highaltitude.log.20080811

[00:11] <fergusnoble> http://www.archive.org/details/1933-08-07_Stratosphere_Balloon_Falls
[00:12] <fergusnoble> the internet archive has quite a lot of interesting old footage
[00:40] icez (n=icez@ip68-98-34-247.ph.ph.cox.net) left irc: Remote closed the connection
[00:40] icez (n=icez@ip68-98-34-247.ph.ph.cox.net) joined #highaltitude.
[00:41] <natrium42> fergusnoble, where is the piano soundtrack?
[00:43] <fergusnoble> the piano soundtrack?
[00:43] <natrium42> all old movies seem to have this piano soundtrack
[00:43] <natrium42> turkish march by mozart, for example :P
[00:49] MetaMorfoziS (n=eee@53d83c5c.adsl.enternet.hu) left irc: Read error: 110 (Connection timed out)
[00:50] edmoore (n=edmoore@37.124-84-212.ippool.ndo.com) left irc: Read error: 113 (No route to host)
[00:55] phatmonkey (n=ben@scooby.firshman.co.uk) left irc:
[02:04] soneil_ (n=soneil@89.100.243.212) joined #highaltitude.
[02:04] soneil (n=soneil@pdpc/supporter/active/soneil) left irc: Read error: 104 (Connection reset by peer)
[02:31] fergusnoble (n=fergusno@88-106-248-135.dynamic.dsl.as9105.com) left irc:
[02:46] fnoble (n=fnoblef@88-106-248-135.dynamic.dsl.as9105.com) left irc: "Leaving"
[04:34] akawaka (n=akawaka@external.treyarch.com) left irc: Read error: 110 (Connection timed out)
[04:55] akawaka (n=akawaka@cpe-76-173-152-142.socal.res.rr.com) joined #highaltitude.
[05:37] akawaka (n=akawaka@cpe-76-173-152-142.socal.res.rr.com) left irc: Read error: 110 (Connection timed out)
[07:07] sloof3 (n=andy@pdpc/supporter/student/sloof3) left irc: Read error: 60 (Operation timed out)
[07:37] akawaka (n=akawaka@cpe-76-173-152-142.socal.res.rr.com) joined #highaltitude.
[07:42] Simon-MPFH (n=simon@phantom.mpfh.co.uk) joined #highaltitude.
[07:44] Ebola (n=Ebola@unaffiliated/ebola) joined #highaltitude.
[08:28] akawaka (n=akawaka@cpe-76-173-152-142.socal.res.rr.com) left irc: Read error: 110 (Connection timed out)
[08:45] Shanuson1 (n=Peter@p54A97C14.dip.t-dialin.net) joined #highaltitude.
[08:53] edmoore (n=edmoore@smtp.vorticity-systems.com) joined #highaltitude.
[09:04] Shanuson (n=Peter@p54A97644.dip.t-dialin.net) left irc: Read error: 110 (Connection timed out)
[09:05] jcoxon (n=jcoxon@85-210-73-114.dsl.pipex.com) joined #highaltitude.
[09:06] Nick change: Shanuson1 -> Shanuson
[09:08] <edmoore> hi jcoxon
[09:11] <jcoxon> morning edmoore
[09:15] <edmoore> all wel?
[09:17] <jcoxon> indeed
[09:17] <jcoxon> just watching hte olympics
[09:17] <jcoxon> and waiting for my results to come out
[09:19] <icez> gah
[09:19] <icez> I so wanted to see the opening ceremony, but all the video sites make it a violation of copyrights
[09:20] <jcoxon> the joy of the bbc and iplayer
[09:20] <icez> lol
[09:21] <edmoore> when do the results come out?
[09:22] <jcoxon> apparently this morning
[09:22] <jcoxon> they had them friday evening but they are mean
[09:22] <edmoore> :)
[09:25] <jcoxon> and yourself?
[09:30] <edmoore> pfff work
[09:31] <edmoore> ups failed over the weekend, lost a bunch of results. will just restart it. am doing other stuff today anyway
[09:43] <jcoxon> oh dear
[09:44] <edmoore> ironic that the power stays on but the ups goes down
[09:44] <jcoxon> i've been persuaded to swap out the webcam and get a canon which can do remote capture
[09:44] <jcoxon> http://capture.sourceforge.net/
[09:44] <edmoore> probably a good call - better glass and that
[09:44] <edmoore> so are you downsampling it?
[09:45] <jcoxon> i'll have to change the resolution down to 320x256 to transmit
[09:45] <edmoore> how will you be doing that- is there a built-in tool?
[09:45] <jcoxon> oh i've got some programs to do that
[09:46] <edmoore> cool, so they should do it right, then. Just when I was wee and didn't know anything about information theory, I downsampled stuff by just taking every 2nd pixel, say, without low pass filtering it first.
[09:46] <edmoore> got ugly results
[09:47] <jcoxon> hehe
[09:48] robert1971 (n=user@212.183.134.211) joined #highaltitude.
[09:48] <robert1971> Morning all
[09:48] <robert1971> Sitting in a cafe in Holland watching it rain
[09:48] <edmoore> why would you do that?
[09:49] <edmoore> and do they speak english to you automatically as if you have a sign above your head announcing your englishness?
[09:49] <robert1971> Meant to be on a cycling holliday around Holland
[09:49] <robert1971> Yes
[09:49] <edmoore> that always spooks me in holland
[09:50] <robert1971> Not that I'm a cyclist in any way. More of an annual event
[09:51] <edmoore> ok, sounds fun. Presumably you've not put your embedded development suitin your panier.
[09:51] <robert1971> I have brought programmer and dev board and a LM55 tempsensor to get working with the adc port
[09:52] <edmoore> That's dedication.
[09:52] <robert1971> I plan to catch up with the diary on my website and do a wiki on programmers for newbies
[09:53] <edmoore> I watched the video - excellent stuff.
[09:54] <robert1971> Put a new bit of video up on www.robertharrison.org/icarus in the diary section. This shows the camera under servo controll shooting pictures and video in various directions.
[09:54] <robert1971> Beat me to it
[09:55] <edmoore> jcoxon - just found an interesting comparison on why you should low pass filter first before downsmapling - http://www-sigproc.eng.cam.ac.uk/~ngk/software/ColourNotes.htm
[09:55] <edmoore> compare the side by sides at the end of the section on 'Clock image with subsampling'
[09:55] <edmoore> Also, I want a sunny day
[09:55] <robert1971> Cheers coming along nicely. I hope to do a teathered flight in the next couple of weeks to test camera setting and general aerodynamics
[09:56] <edmoore> oh wow
[09:56] <edmoore> so not far off a flight then
[09:57] <robert1971> Well I have the tracking and the sensors to sort out
[09:57] <edmoore> that formatted rather wierdly
[09:57] <robert1971> (UMM
[09:57] <robert1971> &UNNY
[09:57] <edmoore> do you know what means of tracking you're going for?
[09:58] <robert1971> )
[09:59] <robert1971> WITH
[09:59] <edmoore> So something like a telit gm863 module?
[09:59] <jcoxon> edmoore, i've got this installed http://netpbm.sourceforge.net/
[10:00] <robert1971> 9EP
[10:00] <jcoxon> wow something wrong with your font robert1971
[10:00] <robert1971> 9EP
[10:00] <robert1971> AND
[10:01] <edmoore> robert1971: not from fergusnoble or I - but there may well be one in the works
[10:01] <edmoore> jcoxon: http://www-sigproc.eng.cam.ac.uk/~ngk/software/ColourNotes_files/image020.jpg
[10:01] <edmoore> I love this - the patch and A and B in are exactly the same intensity
[10:01] <jcoxon> weird
[10:02] <robert1971> )
[10:02] <jcoxon> thats certainly one of hte plans
[10:02] <jcoxon> i think the modules exist - whether there are any dev boards for them is another matter
[10:02] <jcoxon> bbiab
[10:02] <edmoore> putting a pc-grade operating system on a gsm module. whatever next.
[10:03] <robert1971> 3OUNDS
[10:03] <robert1971> 4HIS
[10:03] <robert1971> )
[10:03] <edmoore> well I know some of the telit modules have built in arm9 processors and advertise themselves as linux-ready
[10:03] <icez> looks like you're using a chinese/japanese/korean-input thingy
[10:03] <icez> :P
[10:04] MetaMorfoziS (n=eee@53d83582.adsl.enternet.hu) joined #highaltitude.
[10:06] <robert1971> )M
[10:06] <robert1971> .OTHING
[10:06] <edmoore> robert1971: good luck!
[10:06] <edmoore> where do you get to eventually?
[10:06] <MetaMorfoziS> and evolutionally...
[10:06] <robert1971> !MSTERDAM
[10:07] <robert1971> HAVE
[10:07] MetaMorfoziS (n=eee@53d83582.adsl.enternet.hu) left #highaltitude ("Connection reset by a tapir").
[10:07] <robert1971> TTFN
[10:07] robert1971 (n=user@212.183.134.211) left #highaltitude.
[10:09] icez (n=icez@ip68-98-34-247.ph.ph.cox.net) left irc: "Lost terminal"
[10:12] jcoxon (n=jcoxon@85-210-73-114.dsl.pipex.com) left irc: "Leaving"
[10:18] jnd (n=jenda@unaffiliated/jnd) left irc: Read error: 110 (Connection timed out)
[10:34] Simon-MPFH (n=simon@phantom.mpfh.co.uk) left irc: "Leaving"
[10:38] <gordonjcp> looks like robert1971 has a japanese font
[10:38] <gordonjcp> possibly chinese
[11:49] robert1971 (n=user@212.183.136.195) joined #highaltitude.
[11:49] <robert1971> At the lunch stop ... looks like the font is fixed
[11:50] <gordonjcp> robert1971: yeah, looked like you were using a Japanese or Chinese multibyte font
[11:52] <robert1971> It the trouble was that I had no idea how I managed to turn it on...
[12:30] robert1971 (n=user@212.183.136.195) left irc: Read error: 110 (Connection timed out)
[13:39] Nick change: EI5GTB__ -> EI5GTB
[13:39] jcoxon (n=jcoxon@85-210-73-114.dsl.pipex.com) joined #highaltitude.
[14:04] <jcoxon> they continue to be mean - i won't know till 5pm
[14:08] fergusnoble (n=fergusno@88-106-248-135.dynamic.dsl.as9105.com) joined #highaltitude.
[14:11] <gordonjcp> jcoxon: who does?
[14:11] <jcoxon> oh i'm waiting for my exam results
[14:11] <gordonjcp> ah
[14:46] jiffe98 (n=jiffe@67.158.3.217) joined #highaltitude.
[15:54] jiffe99 (n=jiffe@67.158.3.217) joined #highaltitude.
[15:55] fergusnoble (n=fergusno@88-106-248-135.dynamic.dsl.as9105.com) left #highaltitude.
[15:58] jiffe98 (n=jiffe@67.158.3.217) left irc: Read error: 101 (Network is unreachable)
[16:02] natrium42 (n=alexei@CPE000625d867e2-CM0014045885be.cpe.net.cable.rogers.com) left irc:
[16:14] <jcoxon> dum dee dum dum
[16:17] <SpeedEvil> Don't tell me, don't tell me, I know it!!! ... dum?
[16:17] <jcoxon> yeah!
[16:19] <edmoore> SpeedEvil: this avrFreak2 guy is most unhelpful and getting a bit arsey
[16:19] <jcoxon> oooo irc fight
[16:19] <SpeedEvil> Arsey is common with him.
[16:20] <SpeedEvil> I occasionally get to the putting him on ignore stage.
[16:20] robert1971 (n=user@212.183.134.67) joined #highaltitude.
[16:21] robert1971 (n=user@212.183.134.67) left #highaltitude.
[16:21] <edmoore> he told me a whole bunch of incorrect stuff, I corrected him (sort of politely) and now he makes sarcy comments at everything i saw. Though I am quite enjoying it.
[16:24] <SpeedEvil> Yeah. You can do a kalman filter just fine on an 8 bit micro.
[16:24] <SpeedEvil> If the datarate is low enough.
[16:24] <edmoore> I have done.
[16:25] <edmoore> but.. my data rate is a little higher these days :)
[16:25] <SpeedEvil> edmoore: is this for the rocket?
[16:25] <SpeedEvil> edmoore: or are you meaning for work.
[16:27] <edmoore> no not work, yes rocket.
[16:35] natrium42 (n=alexei@auth7-137.uwaterloo.ca) joined #highaltitude.
[16:35] <natrium42> hi
[16:35] <jcoxon> hey natrium42
[16:35] <natrium42> sup?
[16:36] <jcoxon> still waiting for my results :-p
[16:36] <jcoxon> exams done?
[16:36] <natrium42> lol
[16:36] <natrium42> nope, last one in 1 hour
[16:36] <jcoxon> oh cool
[16:36] <jcoxon> then thats it?
[16:36] <natrium42> for this term :)
[16:36] <jcoxon> sorry, will bbiab
[16:36] <natrium42> k
[16:39] <jcoxon> back
[16:39] <jcoxon> thought someone was at the door
[16:40] <natrium42> it was me
[16:40] <jcoxon> crap!
[16:40] <natrium42> but i left quickly
[16:40] <jcoxon> you'll never make it back to the exam
[16:40] <natrium42> so any luck winning a camera on ebay?
[16:40] <jcoxon> i'm watching a few
[16:40] <jcoxon> i think the ixus400 might be the best for the job
[16:41] <natrium42> yeah, i am also thinking to get one
[16:41] <natrium42> verdex doesn't have usb 2 unfortunately
[16:41] <jcoxon> uub?
[16:41] <jcoxon> hub*
[16:41] <natrium42> so it's a bit slow to transfer pictures
[16:41] <jcoxon> oh right
[16:41] <natrium42> 2.0
[16:41] <jcoxon> i miss read that as 2 usb ports
[16:41] <jcoxon> haha
[16:41] <natrium42> hehe
[16:41] <jcoxon> i'm in too much hurry
[16:42] <natrium42> they better post the results soon
[16:42] <jcoxon> whos results?
[16:42] <natrium42> ixus400 looks good for the transatlantic floater
[16:42] <natrium42> yours
[16:42] <jcoxon> yes
[16:42] <jcoxon> could we try sstv on the TA floater?
[16:43] <natrium42> sure, if you make the subsystems
[16:43] <jcoxon> hmmmm....
[16:43] <natrium42> i wanted to sketch up things to do
[16:43] Action: jcoxon has a little think
[16:43] <natrium42> so that we can divide up the work
[16:43] <jcoxon> feel free to use the ukhas wiki
[16:43] <natrium42> going to do it now that exams will be over
[16:44] <natrium42> k
[16:44] <jcoxon> 17 mins and i'll know if my exams are over
[16:44] <natrium42> do you have any experience with ballasts?
[16:44] <jcoxon> that said if they aren't i'm screwed as that only gives me 4 3 days to prepare
[16:44] <jcoxon> natrium42, ummmm not on my own balloons, yes on CNES ZP balloons
[16:44] <natrium42> so if you don't do well, you have to write more exams?
[16:44] <jcoxon> they use lead ball bearings
[16:45] <natrium42> how did they attach them?
[16:45] <jcoxon> yeah i have to redo my writtens and also redo my practical exams (which i passed previously)
[16:45] <jcoxon> they had them in a container with a valve or something
[16:45] <natrium42> interesting
[16:45] <jcoxon> i've heard of ethanol being used
[16:45] <natrium42> what do you think about dropping empty batteries?
[16:46] <jcoxon> won't make you popular especially lithiums
[16:46] <jcoxon> but then its what happens when we lose payloads
[16:46] <natrium42> and the proper way to dispose lithiums is to discharge them, then put them into salt water :P
[16:46] <jcoxon> is it?
[16:47] <jcoxon> hehe
[16:47] <natrium42> yeah
[16:48] <natrium42> for ethanol, there has to be a valve at the top too, right?
[16:48] <natrium42> to let air in
[16:48] <jcoxon> yeah
[16:48] <jcoxon> i remember Bill Brown said something about this
[16:48] <natrium42> i wonder if valves are reliable
[16:49] <jcoxon> i'm sure a oneway rubber valve would be fine
[16:49] <jcoxon> especially if there was quite a bit of air
[16:49] <natrium42> maybe it's better to use discrete weights?
[16:49] <natrium42> and nichrome wire for each of them
[16:49] <natrium42> to cut them off
[16:50] <jcoxon> i think some tests were done with dripping liquid
[16:50] <jcoxon> as the balloon went up the dripping rate changed appropriately to the pressure
[16:51] <natrium42> interesting, but seems like calibration is required
[16:51] <gordonjcp> I'd be worried about the batteries hitting something
[16:51] <gordonjcp> what about sand?
[16:51] <natrium42> whales?
[16:51] <natrium42> SAVE THE WHALES!!!
[16:51] <jcoxon> Wales?
[16:51] <natrium42> lol
[16:51] <jcoxon> Save Wales!
[16:52] <natrium42> it could, if it comes that far!
[16:52] <gordonjcp> condom full of sand, tied shut with a loop of nichrome ;-)
[16:52] <natrium42> you have an interesting mind :D
[16:52] <edmoore> natrium42: are you pushing for the transatlantic attempt?
[16:52] <natrium42> edmoore, well, i already spent over $2k on it
[16:52] <natrium42> so i guess i have to
[16:53] <jcoxon> haha, thats not a reason!
[16:53] jnd (n=jenda@unaffiliated/jnd) joined #highaltitude.
[16:53] <natrium42> main reason being that it would be freaking cool
[16:53] Ei5GTB_ (n=Paul@213-202-133-145.bas502.dsl.esat.net) joined #highaltitude.
[16:53] <natrium42> edmoore, but i have reliable communications, now i need reliable balloon
[16:53] <jcoxon> natrium42, utarc use liquid ballast
[16:54] <edmoore> natrium42: we are build a weldomatic
[16:54] <edmoore> an automatic ZP balloon cutting/making thing
[16:54] <jcoxon> oh i can speak to mark caviezel if you want
[16:54] <natrium42> edmoore, cool
[16:54] <edmoore> very much our black project, and I am giving it away here, but you will be the first to know/beta test if you want :)
[16:54] <natrium42> jcoxon, would be much appreciated
[16:55] <edmoore> but yeah, mark caveize could probably give you one which works tomorrow
[16:55] <natrium42> maybe you guys build the balloon & ballast & radio system
[16:55] <natrium42> and i build the camera & satphone system
[16:55] <natrium42> and then put everything together here
[16:56] <natrium42> of course, you are invited to come, flight is only $500 or so, right? :P
[16:56] <jcoxon> was talking to mark once - he reckoned there was a time when we could launch to you
[16:56] <jcoxon> but i wasn't convinced
[16:56] <jcoxon> as in a jetstream direction reversal
[16:56] <natrium42> doesn't it happen each 18 months?
[16:56] <jcoxon> i've never seen it
[16:59] <jcoxon> phwwwwww
[16:59] <jcoxon> i dont' need to go back an retake
[16:59] <jcoxon> thanks god
[16:59] <jcoxon> thank*
[17:00] <natrium42> cool, congrats!
[17:00] <natrium42> can you practice medicine yet?
[17:00] <jcoxon> no, not yet :-)
[17:00] <natrium42> not even on pets?
[17:01] <jcoxon> actually it doesn't mean i've passed - just that i don't have to retake the exams
[17:01] <jcoxon> hehe,
[17:01] <jcoxon> funny thing is that its illegal for doctors to treat animals but not illegal in an emergency for vets to treat humans
[17:01] <natrium42> edmoore, when is next night launch?
[17:01] <natrium42> haha, wtf
[17:01] <edmoore> natrium42: hrm.
[17:02] <natrium42> is it soon? :D
[17:02] <natrium42> yes? yes?
[17:02] <edmoore> fergus is in cambridge with all the hardware, annoyingly I've not had the chance to do some much recently being on a summer job elsewhere, so he's probably got a better idea of when exactly. But probably sepember
[17:03] <edmoore> we're building the badger truck - just a general physical payload with the same thinking as the badger board, and we'll probably build the next revision of the badger board too.
[17:03] <natrium42> ah, i see
[17:06] robert1971 (n=user@212.183.136.194) joined #highaltitude.
[17:08] <SpeedEvil> I've wondered about a natural gas balloon.
[17:08] <SpeedEvil> With propane liquid in a cup, which is exposed to the gas at altitude, and then lots of natural gas dissolves in it.
[17:08] <SpeedEvil> Heat the propane, and the natural gas boils out.
[17:09] <natrium42> lol
[17:09] <natrium42> there was a big propane explosion in toronto yesterday :P
[17:09] <SpeedEvil> It wasn't me!
[17:09] <natrium42> 12 000 people evacuated
[17:09] <natrium42> yes, yes :P
[17:10] EI5GTB (n=Paul@78.16.197.131) left irc: No route to host
[17:12] edmoore (n=edmoore@smtp.vorticity-systems.com) left irc:
[17:15] <SpeedEvil> Fuel air explosions are nasty shit.
[17:15] <SpeedEvil> Almost as fun as fertiliser.
[17:15] <SpeedEvil> You're canadian?
[17:16] <natrium42> yes
[17:17] <natrium42> exam time
[17:17] natrium42 (n=alexei@auth7-137.uwaterloo.ca) left irc:
[17:17] <jcoxon> good luck!
[17:17] <jcoxon> oh he's gone
[17:18] <SpeedEvil> Oh - I was thinking of the texas explosion.
[17:18] <SpeedEvil> http://en.wikipedia.org/wiki/Halifax_Explosion
[17:18] <SpeedEvil> http://www.local1259iaff.org/disaster.html I mean
[17:20] <jcoxon> back in a bit
[17:20] jcoxon (n=jcoxon@85-210-73-114.dsl.pipex.com) left irc: "Leaving"
[17:20] robert1971 (n=user@212.183.136.194) left #highaltitude.
[17:42] Ebola (n=Ebola@unaffiliated/ebola) left irc: "home time"
[17:50] edmoore (n=edmoore@37.124-84-212.ippool.ndo.com) joined #highaltitude.
[18:03] jnd (n=jenda@unaffiliated/jnd) left irc: Read error: 110 (Connection timed out)
[18:04] phatmonkey (n=ben@scooby.firshman.co.uk) joined #highaltitude.
[18:40] kamaji (n=kamaji@handtomouse.demon.co.uk) joined #highaltitude.
[18:40] <kamaji> hey all
[18:41] <edmoore> hi
[18:41] <kamaji> How's things?
[18:43] <edmoore> not bad, having a mental fpga learning curve session
[18:48] <kamaji> ooh, funfun
[18:48] <kamaji> what are you planning to do with it?
[18:48] edmoore_ (n=edmoore@37.124-84-212.ippool.ndo.com) joined #highaltitude.
[18:49] edmoore (n=edmoore@37.124-84-212.ippool.ndo.com) left irc: Read error: 104 (Connection reset by peer)
[18:55] Hiena (n=Hiena@81.93.195.181.datatrans.hu) joined #highaltitude.
[19:19] akawaka (n=akawaka@external.treyarch.com) joined #highaltitude.
[19:37] edmoore_ (n=edmoore@37.124-84-212.ippool.ndo.com) left irc:
[20:19] MetaMorfoziS (n=eee@53d83582.adsl.enternet.hu) joined #highaltitude.
[20:21] natrium42 (n=alexei@CPE000625d867e2-CM0014045885be.cpe.net.cable.rogers.com) joined #highaltitude.
[20:50] Hiena (n=Hiena@81.93.195.181.datatrans.hu) left irc: "-=Halt! Hammerzeit!=-"
[21:01] jcoxon (n=jcoxon@85-210-73-114.dsl.pipex.com) joined #highaltitude.
[21:07] <jcoxon> hey all
[21:08] icez (n=icez@ip68-98-34-247.ph.ph.cox.net) joined #highaltitude.
[21:26] <jcoxon> natrium42, oops i missed my ixus auction
[21:26] <natrium42> d'oh
[21:27] <jcoxon> exam go okay?
[21:27] <natrium42> that's why you should use gixen :P
[21:27] <natrium42> yeah, very easy (shouldn't have studied)
[21:27] <jcoxon> haha
[21:27] Laurenceb (n=Laurence@host86-133-69-206.range86-133.btcentralplus.com) joined #highaltitude.
[21:27] <natrium42> not that i studied much
[21:27] <natrium42> only a few hours before exam
[21:27] <Laurenceb> hi everyone
[21:27] <jcoxon> hey Laurenceb
[21:27] <natrium42> hey lb
[21:27] <Laurenceb> hi :D
[21:27] <Laurenceb> natrium: an exam?
[21:28] <natrium42> i just had my last exam
[21:28] <natrium42> it was easy
[21:28] <Laurenceb> congratulations and good luck
[21:28] <natrium42> intro to astrophysics
[21:28] <Laurenceb> ha
[21:29] <natrium42> :)
[21:29] <natrium42> what would *you* know about physics! :P
[21:29] <Laurenceb> so, whats the schwarzschild radius?
[21:29] <natrium42> um, let me check something quickly
[21:30] Action: jcoxon wins!
[21:30] <natrium42> grr, that's exactly what i was doing
[21:30] <natrium42> XD
[21:30] <jcoxon> good old zeusbot
[21:31] <natrium42> i should have remembered it though, since i did take intro to quantum physics a few years ago...
[21:32] <natrium42> bbl food
[21:33] Nick change: Ei5GTB_ -> EI5GTB
[21:34] jnd (n=jenda@unaffiliated/jnd) joined #highaltitude.
[21:41] <Laurenceb> any maths gurus on here?
[21:41] <jcoxon> not me!
[21:42] <Laurenceb> I've got a set of datapoints with errors and want to fit a circle to them
[21:42] Action: SpeedEvil suffers a guru meditation error.
[21:42] <Laurenceb> - so a single datapoint has x,y, and error values
[21:42] Action: SpeedEvil contemplates.
[21:43] <SpeedEvil> an error circle?
[21:43] <Laurenceb> I actually only require the center of the cirle
[21:43] <Laurenceb> no, a best fit cirle to the points
[21:43] <SpeedEvil> I mean - are the error values circular?
[21:43] <Laurenceb> its an idea for improving my UAV glider code
[21:43] <Laurenceb> yes
[21:43] <SpeedEvil> Umm...
[21:44] <SpeedEvil> I'm 37.2% certain that that's a simple average of all teh points.
[21:44] <SpeedEvil> To give you a centre.
[21:44] <Laurenceb> I want to take the past 60 seconds or so of gps velocities and do this to them
[21:44] <Laurenceb> no, they will be just spread along one edge of the cirle is its flying properly
[21:44] <SpeedEvil> ah - right
[21:44] <Laurenceb> as airspeed is ~constant
[21:45] <SpeedEvil> you're assuming that you can fit the past course to a circle?
[21:45] <Laurenceb> but the air bearing fluctuates a bit
[21:45] <SpeedEvil> Of varying radius
[21:45] MetaMorfoziS (n=eee@53d83582.adsl.enternet.hu) left irc: Remote closed the connection
[21:45] <Laurenceb> yes, the airspeed changes with height
[21:45] <Laurenceb> as does wind vector
[21:45] <Laurenceb> but over ~60 seconds it should be ok
[21:45] <SpeedEvil> Why is that useful?
[21:46] <Laurenceb> and I think its possible to get a statistically useful result from 60 datapoints
[21:46] <SpeedEvil> Sorry, I don't follow why you want to do this.
[21:46] <Laurenceb> you need to know the air vector to avoid going hopelessly out of control
[21:47] <SpeedEvil> the relative wind vector?
[21:47] <Laurenceb> as the wind speeds you encouter will be comparible or even larger than the glider airspeed
[21:47] <Laurenceb> gps gives you ground speed, so you need to convert reference frames
[21:48] <SpeedEvil> Of course.
[21:48] <SpeedEvil> Is this still for the parafoil?
[21:49] borism (n=boris@195-50-197-174-dsl.krw.estpak.ee) left irc: Connection timed out
[21:49] <Laurenceb> or rogallo, yes
[21:50] <SpeedEvil> pitot?
[21:50] <SpeedEvil> Hang on.
[21:50] <SpeedEvil> If you've just got one servo, you can't stall it.
[21:51] <Laurenceb> Speed: http://wiki.ukhas.org.uk/code:parafoil_main look down near the bottom
[21:51] <SpeedEvil> yes?
[21:51] <Laurenceb> you cant stall it, but the PID loop could go out of control
[21:52] <Laurenceb> thats the code for changing reference frames
[21:52] <SpeedEvil> ah - right
[21:52] <Laurenceb> it just records wind to eeprom on the ascent atm
[21:53] <SpeedEvil> Well...
[21:53] <SpeedEvil> You've got an accelerometer there?
[21:53] <Laurenceb> nope
[21:53] <SpeedEvil> Ah.
[21:53] <Laurenceb> but how would that help?
[21:54] <Laurenceb> gyro
[21:54] <SpeedEvil> If you're in a crosswind, that'd tell you
[21:54] <SpeedEvil> as with a crosswind, and are maintaining a course, teh wings will be canted
[21:54] <Laurenceb> it would only detect gusts accurately
[21:54] <SpeedEvil> Oops.
[21:55] <Laurenceb> and the existing PID control seems to handle gusts ok
[21:55] <SpeedEvil> Am I assuming coordinated flight
[21:55] <SpeedEvil> So why the circle?
[21:56] <SpeedEvil> you want to command a given rudder input, and use it to measure wind?
[21:56] <Laurenceb> I explained earlier
[21:56] <Laurenceb> no, I think it will wobble about enough as it is
[21:56] <SpeedEvil> I don't think you did - or at least I missed it.
[21:57] <Laurenceb> brb
[21:58] <SpeedEvil> How do you go from matching a segment of a circle to a wind vector, and why do you assume that you're flying in a circle?
[21:58] <SpeedEvil> k
[22:01] Action: SpeedEvil ponders why Laurenceb might have left, and notices the Olympics Beach Volleyball is on.
[22:01] Action: kamaji titters
[22:02] <Laurenceb> haha
[22:02] <Laurenceb> sorry was cooking
[22:03] <Laurenceb> actually I think my idea is flawed - it will pick up on gusts of crosswind
[22:03] <Laurenceb> and the resulting corrections from the PID control
[22:04] <Laurenceb> however, introducing a rudder fluctuation at some knows frequency the using some sort of frequency filtering of the heading data could reveal the wind vector
[22:04] <Laurenceb> *known
[22:04] <SpeedEvil> Can you assume that IAS is constant?
[22:04] <SpeedEvil> With no control input.
[22:04] <Laurenceb> IAS?
[22:04] <SpeedEvil> indicated airspeed
[22:05] <Laurenceb> got you, I think so
[22:05] <SpeedEvil> You don't even have a gyro do you?
[22:05] <Laurenceb> yes
[22:05] <Laurenceb> gps and gyro
[22:05] <SpeedEvil> Given heading, and atmospheric pressure, you have airspeed.
[22:05] <Laurenceb> IAS is ~ constant over ranges of a few 100m altitude if you can filter out effects from gusts of wind
[22:06] <SpeedEvil> Can't you just subtract out the GPS signal?
[22:06] <Laurenceb> I only have ground speed
[22:06] <SpeedEvil> IAS - not airpseed
[22:06] <Laurenceb> from the GPS
[22:06] <Laurenceb> sorry ground vector
[22:06] <Laurenceb> or more correctly velocity
[22:07] <Laurenceb> not my day today
[22:08] <SpeedEvil> http://en.wikipedia.org/wiki/Calibrated_airspeed
[22:08] <Laurenceb> I think you'd need a compass to work out air velocity directly
[22:08] <Laurenceb> bbl, food
[22:11] <SpeedEvil> I'm fairly sure that as your thingy is fixed geometry - it should (after some moments) reach a speed at which the calibrated airspeed is identical - whatever the altitude (till it goes supersonic, when things change).
[22:11] <SpeedEvil> Neglecting manouvering, which will slow it, as it'll wallow like a pig in turns, not fly nice and coordinated.
[22:12] <SpeedEvil> Ah
[22:13] <SpeedEvil> equivalent airspeed may be the term I was looking for.
[22:13] <SpeedEvil> http://en.wikipedia.org/wiki/Equivalent_airspeed
[22:13] <SpeedEvil> Equivalent airspeed (EAS) is the airspeed at sea level which represents the same dynamic pressure as that flying at the true airspeed (TAS) at altitude. It is useful for predicting aircraft handling, aerodynamic loads, stalling etc.
[22:16] jcoxon (n=jcoxon@85-210-73-114.dsl.pipex.com) left irc: "Leaving"
[22:26] edmoore (n=edmoore@37.124-84-212.ippool.ndo.com) joined #highaltitude.
[22:35] callme (n=callme@212-198-248-33.rev.numericable.fr) joined #highaltitude.
[22:36] <Laurenceb> back
[22:36] <Laurenceb> hi edmoore
[22:37] callme (n=callme@212-198-248-33.rev.numericable.fr) left irc: Client Quit
[22:39] <edmoore> Laurenceb: hi
[22:41] jnd (n=jenda@unaffiliated/jnd) left irc: Connection timed out
[22:46] jnd (n=jenda@unaffiliated/jnd) joined #highaltitude.
[22:48] <SpeedEvil> I was basically saying EAS - which can be calculated from the altitude and pressure is known. If you've not commanded large course changes recently, the EAS should be pretty close to the glide-speed near the ground. (even though the true airspeed may be much higher at altitude) So, you simply take the opposite of the heading vector, and add this to the groundspeed, to get windspeed.
[22:48] <SpeedEvil> s/heading/pointing/
[22:49] <SpeedEvil> though on reflection you don't have a nice pointing vector, you've got to compute that from the gyro and GPS output.
[23:18] edmoore (n=edmoore@37.124-84-212.ippool.ndo.com) left irc:
[23:18] <Laurenceb> sorry, got dragged off again
[23:18] <Laurenceb> yes
[23:19] <Laurenceb> you need a compass for that
[23:20] <Laurenceb> ok, I think I've worked something out: you add a sinusoidal fluctuation to the rudder, then use a digital filter to find that frequency in the gps ground heading and kalman filter heading
[23:20] <Laurenceb> from the amplitudes and phases you can find the wind vector
[23:20] <SpeedEvil> You're BARKING MAD.
[23:21] <SpeedEvil> Just add a damn compass.
[23:21] <SpeedEvil> :)
[23:21] <Laurenceb> the sinusoidal fluctuation would so something as low as possible, like 60 second period
[23:21] <Laurenceb> I want to try it :P
[23:21] <Laurenceb> its a firmware rather than hardware solution
[23:21] <Laurenceb> no new hardware is needed
[23:21] <SpeedEvil> Or to go really cheap - a sun sensor.
[23:22] <SpeedEvil> phototransistor and a cylinder lens.
[23:22] <Laurenceb> that doesnt work if its cloudy
[23:22] <SpeedEvil> True.
[23:22] <SpeedEvil> I was assuming you could get up high enough to be in the sun to get a lock, then keep that
[23:22] <Laurenceb> and moving parts always end in disaster
[23:22] <SpeedEvil> moving parts?
[23:22] <SpeedEvil> naah.
[23:22] <SpeedEvil> You circle once and see the sun.
[23:22] <Laurenceb> oh god
[23:23] <SpeedEvil> note teh gyro position, and call it north
[23:23] <Laurenceb> mems gyros arent that good
[23:23] <SpeedEvil> I thought they were a degree a minute or so.
[23:23] <Laurenceb> not in real life
[23:23] <SpeedEvil> Oh - you're not ovening it.
[23:23] <Laurenceb> I have managed that with a few tests
[23:24] <Laurenceb> but once its at repidly changing temperature
[23:24] <Laurenceb> and vibrating, no chance
[23:24] <Laurenceb> *rapidly
[23:24] <SpeedEvil> A neutrino detector, picking up the sun wouldn't suffer from clouds :)
[23:25] <SpeedEvil> You might need to increase the size of the balloon.
[23:25] <Laurenceb> I like the rudder fluctuation technique as it could be quite subtle and you could still pick it out of the noise
[23:25] <Laurenceb> due to the fact you could have a high Q filter
[23:26] <Laurenceb> Q ~ 100 should be possible
[23:26] <SpeedEvil> You have a rudder?
[23:26] <SpeedEvil> Or do you just mean weight shift
[23:28] Action: SpeedEvil wonders about 180 degree teeny lenses.
[23:29] <Laurenceb> mean weight shift
[23:29] <Laurenceb> actually thast was wrong, Q ~ 10 is more likely
[23:29] <SpeedEvil> Let me recap.
[23:30] <SpeedEvil> You're wiggling the rudder at 1/60Hz, and observing the changes in course on the GPS. You're using this to derive pointing?
[23:30] <SpeedEvil> I assume you mean not blindly, but commanding a +-10 degree turn
[23:31] <Laurenceb> correct
[23:31] <Laurenceb> but I'm using it to derive wind vector
[23:31] <Laurenceb> also it'd be wiggled at around 1/6 or 1/10 Hz
[23:31] <SpeedEvil> I don't see how it gives you wind vector straight.
[23:32] <Laurenceb> you need to take the kalman filter output as well
[23:32] <SpeedEvil> I can see pointing.
[23:32] <Laurenceb> it traces an arc
[23:32] <Laurenceb> that is centered on the wind vector
[23:33] <SpeedEvil> And you can go from that + EAS to a airspeed vector, then subtract from the ground vector to get wind
[23:33] <Laurenceb> more or less
[23:34] <Laurenceb> if you just take the "rudder wiggle" frequency you dont have an arc
[23:34] <Laurenceb> just a line in the plane
[23:34] <SpeedEvil> right - oscillating left and right (say0
[23:34] <Laurenceb> you need to use the kalman filtered pointing output to get the magnitude of the oscillation
[23:34] <Laurenceb> then you can do some trig to find the point the arc centers on
[23:35] <SpeedEvil> I really don't like that idea.
[23:35] <Laurenceb> hehe
[23:35] <SpeedEvil> It smells of crashing and burning - trying to get too far from the data into spaces where it might not mean much given errors.
[23:36] <Laurenceb> your basically using the rate gyro data to find the moagnitude of the oscillation
[23:36] <Laurenceb> IMO the frequency specific filtering gets rid of a ton of your noise
[23:37] <SpeedEvil> You need speed too to get the magnitude.
[23:37] <SpeedEvil> TAS
[23:37] <Laurenceb> the rate gyro gives a pretty clean signal in my test flights
[23:37] <SpeedEvil> Yes - you get a clean signal - but the speed will vary with altitude significantly AIUI.
[23:38] <Laurenceb> which speed?
[23:38] <SpeedEvil> the airspeed.
[23:38] <Laurenceb> that doesnt matter
[23:39] <Laurenceb> sorry - that means you have to limit your integration time to a minute or so
[23:39] <SpeedEvil> If you command +-10 degrees every 10 seconds, you will get a sine (or whatever) osicllation in gyro output at .1Hz. The amplitude - in meters that this goes left and right of your base course depends on airspeed.
[23:39] <SpeedEvil> true airspeed
[23:40] <Laurenceb> gyro gives heading
[23:40] <Laurenceb> degrees not meters
[23:40] <SpeedEvil> right
[23:41] <SpeedEvil> your basically using the rate gyro data to find the moagnitude of
[23:41] <SpeedEvil> the oscillation
[23:41] <SpeedEvil> I assumed you were talking meters, as you know the magnitude of the commanded oscillation.
[23:41] <Laurenceb> ok
[23:41] <Laurenceb> youve got it now
[23:42] <SpeedEvil> I don't see that this oscillation describes a segment of a circle though.
[23:42] <Laurenceb> draw a diagram :P
[23:43] <SpeedEvil> Hmm.
[23:43] <SpeedEvil> I think I see.
[23:45] <SpeedEvil> I wonder what happens if you draw a graph of n seconds ago GPS position delta rotated by the then-gyro pointing.
[23:45] <SpeedEvil> Where n is 0-60
[23:45] <Laurenceb> http://wiki.ukhas.org.uk/_media/projects:ukhas_glider_project:fft.png thats a fft of real world test data
[23:46] <Laurenceb> I set the PID loop up so it oscillated
[23:46] <SpeedEvil> Ffft. FFTs.
[23:58] <Laurenceb> IMO the rudder fluctuation will stick out quite well
[23:58] <Laurenceb> its just a question of the noise on the gps
[00:00] --- Tue Aug 12 2008