highaltitude.log.20090131

[00:00] <ShellEvil> what's the L1/L2 freq diff?
[00:01] <Laurenceb> L2 is 1.26GHz
[00:01] <ShellEvil> Hmm.
[00:01] <ShellEvil> That's only 200MHz
[00:01] <Laurenceb> L1 is 1.52 or something..
[00:01] <Laurenceb> yeah about
[00:02] <ShellEvil> The difference beween the top and bottom of L1 is ~1% of that
[00:02] <ShellEvil> I don't know if you can extract meaningful information from that
[00:09] <Laurenceb> there are both made using an onboard pll
[00:09] <ShellEvil> ?
[00:12] <Laurenceb> on tha sat
[00:14] <ShellEvil> No - I mean - you have a ~2MHz wide signal from the sat coming through the IF filter. This will be somewhat smeared by ionospheric delay of the top relative to the bottom. You could in principle skew the inbound frequency spectrum till the recieved signal matched the transmitted - with no skew
[00:14] <ShellEvil> To get you a measure of the ionospheric delay - but only using L1
[00:15] <Laurenceb> theres too much noise to do that
[00:15] <ShellEvil> I can't think of a remotely computationally cheap way to do this though.
[00:15] <Laurenceb> better to use ground station data
[00:15] <ShellEvil> I don't mean over 1ms. More like tens or hundreds of seconds.
[00:15] <Laurenceb> theres ways to reciever L2 without the prn codes
[00:15] <ShellEvil> I know - but that requires more hardware
[00:16] <ShellEvil> (though I can't think of a computationally inexpensive way of doing this - so it may effectively require more CPU)
[00:17] <ShellEvil> http://focus.ti.com/dlpdmd/docs/dlpdiscovery.tsp?sectionId=60&tabId=2235 neat!
[00:21] <josepharmbruster> wow, that is cool :-)
[00:21] <josepharmbruster> http://www.youtube.com/watch?v=7UfarRM0BoM
[00:23] <ShellEvil> I'm trying to reboot my knowledge of opto units.
[00:23] <ShellEvil> 7lm picture.
[00:24] <ShellEvil> and how that compares to a laptop screen
[00:25] <Laurenceb> I dont follow what the msp430 does
[00:25] <ShellEvil> prolly just supervisory
[00:25] <Laurenceb> yeah
[00:25] <ShellEvil> Ok - eeepc does 100cd/m^2
[00:25] <ShellEvil> A cd is a lumen per steradian
[00:26] <Laurenceb> not bad then
[00:26] <Laurenceb> for laptop size images
[00:26] <Laurenceb> maybe up to say 20" on a wall
[00:26] <ShellEvil> so with an output angle of 4 steradians, that's ...
[00:26] <Laurenceb> but your nightvision can adjust well
[00:27] <ShellEvil> not that well - you want a rapid image
[00:27] <ShellEvil> and for the eye not to drop down much in colour rendition
[00:28] <ShellEvil> 100cd/m^2 is 400lm/m^2 I think at 4 steradians viewing angle.
[00:28] <ShellEvil> So that'd be 1/57th of a m^2
[00:28] <josepharmbruster> furgusnoble: you around?
[00:28] <ShellEvil> Or about 10*15cm.
[00:29] <ShellEvil> (at the same brightness as the eeepc, with a wide viewing angle)
[00:39] <Laurenceb> hmm
[00:39] <Laurenceb> I was just looking at residuals
[00:39] <Laurenceb> they went down to < 5m on all sats
[00:39] <Laurenceb> which is good
[00:39] <ShellEvil> Indeed.
[00:39] <ShellEvil> geom?
[00:39] <Laurenceb> I need to plot the residualts v time
[00:40] <Laurenceb> yes
[00:40] <Laurenceb> they seemed to be minial at one point
[00:41] <Laurenceb> which corresponded to an altitude about 10m higher
[00:41] <Laurenceb> than the avergae
[00:43] <Laurenceb> which corresponds to some thing being held at 1.5m altitude around that area
[00:43] <Laurenceb> looks good
[00:44] <Laurenceb> +-1m or so
[01:10] <ShellEvil> Can you think of any keywords for a graph of delay vs frequency range for the ionosphere?
[01:15] <ShellEvil> http://www.google.co.uk/url?sa=t&source=web&ct=res&cd=12&url=http%3A%2F%2Fwww.ncrs.org.au%2Fwars%2Fwars2008%2FRoberts%2520WARS%25202008.pdf&ei=1aWDScWzOeKtjAerlrTNAw&usg=AFQjCNFP2sj5r_dtgr9i1GrhfEM1tF0rHQ&sig2=0PMqtkb8VO4WP5y6-wYlsw
[01:15] <ShellEvil> interesting
[01:15] <ShellEvil> (inverse filter for reversing iono effects)
[01:16] <ShellEvil> (of course only one fixed delay)
[01:19] Action: ShellEvil stabs.
[01:19] <ShellEvil> Bastard stole my idea.
[01:19] <ShellEvil> http://www.patentgenius.com/patent/7477183.html
[01:41] Tigga (n=chatzill@pc-232-235-103.magd.cam.ac.uk) left irc: "ChatZilla 0.9.84 [Firefox 3.0.5/2008120122]"
[01:52] SpeedEvil (n=user@mauve.plus.com) left irc: Read error: 113 (No route to host)
[02:02] jiffe20 (n=jiffe@209.159.247.189) joined #highaltitude.
[02:10] SpeedEvil (n=user@tor/regular/SpeedEvil) joined #highaltitude.
[02:23] natrium42 (i=natrium4@CPE000625d867e2-CM0014045885be.cpe.net.cable.rogers.com) left irc: Remote closed the connection
[02:24] natrium42 (i=natrium4@CPE000625d867e2-CM0014045885be.cpe.net.cable.rogers.com) joined #highaltitude.
[02:41] borism_ (n=boris@195-50-201-33-dsl.krw.estpak.ee) joined #highaltitude.
[02:46] borism (n=boris@195-50-211-134-dsl.krw.estpak.ee) left irc: Read error: 145 (Connection timed out)
[02:48] <fergusnoble> hello
[02:48] <fergusnoble> anyone awake?
[02:48] <SpeedEvil> http://flickr.com/photos/84289229@N00/2903791610/
[02:48] <SpeedEvil> err
[02:48] <SpeedEvil> I mean yes
[02:48] <fergusnoble> try out the wind overlay on the tracker
[02:49] <fergusnoble> its a bit rubbish atm, but its just a proof of concept
[02:49] <fergusnoble> http://www.srcf.ucam.org/~cuspaceflight/predict/index.php
[02:49] <fergusnoble> sorry, predictor, not tracker
[02:51] <SpeedEvil> are you generating the wind arrows yourself?
[02:52] <fergusnoble> yeah
[02:52] <fergusnoble> its done in matplotlib in python
[02:52] <fergusnoble> its just the image is too small really to be spread across that area
[02:53] <fergusnoble> unfortunately the winds are quite constant across the area atm
[02:53] <fergusnoble> ill turn on normalization to make it look pretty, one sec
[02:54] <SpeedEvil> http://www.yr.no/place/United_Kingdom/Scotland/Glenrothes/hour_by_hour.html
[02:54] <SpeedEvil> I like the above style of wind arrows
[02:54] <SpeedEvil> roman numerals++
[02:54] <fergusnoble> try it now, you can see the color range more clearly
[02:55] <SpeedEvil> and I also find it silly that the best weather forcast I've found is from norway
[02:55] <fergusnoble> hehe
[02:55] <fergusnoble> yeah, the ticks on the arrows are nice
[02:56] josepharmbruster (n=josephar@45.144.119.70.cfl.res.rr.com) left irc: Read error: 60 (Operation timed out)
[02:58] <SpeedEvil> what does the colour range show?
[02:58] <fergusnoble> wind speed
[02:58] <SpeedEvil> ah - I see
[02:58] <SpeedEvil> neat
[02:59] <fergusnoble> its supposed to be like wunderground
[02:59] <fergusnoble> but needs a bit more work
[02:59] <fergusnoble> ok, bed for me
[02:59] <fergusnoble> night
[03:00] <SpeedEvil> looking promising
[03:00] <SpeedEvil> :)
[03:05] fergusnoble (n=fergusno@fn217.quns.cam.ac.uk) left irc:
[04:08] Nick change: hallam|zzz -> hallam
[04:24] <hallam> http://www.hobby-lobby.com/video-camera.htm
[04:24] <hallam> 20 grams
[04:28] <SpeedEvil> awaiting delivery of http://www.dealextreme.com/details.dx/sku.18088
[04:31] <hallam> oh now that is quite nice
[04:31] <SpeedEvil> there are even cheaper lower-res pens
[04:34] <hallam> any idea on weight once gutted?
[04:37] <SpeedEvil> I would guess well under 10g
[04:38] <SpeedEvil> It'll prolly be a week yet before I get it though
[06:02] natrium42_ (i=natrium4@CPE000625d867e2-CM0014045885be.cpe.net.cable.rogers.com) joined #highaltitude.
[06:21] natrium42 (i=natrium4@CPE000625d867e2-CM0014045885be.cpe.net.cable.rogers.com) left irc: Read error: 110 (Connection timed out)
[06:25] hallam (i=836fc8c8@gateway/web/ajax/mibbit.com/x-23f83ab5654ac713) left irc: "http://www.mibbit.com ajax IRC Client"
[06:43] Simon-MPFH (n=simon@phantom.mpfh.co.uk) joined #highaltitude.
[07:29] Simon-MPFH (n=simon@phantom.mpfh.co.uk) left irc: "Leaving"
[07:48] rjharrison (n=rharriso@80.176.172.227) left #highaltitude.
[08:04] rjharrison (n=rharriso@80.176.172.227) joined #highaltitude.
[08:29] <rjharrison> CUSF = no launch?
[09:02] hallam (i=836fc8c8@gateway/web/ajax/mibbit.com/x-81664b75fe92a090) joined #highaltitude.
[09:16] Hiena (n=Hiena@81.93.195.181.datatrans.hu) joined #highaltitude.
[09:25] <Laurenceb> good morning habbers
[09:26] <hallam> hello
[09:27] <Laurenceb> http://maps.google.co.uk/maps?f=q&source=s_q&hl=en&geocode=&q=52+12%27+7.57%27%27+N+0+7%27+11.09%27%27+E&sll=52.202117,0.119741&sspn=0.000186,0.000603&ie=UTF8&ll=52.202128,0.119808&spn=0.000186,0.000603&t=h&z=21&iwloc=addr
[09:27] <hallam> why do you always use minutes and seconds?
[09:29] <hallam> got about 15m of error on that
[09:29] <Laurenceb> hmf
[09:31] <Laurenceb> the residuals are small
[09:31] <Laurenceb> <5m
[09:32] <hallam> probably iono then
[09:33] jcoxon (n=jcoxon@host86-158-31-172.range86-158.btcentralplus.com) joined #highaltitude.
[09:33] <Laurenceb> I'll try again with looser dll
[09:34] <jcoxon> morning all
[09:34] <hallam> Hi James
[09:34] <Laurenceb> hello moxon jcoxon
[09:35] mc- (n=mfcastle@cpc3-glfd1-0-0-cust439.glfd.cable.ntl.com) joined #highaltitude.
[09:35] <jcoxon> hey hey
[09:36] <mc-> hi jcoxon
[09:36] <Laurenceb> hi mc-
[09:36] <mc-> been reading about your competition
[09:36] <mc-> hi lb
[09:36] <jcoxon> mc-, yeah, just to kick things along
[09:36] <jcoxon> though only one team right now
[09:37] <mc-> is it allowed to tx 0.5W on 169MHz ?
[09:37] <jcoxon> the uplink yeah
[09:37] <mc-> can't see that in the IR2030.pdf
[09:38] <mc-> apart from assistive listening devices
[09:38] <mc-> ie hearing aids
[09:39] <mc-> found it now
[09:39] <jcoxon> mc-, page 13
[09:40] <mc-> I've got a 458MHz 0.5W tx, so I prefer to use that.
[09:40] <mc-> thanks
[09:40] <jcoxon> we thought you might get more range at 169
[09:40] <mc-> yes, in theory
[09:40] <jcoxon> interesting flight profile today
[09:41] <hallam> yeah, pity nobody's flying
[09:41] <jcoxon> :-(
[09:41] <jcoxon> i was already for a bit of tracking
[09:41] <hallam> well if CUSF are flying they haven't told me
[09:41] edmoore (n=edmoore@pomegranate.chu.cam.ac.uk) joined #highaltitude.
[09:41] <hallam> what ho, E
[09:41] <hallam> d
[09:42] <edmoore> you can never make me decide on a ho tis early in the morning, hallam
[09:42] <edmoore> it's a decision that takes time and consideration
[09:43] <hallam> That's pretty good advice
[09:44] <edmoore> i cannot breathe through my nose. Good god I hate being ill. Imagine if I actually got actually ill? I'm not sure I could deal with it
[09:44] <edmoore> jcoxon: ping
[09:45] <hallam> edmoore: I've been unable to breathe through my nose for the last four years. Finally getting surgery in April with a bit of luck
[09:45] <edmoore> then you can sing n aria too us
[09:45] <edmoore> well, at least your conditions isn't accompanied by random snotty dribble
[09:46] <edmoore> let's see what's happening on the repeater lets
[09:46] <Hiena> http://www.guardian.co.uk/media/pda/2009/jan/30/google-digitalmedia
[09:46] Action: Laurenceb shoves a drill up hallams nose
[09:47] <Hiena> Laurenceb, which diameter?
[09:47] <jcoxon> hey edmoore pong
[09:47] <hallam> why do I keep misplacing useful things like my DC-daylight scanner?
[09:48] <edmoore> hallam: I might have to recover my soldering/hot air station from you soon, if I end up doing some soldering for someone
[09:48] <edmoore> just a heads-up
[09:48] <edmoore> jcoxon: when the cold gets to the dribbly runny snot stage, does that mean the worst is over?
[09:48] <hallam> sure, whenever you like
[09:49] <jcoxon> edmoore, apparently so - though it varys between people and the various cold viruses
[09:49] <edmoore> ok. 5 days of this is just too much
[09:49] <jcoxon> reckon more variation between people rather then virus though
[09:49] <jcoxon> you'd probably get teh same immune reaction to each virus
[09:50] <edmoore> it's been getting 1 berocca the 4 lemsips a day since teusday. it really ought to do the decent thing and diw
[09:50] <edmoore> die*
[09:50] <jcoxon> hehe
[09:50] <jcoxon> got to respect the genius behind viruses
[09:50] <edmoore> you're sounding a little intelligent design there
[09:51] <jcoxon> nah, still a big fan of darwin
[09:51] <hallam> edmoore: considered amputation?
[09:51] <edmoore> of the nose?
[09:51] <hallam> or below the neck
[09:52] <edmoore> that wuld probably do the trick
[09:55] <jcoxon> edmoore, just think that you'll never get that cold virus again
[09:55] <jcoxon> one less to get next year :-p
[09:56] <edmoore> I take comfort from that
[09:56] <edmoore> thanks
[09:57] <jcoxon> ooo bill brown launch today, on 14.077
[09:57] <jcoxon> http://home.hiwaay.net/~bbrown/
[09:57] <edmoore> oh *awesome*
[09:57] <edmoore> what time gmt?
[09:57] <edmoore> the guy is a legend, but he could do with a refresher course on site design
[09:58] <jcoxon> -6 i think
[09:58] <edmoore> hrm, 14mhz will probably be closed for us by then
[09:59] <jcoxon> natrium42_, might get something
[09:59] <jcoxon> i think he was planning to listen last week before it was postponed
[09:59] <hallam> Any of you chaps know where I can find a copy of sx (xmodem send program) for windows, or sources?
[09:59] <Laurenceb> I dont follow why you get ionospheric paths at night
[10:00] <Laurenceb> the electron density is higher during the day
[10:00] <edmoore> Laurenceb: i think it depends on the frequency
[10:01] <edmoore> it's not true for all frequencies, but by received wisdom, it is true for 20m
[10:01] <edmoore> from what I've read anyway
[10:01] <hallam> maybe a lower layer opens up allowing it through to a higher layer?
[10:01] <jcoxon> hallam, it'll be in some gnu package
[10:01] <Laurenceb> hmm yeah
[10:01] <Laurenceb> probably that
[10:01] <edmoore> i read an explanation once but I didn't absorb in thoroughly enough to embarass myself with an explanation
[10:02] <jcoxon> hallam, how about lrzsz
[10:02] <jcoxon> that'll do the job, i've used it on linux before
[10:03] <hallam> that sounds like an eastern european capital
[10:03] <hallam> thanks, I'll check it out
[10:06] <hallam> brb
[10:09] hallam (i=836fc8c8@gateway/web/ajax/mibbit.com/x-81664b75fe92a090) left irc: "http://www.mibbit.com ajax IRC Client"
[10:14] hallam (i=836fc8c8@gateway/web/ajax/mibbit.com/x-9f8ab6bb2299f778) joined #highaltitude.
[10:14] <rjharrison> Shame about no launch!
[10:14] <rjharrison> Hopefully edmoore will be virus free for the next launch
[10:15] <rjharrison> If next w/e is looking good I think I may be up for a lunch then
[10:15] <edmoore> I hope so. this is crap
[10:15] <edmoore> not been ill for years
[10:15] <rjharrison> Man flue its the worst
[10:15] <edmoore> simultaneous launches would be fun
[10:15] <rjharrison> flu
[10:15] <rjharrison> Yep you can go first and i can see what the wind is doing
[10:15] <rjharrison> Perfect
[10:15] <rjharrison> :)
[10:16] <jcoxon> bbl
[10:16] jcoxon (n=jcoxon@host86-158-31-172.range86-158.btcentralplus.com) left irc: "Leaving"
[10:16] <Laurenceb> hallam: I get a very similar point with looser loops
[10:16] <rjharrison> mc- did you get your grmmets
[10:16] <rjharrison> grommets
[10:16] <Laurenceb> but the residuals are closer to 10m
[10:16] <Laurenceb> looks like its time to put in ionosphere
[10:17] <Laurenceb> what time did you do this?
[10:24] <mc-> yes, thanks got the grommets
[10:25] <mc-> was looking at your cutdown pics
[10:25] <Laurenceb> hallam: what time of day was this recorded?
[10:26] <mc-> can you use a few loops of plastic, so you cut through one loop with a thin piece of plastic?
[10:27] <hallam> Laurenceb: you should be able to get that from the GPS data ;)
[10:28] <hallam> but I think it was around 5:30am
[10:28] <mc-> I cut through some fishing line with a small resistor, it took a few secs
[10:28] <Laurenceb> ok...
[10:28] <Laurenceb> hmm there shouldnt have been much ionosphere then
[10:29] <hallam> maybe less than usual?
[10:29] <hallam> or some for one satellite but not for others
[10:29] <hallam> some of the beams would be going through daylit parts of the atmosphere
[10:31] Xenion (n=robert@p579FCD56.dip.t-dialin.net) joined #highaltitude.
[10:31] <Laurenceb> http://www.kn.nz.dlr.de/daily/tec-eu/http://www.kn.nz.dlr.de/daily/tec-eu/
[10:31] <Xenion> Guten Morgen :-)
[10:32] <Laurenceb> hi
[10:32] <Laurenceb> hallam: looks quite small
[10:32] <hallam> said the actress to the bishop
[10:32] <Laurenceb> 3m max on 21
[10:33] <Laurenceb> :)
[10:33] <hallam> would dop have amplified that?
[10:33] <Laurenceb> hmm
[10:33] <Laurenceb> maybe
[10:33] <hallam> back in a bit, uninstalling friggin' AVG because it keeps identifying my blackfin cross-compiler as a virus
[10:33] hallam (i=836fc8c8@gateway/web/ajax/mibbit.com/x-9f8ab6bb2299f778) left irc: "http://www.mibbit.com ajax IRC Client"
[10:33] <Laurenceb> but it doesnt explain why the residuals were around 10m
[10:33] <Laurenceb> cya
[10:34] Tigga (n=chatzill@pc-232-235-103.magd.cam.ac.uk) joined #highaltitude.
[10:34] hallam (i=836fc8c8@gateway/web/ajax/mibbit.com/x-f8143c1eaa3fc169) joined #highaltitude.
[10:34] edmoore (n=edmoore@pomegranate.chu.cam.ac.uk) left irc:
[10:34] <Laurenceb> the behaviour is a bit odd
[10:35] <Laurenceb> its finding the least squares solutuion
[10:35] <Laurenceb> so I guess it can be made to wanted off by smaller errors
[10:35] <Laurenceb> *wander
[10:35] <Laurenceb> so maybe 10m residuals are to be expected
[10:36] <Laurenceb> it was only on two sats, the others were <1m
[10:36] <Laurenceb> maybe ionospheric would get it down to a few m error on all, which would be sensible
[10:40] <Laurenceb> I've got 0.2 to 2m delays from that map
[10:40] <Laurenceb> - guesstimated
[10:55] mc- (n=mfcastle@cpc3-glfd1-0-0-cust439.glfd.cable.ntl.com) left irc:
[10:58] <hallam> man cygwin is kind of annoying sometimes
[11:16] jcoxon (n=jcoxon@host86-158-31-172.range86-158.btcentralplus.com) joined #highaltitude.
[11:30] <Laurenceb> hallam: putting in the ionosphere has only reduced the residuals by about 1m
[11:30] <Laurenceb> I'm about 2m to the east with the position
[11:33] <Laurenceb> looks like the low elevation ones are having too large an effect
[11:35] <Laurenceb> I think theres multipath on 2
[11:37] <hallam> going the right direction
[11:37] <hallam> could be multipath I guess, there are buildings
[11:38] <Laurenceb> yeah
[11:38] <Laurenceb> looking at the map there would be
[11:38] <Laurenceb> its very noisy
[11:38] <Laurenceb> its quite strong on the acquisition
[11:39] <Laurenceb> but then the tracking fades in and out
[11:39] <Laurenceb> thing is the residual on 2 is like 10cm
[11:39] <Laurenceb> so I think 2 is throwing it off
[11:40] <rjharrison> Off to 2e0
[11:40] <Laurenceb> this is where kalman filters help
[11:44] <Laurenceb> http://s1.b3ta.com/host/creative/37069/1233242272/GREATESCAPE.jpg
[11:48] mc- (n=mfcastle@cpc3-glfd1-0-0-cust439.glfd.cable.ntl.com) joined #highaltitude.
[11:51] <mc-> Laurenceb, I've got a GPS qn for you.
[11:51] <Laurenceb> go on then
[11:52] <mc-> can you do correlation before the mixing down to baseband?
[11:53] <Laurenceb> guess you could
[11:53] <Laurenceb> but I dont know why youd want to
[11:53] <Laurenceb> all horrible high frequency stuff
[11:53] <mc-> then you can search the code space with a low IF
[11:53] <mc-> and do a FFT to find the peaks
[11:53] <Laurenceb> hmm
[11:54] <Laurenceb> I'm not really following
[11:54] <Laurenceb> what would you mix with before going down to baseband?
[11:54] <Laurenceb> hallam: wow residuals are really low now
[11:54] <Laurenceb> oh
[11:54] <Laurenceb> :-/
[11:54] <Laurenceb> of course as theres only 4 sats
[11:54] <ShellEvil> mc-: you can in principle correlate at any frequency
[11:55] <mc-> ie do correlation first, and then detect peaks in the frequency domain
[11:55] <Laurenceb> what do you mean by correlation
[11:55] <Laurenceb> multiplication by prn code?
[11:55] <ShellEvil> mc-: for example, you could multiply the incoming sat signal with the PRN, and get a carrier out.
[11:55] <mc-> yes
[11:55] <Laurenceb> ok
[11:55] <ShellEvil> mc-: at 1.5GHz
[11:55] <Laurenceb> but you'd need to keep the prn syncronised
[11:55] <mc-> what's the conventional way of doing it?
[11:56] <Laurenceb> how does this help
[11:56] <Laurenceb> mc-: get an SE4120
[11:56] <Laurenceb> downconverts straight to baseband
[11:56] <Laurenceb> the 4120 has zero IF
[11:56] <mc-> there's always some freq offset due to doppler
[11:57] <Laurenceb> - direct downconversion
[11:57] <Laurenceb> sure
[11:57] <mc-> so I was wondering about doing correlation first, then detecting peaks in the f domain
[11:57] <mc-> or maybe that is how it is done normally
[11:57] <Laurenceb> nope
[11:57] <Laurenceb> you downconvert first
[11:58] <Laurenceb> then do all the processing
[11:58] <mc-> so you search in the freq domain to find a peak, and then search in phase?
[11:58] <Laurenceb> no, you have to search in both
[11:59] <mc-> yes, sorry, that's what I meant
[11:59] <mc-> how do you search in the freq domain?
[12:00] <Laurenceb> local oscillator
[12:00] <Laurenceb> i.e. multiplication digitallyu
[12:00] <mc-> so in effect you try lots of different LO freqs, until you find a peak?
[12:00] <Laurenceb> yes
[12:01] <Laurenceb> about 40
[12:01] <Laurenceb> get yourself kai borres boor
[12:01] <Laurenceb> book
[12:01] <mc-> and you have to search the code space as well for each LO freq?
[12:01] <ShellEvil> Or you use FFT, or you use a FPGA correlator
[12:01] <ShellEvil> mc-: yes
[12:02] <ShellEvil> mc-: barring being _really_ clever
[12:02] <ShellEvil> mc-: there are ways you can search both code and frequency space at the same time. They tend to be much, much more complex and maybe on some hardware actually slower
[12:03] <mc-> can you do correlation first, and then search the frequency space using a FFT?
[12:03] <hallam> you can pretty easily do a parallel search in either code or freq space though, take your pick
[12:03] <hallam> mc - it's not really called correlation at that point
[12:04] <hallam> you can certainly multiply with the codes and then parallel-search frequency with FFT
[12:04] <hallam> or you can multiply with a few different LO frequencies, and then parallel-search code with IFFT
[12:04] <Laurenceb> hallam: due to the geometery I'm wandering north south by 30m
[12:04] <Laurenceb> but east is very stable
[12:04] <Laurenceb> I'll try and find the average
[12:04] <hallam> Laurenceb: why only 4 sats? I tracked 5 or 6
[12:05] <Laurenceb> to avoid multipath
[12:05] <hallam> oh, you rejected the ones you think are bad?
[12:05] <Laurenceb> yeah
[12:05] <hallam> makes sense
[12:05] <mc-> I know there's some clever ways of doing an FFT without needing all the FFT hardware.
[12:06] <ShellEvil> There are about eleventy-one clever schemes.
[12:07] <ShellEvil> Some of these will result in great speedups on some hardware, but on other hardware it'll slowdown.
[12:07] <mc-> exactly
[12:08] <ShellEvil> Laurenceb: did you get that note about the patent I linked you?
[12:08] <Laurenceb> nope
[12:09] <ShellEvil> Someone detecting ionospheric slope by looking at differences between upper and lower sidebands of a signal.
[12:09] <ShellEvil> :/
[12:09] <ShellEvil> I'm not sure if I actually mentioned that idea here before the patent was granted.
[12:17] <Laurenceb> http://maps.google.co.uk/maps?f=q&source=s_q&hl=en&geocode=&q=52+12%27+7.150%27%27+N+0+7%27+11.6941%27%27E&sll=52.201794,0.120286&sspn=0.001486,0.004828&ie=UTF8&ll=52.201886,0.120077&spn=0.001486,0.004828&t=h&z=18&iwloc=addr
[12:17] <Laurenceb> I see
[12:17] <Laurenceb> hallam: thats about as good as I can get
[12:18] <Laurenceb> the geometery really screws things un in the north south direction
[12:18] <hallam> that's pretty good
[12:18] <hallam> it was about 5m NE of there
[12:19] <Laurenceb> nice
[12:19] <Laurenceb> pity about the multipath
[12:19] <Laurenceb> it needs the sats to be weighted
[12:20] <Laurenceb> then hopefully you'd strike the best balance between DOP and reliance on poor sats
[12:22] <Laurenceb> interesting that your window gave better results
[12:22] <Laurenceb> must be multipath
[12:22] <Laurenceb> I might try with different DLL widths
[12:23] <Laurenceb> a narrower one will screen out multipath better
[12:23] <ShellEvil> Will it?
[12:23] <Laurenceb> yep
[12:23] <Laurenceb> as in less difference between early and late
[12:23] <ShellEvil> Surely if your multipath delay is shorter than 300m/4 or so, you can't really resolve it
[12:23] <Laurenceb> yes
[12:24] <Laurenceb> you can use a dithered early late
[12:24] <Laurenceb> then actually map out the multipath
[12:24] <ShellEvil> Umm.
[12:24] <Laurenceb> and find the best solution for the direct path
[12:24] <ShellEvil> That sounds very noisy.
[12:24] <ShellEvil> Also.
[12:24] <Laurenceb> yeah you have to do it over a second or so
[12:24] <Laurenceb> by averaging
[12:24] <ShellEvil> I'm not sure that works.
[12:25] <Laurenceb> fastrax do that on their Se4120/arm9 solution
[12:25] <ShellEvil> Add to the signal, .5 of the signal delayed by 100ns.
[12:25] <Laurenceb> ShellEvil: deconvolution
[12:25] <ShellEvil> The pulse shape will differ, yes.
[12:25] <Laurenceb> you multipath function is convolved with a triangle function
[12:26] <Laurenceb> you can derive the multipath function
[12:26] <ShellEvil> Hmm.
[12:26] <Laurenceb> then find the point where it starts to rise out of the noise floor
[12:26] <Laurenceb> thats the direct path
[12:26] <Laurenceb> you do this in a background task
[12:26] <ShellEvil> Perhaps. Sounds plausible.
[12:26] <Laurenceb> and use it do adjust the DLL
[12:26] <Laurenceb> it works at 0.5Hz or so
[12:27] <ShellEvil> I suppose multipath is actually good for noise.
[12:27] <ShellEvil> Actually no - it can be opposite phase
[12:27] <Laurenceb> also you do kalman filtering based on the singal amplitudes and results from the multipath analysis ect
[12:27] <Laurenceb> 6 component state vector - using doppler for the velocity
[12:28] <Laurenceb> all these techniques get the performance way up in these sort of scituations
[12:28] <Laurenceb> I'd imagine you could get it down to 5m error just with kalman
[12:29] <Laurenceb> multipath analysis enables you to give low elevation sats lower error values in the covariance matrix
[12:31] fergusnoble (n=fergusno@fn217.quns.cam.ac.uk) joined #highaltitude.
[12:31] <fergusnoble> hello
[12:31] <Laurenceb> SheelEvil: every so often you dither the early-late to some other value
[12:31] <Laurenceb> hi fergus
[12:31] <Laurenceb> say every 10th ms you dither it
[12:31] <ShellEvil> I see how it works in principle
[12:32] <Laurenceb> then you try about 20 dithered positions
[12:32] <Laurenceb> and average 10
[12:32] <Laurenceb> giving 0.5Hz multipath map
[12:32] <ShellEvil> The reason I'm wondering is it seems very, very sensitive to the least non-linearity or rounding error, or noise.
[12:32] <Laurenceb> well noise is averaged out
[12:33] <Laurenceb> but yeah you have to consider rounding ect a lot
[12:33] <ShellEvil> And I'm just a bit surprised you can pick out a sine wave from two superimposed sinewaves when the seperation is a degree or five.
[12:34] <Laurenceb> its from the prn
[12:34] <Laurenceb> not the carrier
[12:34] <Laurenceb> it picks up sources of multipath on the order of 10m or more
[12:34] <ShellEvil> Yes - but the fundamental input signal is a BPSK ~1MHz signal
[12:35] <Laurenceb> fastrax have some graphs from driving past multistory buildings
[12:35] <ShellEvil> that's where I was meaning the superposed sinewaves come in
[12:36] <Laurenceb> wel anyway I'm off
[12:36] <Laurenceb> cya all
[12:36] <ShellEvil> See you
[12:36] <ShellEvil> wave
[12:36] <ShellEvil> interesting.
[12:41] Laurenceb (n=laurence@dyres221-140.surrey.ac.uk) left irc: Read error: 104 (Connection reset by peer)
[13:12] edmoore (n=edmoore@pomegranate.chu.cam.ac.uk) joined #highaltitude.
[13:34] <hallam> hi edmoore, fergusnoble
[13:34] <fergusnoble> hallam: hello
[13:34] <edmoore> hi
[13:34] <fergusnoble> edmoore: want to go look at the rotator?
[13:34] <edmoore> anyone want to take a look at the rotator in a bit?
[13:34] <hallam> haha
[13:35] <hallam> sure, why not
[13:35] <edmoore> i guess that answers that one yet
[13:35] <fergusnoble> edmoore: btw, check the predictor, added image overlay
[13:35] <fergusnoble> its a bit rough and ready atm, but its just a proof of concept
[13:35] <hallam> I saw it, very nice
[13:35] <edmoore> cool, yeah
[13:35] <edmoore> definitely prooves concept
[13:35] <hallam> I've started making early efforts to put some GPS stuff on the blackfin
[13:36] <fergusnoble> oh awesome
[13:36] <edmoore> I am going to have a quick soak then i'll come back on and say if people want to go and hceck it out
[13:36] <fergusnoble> what happened to stopping once youve got a fix :p
[13:36] <edmoore> say in about 3/4 of an hr
[13:36] <hallam> I did stop, then I started again
[13:36] <edmoore> glad we didn't launch today
[13:36] <fergusnoble> ok, i need to do a couple of things too
[13:36] <hallam> ok
[13:37] <edmoore> I don't fancy driving to wales
[13:37] <hallam> haha
[13:37] <fergusnoble> well, a trip to bristol would have been nice
[13:37] <hallam> it's rather a trek
[13:37] <hallam> wouldn't it be ironic if it landed in the Irish sea
[13:37] <fergusnoble> hehe
[13:37] <edmoore> why ironic?
[13:38] <hallam> since it usually goes the other way and we pray for it to go west
[13:38] <edmoore> ah I seeee
[13:38] Action: hallam is debating trying to use uClinux
[13:38] <hallam> the bare-metal firmware is really annoying to program for
[13:38] <hallam> especially anything that uses maths
[13:38] <edmoore> loads of people seem to say that for the bigger micros
[13:39] <hallam> I can't get any of the libcs to link properly
[13:39] <hallam> but I really don't know how much overhead there is in uClinux
[13:39] <hallam> potentially quite a lot, I guess
[13:40] <edmoore> #avr32 seem to be quite good at uCLinux
[13:41] mc- (n=mfcastle@cpc3-glfd1-0-0-cust439.glfd.cable.ntl.com) left irc:
[13:41] <hallam> the support for this dev board is a bit crap, as far as I can tell
[13:41] <hallam> they only got the camera working a few weeks ago
[13:42] <hallam> and it's limited to 115200 for the console and firmware upload, which really bites
[13:42] <hallam> unless you have JTAG
[13:42] <hallam> but the JTAG requires a particular brand of JTAG
[13:43] <edmoore> of course
[13:43] <edmoore> apparently inventek have some sirfIII units modded to work up to 135k ft
[13:44] <hallam> interesting
[13:44] <fergusnoble> hallam: maybe use something like newlib but i wouldnt bother with uClinux
[13:45] <hallam> fergusnoble: would you be willing to give me a hand wrangling blackfin-gcc?
[13:45] <fergusnoble> it will eat a lot of time
[13:45] <fergusnoble> yeah sure
[13:45] <hallam> cool
[13:45] <edmoore> i think the bare metal gcc toolchain for blackfin comes with newlib anyway? I know someone said it did somewhere...
[13:45] <fergusnoble> i mean, there are other os's that would probably work but you need realtimeness
[13:45] <fergusnoble> and realtime linux isnt as realtime as the name suggests
[13:46] <fergusnoble> edmoore: yeah i wouldnt be suprised
[13:46] <hallam> edmoore: it does, but the srv build doesn't use it
[13:46] <fergusnoble> hallam: is the problem the calls to standard c functions?
[13:46] <hallam> yeah I have two major issues at the moment
[13:47] <hallam> one is I can't get newlib to link, so I don't have any maths functions
[13:47] <hallam> the other totally separate issue is that I can't get the cache to work
[13:47] <hallam> they just don't use the cache in the existing firmware, which is kind of suicidal for performance
[13:47] <fergusnoble> yeah, sucks
[13:47] <fergusnoble> ok, we can have a go after antenna fun
[13:48] <hallam> cool
[13:48] <hallam> I think I'm going to get a sandwich from smiley's
[13:48] <edmoore> indeed, infact getting the cache bit sorted was the main reason I didn't want to spend ages on the ARM9 board
[13:48] <fergusnoble> yeah, im going to take a shower, brb
[13:48] <hallam> means you have to figure out what the hell a cacheability protection translation lookaside buffer table is
[13:49] <fergusnoble> hallam: oh thats all fun stuff
[13:49] <hallam> I did do a quick test and the existing JPEG compressor, without cache, can manage just over 6 fps at 640x480
[13:49] <fergusnoble> hallam: ill bring my op sys notes from last year, it describes TLBs quite well
[13:49] <hallam> that's not including sending them over uart or anything
[13:49] <hallam> cool
[13:50] <fergusnoble> is basically how to map it into the virtual memory space
[13:50] <hallam> yeah I've done some reading
[13:50] <hallam> I also wrote the core loop of the gps tracker in blackfin assembler, haven't tried to assemble it yet or anything
[13:50] <fergusnoble> ok neat
[13:50] <hallam> but 10MHz / channel is very achievable I think
[13:50] <fergusnoble> is the nav stuff in c now?
[13:50] <hallam> partly
[13:51] <hallam> it extracts nav bits but doesn't decode them into ephemeris
[13:51] <hallam> mainly because I CBA to figure out how the parity system works
[13:51] <hallam> I'll do more on that later
[13:52] <hallam> anyway back in a few
[13:52] Nick change: hallam -> hallam|sandwich
[14:04] Simon-MPFH (n=simon@phantom.mpfh.co.uk) joined #highaltitude.
[14:11] <edmoore> hallam|sandwich: TLBs were the bread and butter of 3F5
[14:11] <edmoore> am cool with them now :)
[14:12] <edmoore> the lab was on chaching schemes and writing cache-friendly software. Useful stuff
[14:24] Nick change: hallam|sandwich -> hallam
[14:24] <hallam> edmoore: Nice. I'll grill you on the way to the rotator then
[14:26] <fergusnoble> edmoore: when are you coming?
[14:26] <edmoore> fergusnoble: you have an annoying habit of asking the question that i'm typing the answer to simultaneously
[14:26] <edmoore> :p
[14:26] <edmoore> now?
[14:27] <fergusnoble> yeah, sounds good to me
[14:27] <edmoore> meet you and hallam by the bollards?
[14:27] <fergusnoble> hallam: shall i meet you by nadias? need to get a snadwich myself
[14:27] <fergusnoble> yeah, sounds good
[14:27] <edmoore> ok, well what time?
[14:27] <edmoore> give me a time to be at the bollards and I will see you there
[14:27] <edmoore> so long as it's > 10 mins
[14:28] <fergusnoble> now is fine
[14:28] <fergusnoble> nadias is on the way
[14:28] <fergusnoble> hallam: that ok with you?
[14:29] <hallam> ok sure
[14:29] <hallam> bollards at 1445
[14:29] <edmoore> ok, 10 mins
[14:29] Nick change: hallam -> hallam|azel
[14:30] <fergusnoble> ok, see you in a sec
[14:33] edmoore (n=edmoore@pomegranate.chu.cam.ac.uk) left irc:
[14:34] hallam|azel (i=836fc8c8@gateway/web/ajax/mibbit.com/x-f8143c1eaa3fc169) left irc: "http://www.mibbit.com ajax IRC Client"
[14:59] <Tigga> is it just me.... or has google gone mental?
[15:10] josepharmbruster (n=josephar@45.144.119.70.cfl.res.rr.com) joined #highaltitude.
[15:13] <Hiena> Nope
[15:14] <Hiena> It's a reality
[15:14] <Hiena> Guess they has a main database falure
[15:46] edmoore (n=edmoore@pomegranate.chu.cam.ac.uk) joined #highaltitude.
[15:52] fergusnoble (n=fergusno@fn217.quns.cam.ac.uk) left irc:
[15:56] <ShellEvil> ?
[15:59] edmoore (n=edmoore@pomegranate.chu.cam.ac.uk) left irc:
[16:30] Laurenceb (n=laurence@dyres221-140.surrey.ac.uk) joined #highaltitude.
[16:32] josepharmbruster (n=josephar@45.144.119.70.cfl.res.rr.com) left irc: Read error: 113 (No route to host)
[16:46] Action: Laurenceb isd making a moxon antenna
[16:52] <ShellEvil> For GPS?
[16:52] Action: ShellEvil passes Laurenceb 30cm of wire.
[16:55] <Laurenceb> no 434
[16:56] <ShellEvil> handheld?
[17:06] Hiena (n=Hiena@81.93.195.181.datatrans.hu) left irc: "-=Halt! Hammerzeit!=-"
[17:15] <Laurenceb> yeah
[17:15] <Laurenceb> think its buildable with one pizza base
[17:15] Simon-MPFH (n=simon@phantom.mpfh.co.uk) left irc: "Leaving"
[17:16] <SpeedEvil> There is one important question you must answer first though.
[17:16] <SpeedEvil> What to have on the pizza.
[17:18] <Laurenceb> hmm annoyingly its a bit too big
[17:18] <Laurenceb> polystyrene pizza base
[17:18] <Laurenceb> they come in the bottom of the packs
[17:19] josepharmbruster (n=josephar@45.144.119.70.cfl.res.rr.com) joined #highaltitude.
[17:24] <SpeedEvil> You clearly need a larger pizza.
[17:25] <Laurenceb> yeah
[17:25] <Laurenceb> my largest pizza base is 4mm too small
[17:27] <jcoxon> a pizza base?
[17:28] <Laurenceb> I'm making a mnoxon antenna
[17:29] <jcoxon> still don't get the pizza base bit
[17:29] <Laurenceb> as a frame
[17:31] <jcoxon> fair enough
[17:31] <jcoxon> get some coax in the end then?
[17:31] <Laurenceb> yeah
[17:31] <Laurenceb> maplins
[17:32] <jcoxon> cool, i'm still suprised how good my moxon is
[17:32] <jcoxon> that said there isn't much on 70cm
[17:34] <Laurenceb> right glue gun time
[17:34] <Laurenceb> I'm going to stick it on the end of some dowel
[17:36] <Laurenceb> does the coax need a balun?
[17:36] <Laurenceb> i.e. do I need to coil the coax
[17:37] <jcoxon> no i don't think you do
[17:37] <SpeedEvil> gluegun on polystyrene - be careful, it may melt it
[17:38] <Laurenceb> thats the plan
[17:39] <Laurenceb> you have to be careful
[17:39] <Laurenceb> but it melts in and gets a good bond
[17:39] <SpeedEvil> copolymers FTW.
[17:42] josepharmbruster (n=josephar@45.144.119.70.cfl.res.rr.com) left irc: Read error: 113 (No route to host)
[17:52] <rjharrison> 2e0 done
[17:54] borism_ (n=boris@195-50-201-33-dsl.krw.estpak.ee) got netsplit.
[17:54] Ebola (i=ebola@unaffiliated/ebola) got netsplit.
[17:54] gordonjcp (n=gordonjc@symmetria.fi) got netsplit.
[17:55] <Laurenceb> 2e0 ?!?
[17:55] Ebola (i=ebola@goatse.co.uk) joined #highaltitude.
[17:56] gordonjcp (n=gordonjc@symmetria.fi) returned to #highaltitude.
[17:56] Nick change: Ebola -> Guest40334
[18:05] borism_ (n=boris@195-50-201-33-dsl.krw.estpak.ee) got lost in the net-split.
[18:12] borism (n=boris@195-50-201-33-dsl.krw.estpak.ee) joined #highaltitude.
[18:19] <Laurenceb> finished :P
[18:24] <rjharrison> Radio exam
[18:26] <Laurenceb> ah
[18:26] <Laurenceb> how did it go?
[18:28] <Laurenceb> jcoxon: ok, what can I find on 434?
[18:36] wickerwaka (n=akawaka@66-215-97-198.dhcp.malb.ca.charter.com) joined #highaltitude.
[18:41] <jcoxon> ummm not much
[18:41] <Laurenceb> wow from the noise it seems to be spot on
[18:41] <jcoxon> but you can use the antenna for all of the 70cm band
[18:41] <Laurenceb> yeah
[18:41] <Laurenceb> the noise goes down either side of 434
[18:41] <jcoxon> so http://www.charlescoverley.co.uk/ham/repeaters.htm
[18:42] <jcoxon> use this to find your nearest repeater and tune to that and listen in
[18:42] <Laurenceb> also vertical polarization has more noise
[18:43] <Laurenceb> thanks
[18:44] <Laurenceb> these are fm ?
[18:44] <jcoxon> yeah
[18:56] jcoxon (n=jcoxon@host86-158-31-172.range86-158.btcentralplus.com) left irc: "Leaving"
[19:06] wickerwaka (n=akawaka@66-215-97-198.dhcp.malb.ca.charter.com) left irc: Read error: 110 (Connection timed out)
[19:14] Nick change: Guest40334 -> Ebola
[19:15] Nick change: Ebola -> Guest26295
[19:16] Nick change: Guest26295 -> Ebola
[19:19] EI5GTB (n=Paul@apollo.paulsnet.org) left irc: Read error: 104 (Connection reset by peer)
[19:25] <rjharrison> Laurenceb: Passed
[19:25] <Laurenceb> nice
[19:25] <rjharrison> Now on to the advanced
[19:26] <Laurenceb> my antenna seems to work
[19:26] <rjharrison> Laurenceb: You found a REPEATER?
[19:26] <rjharrison> oops
[19:26] <Laurenceb> need a hab to track to be 100% sure
[19:26] <rjharrison> hehe
[19:26] <Laurenceb> not sure
[19:26] <rjharrison> Con on CUSF
[19:26] <rjharrison> come
[19:27] <Laurenceb> I found a bit of stuff that didnt appear to be local
[19:27] <rjharrison> Are they having a problem getting it up
[19:27] <Laurenceb> and you could tell by the background noise it was working, the polarization was right, and it was centered on 434
[19:27] <Laurenceb> hehe
[19:28] <rjharrison> Do you have a way of sending voice
[19:28] <rjharrison> Put a radio check out on the local repeater
[19:28] <Laurenceb> I dont have any tx kit
[19:28] <rjharrison> Asuuming you have a licence
[19:28] <rjharrison> Ok
[19:28] <Laurenceb> how do repeaters work?
[19:28] <Laurenceb> I dont have a license
[19:28] <rjharrison> Well you tx to them and they tx out again
[19:29] <Laurenceb> what do you send?
[19:29] <rjharrison> Normally with a shift in frequency
[19:29] <rjharrison> Voice normally
[19:29] <rjharrison> Typeically 600khz diffetence
[19:29] <rjharrison> difference
[19:29] <Laurenceb> neat
[19:30] <rjharrison> inbound on 145.1875 outbound on 145.7875
[19:30] <rjharrison> This is my local repeater
[19:30] <rjharrison> You send a sub audible tone too to activate it
[19:31] <rjharrison> in my case 82.5hz
[19:32] <rjharrison> Usually repeaters are in high places and therfore good for longer distance rather than calling simplex
[19:32] <rjharrison> ie radio to radio
[19:36] Tigga (n=chatzill@pc-232-235-103.magd.cam.ac.uk) left irc: "ChatZilla 0.9.84 [Firefox 3.0.5/2008120122]"
[20:35] rjharrison (n=rharriso@80.176.172.227) left #highaltitude.
[20:46] josepharmbruster (n=josephar@45.144.119.70.cfl.res.rr.com) joined #highaltitude.
[21:18] edmoore (n=edmoore@pomegranate.chu.cam.ac.uk) joined #highaltitude.
[21:49] SpeedEvil (n=user@tor/regular/SpeedEvil) left irc: Read error: 113 (No route to host)
[21:56] shellevi1 (n=user@mauve.plus.com) joined #highaltitude.
[21:58] <shellevi1> Prolly going to be offline for a few hours - power cut
[21:58] <shellevi1> :/
[21:58] Action: shellevi1 stabs the UKs shoddy infrastructure.
[22:03] ShellEvil (n=jkffjkjf@tor/regular/SpeedEvil) left irc: Read error: 113 (No route to host)
[22:10] G8KHW (n=Steve@217.47.75.27) joined #highaltitude.
[22:10] <Laurenceb> hi steve
[22:11] <Laurenceb> I've made a moxon antenna
[22:11] <Laurenceb> but nothing to pick up :-/
[22:12] <G8KHW> such is life
[22:14] <Laurenceb> I'm considering making a power supply for my icom
[22:14] <Laurenceb> but it'll take £25 worth of batteries
[22:15] <edmoore> don't drive 9 pints, add that to the £25, then just buy one
[22:15] <edmoore> 0-20V, 0-30A, cost me about £65
[22:15] <edmoore> drink*
[22:15] <edmoore> not drive
[22:15] <Laurenceb> one what?
[22:15] <Laurenceb> oh a ups
[22:16] <edmoore> well, a PS
[22:16] <Laurenceb> I was going to use AA Ni-MH
[22:16] <edmoore> mine doesn't have any U
[22:16] <Laurenceb> 2500mah
[22:16] shellevi1 (n=user@mauve.plus.com) left irc: Read error: 110 (Connection timed out)
[22:17] <Laurenceb> you can get up to 2850mah now
[22:18] <Laurenceb> I was reading about EEstor
[22:20] <Laurenceb> http://en.wikipedia.org/wiki/EEStor
[22:29] Hiena (n=Hiena@81.93.195.181.datatrans.hu) joined #highaltitude.
[22:29] <natrium42_> Hi Hiena
[22:29] <Hiena> Hi natrium42_ !
[22:29] <Hiena> How is the things on the west side?
[22:30] <natrium42_> lol, pretty good
[22:30] <natrium42_> how are you?
[22:30] <Hiena> In pain. Had a back injury from a self, and out of the painkiller.
[22:31] <natrium42_> is that still from the hang gliding incident?
[22:33] <Hiena> Nope, moved a cable from under the welded steelself and it felt on to me strained my back seriously.
[22:33] <natrium42_> ah, ouch :(
[22:35] <Hiena> Yup, i ended up in the traumatology, got a several xray shot, and sent to the bedrest for two weeks.
[22:54] G8KHW (n=Steve@217.47.75.27) left irc: "Leaving"
[22:58] shellevil (n=user@mauve.plus.com) joined #highaltitude.
[22:59] Hiena (n=Hiena@81.93.195.181.datatrans.hu) left irc: "-=Halt! Hammerzeit!=-"
[23:33] <Laurenceb> wb shellevil
[23:33] <Laurenceb> why th epower cut?
[23:41] Action: Laurenceb is pondering the EEstor concept
[23:43] <Laurenceb> looks like you have barium titanata power in PET film
[23:43] jiffe20 (n=jiffe@209.159.247.189) left irc: "Leaving"
[23:44] <Laurenceb> I'm a bit confused how that works... if you look at the molecular level, each PET or barium titanate molecule can be polarized, so you average the permittivities
[23:45] <Laurenceb> but if you have two capacitors in series, you have 1/c=1/(1/c_1+1/c_2)
[23:46] <Laurenceb> erm c=1/ ...
[23:48] <shellevil> I don't think it's complex
[23:49] <shellevil> it's basically a _really_ big ceramic capacitor operating at insane field strengths.
[23:49] <shellevil> was lots of wind
[23:49] <shellevil> eestor - if they can get it to work cheaply - will be revolutionary.
[23:51] <Laurenceb> well they get very high performance out of it
[23:51] <shellevil> Well...
[23:52] <Laurenceb> they heat to 180C - just below the melting point, thne apply 4Kv to align the titanate crystals
[23:52] <Laurenceb> along the crystal axes
[23:52] <Laurenceb> that seems to be the key
[23:52] <shellevil> One or two lab samples != a hundred thousand tons a year of them.
[23:52] <shellevil> Which is what'd really be needed.
[23:53] <Laurenceb> the thing I dont get is that the crystals wont be touching
[23:53] <Laurenceb> there polymer inbetween
[23:53] <Laurenceb> how does that effect the permitivitty
[23:53] <Laurenceb> isnt it a series capacitor like effect?
[23:54] <shellevil> I th0ught the aim of the PET was simply to fill the voids in the crystals
[23:54] <shellevil> as any teeny little void will flash over
[23:54] <Laurenceb> not as I understand it
[23:55] <Laurenceb> they make titanate powder by chemical means
[23:55] <shellevil> I haven't bothered trying to understand it.
[23:55] <Laurenceb> so its veyr pure single crystals
[23:55] <shellevil> A) I'd need to learn lots.
[23:55] <Laurenceb> then mix with a polymer
[23:55] <Laurenceb> and blow into sheets plastic bag style
[23:55] <shellevil> B) I don't have a massive chemical plant, and anyway they have patents
[23:55] <Laurenceb> coat with alu
[23:56] <Laurenceb> then heat to 180C and apply 4Kv to align the crystals
[23:56] <Laurenceb> in the polymer carrier
[23:57] <shellevil> If they can get it working, it kills all batteries.
[23:57] <Laurenceb> I guess the key to this is the transition point from where you model it as a bunch on mini caps to where you model it as molecules
[23:57] <shellevil> It makes electric cars easy
[23:57] <Laurenceb> which I guess depends on the thickness of the surface change
[23:57] <shellevil> It makes renewable energy cheaper.
[23:57] <shellevil> ...
[23:58] <Laurenceb> god knows how thinck that is
[23:58] <shellevil> Even completely wacky stuff like electric airliners become sort-of-plausible.
[23:58] <Laurenceb> yeah
[23:59] <shellevil> If you can get the price below maybe 100 quid/Kwh
[00:00] <Laurenceb> ok I guess if you have really small crystals the bulk properties of the sheet are just the average of the polymer and titanate
[00:00] --- Sun Feb 1 2009