highaltitude.log.20090120

[00:00] <edmoore> if it turns out to be possible without going to slightly silly lengths just to win the comp, then that's cool
[00:01] <edmoore> but i don't want to waste time on the comp if we get an uplink working satisfactorily and it would require deviating significantly from that. I guess let's suck it and see
[00:02] <fergusnoble> yeah
[00:06] <fergusnoble> so should we put the rx on badger2?
[00:07] <edmoore> well, I did make the footprint for it earlioer :)
[00:07] <edmoore> so may aswell use it in anger
[00:08] <fergusnoble> yeah
[00:08] <fergusnoble> are you going to work on it tonight?
[00:08] <edmoore> this is where having dma on the arm would be super useful
[00:08] <edmoore> no, off to bed
[00:08] <edmoore> i'm going to try and have a week of 12.30-7 sleeps. body clock has got a bit mashed recently
[00:09] <fergusnoble> btw, i got crossworks mac working, for some reson some of the target drivers (i.e. different programmers) were making it crash
[00:09] <edmoore> peculiar
[00:09] <edmoore> oh well
[00:09] <fergusnoble> i just edited the xml file where it decides what to load and got rid of everything but the crossconnect and the simulator
[00:09] <fergusnoble> and it seems to have fixed it
[00:10] <fergusnoble> no lic yet though, sent an email
[00:10] <fergusnoble> he didnt give us proper licences for the mac version did he?
[00:10] <edmoore> no
[00:10] <fergusnoble> ok
[00:10] <fergusnoble> maybe we should ask for some
[00:11] <edmoore> is it released yet?
[00:11] <edmoore> he said at the time he wasn't issuing them until it was released
[00:11] <fergusnoble> nope
[00:11] <fergusnoble> ok
[00:12] <fergusnoble> anyway, should it go on the main board or the expansion board?
[00:12] <fergusnoble> if we have space on the main board, it might be good
[00:12] <edmoore> it would probably need a fair bit of processing time
[00:13] <edmoore> the trouble with us not having DMA is that we'd be forever servicing interrupts
[00:13] <edmoore> it's a bit old-tech
[00:14] <fergusnoble> i think its possible
[00:14] <edmoore> i think the uplink is, for me, in the secondary/experimental camp
[00:14] <edmoore> and not one of the core features we've grown to need
[00:14] <fergusnoble> the interrupts can be made fast
[00:15] <fergusnoble> yeah, so maybe on the expansion oard
[00:15] <edmoore> i'd be totally up for the expansion board
[00:15] <fergusnoble> what about using one of the transcievers?
[00:15] <edmoore> fo shizzle
[00:15] <edmoore> what freq?
[00:16] <fergusnoble> 434
[00:16] <edmoore> most of them seem to be digital
[00:16] <edmoore> but i've not looked properly
[00:16] hallam (i=836f0142@gateway/web/ajax/mibbit.com/x-6253d73ecbf38f90) joined #highaltitude.
[00:16] <fergusnoble> they seem to be digital in so much as you can set the channel and tx power over a serial protocol
[00:17] <hallam> hello
[00:17] <fergusnoble> but they let you disable the modem and drive the TXD yourself
[00:17] <fergusnoble> (i think)
[00:17] <fergusnoble> so we can go for lower data rate
[00:18] <edmoore> if it looks like a goer, am happy to go for it
[00:18] <fergusnoble> not sure if the rx uses the modem always though, but even so we should be able to make 1200baud with 35W of power up
[00:18] <edmoore> would also save the diplexer, which would be something
[00:18] natrium42 (i=natrium4@CPE000625d867e2-CM0014045885be.cpe.net.cable.rogers.com) joined #highaltitude.
[00:18] <edmoore> sure, though that would definitely write us off this comp
[00:19] <edmoore> it may be that a nrx2 with a pre-amp and some very slow baud dominoEX with FEC would do the trick
[00:19] <edmoore> + very very careful SWR tuning
[00:20] <hallam> is it cheating to use the 434.075 repeater?
[00:20] <edmoore> sticking a pre-amp inline is an easy thing to do
[00:20] <fergusnoble> hallam: probabl
[00:20] <fergusnoble> we are supposed to be using unlicensed bands only
[00:21] <fergusnoble> and the repeater you can only use under your callsign
[00:21] <hallam> it would be pretty cheeky
[00:22] <edmoore> ironically, we'd probably cause less interference to other users if we used a yagi on the 10mW ground uplink than an omni
[00:23] <edmoore> right, off to bed
[00:23] <edmoore> am in dept all day tomorrow if anyone's around
[00:23] <edmoore> though actually, got a lab report to do
[00:23] <edmoore> so may be in library
[00:24] <hallam> what's the plan then, ed and fergus? 868?
[00:25] <fergusnoble> we can use 169 at 500mw
[00:25] <fergusnoble> for better fspl, but as ed points out a more cumbersome antenna on the balloon
[00:25] <hallam> 169?
[00:26] <edmoore> I'd like to see if we can get away with 434, as it's an amateur band and we can use the same antenna.
[00:27] <hallam> did you mean 869 rather than 169?
[00:27] <fergusnoble> edmoore: i doubt it would work
[00:27] <fergusnoble> hallam: no i didnt
[00:27] <fergusnoble> http://www.ofcom.org.uk/radiocomms/ifi/tech/interface_req/IR2030final.pdf
[00:27] <hallam> that's what I'm looking at
[00:28] <hallam> 169MHz isn't even mentioned
[00:28] <fergusnoble> edmoore: if you think we can struggle to get a reliable signal on the ground with a far more sensitive reciever
[00:28] <fergusnoble> hallam: think of the balloon as an asset we might want to track
[00:28] <edmoore> fergusnoble: what's your point?
[00:28] <edmoore> oh i see, in terms of 10mW erp still
[00:29] <fergusnoble> yes
[00:29] <fergusnoble> i think if we are going to get james' thing to work it will be with one of the 500mw bands
[00:29] <edmoore> but we could easily get a massive signal with 35W on ssb down a yagi to it
[00:29] <fergusnoble> oh yeah, would work a charm
[00:29] <edmoore> yes indeed, that's my point too
[00:29] <fergusnoble> and i think we should start with that
[00:30] <edmoore> i guess on whether or not it's worth doing jame'ss thing
[00:30] <edmoore> sure, ok
[00:30] <edmoore> well i guess just get experience with an uplink that's actually useful to us as CUSF and then decide if we want to have a crack at this comp
[00:30] <fergusnoble> hallam: table 3.2 cat ix
[00:30] <edmoore> as fergusnoble says
[00:30] <hallam> ok I see that now
[00:31] <hallam> table 3.15 is interesting
[00:31] <hallam> I'm trying to figure out how a social alarm for the elderly and infirm could be used in airborne applications
[00:31] <edmoore> right, catch you all tomorrow
[00:31] <hallam> paratrooping grannies?
[00:31] <hallam> night
[00:31] edmoore (n=edmoore@pomegranate.chu.cam.ac.uk) left irc:
[00:33] <fergusnoble> hallam: allowed 4W 2.4ghz if our balloon counts as an rfid tag
[00:33] <gordonjcp> hallam: you could use it to freak out old people's homes over a wide area
[00:35] <hallam> I think it would be cheeky to try to count it as an rfid tag
[00:35] <hallam> but god knows you don't need 4W
[00:36] <hallam> I've been doing too much GPS stuff lately, makes me want to try code-based spread spectrum
[00:38] <fergusnoble> hehe
[00:38] <fergusnoble> hows the gps stuff going?
[00:38] <fergusnoble> when does the book come?
[00:40] <hallam> in the post
[01:31] jiffe88 (n=weechat@64.251.162.157) left irc: "WeeChat 0.2.6"
[02:21] <hallam> fergusnoble: about?
[02:38] iNatrium (i=natrium4@CPE000625d867e2-CM0014045885be.cpe.net.cable.rogers.com) joined #highaltitude.
[02:39] <iNatrium> Hi
[02:41] <hallam> Hi
[02:57] iNatrium (i=natrium4@CPE000625d867e2-CM0014045885be.cpe.net.cable.rogers.com) left #highaltitude.
[03:13] hallam (i=836f0142@gateway/web/ajax/mibbit.com/x-6253d73ecbf38f90) left irc: "http://www.mibbit.com ajax IRC Client"
[03:15] hallam (i=836f0142@gateway/web/ajax/mibbit.com/x-bcad1f0ef56a44d4) joined #highaltitude.
[03:30] EI5GTB (n=Paul@apollo.paulsnet.org) left irc: Read error: 104 (Connection reset by peer)
[04:09] hallam (i=836f0142@gateway/web/ajax/mibbit.com/x-bcad1f0ef56a44d4) left irc: "http://www.mibbit.com ajax IRC Client"
[04:37] natrium42 (i=natrium4@CPE000625d867e2-CM0014045885be.cpe.net.cable.rogers.com) left irc:
[04:39] natrium42_ (n=natrium4@CPE000625d867e2-CM0014045885be.cpe.net.cable.rogers.com) joined #highaltitude.
[07:11] jiffe88 (n=jiffe2@209.159.247.189) joined #highaltitude.
[07:22] jiffe97 (n=jiffe2@209.159.247.189) left irc: Read error: 113 (No route to host)
[07:38] Simon-MPFH (n=simon@phantom.mpfh.co.uk) joined #highaltitude.
[08:03] RocketBoy (n=Steve@217.47.75.27) joined #highaltitude.
[08:12] edmoore (n=edmoore@pomegranate.chu.cam.ac.uk) joined #highaltitude.
[08:13] RocketBoy (n=Steve@217.47.75.27) left irc: "Leaving"
[08:19] edmoore (n=edmoore@pomegranate.chu.cam.ac.uk) left irc:
[08:36] edmoore (n=ed@chu-gw.churchillcambridge.co.uk) joined #highaltitude.
[08:49] fnoble_lab (i=81a9460e@gateway/web/ajax/mibbit.com/x-4a2978d12bfc7c93) left irc: "http://www.mibbit.com ajax IRC Client"
[09:01] edmoore (n=ed@chu-gw.churchillcambridge.co.uk) left irc: Read error: 110 (Connection timed out)
[09:46] rharrison (n=rharriso@gateway.hgf.com) joined #highaltitude.
[09:46] <rharrison> morning all
[09:46] <rharrison> got lots of work to do today :(
[10:35] edmoore (n=ed@chu-gw.churchillcambridge.co.uk) joined #highaltitude.
[11:14] edmoore (n=ed@chu-gw.churchillcambridge.co.uk) left irc: Read error: 110 (Connection timed out)
[11:16] edmoore (n=edmoore@pomegranate.chu.cam.ac.uk) joined #highaltitude.
[12:14] rharrison (n=rharriso@gateway.hgf.com) left irc:
[13:54] rharrison (n=rharriso@gateway.hgf.com) joined #highaltitude.
[13:58] Laurenceb (i=83e34f19@gateway/web/ajax/mibbit.com/x-ba37bfed63412538) joined #highaltitude.
[13:58] <Laurenceb> Hello
[13:59] <rharrison> hi
[14:00] <rharrison> any launches?
[14:00] <rharrison> Laurenceb U signed for the comp?
[14:00] <Laurenceb> no
[14:00] <Laurenceb> where is it?
[14:02] <Laurenceb> ah wiki
[14:03] <Laurenceb> lol free envelope
[14:05] <Laurenceb> ah a radio relay prize
[14:05] <Laurenceb> tricky
[14:10] <Laurenceb> SpeedEvil: do you have a gps front end working?
[14:13] Action: Laurenceb contemplates compiling fldigi for a linux board
[14:19] <Laurenceb> thing is I'd want to do direct downconversion
[14:20] <Laurenceb> is there a way to compile some application os it looks like the sounds card?
[14:21] <Laurenceb> hmm guess its best to write the code from scratch
[14:22] <rharrison> hi edmoore
[14:23] <gordonjcp> Laurenceb: What exactly are you trying to do?
[14:23] <rharrison> is CUSF up for the comp?
[14:23] <Laurenceb> just contemplating jcoxons challenge
[14:23] <edmoore> rharrison: hi
[14:23] <rharrison> You geeks should walk this challenge
[14:24] <rharrison> I have ordered modules for test purposes today
[14:25] <Laurenceb> I'd use an AD4350 eval board for starters
[14:25] <Laurenceb> then NGW100 board
[14:28] <rharrison> edmoore do you have the official CUSF response or are you still contemplating?
[14:31] <Laurenceb> then an ADL5387
[14:31] <edmoore> it's not so much having an uplink, that's not too bad, it's whether or not doing it within the rules of this challenge is worth it
[14:32] <Laurenceb> yeah
[14:32] <Laurenceb> a basic uplink is fairly doable
[14:32] <edmoore> we're basically going to do our uplink anyway and see from then whether or not it can be made to work within the rules
[14:32] <Laurenceb> but to be worth it, I'd go for something hardcore
[14:32] <Laurenceb> edmoore: whats your uplink plan?
[14:33] <edmoore> we wouldn't be doing a 'basic' uplink
[14:33] <Laurenceb> hardcore rf? off the shelf?
[14:33] <edmoore> we'd be doing one where we could point several watts up a yagi at the balloon and just have simple kit on balloon
[14:33] <Laurenceb> oh that one
[14:33] <Laurenceb> fairdoos
[14:34] <Laurenceb> I'd go for ~10mw up the yagi
[14:34] <Laurenceb> then some AD kit and an NGW100 onboard
[14:34] <edmoore> that's not within the rules
[14:34] <edmoore> 10mW erp
[14:35] Action: Laurenceb reads again
[14:35] <Laurenceb> ah yeah, uplink as well
[14:35] <edmoore> exactly
[14:35] <Laurenceb> hmm need to go for hardcore RF approach
[14:35] <Laurenceb> but there as eval boards...
[14:35] <Laurenceb> *are
[14:36] <edmoore> so that's the only thing holding us back
[14:37] <edmoore> badger 2 was always getting an uplink, it might not be worth adjusting everything to make it capable of winning this competition if we can acheive the same thing much more simply and reliably for normal operation with a different approach
[14:37] <edmoore> that said, 500mW on 169mhz is nice, and can probably do a lot with that. I think Fergus and I will have to have a few beers over it
[14:37] <Laurenceb> hmm NGW100 -- spi--- AD4350 eval --- ADL5387 eval ---- filtering +opamps + ADC on a basic pcb (its <20MHz)
[14:38] <edmoore> Laurenceb: re: hardcore DSP and stuff - that's exactly my point - don't want to have to stick loads of power hungry DSP on the payload just to decode a very weak signal when with 10W up a yagi we could get away with much much less hardware on the payload
[14:38] <Laurenceb> oh and an LNA on the RF into the AD5387
[14:38] <Laurenceb> yeah, but it would be a nice project
[14:39] <Laurenceb> and uits not horrendous <1A
[14:39] <edmoore> for sure, but not really sure we have the time when there's big projects for us on the go currently anyway
[14:39] <Laurenceb> which is possible with lipo
[14:39] <Laurenceb> yeah
[14:39] <Laurenceb> I'm not sure itsd worth it
[14:39] <edmoore> that said, it may turn out it's much easier than i think and I'm being unduly pessimistic. will see in due course
[14:39] <edmoore> but those are my concerns
[14:40] <Laurenceb> well with some eval board its not _too_ bad
[14:40] <edmoore> why specifically eval boards?
[14:40] <Laurenceb> but there will be the inevitable impedance balancing issues
[14:40] <Laurenceb> so the design is quicker and easier
[14:40] <edmoore> oh I see
[14:41] <Laurenceb> stick most of the impedance/RFI issues onto premade boards
[14:41] <edmoore> i think the extent we'd probably go to is some pre-amps in the rx chain. if it took a massive re-design, got better things to be spending design time on, basically.
[14:41] <Laurenceb> then its mostly just RTTY tracking on the NGW100
[14:41] <Laurenceb> yeah, you need a powerful LNA on that design I came up with
[14:42] <Laurenceb> which may have to be on a new PCB .... if your lucky you could reuse a GPS ant
[14:42] <Laurenceb> some GPS LNAs go down to almost DC
[14:46] <Laurenceb> yeah I dont think its really worth it - may as well work on a similar less power hungry design for my USB stick
[14:47] <Laurenceb> what I mensioned was the first usb stick design
[14:47] <edmoore> apparently the telescope/track-a-tron has just arrived
[14:48] <Laurenceb> a telescope aimer?
[14:48] <edmoore> yagi
[14:48] <Laurenceb> well it is now
[14:48] <Laurenceb> my icom prc-1000 has arrived :P
[14:49] <Laurenceb> just need a yagi
[14:49] <Laurenceb> oh and a BNC thingy
[14:54] <edmoore> a BNC thingy?
[14:54] <Laurenceb> to connect onto the icom
[15:04] <Laurenceb> never buy kit for TVs from hardware stores
[15:04] <Laurenceb> its horrendously overprices and really crappy
[15:07] <rharrison> edmoore I think you have 500mw
[15:07] <edmoore> sure, on less helpful non-ham bands
[15:07] <rharrison> BTW you might be able t use a yagi provided the ERP is 500mw
[15:07] <Laurenceb> still not much
[15:08] <edmoore> depends on which band
[15:08] <Laurenceb> unless you have some very good hardware
[15:09] <edmoore> i can do 500mW on 169 and 868, sure. It does suddenly nock you out of ham equipment though
[15:11] <edmoore> I'm having a listen to 169 now, and it's certainly not empty either
[15:12] <rharrison> Can you get 169 on the icom
[15:13] <edmoore> i can hear it, sure
[15:13] <edmoore> can't broadcast though
[15:17] <Laurenceb> there should be several prizes
[15:17] <Laurenceb> I was expecting one for controlled descent
[15:17] <edmoore> I don't want to complicate things. But I am not partiularly convinced that it is illiegal to use amateur kit to transmit from the ground to our equipment
[15:18] <edmoore> I think we're covered under section 10
[15:18] <Laurenceb> yeah - I thought we'd already decided that
[15:19] <edmoore> and I think realistically for a useful and reliable uplink, that's exactly what we'd do. but oh well.
[15:19] <Laurenceb> can always dream of DSPiness
[15:19] <Laurenceb> actually its direct downconversion
[15:27] <rharrison> 169 looks very promising
[15:44] edmoore (n=edmoore@pomegranate.chu.cam.ac.uk) left irc:
[16:05] <Laurenceb> gyuess you can always go for the dead bug approach http://i171.photobucket.com/albums/u314/lugnut808/lfcsp_with_0402_decoupling.jpg
[16:07] <Laurenceb> 1st prize for building an entire receiver like that :P
[16:14] <Laurenceb> guess you could be it with breakout boards, pretty muc everything is QFN
[16:15] <Laurenceb> at 434Mhz - wavelnght is ~70cm , so in a pcb it may be around 15cm
[16:16] <Laurenceb> 15/10=1.5, so as a rough rule of thumb, 1.5cm non matched lines are possible
[16:17] <Laurenceb> but its still going to screw things up a bit
[16:35] Bluenarf (n=Paul@apollo.paulsnet.org) joined #highaltitude.
[16:36] Nick change: Bluenarf -> EI5GTB
[16:37] Action: SpeedEvil passes Laurenceb a SWR meter.
[16:38] <EI5GTB> the only downside of swr is that its so easy to measure :P
[16:39] <Laurenceb> dont you need quite high power rf input?
[16:45] <Laurenceb> EI5GTB: have you looked at trirotors?
[16:46] <Laurenceb> theres also chinook style designs - saves two motors
[16:46] <EI5GTB> hmmm... nope
[16:46] <EI5GTB> i gave up when i realised i was jumping in at the deep stormy end
[16:47] <EI5GTB> im thinking of making a rogallo deal first
[16:47] <Laurenceb> if you have one motor tilting in pitch axis, and one in roll
[16:47] <Laurenceb> yeah, they are nice and easy, but crappy glide ratio
[16:47] <EI5GTB> orly>
[16:47] <EI5GTB> hmm
[16:48] <EI5GTB> i just need somthing to learn about kalman filtering, and desiging autonomus things that fly
[16:48] <Laurenceb> http://wiki.ukhas.org.uk/_media/projects:ukhas_glider_project:p7270025.jpg <-- that was tons of fun
[16:49] <EI5GTB> cool
[16:49] <EI5GTB> yea, i wanna do somthing like that
[16:49] <EI5GTB> maybe with a motor
[16:49] <EI5GTB> i donno
[16:49] <EI5GTB> anything..
[16:49] <Laurenceb> that wouldnt work with a motor
[16:49] <EI5GTB> just an excuse to get going on embedded systems etc..
[16:49] <Laurenceb> you need a "freewing"
[16:50] <EI5GTB> oh
[16:50] <Laurenceb> i.e. free to tilt in pitch
[16:50] <Laurenceb> to get a motor running properly on a rogallo
[16:50] <Laurenceb> and then you need a tail, to stabilise the body
[16:50] <Laurenceb> so it looks luike a normal plane with rogallo wings bolted on the top
[16:50] <EI5GTB> lol
[16:51] <Laurenceb> the advantage is you have a rudder to steer
[16:51] <Laurenceb> but some peple go for vectored thrust
[16:51] <Laurenceb> as you can pull off interesting manovers
[16:51] <EI5GTB> hmm
[16:52] <Laurenceb> theres some threads on rcgroups
[16:53] <Laurenceb> I tried sticking a motor on, and it would fly ok, but not glide
[16:54] <Laurenceb> the problenm is you need to change the CofG to go from powered flight to glide
[16:54] <SpeedEvil> you want a motor that can put the prop in the right position, behind a spar.
[16:54] <Laurenceb> yeah
[16:54] <EI5GTB> hmmm
[16:54] <SpeedEvil> AS one option at least.
[16:54] <gordonjcp> SpeedEvil: easy enough surely?
[16:54] <Laurenceb> but the wing will end up being shredded
[16:54] <SpeedEvil> Laurenceb: pusher?
[16:55] <gordonjcp> SpeedEvil: optical sensor to detect when the prop is in position, and magnetic braking
[16:55] <Laurenceb> what the hell
[16:55] <SpeedEvil> gordonjcp: or a brushless motor that has only several positions, and a hall sensor
[16:55] <Laurenceb> no I dont think it will work anyway
[16:55] <SpeedEvil> pitch problems, yeah
[16:55] <gordonjcp> yeah
[16:56] <Laurenceb> you need to move the CofG whereever the motor is
[16:56] <Laurenceb> to go from level flight to gliding 1:4 descent
[16:56] <SpeedEvil> Umm.
[16:56] <SpeedEvil> Depending on the thrust vector surely
[16:56] <Laurenceb> otherwise when you turn the motor off it goes down like a chute
[16:57] <Laurenceb> nope, thrust vector just makes it worse
[16:57] <Laurenceb> its not the whole problem
[16:57] Action: SpeedEvil ponders.
[16:58] <SpeedEvil> 4 chutes! Normally, it looks like a biplane.
[16:58] <Laurenceb> CofG and Cof Lift are at different places
[16:58] <Laurenceb> lol
[16:58] <SpeedEvil> When under power, it turns two of the chutes backwards, and rotates them like a coaxial heli.
[16:58] <Laurenceb> that would look mad
[16:58] <Laurenceb> oh god
[16:59] <Laurenceb> SpeedEvil: you have to have a wing thats free to move in the pitch axis - and hinged through the C of Lift
[17:00] <Laurenceb> then I think your thrust vector needs to point through the center of gravity
[17:01] <Laurenceb> need to draw some vectors on paper I think
[17:03] rharrison (n=rharriso@gateway.hgf.com) left irc:
[17:03] <SpeedEvil> doesn't the CoL vary with airspeed?
[17:03] <SpeedEvil> doesn't the CoL vary with indicated airspeed?
[17:05] <Laurenceb> dont think so
[17:05] Action: Laurenceb is working on a "bicopter" design
[17:06] <Laurenceb> two counterrotating brushless motors buried inside a foam body
[17:06] <Laurenceb> rectangular body approx 60cm*25
[17:06] <SpeedEvil> dealextreme has cheap ducted fan modules.
[17:06] <Laurenceb> hmmm
[17:06] <Laurenceb> link?
[17:06] <SpeedEvil> though not counterrotating
[17:07] <SpeedEvil> dealextreme.com
[17:07] <SpeedEvil> :)
[17:07] <Laurenceb> problem is I then have to tilt the entire unit
[17:07] <SpeedEvil> search for brushless.
[17:07] Action: SpeedEvil is having issues with DSN ATM.
[17:08] <Laurenceb> I'd use a sparkfun 5DOF imu with an extra MLX90609 for yaw
[17:08] <Laurenceb> cant work out how to interface with the RC receiver
[17:08] <Laurenceb> the corona DSP isnt very hackable
[17:09] <Laurenceb> and 4 PWM streams are virtually impossible to decode
[17:09] <SpeedEvil> :/
[17:09] <SpeedEvil> why?
[17:09] <Laurenceb> well unless you have a uC with 4 ICP units
[17:10] <SpeedEvil> 4 interrupt on port change pins, and timing in interrupt handlers.
[17:10] <Laurenceb> noo
[17:10] <Laurenceb> tried that, its way too glitchy
[17:11] <Laurenceb> may have to use an old school receiver
[17:11] <Laurenceb> and hack it to get PPM
[17:11] <SpeedEvil> if it does each servo at different times, you could diode OR all the servos, and do it that way
[17:11] <Laurenceb> doubt it - it problbly has two 16 bit timers
[17:11] <SpeedEvil> with the interrupt routine only picking up which servo it was
[17:11] <Laurenceb> and dual channel PWM
[17:11] <Laurenceb> or maybe 4 channel off one timer
[17:12] <SpeedEvil> or even cheat, and R-C-R-C filter it then A/D
[17:13] <Laurenceb> lol
[17:13] <Laurenceb> need an open DSP Rx
[17:13] jcoxon (n=jcoxon@host86-158-31-172.range86-158.btcentralplus.com) joined #highaltitude.
[17:16] <Laurenceb> I think it'd be very important there was absolutely no play in the rotor tiling mechanisms
[17:16] Action: SpeedEvil ponders rotor tiling.
[17:16] Action: SpeedEvil has only done bathroom tiling so far.
[17:17] <SpeedEvil> :)
[17:17] <Laurenceb> *tilting
[17:17] <SpeedEvil> yeah.
[17:17] <epictetus> i have had a VERY bad experience with dealextreme
[17:17] <epictetus> i bought a USB thermometer from them 3 months ago and they took my money and still havent delivered it
[17:17] Action: SpeedEvil has had good experiences generally.
[17:17] <epictetus> they give me a different excuse every time
[17:18] <Laurenceb> god those things are powerful
[17:18] <SpeedEvil> Bought maybe 6-10 orders of $2-300 total.
[17:18] <epictetus> at one point they said that since the package didn't go through, they wanted me to pay a second time for shipping for them to resend what I never received
[17:18] <Laurenceb> rather inefficient
[17:18] <Laurenceb> you could do that for ~20 amps normally
[17:18] <SpeedEvil> ducted fans are horribly inefficient.
[17:19] <Laurenceb> yeah
[17:19] <SpeedEvil> They whole mv vs 0.5*mv^2 thing
[17:19] <Laurenceb> probably best to go for a normal motor in a circular hole
[17:19] <Laurenceb> it would be about 4 times as big
[17:20] <SpeedEvil> Hmm.
[17:20] <Laurenceb> maybe 3
[17:20] <Laurenceb> times the diameter
[17:21] <SpeedEvil> If you could put the prop and motor above the chute, counterbalanced by the payload, that would tilt it in the right way
[17:21] <SpeedEvil> but mechanical complexity
[17:21] <Laurenceb> thats why that thing is so inefficient
[17:21] <Laurenceb> no
[17:21] <Laurenceb> sorry, got it wrong earlier
[17:22] <Laurenceb> you need the motor low down
[17:23] <Laurenceb> so it pitches the ting up
[17:23] <Laurenceb> but thats hard to do unless you want a destroyed prop
[17:23] <Laurenceb> or at least it ended up like that with my design
[17:24] <jcoxon> afternoon all
[17:25] <SpeedEvil> afgernogo
[17:25] Action: SpeedEvil blames RFI on his wireless keyboard.
[17:26] <SpeedEvil> It's a much better excuse than being unable to type.
[17:26] <Laurenceb> heh
[17:26] <Laurenceb> hi jcoxon
[17:26] <jcoxon> hey Laurenceb
[17:26] <jcoxon> and SpeedEvil
[17:27] <Laurenceb> I'm tempted to make a flying saucer stlye quadcopter with thoase fans
[17:27] <Laurenceb> but it would suck for efficiency
[17:28] <Laurenceb> 2Kg or lipos could be pricy
[17:29] <Laurenceb> wheres a cheap place for lipo?
[17:31] <SpeedEvil> ebay generally
[17:33] Action: Laurenceb searches for something absolutely huge
[17:35] <Laurenceb> http://cgi.ebay.co.uk/2x-14-8v-4000mAh-15C-LiPo-4-Cell-RC-Battery-14-8-WF_W0QQitemZ270315735335QQcmdZViewItemQQptZUK_ToysGames_RadioControlled_JN?hash=item270315735335&_trksid=p3286.c0.m14&_trkparms=72%3A1298|66%3A2|65%3A12|39%3A1|240%3A1318
[17:35] <Laurenceb> you could use 6 of those
[17:36] <Laurenceb> 24Ah - 10 minutes flight or something
[17:37] <Laurenceb> not very impressive
[17:37] hallam (i=836fc8c8@gateway/web/ajax/mibbit.com/x-24841b16c7e241b7) joined #highaltitude.
[17:37] <Laurenceb> you would have about 1Kg of payload
[17:37] <SpeedEvil> hovering is generally horribly inefficient.
[17:38] Action: Laurenceb wonders of you could get away with using an AVR
[17:38] <hallam> what's the topic?
[17:38] <Laurenceb> quadcopters/bicopters
[17:39] <hallam> fun
[17:39] <Laurenceb> using cheap ducted fans/other crap off ebay
[17:39] <Laurenceb> http://wiki.ukhas.org.uk/code:4_state_extended_kalman_filter_in_matlab
[17:40] <Laurenceb> I need to write a euler angle version of that, and see howe many clock cycles it uses on an AVR
[17:40] <Laurenceb> I've got a euler angle filter working so far, but it need to artificial horizon bit rewriting
[17:41] <Laurenceb> atm it takes 3.5ms on a 20MHz atmega168
[17:42] <Laurenceb> the nice thing is if you temperature compensate, the 5DOF IMU from sparfun is better than the 6DOF
[17:45] <Laurenceb> I guess you may be able to use the euler angles straigh away
[17:45] <Laurenceb> as the quadcopter _should_ be close to horizontal
[17:45] <hallam> wish I still had access to the calibrated turntable at MIT
[17:45] <Laurenceb> hmm neat
[17:45] <Laurenceb> yeah, that was just held be hand
[17:46] <Laurenceb> I was really confused when I tried plotting the euler angles
[17:47] <Laurenceb> but of course they arent intuitive - you really need to move to an artificial horizon
[17:47] <hallam> quaternions are more intuitive
[17:47] <hallam> (never thought I'd say that)
[17:47] <Laurenceb> hehe yeah
[17:47] <hallam> but yeah, euler angles blow goats
[17:47] <Laurenceb> I gave up on euler angles and moved to quaternions
[17:48] <Laurenceb> then used the quaternion to look at an artificial horizon
[17:49] <Laurenceb> but with AVR-libc, euler angles are quite a lot faster
[17:49] <Laurenceb> theres only a few trig functions to evaluate if you cache them properly
[17:50] <hallam> can you do the processing in Euler, then convert to quats?
[17:50] <Laurenceb> I guess so
[17:50] <Laurenceb> as theres no extra inform,ation
[17:50] <Laurenceb> but euler filters explode in certain attitudes
[17:51] <Laurenceb> e.g. at 51 seconds in my dataset I think euler would explode
[17:52] <Laurenceb> but I dont understand it that well
[17:52] <hallam> yeah they do have singularities AIUI
[17:52] <Laurenceb> yeah... I cant remember where offhand :P
[17:52] <hallam> that's why the redundant degree of freedom in quaternions is worth it
[17:54] <Laurenceb> its supriasingly easy to get control loop inputs from the state quaternion for roll and pitch
[17:54] <hallam> http://lea.hamradio.si/~s53mv/index.html this guy is a radio god
[17:55] <Laurenceb> just transform the z axis
[17:55] <Laurenceb> and find its x and y components
[17:55] <Laurenceb> then arcsin for the angles
[17:55] <Laurenceb> but these are != euler angles
[17:56] <Laurenceb> omg
[17:56] Action: Laurenceb gets down on his prayer mat
[17:56] <hallam> yeah I think I want to have his babies
[17:56] <hallam> homebuilt GPS. In 1992.
[17:57] <Laurenceb> ah I've found copies of that page on other sites
[17:57] <hallam> he also made a complete set of avionics for his aeroplane
[17:58] <hallam> including TCAS, which usually costs in the $100k region
[17:58] <hallam> and built 3 copies of Ed's synthesizer
[17:59] <Laurenceb> Ed's synthesizer?
[18:00] <hallam> Ed wants to make a DDS + heterodyne software-defined transmitter for all the amateur bands
[18:00] <Laurenceb> hmm
[18:00] <Laurenceb> interesting - I havent really looked at transmitters
[18:01] <Laurenceb> I've been working on a (hopefully) DC to 2GHz scanner to fit in a USB stick
[18:01] <Laurenceb> direct downconversion then ~ 30MHz bandwidth at 8bit samples over USB2
[18:02] <hallam> you want to actually make it stick-sized?
[18:02] <Laurenceb> yes
[18:02] <Laurenceb> 15x45mm
[18:02] <Laurenceb> 4 layer
[18:03] jcoxon (n=jcoxon@host86-158-31-172.range86-158.btcentralplus.com) left irc: "Leaving"
[18:04] <Laurenceb> the biggest problem is power consumption
[18:04] <Laurenceb> you can go up to 500ma but thats going to turn your stick into a fireball
[18:06] <hallam> aluminium case, it'll be fine
[18:06] <hallam> what needs so many amps, anyway?
[18:06] <Laurenceb> the AD integrated VCO and PLL
[18:06] <hallam> you might use the heat to make a crystal oven good enough for a GPS-locked frequency standard
[18:07] <Laurenceb> but I've ditched that for a Maxim VCO and AD PLL
[18:07] <Laurenceb> problem is its more board space
[18:08] <Laurenceb> also the mixer is power hungry
[18:09] <hallam> are you using such a tiny board just to be masochistic, or is there a good reason to make it that small?
[18:09] <Laurenceb> but linear make some fairly (~60ma) efficient ones
[18:09] <Laurenceb> I want to sell it
[18:10] <Laurenceb> the VCO seems to be one of the main limits on frequency
[18:11] <Laurenceb> you can only tune over an orcer of magnitude or so
[18:11] <hallam> hm, maybe you could use a DDS
[18:11] <Laurenceb> looks too expensive
[18:12] <Laurenceb> also it would only go up to a few hundered MHz
[18:13] <Laurenceb> I think it looks like 50MHz to 1.2 GHz or so is practical
[18:13] <Laurenceb> with direct downconversion
[18:16] <Laurenceb> the problem is the I and Q intermediate frequencies will interfere with the input
[18:19] <hallam> I really don't know much about this kind of stuff, but it does seem a bit ambitious to try to do all that in a tiny board right next to an operating computer and expect the SNR to be good
[18:20] <hallam> would be way cool if you can pull it off though!
[18:20] <Laurenceb> another problem with a DDS scanner is if you're looking for a weak signal in a very large bandwidth with several other sources present, the ADC will not have enough range to enable the signal to be captured
[18:21] <Laurenceb> yeah, theres going to have to be some sheilding
[18:21] <hallam> I guess DDS is easier to do well for TX than RX
[18:22] Hiena (n=Hiena@81.93.195.181.datatrans.hu) joined #highaltitude.
[18:22] <Laurenceb> time for me to head off
[18:22] <Laurenceb> cya
[18:22] <hallam> bye
[18:22] Laurenceb (i=83e34f19@gateway/web/ajax/mibbit.com/x-ba37bfed63412538) left irc: "http://www.mibbit.com ajax IRC Client"
[19:12] rjharrison (n=rharriso@80.176.172.227) joined #highaltitude.
[19:12] <rjharrison> hallam ping
[19:17] <EI5GTB> evening
[19:18] <hallam> Hi rjharrison
[19:18] <hallam> I'm going out shortly though
[19:18] Laurenceb (i=83e3dd23@gateway/web/ajax/mibbit.com/x-f8e9d66fbb9438ef) joined #highaltitude.
[19:18] <hallam> got GPS on the brain
[19:18] <hallam> how are you doing?
[19:18] <Laurenceb> hello
[19:24] <fergusnoble> hallam: ed is having dinner at 8
[19:24] <fergusnoble> hallam: so it will be a bit
[19:24] <fergusnoble> hallam: unless you want to come up here and play with the scope
[19:26] <hallam> think I need to get some grub myself
[19:26] <hallam> will swing by though
[19:26] edmoore (n=ed@chu-gw.churchillcambridge.co.uk) joined #highaltitude.
[19:26] <fergusnoble> edmoore: hello
[19:26] <hallam> what's the sky doing?
[19:26] <edmoore> hi
[19:27] <fergusnoble> hallam: i think its quite clear, can see one star from my window
[19:27] <Laurenceb> hallam: being dark
[19:27] <fergusnoble> Laurenceb: the scope has come :)
[19:28] <Laurenceb> ah
[19:28] <Laurenceb> thought you only got a mount?
[19:28] <edmoore> how do we optically track the balloon with just a mount?
[19:28] <Laurenceb> ah
[19:28] <Laurenceb> I thought it was for a zagi
[19:28] <edmoore> it is for a yagi
[19:28] <Laurenceb> erm yagi
[19:29] <edmoore> but come on, you gotta stick a webcam on the end
[19:29] <Laurenceb> :P
[19:29] <fergusnoble> Laurenceb: were going to mount the yagi but keep the scope on too
[19:30] <edmoore> fergusnoble: am aware that i definitely must prioritise lab report for tomorrow, tonight
[19:31] <edmoore> any cusf is subject to when i get through that, this eve. I have done about 1/3rd of it though and it's the longest third
[19:31] <rjharrison> Sorry bit delayed
[19:31] <rjharrison> Been plauing with R's and payload cutdown
[19:31] <edmoore> R's?
[19:32] <rjharrison> Work great off 9v supply but not enough powere for 9v pp3
[19:32] <Laurenceb> rjharrison: cable ties have a high melting point
[19:32] <rjharrison> Resistors
[19:32] <Laurenceb> try insulation off some solid core
[19:32] <Laurenceb> or insulating tape
[19:32] <edmoore> why are you interested in the melting point of cable ties?
[19:32] <fergusnoble> rjharrison: PP3s are really bad with current
[19:32] <akawaka> yeah
[19:32] <fergusnoble> rjharrison: they have a very low capacity
[19:33] <rjharrison> would 4aa give some more poke
[19:33] <rjharrison> Or amps as we should say
[19:33] <fergusnoble> edmoore: shall i just carry on with henry then, or should we wait for you?
[19:33] <Laurenceb> rjharrison: about 8 thicknesses of insulating tape 2mm wide
[19:33] <Laurenceb> works well
[19:33] <hallam> is it a bog standard PP3 or lithium?
[19:33] <Laurenceb> with 4xaa
[19:33] <fergusnoble> rjharrison: yeah AAs should work fine
[19:33] <rjharrison> litium
[19:33] <rjharrison> lithium
[19:33] <edmoore> fergusnoble: go with henry, i will come if I can
[19:33] <hallam> personally I like the 6V lithium photo cells, lighter than 4xAA
[19:33] <Laurenceb> pp3 lithium is very low current
[19:34] <Laurenceb> its lithium thionol chloride or summit
[19:34] <rjharrison> 4xUltralife
[19:34] <rjharrison> opps
[19:34] <Laurenceb> which is != aa/aaa photo lithium
[19:34] <rjharrison> Ultralife pp3
[19:34] <edmoore> are you using cable ties for load bearing?
[19:34] <Laurenceb> yeah ultralife pp3 is awful
[19:34] <hallam> http://data.energizer.com/PDFs/l522.pdf quotes 120mA
[19:35] <rjharrison> ok i'll have a play with 4xaa and se what I get
[19:35] <Laurenceb> sounds about right
[19:35] <hallam> edmoore, if he's using the method Sal and I were testing, it's only semi load bearing
[19:35] <Laurenceb> I couldnt even power up a gps with one
[19:35] <rjharrison> I was going to try one of those lipos but I thought it might not be the safest thing to do
[19:35] <akawaka> how long do you need to running the current for?
[19:35] <hallam> not nearly under the same sort of stress as the UKHAS attachment was
[19:35] <edmoore> you know my reservations. at least the parachute isn't being held on by one
[19:35] <hallam> yeah, if it fails, the worst that you get is an early cutdown
[19:35] <rjharrison> 40 secs max
[19:35] <edmoore> yeah
[19:36] <hallam> ok food time for me
[19:36] Nick change: hallam -> hallam|fud
[19:36] <rjharrison> have a nice 5ohm 350 deg c R
[19:36] <edmoore> still, the loads on the way up are pretty high still
[19:36] <edmoore> same ballpark as on the way down
[19:37] <rjharrison> it's np ai have tested with 15kg and it works gread
[19:37] <rjharrison> great
[19:37] <Laurenceb> rjharrison: I use 8 seconds
[19:37] <edmoore> rjharrison: no
[19:37] <edmoore> you have not tested it with 15kg at -55
[19:37] <Laurenceb> but you will have higher thermal mass
[19:37] <edmoore> basic materials science
[19:37] <rjharrison> True
[19:37] <edmoore> not just true, very important
[19:37] <Laurenceb> rjharrison: cable ties are very brittle
[19:38] <Laurenceb> at those temperatures
[19:38] <edmoore> just test it well. it got us very close to lots and lots and lots of trouble last time
[19:38] <rjharrison> the ct at worst will break and i'll have early cutdown
[19:38] <Laurenceb> yeah, it helps to make that the weakest link
[19:39] <akawaka> has anyone thought of electro/mechanical solutions for the cutdown?
[19:39] <edmoore> yep
[19:39] <edmoore> that would be my prefered option too
[19:40] <edmoore> maybe with a pyro back-up.
[19:40] <edmoore> we could always just buy some pyro pull pins
[19:40] <edmoore> i don't think there have been any documented cases of those failing
[19:40] <Laurenceb> I've found resistors work well
[19:40] <Laurenceb> no cases of them failing for me
[19:41] <edmoore> sure, but your data set isn't as big as the awerospace industry's
[19:41] <Laurenceb> but its design dependant
[19:41] <Laurenceb> edmoore: they are used on a lot of kit
[19:41] <Laurenceb> not off the shelf resistors, but a similar design
[19:42] <edmoore> there could always be something shit happen, like launching on a wet day and getting a bunch of ice on your resistor
[19:42] <edmoore> which would take tonnes of energy to shake before it melted the link
[19:42] <Laurenceb> maybe
[19:42] <Laurenceb> on exomars they were using some similar devices
[19:42] <Laurenceb> but thats not going to get iced :P
[19:43] <edmoore> on exo they use pullpins for when it counts
[19:43] <edmoore> or frangibles
[19:44] <rjharrison> Where does one buy pull pins?
[19:44] <Laurenceb> frangibles?
[19:45] <edmoore> rjharrison: i'm not actually sure how you go about it as a random punter
[19:45] <edmoore> Laurenceb: nuts which split up into many pieces
[19:45] <edmoore> pyro again
[19:45] <Laurenceb> ok
[19:46] <akawaka> you could probably make a pyro driven pin pretty easily
[19:46] hallam|fud (i=836fc8c8@gateway/web/ajax/mibbit.com/x-24841b16c7e241b7) left irc: "http://www.mibbit.com ajax IRC Client"
[19:47] <edmoore> yep
[19:47] <Laurenceb> gtg
[19:47] <Laurenceb> bbl
[19:48] Hiena (n=Hiena@81.93.195.181.datatrans.hu) left irc: "-=Halt! Hammerzeit!=-"
[19:48] bfirsh (n=ben@host-137-205-75-156.res.warwick.ac.uk) joined #highaltitude.
[19:49] <akawaka> is there anything even simpler though, like some kind of screw mechanism
[19:49] <edmoore> define simple
[19:49] <akawaka> i guess simple isn't the right word
[19:49] <akawaka> non-explosive:)
[19:49] <edmoore> often involves electric motors and drivers, the motors having differing thermal coefficieents so might seive up
[19:50] <edmoore> grease in the servos can freeze solid
[19:51] <edmoore> http://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/19950024937_1995124937.pdf
[19:51] <edmoore> a good read
[19:51] <Laurenceb> http://wiki.ukhas.org.uk/_detail/projects:mihab:p6290015.jpg?id=projects%3Aultrahab_page&cache=cache
[19:52] <Laurenceb> ^ that works for me
[19:52] <Laurenceb> anyway bbl
[19:52] Simon-MPFH (n=simon@phantom.mpfh.co.uk) got netsplit.
[19:52] dh (i=dh@bsd.ee) got netsplit.
[19:53] <rjharrison> Humm#
[19:53] <rjharrison> hehe that's not a group
[19:53] <rjharrison> I'm not sure about elec cut down
[19:53] <rjharrison> Tried and tested and failed
[19:54] <rjharrison> I'm getting to about 200C and it'd not enought
[19:54] <rjharrison> 9v off the PSU is fine as I dare say it can pack the amps
[19:55] <rjharrison> 9/5=I
[19:55] <edmoore> i will be open minded but i think pyros are the solution if you want reliability for this stuff
[19:55] <edmoore> and simplicity
[19:55] <rjharrison> I have applied fot exp. Licence
[19:55] <rjharrison> Old bill due around in a couple of weeks time
[19:56] <rjharrison> Avoided appying in the name of Fawkes
[19:56] <edmoore> :)
[19:56] <rjharrison> 1kg of black powered will be enough
[19:56] <rjharrison> I'm getting the fine stuff
[19:56] <rjharrison> as in small
[19:57] <rjharrison> Apparently bic is the way forward
[19:57] <rjharrison> + elec match
[19:58] <akawaka> semtex + a few bags of fertilizer
[19:59] <rjharrison> touch heavy
[20:01] <EI5GTB> can you get an exp licence over here?
[20:01] <rjharrison> not sure they may be more tetchy
[20:02] <EI5GTB> lol
[20:02] <EI5GTB> you probably have to pass a test
[20:03] dh (i=dh@bsd.ee) got lost in the net-split.
[20:03] Simon-MPFH (n=simon@phantom.mpfh.co.uk) got lost in the net-split.
[20:03] <edmoore> bic tubes are fine yeah
[20:03] <edmoore> keep them shortish
[20:03] <fergusnoble> rjharrison: one tip - dont use bic biros with the angled sides
[20:03] <natrium42_> hai
[20:03] <fergusnoble> rjharrison: must be round or they will fail with a longitudinal crack
[20:03] <edmoore> otherwise they may just split down one sside but not break
[20:03] <natrium42_> got my sarantel geohelix antennas
[20:03] <fergusnoble> rjharrison: and half will stay attached
[20:10] Simon-MPFH (n=simon@phantom.mpfh.co.uk) joined #highaltitude.
[20:10] dh (i=dh@bsd.ee) joined #highaltitude.
[20:41] edmoore (n=ed@chu-gw.churchillcambridge.co.uk) left irc: Read error: 110 (Connection timed out)
[20:52] rjharrison (n=rharriso@80.176.172.227) left irc: Read error: 110 (Connection timed out)
[20:52] rjharrison (n=rharriso@80.176.172.227) joined #highaltitude.
[20:58] RocketBoy (n=Steve@217.47.75.27) joined #highaltitude.
[21:30] Simon-MPFH (n=simon@phantom.mpfh.co.uk) left irc: "Leaving"
[22:04] jiffe97 (n=jiffe2@209.159.247.189) joined #highaltitude.
[22:16] jiffe88 (n=jiffe2@209.159.247.189) left irc: Read error: 110 (Connection timed out)
[22:41] rjharrison (n=rharriso@80.176.172.227) left irc:
[22:46] RocketBoy (n=Steve@217.47.75.27) left irc: "Leaving"
[22:48] RocketBoy (n=Steve@217.47.75.27) joined #highaltitude.
[22:58] <Laurenceb> back
[22:58] <Laurenceb> hmm what do you need to get an explosives license?
[22:58] <Laurenceb> do you need to have special "storage areas" ect
[22:59] edmoore (n=edmoore@pomegranate.chu.cam.ac.uk) joined #highaltitude.
[23:05] <RocketBoy> depends on what you want to get and or store - most rocket folks get by without one these days (by limiting what they buy)
[23:05] <RocketBoy> see http://ukra.org.uk/legal
[23:09] RocketBoy (n=Steve@217.47.75.27) left irc: "Leaving"
[23:32] Tygrys^ (i=tygrys@moo.pl) joined #highaltitude.
[23:44] Tiger^ (i=tygrys@moo.pl) left irc: Read error: 110 (Connection timed out)
[23:54] edmoore (n=edmoore@pomegranate.chu.cam.ac.uk) left irc:
[00:00] --- Wed Jan 21 2009