irclog2html for tuxscreen on 2001.08.01

02:16:14Russdid someone have a tuxscreen and an osscilliscope?
04:28:31sgodsellHello tuxers ;
04:28:34sgodsell;)
07:38:04TangentRuss|werk : Yep.. I got a Tux and a 5MHz scope
08:03:01RussI have to go to sleep...can you hook it up to....
08:03:40Russpin 2 of IC305 on the daughter board?
08:04:16Russer, sorry, pin 3
08:05:11Russthen try some keys (esc, f4)
08:05:21Russand see what they look like
08:05:55Russ(might try in 2 as well, see if the buffer is doing anything)
08:07:31RussIC305 is right to the left of the right edge of the board
08:07:53Russwhere it says "ASSY 848036638"
08:08:50Russmake sense?
08:12:03Russsleep
08:22:21TangentYep.. makes sense.. I'll see if I can find some time today whilst you're asleed... got things to do for a couple of hours first
16:46:07TangentRuss : You awake yet?
16:46:22Russ|werkyes
16:46:22TangentRuss : I got my scope all rigged up, and I'm looking at the keystrokes now...
16:46:31TangentRuss|werk : Ah.. cool :)
16:47:10TangentEverything happens in 0.8ms steps
16:47:30TangentI guess the clock on the keyboard runs at some multiple of 1250Hz
16:47:52Russ|werkIt should be periods of a frequency (~38kHz or ~56kHz) ond periods of silence
16:47:57TangentBTW, pins 2 & 3 give the same signals.. just one at 3.3V, and the other at 5V
16:48:23TangentPicl a key, and I'll draw a picture of what I see for you
16:48:34Russ|werkthe periods of active frequency are multiples of 800us
16:48:50TangentYep
16:49:30Russ|werkin the active periods, what is the frequency?
16:50:49TangentEssentially, we have a 1250Hz square wave, with occasional cycles missed out...
16:51:12TangentI'll draw a diagram of what's on the scope
16:56:18Russ|werkyou want to invert it
16:56:30Russ|werkas the signal is inverted
16:56:50Russ|werkthe square wave should be made up of a smaller square wave
17:03:09Tangenthttp://ipaq.secret.org.uk/piccies/EscKey.gif
17:03:24TangentThat's the keydepressed sequence
17:03:43TangentWhen no key is pressed, the signal is at 3.3V (like you say.. it's inverted)
17:04:02TangentMy scope isn't clever enough to invert it for me
17:17:34TangentBZFlag : Hi
17:19:59tvramanTangent --ibot helped me track you down --
17:20:27tvramanI'm still trying to recover intimate on my ipaq --device stil doesn't see the pcmcia slots on the one card or dual card sleeve.
17:21:29Tangenttvraman : AliC is on #handhelds.org at the moment.. I've just asked him for the 244bootstrap image that he has working...
17:21:59Tangenttvraman : Hopefully that'll get you working again
17:22:53tvramanOkay Tangent --thanks! shall I look for you on channel intimate rather than adding noise here?
19:12:24nikosis away: Errand
19:34:24nikosis back
19:45:23nikosis away: Errand again
20:00:08nikosis back
20:12:36nikostvraman: Did you catch that URL?  Kind of noisy in #handhelds.org right now.
20:14:51nikostvraman: You there?
20:14:56nikosraman: Hello?
20:21:35Tangenthttp://ipaq.secret.org.uk/TuxScreen/KeyStrokes.html
20:24:21TangentShame that Russ|Werk isn't here to see that
20:24:46nikosBut Russ is.
20:25:52TangentYeah.. but I think that Russ is at Werk.. I was chatting to him and then... Ping Timeout
20:25:56Russ|werksorry about that tangent
20:25:57TangentBut speak of the devil
20:26:05TangentRuss|werk : http://ipaq.secret.org.uk/TuxScreen/KeyStrokes.html
20:26:34Russ|werkpeople at my office decieded to "Hi! How are you?"
20:27:16TangentRuss|werk : Pesky Real-Life interruptions :)
20:27:20Russ|werkI have a drive copying box here that I use for tech work
20:27:33Russ|werkit makes the drives availabe on the network
20:27:45TangentRuss|werk : As you can see, I found the cable and software for my scope
20:27:46Russ|werkif infected the drives I was copying
20:27:57Russ|werkthen opened the "halt" share and shut it down
20:28:47Russ|werkif you zoom in on those pulses, can you see the 38kHz or 56kHz carrier?
20:29:09TangentRuss : No.. that's the finest I can see
20:29:17Russ|werknice drawing btw
20:29:40TangentIt may just be that there's an RC filter that's turning the 56KHz signal into something easier to decode
20:30:20Russ|werkit goes from the ir diode, to the buffer, to the sa1100
20:32:57TangentRuss|werk : Judging by this circuit someone built.. (http://www.lirc.org/receivers.html), the TK19 IR transceiver deals with the 56KHz bit for us
20:33:25Russ|werkwhat is the exact timing of the pluses
20:33:31Russ|werkor, pulses
20:33:44Russ|werk800? 833.3? (usec)
20:34:01TangentRuss|werk : Everthing happens in 0.8ms steps
20:34:22Russ|werkto what accuracy
20:34:53TangentRuss : I'll see if I can sample a little more closely...
20:35:19Russ|werk1200 baud is 833.33usec  per bit
20:35:55Russ|werkBZFlag: is it possible to do a register dump in inferno (peek some memory locations)?
20:37:27TangentRuss|werk : OK.. looks like the pulses are slightly asymettric (or may just be scope error)... I get the 3.3v part of the signal as 0.75ms, and the 0v part as 0.85ms
20:40:09Russ|werktake a pulse + a space and divide by 2
20:41:13Russ|werkI'm spoiled with the lab oscilliscopes that do all sorts of automesurements
20:41:48TangentA pulse+space is 1.6ms... every time... tho that looks like 740us high, and 860us low
20:42:15TangentRuss : I seem to remember something like 741 85x being a common sequence
20:42:32TangentWhen we tested with mode2 on the iPAQ
20:42:35Russ|werkcan you tell if its 1.66666 or 1.600000
20:43:51Tangentcan't really tell... is LCD display on my scope... ( http://www.velleman.be/Product.asp?lan=1&id=9373 )
20:46:20Russ|werkhmm
20:46:42Russ|werkso it looks like its a CIR receiver (will have to try to look it up)
20:46:53Russ|werkand the signalling is somewhere around 1200 baud
20:47:06Russ|werkto the regular uart should be used
20:47:38Russ|werkhow many us from the first drop to the first rise?
20:47:57Russ|werk3 * 800?
20:49:12TangentLooks that way
20:49:18TangentScope says 2.4ms
20:49:23Russ|werkit might be 8N2 ?
20:49:32Russ|werktwo bytes
20:50:22Russ|werkie, esc, start LLHLHLLHSS start HHLHHLLLSS
20:50:36Russ|werkor maybe 8E1/8P1
20:50:59Tangentneeds to read up on different serial encoding
20:51:07Russ|werkstart LLHLHLLHpS start HHLHHLLLPs
20:51:57Russ|werkon the esc at least, the parity would match up
20:51:57Russ|werka start bit is low
20:51:57Russ|werka stop bit is high
20:52:18Russ|werkand its LRZ, so LLHLHLLH becomes 00111101 I think
20:52:29Russ|werkcould be off on my LRV technique
20:52:34Russ|werker, LRZ
20:52:48TangentWhat's LRZ?
20:53:08BZFlagRuss|werk: from sboot you can dump memory, that should handle all the register values you need to see, no?
20:53:24Russ|werkIt means that a transition is logical one, and no transition is logical zero (used for rs232)
20:53:47TangentNod.. but with inverted signals of course...
20:55:38Russ|werkBZFlag: can you look up the register values for the second serial port?
20:55:47Russ|werkno, its not inverted
20:55:54Russ|werkbecause the default state is high
20:56:15Russ|werkso it drops low for the start bit (triggers the oscilliscope)
20:56:34Russ|werkthen remains high after the last stop bit
20:56:43Russ|werka break is when you keep it low
20:56:46TangentYep.. that's certainly what it does
20:57:07TangentSadly I can't capture the key-release sequences
20:57:24Russ|werkthats ok
20:59:09Russ|werkparity matches up on all of them
20:59:21TangentCool :)
21:00:01Russ|werker....its nrz I don't know if you can still apply paritiy by counting the highs while its still nrz
21:01:20Russ|werkand I think the register dumps from the inferno will help to
21:01:24Russ|werkBZFlag: if you don't know which registers to dump, I'll look them up tonight
21:08:27BZFlagRuss|werk: I don't know off hand, but I'm leaving for NY tonight. I do have a tuxscreen here though.
21:08:48BZFlagI will be bringing at least one to NY, but It may have Linux on it my then ;-)
21:08:59Russ|werkhmm...
21:11:13Russ|werkox
21:11:16Russ|werker, ok
21:11:51Russ|werkcan you give me 0x80020060 to 0x80020084
21:12:01Russ|werker
21:12:14Russ|werk0x80020060 to 0x80020070
21:12:38Russ|werk0x80030000 to 0x80030010
21:13:20Russ|werk0x80060000
21:13:52Russ|werk0x80070060 to 0x80070064
21:15:21Russ|werkand the cleverly hidden, barely documented 0x90060028
21:15:28BZFlagslow down... ;-)
21:15:51Russ|werkso, just those 40 registers...
21:16:18Russ|werker, 48 registers
21:16:25BZFlagcrap, where's a null modem cable when I want one?
21:16:26Russ|werker, 40
21:16:35Russ|werkno, 36 registers
21:17:03Russ|werkno, I'm changing my mind again, thats only 12 registers
21:21:08BZFlagfound a cable...
21:21:58BZFlagfrom: >>> e 0x80070060
21:22:31BZFlagwishes that the addresses were absolute, no relative
21:22:47BZFlag>>> e 0x80070060
21:22:47BZFlag00000000: 00 00 00 00 00 00 00 00 00 00 00 00 ff ff 00 00  ................
21:22:47BZFlag00000010: 00 00 00 00 02 00 00 00 00 00 00 00 01 00 00 00  ................
21:22:47BZFlag00000020: 00 00 00 00 00 00 00 00 00 00 00 00 ff ff 00 00  ................
21:22:47BZFlag00000030: 00 00 00 00 02 00 00 00 00 00 00 00 01 00 00 00  ................
21:22:57Russ|werkthats the MCP
21:23:08Russ|werktry 0x90060028
21:23:27BZFlag>>> e 0x90060028
21:23:27BZFlag00000000: 00 00 0c 00 ff ff 03 00 00 30 00 00 00 00 00 00  .........0......
21:23:27BZFlag00000010: 00 00 00 00 00 00 00 00 00 00 00 00 92 d1 03 00  ................
21:23:28BZFlag00000020: 00 00 00 00 ff ff 3f 00 00 30 00 00 00 00 00 00  ......?..0......
21:23:28BZFlag00000030: 00 00 00 00 00 00 00 00 00 00 00 00 49 d1 03 00  ............I...
21:23:28Tangentwonders how geeks get a reputation for speaking in code ;)
21:23:50Russ|werkthat one makes sense to me
21:23:54BZFlagTangent needs to RTFM or RTJF
21:24:07Russ|werktry 0x80030000
21:24:24BZFlag>>> e 0x80030000
21:24:24BZFlag00000000: 09 00 00 00 00 00 00 00 b5 00 00 00 03 00 00 00  ................
21:24:25BZFlag00000010: 00 00 00 00 1e 00 00 00 00 00 00 00 1d 00 00 00  ................
21:24:25BZFlag00000020: 04 00 00 00 32 00 00 00 10 00 00 00 00 00 00 00  ....2...........
21:24:25BZFlag00000030: 1b 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00  ................
21:25:25Russ|werkthat one doesn't make sense
21:25:46Russ|werkI think you might need to know what the virtual mapping
21:26:07Russ|werkunless you can get absolute address with that inferno remote debugger thing
21:27:46BZFlaghmmm... I've found other regs that do make sense. where is mem timing again?
21:29:43BZFlag>>> e 0xa0000000
21:29:43BZFlag00000000: 1f 99 26 03 cf cc cc cc fc ff ff ff ff ff ff ff  ..&.............
21:29:43BZFlag00000010: 30 42 30 42 18 20 18 20 00 00 00 01 00 00 00 00  0B0B. . ........
21:29:44BZFlag00000020: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00  ................
21:29:44BZFlag00000030: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00  ................
21:29:50Russ|werksorry, daily disconnect (3pm)
21:29:52BZFlagthose look at least workable.
21:30:18Russ|werkMECR is blank
21:30:19BZFlag(no prob) ... so I expect that the MMU is setup flat.
21:30:57Russ|werkdo 0x90010004
21:31:02Russ|werksee if it changes
21:31:10Russ|werk(real-time clock count)
21:31:13BZFlag>>> e 0x90010004
21:31:14BZFlag00000000: 50 fe 25 51 00 00 00 00 00 00 00 00 03 00 00 00  P.%Q............
21:31:14BZFlag00000010: 00 00 00 00 00 00 00 00 00 00 00 00 ff ff ff ff  ................
21:31:14BZFlag00000020: b2 04 26 51 00 00 00 00 00 00 00 00 03 00 00 00  ..&Q............
21:31:14BZFlag00000030: 00 00 00 00 00 00 00 00 00 00 00 00 ff ff ff ff  ................
21:31:33BZFlag>>> e 0x90010004
21:31:33BZFlag00000000: c9 b4 33 51 00 00 00 00 00 00 00 00 03 00 00 00  ..3Q............
21:31:33BZFlag00000010: 00 00 00 00 00 00 00 00 00 00 00 00 ff ff ff ff  ................
21:31:33BZFlag00000020: db ba 33 51 00 00 00 00 00 00 00 00 03 00 00 00  ..3Q............
21:31:34BZFlag00000030: 00 00 00 00 00 00 00 00 00 00 00 00 ff ff ff ff  ................
21:31:42BZFlagyep/
21:32:47Russ|werklemme look over the 0x80030000 stuff again then
21:33:28BZFlag>>> e 0x80030000
21:33:28BZFlag00000000: 09 00 00 00 00 00 00 00 b5 00 00 00 03 00 00 00  ................
21:33:28BZFlag00000010: 00 00 00 00 1e 00 00 00 00 00 00 00 1d 00 00 00  ................
21:33:28BZFlag00000020: 04 00 00 00 32 00 00 00 10 00 00 00 00 00 00 00  ....2...........
21:33:28BZFlag00000030: 8d 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00  ................
21:36:20Russ|werkguess what the baud rate divisor b5 becomes
21:37:19Russ|werk1266 baud, or 780usec per bit
21:38:16Russ|werkusing odd parity
21:38:45Russ|werker
21:38:53Russ|werkhard to tell with the endianness
21:39:22Russ|werkyah, 1266 8O1
21:39:37BZFlagthat is strange.
21:39:49BZFlagis that clock dependant?
21:39:59Russ|werkyes, but it fits the signals tangent took *exactly*
21:40:17Russ|werkthe baud rate calculation goes like this:
21:41:04Russ|werkbaudrate = 3.6864E6 / (16 * (divisor + 1))
21:41:40Russ|werkthe divisor is 0xB5, or 181
21:41:49Russ|werk: huh?
21:41:49Russ|werkibot: 3.6864E6 / (16 * 182)
21:42:031265.93406593407
21:42:03Russ|werkibot: 3686400 / (16 * 182)
21:42:22789.930555555556
21:42:22Russ|werkibot: 1000000 / (3686400 / (16 * 182))
21:42:43tvramantangent --see you back in #intimate land?
21:43:16Russ|werkthe receiver and transmitter interrupt are both set
21:43:39BZFlagyou think there is a transmitter?!?
21:43:43Russ|werkHP-SIR is disabled
21:44:53Russ|werkmaybe its just generic routines
21:46:17BZFlaghttp://TuxScreen.net/img/rimg0003.jpg - I think the current is just a receiver, but it would be nice to be able to swap in a tranceiver. ;-)
21:46:26BZFlaghttp://TuxScreen.net/img/rimg0003.jpg - I think the current is just a receiver, but it would be nice to be able to swap in a tranceiver. ;-)
21:46:49Russ|werkhave you looked that up anywhere?
21:47:19Russ|werkbtw, don't let the hot glue flake off anything, thats bad
21:47:42jacques_workdoes the tuxscreen have a speaker other than in handset?
21:47:56Russ|werkit has a speaker for speakerphone
21:48:10Russ|werkas well as a mic for speakerphone
21:48:44BZFlaglooking... TSOP1738?
21:49:14jacques_workexcellent
21:49:33Russ|werkI get german stuff about IR
21:49:43TangentThe TK19 is just an IR receiver
21:50:09TangentIt's the same one that's in the Lego Mindstorms stuff.. You need extra LEDs to transmit
21:51:21BZFlaghttp://batronix.com/pdf/tsop17xx.pdf
21:51:26Russ|werkdatasheet is here http://www.sembs.rv.bw.schule.de/dokumente/mikrocontroller/tsop17xx.pdf
21:51:29Russ|werkor there
21:52:32Russ|werkthe "application circuit" looks familiar
21:53:32Russ|werkok, so the keyboard will be cracked tonight
21:53:39TangentWoohoo :)
21:55:21Russ|werksejin sent me an annyoing email btw
21:55:36Russ|werkMr. Dill:
21:55:37Russ|werkUnfortunately at this time I cannot release the required information to you.
21:55:37Russ|werkI need some more detailed information as to what your application is and the
21:55:37Russ|werkquantity of keyboards that your looking at purchasing from us.  Please get
21:55:37Russ|werkback to me when you get a chance.
21:56:35BZFlaghttp://TuxScreen.net/wiki/view//InfraredPort
21:57:08BZFlagnice, post it on the wiki page. ;-)
21:57:51Russ|werkI don't think it will matter as now the keyboard will just start sending scancodes
21:58:32BZFlagyeah, I just meant to post the email so others can know who is/isn't responsive on the net.
22:00:49Russ|werkCONFIDENTIALITY NOTICE:
22:00:49Russ|werkThis Email and any files transmitted with it are confidential and
22:00:49Russ|werkintended solely for the use of the individual or entity named to
22:00:49Russ|werkwhom
22:00:49Russ|werkthey are addressed.  If you are not the intended recipient, you
22:00:49Russ|werkare notified that disclosing, copying, distributing or taking any
22:00:51Russ|werkaction in reliance on the contents of this information is
22:00:53Russ|werkstrictly prohibited.
22:00:55Russ|werkIf you are not the intended recipient of this Email, please
22:00:57Russ|werknotify the Mail Administrator at postmaster@sejin.com
22:00:59Russ|werkimmediately.
22:01:07Russ|werkoops...oh well
22:01:23Russ|werkI feel so legally bound be that
22:01:44jacques_workdamn, who *are* those ppl?
22:01:52jacques_workare you emailing the NSA?
22:02:01Russ|werkoh wait, it doesn't say anything about the intended recepient
22:02:28Russ|werkthe intended recepient can put it in the WSJ
22:02:50Russ|werkjacques_work: they make IR keyboards....I think they are all webtv compatible
22:03:10jacques_workRuss|werk: interesting
22:03:39jacques_workRuss|werk: is the tuxscreen kb webtv compat?
22:03:53Russ|werkthere is a good chance it is
22:04:14Russ|werkwe can most likely make a driver for the mice that come on some webtv keyboards
22:04:28jacques_workheh, super cool
22:05:06Russ|werkas well as provide scancodes from many remotes
22:05:26jacques_work:-D
22:06:16Russ|werkthe remotes might be harder though...
22:27:46BZFlagcourse the speaker will suck for mp3's but a remote control mp3 terminal would be cool. ;-)
22:30:28jacques_worki just want it to talk
22:34:36TangentBe nice if it could listen too :)
22:43:15jacques_workhi Tangent!
22:43:48jacques_workI just don;t know if the cpu is powerful enough for speech recognition
22:44:33TangentIf a 468DX40 with 32MB can do it.. surely the SA1100 must be able to do something worthwhile
22:45:41TangentNot speech recognition, but pre-defined words
22:46:37TangentOne of my old mobile phones could do that
22:46:48jacques_workTangent: ah right, good point - even my nokia can do it
22:47:16TangentJust simple stuff like 'Lights On' would be good :)
22:48:20Russ|werkhas anyone ever made a combination of two or more mics, that by positional audio (delays, etc) can pull out a single source?
22:48:20jacques_work"room lights up 50%"
22:48:21BZFlagThe folks at Fonix have an engine running on the iPAQ, they say it has enough power.
22:48:44jacques_workcool

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