highaltitude.log.20090407

[00:06] <Laurenceb> there was an article about this in physics world a couple of months ago
[00:06] <Laurenceb> there is a theoretic mechanism involving oxygen ions IIRC
[00:07] <Laurenceb> but lots of people have questioned how any effect could come about when you have tens of Km of waterlogged crust inbetween the area where the stress in concentrated and the surface
[00:07] <stilldavid> Does anyone have any experience with the Beagle Board on a HAB?
[00:09] <Laurenceb> Beagle Board?
[00:09] <Laurenceb> whats that
[00:10] <stilldavid> a single board computer with pretty beefy specs: http://beagleboard.org
[00:14] sYx66 (n=wooh@bas1-kitchener06-1177615372.dsl.bell.ca) joined #highaltitude.
[00:14] <edmoore> stilldavid: not with HAB, but it looks like a nice thing
[00:15] <edmoore> was looking at it for a forthcoming project
[00:15] <edmoore> it goes without saying, catastrophically sledge-hammer to a nail for HAB stuff, unless you have a specific use in mind
[00:16] <stilldavid> well, I want to experiment with long-range wifi equipment, so I need something more capable than the standard micro
[00:16] <stilldavid> but I wouldn't plan on having anything essential on the board
[00:17] <edmoore> true, though things like an NGW100 may help with the bank balance and power consumption
[00:17] <edmoore> + still having the linux kernel
[00:17] <edmoore> Laurenceb has experience with them
[00:18] <Laurenceb> yeah its pretty low power
[00:18] <Laurenceb> processor doesnt even get warm
[00:19] <stilldavid> wow, looks promising indeed, and only $89 usd?
[00:20] <edmoore> amazing isn't it
[00:20] <edmoore> avrfreaks.net is a good place
[00:20] <edmoore> excellent forum, with quite a large and active section on avr32 (the cpu) linux
[00:21] <edmoore> avrfreaks.net is an excellent place generally
[00:21] <edmoore> and the avr atmega microcontrollers are bloody marvellous as 8-bit micros go
[00:25] <Laurenceb> I saw someone had fixed a psp screen onto a ngw100
[00:25] <Laurenceb> the inbuilt lcb controller can drive it with minimal hacking
[00:26] <edmoore> yes
[00:26] <Laurenceb> theres some cool videos of it playing quake
[00:26] <edmoore> youtube has a vid of it with psp screens
[00:26] <stilldavid> Laurenceb: how is the dev environment for it?
[00:26] <Laurenceb> avr32 studio is good
[00:26] <Laurenceb> but I've been working with python
[00:26] <SpeedEvil> Laurenceb: sounds a bit dodgy (O2)
[00:26] <stilldavid> does it come with python out of the box?
[00:27] <Laurenceb> nope
[00:27] <Laurenceb> apparently theres some rumors that the shipped kernel screws up the flash
[00:27] <Laurenceb> so I compiled from source
[00:28] <SpeedEvil> Flash. Ahahaaaaa!
[00:28] <Laurenceb> theres 16MB (IIRC) of flash
[00:28] <SpeedEvil> this the ngw100?
[00:28] <Laurenceb> yes
[00:28] <Laurenceb> theres uboot on the flash, it boots that then looks for the kernel
[00:29] <stilldavid> I've got a bit of experience with gumstix, which is very similar
[00:29] <Laurenceb> that I stuck on the flash as well
[00:29] <Laurenceb> then once its booted (~4 seconds) you can use the mmc card, so I had python on there
[00:30] Tigga (n=chatzill@dsl-217-155-6-122.zen.co.uk) left irc: Read error: 110 (Connection timed out)
[00:30] <stilldavid> yeah, I wound up getting a microsd card for my gumstix just for python
[00:30] <stilldavid> the build was ~7MB, and the gumstix only had 8MB flash :-/
[00:30] <Laurenceb> its not too hard as theres plenty of people to help you
[00:30] <Laurenceb> oh well I've gtg, cya
[00:31] <stilldavid> thanks! peace
[00:31] Laurenceb (n=laurence@host86-133-67-102.range86-133.btcentralplus.com) left irc: "The day microsoft make something that doesnt suck is the day they make a vacuum cleaner"
[00:54] edmoore (n=edmoore@pomegranate.chu.cam.ac.uk) left irc:
[01:25] bfirsh (n=ben@scooby.firshman.co.uk) left irc:
[07:41] jcoxon (n=jcoxon@host86-144-100-209.range86-144.btcentralplus.com) joined #highaltitude.
[07:46] Simon-MPFH (n=simon@phantom.mpfh.co.uk) joined #highaltitude.
[07:48] trialex (n=1033ECF9@124-170-218-242.dyn.iinet.net.au) joined #highaltitude.
[08:27] mc- (n=zsirc@cpc3-glfd1-0-0-cust439.glfd.cable.ntl.com) joined #highaltitude.
[08:37] edmoore (n=edmoore@pomegranate.chu.cam.ac.uk) joined #highaltitude.
[08:37] edmoore (n=edmoore@pomegranate.chu.cam.ac.uk) left irc: Client Quit
[08:57] jcoxon (n=jcoxon@host86-144-100-209.range86-144.btcentralplus.com) left irc: "Leaving"
[09:15] mc- (n=zsirc@cpc3-glfd1-0-0-cust439.glfd.cable.ntl.com) left irc: Read error: 110 (Connection timed out)
[09:49] icez (n=icez@unaffiliated/icez) left irc: Remote closed the connection
[10:15] edmoore (n=ed@pomegranate.chu.cam.ac.uk) joined #highaltitude.
[10:32] rjharrison (n=rharriso@gateway.hgf.com) joined #highaltitude.
[10:34] <SpeedEvil> hi edmoore rjharrison
[10:35] Action: SpeedEvil is leaving for passport interview soon.
[10:35] <SpeedEvil> Fun.
[10:35] <rjharrison> Hey SpeedEvil
[10:35] <rjharrison> It's not too bad
[10:35] <edmoore> hi SpeedEvil
[10:35] <SpeedEvil> rjharrison: no waterboarding then?
[10:35] <edmoore> SpeedEvil: going anywhere fun?
[10:35] <SpeedEvil> edmoore: alas no.
[10:35] <SpeedEvil> I actually don't need a passport.
[10:36] <SpeedEvil> I just need it to apply for a driving license.
[10:36] <edmoore> ah right
[10:36] <SpeedEvil> As it's been long enough that my provisional hasn't gotten my pic on it.
[10:36] <edmoore> then you can come to cambridge for a launch
[10:36] <SpeedEvil> Maybe.
[10:39] <SpeedEvil> Even at 40MPG, it's ~ 90 quid - assuming I sleep in the car. :)
[10:51] Laurenceb (n=laurence@host86-133-67-102.range86-133.btcentralplus.com) joined #highaltitude.
[10:59] <edmoore> SpeedEvil: was thinking more about your audio streaming idea for listeners
[10:59] <edmoore> if peeps could record the wav then send it through, say for transtantic, we could cross correlate all the recordings
[11:00] <SpeedEvil> yeah
[11:00] <SpeedEvil> phased array shit
[11:00] <edmoore> and as a sample of noise in theory has zero correlation with someone elses sample of noise, we should get some good signal out
[11:01] <edmoore> would be very useful for those early morning when readings when propagation is only just coming through
[11:01] <SpeedEvil> frequency skew makes it a bit annoying
[11:01] <SpeedEvil> Is this something that's really ever been done hamwise?
[11:03] <edmoore> I'm not sure it has
[11:06] mc- (n=mfcastle@cpc3-glfd1-0-0-cust439.glfd.cable.ntl.com) joined #highaltitude.
[11:09] <mc-> edmoore, a thought on frequency skew between different receivers
[11:10] <edmoore> that can be normalised, I'm pretty sure
[11:10] <mc-> if you AM modulate the transmitter, then frequency skew can be removed
[11:10] <mc-> ans=sqr(I*I+Q*Q)
[11:11] <mc-> need lots of filtering before
[11:12] <mc-> CW is a form of AM modulation
[11:12] <mc-> am I making sense?
[11:19] Tigga (n=chatzill@dsl-217-155-6-122.zen.co.uk) joined #highaltitude.
[11:25] shellevil (n=user@tor/regular/SpeedEvil) left irc: Read error: 113 (No route to host)
[11:36] <Laurenceb> nope
[11:36] <Laurenceb> how do you remove frequency squew?
[11:37] <Laurenceb> using AM modulation
[11:39] <mc-> if there's a frequency offset, then the sqr(II+QQ) removes it
[11:39] <mc-> and just returns the amplitude
[11:40] <Laurenceb> yes
[11:40] <Laurenceb> but how does that help
[11:41] <mc-> if receiver A is tuned to 100Hz and receiver B is tuned to 200Hz, then the signals can be combined using sqr(II+QQ)
[11:41] <Laurenceb> yeah but that leads to more noise entering the system
[11:41] <mc-> you need filters before the sqr process
[11:42] <mc-> so you know there's a signal within a 50Hz bandwidth, but not sure of the exact freq
[11:42] <Laurenceb> hmm
[11:43] <Laurenceb> well if you know where the signal is to apply the filter, youd be better shifting it
[11:43] <Laurenceb> tho that leads to a phase problem
[11:43] <mc-> yes, I'm assuming the signals are downconverted to baseband
[11:44] <mc-> there's no phase problem, if the signal is AM modulated.
[11:46] <Laurenceb> yeah
[11:47] <mc-> it might work well with HF signals since some receivers would get signals, while others get nothing. As the propagation changed, different rxers would hear signals.
[11:48] <Laurenceb> hmm but theres time syncronisation as well
[11:48] <mc-> yes, you would need a timestamp for each listener's file, so the files could be merged.
[11:49] <mc-> a GPS timestamp would be accurate enough
[11:50] <Laurenceb> from the station?
[12:00] <Laurenceb> hmm that places extra demands on who can listen in
[12:00] <Laurenceb> how about dividing time up into bins of 30s or so, then finding the frequency with max intensity, then trying to correlate with other stations in time space
[12:01] <Laurenceb> there may be a way to do it with only rought time, e.g. overlapping 30s bins
[12:01] <Laurenceb> say a 30s bin starting every 15s
[12:02] sbasuita (n=sbasuita@unaffiliated/drebellion) joined #highaltitude.
[12:03] <Laurenceb> then you search for the frequency offset using amplitude of each frequency bin, then shift the frequency to some constant IF
[12:04] <Laurenceb> frequency offset drift is slow so you could use a filtering technique to track it better
[12:05] <Laurenceb> once you get 30s bins of observations at the constant IF, you correlate them
[12:05] <mc-> You don't know the transmitter frequency offset if it's on a balloon
[12:06] <mc-> so you need to have a 50Hz or more bandwidth
[12:06] <Laurenceb> that stage is very computationally demanding, so you could start with the three highest intensity stations
[12:06] <Laurenceb> yeah
[12:06] <mc-> how do you know the highest intensity station, if there's only noise
[12:07] <Laurenceb> you have to have a strong enough signal that a 30s integration will reveal it
[12:07] <Laurenceb> - well 30s FFT
[12:07] <mc-> if you have a signal from a single station, then you don't need distributed listeners...
[12:07] <Laurenceb> nope
[12:07] <Laurenceb> as 1/30 baud is a tad slow
[12:08] <mc-> implementing sqr(II+QQ) on the server for all the incoming data should be simple.
[12:08] <Laurenceb> I'm not convinced it helps
[12:08] <Laurenceb> of course there is a simple way to do this
[12:09] <Laurenceb> generate a PRN code from the balloon and transmit it with the data
[12:09] <mc-> a few lines of matlab code would prove it
[12:09] <Laurenceb> then you just do a time freuqency search for the prn code in each stations data
[12:10] <mc-> that's solving the problem with a single listener
[12:10] <Laurenceb> nope
[12:10] <Laurenceb> as a prn code would span an entire packet
[12:12] <Laurenceb> the only tricky bit is how to sent it at the same time as the data
[12:12] <mc-> My simulation is: create 2 sine waves (at 10Hz and 11Hz) which are both modulated with a 1Hz on-off pulse
[12:13] <mc-> then pass them through sqr(I*I+Q*Q)
[12:13] <mc-> do the same test with a 1 sine wave, and compare the amplitude
[12:14] <Laurenceb> right
[12:15] <Laurenceb> anyone know how rocketboys HF radio is modulated?
[12:15] <mc-> alternating between RTTY CW and MFSK over a few minutes
[12:15] <Laurenceb> right
[12:16] <Laurenceb> so RTTY is AM and MFSK is?
[12:16] <mc-> RTTY and MFSK are both FSK
[12:17] <Laurenceb> ah ok
[12:17] <Laurenceb> so you receive with ssb ?
[12:17] <mc-> yes
[12:18] <Laurenceb> nice
[12:19] <mc-> alternating between RTTY and CW and MFSK over a few minutes - so there's no confusion
[12:19] <Laurenceb> hmm theres got to be a way to transmit PRN at the same time
[12:19] <Laurenceb> maybe amplitude modulation
[12:20] <mc-> a PRN signal would be take up lots of space in the HF band
[12:20] <Laurenceb> not if its al low baud rate
[12:21] <Laurenceb> ok if you have 4 level FSK
[12:21] <Laurenceb> that would probably work more easily
[12:22] <Laurenceb> but for MFSK 2 levels is taking it a bit far
[12:23] <Laurenceb> you could also double the baudrate
[12:23] <Laurenceb> or use amplitude modulation for the prn
[12:25] <Laurenceb> for RTTY it would work well to use a 4 level fsk scheme
[12:26] <Laurenceb> however... you want this to work with existing spftware like fldigi, so amplitude modulation would prob be best - but would involve hacking the transmitted
[12:28] <Laurenceb> - as its already built and it'd be a bit of a pain to have to modify it
[12:31] <mc-> if CW signals are sent, then the transmitter doesn't need modifying
[12:32] <Laurenceb> I'm guessing its just on/off
[12:32] <mc-> yes
[12:32] <Laurenceb> hmm maybe you could use pwm or something
[12:32] <Laurenceb> havent got a schematic to see
[12:32] <mc-> the rockmite has CW shaping
[12:33] <Laurenceb> is there a schematic?
[12:34] <mc-> should be something on the web, search for rockmite
[12:35] <Laurenceb> http://www.w5usj.com/images/rockmite/rockmite.gif
[12:36] <Laurenceb> pwm into shift may work
[12:37] <mc-> doesn't seem to be much shaping in the circuit, but assume it's enough.
[12:37] <Laurenceb> if its fast enough
[12:38] <Laurenceb> not sure if it would work with the pic
[12:39] <Laurenceb> but amplitude mod with prn would surely be the best way
[12:39] <mc-> the shift input adjust the freq between tx and rx
[12:41] <Laurenceb> whats the T-R line for?
[12:42] <mc-> to switch between transmit / receive
[12:42] <Laurenceb> yeah, so if you can control that you can adjust the amplitude
[12:44] <Laurenceb> surely the push switch controls that
[12:49] <Laurenceb> youd still need to do phase correction before combining the signal
[12:49] <Laurenceb> maybe have a c program that crancks out wav files
[12:52] <Laurenceb> right I've gtg, cya
[12:52] Laurenceb (n=laurence@host86-133-67-102.range86-133.btcentralplus.com) left irc: "The day microsoft make something that doesnt suck is the day they make a vacuum cleaner"
[12:54] DanielRichman (n=DanielRi@89.243.213.125) joined #highaltitude.
[12:57] trialex (n=1033ECF9@124-170-218-242.dyn.iinet.net.au) left irc: Read error: 104 (Connection reset by peer)
[13:08] <mc-> I wrote a few lines to test sqr(II+QQ) with 2 listeners on different freq, and it works
[13:09] mc- (n=mfcastle@cpc3-glfd1-0-0-cust439.glfd.cable.ntl.com) left irc:
[13:26] rjharrison (n=rharriso@gateway.hgf.com) left irc:
[13:44] rjharrison (n=rharriso@gateway.hgf.com) joined #highaltitude.
[14:24] edmoore_ (n=ed@pomegranate.chu.cam.ac.uk) joined #highaltitude.
[14:24] edmoore (n=ed@pomegranate.chu.cam.ac.uk) left irc: Read error: 104 (Connection reset by peer)
[14:38] rjharrison (n=rharriso@gateway.hgf.com) left irc:
[16:17] jcoxon (n=jcoxon@host86-144-100-209.range86-144.btcentralplus.com) joined #highaltitude.
[16:17] <edmoore_> good afternoon jcoxon
[16:18] <jcoxon> good afternoon edmoore_
[16:19] <edmoore_> my revision is going swimmingly well
[16:19] <edmoore_> managed to spend 1hour on one part of a question thatnks to a wikipedia wonder. I give up
[16:19] Nick change: edmoore_ -> edmoore
[16:20] <edmoore> all well with you?
[16:20] <jcoxon> yeah
[16:20] <jcoxon> just started my new rotation - catching babies
[16:20] <jcoxon> not that i've seen any patients - easter is messing up the timetable
[16:25] <jcoxon> edmoore, how long till exams?
[16:25] <edmoore> 2 weeks
[16:25] <edmoore> catching babies!?
[16:26] <jcoxon> obs and gynae
[16:27] <edmoore> i think I can figure that one out
[16:27] <jcoxon> :-)
[16:29] <jcoxon> i've been thinking, looking at the ARHAB flights there is a PBH-9 launch on 18/04/09
[16:29] <jcoxon> might be a good chance to test teh DL Client on a HF, long range mission
[16:29] <jcoxon> its a 50 hour duration ZP flight
[16:30] <edmoore> from where to where roughly?
[16:30] <jcoxon> and will be transmitting on 30m and 40m
[16:30] <jcoxon> launching from New York State
[16:32] <edmoore> do they have a direction in mind?
[16:32] <jcoxon> that i don't know
[16:32] <edmoore> cool
[16:32] <edmoore> well I'll be in exam panic mode for each of those 50 hours
[16:32] <jcoxon> i think its a duration attempt rather then trans-a
[16:32] <jcoxon> fair enough
[16:33] <edmoore> right, tesco time
[16:33] <edmoore> bbl
[16:34] edmoore (n=ed@pomegranate.chu.cam.ac.uk) left irc:
[16:56] sbasuita (n=sbasuita@unaffiliated/drebellion) left irc: "Leaving"
[17:10] DanielRichman (n=DanielRi@89.243.213.125) left irc: "Leaving"
[17:39] sbasuita (n=sbasuita@unaffiliated/drebellion) joined #highaltitude.
[17:54] Tigga (n=chatzill@dsl-217-155-6-122.zen.co.uk) left irc: Read error: 104 (Connection reset by peer)
[17:54] sbasuita (n=sbasuita@unaffiliated/drebellion) left irc: "Leaving"
[18:05] Hiena (n=Hiena@81.93.195.181.datatrans.hu) joined #highaltitude.
[18:08] rjharrison (n=rharriso@80.4.177.225) joined #highaltitude.
[18:09] edmoore (n=edmoore@pomegranate.chu.cam.ac.uk) joined #highaltitude.
[18:09] edmoore (n=edmoore@pomegranate.chu.cam.ac.uk) left irc: Client Quit
[18:28] rjharrison (n=rharriso@80.4.177.225) left irc:
[18:47] icez (n=icez@unaffiliated/icez) joined #highaltitude.
[19:40] rjharrison (n=rharriso@80.176.172.227) joined #highaltitude.
[20:06] natrium42 (n=alexei@CPE000625d867e2-CM0014045885be.cpe.net.cable.rogers.com) left irc:
[20:08] edmoore (n=edmoore@pomegranate.chu.cam.ac.uk) joined #highaltitude.
[20:24] Tigga (n=chatzill@dsl-217-155-6-122.zen.co.uk) joined #highaltitude.
[20:27] edmoore (n=edmoore@pomegranate.chu.cam.ac.uk) left irc:
[20:27] Hiena (n=Hiena@81.93.195.181.datatrans.hu) left irc: "-=Mi ez a szag? Hippit égettéll?=-"
[20:30] edmoore (n=ed@pomegranate.chu.cam.ac.uk) joined #highaltitude.
[20:38] Laurenceb (n=laurence@host86-133-67-102.range86-133.btcentralplus.com) joined #highaltitude.
[20:38] <Laurenceb> hello
[20:38] <Laurenceb> was anything worked out wrt multiple stations and wav files?
[20:38] <Laurenceb> brb
[20:45] <rjharrison> not yet
[20:45] <rjharrison> I'm working on capturing all raw data
[20:45] <rjharrison> This way people can do eyeball debug
[20:45] <edmoore> i think a good enough time sync will be the issue
[20:46] <edmoore> Laurenceb: you could of course cross-correlate the amples together in time to sync them, but then you have a bit of a chicken/egg thing
[20:46] <rjharrison> I don't think computers are going to be good at decyphering the bad from the good
[20:46] <rjharrison> Well not by my programming
[20:47] <edmoore> we'd get 9dB from 8 receivers if we got it right
[20:47] <rjharrison> For the time when it's going to be usefull I think keeping a central log of all the raw data is going to be the easiest way forward
[20:48] <rjharrison> It's a whole different matter if we are going to capture the raw wave files and sync them and then decode
[20:48] <Laurenceb> edmoore: yeah, once you go past three or so it gets very demanding
[20:49] <edmoore> rjharrison: that would probably have to run on its own server
[20:49] <Laurenceb> fiding a way to embed PRN in the signals would be a lot easier
[20:49] <edmoore> and would be more than a php job
[20:49] <edmoore> :)
[20:49] <rjharrison> Though I can see that this would be usefull but quite frankly the coms system needs an upgrade if it's going to perform that poorly
[20:49] <Laurenceb> gives you something to search for in time frequency space
[20:49] <edmoore> it's a systems thing
[20:50] <edmoore> sometime you have finite power on the balloon
[20:50] <Laurenceb> of course you could switch between PRN and data
[20:50] <Laurenceb> over a ~20 second timescale
[20:51] <rjharrison> Yep but I would still design a coms system such that it's fit for purpose
[20:51] <edmoore> Laurenceb: was wondering if you could modulate your bitstream ontop of something much slower
[20:51] <edmoore> so maybe a 5-baud thing that just sent you the gps loaction over the course of 30 secs
[20:51] <Laurenceb> yeah, but that would lead to uncertainty
[20:51] <Laurenceb> to your position in the correlation peak
[20:51] <rjharrison> Could't you pass it out at a higher frequency
[20:52] <Laurenceb> it needs to be of the same order as the baud rate
[20:52] <Laurenceb> but the actual code lenght can be very long
[20:52] <edmoore> I don't see that that's the case
[20:53] <edmoore> although actually a radiometrix might struggle to do it
[20:53] <edmoore> hmm
[20:53] <Laurenceb> you have frequency drifts that mean your prn code cant be too long in time
[20:53] <Laurenceb> maybe 20 or 30s duration
[20:53] <rjharrison> OK what is prn again
[20:53] <rjharrison> It's in GPS
[20:53] <Laurenceb> and unless you want a second level of searching you want your prn peak to localise something like the first bit of the packet
[20:54] <rjharrison> Psudo random number?
[20:54] <edmoore> yep
[20:54] <Laurenceb> rjharrison: its a "gold code" that has a single peak in its autocorrelation function
[20:54] <rjharrison> Basically that math suff you did for GPS
[20:54] <rjharrison> decode
[20:54] sbasuita (n=sbasuita@unaffiliated/drebellion) joined #highaltitude.
[20:55] <Laurenceb> hehe yes its like gps
[20:55] <Laurenceb> you find a way to embed this into the transmission and you can do a search for each station and find its time and frequency offsets
[20:57] <Laurenceb> then you can add the stations coherently
[20:57] <Laurenceb> the embedding might take the form of +-20% or something amplitude modulation
[20:58] <Laurenceb> or we could double the baudrate or the number of frequency levels
[20:58] <Laurenceb> but those approaches are harder and make it noncompatible with fldigi
[20:59] <edmoore> gnuradio might be a nice framework in which to try some of this
[20:59] <Laurenceb> hehe yeah
[21:00] <Laurenceb> if we all had a USRP
[21:00] <Laurenceb> grrr I should make that usb stick SDR
[21:00] <Laurenceb> I've got as far as paper schematics
[21:01] shellevil (n=user@tor/regular/SpeedEvil) joined #highaltitude.
[21:01] <Laurenceb> mixture of AD, linear, cypress and Ti products... it got rather complicated keeping good performance whilst having very large operational frequency range
[21:02] <edmoore> Laurenceb: i mean the softeware
[21:02] <edmoore> it's a decent framework
[21:02] <Laurenceb> yeah suppose so
[21:02] <Laurenceb> this sort of thing isnt exactly hard
[21:03] <Laurenceb> just throw an fft library at it
[21:03] <shellevil> hlo
[21:03] <shellevil> this the multi-reciever idea?
[21:03] <shellevil> or something else
[21:03] <Laurenceb> multi-receiver yes
[21:04] <Laurenceb> I'm not sure how practical AM is for the PRN
[21:12] <Laurenceb> also if your doing coherent integration, you wnat to know the receiver coherency period
[21:12] <Laurenceb> as phase changes will screw up the integration
[21:28] <SpeedEvil> FWIW - had a talk to someone that was managing the cell-tower nets for t-mobile. - one of the aforementioned cousins. There shouldn't be any consequences ohter than not getting signal if the phone goes too high - not going to ban your SIM forex. Just not connect.
[21:28] RocketBoy (n=Steve@217.47.75.8) joined #highaltitude.
[21:31] Simon-MPFH (n=simon@phantom.mpfh.co.uk) left irc: "Leaving"
[21:31] natrium42 (n=alexei@CPE000625d867e2-CM0014045885be.cpe.net.cable.rogers.com) joined #highaltitude.
[21:32] <RocketBoy> Does anyone know what this type of connector is called:- http://imagebin.org/44688 - its sort of latching 0.1"
[21:37] hallam (i=126f3681@gateway/web/ajax/mibbit.com/x-1a24d35325e27b23) joined #highaltitude.
[21:46] <edmoore> re-ask your question RocketBoy
[21:46] <edmoore> I know hallam likes his connectors
[21:47] <hallam> RocketBoy: got a photo from another angle?
[21:50] <SpeedEvil> RocketBoy: and what's it in?
[21:53] edmoore_ (n=edmoore@pomegranate.chu.cam.ac.uk) joined #highaltitude.
[21:53] edmoore (n=ed@pomegranate.chu.cam.ac.uk) left irc: Read error: 104 (Connection reset by peer)
[21:56] <RocketBoy> Re Q: Does anyone know what this type of connector is called:- http://imagebin.org/44688 - its sort of latching 0.1"
[21:56] <RocketBoy> its on a fan
[21:57] <SpeedEvil> what do you want to do?
[21:57] <RocketBoy> find a plug for it
[21:57] <RocketBoy> sorry no other photos at the mo
[21:58] <SpeedEvil> why?
[21:58] edmoore_ (n=edmoore@pomegranate.chu.cam.ac.uk) left irc: Read error: 104 (Connection reset by peer)
[21:58] <SpeedEvil> Pair of scissors, job done
[21:58] <RocketBoy> its for someone elses project - they have lots of these fans
[21:58] <SpeedEvil> looks like it fits to 0.1" header - is that the right spacing?
[21:59] <SpeedEvil> though the lock won't of course work
[21:59] <hallam> I don't immediately recognise it, so best I can suggest is to trawl through the samtec and molex websites
[21:59] edmoore (n=ed@pomegranate.chu.cam.ac.uk) joined #highaltitude.
[21:59] <SpeedEvil> digikey catalog - the actual physical one - or the shitty electronic flash copy
[21:59] <RocketBoy> I think so but it would be good if it latched
[21:59] <hallam> samtec have an excellent free sample service
[21:59] <SpeedEvil> and look for a similar shape
[21:59] <hallam> maybe actually emailing the photo to samtec or molex would get results
[22:00] <edmoore> latching connectors are good
[22:00] <edmoore> lets start some ballooning axioms
[22:01] <RocketBoy> There are some comming to me shortly so I'll get a better idea then
[22:03] <SpeedEvil> http://www.linuxdevices.com/news/NS3314447002.html
[22:03] <SpeedEvil> interesting
[22:06] <SpeedEvil> http://bifferos.bizhat.com/
[22:06] <SpeedEvil> 26 quid 486/133 with USB
[22:07] <SpeedEvil> And NEEDS MOAR GPIO!
[22:08] <SpeedEvil> Otherwise - very nearly ideal for some 'if only I could just run linux' things
[22:08] <jcoxon> evening all
[22:09] <SpeedEvil> evening.
[22:11] <hallam> hi James
[22:13] <jcoxon> hey henry
[22:17] sbasuita (n=sbasuita@unaffiliated/drebellion) left irc: Read error: 104 (Connection reset by peer)
[22:22] <Laurenceb> I'll have to try samtec
[22:23] <Laurenceb> I've been using toby.co.uk , searching around on farnell for connectors can take all day
[22:30] edmoore_ (n=ed@pomegranate.chu.cam.ac.uk) joined #highaltitude.
[22:30] edmoore (n=ed@pomegranate.chu.cam.ac.uk) left irc: Read error: 104 (Connection reset by peer)
[22:31] <Laurenceb> that is crazily cheap
[22:31] <SpeedEvil> the above board?
[22:31] <SpeedEvil> yeah
[22:31] <Laurenceb> I cant make out the pcb, is there a large heatsink?
[22:31] <SpeedEvil> seems to be a small one, yes
[22:31] <Laurenceb> although I paid £28 for an NGW100
[22:31] <SpeedEvil> it's large in context to the PCB - but
[22:32] <SpeedEvil> that's about 30*60mm
[22:33] <SpeedEvil> I wonder about losing the I/O board, and the resultant cost.
[22:38] <Laurenceb> the website is a bit basic
[22:39] <Laurenceb> but they probably spend more time on more useful things :P
[22:44] <stilldavid> can anyone give me some background on how the balloon trajectory programs work?
[22:45] <SpeedEvil> Basically
[22:45] <SpeedEvil> it takes weather forecast models for the future
[22:46] <SpeedEvil> and simulates how a massless particle is blown around by the wind
[22:46] <stilldavid> I guess I'm just looking for a link to the forecasts, just to see how that's translated into the results
[22:48] Action: SpeedEvil doesn't remember - ask again tomorrow or in a bit.
[22:48] <Tigga> http://nomads.ncep.noaa.gov/
[22:48] <Tigga> stilldavid:
[22:48] <SpeedEvil> or look there.
[22:49] <stilldavid> Tigga: perfect, thanks!
[22:49] <edmoore_> pluck a simulation step out of your bum, eg 1hz. then just numerically integrate your position by looking up what the wind is doing. The end.
[22:50] <SpeedEvil> Laurenceb: bifferboard - USB hub - USB soundcard (to modem) - USB SD adaptor - couple of USB mice with new lenses, ...
[22:51] <Laurenceb> oh god
[22:51] <Laurenceb> think is real time control with linux just isnt going to happen
[22:51] Action: SpeedEvil has done real-time DSP in awk.
[22:52] <SpeedEvil> And as long as you're not running multiple processes, or no carefully vetted processes - it's not bad at all.
[22:52] <Laurenceb> if you want a control loop, its not too easy
[22:52] <Laurenceb> hmm
[22:52] <SpeedEvil> Maybe.
[22:53] <SpeedEvil> It's just a bit tempting in some ways.
[22:53] <Laurenceb> say you want a control loop running every 20ms
[22:53] <Laurenceb> how do you make sure its 20ms +-1% or something
[22:53] <SpeedEvil> timers, yes.
[22:53] <Laurenceb> or lack therof in linux
[22:54] <hallam> you can do realtime-ish stuff, how do you think fldigi etc work?
[22:54] <SpeedEvil> The hardware is of course trivially capable of it.
[22:54] <SpeedEvil> Configuring the software is the fun part.
[22:54] <Laurenceb> fldigi has buffered soundcard data
[22:55] <edmoore_> what laurenceb says
[22:55] <Laurenceb> that solves the real time issue
[23:01] <hallam> hm, I see
[23:01] <hallam> but what stops you just doing one loop iteration then sleeping until the global millisecond timer ticks to whatever you want?
[23:02] <SpeedEvil> Nothing.
[23:03] <hallam> I guess if the scheduler isn't fine-grained enough there'll be jitter
[23:04] <Laurenceb> yeah, typically its of order 10ms
[23:08] <Laurenceb> http://www.neogaf.com/forum/showthread.php?t=307605
[23:09] <SpeedEvil> italy?
[23:10] <Laurenceb> I'll have to fire up MIDAS
[23:10] <SpeedEvil> ...
[23:10] <SpeedEvil> the GPS error residual thigny?
[23:11] <Laurenceb> http://www.cosis.net/abstracts/EGU04/00336/EGU04-A-00336.pdf
[23:16] <Laurenceb> it reconstructs 3D ionospheric structure from GPS TEC measurements
[23:19] <hallam> later!
[23:19] hallam (i=126f3681@gateway/web/ajax/mibbit.com/x-1a24d35325e27b23) left irc: "http://www.mibbit.com ajax IRC Client"
[23:21] <Laurenceb> http://hal.archives-ouvertes.fr/docs/00/29/94/51/PDF/nhess-7-543-2007.pdf
[23:21] <Laurenceb> hmm that paper actually looks convincing... maybe
[23:36] <Laurenceb> gtg, cya
[23:37] Laurenceb (n=laurence@host86-133-67-102.range86-133.btcentralplus.com) left irc: "The day microsoft make something that doesnt suck is the day they make a vacuum cleaner"
[23:38] shellevil (n=user@tor/regular/SpeedEvil) left irc: Read error: 110 (Connection timed out)
[23:38] SpeedEvil (n=jfjfjf@tor/regular/SpeedEvil) left irc: Read error: 110 (Connection timed out)
[23:49] RocketBoy (n=Steve@217.47.75.8) left irc: Read error: 110 (Connection timed out)
[23:57] jcoxon (n=jcoxon@host86-144-100-209.range86-144.btcentralplus.com) left irc: "Leaving"
[23:58] edmoore_ (n=ed@pomegranate.chu.cam.ac.uk) left irc: Read error: 113 (No route to host)
[00:00] --- Wed Apr 8 2009