highaltitude.log.20060912

[00:04] <rikstah> how did the weekend launches go?
[00:04] <jcoxon> 2 payloads were launched
[00:04] <jcoxon> Nova 1 (which HenryH is part of the team)
[00:05] <rikstah> how high and stuff? any pics?
[00:05] <jcoxon> and MiHAB 2
[00:05] <rikstah> im thinking about trying it when i get home to england
[00:05] <jcoxon> Nova 1 made 105,000ft
[00:05] <jcoxon> and MiHAB 2 76,000ft
[00:05] <rikstah> sweet!
[00:05] <jcoxon> http://www.srcf.ucam.org/~hmh33/cuspaceflight/nova1launch.html
[00:06] <rikstah> thats awesome, congrats
[00:06] <rikstah> i find it completely fascinating that you can see so high over the horizon!
[00:07] <jcoxon> was a really good weekend for launches
[00:07] <jcoxon> wish i had a payload ready
[00:07] <jcoxon> oh well :-D
[00:07] <rikstah> is there anyone doing this stuff near manchester ?
[00:08] <jcoxon> hmmmm, not that i know of
[00:08] <HenryH> thanks for the kind words :)
[00:08] <HenryH> don't know of anyone up that way but you're welcome to come to one of our launches and/or use our CAA waiver
[00:09] <rikstah> HenryH, pleasure, it's fascinating
[00:09] <rikstah> CAA waiver?
[00:09] <HenryH> you need / it is a very good idea to get permission from the CAA to launch a balloon
[00:09] <rikstah> i only made 13,000 ft yesterday when i jumped out of a plane :)
[00:09] <rikstah> oh i see
[00:09] <HenryH> nice :) where did you do that?
[00:10] <rikstah> here in seattle, WA
[00:10] <rikstah> done it a few times, been learning to get qualified
[00:10] <HenryH> would you recommend going to the US to learn?
[00:10] <rikstah> yesterday i was doing my own turns and stuff, and pulled my own chute
[00:10] <HenryH> sweet
[00:11] <rikstah> well, it's cheaper :) but its just because im out here working right now
[00:11] <rikstah> so how much is a 800g balloon
[00:12] <jcoxon> rikstah, cost?
[00:12] <rikstah> yeah
[00:12] <jcoxon> £25
[00:12] <rikstah> oh sweet!!
[00:14] <rikstah> is it safe t just unplug the gumstix
[00:15] <rikstah> aha! no need
[00:15] <rikstah> i can connect to it now :D :D
[00:15] <rikstah> sweet!!
[00:15] <jcoxon> hehe
[00:15] <rikstah> jcoxon, you were right, just needed to enable a dhcp server on the laptop
[00:16] <rikstah> oh that's sweeeet
[00:16] <rikstah> i just used sweet 3 times in 2 minutes, damn...americans polluting my brain
[00:16] <jcoxon> yup
[00:16] <jcoxon> i'd worry about that if i were you!
[00:16] <jcoxon> right i'm gonna have to go to bed
[00:16] <jcoxon> night all
[00:16] <rikstah> mate, i'm finding myself saying elevator and stuff
[00:16] <rikstah> hehe
[00:17] <rikstah> ok gnite
[00:17] jcoxon (n=jcoxon@81-178-160-64.dsl.pipex.com) left irc: "Leaving"
[00:56] <HenryH> night
[00:56] HenryH (n=H@natreg.pem.cam.ac.uk) left irc: "I can't believe what a lame quit message I had before"
[02:05] defy (i=defy@port-60-234-107-209.orcon.net.nz) joined #highaltitude.
[05:13] Nick change: rikstah -> riksta
[05:13] riksta (n=rick@c-71-227-234-92.hsd1.or.comcast.net) left #highaltitude ("Leaving").
[05:13] riksta (n=rick@c-71-227-234-92.hsd1.or.comcast.net) joined #highaltitude.
[05:13] defy (i=defy@port-60-234-107-209.orcon.net.nz) left irc: Read error: 110 (Connection timed out)
[06:13] icez (n=icez@ip68-3-56-121.ph.ph.cox.net) left irc: Remote closed the connection
[08:46] jcoxon (n=jcoxon@81-178-160-64.dsl.pipex.com) joined #highaltitude.
[14:01] icez (n=icez@ip68-3-56-121.ph.ph.cox.net) joined #highaltitude.
[15:26] icez (n=icez@ip68-3-56-121.ph.ph.cox.net) left irc: Remote closed the connection
[16:46] rocketboy (n=steve@217.47.75.5) joined #highaltitude.
[17:12] phatmonkey (i=nobody@81.2.121.150) joined #highaltitude.
[17:30] <phatmonkey> http://www.sparkfun.com/commerce/product_info.php?products_id=7917
[17:30] <phatmonkey> wow!!
[17:33] <rocketboy> seems to have everything
[17:34] <jcoxon> makes us pretty useless hehe
[17:35] <jcoxon> just wiring up my i2c compass
[17:35] <phatmonkey> that'd make a really nice tiny payload
[21:31] Action: jcoxon has got his digital compass work :-)
[21:32] <rocketboy> spiffing
[21:32] <jcoxon> i2c is a very simple interface
[21:33] <rocketboy> I havn't looked at it much - just a couple of wires - clock and data?
[21:33] <jcoxon> yup
[21:34] <jcoxon> and its on linux its easy to access
[21:34] <jcoxon> (though i did cheat as i found the code for the compass)
[21:34] <rocketboy> ah
[21:34] <jcoxon> oh laurence's photos are here: http://www.pegasushabproject.org.uk/mihab/
[21:34] <jcoxon> but actuallly its a matter of write() and read() of hte device file
[21:40] <rocketboy> so does it actually give a compass bearing?
[21:41] <jcoxon> yeah
[21:41] <jcoxon> the software outputs a number
[21:41] <jcoxon> i haven't calibrated it say to north
[21:41] <rocketboy> sort of 0 to 359?
[21:41] <jcoxon> yeah
[21:42] <rocketboy> how fast?
[21:42] <jcoxon> the software was getting it at about twice a second
[21:42] <rocketboy> ah - I thought it was faster than that
[21:43] <jcoxon> well its the software which is controlling it
[21:43] <jcoxon> "usleep(200000);"
[21:43] <rocketboy> ok - how fast do you think you will need?
[21:44] <jcoxon> that i'm not sure
[21:44] <jcoxon> i reckon Peg IV will just record the bearing
[21:44] <jcoxon> i was thinking of adapting the code
[21:45] <jcoxon> say to timestamp every minute
[21:45] <jcoxon> then record as quick as possible the output of the compass
[21:46] <rocketboy> it looked to me that the payload might have been spining quite fast in places - say as much as once every few seconds?
[21:47] <jcoxon> what do you think is the best method of record that?
[21:49] <rocketboy> either calculate the rate of spin from the data and then record that once per min or occasionally record the full data - say 10 consectutive readings per minute
[21:50] HenryH (n=H@natreg.pem.cam.ac.uk) joined #highaltitude.
[21:50] <jcoxon> could do both - 10 mins of full data recording
[21:50] <HenryH> evening
[21:50] <jcoxon> then 10 mins of calculations
[21:50] <jcoxon> i've got hte processing power
[21:50] <jcoxon> hey HenryH
[21:53] <jcoxon> got my i2c digital compass today
[21:54] <HenryH> cool
[21:54] <HenryH> how big is it?
[21:55] <jcoxon> 30mm x 30mm
[21:56] <HenryH> ouch
[21:56] <HenryH> that's nearly as big as the whole of the version 3 tracker
[21:57] <HenryH> is it very precise then?
[21:57] <jcoxon> tracker v3 is 30mm by 30mm!!!
[21:57] <jcoxon> its not that precise
[21:57] <jcoxon> i guess most of it is the i2c interfacing
[21:58] <jcoxon> the actually mag field detectors are tiny
[22:00] <HenryH> 38x38
[22:00] <jcoxon> wow thats small
[22:00] <HenryH> it may grow ;) I'm producing it in several versions
[22:01] <HenryH> single radio, dual radio, radio + nokia phone, radio + gsm/gps module
[22:01] <jcoxon> cool
[22:01] <jcoxon> is this to be used in all 3 projects then
[22:05] <jcoxon> rocketboy, HenryH where do you order your electrical parts from
[22:05] <jcoxon> ?
[22:07] <rocketboy> mostly rs (rswww.com), rapid(rapidonine.com), farnell (uk.farnell.com/)
[22:08] <HenryH> same here
[22:09] <HenryH> some combination will be used in the various projects but I expect it will evolve a bit
[22:09] <jcoxon> think i'm ready to move away from maplin
[22:09] <HenryH> oh god not maplin
[22:09] <jcoxon> HenryH, right
[22:09] <jcoxon> hehe
[22:09] <jcoxon> only cause its accessible
[22:10] <rocketboy> yeah they have got that going for them - but they are expensive and carry poor quaility components
[22:11] <jcoxon> yeah i've learnt that
[22:11] <rocketboy> fortunatly I have rapid only about 30mins drive away - which is just as close as the nearest maplins
[22:11] <jcoxon> wheres your rapid (considering i'm in suffolk
[22:11] <jcoxon> )
[22:12] <rocketboy> its near colchester - just off the A12
[22:12] <jcoxon> oh right
[22:13] <HenryH> rapid do next day delivery for free though
[22:14] <rocketboy> http://www.streetmap.co.uk/newmap.srf?x=600815&y=229134&z=0&sv=CO4+5JS&st=2&pc=CO4+5JS&mapp=newmap.srf&searchp=newsearch.srf
[22:14] <jcoxon> oh right
[22:15] <rocketboy> yeah but if you need it right *now* go to their trade counter and they pick it while you wait
[22:15] <jcoxon> yeah i know roughly where that is
[22:15] <jcoxon> i'm in no hurry
[22:38] <HenryH> Steve, how possible is it to connect two antennae to one of those radiometrix modules?
[22:38] <HenryH> will it decrease the signal strength by just half, or more than half?
[22:39] <rocketboy> to do it properly you either need a RF (co-ax) relay - or a power splitter circuit
[22:39] <rocketboy> if you just put the 2 antenna together you will get an impedance mis-match
[22:39] <HenryH> the relay would allow you to swich between them? that could work too
[22:40] <HenryH> ok
[22:40] <HenryH> so how complex would a power splitter circuit be?
[22:40] <rocketboy> yes the relay allows you to switch
[22:40] <HenryH> that's probably the best bet then
[22:41] <HenryH> much power loss in those?
[22:41] <rocketboy> yes if its a good rf switch
[22:41] <rocketboy> a good switch will be down at 0.1db level
[22:41] <HenryH> ok cool
[22:42] <HenryH> so you'd probably leave it transmitting on one antenna for the flight then on touchdown it would either sense which way up it was and use the best antenna, or just alternate between them
[22:42] <HenryH> I just think that having two separate transmitters seems a bit wasteful
[22:43] <HenryH> reading some interesting stuff on turbo codes
[22:43] <rocketboy> yeah - its a fine balance though rf relays don't come cheap
[22:43] <HenryH> my compsci friends think they can write a decoder program
[22:43] <HenryH> ah
[22:43] <jcoxon> hehe have a flipper mechanism like the robots in robot wars :-p
[22:43] <HenryH> hah
[22:43] <HenryH> if the rf relay is bigger and heavier....
[22:44] <HenryH> btw
[22:44] <rocketboy> exactly - they can be quite small though - about half the size of the circuit design module
[22:44] <HenryH> not bad
[22:45] <HenryH> could you theoretically do FSK with the circuit design transmitter and have two separate receivers tuned to each of the frequencies?
[22:45] <HenryH> not that I think that's a good idea really
[22:45] <rocketboy> yes I was thinking a dedicated decoder (auto tracking) would be a good software project for someone
[22:46] <rocketboy> I don't think it would need to much processing power - perhaps a PIC18F could just about do it
[22:46] <HenryH> for turbo codes do you mean?
[22:46] <rocketboy> the 18F have an 8x8 multiplier dont they?
[22:46] <HenryH> yes
[22:46] <HenryH> from what I've read the encoding isn't too heavy on the processor, the decoding is quite taxing but that can be done on a laptop
[22:47] <HenryH> I'm thinking of moving to the 16-bit PICs anyway
[22:47] <HenryH> the only thing holding me back is that they consume about 5 more mA
[22:47] <HenryH> I'd quite like to do turbo codes just for the sake of it really, they look interesting
[22:48] <rocketboy> I don't think the two recivers is that practical (it could be done though)
[22:48] <HenryH> yeah it's not worth it just to save a couple of cubic cm
[22:48] <HenryH> maybe if we had receivers out the wazoo
[22:50] <rocketboy> the best way is to give the recieve channel specturm to a DSP (rather than going through the audio)
[22:51] <HenryH> the direct RF spectrum or the audio output from the receiver?
[22:53] <HenryH> does that mean the DSP would need to take samples at some multiple of 434MHz? that's some DSP
[22:54] <rocketboy> No - you would use an IF output (typically at 10.7MHz or more probably at 455KHz) and output that to the DSP
[22:55] <rocketboy> - its like taking a normal reciver and chopping the detector and audio stages off
[22:56] <rocketboy> so the DSP would need to sample at say 1MHz for a 455KHz IF
[22:56] <HenryH> ok that's more doable
[22:57] <HenryH> I think I need to read more about the IF
[22:57] <rocketboy> it would need some grunt - but 300 and 400MHz DSPs are arround
[22:58] <rocketboy> - its just a way of taking a RF signal and dropping it down (using mixing) to a more managable frequency
[22:59] <HenryH> right I'm looking at a diagram for a superheterodyne receiver
[22:59] <HenryH> it's starting to make sense
[22:59] <HenryH> http://en.wikipedia.org/wiki/Superheterodyne
[22:59] <HenryH> you tune the local oscillator to 434.075MHz
[22:59] <HenryH> the mixer multiplies that with the incoming RF signal
[22:59] <HenryH> well maybe adds not multiplies
[23:00] <HenryH> then you put a 455KHz low-pass filter on that and the output of that is the IF?
[23:00] <jcoxon> night all
[23:00] jcoxon (n=jcoxon@81-178-160-64.dsl.pipex.com) left irc: "Leaving"
[23:01] <rocketboy> yes you would normally mix (multiply) the signal with 423.375 (or 444.775) and pass it though a 10.7MHz IF filter
[23:03] <HenryH> is the IF filter low-pass or bandpass?
[23:03] <rocketboy> and then mix it with 10.245MHz (or 11.155MHz) and pass it through a 455KHz filter
[23:03] <rocketboy> band pass
[23:03] <HenryH> ok
[23:03] <HenryH> why the two stages?
[23:04] <rocketboy> its all to do with selectivity - its difficult to make a input stage selective enough to reject the signal 455KHz above or below at 434MHz
[23:05] <rocketboy> its easier if its 10.7MHz above/below
[23:05] <HenryH> right
[23:05] <rocketboy> 10.7 /455 are just standard IF frequences - lots of filters available
[23:06] <HenryH> so is the best bet to make one of these from scratch or just open up a commerical receiver and tap in before the detector?
[23:07] <rocketboy> I'd go with the latter
[23:07] <HenryH> ok
[23:07] <HenryH> the DSP-ness should be interesting
[23:08] <HenryH> actually don't pronounce that word
[23:08] <HenryH> you pretty much want to carry out an FFT on it right?
[23:09] <rocketboy> yes it would be quite an interesting project - somthing I have thought about for ages
[23:10] <HenryH> actually is it an FFT or just two bandpass filters
[23:12] <rocketboy> two bandpass filters (even these days) DSP techniquesarn't fast nough
[23:14] <HenryH> well
[23:14] <HenryH> interesting stuff
[23:14] <HenryH> I think for now though I'd like to do something that doesn't require dedicated receiver hardware other than a standard receiver and computer
[23:15] <HenryH> so I'll go with writing a better replacement for TrueTTY, perhaps supporting multiple frequencies instead of just two, looks like you could fit maybe 8 in the bandpass
[23:15] <HenryH> and perhaps support for turbo codes
[23:15] <HenryH> I think that would give us a few more dB
[23:15] <HenryH> does that sound sensible to you?
[23:16] <rocketboy> sort of - you won't be able to use them with the TNC I used on nova 1 - but they would work with the system I used on MiHAB2
[23:17] <rocketboy> have you got any turbo code examples
[23:18] <HenryH> not yet, actual examples of algorithms seem to be hard to find but I think there are some guys in the university working on them so maybe they could help out
[23:18] <HenryH> I'm plannign to build something similar to the one you used on MiHAB2 with our radiometrix transmitter and an actual DAC
[23:19] <rocketboy> sounds good - I started some work using reed-sol codes but diddn't get round to finishing it
[23:19] <HenryH> the only practical benefit will be being able to send a low-res photo in 1 minute instead of 5
[23:19] <HenryH> but that's still nice :)
[23:19] <rocketboy> also multi frequency FSK would give another db or so of noise margin
[23:20] <rocketboy> thats about the best that could be done using standard licence exempt transmitters
[23:22] <HenryH> it would be good to get maybe 100 bytes/sec at 400km range
[23:22] <HenryH> and that seems doable
[23:22] <rocketboy> BTW - i'm having trouble with the PIC sumulator reading the USART - it keeps comming back witn 0 rather than the data that is in the input file - any ideas
[23:22] <HenryH> hm
[23:22] <HenryH> the PIC simulator is rubbish
[23:23] <rocketboy> i was wondering if it was a bug
[23:23] <rocketboy> in the simulator
[23:23] <HenryH> I've never used input files with the simulated usart
[23:24] <rocketboy> the output works ok
[23:24] <HenryH> mplab has some issues with long folder names, try putting it in a directory near the root
[23:24] <rocketboy> ah ok it is rather long
[23:25] <HenryH> otherwise I'd just put a PIC on a breadboard and connect it to the serial port
[23:25] <rocketboy> I think 100 bytes/sec at 400km range is doable - but it would be taking everything and pushing it to the limits - multi-frequency + error correction codes
[23:27] <HenryH> it might be worth scaling the bitrate with the range
[23:27] <HenryH> i.e. it would work out how far it is from the base station and adjust the transmission rate accordingly
[23:28] <rocketboy> yeah - thats what they did with Voyager I think
[23:29] <HenryH> it's a bit more annoying when you don't have any ability to uplink "speak slower" etc
[23:29] <HenryH> then again a really slow uplink might be doable but it means more weight on the payload
[23:29] <rocketboy> The SSB channel should push up to about 1200 bps FSK max
[23:30] <HenryH> the channel is 3kHz right?
[23:30] <rocketboy> uplink is a problem without a HAm licence
[23:31] <rocketboy> yes normally 300Hz to 3.3KHz or there abouts - somtimes its cut down to 2.4KHz
[23:31] <HenryH> my idea was to get a really directional antenna, aim it very accurately and plug a 10mW module into it
[23:31] <HenryH> how were you sending 4800bps though a 3kHz channel?
[23:32] <rocketboy> thats the problem - most licence exempt stuff prohibits the use of an antenna with gain on the transmit end
[23:32] <HenryH> ah
[23:32] <HenryH> spoilsports
[23:32] <rocketboy> yeah
[23:32] <HenryH> lasers it is then
[23:33] <HenryH> but am I right in thinking the unit you flew on saturday was transmitting at 4800 baud? or was that just the buffer on the TNC
[23:33] <rocketboy> there are restrictyions on sky piinting lasers too (aircraft)
[23:34] <rocketboy> saturday was 1200 baud to air - the link to the PC was 4800
[23:34] <HenryH> ok
[23:34] <HenryH> a really long wire?
[23:34] <rocketboy> sunday was 300 baud to air - 4800 to the FC
[23:35] <rocketboy> I'll have a look - I think ther is some long range licence exempt stuff that may cut it
[23:35] phatmonkey (i=nobody@81.2.121.150) left irc: Remote closed the connection
[23:36] <HenryH> an uplink isn't that critical though
[23:36] <rocketboy> the shorter filename didn't work :-(
[23:36] <HenryH> oh.. not sure then, sorry
[23:37] <HenryH> thanks for all the radio help, I am learning, slowly
[23:37] <rocketboy> yeah - I'm sure I can do somthing on the uplink - I just havn't thought of it yet
[23:37] <rocketboy> no probs
[23:38] <rocketboy> I'll just have to buy one of thse ICD2 clones - what was that .nl link again
[23:38] <HenryH> www.voti.nl
[23:39] <rocketboy> ah ta - saves a trip to the car
[23:39] <HenryH> the uplink can be as slow as you like really
[23:40] <HenryH> do the signal-to-noise/range benefits continute to get better even when you're down to several seconds per bit?
[23:40] <HenryH> continue*
[23:40] <rocketboy> yeah and the reciver doesn't have to be standard
[23:41] <rocketboy> yes - each time you half the bandwidth you halve the noise taken in - also a bit is longer - so you can integrate over the longer period
[23:51] <rocketboy> I'm off 2 bed - c u
[23:51] rocketboy (n=steve@217.47.75.5) left irc: "Leaving"
[23:53] HenryH (n=H@natreg.pem.cam.ac.uk) left irc: "I can't believe what a lame quit message I had before"
[00:00] --- Wed Sep 13 2006