[00:00] <edmoore> I just remember a pretty funkily routed board of his
[00:00] <natrium42> i know that laurence likes to route that way
[00:00] <natrium42> hehe :)
[00:00] <Laurenceb> yes
[00:00] <edmoore> it's how spiderman would probably do it
[00:00] <Laurenceb> I tried it
[00:00] <Laurenceb> but couldnt get it running properly
[00:00] <edmoore> right, see you all
[00:00] <Laurenceb> you have to do file type conversions
[00:01] <Laurenceb> cya
[00:01] <natrium42> later
[00:01] edmoore (n=edmoore@pomegranate.chu.cam.ac.uk) left irc:
[00:03] <jcoxon_> fergusnoble, what sort of track width is reasonable?
[00:03] <fergusnoble> just for signals? not high-current?
[00:03] <natrium42> 8 mil is good
[00:03] <jcoxon_> just signals
[00:03] <Laurenceb> IIRC olimex cant go that small
[00:03] <fergusnoble> jcoxon_: the real answer is as big as you can make them
[00:04] <Laurenceb> batchpcb can
[00:04] <Laurenceb> yeah
[00:04] <Laurenceb> your not doing anything tricky
[00:04] <fergusnoble> but 10mil is a lower limit for basic stuff
[00:04] <fergusnoble> doing any smt?
[00:04] <Laurenceb> its just an adaptor right?
[00:04] <jcoxon_> i'm playing with a breakoutboard thingly
[00:05] <jcoxon_> the only issue is that the connector is horrible
[00:05] <jcoxon_> as its for a 0.5mm flex cable
[00:06] <fergusnoble> ok, well go for roughtly the same, maybe slightly smaller than the pads on your nasty connector
[00:06] <Laurenceb> no so hard to solder with the right technique
[00:06] <fergusnoble> jcoxon_: what is on the other end of the flex cable? :)
[00:07] <jcoxon_> the gumstix verdex has a 24pin flex connector
[00:07] <jcoxon_> and i was going to use a ffc and then a pcb breakout
[00:07] <fergusnoble> cool
[00:08] <Laurenceb> fergus, whats in overlay_icons.csv ?
[00:08] <Laurenceb> is flight_so_far a record in case the tracker crashes?
[00:09] <fergusnoble> flight_so_far.csv is an input to the prediction software
[00:10] <fergusnoble> its just a list of all the recieved points
[00:10] <fergusnoble> overlay_icons.csv is a list of coords against image file names that my modified tangogps reads in and uses to plot icons on th map
[00:12] Action: SpeedEvil arises from his dip, sparkling with droplets, as Adonis* rising from a pool. *(mythological figure may vary)
[00:14] <fergusnoble> ok, sleep time
[00:14] <fergusnoble> night all
[00:15] <SpeedEvil> night
[00:16] <SpeedEvil> jcoxon_: generally, the best approach is to have a play. slap the components down. Drag them around to where you have the least number of crossing air-wires. Hit 'autoroute'. Try to figure out what's giving the autorouter problems. Maybe rip up all the tracks, do a couple manuyally and retry. After doing a few iterations on your board, it gets a lot neater.
[00:17] <SpeedEvil> There are lots of tricks
[00:20] <jcoxon_> SpeedEvil, yeah thats what i'm doing
[00:20] <jcoxon_> its a simple board
[00:20] <SpeedEvil> generally ##electronics is willing to help with this sort of question
[00:21] <jcoxon_> okay cool
[00:21] <SpeedEvil> if you get to a state you consider 'OK' - post it here or there - images - and people can look at it and point out why it's ALL WRONG!
[00:21] <SpeedEvil> :)
[00:21] <jcoxon_> hehe
[00:22] <SpeedEvil> Generally, for low-speed, low power circuits, if it all routes OK, it's a matter of asthetics.
[00:22] hallam (i=836fc8c8@gateway/web/ajax/mibbit.com/x-b8136b5173966818) joined #highaltitude.
[00:22] <SpeedEvil> Any (or worse, some of) high frequency, low noise, low signal levels, high power, high switching speed, large processors, and it gets messy
[00:23] <jcoxon_> hmmm does eagle work in inches?
[00:24] gregHome (n=gleblanc@ joined #highaltitude.
[00:26] <Laurenceb> say hallam: local osc is a function of data point number
[00:26] <Laurenceb> jcoxon: yes
[00:27] <Laurenceb> hallam: so and correction to th frequency caused by the loop filter should be doubled
[00:27] <Laurenceb> or just make your frequency take input in rads/s
[00:28] <Laurenceb> then convet to the appropriate for you samples
[00:30] jcoxon_ (n=jcoxon@ left irc: "Leaving"
[00:36] <Laurenceb> *any correction
[00:38] <hallam> the frequency already takes input in rad/s
[00:40] <Laurenceb> theres no difference then
[00:41] <Laurenceb> its got to be a bug somewhere
[00:42] <hallam> yeah
[00:42] <hallam> I've now replicated it in the 8MS/s version
[00:42] <hallam> I thinkl
[00:44] <Laurenceb> ?!
[00:44] <Laurenceb> how
[00:44] <hallam> well, I fiddled with the carrier phase and the chucking so that it matches the 4MS/s version
[00:44] <hallam> and it no longer locks
[00:45] <Laurenceb> odd
[00:45] gregHome (n=gleblanc@ left irc: Read error: 104 (Connection reset by peer)
[00:48] <Laurenceb> are there badger logfiles online somewhere?
[00:50] <Laurenceb> ah got one nvm
[00:58] <Laurenceb> fergusnoble: ok got it running, but it doesnt do anything - just sits there
[00:58] <Laurenceb> does it need to detect new lines in the log file?
[01:00] EI5GTB (n=Paul@apollo.paulsnet.org) left irc: Read error: 104 (Connection reset by peer)
[01:00] <Laurenceb> if data[0] == callsign: <-- is that robust?
[01:01] <Laurenceb> surely it should look for the end of data[0]
[01:20] <Laurenceb> data[0][len(data[0])-len(callsign):len(data[0])]
[01:20] <Laurenceb> or something....
[02:12] akawaka (n=akawaka@66-215-97-198.dhcp.malb.ca.charter.com) left irc: Remote closed the connection
[02:58] kc0wys (n=kc0wys@mo-67-77-184-156.sta.embarqhsd.net) joined #highaltitude.
[03:12] kc0wys (n=kc0wys@mo-67-77-184-156.sta.embarqhsd.net) left irc:
[04:13] Laurenceb (n=laurence@dyres221-140.surrey.ac.uk) left irc: Read error: 104 (Connection reset by peer)
[05:16] hallam (i=836fc8c8@gateway/web/ajax/mibbit.com/x-b8136b5173966818) left irc: "http://www.mibbit.com ajax IRC Client"
[06:21] Simon-MPFH (n=simon@phantom.mpfh.co.uk) joined #highaltitude.
[07:01] EarlJr (n=earljr@66-230-111-65-rb1.nwc.dsl.dynamic.acsalaska.net) left irc: Read error: 104 (Connection reset by peer)
[07:01] EarlJr (n=earljr@66-230-115-119-rb1.nwc.dsl.dynamic.acsalaska.net) joined #highaltitude.
[07:10] EarlJr (n=earljr@66-230-115-119-rb1.nwc.dsl.dynamic.acsalaska.net) left irc: Read error: 104 (Connection reset by peer)
[07:14] rjharrison (n=rharriso@ left #highaltitude.
[08:20] edmoore (n=edmoore@pomegranate.chu.cam.ac.uk) joined #highaltitude.
[08:25] edmoore (n=edmoore@pomegranate.chu.cam.ac.uk) left irc: Client Quit
[09:53] rharrison (n=rharriso@gateway.hgf.com) joined #highaltitude.
[10:39] hallam (i=836fc8c8@gateway/web/ajax/mibbit.com/x-f5451f8a08d41613) joined #highaltitude.
[10:39] <hallam> morning lal
[10:39] <hallam> all*
[10:44] <fergusnoble> hallam: hello
[10:50] <hallam> how's it going?
[10:51] <hallam> I think I narrowed down the GPS problem a bit
[10:51] <hallam> all that chucking stuff was fairly pointless
[10:52] <hallam> I have a 4MSps, square-wave carrier, integer arithmetic matlab version working nicely now
[10:52] <hallam> so I'm just translating it to C
[10:55] <hallam> and I improved the loop filters so the pseudorange accuracy is about 3% of a chip, = 10m
[10:56] <hallam> carrier aiding helps enormously
[11:04] jiffe20 (n=jiffe@ left irc: "Leaving"
[11:11] edmoore (n=edmoore@pomegranate.chu.cam.ac.uk) joined #highaltitude.
[11:23] <hallam> hi ed
[11:23] <edmoore> hi henry
[11:23] <edmoore> how's it going?
[11:23] <hallam> not too bad, think I'm nearly at the point of getting the C code to work
[11:23] <edmoore> ok awesome
[11:24] <edmoore> fergus was telling me about some fun bit manipulations to take advantage of the instruction set
[11:24] <edmoore> that you did
[11:24] <hallam> oh yes
[11:24] <hallam> very much so
[11:24] <edmoore> You'll have to talk me through that some time soon - I love that kind of stuff
[11:31] <edmoore> would you be up for talking about this lot at the talk?
[11:31] <edmoore> it should make some pretty cool content
[11:33] <hallam> sure
[11:33] <hallam> it's pretty damn fascinating once you get into it
[11:33] <hallam> when is the talk?
[11:34] <edmoore> march some time
[11:34] <edmoore> ages
[11:34] <edmoore> end of term
[11:39] <rharrison> How did the baking go guys?
[11:39] <rharrison> I'm not sure we heard about the results?
[11:39] <hallam> baking eh?
[11:40] <edmoore> baking?
[11:40] <edmoore> oh the rocket tube. yeah it was a good improvement on previous ones
[11:40] <edmoore> the latex tube worked a lot better
[11:40] <rharrison> Cool, did any one snap any pics
[11:40] <edmoore> i don't think so
[11:41] <edmoore> will try to on the next one
[11:41] <edmoore> there are always more being made :)
[11:41] <rharrison> I guess symmetry is what it's all about
[11:41] <edmoore> on one axis anyway
[11:42] <rharrison> Is the cone part of the process or is that added at a later stage?
[11:42] <edmoore> it's part of the mould
[11:42] <edmoore> but the very tip will probably be replaced with something else
[11:42] <edmoore> perhaps that the gps signals can propagate through
[11:43] <rharrison> hehe guidance tip
[11:43] <edmoore> and depends on how recovery is going to be done too
[11:43] <hallam> edmoore, I think I favour some fiberglass windows in the sides for the GPS
[11:43] <rharrison> I guess you are following a standard'ish procedure for making the rocket body
[11:45] <edmoore> is there such a thing?
[11:45] <edmoore> hallam: yeah - as you say there should be plenty of sats in view
[11:47] <edmoore> thoughfor an altitude fix, the geometry would favour an antenna that can sea more above - would it not?
[11:47] <hallam> sort of
[11:47] <hallam> really you want plenty from all directions
[11:47] <edmoore> though equally again I suppose it's a function of how recovery is done. That will have a large bearing on the internal layour
[11:47] <hallam> and there is more solid angle out the sides than out the tip of the nose
[11:48] <hallam> we can experiment easily enough on the ground
[11:48] <edmoore> sure
[11:49] <edmoore> the lpc2368 is cheaper than the 2148
[11:49] <edmoore> go figure
[11:53] <hallam> oh, has Fergus been excited about it to you?
[11:54] <edmoore> I looked at using it for a Vorticity project
[11:54] <edmoore> but ended up working on different stuff
[11:54] <edmoore> but he told me about the fast SPI
[11:54] <edmoore> mooted using it for badger but the 2148 works and has code written for it
[11:54] <edmoore> for a new design we obviously have more flexibility. I might try it for my radio board
[11:56] <hallam> not just fast SPI, but a dedicated SD card interface (4-bit data)
[11:56] <hallam> it makes a really nice rocket board design possible
[11:56] <edmoore> does it have all the crchardware built in?
[11:57] <edmoore> hang on, will yoink the datasheet and have a skim
[11:57] <edmoore> hallam: heard the ISS 2m repeater yesterday as it came over cambridge
[11:58] <hallam> oh, that's pretty cool
[12:11] <SpeedEvil> I note that many SD cards can write at 12Mbytes/second
[12:11] <SpeedEvil> So, you can log the raw GPS too, if you feel like that.
[12:12] <hallam> we plan to
[12:12] <SpeedEvil> Also.
[12:13] <hallam> hence the 2368
[12:13] <SpeedEvil> As you're already doing the nasty bits of the GPS in software.
[12:13] <hallam> found a benchmark with demonstrated write speeds >5MB/s
[12:13] <SpeedEvil> It seems plausible that it's relatively simple to do differential GPS on two antennas, for orientation
[12:13] <SpeedEvil> as long as tehy are a bit away from each other.
[12:16] <SpeedEvil> Did I raise my idea of multiple antennas cardinally oriented, outputting X, -X, Y,-Y, Z,-Z, and seperate GPS channels, so you have a direction for each sat too?
[12:19] <hallam> what does that get you?
[12:20] <SpeedEvil> orientation
[12:22] <hallam> oh, without needing even a ~1 meter baseline
[12:22] <SpeedEvil> which'd be hard for diameter
[12:23] <edmoore> this is relying on the directionality of the ants, or am I missing something?
[12:23] <hallam> are you saying you can't get full orientation with one ant in the nose and one in the tail of the rocket?
[12:23] <SpeedEvil> no.
[12:24] <SpeedEvil> you don't get rotation
[12:24] <edmoore> I think I see what speedevil is saying - just yaw
[12:24] <hallam> and pitch
[12:24] <SpeedEvil> pitch too
[12:24] <hallam> ok, makes sense
[12:25] <hallam> prolly a magnetometer is cheaper than a bunch more antennas and tracking channels :P but it's a cool idea.
[12:25] <SpeedEvil> Magnetometers are lots easier if they work
[12:26] <SpeedEvil> Magnetometers don't give you absolute orientation though, just one vector
[12:26] <hallam> yeah, that and the pitch baseline
[12:27] <hallam> I know they can line up, but you avoid that or dead-reckon
[12:35] <SpeedEvil> Sure.
[12:36] <SpeedEvil> It also - for more ambitious stuff - means you need a magnetic field model, which adds to complexity
[12:37] <SpeedEvil> I think with the right antennas, if I'm not wrong, connecting a +X, -X antenna to the same frontend gives you absolute X direction
[12:37] <SpeedEvil> But, that's still 3 frontends, or at the very minimum, 3 switchable antenna pairs
[12:44] <hallam> SpeedEvil, can I ask what you do for a living?
[12:44] <hallam> I was considering going into GPS when I graduate, but I have a very short attention span so that might not be a good idea
[12:50] <SpeedEvil> hallam: :(
[12:50] <SpeedEvil> hallam: currently I'm on disability living allowance.
[12:50] <SpeedEvil> I've got Post Viral Fatigue Syndrome.
[12:51] <SpeedEvil> Which basically means that I have maybe a couple of hours a day when I can do creative stuff, and the rest of the time just barely existing.
[12:51] <SpeedEvil> And also horribly weak.
[12:52] <SpeedEvil> I'm trying to learn stuff, and get stuff designed for niches that can go on ebay.
[12:52] <hallam> I'm sorry to hear that
[12:53] <hallam> but good luck with the niche stuff designing, it's a lot of fun
[12:53] <SpeedEvil> Deeply annoying, if I was fit, I'd expect to be doing something in the RF/chip-design field.
[12:55] <SpeedEvil> Or possibly satellite design - who knows.
[12:58] <hallam> well, I hope you get better in time for GPS Block III
[12:59] <SpeedEvil> :/ Yeah. Problem then is starting a career from a non-existant CV. Oh well. Can only go one step at a time.
[13:05] jcoxon_ (n=jcoxon@host86-158-31-172.range86-158.btcentralplus.com) joined #highaltitude.
[13:46] <jcoxon_> ooo finally a connector for the badass ublox gps from sparkfun
[13:46] <jcoxon_> http://www.sparkfun.com/commerce/product_info.php?products_id=9079
[14:03] jcoxon_ (n=jcoxon@host86-158-31-172.range86-158.btcentralplus.com) left irc: "Leaving"
[15:34] hallam (i=836fc8c8@gateway/web/ajax/mibbit.com/x-f5451f8a08d41613) left irc: "http://www.mibbit.com ajax IRC Client"
[15:52] edmoore (n=edmoore@pomegranate.chu.cam.ac.uk) left irc:
[16:19] Bluenarf (n=Paul@apollo.paulsnet.org) joined #highaltitude.
[16:20] Nick change: Bluenarf -> EI5GTB
[17:14] <SpeedEvil> http://www.x-rates.com/d/USD/GBP/graph120.html
[17:14] <SpeedEvil> ow.
[17:26] <rharrison> SpeedEvil: You seem to know your stuff. You're like the hidden robot voice of HA. How far did you get with your studies?
[17:27] <SpeedEvil> I got to basically 2nd year of a physics/math/CS degree, before dropping out due to health.
[17:28] <SpeedEvil> Electronics is something I've not really done any formal stufy on
[17:28] <SpeedEvil> That was about 5 years ago.
[17:39] Bluenarf (n=Paul@apollo.paulsnet.org) joined #highaltitude.
[17:40] rharrison (n=rharriso@gateway.hgf.com) left irc:
[17:44] edmoore (n=ed@chu-gw.churchillcambridge.co.uk) joined #highaltitude.
[17:56] EI5GTB (n=Paul@apollo.paulsnet.org) left irc: Read error: 104 (Connection reset by peer)
[18:39] edmoore (n=ed@chu-gw.churchillcambridge.co.uk) left irc: Read error: 110 (Connection timed out)
[18:40] edmoore (n=ed@chu-gw.churchillcambridge.co.uk) joined #highaltitude.
[19:31] Simon-MPFH (n=simon@phantom.mpfh.co.uk) left irc: Read error: 104 (Connection reset by peer)
[19:31] simon__ (n=simon@phantom.mpfh.co.uk) joined #highaltitude.
[19:44] <edmoore> hi simon__
[20:02] edmoore_ (n=edmoore@pomegranate.chu.cam.ac.uk) joined #highaltitude.
[20:05] edmoore (n=ed@chu-gw.churchillcambridge.co.uk) left irc: Read error: 110 (Connection timed out)
[20:54] rjharrison (n=rharriso@ joined #highaltitude.
[20:54] rjharrison (n=rharriso@ left #highaltitude.
[21:07] Laurenceb (n=laurence@dyres221-140.surrey.ac.uk) joined #highaltitude.
[21:07] <Laurenceb> hello
[21:27] <SpeedEvil> hellow
[21:28] <Laurenceb> http://www.sandpiper.entadsl.com/
[21:28] <Laurenceb> ^ they look nice and cheap
[21:29] <Laurenceb> I'm working on tracker code atm
[21:30] <Laurenceb> probably going to incorporate a simple map into the tracker
[21:30] <Laurenceb> if its possible to load a jpeg large than the window without scaling then its easy
[21:30] simon__ (n=simon@phantom.mpfh.co.uk) left irc: "Leaving"
[21:33] <fergusnoble> Laurenceb: i recommend looking at tangogps, its really cood
[21:33] <fergusnoble> *good
[21:33] <fergusnoble> and easy to mod
[21:33] <Laurenceb> I had a read
[21:34] <fergusnoble> also, if you make the parsing of the radio data more robust i would be really interested to know, because the way im doing it is pretty stupid
[21:34] <Laurenceb> couldnt really follow what you'd done
[21:34] <SpeedEvil> what's this for?
[21:34] <SpeedEvil> tracker?
[21:34] <Laurenceb> yes
[21:34] <Laurenceb> fergus: did you mode the tangogps sourcecode?
[21:34] <fergusnoble> Laurenceb: yes
[21:34] <Laurenceb> *modify
[21:34] <Laurenceb> ah
[21:34] <SpeedEvil> I mean - have you considered openlayers?
[21:34] <Laurenceb> whats that?
[21:34] <fergusnoble> i can send it to you iuw, its quite ugly but it works
[21:35] <SpeedEvil> http://openlayers.org/
[21:35] <Laurenceb> http://pastebin.com/dfa4f5bf
[21:35] <fergusnoble> Laurenceb: i used a lot of open source maps, like openstreetmap etc before but they dont really have the detail yet to be very useful for tracking
[21:36] <SpeedEvil> fergusnoble: really depends where
[21:36] <Laurenceb> yeah, and we want it to work with no web access
[21:36] <SpeedEvil> fergusnoble: london and environs they are quite good actually
[21:36] <Laurenceb> fergusnoble: that code still needs something to wipe junk off the end
[21:36] <SpeedEvil> cambridge is almost 100% mapped AIUI too
[21:36] <fergusnoble> Laurenceb: are you making it comaptible with the ukhas standard protocol?
[21:36] <edmoore_> those places that we definitely don't want to launch balloons in
[21:36] <Laurenceb> and check the checksum
[21:36] <SpeedEvil> edmoore_: of course. That's one problem - it's a streetmap
[21:37] <Laurenceb> fergusnoble: no, the mini rogallo is a bit different
[21:37] <SpeedEvil> edmoore_: if there are no streets, there tend to be blank bits :)
[21:37] <fergusnoble> SpeedEvil: yeah its really good in built up areas and its always getting better, but its just not got all the little lanes and footpaths in the middle of owhere that makes google maps so useful
[21:37] <Laurenceb> fergusnoble: theres a checksum
[21:38] <Laurenceb> I may try and use data even if the checksum doesnt match
[21:38] <fergusnoble> Laurenceb: the ukhas format only specifies the first few fields and the rest you can do what you want with
[21:38] <Laurenceb> yeah
[21:38] <fergusnoble> but its up to you of course
[21:38] <Laurenceb> I'll see about chenging it
[21:39] <SpeedEvil> Also - tangogps may get your IP banned if you do google with it
[21:39] <fergusnoble> if you come up with a good checksumming method then maybe we could incorporate it into the standard
[21:39] <SpeedEvil> MD5! :)
[21:39] <Laurenceb> hehe
[21:39] <Laurenceb> I just sum
[21:39] <fergusnoble> SpeedEvil: ive found its ok if you do smallish chunks at a time, but you do have to be careful
[21:39] <SpeedEvil> NMEA seems to work OK
[21:39] <Laurenceb> yes
[21:40] <SpeedEvil> (the sum)
[21:40] <Laurenceb> just sum into a u08
[21:40] <SpeedEvil> no need to reinvent the wheel.
[21:40] <Laurenceb> it runs from the rx ISR
[21:40] <Laurenceb> *tx
[21:41] <fergusnoble> Laurenceb: also what would be really nice is to get it to load google maps satellite data
[21:41] <Laurenceb> hehe
[21:41] <SpeedEvil> Naah.
[21:42] <edmoore_> just xoring all the bytes of data is pretty easy and standard
[21:42] <SpeedEvil> What you need is to launch a balloon with real-time video downlink
[21:42] <edmoore_> trivial to do on a micro
[21:42] <Laurenceb> yeah
[21:42] <fergusnoble> it cant at the moment as for sat it uses a different format for specifing a location, i.e. not simple lat and long and oom
[21:42] <fergusnoble> *zoom
[21:42] <Laurenceb> freesat kit is very cheap
[21:42] <fergusnoble> ok, bbl
[21:43] <Laurenceb> not sure if the front ends work on 2.4GHz
[21:43] <Laurenceb> but you can pick up a large dish and front end for ~£50
[21:43] <SpeedEvil> freesat?
[21:43] <SpeedEvil> oh
[21:43] <SpeedEvil> I got a couple of 80cm dishes from Lidl.
[21:43] <Laurenceb> http://www.freesat.co.uk/
[21:44] <SpeedEvil> 7.99 each
[21:44] <Laurenceb> haha
[21:44] <Laurenceb> that _is_ cheap
[21:44] <SpeedEvil> They had a _huge_ stack of them.
[21:46] <SpeedEvil> I have a problem.
[21:46] <SpeedEvil> My name is speedevil, and I am addicted to cheap electronics stuff.
[21:47] <Laurenceb> looks like the channels are around 11GHz
[21:47] <Laurenceb> doubt the LNA will work at 2.4
[21:47] <SpeedEvil> Today, morrisons, I bought, solely because they were 48p for 2, with 2 free packs (7p/bulb!!!) 11W CFLs.
[21:47] <Laurenceb> a bit dim
[21:48] <SpeedEvil> I have a bagfull of CFLs in the attic.
[21:48] <Laurenceb> ooh I know what would work - a gps front end
[21:48] <SpeedEvil> Having said that.
[21:48] <SpeedEvil> yeah
[21:48] <Laurenceb> active antenna disassembled
[21:48] <SpeedEvil> I partially bought mine for GPS experiments
[21:48] <SpeedEvil> On that - they are great for lighting.
[21:48] <Laurenceb> then you need a 2.4GHz antenna on it
[21:49] <Laurenceb> with polarization to match the one on the balloon
[21:49] <SpeedEvil> I took a big pile of the 48p CFLs that morrisons were doing a couple of years ago, strung them at 50cm intervals on a cable, and put 25 of them up in my garage.
[21:49] <SpeedEvil> 'energy saving' 500W.
[21:50] <SpeedEvil> But I can see in there.
[21:50] <Laurenceb> even so, with a 80cm dish you'll only get a few Km range
[21:50] <Laurenceb> from a 10mw 2.4GHz tx
[21:50] <SpeedEvil> yeah.
[21:51] <Laurenceb> need one of the dodgy chinese 500mw transmitters off ebay
[21:51] <Laurenceb> I know some guys who go about 1Km with a huge yagi
[21:52] <Laurenceb> they had a 10mw camera on a red kite
[21:52] <Laurenceb> - bird
[21:55] <SpeedEvil> I was looking at - for my UAV thingy - 802.11A link with the dish
[21:55] <SpeedEvil> I figure it should be good to at least a couple of Km
[21:55] <SpeedEvil> and 433MHz yagi for beyond that
[21:56] <Laurenceb> interesting
[21:56] <Laurenceb> http://www.thefreelibrary.com/Crowcam:+camera+on+bird%27s+tail+captures+bird+ingenuity.-a0169754109
[21:56] <SpeedEvil> Did I mention DX has several watt 2.4G video senders?
[21:56] <Laurenceb> ^ I think that used the 10mw cameras as well
[21:56] <Laurenceb> ok
[21:57] <Laurenceb> not very legal :P
[21:59] <SpeedEvil> You only power it up when in international waters :)
[22:00] <SpeedEvil> You only power it up when above international waters :)
[22:01] <Laurenceb> hehe
[22:02] <SpeedEvil> Currently wondering about sensor fusion for my car. :)
[22:02] <SpeedEvil> 4 cameras, presenting on one screen.
[22:02] <SpeedEvil> wide-angle.
[22:03] <SpeedEvil> To do a fake overhead view.
[22:05] <SpeedEvil> It's a pity dvb-s isn't trivial to transmit.
[22:05] <SpeedEvil> http://www.dealextreme.com/details.dx/sku.15820
[22:07] <Laurenceb> you need to generate ofdm ?
[22:07] <SpeedEvil> That sort of thing
[22:08] <SpeedEvil> Did I link to that 'VGA camera pen' ?
[22:08] <Laurenceb> nope
[22:08] <Laurenceb> I want something similarly sized with onboard storage
[22:09] <SpeedEvil> http://www.dealextreme.com/details.dx/sku.18088
[22:09] <edmoore_> fergusnoble: what do your monday and wednesday afternoons look like - do you have labs?
[22:09] <fergusnoble> edmoore_: usually goo
[22:09] <fergusnoble> good
[22:10] <fergusnoble> i have a supervision wed 4pm
[22:10] <fergusnoble> why?
[22:10] <edmoore_> that's hom
[22:10] <edmoore_> poo
[22:10] <edmoore_> just wondering
[22:10] <edmoore_> trying to assess the shape of term and may be able to jiggle some stuff
[22:11] <fergusnoble> mondays i often have to do work for my tues supervisions too, but i can be free
[22:13] <fergusnoble> are you thinking of having a time to meet to do cusf stuff?
[22:15] <edmoore_> possibly yeah
[22:17] <edmoore_> have been called to last orders castle
[22:17] <edmoore_> bbl
[22:17] <edmoore_> oh fergusnoble sorry i've been meaning to tell you
[22:17] <edmoore_> i had a brain blank
[22:17] <edmoore_> i remember why I asked
[22:17] <edmoore_> I'm attanding Mackays Part III course
[22:17] <edmoore_> it's the lecture course of his book
[22:17] <edmoore_> in Cavendish
[22:18] <edmoore_> first one today
[22:18] <edmoore_> next on wednesday
[22:18] <edmoore_> be excited
[22:18] <edmoore_> if your timetable allows it
[22:18] <edmoore_> ... or just read the book, I guess :)
[22:19] <edmoore_> it's a shared course with engineering
[22:19] <edmoore_> although it was mainly engineers, it seems
[22:19] <edmoore_> based on his quick show of hands
[22:21] <edmoore_> ok well am off to castle
[22:21] <edmoore_> will talk later
[22:21] <edmoore_> my monologue co-efficient just sky-rocketed
[22:21] Nick change: edmoore_ -> edmoore|away
[22:28] <Laurenceb> http://pastebin.com/d6bc7b7ec
[22:28] <Laurenceb> ^ fergusnoble
[22:31] <Laurenceb> http://pastebin.com/d7d7c013d
[22:31] <Laurenceb> it only looks if the checksum matches
[22:31] <Laurenceb> but it wont mind junk either side of the callsign
[22:33] <Laurenceb> http://pastebin.com/d4dc7617d I'll never get this right
[22:46] <Laurenceb> fergusnoble: whats with the tail function?
[22:55] <fergusnoble> Laurenceb: it returns the last line of a file
[22:55] <fergusnoble> without reading through all of it
[22:55] <fergusnoble> i didnt write it, got it from a cookbook
[22:56] <Laurenceb> ok
[22:56] <Laurenceb> yeah I cant work out how it work :P
[22:57] <Laurenceb> bbl
[23:12] Nick change: Bluenarf -> EI5GTB
[23:26] natrium42 (i=natrium4@CPE000625d867e2-CM0014045885be.cpe.net.cable.rogers.com) left #highaltitude.
[23:31] natrium42 (i=natrium4@CPE000625d867e2-CM0014045885be.cpe.net.cable.rogers.com) joined #highaltitude.
[23:43] hallam (i=836fc8c8@gateway/web/ajax/mibbit.com/x-7c44751c262f9479) joined #highaltitude.
[23:43] <Laurenceb> hi hallam
[23:44] <Laurenceb> I thought about the tracking issue a bit more
[23:44] <Laurenceb> if it is real, maybe you need a longer integration time
[23:47] <hallam> hey Laurenceb
[23:47] <hallam> I don't think it's real
[23:47] <Laurenceb> thats gppd
[23:47] <hallam> I got the matlab version running on exactly the same data
[23:47] <hallam> successfully
[23:47] <hallam> I think it's a bug in the carrier or something
[23:47] <Laurenceb> I'm just thinking the lassen iq has a 10ms mode
[23:49] <Laurenceb> if you could replicate
[23:49] <hallam> hm
[23:49] <hallam> I would have thought that was less stable, but maybe not
[23:50] <Laurenceb> you'd need to align with the data bits
[23:51] <hallam> btw
[23:51] <hallam> carrier aiding is fantastically effective
[23:51] <hallam> makes the DLL so much more stable
[23:51] <Laurenceb> yeah
[23:51] <hallam> also
[23:51] <hallam> there was no need to do that chucking thing
[23:51] <hallam> you just advance the data pointer by a bit less than usual
[23:51] <hallam> it amounts to the same thing but it's much less complicated
[23:52] <hallam> and you still move forwards through the data buffer
[23:54] <Laurenceb> yeah
[23:54] <Laurenceb> how does the DMA work?
[23:54] <Laurenceb> is it a circular buffer?
[23:55] <hallam> it's very flexible
[23:55] <hallam> it can be made into a circular buffer
[23:55] <Laurenceb> if it was you'd need seperate pointers for each tracker
[23:55] <Laurenceb> to use that technique
[23:55] <hallam> yeah, that's kind of expected though
[23:55] <hallam> it's the same if you chuck
[23:56] <Laurenceb> hmm guess os
[23:56] <Laurenceb> I was wondering how you could achieve dynamic buffer size
[23:56] <hallam> well
[23:56] <hallam> I figure the trackers will not run at exactly real time speed
[23:56] <hallam> they'll either be slower, in which case the whole system won't work
[23:56] <hallam> or faster, in which case you can pause them to shuffle buffers around
[23:57] <Laurenceb> if the DMA pointer and all the buffer pointers are up near the top of the buffer, you can remove the bottom
[23:58] <Laurenceb> dam tracker1 tracker2 bottom
[23:58] <Laurenceb> dam tracker1 tracker2 bottom
[23:58] <Laurenceb> like so
[23:58] <hallam> yeah
[23:58] <Laurenceb> *dma
[23:58] <Laurenceb> ie where the dma sends data to
[23:58] <hallam> the Blackfin also has this nice circular pointer thing
[23:59] <hallam> it's got a couple of pointer registers, each with an associated top and bottom register
[23:59] <Laurenceb> thats kind of handy :P
[23:59] <hallam> yeah there's also the CORRELATE8184 instruction
[00:00] --- Tue Jan 27 2009