highaltitude.log.20100225

[00:00] <Lunar_Lander> cool
[00:00] <Lunar_Lander> congratulations James
[00:01] <Lunar_Lander> http://www.bod.de/index.php?id=1132&objk_id=332866
[00:01] <jcoxon> oooo cool
[00:02] <Lunar_Lander> thanks :)
[00:02] <Lunar_Lander> thing is that they'll only sell it in Germany
[00:02] <Lunar_Lander> as far as I understood the contract
[00:03] <jcoxon> Lunar_Lander, http://www.flickr.com/photos/jcoxon77/4383328360/
[00:03] <Lunar_Lander> whoooo
[00:04] <Lunar_Lander> that is close!
[00:04] <Lunar_Lander> that is even closer than BEXUS 7
[00:05] <Lunar_Lander> amazing
[00:07] <jcoxon> :-)
[00:08] <Lunar_Lander> "I am proud to be a pioneer. Even if I cannot win, my participation is a gain. For me and my country"
[00:08] <Lunar_Lander> that is from Philip Boit
[00:08] <Lunar_Lander> Kenya's first olympic skiier
[00:09] <jcoxon> thats a nice quote
[00:09] <Lunar_Lander> yeah
[00:11] <Lunar_Lander> I mean
[00:11] <Lunar_Lander> when he came to finland to begin his training
[00:11] <Lunar_Lander> Feb 2nd, 1996
[00:12] <Lunar_Lander> that was the first day of his life he saw snow
[00:12] <jcoxon> :-)
[00:12] <jcoxon> Lunar_Lander, time for bed for me, catch you some other time.
[00:12] <jcoxon> nigh
[00:12] <jcoxon> t
[00:12] jcoxon (~jcoxon@host86-161-247-1.range86-161.btcentralplus.com) left irc: Quit: Leaving
[00:12] <Lunar_Lander> ok, good night
[00:12] <Lunar_Lander> night all
[00:12] Lunar_Lander (~lunar_lan@p54884F55.dip.t-dialin.net) left irc: Quit: Lunar_Lander
[00:14] ejcweb (~chatzilla@constantine.chu.cam.ac.uk) left irc: Quit: ChatZilla 0.9.86 [Firefox 3.6/20100115144158]
[00:15] juxta_ (~blah@ppp118-210-251-165.lns20.adl6.internode.on.net) joined #highaltitude.
[00:21] edmoore (~ed@2002:836f:e483:d:21b:63ff:feac:4133) joined #highaltitude.
[00:21] <juxta_> hi edmoore
[00:22] <edmoore> hi
[00:23] <edmoore> just off to bed infact. going to digest the data tomorrow
[00:26] LazyLeopard (~irc-clien@chocky.demon.co.uk) left irc: Quit: Bye
[00:29] edmoore (~ed@2002:836f:e483:d:21b:63ff:feac:4133) left irc: Quit: Bye
[00:30] allenk (none@unaffiliated/anna) left #highaltitude ("Leaving").
[01:22] juxta_ (~blah@ppp118-210-251-165.lns20.adl6.internode.on.net) left irc: Ping timeout: 265 seconds
[01:57] ms7821 (~Mark@flat.ms) left irc: Disconnected by services
[01:57] ms7821 (~Mark@flat.ms) joined #highaltitude.
[03:19] G8DSU_ (~chatzilla@cpc3-mort4-0-0-cust192.croy.cable.virginmedia.com) joined #highaltitude.
[03:23] G8DSU (~chatzilla@cpc3-mort4-0-0-cust192.croy.cable.virginmedia.com) left irc: Ping timeout: 265 seconds
[03:23] Nick change: G8DSU_ -> G8DSU
[03:34] natrium42 (~natrium@CPE000625d867e2-CM0014045885be.cpe.net.cable.rogers.com) joined #highaltitude.
[03:43] jasonb (~jasonb@dsl027-180-244.sfo1.dsl.speakeasy.net) left irc: Ping timeout: 265 seconds
[04:10] natrium42 (~natrium@CPE000625d867e2-CM0014045885be.cpe.net.cable.rogers.com) left irc: Quit: Leaving
[04:17] natrium42 (~alexei@CPE000625d867e2-CM0014045885be.cpe.net.cable.rogers.com) joined #highaltitude.
[04:23] jasonb (~jasonb@m450536d0.tmodns.net) joined #highaltitude.
[04:42] jasonb (~jasonb@m450536d0.tmodns.net) left irc: Ping timeout: 256 seconds
[05:02] blomlet (~quassel@cpc4-dals16-2-0-cust522.hari.cable.virginmedia.com) left irc: Ping timeout: 245 seconds
[05:11] blomlet (~quassel@cpc4-dals16-2-0-cust522.hari.cable.virginmedia.com) joined #highaltitude.
[05:29] natrium42 (~alexei@CPE000625d867e2-CM0014045885be.cpe.net.cable.rogers.com) left irc: Quit: My other car is a cdr.
[06:05] jasonb (~jasonb@adsl-66-124-73-250.dsl.sntc01.pacbell.net) joined #highaltitude.
[06:32] Spetznaz (~spetznaz@195.234.148.29) joined #highaltitude.
[07:44] G8DSU (~chatzilla@cpc3-mort4-0-0-cust192.croy.cable.virginmedia.com) left irc: Quit: ChatZilla 0.9.86 [Firefox 3.5.8/20100202165920]
[08:10] junderwood (~John@adsl.jcu.me.uk) joined #highaltitude.
[08:18] Spetznaz (~spetznaz@195.234.148.29) left irc: Read error: Connection reset by peer
[08:21] b3cft (~abrock@nat/yahoo/x-dgzlpcntghfrnuwu) joined #highaltitude.
[08:26] GW8RAK (~chatzilla@host86-134-240-133.range86-134.btcentralplus.com) joined #highaltitude.
[08:27] Spetznaz (~spetznaz@195.234.148.29) joined #highaltitude.
[09:03] icez (~icez@unaffiliated/icez) left irc: Remote host closed the connection
[09:36] Laurenceb (~laurence@host81-157-248-153.range81-157.btcentralplus.com) joined #highaltitude.
[10:04] Anotia50 (~djames_00@41.155.54.178) joined #highaltitude.
[10:08] jcoxon (~jcoxon@212.183.140.21) joined #highaltitude.
[10:08] <jcoxon> nothing to do
[10:08] Action: jcoxon is bored
[10:09] <rjharrison> hehe
[10:09] <rjharrison> I'm booldy busy]
[10:09] <rjharrison> I have day off
[10:10] <rjharrison> Doing a talk on icraus tonight
[10:10] <rjharrison> About 60 people apparently so I'm thinking 40
[10:12] Anotia50 (djames_00@41.155.54.178) left #highaltitude.
[10:12] <jcoxon> oooo ive fixed dl-fldigi for you
[10:12] <rjharrison> cool
[10:13] <rjharrison> what was the bugger
[10:13] <jcoxon> + are now properly sent
[10:13] <jcoxon> basically + is a reserved character
[10:14] <rjharrison> for curl
[10:14] <jcoxon> so you need to send %2b or something like that, so now search the string and swaps them out
[10:16] <rjharrison> Ok so I need to search the string and swap them put on the server too?
[10:16] <jcoxon> nope, dl-fldigi seems to do it all
[10:17] <rjharrison> Ok so + is reserved in fldigi
[10:17] <rjharrison> not curl
[10:18] <jcoxon> well you'll need to turn off your corrrction and see if I've got it right
[10:18] <rjharrison> will do in a sec
[10:18] <rjharrison> In hot wire hell atm
[10:18] <jcoxon> hmmm yeah its for html so perhaps you'll need to search for it as well
[10:24] edmoore (~836f0142@gateway/web/freenode/x-uyqvzjwohcrtemux) joined #highaltitude.
[10:25] <jcoxon> morning edmoore
[10:25] <edmoore> good mooning
[10:25] <rjharrison> morning ed
[10:26] <edmoore> hello
[10:27] <jcoxon> rjharrison, ive also branched so I can get on and fix the gps setup
[10:28] <rjharrison> cool
[10:28] <rjharrison> wiil chat in 5
[10:29] <jcoxon> no
[10:29] <jcoxon> np *
[10:29] <jcoxon> stupid autocomplete on my phone
[10:32] <rjharrison> btw natrium and I car going to conf on sun @ 7 re listener and tracker database would be good if you an ed wanted to come in
[10:33] <jcoxon> oh okay, if I'm free ill listen in
[10:33] <rjharrison> contribute ;)
[10:39] jcoxon (~jcoxon@212.183.140.21) left irc: Quit: Bye
[10:41] Spetznaz (~spetznaz@195.234.148.29) left irc: Read error: Connection reset by peer
[10:43] <edmoore> rjharrison: sorry was away
[10:43] <edmoore> if i'm back from lond by then I'd be v interested to listen in
[10:43] <edmoore> going to see man u vs villa
[10:44] <edmoore> which i understand is a football match.
[10:47] <rjharrison> hehe I'm the same
[11:19] jcoxon (~jcoxon@212.183.140.21) joined #highaltitude.
[11:20] <jcoxon> new keyboard for my phone
[11:20] <jcoxon> much more accurate
[11:23] <juxta> new keyboard?
[11:25] <gordonjcp> ah
[11:25] <gordonjcp> this is a channel full of clever people
[11:25] <gordonjcp> anyone got any good info on the maths behind SDR?
[11:30] <edmoore> what sort of bits specifically?
[11:30] <edmoore> I *should* know about this stuff if you look at the courses I'm taking, whether or not I actually do is another matter
[11:31] <juxta> heh
[11:31] <juxta> I know someone who just built a SDR rig, but that's about it :)
[11:32] <juxta> edmoore: has anybody done any work on coding schemes for balloon telem?
[11:32] <juxta> also, how did yesterday go? :)
[11:32] <edmoore> in reverse order - very well, thanks
[11:32] <edmoore> it behaved nicely
[11:32] <edmoore> uplink worked a treat
[11:33] <edmoore> coding schemes: a few emails went round about this recently
[11:33] <edmoore> the DominoEX standard has and FEC mode defined
[11:33] <juxta> I was looking into it yesterday, there's a C library for FEC floating around, libfec - dunno how it'd go on a micro though
[11:33] <edmoore> although it's hard to actually track down the standard
[11:33] <juxta> yeah, I looked at that too
[11:33] <juxta> what about things like AX.25?
[11:34] <edmoore> I think you just need to find an FEC scheme that will work specifically for the sort of thing we do
[11:34] <edmoore> I don't think ax.25 has any fec defined
[11:35] <edmoore> so the considerations of an FEC scheme for us are things like: we have tiny data block sizes - the best modern codes like LDPC assume tens of thousands of data bits to be encoded at a time
[11:35] <edmoore> and that our wireless channels are not very gaussian
[11:36] <edmoore> so lots of things that sim well on gaussian channels may not work so well for us.
[11:36] <edmoore> then as you say there's the fact that the encoder probably has to fit on a small 8-bit micro
[11:36] <juxta> yeah
[11:36] <juxta> the block size was the main one when I was looking around
[11:37] <juxta> but hamming codes seemed potentially OK I think
[11:37] <edmoore> so i reckon you probably want to use something like a satellite turbo-code from the 80s - turbo codes work well in real channels, 80's sattelite hardware is probably comparable to an avr now
[11:37] Spetznaz (~spetznaz@195.234.148.29) joined #highaltitude.
[11:37] <edmoore> i guess hamming is better than nothing :p
[11:37] <juxta> re ax.25 - there's a paper floating around called fx.25, some modification of ax.25 with fec or something
[11:38] <edmoore> but i mean an LDPC code is just a hamming code with a bit more rigorous thought applied to the parity check matrix
[11:38] <juxta> hmm, my unlce works in a sat program we have here doing telecoms, I'll try asking him for some input too
[11:39] <edmoore> that would be cool
[11:39] <edmoore> so a balloon packet currently is of the order of 500 data bits
[11:39] <rjharrison> jcoxon PM
[11:40] <edmoore> which is small but not too small - there exist efficient LDPC codes with block sizes on about 1000 bits (so ~500 data bits) and there are off the shelf designs for those - they might be worth looking at. you could seriously crank up the baud rate with one of those
[11:41] <juxta> I mentioned it to him briefly in the past, he mentioned maybe RS, simple LDPC or Golay - and also suggested using the multiple listeners to collaborate their RX data for better TX estimates
[11:41] <edmoore> that's definitely a good thing
[11:42] <edmoore> 3 listeners is something akin to having an R3 code for free, if you assume the channels to each listener are uncorrelated. So there's a huge amount of information there
[11:43] <edmoore> I think laurenceb wrote an RS implementation actually
[11:43] <edmoore> but I'm not sure it ever flew
[11:43] <juxta> ooh
[11:43] <edmoore> Laurenceb: ping
[11:43] <edmoore> we need you
[11:43] <juxta> hehe
[11:43] <edmoore> this was a few years ago
[11:43] <edmoore> i *think it might still be on t'wiki
[11:43] <juxta> i'll look again :)
[11:44] <edmoore> http://ukhas.org.uk/code:python_reedsolomon
[11:44] <edmoore> but he was flying a linux sbc so could use python
[11:46] <edmoore> bbiab
[11:46] <juxta> very nice
[11:46] <juxta> it's actually much less code than i imagined
[11:50] edmoore (~836f0142@gateway/web/freenode/x-uyqvzjwohcrtemux) left irc: Ping timeout: 252 seconds
[12:00] <Laurenceb> edmoore yes
[12:00] <Laurenceb> its on the wiki
[12:00] <Laurenceb> oh you found it nvm :P
[12:00] Action: Laurenceb learns to read scrollback
[12:00] <juxta> did you ever use it Laurenceb?
[12:01] edmoore (~836f0142@gateway/web/freenode/x-luvvscrrghdjmons) joined #highaltitude.
[12:01] <Laurenceb> not in flight
[12:01] <edmoore> hi Laurenceb
[12:01] <Laurenceb> its fairly easy to test from the terminal
[12:01] <Laurenceb> hi ed
[12:01] <edmoore> as I say though, you really want to test the code performance over a representative chanel
[12:01] <Laurenceb> juxta: the only problem is if you lose characters
[12:02] <Laurenceb> fldigi can do that if it misses stop start bit transitions
[12:02] <edmoore> infact on that point, if people want to get into the habbit of recording uncompressed audio clips during their decodes, that could be an excellent source of data
[12:02] <SpeedEvil> yeah.
[12:02] <Laurenceb> I was working on a RTTY decoder in c
[12:02] <Laurenceb> but other things got in the way
[12:02] <SpeedEvil> Get an idea of channel + radio errors
[12:02] <edmoore> yeah
[12:03] <edmoore> so for example my initial fsk decoder build of bayesian changepoint analysis worked *great* with simulated gaussian noise
[12:03] <edmoore> and less well on real audio recorded from BH3
[12:04] <edmoore> i think the noise is more laplacian. unfortunately the maths becomes fuglier though.
[12:05] <juxta> hmm, the math is a bit beyond me :)
[12:05] <edmoore> it still beats fldigi around the block though :p
[12:05] <jcoxon> I'm waiting for you new advanced modem for fldigi
[12:05] <edmoore> juxta: my lecturer has an incredible book - a bit of an eye-opener. If you find it in a lib, it's worth a flickthrough: Numerical Bayesian Methods Applied to Signal Processing by Fitzgerald
[12:06] <edmoore> jcoxon: right now it doesn't run properly in real time, and takes up an entire cpu core :) it'll be a while yet before it's just plug and play
[12:06] <SpeedEvil> Even something trivial like just ignoring start/stop errors would help a lot for fldigi
[12:06] <edmoore> and will def wait until after thesis
[12:06] <SpeedEvil> And comparing with the preceding line(s)
[12:06] <jcoxon> bbl
[12:06] jcoxon (~jcoxon@212.183.140.21) left irc: Quit: Bye
[12:07] <juxta> cheers edmoore, I will keep an eye out
[12:07] <edmoore> it's basically bayes rule applied to all signal processing problems. if you have enough computing power, you can do way better than current methods.
[12:15] zeusbot joined #highaltitude.
[12:15] -:#highaltitude- [freenode-info] why register and identify? your IRC nick is how people know you. http://freenode.net/faq.shtml#nicksetup
[12:15] <rjharrison> The camera case is designed to slide on from the fron and then put a back peice in
[12:15] <juxta> I had a shot with my knife, i'll try lubrication next time ;p
[12:16] <juxta> i made up a hot wire cutter somewhat like a bandsaw, so I could just slide the foam around on a flat working surface
[12:16] <SpeedEvil> make sure it's nice and sharp.
[12:17] <SpeedEvil> It has to be sharp enough that you get a nice clean cut that means when the knife applies pressure sideways to open the cut for the body of the knife, it will slide without tearing.
[12:17] <juxta> yeah, so the thinner the better
[12:18] <juxta> scissors are not brilliant from experience :D
[12:19] <juxta> edmoore: what kind of TX power did you guys use for uplink? I saw in the last test you did you managed around 190km with 35w from memory?
[12:20] <juxta> oh, also - the checksum you guys were using, is the algorithm for that floating around somewhere, or is it a widely used method?
[12:20] <SpeedEvil> Also. The mess is managable if you spray everything often with water.
[12:21] <juxta> to make it stick to things?
[12:22] <SpeedEvil> to make the beads and debris just behave normally
[12:22] <SpeedEvil> rather than electrostatic cling.
[12:22] <juxta> ah
[12:22] <juxta> good point
[12:23] <juxta> rjharrison: is that foam you used prone to mess?
[12:40] <gordonjcp> edmoore: my code grabs a bunch of I/Q samples and munges them into an array of complex numbers
[12:41] <gordonjcp> now, I get how you calculate either the sum or difference to get the appropriate sideband
[12:41] <gordonjcp> what I don't get is how you actually tune the SDR by (seemingly) multiplying by what I presume is a rotating vector in the complex plane
[12:43] <SpeedEvil> You can view any signal as a rotation.
[12:43] <SpeedEvil> So a 1000Hz signal in the input is a rotation (say) clockwise.
[12:44] <SpeedEvil> If you take your frame of reference to be static - it's a 1000Hz input.
[12:44] <SpeedEvil> If you take your frame of reference to be rotating at 100Hz - it's a 900Hz input
[12:44] <SpeedEvil> rotating clockwise at 100Hz
[12:52] <SpeedEvil> So if you simply rotate the input at 100Hz - it tunes it by 100Hz
[13:00] g8khw-iPhone (~RocketBoy@82.132.139.206) joined #highaltitude.
[13:12] <gordonjcp> SpeedEvil: right
[13:12] <gordonjcp> so by multiplying by a rotating vector, I'm mixing it with another signal
[13:13] <juxta> hi g8khw-iPhone
[13:13] <SpeedEvil> yes
[13:13] <gordonjcp> g8khw-iPhone: seen the iPhone echolink client?
[13:13] g8khw-iPhone (~RocketBoy@82.132.139.206) left irc: Quit: Bye chaps
[13:13] <gordonjcp> SpeedEvil: aaaah
[13:13] <SpeedEvil> Though because it's complex, you just get one resultant
[13:13] <SpeedEvil> AIUI
[13:14] <gordonjcp> aye, I only get the sum
[13:14] <gordonjcp> well no, I don't, when I demodulate I only calculate the sum (or difference)
[13:14] <gordonjcp> and the rest can go hang
[13:16] g8khw-iPhone (~RocketBoy@82.132.248.62) joined #highaltitude.
[13:17] <g8khw-iPhone> Gordoncp: Nope tar for the info thoughg
[13:18] <gordonjcp> g8khw-iPhone: I'm planning on getting Mrs Gordonjcp to install it on her iPhone
[13:35] <g8khw-iPhone> Cool
[13:35] <g8khw-iPhone> Bbl
[13:35] g8khw-iPhone (~RocketBoy@82.132.248.62) left irc: Quit: Bye chaps
[13:57] GW8RAK (~chatzilla@host86-134-240-133.range86-134.btcentralplus.com) left irc: Quit: ChatZilla 0.9.86 [Firefox 3.5.8/20100202165920]
[14:06] Laurenceb (~laurence@host81-157-248-153.range81-157.btcentralplus.com) left irc: Ping timeout: 245 seconds
[14:15] jcoxon (~jcoxon@host86-161-247-1.range86-161.btcentralplus.com) joined #highaltitude.
[14:22] <jcoxon> afternoon all
[14:23] <edmoore> hi jcoxon \
[14:23] <jcoxon> pd3em has made a video about his recovery
[14:23] <jcoxon> http://www.youtube.com/watch?v=B-o9fDeFYkE&hd=1
[14:23] <jcoxon> hehe its quite fun
[14:26] Simon-MPFH (~simon@phantom.mpfh.co.uk) joined #highaltitude.
[14:27] edmoore (~836f0142@gateway/web/freenode/x-luvvscrrghdjmons) left irc: Ping timeout: 252 seconds
[14:36] edmoore (~836f0142@gateway/web/freenode/x-gdbxnkdidsazydrn) joined #highaltitude.
[14:36] <edmoore> what's the vid?
[14:36] <jcoxon> its just a video about his recovery
[14:36] <jcoxon> showing some pictures and footage of the payload hanging over the water
[14:40] <jcoxon> edmoore, oh when pd3em opened up the payload the ballast tank was empty apart from a little which was just below the exit hole
[14:42] Jasperw (~jasperw@skeleton2.london.iofc.org) joined #highaltitude.
[14:45] <juxta> nice video jcoxon
[14:54] <edmoore> can you link it jcoxon ?
[14:54] <edmoore> missed it
[15:10] edmoore (~836f0142@gateway/web/freenode/x-gdbxnkdidsazydrn) left irc: Ping timeout: 252 seconds
[15:13] <N900evil> http://www.youtube.com/watch?v=B-o9fDeFYkE&hd=1
[15:13] <N900evil> link
[15:13] <jcoxon> he's gone
[15:13] <N900evil> oh. missed him
[15:13] <jcoxon> :-)
[15:25] Spetznaz (~spetznaz@195.234.148.29) left irc:
[15:32] <jcoxon> N900evil, did you ever make any progress wit that HF radio?
[15:33] <SpeedEvil> I keep finding important stuff to do. I need to try hooking it to a proper antenna. I suspect now that it was working - I just wasn't hearing anything.
[15:33] <SpeedEvil> I'll try to get that done this week.
[15:34] <jcoxon> hehe no hurry - not like we are launching an HF capable payload
[15:34] <SpeedEvil> True.
[15:34] <SpeedEvil> 70cm is a bit out of my range ATM.
[15:34] <jcoxon> :-p
[15:34] Action: SpeedEvil does not have a 60km antenna tower.
[15:35] <jcoxon> hehe
[15:35] <juxta> foolish HP, why did you put torx screws on this printer I wish to dismantle
[15:36] <SpeedEvil> Way too many projects, and too little energy. Post viral fatigue isn't fun. :/
[15:36] <SpeedEvil> juxta: angle grinder
[15:36] <jcoxon> BH4 is being sent back on saturday - so should have it next week
[15:36] <juxta> I'm going to get the motors out anyway, it's just going to make me more angry at HP
[15:36] <SpeedEvil> jcoxon: great.
[15:37] <juxta> SpeedEvil: oh its ok, I have some pretty advanced tools for dismantling printers
[15:37] <juxta> ie concrete floors or bricks
[15:37] <jcoxon> juxta, am about to start work on intergrating gps into dl-fldigi so your chase car automatically updates via dl-fldigi
[15:37] <juxta> oh nice jcoxon :)
[15:37] <jcoxon> how did you do your chase car software to gps
[15:37] <jcoxon> did you just read the serial port
[15:37] <jcoxon> ?
[15:38] <juxta> I wrote a little passthrough app
[15:38] <SpeedEvil> gpsd?
[15:38] <juxta> because we needed the GPS in our offline tracking software
[15:38] <juxta> windows unfortunatley
[15:38] <SpeedEvil> make sense to use something standard
[15:39] <SpeedEvil> oops - i thought it supported windows
[15:39] <SpeedEvil> seems not
[15:39] <juxta> so basically I just used com0com to setup a virtual serial port pair, then wrote a little passthrough app to read the GPS, upload to the tracker etc, and then pass it on to the mapping software
[15:39] <jcoxon> rjharrison was suggesting just to have a file to read
[15:39] <jcoxon> and have you pipe what ever gps into that file
[15:40] Action: SpeedEvil hugs named pipes.
[15:40] <juxta> yeah thats another way to do it
[15:40] <jcoxon> so that people can do what they want to get the gps to that point
[15:40] <juxta> windows doesn't really have a concept of named pipes does it SpeedEvil? ;p
[15:40] <juxta> oh wait.. maybe it does
[15:40] <jcoxon> or i could just find a cross platform serial lib and access the ports themselves
[15:40] <juxta> but they're called something else?
[15:40] <SpeedEvil> no
[15:40] <SpeedEvil> not AFAIK
[15:41] <juxta> jcoxon: I wrote my apps in java so I used libRXTX or something
[15:41] <juxta> I think reading the serial port is probably the better way
[15:41] <jcoxon> yeah, much cleaner
[15:41] <juxta> virtual ports are easy enough under windows, and I'm sure much easier under unix oses
[15:42] <juxta> (ie if you need the GPS data elsewhere)
[15:43] <jcoxon> okay i'll do the whole hog in dl-fldigi
[15:44] <juxta> doesnt fldigi have some GPS support already?
[15:45] <juxta> I seem to remember it in the menus somewhere along the line
[15:45] <jcoxon> nah just the graphic bits to set it up
[15:45] <jcoxon> that was me in preparation a long time ago
[15:45] <juxta> ahh :)
[15:45] <juxta> good thing I didn't try to get it to work then, hehe
[15:46] <juxta> by the way jcoxon, I was meaning to ask you - are yours mods to tinygps published somewhere? I'd like to use the code to number of sats from it if I could :)
[15:46] <jcoxon> oh they aren't but i can if you want
[15:46] <jcoxon> i'll stick them up now
[15:46] <juxta> whenever you have a chance
[15:46] <juxta> I should get to bed
[15:46] <juxta> it's 2:15am
[15:47] <juxta> I'm sort of tired after just riding home from work too
[15:47] <jcoxon> hehe
[15:47] <jcoxon> get some sleep
[15:47] <juxta> :)
[15:47] <juxta> one more question - does the rig cat stuff in dl-fldigi work?
[15:49] <jcoxon> should do
[15:49] <juxta> alrighty
[15:49] <jcoxon> http://www.pegasushabproject.org.uk/wiki/doku.php/missions:ballasthalo:ballasthalo4#pseudo_code
[15:49] <jcoxon> well i haven't messed with it
[15:49] <jcoxon> :-p
[15:49] edmoore (~ed@chu-gw-a.churchillcambridge.co.uk) joined #highaltitude.
[15:49] <juxta> thanks jcoxon :)
[15:49] <edmoore> hey again peeps
[15:50] <jcoxon> hey
[15:51] Action: juxta is off to bed
[15:51] <juxta> night :)
[15:52] <SpeedEvil> night
[15:56] juxta (fourtytwo@ppp118-210-194-250.lns20.adl6.internode.on.net) left irc: Ping timeout: 268 seconds
[15:56] edmoore (~ed@chu-gw-a.churchillcambridge.co.uk) left irc: Quit: edmoore
[16:05] edmoore (~ed@pluto.trinhall.cam.ac.uk) joined #highaltitude.
[16:33] jcoxon (~jcoxon@host86-161-247-1.range86-161.btcentralplus.com) left irc: Quit: Leaving
[16:38] edmoore_ (~ed@pluto.trinhall.cam.ac.uk) joined #highaltitude.
[16:38] edmoore (~ed@pluto.trinhall.cam.ac.uk) left irc: Read error: Connection reset by peer
[16:38] Nick change: edmoore_ -> edmoore
[16:39] jasonb (~jasonb@adsl-66-124-73-250.dsl.sntc01.pacbell.net) left irc: Ping timeout: 240 seconds
[16:41] Simon-MPFH (~simon@phantom.mpfh.co.uk) left irc: Quit: Leaving
[16:46] DanielRichman (~DanielRic@unaffiliated/danielrichman) joined #highaltitude.
[16:52] ejcweb (~chatzilla@constantine.chu.cam.ac.uk) joined #highaltitude.
[16:56] jasonb_ (~jasonb@m450536d0.tmodns.net) joined #highaltitude.
[17:02] <MikeMc68> evening
[17:11] edmoore_ (~ed@pluto.trinhall.cam.ac.uk) joined #highaltitude.
[17:11] edmoore (~ed@pluto.trinhall.cam.ac.uk) left irc: Read error: Connection reset by peer
[17:11] Nick change: edmoore_ -> edmoore
[17:23] gordonjcp (~gordonjcp@aramaki.gjcp.net) left irc: Remote host closed the connection
[17:25] blomlet_ (~quassel@cpc4-dals16-2-0-cust522.hari.cable.virginmedia.com) joined #highaltitude.
[17:29] blomlet (~quassel@cpc4-dals16-2-0-cust522.hari.cable.virginmedia.com) left irc: Ping timeout: 264 seconds
[17:34] jasonb_ (~jasonb@m450536d0.tmodns.net) left irc: Ping timeout: 248 seconds
[17:44] blomlet_ (~quassel@cpc4-dals16-2-0-cust522.hari.cable.virginmedia.com) left irc: Ping timeout: 265 seconds
[17:58] jasonb_ (~jasonb@dsl027-180-244.sfo1.dsl.speakeasy.net) joined #highaltitude.
[18:04] Hiena (~Hiena@81.93.195.181.datatrans.hu) joined #highaltitude.
[18:05] jasonb_ (~jasonb@dsl027-180-244.sfo1.dsl.speakeasy.net) left irc: Quit: Leaving
[18:05] jasonb_ (~jasonb@dsl027-180-244.sfo1.dsl.speakeasy.net) joined #highaltitude.
[18:06] jasonb_ (jasonb@dsl027-180-244.sfo1.dsl.speakeasy.net) left #highaltitude.
[18:08] ejcweb (~chatzilla@constantine.chu.cam.ac.uk) left irc: Ping timeout: 264 seconds
[18:21] edmoore (~ed@pluto.trinhall.cam.ac.uk) left irc: Quit: edmoore
[18:32] RocketBoy (~Steve@217.47.75.8) joined #highaltitude.
[19:50] icez (~icez@unaffiliated/icez) joined #highaltitude.
[20:06] Laurenceb (~laurence@host81-157-248-153.range81-157.btcentralplus.com) joined #highaltitude.
[20:10] <Laurenceb> sup
[20:10] <Laurenceb> brohemians
[20:19] <Laurenceb> SpeedEvil: Im looking at a SEPIC convertor for the autopilot board
[20:19] <SpeedEvil> yeah
[20:20] <Laurenceb> to 3.6v, then LDO to 3.3v - three seperate rails
[20:20] <Laurenceb> just simmed it in spice
[20:20] <SpeedEvil> many topologies enter, one leaves.
[20:20] <Laurenceb> using LT3467
[20:20] <SpeedEvil> Sorry - rewatched mad max.
[20:20] <Laurenceb> if I worked it out right I'm getting 89% efficiency
[20:20] <Laurenceb> heh
[20:20] <Laurenceb> at 8v in
[20:21] <SpeedEvil> how much is the BOM for the smps?
[20:21] <Laurenceb> the LT3467 is pretty awesome
[20:21] <Laurenceb> hmm maybe $3.5
[20:21] <SpeedEvil> not bad
[20:22] <Laurenceb> its SOT23
[20:22] <Laurenceb> so very nice and small
[20:23] <Laurenceb> theres other cool stuff you can do with the LT smps ICs
[20:24] <Laurenceb> camera flash and APD drivers
[20:25] <Laurenceb> but I still dont get the idea with SEPIC - why cant you just have a H bridge with the inductor in the middle?
[20:26] GW8RAK (~56172745@gateway/web/freenode/x-iaftvspvsdrfishs) joined #highaltitude.
[20:27] <SpeedEvil> http://rnzaf.proboards.com/index.cgi?board=civil&action=display&thread=6108
[20:27] <SpeedEvil> on a silly planes topic
[20:27] <SpeedEvil> not really looked at sepic much
[20:28] <Laurenceb> hah crazy
[20:29] <Laurenceb> the cool thing is it can run off one, two or three lipo cells - or of the 5v servo rail
[20:30] <SpeedEvil> neat!
[20:30] <SpeedEvil> flexible is good.
[20:30] GW8RAK (~56172745@gateway/web/freenode/x-iaftvspvsdrfishs) left irc: Ping timeout: 252 seconds
[20:32] <Laurenceb> LT have their own freeware spice
[20:33] <Laurenceb> seems to run fine under wine and has all their parts and application examples etc
[20:35] GW8RAK (~56172745@gateway/web/freenode/x-jrraqpyypotmapwb) joined #highaltitude.
[20:48] <Laurenceb> bbl
[20:55] jasonb (~jasonb@dsl027-180-244.sfo1.dsl.speakeasy.net) joined #highaltitude.
[21:02] jonsowman (~jonsowman@93-97-184-163.zone5.bethere.co.uk) left irc: Ping timeout: 240 seconds
[21:03] jonsowman (~jonsowman@93-97-184-163.zone5.bethere.co.uk) joined #highaltitude.
[21:03] Lunar_Lander (~lunar_lan@p5488722C.dip.t-dialin.net) joined #highaltitude.
[21:04] <Lunar_Lander> hello
[21:04] <Lunar_Lander> how was the Nova17 launch?
[21:08] GW8RAK (~56172745@gateway/web/freenode/x-jrraqpyypotmapwb) left irc: Quit: Page closed
[21:11] DanielRichman (~DanielRic@unaffiliated/danielrichman) left irc: Quit: Leaving
[21:18] <jonsowman> hi Lunar_Lander
[21:18] <jonsowman> nova17 went pretty well all things considered
[21:19] <Lunar_Lander> cool
[21:19] <jonsowman> badgercub's gps went a bit weird
[21:19] <Lunar_Lander> oh not good
[21:24] F4EIR (~5a037936@gateway/web/freenode/x-wupcczsfjynvwcju) joined #highaltitude.
[21:25] rjmunro (~chatzilla@87.127.91.51) joined #highaltitude.
[21:30] <rjmunro> I've improved my triangulator - it now uses the right formulas, and draws an translucent area rather than just a line. http://maps.google.co.uk/?q=http:%2F%2Farjam.net%2Ftriangulate.kml&ie=UTF8&ll=51.549751,2.581787&spn=3.914705,9.678955&z=7
[21:30] <rjmunro> So if we loose GPS again, it's ready :-)
[21:31] <jonsowman> rjmunro: good work :)
[21:32] <rjmunro> I've made them 500km long - is that a reasonable range?
[21:32] <sbasuita> rjmunro, very nice
[21:32] <jonsowman> sounds about right
[21:34] <rjmunro> We'll have to try it next time there is a flight and compare it's output to the GPS.
[21:52] rjmunro (~chatzilla@87.127.91.51) left irc: Remote host closed the connection
[22:15] Laurenceb (~laurence@host81-157-248-153.range81-157.btcentralplus.com) left irc: Ping timeout: 245 seconds
[22:15] Laurenceb (~laurence@host81-157-248-153.range81-157.btcentralplus.com) joined #highaltitude.
[22:25] tittux (~root@205.73-64-87.adsl-dyn.isp.belgacom.be) left irc: Read error: Connection reset by peer
[22:27] tittux (~root@148.235-65-87.adsl-dyn.isp.belgacom.be) joined #highaltitude.
[22:38] <Lunar_Lander> hi tittux
[22:39] N900evil (~Speedevil@tor/regular/SpeedEvil) left irc: Read error: Connection reset by peer
[22:46] sbasuita (~sbasuita@unaffiliated/drebellion) left irc: Quit: Leaving
[22:46] Hiena (~Hiena@81.93.195.181.datatrans.hu) left irc: Quit: -=Got bored from the net. Gone blowing up things.=-
[22:48] F4EIR (~5a037936@gateway/web/freenode/x-wupcczsfjynvwcju) left irc: Ping timeout: 252 seconds
[22:57] junderwood (~John@adsl.jcu.me.uk) left irc: Quit: Leaving
[23:00] fsphil (~phil@2001:470:1f09:483:21f:c6ff:fe44:b25b) joined #highaltitude.
[23:08] jcoxon (~jcoxon@host81-129-141-46.range81-129.btcentralplus.com) joined #highaltitude.
[23:11] SpeedEvil1 (~user@tor/regular/SpeedEvil) joined #highaltitude.
[23:12] SpeedEvil (1000@tor/regular/SpeedEvil) left irc: Ping timeout: 245 seconds
[23:12] Nick change: SpeedEvil1 -> SpeedEvil
[23:13] Laurenceb (~laurence@host81-157-248-153.range81-157.btcentralplus.com) left irc: Ping timeout: 245 seconds
[23:15] Lunar_Lander (~lunar_lan@p5488722C.dip.t-dialin.net) left irc: Quit: Lunar_Lander
[23:16] <fsphil> hmm. is it normal for the ntx2 frequency to jump about a bit now and then?
[23:17] <SpeedEvil> with temp you mean?
[23:17] <fsphil> room temp
[23:17] <SpeedEvil> yby how myuch?
[23:17] blomlet (~quassel@cpc4-dals16-2-0-cust522.hari.cable.virginmedia.com) joined #highaltitude.
[23:18] <fsphil> um. about 30 hz I think
[23:19] <fsphil> it drifts a bit too, but I expected that
[23:20] <Randomskk> 30hz?
[23:20] <SpeedEvil> 30hz isn't really that unexpected
[23:20] <SpeedEvil> are you sure your reciever is good to that?
[23:21] <fsphil> fair question.. though I don't see shifts like that on other frequencies
[23:21] <SpeedEvil> Also, you're sure it isn't moving at 30MPH randomly around the room?
[23:21] <jcoxon> fsphil, its a little unusual - usiely quite stable
[23:22] <jcoxon> usely*
[23:22] <fsphil> seems to have stopped now, just slowly wandering up and down a bit
[23:22] <jcoxon> how are you providing the data line?
[23:22] <fsphil> tied to gnd
[23:22] <SpeedEvil> how are you supplying power?
[23:22] <jcoxon> oh so no input?
[23:23] <fsphil> no input, the most basic circuit just to test it - only got it today
[23:23] <fsphil> power is via 12v psu -> 3.3v regulator
[23:23] <Randomskk> 12V to 3.3V? ouch
[23:24] <jcoxon> fsphil, i wounder if you were to input say 1v in that it would be a lot more stable
[23:24] <Randomskk> enjoy your 8.7I watt losses
[23:25] <Randomskk> I wouldn't have thought 30hz was much?
[23:25] <SpeedEvil> The thing is what - 10KHz full output?
[23:25] <SpeedEvil> 30Hz drift is well within spec
[23:27] <fsphil> seems very sudden though
[23:28] <fsphil> here's a smaller one: http://www.sanslogic.co.uk/files/ntx2-shift.jpeg
[23:28] <SpeedEvil> various mechanisms in oscillators do that
[23:28] RocketBoy (~Steve@217.47.75.8) left irc: Read error: Connection reset by peer
[23:28] <SpeedEvil> sudden random jumps that is
[23:29] <fsphil> normal enough then
[23:29] <fsphil> would fldigi be able to track a shift like that?
[23:29] <SpeedEvil> Are you sure your powersupply is stable to a couple of millivolts?
[23:29] <SpeedEvil> yes
[23:30] <fsphil> the psu was one I got with a usb drive -- nothing special
[23:31] <fsphil> might try again with a mattery
[23:31] <fsphil> er, battery
[23:31] <SpeedEvil> most batterys are.
[23:31] <jcoxon> fsphil, i'd see how it is with actually input
[23:31] <SpeedEvil> unless they are antimattery.
[23:31] <fsphil> lol
[23:31] <fsphil> I've a few bits to get before I can put data to it
[23:33] <fsphil> still a nice module -- far far better than the last one I tried
[23:33] <jcoxon> yeah its a good one, and not too expensive
[23:35] edmoore (~ed@2002:836f:e483:d:21b:63ff:feac:4133) joined #highaltitude.
[23:37] <jcoxon> evening edmoore
[23:37] <edmoore> hey
[23:38] <jcoxon> WB8ELK launch on saturday at 19:00UTC - will hopefully be using dl-fldigi
[23:39] <Randomskk> launch location/path?
[23:41] <jcoxon> not sure
[23:41] <jcoxon> will be on arhab i suspect
[23:44] <edmoore> just went to an excellent talk on superresolution direction finding
[23:44] <edmoore> of radio signals, that is
[23:45] <edmoore> now that would have been able to really pinpoint BH4
[23:45] <Randomskk> involving dsp?
[23:45] <edmoore> yes
[23:45] <edmoore> lots
[23:45] <edmoore> and lots and lots of linear algebra
[23:45] <Randomskk> saw the announcement on cuws
[23:45] <Randomskk> heh
[23:45] <russss> heh
[23:45] <edmoore> solutions are the intersection of subspaces of the signal covariance matrix
[23:45] <edmoore> joyous!
[23:46] <jcoxon> :-p
[23:46] <fsphil> did it involve reversing any polarities?
[23:46] Action: russss 's brain explodes
[23:46] <edmoore> quite a bit actualld fsphil
[23:46] <fsphil> cause it's not proper science if you don't at least reverse a polarity :)
[23:46] <edmoore> lots of stuff on polarisation detection of incoming signals
[23:46] <edmoore> they can do HF ranging just with one station
[23:47] <edmoore> as they can get the azimuth and elevation of the signal coming in, figure out based on parametric models of the ionosphere where it has bounched, and therefore where it has come from
[23:47] <edmoore> neat stuff
[23:47] <fsphil> oh that's a nice trick
[23:48] <fsphil> they can tell the angle of the signal to that accuracy?
[23:48] <Jasperw> presumably that only works with a knows transmitter power?
[23:48] <Jasperw> *known
[23:48] <edmoore> no
[23:48] <edmoore> it's geometry
[23:48] <edmoore> so you don't need to know tx power
[23:48] <SpeedEvil> edmoore: neat
[23:48] <edmoore> this is assuming a bouncing signal
[23:49] <Jasperw> right, but how can you tell how many bounces it's bounced?
[23:49] <SpeedEvil> So you get a huge eliptical footprint if it's near the horizon?
[23:49] <edmoore> but yeah 8 monopole antennas and you can get <1 degree accuracy
[23:49] <Jasperw> i.e. are you seeing a weak singal that bounced once, or a strong signal thats bounced several times and is now weak?
[23:50] <edmoore> and decode signals on the same channel simultaneously if they're coming from different directions
[23:50] <jcoxon> hehe, we should have all gone to this talk
[23:50] <edmoore> so we could fly many high alt balloons on 434.650 and decode them all, provided they're not all in the same place
[23:51] Action: SpeedEvil ponders maximal seperation of recievers.
[23:51] <fsphil> hmm.. actual cloud computing
[23:51] <edmoore> you can also actively block interference from other sources by putting a null in the direction it is coming from
[23:51] <SpeedEvil> I suppose you would actually need coherent reception
[23:51] <jcoxon> edmoore, so its very active receiving
[23:51] <SpeedEvil> that is teh actual IFs of the recievers in sync
[23:51] <edmoore> jcoxon: yes - lots and lots of digital signal processing
[23:51] <jcoxon> adapting your receiving environment to sort of filter out rogue signals
[23:52] <edmoore> and neat math
[23:52] <edmoore> s
[23:52] <edmoore> SpeedEvil: yes, you need a common clock for all the recivers
[23:52] <jcoxon> very cool
[23:52] <russss> phased array receiving basically?
[23:52] <russss> computationally fun
[23:52] <SpeedEvil> It would be interesting to get as many recievers recieving the signal at once, and see what could be done with recordings.
[23:53] <russss> and by fun I mean not terribly fun :P
[23:53] <SpeedEvil> russss: yup
[23:53] <edmoore> it's a bit cleverer than phased array
[23:53] <SpeedEvil> Doesn't it reduce to phased array - with arbitrary phases on the array and potentially multiple outputs?
[23:54] <edmoore> only if you're being very very general
[23:54] <russss> yeah I guess phased array is just a specific variant of an array antenna
[23:55] <SpeedEvil> In that the output is the sum of products of a per-antenna phase and amplitude function?
[23:56] <edmoore> yes. in that each antenna is Asin(wt+n) + Bcos(wt+n) and you can fiddle the knobs A, B and N for each antenna
[23:56] <edmoore> but the source seperation and tracking is really yummy
[23:57] <edmoore> and it also doesn't use mutual coupling between elements
[23:57] <edmoore> it deliberately uses non resonant antennas
[23:57] <russss> I'm going to go to bed before my brain explodes.
[23:57] <russss> night.
[23:57] <russss> :P
[23:57] <fsphil> night russss
[23:57] edmoore (~ed@2002:836f:e483:d:21b:63ff:feac:4133) left irc: Read error: Connection reset by peer
[23:57] edmoore_ (~ed@pluto.trinhall.cam.ac.uk) joined #highaltitude.
[23:58] <SpeedEvil> Umm. Resonance means you lose reception fidelity of the signal in some ways?
[23:58] <edmoore_> It's definitely the most complicated thing to do with radios I've seen. but damn cool
[23:58] <edmoore_> SpeedEvil: yes
[23:58] <fsphil> it's the most complicated thing I've heard all week
[23:58] <fsphil> and I thought I was doing well
[23:58] <edmoore_> but in most situations you are just interested in SNR
[23:58] <edmoore_> rather than just S
[23:58] <SpeedEvil> Sounds like some of the sonar stuff I've read about
[23:59] <SpeedEvil> and bistatic radar.
[23:59] <edmoore_> and if you can very specifically beam shape to both maximise the gain of the signal you want and knock any sources of noise off by 40dB, then you can get back to pretty good snr
[23:59] <edmoore_> as he was saying in almost al practical cases, noise is not spatially white
[00:00] --- Fri Feb 26 2010