevren@207.199.248.164) [NETSPLIT VICTIM]
06:42.00*** join/#edev dfarnsworth (n=dfarnswo@65.200.49.158) [NETSPLIT VICTIM]
06:42.01*** join/#edev ccjoe (n=ccjoe@www.clearcore.com) [NETSPLIT VICTIM]
06:42.01*** mode/#edev [+o jbevren] by irc.freenode.net
07:44.58*** join/#edev hw (n=hw@87.139.57.5)
08:05.24*** part/#edev T0mW (n=Tom@70.44.166.161.res-cmts.sth.ptd.net)
08:05.38*** join/#edev T0mW (n=Tom@70.44.166.161.res-cmts.sth.ptd.net)
08:10.17*** join/#edev hw (n=hw@87.139.57.5)
11:47.05*** join/#edev sjhill (n=sjhill@real.realitydiluted.com)
13:18.41*** join/#edev GPSFan (n=kenm@12.10.255.248)
13:18.41*** mode/#edev [+o GPSFan] by ChanServ
13:21.46*** join/#edev prpplague (n=dave@mail.americanmicrosystems.com)
13:32.26*** mode/#edev [+o prpplague] by ChanServ
13:58.05*** join/#edev purl (i=ibot@pdpc/supporter/active/TimRiker/bot/apt)
13:58.05*** topic/#edev is a channel for the discussion of hacking small consumer electronic devices, mostly ARM based
14:03.15*** join/#edev redblue (i=star@ppp089.108-253-207.mtl.mt.videotron.ca)
14:29.32chouimatmorning
14:59.54*** part/#edev wmat (n=wmat@CPE000f6641f21a-CM00111adf5f4a.cpe.net.cable.rogers.com)
15:09.20*** join/#edev AMDPenguin (i=nobody@nat/amd/x-17354e070241df61)
15:10.07*** mode/#edev [+o CosmicPenguin] by ChanServ
15:10.16*** join/#edev CP|too (n=nobody@www.crouse-house.com)
15:13.09*** join/#edev lgn (i=lgn@gateway/tor/x-631f6f8a9e6d9332)
17:21.57*** join/#edev T0mW (n=Tom@70.44.166.161.res-cmts.sth.ptd.net)
17:22.33T0mWprpplague: google on "linux openocd jtag" what is the second hit?
17:43.59chouimatprpplague: http://smtpd.develooper.com/
17:44.16chouimatooops wrong links
17:44.19chouimatprpplague: good link  http://despair.com/effort.html
17:50.40prpplagueT0mW: hehe
17:51.33T0mWprpplague: ok, so you see the same thing. cool.  I got an email from a company asking if I wanted to play with their USB JTAG thingy and then put the information on my webpage about what I found in reviewing their device.
17:51.57prpplagueT0mW: which company?
17:52.01T0mWprpplague: so, I was wondering "how the fsck did they find this webpage?".  so, I googled.
17:52.08T0mWprpplague: Amontec
17:52.30prpplagueT0mW: ahh
17:53.09T0mWprpplague: they are going to send me one of these: http://www.amontec.com/jtagkey-tiny.shtml
17:53.26T0mWprpplague: not an expensive thing, but looks like it might be cool to use.
17:54.22T0mWprpplague: I recently played with the Olimex ARM-USB-OCD device + OpenOCD.  They work pretty darned well together, but not as fast as the BDI2000
17:54.46T0mWprpplague: maybe I'll put my BDI2000 on ebay? heh
17:55.23prpplaguehehe
17:55.35prpplagueT0mW: yea we have our own ft2232 we designed
17:55.42prpplagueT0mW: works pretty darn good with openocd
17:56.35T0mWprpplague: openocd surprised me with the completeness of the telnet interface.  They even allow you to program flash and dump memory.
17:56.48Crofton|laptopdoes openocd work on linux?
17:56.57T0mWCrofton|laptop: YES!
17:57.25T0mWCrofton|laptop: http://www.openhardware.net/Embedded_ARM/OpenOCD_JTAG/
17:58.03T0mWCrofton|laptop: I used it on Mandriva 2007 with Insight
18:01.06prpplague~lart Crofton|laptop
18:01.40Crofton|laptophttp://www.sparkfun.com/commerce/product_info.php?products_id=7834
18:01.40Crofton|laptopout of stcok
18:01.41prpplagueCrofton|laptop: yea pretty popular item
18:25.51*** join/#edev ico2 (n=ico2@modem-3128.cougar.dialup.pol.co.uk)
18:56.24*** join/#edev lyakh (n=lyakh@dynamic-unidsl-85-197-31-110.westend.de)
19:14.00flyback:P
19:14.06flybackwith a max232 on it also
19:14.30flybackthen again this one is so easy even I can do it
19:15.02flybackfinally a juicebox mod that doesn't suck
19:39.43T0mWprpplague: so, what are you working on?
19:46.36*** join/#edev Crofton (n=balister@hc652174f.dhcp.vt.edu)
19:50.10prpplagueT0mW: sdio wifi card
19:50.44T0mWprpplague: ah, that 54mb/s stuff you were looking to do..
19:53.28prpplagueyea, marketing ya know
19:53.42prpplaguenot like any of the applications use more than around 5mbps
19:53.47T0mW~lart marketing weenies
19:54.25prpplaguei think i have everything to do the job now
19:54.33prpplaguejust a little iffy about some of documentation
19:54.41T0mWprpplague: engineer: "Yes, it is possible, but not with todays technologies"
19:55.26T0mWprpplague: marketing slime: "so, will the customer notice?"
19:55.54flybackavctually that would feel pretty damn good right now
19:56.06flyback~lart flyback's internal hemmroids
19:56.14flybackah that's better
19:56.43flybackor is that lance
19:56.46flybackhmmm
19:57.02T0mWprpplague: I'm just getting these three programs to act in concert: bootstrap, loader and runtime binaries.  The problem is the software authentication mechanism that I devised to keep customers from purchasing one copy and running it "anywhere".
19:57.09T0mWprpplague: fun
19:57.27T0mWprpplague: fun tracing three two flash images and one in RAM
19:57.27flybackso prpplague have you reviewed the pixcutor hack for the juicebox?
19:57.31flybackif so any gotcha's
19:57.40T0mWs/three/thru
19:57.48flybackcause from what I see basiclaly you can use stock mmc/sd no, soldering to the cards etc
19:58.02flybackso if i get sick of it I am only out like 50 cents for th epic
19:58.34T0mW<sigh> I broke something...
19:58.49flybacktom what did you do?
19:59.20prpplagueflyback: who the hell let you back in here?
19:59.36flybackGod did
20:00.03prpplagueflyback: yea i looked at the pixcutor stuff, smart hack, but not very useful
20:00.09flybackwhy not useful?
20:00.15flybackcan't it be used to exec code
20:00.41flybackI been reading the forum posts
20:00.42prpplagueyea, but without getting a ton of other work done, you can't boot an os properly
20:00.53T0mWprpplague: I just wish that Insight has user programmable keys.  I get tired of typeing in "symbol-file ../loader/loader.elf"
20:00.56flybackbut that's all coding correct?
20:01.14prpplagueT0mW: hehe
20:01.16prpplagueflyback: yea
20:01.22T0mWsux
20:01.28flybackso I think it's damn useful
20:01.39flybackcompared to trying to solder 20 wires to a sd card and trashing it in the process
20:01.59flybackand too much of a pain in the ass
20:02.41flybacki've suffered a 10-20% loss of close up vision in the last yr so my chances of small smt soldering are going down before I even try :P
20:03.03flybacknice 33 yr bday present :/
20:03.19T0mWprpplague: hmmm, I can reprogram the keyboard from /etc/inputrc , correct?
20:04.08flybackbtw I been learning virtual pc and vmware stuff at my new job the last 3-4 months
20:04.10flybackWOW!
20:04.46flybackI could setup cross compilers linux setups a lot easier now
20:08.38flybackI don't give a crap if it's slower it's a lot less hassle
20:09.19flybackanyways I need sleep I been at work since 6am-3pm and haven't slept in 2 days
20:12.43prpplagueT0mW: been awhile, but you make some mods in there
20:13.14T0mWprpplague: I'm looking at loadkeys & dumpkeys to modify the kernel keyboard string table
20:13.19prpplagueT0mW: more likely you want to modify your keymap and do something like add the "winblows" key to type out the script
20:21.07T0mWprpplague: <sigh> this is going to take some time, as usual.  :-(
20:37.15*** join/#edev redblue (i=star@ppp042.108-253-207.mtl.mt.videotron.ca)
21:12.16CosmicPenguinargh!
21:12.36prpplagueCosmicPenguin: spill your coffee again?
21:12.45prpplagueCosmicPenguin: or worse yet, your beer?
21:21.40prpplagueCosmicPenguin: are you familiar with the CMD25 for sd specs? i see you've posted some things on the alchemy lists
21:22.31prpplagueCosmicPenguin: actually you were talking with ossman
21:22.32prpplaguehmm
21:22.43CosmicPenguinI can't remember CMD25 right off - refresh my meory?
21:22.52prpplaguemulti block write
21:25.26CosmicPenguinahh - yeah
21:27.53prpplagueCosmicPenguin: you got a second to discuss it? i'm totally confused about the actual logic flow
21:28.16CosmicPenguinI'll do what I can, my memory is fuzzy
21:28.54prpplagueCosmicPenguin: so you issue a CMD25 with the address, block size, and block count, then what?
21:29.15CosmicPenguinthen you keep sending blocks until a you send a stop
21:29.55prpplagueCosmicPenguin: ok, so the question is, how do you send them?
21:30.21prpplagueCosmicPenguin: is the data encapulated in another cmd25 with crc?
21:30.34prpplagueCosmicPenguin: or do you just start writing to the fifo?
21:30.46CosmicPenguinThat I can't remember
21:31.02prpplaguedoh
21:31.41prpplagueCosmicPenguin: know of any code that uses a cmd25 in it that i could browse?
21:42.18CosmicPenguinthe kernel code should support it
21:42.25T0mWprpplague: doesn't that work that you have to send a block, then wait for the status byte(s) of "OK", then simply pump the next block, wait for status, etc, etc, ?
21:43.05T0mWs/wait/poll for/
21:43.24T0mWs/wait/poll for/g
21:43.29T0mWheh
21:43.39MonMothapurl: botsmack
21:43.41purlOWW!
21:44.11MonMothaT0mW: you know much about SSTL?
21:44.23prpplagueT0mW: yes in theory, the question is what does "you have to send a block" mean
21:44.38prpplagueCosmicPenguin: yea found it, had an old mmc dir
21:44.47prpplagueCosmicPenguin: looking now
21:45.12T0mWMonMotha: SSTL?
21:45.16T0mWMonMotha: wassat?
21:45.20MonMothaT0mW: evidently no :)
21:45.25T0mW:)
21:45.27MonMothaSSTL is the signaling method used for DDR SDRAM
21:45.33T0mWoh
21:45.36MonMothamostly specifies drive levels and termination options
21:46.04T0mWMonMotha: I thought prpplague was talking about mmc stuff
21:46.14T0mWMonMotha: stop
21:46.19MonMothahe is, I had a completely different question :)
21:46.49T0mWMonMotha: I'm shell-shocked, yeah, I *finally* saw that you had a different subject. :P
21:47.23prpplaguehehe
21:47.24T0mWMonMotha: afaik, the SD/DDR uses a command set
21:47.50MonMothayeah.  SSTL is the physical level - output buffers, inputs, termination, etc.
21:48.10T0mWMonMotha: ah, well, keep 'em short
21:48.12MonMothaI was just curious if you knew much about it since there's something I'm scratching my head over and can't find much info on
21:48.13T0mWheehee
21:49.03T0mWMonMotha: should be the same considerations as with any other RAM device: input capacitance, trace impedance, etc.?
21:49.44T0mWMonMotha: but no, I have not looked at DDR operation, so, I really cannot comment on it.
21:49.57T0mWMonMotha: SDRAM, yes.
21:50.08MonMothawell, the issue is that SSTL specifies two classes, which appear to differ only in the drive level of their outputs (current source/sink).  They then specify different recommended termination schemes.  I was curious if it is possible to mix, since I have a device designed for class II operation that I want to hook up to a Spartan 3E, which only supports class I outputs
21:50.46MonMothafrom what I've seen others do, it is evidently possible, but I can't find any information related to it
21:51.13T0mWMonMotha: what is the difference in the specs between class I & II? logic levels, output drive levels?
21:51.42MonMothaoutput drive levels, that's all
21:51.56MonMothaand due to the output drive differences, the termination recommendations are different
21:52.10T0mWyeah, they woould be
21:52.44T0mWMonMotha: drive current levels between class I & II are that radically different?
21:52.58MonMothaclass II is 2x class 1 to account for the double termination
21:53.36T0mWMonMotha: Ah, so, sink + source termination with class II where class I only used a sink (pullup) terminator?
21:53.39MonMothahttp://gigapet.monmotha.net/~monmotha/sstl_2spec.pdf
21:53.47T0mWlooking
21:53.48MonMothacorrect, though there's always a series termination
21:54.26T0mWMonMotha: You don't have permission to access /~monmotha/sstl_2spec.pdf on this server.
21:54.43T0mWheh
21:54.54MonMothareload
21:55.16T0mWgot it
21:55.16MonMothafortunately, it's not very long
21:55.28MonMothaunfortunately, it doesn't go into much detail, other than noting that oddball schemes are possible
21:55.36T0mWMonMotha: yeah, just chock full of examples then...
21:56.27MonMothaI could just use SDRAM for this, though there is a potential bandwidth issue
21:56.33MonMothaand DDR is almost easier to find, now
21:57.39MonMothaeven DDR266 (133MHz clock) would be pushing it for a 32bit data bus in some cases (that will likely never get used)
22:01.46T0mWMonMotha: looks like table 4 on page 12 is most pertinent info
22:03.43MonMothayeah, still hard to tell what happens when you want to mix them on bidirectional lines
22:03.51T0mWMonMotha: this is a educated guess: I think that the limitation on the Spartan logic using Class I only is due to the output buffer impendance being higher, therefore, they are disqualifying operation with class II due to the higher current demands that some terminations could result in.
22:05.15T0mWMonMotha: the pdf does state that the values of the resistors are/can be determined empirically.  So, a Class II device is not necessarily incapable of interfacing with a Class I. Given that you don't have to exceed the Class I termination / buss drive limits?
22:05.55MonMothaT0mW: that's what I was thinking.  Basically, just treat the whole thing as a class I system
22:06.24MonMothaI actually just noticed that the parts I'm using have a reduced drive strength option (54% of class II, which is almost exactly class I), which would make such an assumption quite valid
22:06.38T0mWMonMotha: IMHO, the controlling factor in the use of the Class II termination / drive levels would be the amount of ringing (over / under shoot) of the signals.
22:07.05MonMothayes.  Class II seems designed for longer runs where more "aggressive" termination schemes are required
22:07.20T0mWMonMotha: one way to reduce the termination requirement is to run the DDR at a lower clockrate?
22:07.24MonMothaprpplague: it's always fun when wires are no longer wires
22:07.40prpplaguehehe
22:07.42MonMothaT0mW: well, in this case I almost need the extra bandwidth
22:07.43T0mWMonMotha: yeah, they become transmission lines.
22:07.52MonMothaT0mW: yup
22:08.10T0mWsquare waves have all odd multiples of harmonics on them
22:09.05MonMothaalways fun when you get components up in the GHz range.  A 1" long trace is suddently a rather lengthy transmission line
22:09.07T0mWthe base frequency, IIRC, of the fundemental is the period of the square wave
22:09.18T0mWheh
22:09.45T0mWMonMotha: that's why they pay you the big money, to learn how to use a spectrum analyzer
22:10.01MonMothaT0mW: yeah...signals was a, um, fun?, class
22:10.23T0mWMonMotha: well, for me maybe: callsign WN3L
22:10.48MonMothaT0mW: I didn't mind signals actually (also an amateur radio operator, albeit with a much lenghtier call)
22:11.02prpplaguesometimes i really hate implementing stuff that no one can help me with
22:11.22MonMothaprpplague: the problem with SD is that the publicly available specs are so minimal
22:11.34T0mWprpplague: tired of playing 'Columbus' ?
22:11.49T0mWprpplague: "go where no man has gone before..."
22:11.53T0mW:P
22:12.40T0mWMonMotha: well, good luck with that, I think I'll stay below 400MHz for a while yet.
22:12.49MonMothaoh, this is below 400MHz
22:13.14MonMothajust the edges are so fast that things still misbehave
22:15.08prpplagueMonMotha: yea, i have plenty of specs, problem is that none of them really explain the procedure very well
22:19.04prpplaguelooks like for each block transmission you have 8 bits for the start token, 7 bits for the crc and a stop bit
22:21.31prpplaguequestion is whats in the start token, i assume its the two standard start bits, but what of the other 6?
22:22.55*** join/#edev T0mW (n=Tom@70.44.166.161.res-cmts.sth.ptd.net)
22:54.27*** join/#edev Crofton (n=balister@66-207-66-26.black.dmt.ntelos.net)
23:09.26*** join/#edev joe_bleau (n=joe_blea@adsl-69-152-202-58.dsl.fyvlar.swbell.net)
23:12.24*** join/#edev mattastrophe (n=bigboota@CPE0004e28a9289-CM000a73a14ed4.cpe.net.cable.rogers.com)

Generated by irclog2html.pl by Jeff Waugh - find it at freshmeat.net! Modified by Tim Riker to work with blootbot logs, split per channel, etc.