highaltitude.log.20070323

[00:00] edwardmoore (n=edwardmo@88.212.167.121) joined #highaltitude.
[00:01] <edwardmoore> hi jcoxon
[00:01] <edwardmoore> sorry i'm late- had a bit of an eve
[00:01] <jcoxon> hey ed
[00:01] <jcoxon> its okay but i can't stay for long
[00:01] <edwardmoore> just read steve's email
[00:01] <jcoxon> need sleep
[00:01] <edwardmoore> he has a point about weather...
[00:01] <edwardmoore> same!
[00:02] <jcoxon> wunderground forcasts rain around lunch
[00:02] <jcoxon> but thats about it
[00:02] <edwardmoore> \what do you reckon?
[00:02] <jcoxon> but the winds will be quite strong
[00:02] <jcoxon> bbc say 20 mph
[00:03] <jcoxon> next week does look nicer
[00:03] <edwardmoore> aye. warmer too!
[00:03] <edwardmoore> i was outside on my car all of today and work was a real pain
[00:03] <edwardmoore> forgot what it does to your fingers
[00:03] <jcoxon> well i'm pretty easy
[00:03] <edwardmoore> like a sunday morning
[00:04] <edwardmoore> tis your call as it's your house etc
[00:04] <edwardmoore> i would say my car might appreciate a lil more TLC before it goes in a trip to camby, but it would probably make it anyway
[00:05] <jcoxon> hmmho
[00:05] <jcoxon> lets postpone
[00:05] <jcoxon> its to windy
[00:05] <jcoxon> 20 mph is over even what i'd be happy to launch in with the real thing
[00:05] <edwardmoore> there's rain passing over the weekend but then it goes back to high pressure next week, which does look much beter
[00:06] <jcoxon> tues looks good on the bbc
[00:06] <icez> you guys are just living at the worst place for launching balloons
[00:06] <icez> :P
[00:06] <jcoxon> yup
[00:06] <edwardmoore> am busy monday btw
[00:06] <edwardmoore> but yeah, tues looks good
[00:08] <jcoxon> okay i'll email steve and text him
[00:09] <edwardmoore> ok cool
[00:14] <jcoxon> done
[00:16] <edwardmoore> right, i need some sleep
[00:16] <edwardmoore> cya all
[00:18] jcoxon (n=jcoxon@85-210-179-186.dsl.pipex.com) left irc: "Leaving"
[00:19] edwardmoore (n=edwardmo@88.212.167.121) left irc: "toodle-pip"
[02:10] VA3NEE (n=Snoopl3s@209.105.206.171) left irc: Remote closed the connection
[02:29] VA3NEE (n=Snoopl3s@209.105.206.171) joined #highaltitude.
[02:50] VA3NEE (n=Snoopl3s@209.105.206.171) left irc: Remote closed the connection
[02:50] VA3NEE (n=Snoopl3s@209.105.206.171) joined #highaltitude.
[03:11] VA3NEE (n=Snoopl3s@209.105.206.171) left irc: Remote closed the connection
[05:17] icez (n=icez@ip68-3-56-121.ph.ph.cox.net) left irc: Remote closed the connection
[08:12] rocketboy (n=steve@host86-132-42-48.range86-132.btcentralplus.com) joined #highaltitude.
[08:12] rocketboy (n=steve@host86-132-42-48.range86-132.btcentralplus.com) left irc: Client Quit
[08:14] jcoxon (n=jcoxon@85-210-179-186.dsl.pipex.com) joined #highaltitude.
[10:25] edwardmoore (n=edwardmo@88.212.167.121) joined #highaltitude.
[10:25] <edwardmoore> morning jcoxon
[10:43] LaurenceB (n=laurence@host81-132-17-240.range81-132.btcentralplus.com) joined #highaltitude.
[12:02] <jcoxon> hey edwardmoore
[12:02] <edwardmoore> ello fella
[12:02] <edwardmoore> ice-cube toes i hear?
[12:03] <jcoxon> tis a little cold
[12:03] <edwardmoore> good call
[12:03] <jcoxon> will be nice next week
[12:03] <jcoxon> woohoo
[12:04] <jcoxon> got my 386 to be headless and have a serial console
[12:04] <edwardmoore> i didn't get round to finishing my car yesterday either, partly through cold and partly because you need to replace the crankshaft fulley retaining nut everytime you use it, which i didn't know
[12:04] <edwardmoore> pulley*
[12:04] <edwardmoore> oooooh- niceness
[12:04] <edwardmoore> got any steppers yet? :p
[12:05] <edwardmoore> oh i have a linux q
[12:05] Superkuh (n=synthase@c-24-245-25-187.hsd1.mn.comcast.net) joined #highaltitude.
[12:05] <jcoxon> i've got steppers but getting them to work is a different matter
[12:05] <jcoxon> fire away with a linux q
[12:05] <edwardmoore> i want to make a really really quiet linux box as a server for the home, and to back up onto and stuff like that. can u recommend a distro?
[12:05] <edwardmoore> or will anything do
[12:06] <jcoxon> have you used linux before?
[12:06] <edwardmoore> ubuntu briefly....
[12:06] <edwardmoore> i am happy to go on a learnign curve though
[12:06] <jcoxon> was about to recommend ubuntu
[12:06] <edwardmoore> that's partly the idea behind doing this
[12:06] <jcoxon> i use zenwalk
[12:06] <jcoxon> but that suits me
[12:07] <edwardmoore> what differentiats it from, say, ubuntu?
[12:07] <jcoxon> people like ubuntu as its based on debian
[12:07] <jcoxon> therefore it has the apt system for installation
[12:08] <jcoxon> so if you want a program you literally type
[12:08] <jcoxon> apt-get install "program"
[12:08] <jcoxon> and it'll do everything for you
[12:08] <edwardmoore> yeah that's a little spooky
[12:08] <jcoxon> zenwalk is based on slackware i think
[12:08] <edwardmoore> where is the lookup table for all that stuff- t'internet or is it all sitting on the hdd somewhere waiting for the limelight?
[12:09] <jcoxon> and instead of gnome uses XCFE
[12:09] <jcoxon> sort of both
[12:09] <edwardmoore> what's the difference?
[12:09] <jcoxon> XCFE is a much lighter cleaner window manager
[12:10] <jcoxon> but isn't as fancy
[12:10] <jcoxon> http://www.zenwalk.org/images/shots/shot-office.png
[12:10] <Superkuh> edwardmoore: There are many repositories online from which to update the local cache for aptitude.
[12:10] <jcoxon> the computer i'm on is dual monitor - XFCE works nicely with it
[12:11] <edwardmoore> i used beryl the other day. it made me giigle for about 2 minutes, before i wondered why i needed it
[12:12] <jcoxon> hehe beryl is pretty but also pretty useles
[12:12] <jcoxon> s
[12:13] <edwardmoore> so.... the pretty thing about ubuntu was that i stuck the cd, then it just installed on startup etc
[12:13] <edwardmoore> if i downloaded zenwalk, would it be like that or do i have to do a bit more groundwork?
[12:13] <jcoxon> nah zenwalk would be the same
[12:13] <jcoxon> if you want to go hardcore from the ground up
[12:13] <jcoxon> the use gentoo
[12:13] <jcoxon> then when you get bored move on to LFS :-D
[12:14] <edwardmoore> lfs?
[12:14] <jcoxon> linux from scratch
[12:14] <jcoxon> literally compile everything
[12:14] <edwardmoore> also zenwalk doesn't have any of this 'i can't support any common vid/music formats' licensing issues out of the box, or does it?
[12:15] <edwardmoore> yeah i've heard lots about gentoo. it sounds a little scary
[12:15] <jcoxon> the best place to look is http://distrowatch.com/
[12:15] <jcoxon> i think there are better distros for you then zenwalk
[12:16] <jcoxon> basically i use zenwalk, crux ppc on my laptop and then small linux 0.9 on my 386
[12:16] <edwardmoore> well someone gave me a liveCD for ubuntu but it kept giving me CRC errors, so i'll download it and reburn it ultra slowly
[12:16] <edwardmoore> and see how i get on
[12:16] <edwardmoore> it's going on my old centrino laptop
[12:17] <jcoxon> oh cool
[12:28] <edwardmoore> back in half an hour or so
[13:14] <edwardmoore> 1/3rd of the way through ubuntu...
[13:14] <edwardmoore> i need a better connection
[13:22] LaurenceB (n=laurence@host81-132-17-240.range81-132.btcentralplus.com) left irc:
[14:40] LaurenceB (n=laurence@host81-132-17-240.range81-132.btcentralplus.com) joined #highaltitude.
[14:47] <LaurenceB> Rocketboys diagram for directional control is good
[14:47] <LaurenceB> but it doesnt do wind compensation
[14:48] <LaurenceB> http://wiki.ukhas.org.uk/_media/projects:parafoil:control.gif?cache=cache
[14:49] <LaurenceB> if the wind is of similar strenght to the speed of flight then the response is nonlinear and the PID loop will lose control
[14:49] <LaurenceB> strenght=speed
[14:59] <edwardmoore> i'm not totally convinced the PID loop will actually loose control
[15:00] <edwardmoore> i mean in that situation, you have control saturation anyway and the pid loop is the least of your problems
[15:08] Action: jcoxon really shouln't try and set up lilo over a serial console
[15:08] <LaurenceB> well if you imagine drawing vectors...
[15:08] <LaurenceB> then the wind wound be some line form the origin
[15:09] <LaurenceB> and the possible headings the glider could have would form a circle around it
[15:10] <LaurenceB> the absolute heading of any point on that circle is a nonlinear function of the gliders heading relative to the air
[15:10] <LaurenceB> ie relative to the end of the wind vector
[15:10] <LaurenceB> wound=would
[15:13] Action: jcoxon is going to put his baby 386 in to storage for a while - sadly thats the end of the stepper motor driven hot wire cutter :-D
[15:13] <LaurenceB> PID+nonlinear=BAD
[15:27] <edwardmoore> jcoxon why??????
[15:27] <edwardmoore> what did the poor thing do?
[15:27] <LaurenceB> of course when you have gusts of wind it screws it all, thats why a fast flying glider is good, but yesterdays tests were in gusty wind and the PID seemed to be recovering once it had be retuned a few times
[15:27] <jcoxon> well i changed the lilo settings and broke it
[15:27] <jcoxon> its fixable but i need to get a monitor attached
[15:27] <LaurenceB> oh jcoxon I have an XYZ table I pulled out of a skip the other day
[15:28] <jcoxon> which means removing my dodgey resistor hack at the back which is faking a monitor to allow it to boot
[15:28] <edwardmoore> lol
[15:28] <jcoxon> and while i do have a monitor in the attic its far to much effor
[15:28] <jcoxon> t
[15:29] <LaurenceB> do you have some sort of xyz device?
[15:30] <jcoxon> nah
[15:30] <jcoxon> was going to make one
[15:30] <LaurenceB> this thing I found in a skip is not very powerful at all, seems to be a graphics device
[15:30] <jcoxon> but to tell the truth it was only a crazy project
[15:30] <LaurenceB> picks up pens and draws with them
[15:30] <jcoxon> and i've got loads of them
[15:31] <LaurenceB> Also found a radioactive fridge in the same skip if anyones interested
[15:32] <jcoxon> LaurenceB, awfully kind but... :-D
[15:32] <LaurenceB> it was wrapped in "danger radiation" tape
[15:35] <jcoxon> i really need to get on and finish peg V
[15:35] <LaurenceB> wish it would stop raining so I could do some tests
[15:35] <jcoxon> it'll be nicer next week
[15:35] <LaurenceB> when is peg V launch?
[15:36] <jcoxon> there isn't a date yet
[15:36] <jcoxon> plan to finish first before setting a date
[15:36] <jcoxon> nice weather i guess
[15:36] <LaurenceB> I'm thinking of launching the test glider from 4km in 2 weeks
[15:36] <LaurenceB> it only needs a radio
[15:36] <jcoxon> cool
[15:36] <LaurenceB> already got a cut down for it
[15:37] <LaurenceB> reacon I can get the PID sussed with the next test series
[15:37] <jcoxon> hope the cutdown works all you'll be testing from 25km :-D
[15:37] <jcoxon> all -> or
[15:37] <LaurenceB> yes its the resistor type
[15:38] <LaurenceB> just check it gets hot enough
[15:38] <LaurenceB> there is limited space for the radio, only 4x3x1.5cm
[15:42] <jcoxon> right bbl
[15:42] jcoxon (n=jcoxon@85-210-179-186.dsl.pipex.com) left irc: "Leaving"
[15:44] <LaurenceB> looking at rf solutions, the modules are only 31.5mm long, so should work :)
[15:45] jcoxon (n=jcoxon@85-210-179-186.dsl.pipex.com) joined #highaltitude.
[15:45] <LaurenceB> wish the weather was like it says on the forecast
[15:45] <edwardmoore> back so son?
[15:46] <jcoxon> edwardmoore, swapping computers
[15:46] <LaurenceB> its too windy and wet to test the glider grrr
[15:46] <edwardmoore> MI:3 is utter bs
[15:47] <jcoxon> she is hot though
[15:47] <LaurenceB> lol
[15:47] <edwardmoore> they're doing a helicopter chase through a wind-farm. an unrealistically dense one too
[15:47] <LaurenceB> not as bad as the channel tunnel with two tracks
[15:47] <edwardmoore> and insist on flying between blades rather than at 200ft above them all
[15:48] <LaurenceB> at least there not flying down the channel tunnel that was bs
[15:51] <LaurenceB> apparently rocketboy is making a pcb version of the radio used on MiHAB2, but not sure how big it will be
[15:51] <LaurenceB> hopefully it will fit....
[16:02] <edwardmoore> laurence have you read the canadian guy's glider prject thingy?
[16:04] <LaurenceB> yes, he doesnt explain every espect of the code tho
[16:04] <edwardmoore> the git :)
[16:04] <edwardmoore> did you read his stuff about rudder vs aileron control tho?
[16:04] <LaurenceB> has some interesting sensors
[16:04] <LaurenceB> probably but forgotten
[16:04] <edwardmoore> i think it was actually just a sentence about it
[16:04] <LaurenceB> what was the conclusion?
[16:05] <edwardmoore> airleron control worked better than rudder control for direction changing
[16:06] <LaurenceB> rudder works fine in my tests
[16:06] <LaurenceB> but if I was to go with thermopile stabilization I'd use airleron
[16:06] <edwardmoore> uhuh
[16:07] <LaurenceB> with the weather would calm down...
[16:07] <LaurenceB> -wish
[16:07] <edwardmoore> give it the weekend
[16:07] <edwardmoore> next week looks better
[16:07] <LaurenceB> according to the forecast at 6pm it should be dry with 9mph wind
[16:07] <LaurenceB> thats good enough
[16:08] <LaurenceB> so why isnt it calming down?
[16:08] <edwardmoore> because forcasts really tend not to be that accurate?
[16:08] <LaurenceB> test J yesterday looked promising
[16:08] <edwardmoore> i don't really trust them beyond 'next week should better better than this week'
[16:09] <LaurenceB> think with a few more refinements it'll be sorted
[16:09] <LaurenceB> the D term turns out to be the most important, no wonder it wouldnt work before
[16:11] <edwardmoore> well of course! the thing was oscillating!
[16:11] <edwardmoore> very pretty oscillation too
[16:12] <LaurenceB> if P was too small, it only placed priority on going in a straight line - tests F and G
[16:12] <LaurenceB> well actually F started to turn back
[16:12] <LaurenceB> before crashing into a mound of cow muck
[16:13] <edwardmoore> lol
[16:13] <edwardmoore> nice
[16:13] <LaurenceB> the plane still smells
[16:14] <LaurenceB> G went into the top of a tree, at which point i decided it really needed P increasing!
[16:15] <LaurenceB> H is a bit off, there was a strong gust of side wind at 8 seconds
[16:18] <LaurenceB> test I, overdid D and I and then test J was the most promising, dont think there was much oscillation, instead there were gusts of wind at 4 seconds and 10 seconds
[16:19] <edwardmoore> how about adding RC control?
[16:19] <LaurenceB> not enough space in the plane
[16:20] <LaurenceB> also the glide tests are quite short, so wan to use all the avaliable time
[16:20] <edwardmoore> fair enough
[16:20] <LaurenceB> oh in test J the rudder was larger
[16:20] <edwardmoore> how about find a model aeroplane club and get someone to give you a tow?
[16:20] <LaurenceB> think that helps
[16:21] <LaurenceB> hmm good idea, wouldnt want it to go too far though
[16:21] <LaurenceB> could stick the motors back on, but havent got any spare props
[16:25] <LaurenceB> in tests H and I the servo was reaching its limits- set at +-30 degrees
[16:26] <LaurenceB> with a larger rudder that doesnt have to happen, the there is more control authority, and the response is more linear
[16:26] <LaurenceB> > main reason test J was better
[16:28] <LaurenceB> just looking at the servo data, in test J servo doesnt go beyond 20 degrees
[16:31] <LaurenceB> yay weather is improving like forecast said right more tests cya all
[16:32] LaurenceB (n=laurence@host81-132-17-240.range81-132.btcentralplus.com) left irc:
[17:57] <edwardmoore> jcoxon are you about?
[18:16] <jcoxon> i am now
[18:16] <edwardmoore> yo
[18:16] <edwardmoore> the excitement of new OSs!
[18:17] <edwardmoore> i haven't got to the stage of testing yet (still installing) but do you reckon the wireless should work without too much hassle?
[18:19] <jcoxon> shouldn't be difficu.t
[18:19] <jcoxon> difficult
[18:20] <edwardmoore> yippeee. it's yummy to rejuvinate the craptop
[18:20] <jcoxon> i think intels are quite supported as they appear on centrino laptios
[18:20] <jcoxon> laptops
[18:21] <edwardmoore> awesome. nearly installed
[18:21] flowolf (n=flowolf@host81-187-static.13-87-r.retail.telecomitalia.it) joined #highaltitude.
[18:21] <jcoxon> if there isn't a native driver check out NDISwrapper
[18:22] <jcoxon> very clever basically uses the windows driver
[18:22] <jcoxon> but first look for a native driver
[18:22] <jcoxon> a quick google will reveal the answer
[18:22] <jcoxon> right will bbl
[18:22] <edwardmoore> ok cya later
[19:06] Ebola (n=Ebola@host86-143-156-147.range86-143.btcentralplus.com) joined #highaltitude.
[19:10] edwardmoore (n=edwardmo@88.212.167.121) left irc:
[19:23] edwardmoore (n=edwardmo@88.212.167.121) joined #highaltitude.
[19:37] LaurenceB (n=laurence@host86-133-67-59.range86-133.btcentralplus.com) joined #highaltitude.
[19:37] <LaurenceB> Hi all
[19:38] <LaurenceB> just been doing glider tests
[19:39] <LaurenceB> think I've got the oscillation under control, but its vulnerable to gusts of wind
[19:40] <LaurenceB> the system takes 3 or 4 oscillations to stabilise, and oscillates at around 0.1Hz, around the frequency of gusts of wind
[19:43] <LaurenceB> thats the only explanation i can think of for why it was stable on some flights, but started to oscillate on others
[19:53] edmoore (n=ed@88.212.167.121) joined #highaltitude.
[19:54] <edmoore> hi edwardmoore
[19:54] <edwardmoore> hi edmoore
[19:54] <LaurenceB> its flying too far now for the test site, its going into a housing estate, crashed into someones house
[19:54] <LaurenceB> WTF?
[19:54] <edwardmoore> what OS are you using? I'm using OS X
[19:54] <LaurenceB> aha linux
[19:54] <edmoore> funny you ask, i'm using ubuntu
[19:55] <edmoore> i'm running on a craptop, you?
[19:55] <edwardmoore> a sexy macbook pro, me
[19:55] <edmoore> cocky git
[19:55] <LaurenceB> cacbook pro
[19:55] <edwardmoore> yeah, i know. but a brushed aluminium cocky git
[19:56] <edmoore> i'm going to be the LAMP box for the macbook pro to run remotely
[19:56] <edwardmoore> yes, you are. and my file server
[19:56] <LaurenceB> think i'll make a simulation of the glider to see if I can replicate the behaviour
[19:57] <LaurenceB> also thinking of "carrying forward" adding the rate of change of heading to the heading in an attempt to reduce the lag issue
[19:57] <edmoore> that's getting funky
[19:57] <edmoore> the parafoils are 9DOF systems. headache
[19:58] <edmoore> um...... isn't there a wee filter that helps with that?
[19:58] <LaurenceB> DOF=dimension of flight?
[19:58] <edmoore> a recursive state-space filter
[19:58] <edmoore> begins with a k.......
[19:58] <edmoore> damnit, forgotten!
[19:59] <edmoore> degrees of freedom
[20:01] <LaurenceB> Kalman filters only really work when your oversampling is my impression
[20:01] <LaurenceB> they smooth it better than a low pass filter
[20:01] <edmoore> and of course without phase lag
[20:02] <LaurenceB> will try and write a simulation and see if it can reproduce the behaviour
[20:04] <LaurenceB> wind is tricky as there is a torque at the start and then the effect of the addition of the wind vector
[20:04] <LaurenceB> will forget about that to start with and just start it at nonzero offset
[20:05] <edmoore> yup
[20:06] <LaurenceB> it seems that there is a limit to how well it can be stabilized against oscillation
[20:06] <LaurenceB> as the lag acts like a driving term
[20:07] <LaurenceB> hence the carry forward idea
[20:08] <LaurenceB> hey just thought, is there a way to make a digital notch filter to filter out the frequency of oscillation from the heading error datastream
[20:09] <edwardmoore> no t in that kinda noch
[20:10] <edwardmoore> and yes...... but i never understood the maths. ask me when i've done the control course!
[20:11] <edmoore> i think google will be your friend on this one
[20:11] <edwardmoore> agreed
[20:12] <LaurenceB> hmm to be responsive it would have to have a wide bandwidth, and so would interfere with things
[20:16] <LaurenceB> noch filter is effectivly an oscillator that oscillates 180 degrees out of phase, so it would have to be heavily damped to be responsive-large bandwith
[20:16] <LaurenceB> bbl
[20:53] flowolf (n=flowolf@host81-187-static.13-87-r.retail.telecomitalia.it) left irc: "Leaving"
[21:05] edmoore (n=ed@88.212.167.121) left irc: Read error: 104 (Connection reset by peer)
[21:14] edwardmoore (n=edwardmo@88.212.167.121) left irc:
[21:16] edwardmoore (n=edwardmo@88.212.167.121) joined #highaltitude.
[22:11] Ebola (n=Ebola@host86-143-156-147.range86-143.btcentralplus.com) left irc: ""Don't let worry kill you -- let the church help.""
[22:24] <LaurenceB> latest test series up on the wiki
[22:39] <LaurenceB> I'm off cya all
[22:39] LaurenceB (n=laurence@host86-133-67-59.range86-133.btcentralplus.com) left irc:
[00:00] --- Sat Mar 24 2007