highaltitude.log.20080206

[00:52] flowolf (n=flowolf@host178-236-dynamic.2-79-r.retail.telecomitalia.it) left irc: "Leaving"
[06:11] edmoore (n=edmoore@pomegranate.chu.cam.ac.uk) joined #highaltitude.
[06:26] <natrium42> edmoore!!1one
[06:27] <edmoore> too early
[06:27] <natrium42> it's too late here
[06:27] <natrium42> 1:30 am
[06:27] <natrium42> gnite :)
[06:27] <edmoore> what time?
[06:27] <edmoore> lol
[06:27] <edmoore> ok, leep well
[06:27] <edmoore> sleep*
[06:28] <natrium42> hehe, thanks
[08:36] Simon-MPFH (n=simon@phantom.mpfh.co.uk) joined #highaltitude.
[09:40] flowolf (n=flowolf@host178-236-dynamic.2-79-r.retail.telecomitalia.it) joined #highaltitude.
[10:40] Laurenceb (n=Laurence@fwnat-pub-1.physics.ox.ac.uk) joined #highaltitude.
[10:55] Laurenceb (n=Laurence@fwnat-pub-1.physics.ox.ac.uk) left irc: Read error: 104 (Connection reset by peer)
[11:21] Laurenceb (n=laurence@fwnat-pub-1.physics.ox.ac.uk) joined #highaltitude.
[11:21] <Laurenceb> hello
[11:21] Action: Laurenceb has a working radio
[11:22] <edmoore> awesome
[11:22] <Laurenceb> not very good
[11:22] <Laurenceb> but proves I'm on the right lines
[11:23] <Laurenceb> I'm using fldigi
[11:23] <edmoore> what's that?
[11:23] <Laurenceb> think truetty is better
[11:23] <Laurenceb> linux rtty app
[11:23] <Laurenceb> its the only thing that can talk to my line in
[11:23] <Laurenceb> using mic results in frequency splitting that utterly screws the singal
[11:24] <Laurenceb> I'm sure we could get a really nice signal with Rocketboys setup, the FFT looks good
[11:25] Action: Laurenceb realises that typing in obscene comments to test the radio is a bad idea, hams could be listening
[11:27] <Laurenceb> grr minicom cant upload files from anywhere other than its install directory it would appear
[11:30] <Laurenceb> hmm looks like a clock problem
[11:30] <Laurenceb> either my radio isnt 300 baud, or fldigi doesnt think it is
[11:31] <Laurenceb> but still, it should syncronise itself of the start/stop bit transition
[11:33] <Laurenceb> doh you have to calibrate fldigi before using it, probably the cause of the problem
[11:44] <Simon-MPFH> Question about radios and serial - probably displaying a huge ammount of ignorance
[11:44] <Simon-MPFH> When I use serial to ssh ot telnet between Linux boxes it is duplex?
[11:44] <Simon-MPFH> What happens with radios - is it just one way traffic?
[11:45] <edmoore> we just use one way at the moment, yes
[11:45] <Simon-MPFH> Ah
[11:45] <edmoore> we're working on an uplink though
[11:45] <edmoore> but it's been fine so far
[11:45] <Simon-MPFH> I've bought a pair of TDi2s from radiometrix
[11:45] <Simon-MPFH> I will have to test them out
[11:45] <edmoore> cool
[11:46] <Simon-MPFH> Still waiting for the microprocessor from Tin Can
[11:59] <edmoore> the hammer?
[12:07] <Simon-MPFH> Yes
[12:07] <Simon-MPFH> Don't know much about it yet
[12:09] <Laurenceb> hmm theres some sort of timing problem
[12:10] <edmoore> it looks pretty funky
[12:10] <Laurenceb> if I send blah blah blah blah blah
[12:10] <Laurenceb> I get blah ba£^TRR% blah blah £$£$% and so on
[12:11] <Laurenceb> its odd, as if its not using the stop to start bit transitions to syncronise correctly
[12:11] <Laurenceb> maybe a fault in fldigi, I got it to output a discriminator trace and that seems to be ok
[12:13] <Laurenceb> I can get strings of 5 or 6 characters through, so it obviously working pretty much ok
[12:14] <Laurenceb> its really annoying that truetty isnt talking to my sound card properly
[12:15] flowolf (n=flowolf@host178-236-dynamic.2-79-r.retail.telecomitalia.it) left #highaltitude ("Leaving").
[13:11] flowolf (n=flowolf@host178-236-dynamic.2-79-r.retail.telecomitalia.it) joined #highaltitude.
[13:20] <Laurenceb> http://wiki.ukhas.org.uk/_media/projects:dscn0979.jpg
[13:21] <edmoore> that looks really good laurenceb
[13:21] <edmoore> nice job
[13:21] <Laurenceb> thanx
[13:22] <Laurenceb> I got the power saving to work
[13:22] <Laurenceb> uses 10ma most of the time
[13:22] <Laurenceb> then 30ma transmitting
[13:23] <Laurenceb> I think it can be reduced to about 7 to 27 by turning off the unneeded avr hardware
[13:23] <edmoore> from what to what?
[13:24] <Laurenceb> from 7 to 27ma sleep to transmitting
[13:25] <edmoore> wow that's not bad
[13:25] <edmoore> tho- what's the 7ma when it's sleeping?
[13:25] <Laurenceb> it has to keep a bit of hardware ticking over at 12MHz
[13:26] <edmoore> wassat?
[13:26] <Laurenceb> to timeout - timer1
[13:26] <Laurenceb> and that means a load of other logic is running at 12Mhz
[13:26] <Laurenceb> but... if I used a clk divider
[13:26] <edmoore> could you not send the whole thing to sleep?
[13:27] <Laurenceb> no, it has to wake up
[13:27] <Laurenceb> in case the main flight computer dies
[13:27] <edmoore> ext int from usart?
[13:27] <edmoore> oh ok
[13:27] <edmoore> hmm.....
[13:27] <Laurenceb> using the clk divider will get that way down, I should have thought of that one before
[13:28] <edmoore> yeah
[13:28] <edmoore> you could make an insanely low power module
[13:28] <edmoore> or rather, entire balloon
[13:28] <Laurenceb> use a mega168p
[13:28] <edmoore> have it sit on seal island off hunstanton for weeks :)
[13:28] <Laurenceb> hehe
[13:29] <Laurenceb> I was thinking of a mini zero pressure with two powerfilm cells
[13:29] <Laurenceb> and ethanol ballast
[13:29] <edmoore> i can't wait to launch ours
[13:29] <Laurenceb> whats your next launch?
[13:30] <edmoore> the zp
[13:30] <edmoore> and new flight comp
[13:30] <Laurenceb> ah sweet
[13:30] <Laurenceb> ballast?
[13:30] <edmoore> 350g payload
[13:30] <edmoore> nope, just want to do a sort of latex balloon flight profile, but with zp
[13:30] <Laurenceb> wow thats good, almost as light as mihab1
[13:30] <Laurenceb> hmm
[13:31] <Laurenceb> it might land in the north sea
[13:31] <edmoore> well it's what the balloon is rated for to get to 35km ish
[13:31] <Laurenceb> with no burst it will cruise until sunset
[13:31] <edmoore> the new payload ha slanding prediction
[13:31] <edmoore> has*
[13:31] <Laurenceb> well of course
[13:31] <edmoore> and the balloon itself will have cutdown
[13:32] <edmoore> or rather, he vent
[13:32] <Laurenceb> http://wiki.ukhas.org.uk/projects:aerosol_code
[13:32] <Laurenceb> beat that :P
[13:32] <Laurenceb> well its only 50% done
[13:32] <Laurenceb> but it will read a kml file for cutdown
[13:32] <Laurenceb> and do prediction
[13:33] <edmoore> that is nice - I'm hoping to put kml files on ours
[13:33] <edmoore> but doing it all in c from scratch is more tricky
[13:33] <Laurenceb> :(
[13:33] <edmoore> might have to pass on that one
[13:33] Action: Laurenceb is now a python fan
[13:33] <Laurenceb> I'm going to use reed solomon with MD5 checksum
[13:33] <edmoore> it basically is a pseudocode interpreter
[13:34] <Laurenceb> and live google earth view :P
[13:34] <Laurenceb> you've seen my reed solomon encoder?
[13:34] <edmoore> ha
[13:34] <edmoore> got to, really!
[13:35] <Laurenceb> http://wiki.ukhas.org.uk/code:python_reedsolomon
[13:36] <edmoore> impressive
[13:36] <Laurenceb> I'm going to add a parity bit for each byte
[13:37] <Laurenceb> then you have error probability data for the decoder
[13:37] <Laurenceb> which results in a doubling of performance
[13:38] <Laurenceb> god knows why that isnt used as standard... but in CD drives the error probability data comes from the tracking electronics
[13:42] <Laurenceb> the cool thing about reed solomon is that the start of each packet is human readable
[13:43] <Laurenceb> you'll need a list of irreducable polynomial vectors to make that work with different character sizes
[13:44] <Laurenceb> so I'll build that into the code
[13:51] <Laurenceb> bbl lunch
[14:35] <Laurenceb> back :D
[14:36] <Laurenceb> I might email the fldigi developer, I'm sure he could offer some advice
[15:15] jcoxon (n=jcoxon@host86-129-62-157.range86-129.btcentralplus.com) joined #highaltitude.
[15:17] <Laurenceb> hey there
[15:18] <jcoxon> hello Laurenceb
[15:20] <Laurenceb> I got the radio working
[15:20] <jcoxon> excellent
[15:20] <Laurenceb> sort of, theres bursts of corruption
[15:20] <jcoxon> i fried my data logger it seems :(
[15:20] <Laurenceb> I've emailed the fldigi developer
[15:20] <Laurenceb> jcoxon nooo :(
[15:20] <jcoxon> yeah, seems i managed to short out one of the pins by a mistake
[15:21] <jcoxon> and it died
[15:21] <Laurenceb> truetty wont talk to my sound card, and I think fldigi might need some sort of calibration
[15:21] <jcoxon> isn't the end of the world though
[15:21] <Laurenceb> :(
[15:21] <jcoxon> on linux?
[15:21] <Laurenceb> yes
[15:22] <jcoxon> what sound card drivers are you using?
[15:22] <jcoxon> alsa?
[15:22] <Laurenceb> OSS
[15:22] <jcoxon> hmmm
[15:22] <Laurenceb> yes
[15:22] <Laurenceb> it recommends portaudio
[15:22] <Laurenceb> but it wouldnt compile
[15:23] <jcoxon> oh
[15:25] <Laurenceb> theres a binary, but for amd32, not amd64
[15:30] <Laurenceb> jcoxon: always work with the power off, and quadrouple check everything before turning it on. If in doubt, wait until after lunch before turning it on
[15:30] <Laurenceb> thats my technique
[15:31] <jcoxon> oh i know
[15:31] <jcoxon> its more that i placed one of the resistors onto the wrong pin
[15:32] <jcoxon> well thats what i think did it
[15:33] <jcoxon> oh well, c'est la vie
[15:34] <jcoxon> its not the end of the world, can still log directly to the gm862
[15:36] <Laurenceb> get an ngw100
[15:39] <Laurenceb> ooh I have a reply from the fldigi developer :d
[15:40] <jcoxon> ngw100, pah
[15:40] <jcoxon> hehe
[15:40] <jcoxon> i'm quite happy
[15:49] <jcoxon> Laurenceb, so, where do you see the future of ballooning going?
[15:49] <jcoxon> I'm redoing my website and its made me think
[15:50] <Laurenceb> I want to do long duration
[15:50] <Laurenceb> zero pressure with ballast
[15:50] <Laurenceb> then ir
[15:51] <Laurenceb> miir*
[15:51] <Simon-MPFH> It's harmless fun - I would imagine that the government will make it illegal
[15:51] <jcoxon> :-)
[15:51] <Laurenceb> lol
[15:51] <jcoxon> Simon-MPFH, nah they won't make it illegal
[15:51] <jcoxon> its nothing new
[15:51] <Laurenceb> not enough people doing it
[15:51] <Simon-MPFH> Hmmmm, they will say that a terrorist could use it
[15:51] <jcoxon> the CAA still seem happy with us
[15:52] <Simon-MPFH> The Campaign Against Aviation
[15:52] <Laurenceb> not enough people doing it for them to care
[15:52] <Simon-MPFH> Cool - lets hope it stays that way
[15:52] <Laurenceb> hehe
[15:52] <jcoxon> indeed
[15:52] <Simon-MPFH> Question
[15:52] <Simon-MPFH> Why one big balloon and not lots of smaller ones?
[15:52] <jcoxon> (we just have to be sensible so they stay happy)
[15:53] <jcoxon> Simon-MPFH, quite a few answers
[15:53] <jcoxon> you can't be sure about them all bursting at the same time
[15:53] <Simon-MPFH> Aye - but then you have cutdown?
[15:53] <jcoxon> (though you could of course cutdown)
[15:53] <Simon-MPFH> snap
[15:53] <jcoxon> cutdowns aren't perfect
[15:54] <jcoxon> actually they usually don't work
[15:54] <Simon-MPFH> Ha ha - there's confidence!
[15:54] <Laurenceb> jcoxon: did you see my ballast design?
[15:54] <jcoxon> also there would be problems about balloons rubbing
[15:54] <jcoxon> there has been a lot of discussion about damage to balloons
[15:54] <jcoxon> for example US scientists don't really mind how you handle balloons
[15:54] <jcoxon> but CNES do
[15:55] <Simon-MPFH> I guess a big balloon is more effitient - weight/volume
[15:55] <jcoxon> that as well
[15:55] <Simon-MPFH> and big balloons can stretch more
[15:55] <jcoxon> however a multiballoon design has its uses
[15:55] <jcoxon> would be good for a platform say
[15:55] <Laurenceb> basically a ball bearing resting on an o ring, then a small coil to lift it to release ethanol, good for zero pressure ballast
[15:55] <jcoxon> for perhaps a rockoon launch
[15:55] <jcoxon> yup, i saw
[15:55] <jcoxon> i think ethanol is a good choice
[15:55] <jcoxon> not going to cause too much damage
[15:56] <Simon-MPFH> drown your sowwows with the remains when balloon lost?
[15:56] <Simon-MPFH> sorrows
[15:56] <jcoxon> and go blind due to it being industrial ethanol
[15:57] <jcoxon> :-D
[15:57] <Simon-MPFH> Hey Laurenceb - you work for TinCan? Where's my hammer? I'm soooooo excited
[15:57] <jcoxon> Simon-MPFH, it ain't Laurenceb
[15:58] <Simon-MPFH> Ah, soz
[15:58] <Simon-MPFH> Was never any good with names
[15:58] <jcoxon> you want pplague
[15:58] <Simon-MPFH> Still, it's better than a bumstix
[15:58] <jcoxon> ouch
[16:00] Laurenceb (n=laurence@fwnat-pub-1.physics.ox.ac.uk) left irc: Remote closed the connection
[16:01] <Simon-MPFH> The ngw100 looks good
[16:01] <jcoxon> indeed it does
[16:02] Laurenceb (n=laurence@fwnat-pub-1.physics.ox.ac.uk) joined #highaltitude.
[16:02] <Simon-MPFH> Bah, too many choices
[16:02] <jcoxon> there are lots of good modules, but i reckon its best to choose one and stick to it for a bit
[16:02] <jcoxon> constantly changing doesn't make things reliable
[16:02] <Simon-MPFH> I've got plenty done with 8bit Atmels!
[16:02] <Laurenceb> you can do anything with avrs
[16:03] <Laurenceb> Simon-MPFH: you have avr experience?
[16:03] <Simon-MPFH> Yes but quite basic
[16:03] <Simon-MPFH> Assembler hurts my head so I use C from Linux
[16:03] <Simon-MPFH> Driving servos to take photos from r/c gliders and suchlikje
[16:04] <Laurenceb> cool
[16:04] <Simon-MPFH> I'm looking for a 433 MHz Yagi with SMA connector - any sites recommended?
[16:05] <Simon-MPFH> http://www.lprs.co.uk/product_info.php?cPath=3&products_id=108
[16:05] <Simon-MPFH> This seems OK but no data
[16:06] <Laurenceb> http://wiki.ukhas.org.uk/code:python_reedsolomon
[16:06] <Laurenceb> updated it :P
[16:06] <Laurenceb> uses a library to decode the encoded data
[16:08] <Simon-MPFH> So LaurenceB, for the 'hard of radios' what does it do?
[16:10] <Laurenceb> it encodes your data into packets
[16:10] <Laurenceb> errors in a packet can then be corrected
[16:11] <Simon-MPFH> kinda like parity in serial coms?
[16:12] <Laurenceb> no
[16:13] <Laurenceb> it can correct errors, as well as detecting them
[16:13] <Simon-MPFH> Ah, like TCP
[16:15] <Laurenceb> not sure about that
[16:15] <Laurenceb> I dont know how TCP works
[16:15] <Laurenceb> but its FEC, it doesnt have to have the data resent, which I'm guessing TCP does
[16:15] <Simon-MPFH> AFAIK TCP gives you a guarenteed end to end connection - it will keep asking for packets that are lost or corrupt
[16:15] <Laurenceb> ok its not like TCP
[16:18] <Laurenceb> as everything you need to know is contained in one packet
[16:18] <Laurenceb> look up FEC
[16:20] <jcoxon> i'm quite nostalgic :-)
[16:20] <jcoxon> re reading my write up about pegasus 3
[16:22] <Laurenceb> jcoxon: how do I include python code?
[16:22] <jcoxon> ?
[16:22] <Laurenceb> ie whats equivalent to #include
[16:22] <jcoxon> import
[16:22] <Laurenceb> ok
[16:22] <Laurenceb> so import "reed_solomon.py" ?
[16:23] <jcoxon> ummmm
[16:23] <jcoxon> on my gm862 it compiles them as binaries on start up
[16:23] <jcoxon> and so i do
[16:23] <jcoxon> import reed_solomon
[16:24] <jcoxon> ignoring the file ending bit thingy
[16:24] <jcoxon> but you should check first!
[16:27] <Simon-MPFH> Oooooh, so this can run on a GM862 and use a GPIO line to send the GPS readout by Radio
[16:29] <jcoxon> what can?
[16:29] <jcoxon> i'm confused Simon-MPFH
[16:30] <Laurenceb> reed solomon?
[16:30] <Simon-MPFH> Is that not what the "reed_solomon.py could do?
[16:30] <Laurenceb> yes :P
[16:30] <Simon-MPFH> Smart
[16:30] <Laurenceb> in fact, with parity checking, it can correct 23 errors in a 128 byte packet
[16:31] <Simon-MPFH> You hoping to get 3-400Km I should think?
[16:31] <Laurenceb> yes
[16:31] <Laurenceb> with 256 byte packets you could get 1000km
[16:31] <jcoxon> but its not for a gm862
[16:32] <Simon-MPFH> What's your base reciever?
[16:32] <jcoxon> its more for a serial port based radio
[16:32] <Simon-MPFH> Ah
[16:32] <Laurenceb> yes sorry
[16:33] <Laurenceb> http://wiki.ukhas.org.uk/projects:aerosol
[16:33] <Laurenceb> there
[16:33] <jcoxon> sweet got cocoamodem working
[16:42] <Laurenceb> hmm
[16:42] <Laurenceb> say could I send you some wav files?
[16:44] <jcoxon> yup
[16:44] <jcoxon> you want me to test teh rtty?
[16:45] <Laurenceb> maybe, i'm trying fldigi with some samples
[16:49] Simon-MPFH (n=simon@phantom.mpfh.co.uk) left irc: "Leaving"
[16:51] <Laurenceb> aha
[16:51] <Laurenceb> with sample files I get similar behaviour from fldigi
[16:52] <Laurenceb> seems to be a bug in the way it recognises the stop/start bit transition
[16:52] <Laurenceb> and uses it to syncronise itself
[16:53] Simon-MPFH (n=simon@lancaster.mpfh.co.uk) joined #highaltitude.
[16:55] <Laurenceb> yes, its certainly buggy
[16:56] <Laurenceb> which is annoying...
[16:56] <Laurenceb> right time to pack up here
[16:56] <Laurenceb> cya
[17:00] Laurenceb (n=laurence@fwnat-pub-1.physics.ox.ac.uk) left irc: Remote closed the connection
[17:57] Hiena (n=Hiena@81.93.195.181.datatrans.hu) joined #highaltitude.
[17:58] <Hiena> Good eveningt!
[18:53] Hiena (n=Hiena@81.93.195.181.datatrans.hu) left irc: "-=Alkoholmentes-sör és repülõgép-szimulátor, biztos út a guminõ felé=-"
[19:20] <edmoore> hi jcoxon
[19:23] jatkins (n=jatkins@79-73-124-40.dynamic.dsl.as9105.com) joined #highaltitude.
[19:59] MetaMorfoziS (n=khmhm@3e44a429.adsl.enternet.hu) joined #highaltitude.
[20:03] <natrium42> ohai
[20:49] jatkins (n=jatkins@79-73-124-40.dynamic.dsl.as9105.com) left irc: " I love my HydraIRC -> http://www.hydrairc.com <-"
[21:30] Simon-MPFH (n=simon@lancaster.mpfh.co.uk) left irc: "Leaving"
[22:08] edmoore (n=edmoore@pomegranate.chu.cam.ac.uk) left irc: Read error: 113 (No route to host)
[22:13] mc- (n=mfcastle@cpc4-glfd1-0-0-cust538.glfd.cable.ntl.com) joined #highaltitude.
[22:28] mc- (n=mfcastle@cpc4-glfd1-0-0-cust538.glfd.cable.ntl.com) left irc:
[23:00] edmoore (n=edmoore@fn217.quns.cam.ac.uk) joined #highaltitude.
[23:10] <jcoxon> hey edmoore
[23:14] <edmoore> hi jcoxon
[23:14] <jcoxon> hows things?
[23:14] <edmoore> am with fnoble (again)
[23:14] <edmoore> fpga time
[23:15] <jcoxon> sounds fun
[23:15] <edmoore> we're looking at some waterproof enclosures for nova 6
[23:15] <jcoxon> interesting
[23:16] <jcoxon> i was going to make firefly float
[23:16] <jcoxon> but not be fully waterproof
[23:16] <edmoore> well probs same for us
[23:16] <edmoore> not all will be waterproof
[23:18] <jcoxon> really getting there with firefly
[23:29] <edmoore> what's left?
[23:30] <jcoxon> ummmm payload case
[23:30] <jcoxon> and the cutdown code
[23:30] <edmoore> cool
[23:30] <edmoore> our igniters now have avrs on them
[23:31] <edmoore> so they can all be daisychained up on an i2c loop
[23:31] <edmoore> they will have normaly digital 1/0 overide though
[23:32] <jcoxon> nice
[23:33] <jcoxon> the gm862 does't do floats
[23:33] <jcoxon> so its needs some work to change them into usable ints
[23:34] <jcoxon> (i know its a matter of just shifting the decimal point)
[23:35] edmoore (n=edmoore@fn217.quns.cam.ac.uk) left irc:
[00:00] --- Thu Feb 7 2008