highaltitude.log.20090325

[01:36] SteamAtom (n=steamato@74-132-3-136.dhcp.insightbb.com) joined #highaltitude.
[02:02] SteamAtom (n=steamato@74-132-3-136.dhcp.insightbb.com) left irc: "Get Colloquy for iPhone! http://mobile.colloquy.info/"
[02:21] <Laurenceb> hmm how close can you have cables to patch antennas?
[02:22] <Laurenceb> without any ill effects
[02:33] Laurenceb (n=laurence@dyres221-35.surrey.ac.uk) left irc: "The day microsoft make something that doesnt suck is the day they make a vacuum cleaner"
[03:32] sYx66 (n=wooh@bas1-kitchener06-1167865234.dsl.bell.ca) joined #highaltitude.
[05:33] natrium (n=natrium4@CPE000625d867e2-CM0014045885be.cpe.net.cable.rogers.com) joined #highaltitude.
[05:51] natrium_ (n=natrium4@CPE000625d867e2-CM0014045885be.cpe.net.cable.rogers.com) left irc: Read error: 110 (Connection timed out)
[06:43] Simon-MPFH (n=simon@phantom.mpfh.co.uk) joined #highaltitude.
[07:00] rjharrison (n=rharriso@80.176.172.227) joined #highaltitude.
[07:16] jcoxon (n=jcoxon@host86-145-224-2.range86-145.btcentralplus.com) joined #highaltitude.
[07:26] <jcoxon> morning
[07:35] edmoore (n=edmoore@pomegranate.chu.cam.ac.uk) joined #highaltitude.
[08:17] <jcoxon> hey guys what is a good way of controlling a low voltage with a micro, a mosfet?
[08:17] <jcoxon> if i want to control say the ballast tank valve with an arduino would a low power mosfet do?
[08:41] icez (n=icez@unaffiliated/icez) left irc: "Lost terminal"
[08:59] <edmoore> jcoxon: yes
[09:00] <edmoore> well, depends on how low poer.
[09:00] <edmoore> power*
[09:05] <jcoxon> edmoore, well the valve requires 3v
[09:05] <edmoore> how much current?
[09:05] <jcoxon> just looking that up right now
[09:05] <jcoxon> one sec
[09:06] <jcoxon> 0.5W on the data sheet
[09:07] <jcoxon> 0.2A then
[09:07] <edmoore> cool. well one of rapid's logic level mosfets will do fine.
[09:07] <edmoore> You can get away with probably much smaller, but these are cheap and easy to solder
[09:08] <jcoxon> i'm suprised - thought it would require more current
[09:08] <jcoxon> these sort: http://www.rapidonline.com/Electronic-Components/Discrete-Semiconductors/MOSFETs/TO-220-Logic-level-power-MOSFETs-N-Channel/77687/kw/mosfet
[09:11] <edmoore> those ones yes
[09:11] <jcoxon> cool, thanks edmoore
[09:11] <edmoore> np
[09:12] <edmoore> sound advice from mark C
[09:12] <jcoxon> yeah
[09:12] <edmoore> we've been doing our fills with horizonal bottles of late
[09:12] <edmoore> well, for N10 anyway
[09:12] <jcoxon> i think its sensible
[09:13] <jcoxon> i think we've always done it but only cause the ground was uneven
[09:13] <jcoxon> but its very true - more care needs to be taken
[09:13] <jcoxon> and safety googles with cutdowns :-D
[09:13] <jcoxon> its a good look
[09:14] <edmoore> Henry had one go off infront of him on Saturday night
[09:15] <jcoxon> i know
[09:15] <edmoore> lucky he has glasses
[09:15] <jcoxon> yeah, didn't think of that
[09:16] <jcoxon> urgh i hate revision
[09:17] <edmoore> snap
[09:19] <jcoxon> edmoore, http://spacenear.us/gfs/2009-03-25-2.png
[09:20] <edmoore> oh man
[09:20] <edmoore> look at that
[09:20] <jcoxon> would be a seriously cool flight
[09:21] <jcoxon> it would reach the UK in just under 48hrs
[09:27] <edmoore> there'll be other good winds
[09:27] <edmoore> it's an exciting planet
[09:28] <jcoxon> oh of course - i've got an enormous directory of them
[09:28] <jcoxon> http://spacenear.us/gfs/
[09:28] <jcoxon> :)
[10:11] <edmoore> jcoxon: when's the next big testing event then?
[10:13] <jcoxon> next test flight?
[10:31] <SpeedEvil> For MOSFETs
[10:32] <SpeedEvil> A good plan for critical stuff is possibly not to simply trigger the FET directly from the pin.
[10:33] <SpeedEvil> But - to the pin connect a small capacitor. Then, from the other end of this cap,connect a diode to prevent it going negative, and another diode to a cap connected between gate and ground. A resistor across this second cap
[10:34] <SpeedEvil> this means that a static level on the pin can't trigger the FET
[10:35] <SpeedEvil> only a sustained burst of transistions - frequency and duration depending on the design
[12:30] edmoore (n=edmoore@pomegranate.chu.cam.ac.uk) left irc:
[13:37] jcoxon (n=jcoxon@host86-145-224-2.range86-145.btcentralplus.com) left irc: "Leaving"
[13:56] edmoore (n=edmoore@pomegranate.chu.cam.ac.uk) joined #highaltitude.
[14:00] bfirsh (n=ben@scooby.firshman.co.uk) joined #highaltitude.
[14:05] Laurenceb (i=83e34f19@gateway/web/ajax/mibbit.com/x-c1d3b9a96ba95e95) joined #highaltitude.
[14:05] <Laurenceb> hello
[14:05] <shellevil> o
[14:08] <Laurenceb> I've been looking at wind estimation techniques
[14:09] <Laurenceb> its quite interesting, moving from layerd averages to layers with line fits only leads to a 1% improvement
[14:10] <Laurenceb> the main improvements are gained by choosing the right bin thickness
[14:10] <Laurenceb> error on windspeed looks like a "bathtub" curve
[14:10] <Laurenceb> if you plot it on a log linear scale
[14:11] <Laurenceb> minimum error on windspeed is 28% at approx 350m altitude bins
[14:12] <shellevil> as the skew - m/s /m is quite small?
[14:12] <shellevil> unless you're close to the ground?
[14:12] <Laurenceb> which is suprisingly large... I suspect using velocity data from the gps will make it work better with thinner layers
[14:12] <Laurenceb> well the skew appears to be quite large
[14:12] <Laurenceb> but only in a few places
[14:13] <shellevil> I suppose for when the skew is large, around that area it's very likely to change radically
[14:13] <Laurenceb> moving to small layers captures the skew, but leads to noisy measurements
[14:13] <shellevil> windshear? Or does that mean something else.
[14:13] <Laurenceb> the lassen iq kalman filters the doppler and pseudorange to get position
[14:13] <Laurenceb> windshear is the same I think...
[14:14] Tigga (n=chatzill@apple3feet.plus.com) left irc: Read error: 110 (Connection timed out)
[14:14] <shellevil> contemplating going from the psuedorange outputs into your kalman?
[14:14] <Laurenceb> nope
[14:14] <Laurenceb> there is a bit of lag, but according to the datasheet you can reduce it
[14:15] <Laurenceb> what I cant quite see is why using velocity from the gps would give better results than position
[14:15] <Laurenceb> if the gps receiver is kalman filtering pos and velocity
[14:16] <shellevil> umm
[14:16] <Laurenceb> hmf... I can see it with say 2 measurements
[14:16] <shellevil> rounding?
[14:16] <Laurenceb> velocity error is around 20cm/s
[14:16] <Laurenceb> position is around 5m error
[14:17] <Laurenceb> so from 2 measuremnts youd get better velocity if you ignore the position info
[14:17] <shellevil> umm
[14:17] <Laurenceb> but if you use kalman filtered data and the filter has been running for some time...
[14:17] <Laurenceb> differentiating position should be better
[14:18] <shellevil> yeah
[14:18] <shellevil> the gps has a seperate position and velocity out?
[14:18] <Laurenceb> yes
[14:18] <shellevil> you've tried integrating the velocity out/differentiating the position to make sure it's not a stupid in your code?
[14:18] <Laurenceb> 3D velocity and pos in local ENU frame
[14:18] <Laurenceb> no
[14:19] <Laurenceb> but I have tracks on the ground
[14:19] <Laurenceb> and the filtering on the gps seems to work well
[14:19] <shellevil> I suppose velocity out is the 'purer' output
[14:19] <shellevil> position out would normally effectively be the integration of this
[14:20] <shellevil> but...
[14:20] <Laurenceb> position does tend to wander about by a few m
[14:20] <Laurenceb> look at the kml on the wiki
[14:20] <shellevil> maybe it weights the sat-noises differently
[14:20] <Laurenceb> hmm yeah maybe its not very clever
[14:21] <Laurenceb> your "locked in" to your sats with a pll
[14:21] <Laurenceb> so velocity should be extremely accurate
[14:21] <shellevil> or possibly, the residuals are screwing things up.
[14:21] <Laurenceb> <20cm/s error
[14:21] <shellevil> you've got a meter or three of error, the solution can bounce around this sphere/... quite easily
[14:21] <Laurenceb> position should use pseudorange measurements and velocity info
[14:22] <Laurenceb> so it shouldnt bounce about like that....
[14:22] <shellevil> as sats pop in and out of the solution due to low signal.
[14:22] <Laurenceb> yeah.... there appears to be glitches due to changing constellation
[14:22] <shellevil> For a position solution, you're always playing catch-up.
[14:23] <shellevil> Or predicting the future behaviour to some extent
[14:23] <Laurenceb> yeah.... maybe they fundged the kalman filter
[14:23] <shellevil> if you see what I mean.
[14:23] <Laurenceb> in most applications you only need pos to 3m
[14:23] <shellevil> With velocity - it's just simple maths.
[14:23] <Laurenceb> but having a nice velocity is good
[14:23] <Laurenceb> so maybe with they kalman filter for pos they put in a larger velocity error than is the case
[14:24] <shellevil> I wonder if sometimes velocity can screw up badly in some multipath conditions with poor sat geomerty
[14:24] <shellevil> so they've weighted it ...
[14:25] <Laurenceb> that probably leads to better position solution performance in conditions
[14:25] <Laurenceb> yeah like multipath
[14:25] <Laurenceb> thats probably their reasoning
[14:25] <Laurenceb> for sat nav and stuff you only need +-3m in really good signal conditions
[14:25] <shellevil> That, and the manoevering.
[14:26] <Laurenceb> but you dont want screwy velocities in multipath areas to throw you off badly
[14:26] <shellevil> if you're averaging over 1s position say, you've got to work out the accellerration
[14:26] <Laurenceb> yeah... guess that adds up
[14:26] <shellevil> If you've got a lot of vibration and jerk going on, as you might well have had...
[14:26] <Laurenceb> in cars ect
[14:26] <Laurenceb> hmm
[14:27] <shellevil> Why not just use velocity?
[14:27] <Laurenceb> no, you use integrated carrier phase
[14:27] <shellevil> I mean - the velocity out of the GPS - if it works better
[14:27] <Laurenceb> for velocity, so its integrated over the 1 second
[14:28] <shellevil> also - given that others have reported no problems with the IQ, I'd contemplate getting another and comparing
[14:28] <shellevil> or...
[14:30] <Laurenceb> well... the parity was wrong on my configuration packet
[14:30] <Laurenceb> and the BBRAM only had power removed for ~15s, not 35 as recommended in the datasheet
[14:31] <Laurenceb> theres a reset to factory setting command as well, might try that
[14:32] <Laurenceb> but yeah the mulitpath idea is a good one
[14:32] <Laurenceb> probably explains the poor position solution considering that its being filtered
[14:35] <Laurenceb> so I think for using velocity info from the gps, I think I'll stay with 100m layers
[14:35] <Laurenceb> rather than increasing to a few hundered
[14:37] <Laurenceb> with the position based soltuion, theres an error of about 29.5% with 100m layers
[14:38] <Laurenceb> I'm going to get this asap http://www.sparkfun.com/commerce/product_info.php?products_id=8889
[14:39] <edmoore> Laurenceb: drop hallam an email about that
[14:39] <Laurenceb> ok
[14:40] <Laurenceb> he has ublox5 modules?
[14:40] <edmoore> he may have a 406
[14:40] <Laurenceb> omg
[14:40] <edmoore> don't quote me. just drop him an email
[14:40] <Laurenceb> :P right
[14:40] <Laurenceb> yeah, I'd be happy just to borrow one
[14:44] <Laurenceb> tho I'd need a cable if I was going to do that
[14:45] <Laurenceb> well I need a bcable anyway
[14:46] <Laurenceb> some fun soldering
[14:48] <Laurenceb> about 0.3mm pitch by the look of it
[15:03] <Laurenceb> hmm apparently 0.5mm so should be solderable
[15:04] <Laurenceb> ribbon cable onto cable, then epoxy
[15:08] <Laurenceb> I still havent solved the radio issue.... I'm not sure if its worth resetting the radio...
[15:09] <SpeedEvil> :/
[15:09] <SpeedEvil> have you stuck it in the freezer - how cold did it get? or hot?
[15:09] <Laurenceb> havent tried that
[15:10] <Laurenceb> but it went dodgy about 1 minute after launch
[15:10] <Laurenceb> the only thing I can thinkl is that there was an RFI issue when it was first assembled, and I had to stick a cap on the enable line to the radio
[15:14] <Laurenceb> its a bit complex resetting the radio with no glitches... but should be possible
[15:17] <SpeedEvil> 1 min... that's only 300m or so isn't it?
[15:18] <Laurenceb> yeah
[15:18] <SpeedEvil> don't suppose there are sealed cans?
[15:19] <Laurenceb> it may have been dodgy all the time - we tested it with eds macbook about 30 minutes before launch, but then I turned it off
[15:19] <SpeedEvil> but 300m is a very small pressure drop to get upset with
[15:19] <SpeedEvil> yeah
[15:19] <Laurenceb> I was too busy to check it before launch
[15:19] <Laurenceb> just tested the flight computer was running over serial
[15:19] <Laurenceb> I tuned in the radio when it was about 300m altitude
[15:20] <Laurenceb> and it seemed to be modulating 0xoF or 0x00
[15:20] <SpeedEvil> this is a 'smart' radio?
[15:21] <SpeedEvil> or a dumb modulation volts in
[15:21] <Laurenceb> avr modulating a radiometrix
[15:21] <SpeedEvil> ah
[15:21] <Laurenceb> http://wiki.ukhas.org.uk/_media/projects:dscn0979.jpg
[15:22] <Laurenceb> a more recent version of that
[15:24] <SpeedEvil> worked back on ground?
[15:25] edmoore (n=edmoore@pomegranate.chu.cam.ac.uk) left irc:
[15:27] <Laurenceb> when I power cycled it
[15:33] Laurenceb (i=83e34f19@gateway/web/ajax/mibbit.com/x-c1d3b9a96ba95e95) left irc: "http://www.mibbit.com ajax IRC Client"
[16:02] hallam (i=4e20ad96@gateway/web/ajax/mibbit.com/x-6b8d98dd65931513) joined #highaltitude.
[16:18] DRebellion (n=DRebelli@unaffiliated/drebellion) joined #highaltitude.
[16:30] <DRebellion> Do we need to make any consideration to ropes freezing up and snapping or similar?
[16:30] <DRebellion> Or will any rope do?
[16:33] <hallam> it's worth considering, I think polypropylene is meant to be better than nylon
[16:33] <hallam> you do definitely get cold effects on some plastics
[16:35] <DRebellion> "Polypropylene is liable to chain degradation from exposure to UV radiation such as that present in sunlight."
[16:35] <DRebellion> =/
[16:35] <gordonjcp> that's over a fairly long time
[16:37] <hallam> yeah weeks-months
[16:37] <DRebellion> makes sense
[16:38] <hallam> you do see some cool UV effects though
[16:38] <hallam> some superglue came back green on nova 10
[16:38] <DRebellion> hehe
[16:38] <hallam> are you guys still planning to measure UV levels? that would be pretty interesting
[16:38] <hallam> I'm sure the data is available somewhere but it would be fun to measure it
[16:38] <DRebellion> hallam, we've sort of dropped it
[16:38] <DRebellion> It depends
[16:38] <hallam> ok not to worry
[16:39] <DRebellion> We need to write up a more formal proposal tonight for the school
[16:39] <DRebellion> Maybe we'll think about adding some more
[16:50] aLeXBrEtOn (n=Alexande@client-80-5-40-149.cht-bng-014.adsl.virginmedia.net) joined #highaltitude.
[17:12] Laurenceb (n=laurence@dyres221-35.surrey.ac.uk) joined #highaltitude.
[17:12] <Laurenceb> hello
[17:13] <hallam> hey
[17:13] <Laurenceb> get my email?
[17:14] <Laurenceb> ah
[17:14] <hallam> yes, get my reply?
[17:14] <Laurenceb> sure, ok I'll email the manufacturer
[17:44] icez (n=icez@unaffiliated/icez) joined #highaltitude.
[18:03] bfirsh (n=ben@scooby.firshman.co.uk) left irc:
[18:19] aLeXBrEtOn (n=Alexande@client-80-5-40-149.cht-bng-014.adsl.virginmedia.net) left irc:
[18:39] Hiena (n=Hiena@81.93.195.181.datatrans.hu) joined #highaltitude.
[18:39] aLeXBrEtOn (n=Alexande@client-80-5-40-149.cht-bng-014.adsl.virginmedia.net) joined #highaltitude.
[18:41] Action: Laurenceb has alm ost finished hardware improvements...
[18:41] <Laurenceb> retrimmed it so the COG is ~10mm further forward, and rebuilt the front spar assembly
[18:43] <SpeedEvil> :)
[18:47] <Laurenceb> I discovered it was out of alignment by about 20mm before, which caused a load of stree in the front of the electronics enclosure
[18:47] <DRebellion> Can anybody spot anything missing from the Hardware section here? http://wiki.ukhas.org.uk/projects:alien . I'm running out of things to research ; P
[18:47] <Laurenceb> probably the main reason it cracked
[18:49] <SpeedEvil> DRebellion: laser cannon.
[18:49] <Laurenceb> compactflash is for the camera?
[18:50] <DRebellion> Laurenceb, yes
[18:50] <DRebellion> aLeXBrEtOn, how did you get the 8gb figure?
[18:50] <aLeXBrEtOn> jessops
[18:51] <DRebellion> aLeXBrEtOn, right...
[18:51] <aLeXBrEtOn> cheapest one
[18:51] <DRebellion> so you didn't calculate
[18:51] <DRebellion> in the air for this amount of time
[18:51] <DRebellion> want to do this interval
[18:51] <aLeXBrEtOn> went up to 90 quid
[18:51] <DRebellion> size of image...
[18:51] <aLeXBrEtOn> oh that
[18:51] <aLeXBrEtOn> wait
[18:51] <aLeXBrEtOn> how long will the flight be?
[18:51] <DRebellion> #ha ?
[18:51] <DRebellion> 2.5 hours?
[18:51] <aLeXBrEtOn> OK
[18:52] <aLeXBrEtOn> depends on the quality/size setting you want
[18:52] <aLeXBrEtOn> I worked it out with highest of both
[18:52] <DRebellion> aLeXBrEtOn, we want highest quality yes
[18:52] <DRebellion> so how big is that sort of high quality image?
[18:52] <DRebellion> what format?
[18:52] <aLeXBrEtOn> in that case around 2MB for 1 pic
[18:53] <aLeXBrEtOn> jpeg
[18:53] <DRebellion> fine
[18:53] <SpeedEvil> also - most cameras of that era will probably not take that many pictures
[18:53] <aLeXBrEtOn> 2272x1704 pics
[18:53] <DRebellion> SpeedEvil, why not?
[18:53] <aLeXBrEtOn> 'superfine' setting
[18:53] <SpeedEvil> you're unlikely to be able to take 4000 shots before battery runs out
[18:53] <aLeXBrEtOn> it's not THAT old
[18:53] <aLeXBrEtOn> but you have a point
[18:53] <aLeXBrEtOn> though coming down...we don't care
[18:53] <SpeedEvil> CF -> relatively old
[18:53] <DRebellion> SpeedEvil, it takes AA i think?
[18:53] <DRebellion> aLeXBrEtOn, ?
[18:53] <aLeXBrEtOn> as long as we have apogee pics and going up
[18:53] <SpeedEvil> yeah
[18:54] <SpeedEvil> Ideally, you want to check that.
[18:54] <SpeedEvil> stick some NiMH in it, charged fully, and hold the shitter button down.
[18:54] <DRebellion> SpeedEvil, so we can string together a tonne of AAs
[18:54] <SpeedEvil> See how many pics you get
[18:54] <SpeedEvil> that works
[18:54] <DRebellion> SpeedEvil, we're going for Lithium Iron DiSulfide (Li/FeS2)
[18:54] <aLeXBrEtOn> well a 4GB will still hold 2000 pics
[18:54] <aLeXBrEtOn> lol@ the shitter button
[18:55] <SpeedEvil> Also
[18:55] <aLeXBrEtOn> we can cut down to 4GB, 25 quid
[18:55] <SpeedEvil> yes, 'high' detail is 'obvious'
[18:55] <DRebellion> Well, if the flight lasts 2.5 hours
[18:55] <SpeedEvil> but - look at the shots on the wiki
[18:55] <DRebellion> that's only 9000 seconds
[18:55] <SpeedEvil> thre aren't really any significant high detail bits once you're up.
[18:56] <aLeXBrEtOn> yeah, but we want to be able to enlarge the pics
[18:56] <DRebellion> indeed
[18:56] <aLeXBrEtOn> 2272x1704=more than A4?
[18:56] <DRebellion> aLeXBrEtOn, you need to do a test on how long the batteries will last then
[18:56] <aLeXBrEtOn> trouble is my NIMHs are crap to the max
[18:56] <DRebellion> aLeXBrEtOn, grab some of those L91/L92s and see how many
[18:56] <aLeXBrEtOn> the ones I charged yesterday lasted 0 seconds
[18:56] <SpeedEvil> aLeXBrEtOn: worst case - Google!
[18:57] <SpeedEvil> aLeXBrEtOn: some review sites will tell you the number of pics
[18:57] <aLeXBrEtOn> I believe lithium batts are rated at around 700 pics
[18:57] <SpeedEvil> Or failing that, RTFM.
[18:57] <aLeXBrEtOn> they probably test that on regular cameras like mine^
[18:57] <SpeedEvil> Umm.
[18:57] <SpeedEvil> They test them on whatever'll give the best numbers of course :)
[18:58] <aLeXBrEtOn> well they want a realistic number
[18:58] <aLeXBrEtOn> to minimise complaints ;-)
[18:58] <DRebellion> aLeXBrEtOn, look man! Just do a test!
[18:58] <aLeXBrEtOn> I ONLY GOT 699!
[18:58] <SpeedEvil> But - look at the makers website, get the manual, and look at it
[18:58] <SpeedEvil> it'll at least give you an optimistic idea
[18:58] <aLeXBrEtOn> I will need to splash out 7 quid on batteries
[18:59] <SpeedEvil> aLeXBrEtOn: or get some new AA NiMH ones. They'll be broadly similar - though won't work for flight bats
[18:59] <aLeXBrEtOn> aha
[18:59] <aLeXBrEtOn> with NIMH, manual says 1000 shots
[18:59] <aLeXBrEtOn> so more with the lithiums
[18:59] <SpeedEvil> with what capacity nimh?
[18:59] <aLeXBrEtOn> maybe 1200?
[18:59] <DRebellion> aLeXBrEtOn, so let's say take a picture every 10 seconds
[18:59] <DRebellion> simple
[19:00] <DRebellion> that's really conservative
[19:00] <SpeedEvil> also - you can't take 1000 shots with 10 minutes between
[19:00] <SpeedEvil> it'll use significant power when on probably
[19:00] <DRebellion> SpeedEvil, surely at one every 10 seconds it won't be wasting that much power idling?
[19:00] <aLeXBrEtOn> those are 1600mA batts
[19:00] <DRebellion> the specs would have taken that into account
[19:00] <DRebellion> aLeXBrEtOn, we're going to have to test
[19:00] <aLeXBrEtOn> the lithiums are more, no?
[19:01] <aLeXBrEtOn> gimme 7 quid then :P
[19:01] <SpeedEvil> DRebellion: you need to read the small print of what the '1000 shots' in the manual is
[19:01] <aLeXBrEtOn> I did
[19:01] <aLeXBrEtOn> with 1600mA NIMH batteries, LCD off
[19:01] <DRebellion> aLeXBrEtOn, that probably includes flash?
[19:01] <aLeXBrEtOn> the canon batteries
[19:01] <aLeXBrEtOn> no I don't think it does
[19:01] <aLeXBrEtOn> or maybe they factor flash for one in 10 shots or something
[19:01] <SpeedEvil> I'd offer you money, but offering schoolkids money tends to get frowned on :) I could swap you 4 lithium AAs for moving 20 tons of gravel :)
[19:02] Action: SpeedEvil is trying to work out how to do his drive.
[19:02] <DRebellion> hah
[19:02] <aLeXBrEtOn> hehe
[19:02] <SpeedEvil> (without lots of work myself)
[19:02] <aLeXBrEtOn> dinner calls
[19:02] <DRebellion> aLeXBrEtOn, have fun
[19:02] <aLeXBrEtOn> an omelette awaits ;-)
[19:03] Action: SpeedEvil has sausage and proto-mash.
[19:03] <DRebellion> Finally!
[19:04] <DRebellion> mum has bothered to dig up her old ericsson
[19:04] <DRebellion> w800i
[19:04] <SpeedEvil> I'd suggest NiMH - you're going to want to do more than one test.
[19:04] Action: DRebellion checks the ericsson AT commands developer reference
[19:04] <SpeedEvil> For example - which uses more power - 1000 super, or 2000 'fine' pics
[19:04] <DRebellion> SpeedEvil, and we can expect very similar results to LiFeS2?
[19:04] <SpeedEvil> for example
[19:05] <SpeedEvil> DRebellion: not hugely similar - but they will be broadly comparably. You might get 30% more with lithium for example - but if you get 1.4* more pics with 'fine' resolution on NiMH, you're likely to get similar on Lithium
[19:12] RocketBoy (n=Steve@217.47.75.8) joined #highaltitude.
[19:13] <Laurenceb> hi steve
[19:13] <RocketBoy> hi
[19:14] <Laurenceb> the main reason the front smashed off when we dropped it out of the tree was the front spar was out by 20mm
[19:14] <Laurenceb> it was being bent forwards, so there was a load of stress in the foam enclosure
[19:15] <RocketBoy> ah - I noticed it was bending slightly - but assumed that was designed
[19:17] <hallam> DRebellion: use lithium AAs, they'll be fine for a flight taking pictures constantly
[19:17] <Laurenceb> I wonder if the enclosure could be lagged in fibreglass
[19:17] <hallam> bench test it if you want, but be aware that you won't get quite as much life out of them in the cold
[19:23] <Laurenceb> are CU spaceflight still planning on a pointed antenna?
[19:23] <SpeedEvil> There is a GPLEMU engine managment unit?
[19:23] <SpeedEvil> oops
[19:24] <hallam> Laurenceb: yes, it's in the works
[19:25] <hallam> need to poke the newbies we assigned to it
[19:25] <Laurenceb> considered using standard antenna pointing gear?
[19:25] <hallam> we are
[19:25] <Laurenceb> cool
[19:25] <hallam> SPID antenna rotator
[19:25] <hallam> just need to make up cables and get it on the tower
[19:25] <Laurenceb> they have a rig here using pointing gear with a hacked up control board
[19:26] <Laurenceb> then you ssh into the machine an slew the antennas
[19:26] <Laurenceb> uses two antenna rotators and some alu bar
[19:26] edmoore (n=edmoore@pomegranate.chu.cam.ac.uk) joined #highaltitude.
[19:27] <Laurenceb> IIRC the yagis are 434 or something close
[19:27] <Laurenceb> then it runs a script with all the sats it needs to talk to
[19:27] <hallam> "here" = SSTL?
[19:27] <hallam> hi edmoore
[19:27] <Laurenceb> yes
[19:27] <edmoore> hi
[19:29] <Laurenceb> theres four yagis on each rig, and four different rigs
[19:29] <edmoore> what film did Bill Murray play a doctor in?
[19:31] <SpeedEvil> Doctor Who and the Golf Course.
[19:32] <Laurenceb> http://cgi.ebay.co.uk/linear-actuator_W0QQitemZ260381243890QQcmdZViewItemQQptZUK_BOI_Industrial_Automation_Control_ET?hash=item260381243890&_trksid=p3286.c0.m14&_trkparms=72%3A1683|66%3A2|65%3A12|39%3A1|240%3A1318
[19:33] <Laurenceb> alternatively one rotator and a linac for elevation
[19:33] <hallam> the spid we have is 2-axis
[19:33] <Laurenceb> ah neat
[19:33] <Laurenceb> then your sorted
[19:34] Action: Laurenceb ponders a linac controlled parafoil
[19:34] Action: SpeedEvil votes for a lineac firing protons.
[19:35] <hallam> they're not too speedy
[19:35] <Laurenceb> or small...
[19:36] <Laurenceb> my other idea was a winch servo, but speed*turns seems to be very limited
[19:36] <DRebellion> Guys, I have a Sony Ericsson w800i sitting in front of me. There are four metal pads for interfacing with it, arranged in a curve. Anybody know what corresponds to each pad? (Tx, Rx, etc)
[19:37] <Laurenceb> pinouts.ru
[19:37] <DRebellion> Laurenceb, tried it
[19:37] <Laurenceb> multimeter?
[19:37] <DRebellion> Laurenceb, pain; )
[19:38] <aLeXBrEtOn> oh make some sacrifices would you?
[19:38] <edmoore> Ghost Busters, fyi
[19:38] <SpeedEvil> DRebellion: are these the normal contacts on the bottom?
[19:38] <SpeedEvil> edmoore: don't cross the streams!
[19:38] <DRebellion> SpeedEvil, just flat pads
[19:38] <edmoore> bbl
[19:38] <SpeedEvil> DRebellion: those are probably usually power, for charging, and handset
[19:38] <DRebellion> SpeedEvil, about a mm across
[19:38] <DRebellion> gold
[19:39] <DRebellion> no
[19:39] <DRebellion> they're not the ones on the bottom
[19:39] <SpeedEvil> ebay
[19:39] <DRebellion> that is 12-pin
[19:39] <SpeedEvil> search yourphone cable
[19:39] <DRebellion> these are on the circuit board itself
[19:39] <Laurenceb> hmm is there such a thing as a linear actuator with basically a rod of some sorts thats driven through a control unit from one end to the other?
[19:39] <SpeedEvil> ah
[19:39] <Laurenceb> i.e. control unit is central
[19:39] <SpeedEvil> Laurenceb: yes
[19:39] <SpeedEvil> Laurenceb: but I've not seen small ones
[19:39] <SpeedEvil> Laurenceb: or cheap ones
[19:39] <Laurenceb> :-/
[19:40] <SpeedEvil> you want to lose the cable?
[19:40] <DRebellion> this looks promising http://pinouts.ru/CellularPhonesCables/sony_ericsson_service_cable_pinout.shtml
[19:40] <SpeedEvil> string
[19:40] <DRebellion> but no layout thing
[19:40] <Laurenceb> not necessarily
[19:40] <SpeedEvil> Laurenceb: cable spool I mean
[19:40] <Laurenceb> no
[19:41] <Laurenceb> but winch servos are slow
[19:41] Action: SpeedEvil is getting his driving licence - for various reasons postponed till now.
[19:41] Action: SpeedEvil has been checking insurance online.
[19:41] <SpeedEvil> First reactions - how many people do I get to kill for that amount?
[19:41] <gordonjcp> hah
[19:42] <gordonjcp> SpeedEvil: thought you were an old fart/
[19:42] <SpeedEvil> ~390 for a 1l micra
[19:42] <gordonjcp> o_O
[19:42] <Laurenceb> SpeedEvil: how have you been driving?
[19:42] <SpeedEvil> gordonjcp: I am - due to not having a car of my own, no great reason to bother passing test.
[19:42] <gordonjcp> hm
[19:42] <aLeXBrEtOn> oh my god.
[19:42] <gordonjcp> significantly over 25 though?
[19:43] <aLeXBrEtOn> they found a japanese guy who survived both atom bombs ;-)
[19:43] <SpeedEvil> Laurenceb: till recently, I've been driving my mum around.
[19:43] <SpeedEvil> gordonjcp: yes, a bit.
[19:43] <gordonjcp> SpeedEvil: should come down with even one year's no-claims
[19:43] <SpeedEvil> gordonjcp: yeah - brand new shiny licence - which seems to boost it into the stratosphere.
[19:43] <gordonjcp> aye
[19:44] <hallam> is this time to grumble about trying to get balloon insurance? far, far more money for far, far less risk than car ins.
[19:44] <SpeedEvil> Which would be - strangely - on-topic.
[19:44] <gordonjcp> SpeedEvil: have you got a car yet?
[19:44] <Laurenceb> http://www.cornwallmodelboats.co.uk/acatalog/hitec_hs785.html
[19:44] <Laurenceb> that would do, but so slowwwwww
[19:44] <SpeedEvil> gordonjcp: well - soon - above micra
[19:44] <SpeedEvil> Laurenceb: rack and pinion and motor?
[19:45] <gordonjcp> SpeedEvil: heh, if you're not set on *that* car, shop around and find a cheaper-to-insure car ;-)
[19:45] <Laurenceb> hmmm too much to go wrong unless its made very well
[19:45] <gordonjcp> you'd be surprised what comes out as "cheap"
[19:45] <Laurenceb> I dont have enough kit
[19:45] <SpeedEvil> gordonjcp: it's a poverty spec 8 year old 1l micra.
[19:45] <hallam> Laurenceb: maybe lego technic, glued with CA once you get it right?
[19:45] <SpeedEvil> gordonjcp: there can be cheaper to insure?
[19:45] <gordonjcp> SpeedEvil: yup
[19:45] <SpeedEvil> gordonjcp: wacky
[19:45] <Laurenceb> hehe lego
[19:46] <gordonjcp> SpeedEvil: I can insure a '91 Jag for about half as much as a '91 Escort
[19:46] <gordonjcp> SpeedEvil: adding me to my gf's Polo insurance made it cheaper overall
[19:47] <SpeedEvil> gordonjcp: Okaay. Jaguar - what model - for giggles?
[19:48] <gordonjcp> oh, can't remember now, some sort of XJ40
[19:48] <gordonjcp> it was a 4.2
[19:49] <gordonjcp> more seriously it's probably worth looking at Volvos and Peugeot estates and stuff
[19:50] <hallam> I guess the idea is that micras are often bought by reckless 17-year-olds?
[19:51] <gordonjcp> hallam: in the belief that they will be cheap to insure
[19:51] <gordonjcp> but actually turn out to be less forgiving of "spirited" driving than the sporty little cars they wish they had
[19:52] <SpeedEvil> Micra is surprisingly capable in some ways.
[19:52] <gordonjcp> so in fact, said reckless 17-year-old is potentially less likely to have a prang in a Saxo VTS than an ordinary one
[19:52] Action: SpeedEvil has had bath, 40 fence-posts, ... in the back, ...
[19:52] <gordonjcp> SpeedEvil: I had two of the old boxy-shaped ones
[19:52] <gordonjcp> a D-reg and a E-reg
[19:52] <gordonjcp> wish I still did, too
[19:52] <SpeedEvil> (a plastic bath, I have not filled it with water)
[19:52] <gordonjcp> I could sling one on davits off the roof of the CX
[19:53] <gordonjcp> winch it down when I need something little for nipping into town
[19:53] <SpeedEvil> :)
[19:53] <gordonjcp> they had the most comfortable seats and driving position of any car I've ever driven
[19:53] <gordonjcp> and I've driven a lot of cars
[19:53] <gordonjcp> some of them quite good cars, too
[19:53] <SpeedEvil> I actually also have a broken renault laguna too. And Jag XJR-4 was 763 :)
[19:54] <gordonjcp> SpeedEvil: so roughly twice as much for a *silly* jag
[19:54] <gordonjcp> what's wrong with the Creature from the Black Laguna?
[19:54] <SpeedEvil> gordonjcp: it has no ECU
[19:54] <gordonjcp> that would be an issue
[19:54] edmoore (n=edmoore@pomegranate.chu.cam.ac.uk) left irc: Read error: 113 (No route to host)
[19:54] <SpeedEvil> gordonjcp: I need to install a DIY one. Or rather - the cost of getting it fixed exceeds the cost of the car.
[19:54] <gordonjcp> SpeedEvil: scrappy?
[19:54] Action: gordonjcp has worked out a belter of a route up to Dunkeld
[19:55] <gordonjcp> it's not like there's any shortage of donor Lagunas
[19:55] <SpeedEvil> gordonjcp: I have ECU kit. I may try to install it.
[19:55] <gordonjcp> okay
[19:55] <SpeedEvil> gordonjcp: partly 'just because'.
[19:55] <gordonjcp> yeah
[19:55] <gordonjcp> I quite fancy fitting the breathing apparatus and injection kit off an Espace to the CX "just because"
[19:57] <gordonjcp> jeez, they did a 2.1 naturally-aspirated Douvrin engine
[19:57] <gordonjcp> 64bhp, that would be somewhat pedestrian in the CX
[19:58] <SpeedEvil> And laguna is more than micra. Oh well.
[19:58] <DRebellion> Judging by this page, http://stefans.datenbruch.de/k750i/#interfacing , we should be able to use AT commands via the 12-pin on the bottom - is this plausible?
[19:59] <SpeedEvil> DRebellion: possible - look for 'cables' on ebay - they may suit too
[19:59] <hallam> the "virtual" serial interface over USB doesn't sound like it's what you want, but that doesn't mean there aren't real serial pins there too
[20:04] <DRebellion> hallam, there are real serial pins
[20:04] <DRebellion> I just can't find any layout info
[20:05] <DRebellion> hallam, the four underneath the battery/SIM would be likely candidates, no?
[20:06] <aLeXBrEtOn> gents, precisely how necessary is GSM?
[20:06] <SpeedEvil> It's backup.
[20:06] <DRebellion> aLeXBrEtOn, it's made of pure win
[20:06] <aLeXBrEtOn> thank you speedevil, for explaining that
[20:07] <aLeXBrEtOn> 'pure win' ain't too descriptive :P
[20:07] <DRebellion> ; )
[20:07] <aLeXBrEtOn> I think we can get away with a 4GB CF card
[20:07] <aLeXBrEtOn> perhaps even 2GB
[20:08] <DRebellion> http://pinouts.ru/CellularPhones-P-W/sony_ericsson_j100_pinout.shtml
[20:08] <DRebellion> this looks promising
[20:08] <SpeedEvil> ebay works well sometimes for small CF cards
[20:08] <aLeXBrEtOn> 2GB will give us 1k pics
[20:08] <DRebellion> although it is for 'j100'
[20:08] <DRebellion> it uses the k750i pins
[20:08] <DRebellion> and the k750 is identical to the w800i
[20:08] <DRebellion> apart from cosmetics and software
[20:08] <hallam> DRebellion: I really couldn't say whether the ones behind the battery or on the bottom would be more likely, could be either
[20:09] <aLeXBrEtOn> lol
[20:09] <hallam> aLeXBrEtOn: make sure to allow about an hour of on-time for all the electronics before launch, in your battery life and card space calcs
[20:09] <DRebellion> "J100i has a very limited AT commands set. Strange but it responds mostly with OK only. Pin out is correct but not complete. Microphone and speaker connection is possible like in W800i.
[20:09] <DRebellion> Boris. "
[20:09] <aLeXBrEtOn> Rx: may be tx / tx: may be rx
[20:09] <DRebellion> very encouraging
[20:09] <DRebellion> aLeXBrEtOn, indeed
[20:09] <aLeXBrEtOn> very useful indeed
[20:10] <aLeXBrEtOn> may as well pick one at random :P
[20:10] <hallam> you don't want to start filling helium until you're sure the electronics are working, so they need to be able to run for a while before liftoff
[20:10] <DRebellion> aLeXBrEtOn, nah
[20:10] <DRebellion> Will have to design a simple connector in tech tomorrow
[20:10] <aLeXBrEtOn> I was joking ;-)
[20:10] <hallam> DRebellion: just get a usb-to-TTL converter and experiment already
[20:10] <DRebellion> hallam, need to create something to connect to the pins though
[20:10] <DRebellion> they're tiny
[20:11] <DRebellion> and right next to each other
[20:11] <DRebellion> ; )
[20:11] <Laurenceb> ebay, data connector
[20:12] <aLeXBrEtOn> just found a shit CF card
[20:13] <aLeXBrEtOn> 32MB
[20:13] <DRebellion> Laurenceb, they're all usb ones
[20:15] <Laurenceb> if theres a serial pin, your ok
[20:15] <hallam> you can probably solder kynar wire direct to the pins
[20:15] edmoore (n=edmoore@pomegranate.chu.cam.ac.uk) joined #highaltitude.
[20:21] Action: Laurenceb needs a parametric servo search tool
[20:22] <Laurenceb> http://cgi.ebay.co.uk/SAIL-WINCH-SERVO-HITEC-785HB-FOR-RADIO-CONTROL-BOAT-_W0QQitemZ220379042547QQcmdZViewItemQQimsxZ20090317?IMSfp=TL090317194003r30042
[20:22] <edmoore> Laurenceb: try me
[20:22] <Laurenceb> thats nice, just too slow
[20:22] <edmoore> I evaluated about 20 for walking robots
[20:22] <edmoore> + antweight robots on robot wars
[20:22] <edmoore> oh the memories
[20:22] <Laurenceb> I'm looking for something for a parafoil
[20:23] <Laurenceb> so winch servo with several turns
[20:23] <edmoore> do you really need several turns?
[20:23] <Laurenceb> but faster than what I linked
[20:23] <edmoore> are longer arms not an option?
[20:23] <Laurenceb> yeah, keeps it compact
[20:23] <aLeXBrEtOn> camera works =)
[20:23] <edmoore> ok. well I'm not too hot on multiple turn ones
[20:23] <aLeXBrEtOn> after years of disuse
[20:23] <Laurenceb> I want to mount everything in a cf tube
[20:23] <edmoore> the hitec robot servos are multiple turn
[20:23] <edmoore> as are the bioloid ax-things that robotsavvy sell
[20:24] <Laurenceb> ooh yeah, forgot about those
[20:24] <edmoore> serial interface too
[20:24] <Laurenceb> I'll look at ze specs
[20:24] <edmoore> current feedback
[20:24] <edmoore> etc etc
[20:26] <Laurenceb> serial would be a pain, no spare ports
[20:26] <edmoore> I *think* it may be able to do normal stuff too
[20:26] <edmoore> don't quote me
[20:28] Action: Laurenceb hates people using serial for everything
[20:28] <Laurenceb> i2c or spi
[20:28] <edmoore> agreed
[20:29] <hallam> mm, lpc2368 with 4 uarts
[20:29] <hallam> serial tx is possibly easier to bit-bang than servo pwm though
[20:30] <Laurenceb> its daft getting a uC specifically for the uarts
[20:30] <Laurenceb> hardware pwm
[20:30] <Laurenceb> but yeah i2c or spi would be the sensible option
[20:30] <Laurenceb> edmoore: who sells those>
[20:31] <Laurenceb> oh robosavvy sorry
[20:31] DanielRichman (n=DanielRi@89.243.213.125) joined #highaltitude.
[20:32] <hallam> Laurenceb: 2368 has a bunch of other nice stuff too
[20:32] <Laurenceb> fpu?
[20:33] <edmoore> if only
[20:33] <edmoore> 60MFLOPS
[20:33] <edmoore> gosh
[20:33] <hallam> nah, it's still a cheap microcontroller
[20:33] <hallam> but still
[20:33] <edmoore> Laurenceb: in the arm9 camp you've the ep9302
[20:33] <edmoore> tqfp package
[20:33] <edmoore> ~200Mhz, FPU
[20:33] <Laurenceb> yeah, for imu time
[20:33] <hallam> proper SD card interface, usb, dac, adc, ethernet, bunch o' sync serial ports
[20:34] <hallam> it's a pity the ep9302's peripherals are a bit outdated
[20:34] <edmoore> agreed
[20:34] <edmoore> you need to put an fpga on the mem bus :)
[20:34] Action: Laurenceb was planning on a mega324p for a parafoil
[20:34] <Laurenceb> all you need really
[20:35] <edmoore> they can shot down mach 2 migs with mach 3 missiles using less than a Z80 :)
[20:36] <hallam> yeah I don't see why plain IMUs need anything fancy at all
[20:36] <Laurenceb> oh I was talking to some guys from kinetiq about rocket gps
[20:36] <edmoore> you can land a 10m/s toy kite in a field with a mega324!
[20:36] <edmoore> oh?
[20:36] <Laurenceb> yeah, apparently pll is a bad idea
[20:36] <Laurenceb> altho biasing it with accels will work
[20:36] <hallam> IMO the only thing you need serious number crunching for is soft gps, soft radio and image processing
[20:36] <Laurenceb> the proper way to do it is with a fll
[20:36] <hallam> what do they suggest instead, fll?
[20:36] <edmoore> well that's slightly disingenuous, 'proper' imu's don't need nearly as much cpu as the characteristics of the sensors are so much better
[20:36] <Laurenceb> yeah
[20:37] <hallam> true, but with the bandwidths of the sensors we have, I can't see how you need high-speed processing
[20:37] <hallam> how many cycles can you possibly use to process 1kHz data?
[20:37] <Laurenceb> you can crunch quaternions from an accel/magno/3gyro unit at 50Hz with an avr
[20:37] <edmoore> thousands!
[20:38] <hallam> but not millions, right?
[20:38] <edmoore> sure
[20:38] <edmoore> and you can do cool things
[20:38] <Laurenceb> depends on the size of your state vector
[20:38] <hallam> Laurenceb: you'll have to explain to me again how FLLs work. but I'll be back in a few
[20:38] Nick change: hallam -> hallam|dinner
[20:38] <edmoore> normalize your quaternions with Carmack's fast inverse square root
[20:38] <Laurenceb> some military stuff is using ~35 conponent state vector
[20:38] <edmoore> and other sexy bits
[20:38] <Laurenceb> thats a lot of flops
[20:39] <Laurenceb> n!(n^3)
[20:40] <Laurenceb> 10^41 times more than the basic one I tried in matlab :P
[20:40] <Laurenceb> they must have a lot of shortcuts
[20:40] <Laurenceb> even with a gigaflop fpga
[20:42] <edmoore> I wouldimagine so
[20:42] <edmoore> though... a bunch of cascaded filters are a lot quicker than one big one
[20:47] <fergusnoble> hey guys
[20:48] <fergusnoble> who is actually in cambridge for this thing on saturday?
[20:48] <fergusnoble> is it just dan atm?
[20:49] <edmoore> I'm not sure I can be. But apparently this weekend is the only possible time it can be done (I don't recall ever being told why this is the case).
[20:49] <edmoore> Sal is
[20:49] <DRebellion> What's happening on saturday?
[20:49] <edmoore> Don't know about Rob
[20:49] <fergusnoble> Rob is not around, he just suggested it as he ran the sim
[20:50] <fergusnoble> DRebellion: a launch apparently
[20:50] <fergusnoble> but its looking like we dont have the people for it
[20:50] <fergusnoble> edmoore: i guess i could schlep up to cam for the w/e
[20:50] <edmoore> I'm not even sure badger is in cambridge
[20:50] <fergusnoble> no, jeremy has it
[20:50] <edmoore> exactly
[20:53] <edmoore> I've no idea about helium access with me.
[20:54] <edmoore> I'm not sure I can face asking after Peter made a big hoo-ha on Saturday night about how it could only be Churchill members.
[20:54] <edmoore> Carl has gone home...
[20:54] <edmoore> (obviously I will... just like an arcade game, I have limited life before I just sigh and fall over)
[20:56] <fergusnoble> well, i cant be bothered to push for it myself
[20:58] <fergusnoble> but obviously will help if i can
[20:59] <edmoore> I'll ask him, it's just going to be interesting. When I phoned him up on Saturday night (interupting his whisky) I got half an hour about how we can't let 'any old' people from the university come in and get keys to secure churcill places 'willy-nilly'. You've dealt with him enough to extrapolate the rest
[20:59] Nick change: hallam|dinner -> hallam
[21:00] <fergusnoble> yeah, maybe its best to stay out of his way for a bit
[21:00] <hallam> I think Jeremy is going back up tomorrow or something
[21:00] <hallam> Kirwan is so exasperating
[21:00] <fergusnoble> especially if the launch ends up not happening anyway
[21:00] <hallam> way more so for you Ed, of course
[21:00] <edmoore> I don't mind. sailors complaining about the sea etc. he usually comes through in the end
[21:01] <hallam> I wish he could be made to understand that CUSF is good for Churchill
[21:01] <edmoore> but we do seem hell-bent on launching it this weekend when just about all key team members aren't here
[21:01] <hallam> is it really hell-bent atm?
[21:01] <edmoore> everyone is talking about how it has to be this weekend all of a sudden
[21:01] <edmoore> and that the contract has to terminate on april 6th. In each of these cases, I have no idea why.
[21:02] <fergusnoble> edmoore: for cartoon network that is when they are launching the rest of their publicity campaign
[21:02] <edmoore> Just like the 800 competition suddenly had to be launched this holiday (now no longer the case, thankfully). It seems like self flagilation
[21:02] <hallam> I thought the latter is to do with when the cartoon series launches
[21:02] <fergusnoble> so its fairly fixed
[21:02] <edmoore> i heard from rob they could deal with another week
[21:03] <edmoore> hallam: when are you back?
[21:03] <fergusnoble> edmoore: i think the point was if there is a chance now, take it, as next week might not work out
[21:03] <edmoore> Does Jeremy have my eee?
[21:03] <edmoore> fergusnoble: that's indeed correct, but if there's no one around we have issues. we'll hammer it out in the meeting tomorrow, no doubt
[21:03] <fergusnoble> what meeting tomorrow?
[21:03] <edmoore> check about 14 emails
[21:04] <edmoore> just a meeting
[21:04] <hallam> edmoore: not until 14th april I'm afraid
[21:04] <edmoore> the more important question is the 2nd one
[21:04] <hallam> edmoore: yes, Jeremy has your eee
[21:04] <edmoore> cool
[21:04] <fergusnoble> edmoore: are you still in camb then?
[21:04] <hallam> and the yaesu, and the badger, and the yagi
[21:05] <edmoore> I am right now, yes
[21:05] <edmoore> but may/will probs be home this weekend
[21:05] <edmoore> depending on when the funeral is finally set
[21:05] <fergusnoble> yup, ok
[21:05] <edmoore> got a couple of supervisions this week so am up
[21:05] <hallam> is it fair to say that three reasonably experienced people can handle a relatively simple launch, chase issues and helium access aside?
[21:06] <fergusnoble> well the chase people can just use fldigi
[21:06] <edmoore> If Rob is up it should be ok
[21:06] <edmoore> he can fill a balloon and can chase
[21:06] <edmoore> the radio bit is easy to teach
[21:06] <hallam> fergusnoble: fldigi worked so much better on Ed's eee than Jeremy's laptop
[21:07] <fergusnoble> odd, any ideas why?
[21:07] <edmoore> they'll want a satnav between them
[21:07] <edmoore> or a very good GE cache
[21:07] <hallam> satnav essential IMO
[21:07] <edmoore> I would say so
[21:07] <edmoore> well no
[21:07] <edmoore> just very very very useful
[21:07] <edmoore> but not essential
[21:07] <hallam> very desireable
[21:07] <fergusnoble> im afraid i dont think it will be easy for them to set up the proper tracking software
[21:07] <hallam> Jeremy's laptop seemed to have the mic gain set very high
[21:07] <edmoore> essential means you'd go and buy one without telling anyone, and we can't have that :p
[21:08] <hallam> not with new expenses system ;)
[21:08] <hallam> the proper tracking software, while useful and reassuring, isn't essential with remote people doing tracking + prediction
[21:08] <fergusnoble> anyway, nobody has mentioned how well the fish scales worked?
[21:08] <edmoore> yes - spacenearus is good
[21:08] <edmoore> oh they were sort of as expected
[21:09] <edmoore> they said what we thought they would
[21:09] <edmoore> i.e. ballast + fill tube +/- 100g
[21:09] <hallam> still best to have the ballast weights, you think?
[21:09] <fergusnoble> can you give a bit more detail on how you used them?
[21:09] <edmoore> both is useful. but ballast is useful as it can give you an actual feel for ascent rate
[21:09] <edmoore> we turned them upside down and let the balloon pull them up
[21:09] <edmoore> there was not a breath of wind
[21:10] <edmoore> i think they would be less useful in windy conditions
[21:10] <fergusnoble> ok, and ballast on the balloon side to reduce the measured pull?
[21:10] <edmoore> at least with ballast you're only talking about vertical components
[21:10] <edmoore> we took the ballast off
[21:10] <edmoore> the measurements were seperate
[21:10] Hiena (n=Hiena@81.93.195.181.datatrans.hu) left irc: "-=Mi ez a szag? Hippit égettéll?=-"
[21:10] <hallam> you can mentally integrate the feel of the ballast over a longer period
[21:11] <fergusnoble> edmoore: remember i wasnt there, you will have to take it back a couple of steps
[21:11] <fergusnoble> what ballast where?
[21:11] <edmoore> so you know how we put weights in a bag
[21:11] <fergusnoble> yes
[21:11] <edmoore> weights = payload weight + free lift - fill tube
[21:11] <edmoore> we did that as usual
[21:12] <fergusnoble> ok, so where do the scales fit in with that?
[21:12] <edmoore> typing....
[21:12] <Laurenceb> what are you launchng?
[21:12] <edmoore> then we sealed up the balloon
[21:12] <edmoore> Laurenceb: not now
[21:12] <edmoore> and made a loop of string to hang from balloon neck
[21:13] <edmoore> we then put the fish scale hook into the loop (upside down) and made the fish scales take the full lift of the balloon
[21:13] <edmoore> that was the case for the lesser filled balloon
[21:14] <fergusnoble> ok, so you didnt actually use the scales to fill the balloon, just to work out what fill you had given it after the fact
[21:14] <edmoore> for the main balloon, we did the same thing but more simply substituded the ballast bag for
[21:14] <edmoore> oh gosh fergus just listen
[21:14] <edmoore> you're one of those people who asks questions in lectures 3 seconds before the topic moves to that :p
[21:14] <edmoore> for the main balloon, we did the same thing but more simply substituded the ballast bag for the scales
[21:14] <fergusnoble> well you just said thats what you did for the lesser filled balloon
[21:15] <edmoore> so the scales showed total lift - fill tube
[21:15] <fergusnoble> ok, thats what i imagined
[21:15] <edmoore> so we have two numbers - one arrived at during the fill, and one just measuring it afterwards
[21:15] <fergusnoble> so why didnt you do that for the first balloon!?
[21:15] <edmoore> we did
[21:16] <edmoore> and for the second we just measured the lift afterwards, as we wanted to get it up and away as the winds were picking up, the bottle lid froze up and seperated from the actual screw, and a bunch of other niggles
[21:16] <edmoore> oh yes, i forgot to mention
[21:16] <edmoore> the bottle valve handle actually parted company from the value itself
[21:16] <edmoore> that is bed
[21:16] <edmoore> we should tell jamie
[21:16] <edmoore> bad*
[21:17] <fergusnoble> if its the one on the bottle and not the reg they will come and take the bottle away and replace it
[21:18] <edmoore> I'm not concerned about it being replaced, I'm concerned that they should actually know it's an issue
[21:18] <edmoore> it's dangerous
[21:18] <fergusnoble> only a little i think
[21:18] <edmoore> if it broke in the on position, we'd be stuffed
[21:19] <fergusnoble> no, you still have the reg inline
[21:19] <fergusnoble> worst case you have to empty the bottle
[21:19] <gordonjcp> you wouldn't want to undo the regulator though, until it was properly empty ;-)
[21:19] <edmoore> exactly
[21:19] <edmoore> but we'd be stuffed for a launch, I mean
[21:19] <fergusnoble> but likely you turn it off at the reg and then find something to turn off the main with, like a spanner
[21:20] <edmoore> we had the 2nd bottle thankfully
[21:20] <edmoore> no, this is the annoying bit
[21:20] <edmoore> with the reg in place and with the protecting handling collar, you can't actually get a spanner to it
[21:20] <edmoore> you'd have to give you 2 90-degree bends
[21:20] <fergusnoble> ok, sucks, anyway, yeah let jamie know and get a new bottle
[21:20] <edmoore> you'd have to give one* (a spanner) two bends, even
[21:21] <fergusnoble> how much gas do we have left?
[21:21] <edmoore> 1 bottle - about 2 cubic meters
[21:21] <edmoore> - = 'minus'
[21:21] <fergusnoble> ok, cool, enough for a 1200g then probably
[21:22] <fergusnoble> with a lightish payload
[21:22] <edmoore> shame we don't have a 1.2
[21:22] <fergusnoble> oh yeah, we only have a 1.5 now?
[21:22] <edmoore> 1 x 1.5
[21:22] <edmoore> and 1 x 3kg
[21:22] <edmoore> i think iain has a point
[21:22] <fergusnoble> well, can just get another bottle sent up
[21:22] <fergusnoble> is the two bottles thing permanent arrangement now?
[21:22] <edmoore> we massively increase our envelope if we just stick it on a 3kg at 10m/s
[21:22] <fergusnoble> or was that too far to push peter :)
[21:23] <edmoore> no, it's not
[21:23] <hallam> I think we should try to keep the 2 bottle arrangement if poss
[21:23] <edmoore> well yes of course
[21:23] <hallam> it was really good to have a backup when the first bottle failed (even in a 1 balloon scenario)
[21:23] <edmoore> but as far as peter's concerned, it isn't. we'd have to adjust storage arrnagements
[21:23] <edmoore> i will add to to my martydom mission
[21:23] <hallam> what if we volunteered to tart the storage area up a bit?
[21:24] <hallam> add a new concrete platform or something
[21:24] <hallam> can't see him liking the sound of it, but apart from that minor problem
[21:24] <edmoore> he'd end up having to organize it
[21:24] <fergusnoble> hallam: i think the main thing peter doesnt like are things making his job more complicated
[21:24] <edmoore> I want to see if we can get sucure storage up top
[21:24] <edmoore> maybe with grounskeeping
[21:25] <hallam> that would be good
[21:25] <edmoore> much less far to push the bottle
[21:25] <edmoore> and they have sheds and stuff up there
[21:25] <fergusnoble> yeah, that would be good
[21:25] <hallam> a locked CUSF garden shed?
[21:25] <fergusnoble> we only need a blank patch of wall really
[21:25] <edmoore> if you think we could ever get a ehole shed, you've another thing coming
[21:25] <fergusnoble> in a secure area
[21:25] <edmoore> they're like gold around this place
[21:26] <hallam> if we paid for it?
[21:26] <fergusnoble> hallam: its never going to happen
[21:26] <hallam> they can be less than a grand, and well worth it to store stuff at launch site
[21:26] <fergusnoble> aside from the fact that we really dont need it
[21:26] <edmoore> no henry. housekkeping want them, ball committees want them, the casino wants them
[21:27] <edmoore> we'd have to pay not-actually-at-all-good-value-for-us amounts to get one
[21:27] <edmoore> oh an actual shed
[21:27] <edmoore> i see
[21:27] <edmoore> sorry i thought you meant renting a pre-existing one
[21:27] <edmoore> well i was going to look into one of the containers like archery have
[21:27] <hallam> nah, B&Q special
[21:27] <hallam> what's the archery one like?
[21:27] <edmoore> big and green
[21:28] <hallam> wood/metal/plastic?
[21:28] <edmoore> metal
[21:28] <edmoore> folded steel
[21:28] <edmoore> cheap
[21:28] <edmoore> but it has a lock
[21:28] <hallam> proper job
[21:28] <edmoore> basically, i will investigate
[21:28] <edmoore> i am sort of hoping we just just lock the bottles in some pre-existing secure place the groundkeeping have
[21:28] <edmoore> kiss
[21:29] <hallam> would be nice
[21:29] <hallam> http://www.homebase.co.uk/webapp/wcs/stores/servlet/ProductDisplay?langId=-1&storeId=20001&partNumber=1728525&Trail=searchtext%3ESHED
[21:29] <fergusnoble> yeah, im sure there is somewhere they can be tucked
[21:29] <fergusnoble> remember jamie said we can have any mountings and fittings we want for free
[21:29] <hallam> ok, off to watch the apprentice
[21:29] <hallam> ttyl
[21:29] <fergusnoble> he will even get them fitted for us
[21:30] <fergusnoble> so if its a case of just putting some brackets on a wall then thats no problem
[21:30] <edmoore> I am off too
[21:30] <edmoore> fergusnoble: what do you think re-cutdown suggestion?
[21:30] Simon-MPFH (n=simon@phantom.mpfh.co.uk) left irc: "Leaving"
[21:30] <fergusnoble> using nylon cord instead?
[21:31] <fergusnoble> sure, if it works it would make me more happy
[21:31] <edmoore> trying it out, yeah
[21:31] <fergusnoble> but is obviosuly not so convenient
[21:31] <fergusnoble> i think we should test some cable ties
[21:31] <hallam> woo, wifi in living room
[21:32] <fergusnoble> as its not clear how bad they are
[21:32] <edmoore> we'll have to see if we can find a decent make. there were a lot of duff ones we tried on sunday
[21:32] <edmoore> might be a dodgy batch
[21:32] <hallam> I feel way happier about cable ties in the sal cutdown versus cable ties directly as load-bearing in the line
[21:32] <fergusnoble> e.g. they always hold up when it comes to closing the balloon
[21:33] <edmoore> fergusnoble: no
[21:33] <edmoore> i just said exactly the opposite
[21:33] <edmoore> that's incorrect
[21:33] <edmoore> whether we had a duff bacth or not, we had two or three fail on us on sunday. I've not come across that before so I'm willing to assume there was something wrong with these particular ones
[21:34] <fergusnoble> edmoore: what happened
[21:34] <edmoore> an old mould or something.
[21:34] <edmoore> during tightening, in two of them the sharp tip of the ratchet-tooth broke
[21:35] <edmoore> and it just sprang open again
[21:35] <fergusnoble> sucks
[21:35] <edmoore> the third one it just was loose to begin with
[21:35] <fergusnoble> ok, anyway, when i said they hold up i mean ones that are working to begin with
[21:35] <fergusnoble> :p
[21:35] <edmoore> well in 2 of the three, they were
[21:36] <edmoore> hallam: I don't so how they aren't load-bearing in sal's ones
[21:36] <edmoore> they have to transmit the tension from balloon to payload through them.
[21:36] <fergusnoble> edmoore: its the less torsion that i think makes the difference
[21:36] <edmoore> sure there's some friction
[21:37] <hallam> it's more the repeated flexing I'm concerned about
[21:37] <hallam> fatigue
[21:37] <hallam> it's definitely not the same as e.g. the one that we think failed in nova 4
[21:38] <edmoore> easy tiger
[21:38] <edmoore> whilst it's not the same setup, you don't know what the exact failure mode was
[21:38] <hallam> of course not
[21:38] <hallam> I agree that trying cord or something else that's not a cable tie is worthwhile
[21:39] <edmoore> so we can't conclude that the problem goes away with this new design. It definitely needs some decent testing
[21:39] <hallam> sure
[21:39] <hallam> we do rely on cable ties in a lot of other places though
[21:39] <hallam> proper testing is worth doing
[21:39] <hallam> cable ties have the nice feature of being ratchetable tight, which I think is hard to do with cord - or can you make some special knot?
[21:39] <edmoore> generally not load bearing. This is the vid I showed fergus when he was arguing flexing killed ukhas 1: http://vimeo.com/1598522
[21:40] <edmoore> all I mean is, whilst there may not be torsion, there are still some non-triv dynamic loading effects, hanging underneath that thing
[21:40] <edmoore> hallam: you are correct re that - i have a thought though
[21:40] DanielRichman (n=DanielRi@89.243.213.125) left irc: "Leaving"
[21:40] <edmoore> i will explain when i get back from going out
[21:40] <hallam> fair point
[21:40] <edmoore> bbl
[21:40] <hallam> see you
[21:41] edmoore (n=edmoore@pomegranate.chu.cam.ac.uk) left irc:
[21:46] aLeXBrEtOn (n=Alexande@client-80-5-40-149.cht-bng-014.adsl.virginmedia.net) left irc:
[21:48] <Laurenceb> I like the pre dawn video
[21:48] <Laurenceb> you can see all the towns laid out in east anglia
[21:53] <Laurenceb> hmm the Dynamixel RX-64 Servo is good
[21:55] hallam (i=4e20ad96@gateway/web/ajax/mibbit.com/x-6b8d98dd65931513) left irc: "http://www.mibbit.com ajax IRC Client"
[21:59] aLeXBrEtOn (n=Alexande@client-80-5-40-149.cht-bng-014.adsl.virginmedia.net) joined #highaltitude.
[22:02] aLeXBrEtOn (n=Alexande@client-80-5-40-149.cht-bng-014.adsl.virginmedia.net) left irc: Client Quit
[22:17] DRebellion (n=DRebelli@unaffiliated/drebellion) left irc: Read error: 110 (Connection timed out)
[22:21] <Laurenceb> http://www.gpdealera.com/cgi-bin/wgainf100p.pgm?I=FUTM0099
[22:21] <Laurenceb> would I fry that with two lipo cells?
[22:23] <Laurenceb> bbl
[22:37] hallam (i=4e20ad96@gateway/web/ajax/mibbit.com/x-1b5e895d58e6cd83) joined #highaltitude.
[23:05] RocketBoy (n=Steve@217.47.75.8) left irc: Read error: 104 (Connection reset by peer)
[23:24] Tigga (n=chatzill@pc-232-235-103.magd.cam.ac.uk) joined #highaltitude.
[23:40] Action: Laurenceb back
[23:44] <Laurenceb> hallam: any experience with servos?
[23:54] rjharrison (n=rharriso@80.176.172.227) left irc: Read error: 110 (Connection timed out)
[00:00] --- Thu Mar 26 2009