highaltitude.log.20121211

[00:03] johnboiles (~Adium@bastion.sfo1.yelpcorp.com) left irc: Ping timeout: 255 seconds
[00:05] radim_OM2AMR (~radimmuti@213.215.84.62) left irc: Quit: radim_OM2AMR
[00:05] lz1dev (~lz1dev@46.47.80.192) joined #highaltitude.
[00:08] lz1dev (~lz1dev@46.47.80.192) left irc: Client Quit
[00:12] LazyLeopard (~irc-clien@chocky.demon.co.uk) left irc: Quit: Bye
[00:15] <hyte> Hi, if i just power the ublox6 should it spit out serial data all on its own?
[00:16] <craag> hyte: Yes.
[00:17] <hyte> thanks, I must have a problem receiving it. cheers
[00:18] <heathkid> how do you program the ublox6 to "flight mode" and does that do away with the 50k feet limit?
[00:19] <heathkid> sorry... 50km limit
[00:19] <heathkid> whatever it is
[00:19] <Darkside> look on the ukhas wiki
[00:19] <Darkside> theres code there
[00:19] <heathkid> thanks
[00:19] <Darkside> and it does away with the 18km limit
[00:19] <Darkside> i think it still caps out at 50km though
[00:19] <craag> hyte: You can try using ucenter on the PC to check the ublox is working.
[00:20] <hyte> yep, have done that and it works. Just thought it might be throwing the module a few bytes to get the data
[00:21] Dan-K2VOL (~Dan-K2VOL@74-134-73-252.dhcp.insightbb.com) joined #highaltitude.
[00:21] <fsphil> if the 50km limit is causing a problem, then congratulations :)
[00:23] <heathkid> I'm tired...
[00:23] <heathkid> sorry
[00:24] <heathkid> has anyone used UI-View32 for APRS?
[00:25] Dan-K2VOL (~Dan-K2VOL@74-134-73-252.dhcp.insightbb.com) left irc: Client Quit
[00:27] chris_99 (~chris_99@unaffiliated/chris-99/x-3062929) left irc: Quit: Leaving
[00:28] <arko_> jesus their website looks like a geocities prototype
[00:28] Nick change: arko_ -> arko
[00:28] <Darkside> the guy that write uiview died a few years back
[00:28] <Darkside> and in his will he requested that all the source code be deleted
[00:28] <Darkside> so uiview32 hasn't changed for quite a while
[00:29] <arko> wtf
[00:29] <fsphil> why would he do that
[00:29] <arko> maybe open source licensing is different in his religon
[00:29] <fsphil> someone should write a uiview33
[00:30] <arko> or throw it into IDA and reverse it
[00:30] <arko> ^^^ hax
[00:30] <fsphil> man I've not reverse engineered anything for years
[00:30] <Darkside> the other option is make xastir usable
[00:30] <Darkside> at the moment it looks like something from 1992
[00:30] <fsphil> or aprsmap
[00:30] <Darkside> i mean, it uses motif ffs
[00:30] <fsphil> it's much nicer than xastir
[00:31] <arko> i respect folks who dont want to release source, but when you die and one one else is around, why not pass it down
[00:31] <arko> he didnt have to kill all the source code
[00:31] <craag> APRSISCE/32 is pretty good for windows btw.
[00:31] <arko> maybe pass it down to another dev
[00:31] <Darkside> it seems to be a thing in the amateur radio community
[00:31] <Darkside> well, amongst the older members
[00:31] <arko> cranky old farts
[00:31] <Darkside> you don't find a lot of open source amateur radio software
[00:31] <fsphil> it's getting better. thinks like codec2 and fldigi
[00:31] <arko> all the sites are still in 1998
[00:31] <Darkside> yeah
[00:31] <fsphil> things*
[00:32] <arko> agree
[00:32] <Darkside> a lot of that is the new generation pushing through
[00:32] <arko> youngers!
[00:32] <Darkside> get off my lawn!
[00:32] Action: arko shakes his non existant kane
[00:32] <fsphil> why when I was a lad, we had to whistle the RTTY tones
[00:33] <arko> lol
[00:34] <arko> i wish we had one of the bots we have on efnet in here
[00:34] <arko> it would take quotes
[00:34] <arko> that would be a quote right there
[00:34] <fsphil> there was a stats bot for a while
[00:34] <fsphil> used to display random quotes from people
[00:35] <fsphil> not sure where that got to
[00:35] <fsphil> ah here, http://habhub.org/zeusbot/pisg.html
[00:35] <Darkside> ahh pisg
[00:35] <Darkside> wonderful
[00:35] <Darkside> i wondered if there was one of those for this channel
[00:36] johnboiles1 (Adium@bastion.sfo1.yelpcorp.com) left #highaltitude.
[00:51] signaleleven (~signalele@p5DE47457.dip0.t-ipconnect.de) left irc: Quit: Leaving
[01:27] johnboiles (~Adium@bastion.sfo1.yelpcorp.com) joined #highaltitude.
[01:32] johnboiles1 (~Adium@bastion.sfo1.yelpcorp.com) joined #highaltitude.
[01:33] johnboiles (~Adium@bastion.sfo1.yelpcorp.com) left irc: Ping timeout: 260 seconds
[01:49] mrShrimp (4c796894@gateway/web/freenode/ip.76.121.104.148) joined #highaltitude.
[01:50] <mrShrimp> Hi can someone help me set up SDR# from this tutorial: http://ukhas.org.uk/guides:sdr_tracker ?
[01:50] johnboiles1 (Adium@bastion.sfo1.yelpcorp.com) left #highaltitude.
[01:53] <KT5TK_QRL> Don't have Windows here, but what's the problem?
[01:55] <mrShrimp> When we open up SDR# and try to select a device from the dropdown menu, it does not have an option for our ezcap dongle
[01:56] <mrShrimp> I tried using the RTL-SDR/USB option
[01:56] <mrShrimp> and all of the other ones
[01:56] <mrShrimp> none of them work
[01:56] <KT5TK_QRL> Did you go through the Driver install with Zadig?
[01:56] <mrShrimp> and I have gotten them to work before
[01:56] <mrShrimp> yes
[01:56] <mrShrimp> the only thing unusual that I noticed
[01:56] <mrShrimp> was that the computer was able to install drivers of its own
[01:57] <mrShrimp> before I used zadig
[01:57] <KT5TK_QRL> Is the RTL still plugged into the same USB port?
[01:57] <mrShrimp> yes
[01:57] <mrShrimp> I tried it in both ports
[01:57] <mrShrimp> This is what Zadig says the driver is now
[01:57] <mrShrimp> RTL2832UUSB (v64.1.802.2010)
[01:57] <mrShrimp> the driver that the computer assigned to it
[01:58] <KT5TK_QRL> checked the config file for the manual changes?
[01:58] <mrShrimp> yes
[01:58] <mrShrimp> i'll check again
[02:00] <KT5TK_QRL> Is the SDR# RTL Plugin installed? http://sdrsharp.com/downloads/sdr-nightly-rtlsdr.zip
[02:00] <KT5TK_QRL> Added 07/09/12 Kevin Walton. As of the 07/09/12 the latest Dev versions of SDRSharp now also need the following: Download the SDR# RTL Plugin and copy the contents contents of the zip file to the SDR# directory. When I copied the contents of the zip file into the SDR directory, it asked me if I wanted to overwrite several files - I said yes to overwrite all files and SDRSharp works.
[02:00] <mrShrimp> I did that
[02:00] <mrShrimp> but I can recheck it
[02:01] <mrShrimp> to see if something went wrong
[02:01] <mrShrimp> the config file seems to be accordingly changed
[02:02] <KT5TK_QRL> OK, well then I'm out of ideas. I use gqrx with my RTL
[02:03] <mrShrimp> does it only work on a 32 bit pc?
[02:04] <KT5TK_QRL> I used Windows last time when it was called Windows2000 ;)
[02:04] <mrShrimp> this is a 64
[02:04] <mrShrimp> Oh
[02:04] <KT5TK_QRL> maybe
[02:04] <mrShrimp> we oh
[02:04] <mrShrimp> hold on
[02:16] mrShrimp (4c796894@gateway/web/freenode/ip.76.121.104.148) left irc: Quit: Page closed
[02:26] pjm (~pjm@uhfsatcom.plus.com) joined #highaltitude.
[02:26] pjm__ (~pjm@uhfsatcom.plus.com) left irc: Read error: Connection reset by peer
[02:29] Dan-K2VOL (~Dan-K2VOL@74-134-73-252.dhcp.insightbb.com) joined #highaltitude.
[02:46] Lunar_Lander (~gd-compto@p548833B0.dip.t-dialin.net) left irc: Quit: Verlassend
[03:04] heathkid (~heathkid@unaffiliated/heathkid) left irc: Disconnected by services
[03:04] heathkid|2 (~heathkid@unaffiliated/heathkid) joined #highaltitude.
[03:09] <heathkid|2> make fun of it... I'm still trying to find something that works reliably.
[03:09] Nick change: heathkid|2 -> heathkid
[03:09] <heathkid> okay?
[03:11] cuddykid (~acudworth@cpc2-basf8-2-0-cust23.12-3.cable.virginmedia.com) joined #highaltitude.
[03:11] cuddykid (~acudworth@cpc2-basf8-2-0-cust23.12-3.cable.virginmedia.com) left irc: Client Quit
[03:11] cuddykid (~acudworth@cpc2-basf8-2-0-cust23.12-3.cable.virginmedia.com) joined #highaltitude.
[03:21] cuddykid (~acudworth@cpc2-basf8-2-0-cust23.12-3.cable.virginmedia.com) left irc: Quit: cuddykid
[03:23] BrainDamage (~BrainDama@dynamic-adsl-78-12-250-165.clienti.tiscali.it) left irc: Quit: Leaving.
[04:20] Dan-K2VOL1 (~Dan-K2VOL@43.sub-70-194-132.myvzw.com) joined #highaltitude.
[04:22] Dan-K2VOL (~Dan-K2VOL@74-134-73-252.dhcp.insightbb.com) left irc: Ping timeout: 252 seconds
[04:25] SamSilver (2985f4d6@gateway/web/freenode/ip.41.133.244.214) joined #highaltitude.
[04:31] pjm (~pjm@uhfsatcom.plus.com) left irc: Read error: Connection reset by peer
[04:31] pjm (~pjm@uhfsatcom.plus.com) joined #highaltitude.
[05:06] Dan-K2VOL1 (~Dan-K2VOL@43.sub-70-194-132.myvzw.com) left irc: Quit: Leaving.
[06:17] radim_OM2AMR (~radimmuti@2a02:dd8:8aef:46:6454:e85:756:5aaf) joined #highaltitude.
[06:46] <x-f> morning
[06:46] <x-f> with fresh snow 10 cm and counting
[06:50] <SpeedEvil> I would be happy with 3cm of snow on Christmas, and none the rest of the year
[06:53] <x-f> two weeks still, anything can happen
[06:55] <x-f> when there is no snow yet in the autumn, i want it to fall as late as possible, because then i can do bike riding, work in the garden and so on, everything that snow makes impossible
[06:56] <x-f> but when the snow has arrived, i don't want it to melt, because then i can do country skiing, alpine skiing, snow kiting..
[06:58] <x-f> i guess, i just don't like changes
[07:03] vladimirek (~vladimire@95.105.250.72) joined #highaltitude.
[07:07] radim_OM2AMR (~radimmuti@2a02:dd8:8aef:46:6454:e85:756:5aaf) left irc: Quit: radim_OM2AMR
[07:18] nosebleedkt (~kostas@213.140.128.74) joined #highaltitude.
[07:18] <nosebleedkt> Allo !
[07:19] radim_OM2AMR (~radimmuti@2a02:dd8:8aef:46:5117:bdae:dd7:ed7f) joined #highaltitude.
[07:20] <SamSilver> good moaning nosebleedkt
[07:20] <nosebleedkt> moan
[07:20] <nosebleedkt> lol
[07:20] nosebleedkt (~kostas@213.140.128.74) left irc: Client Quit
[08:29] rmp (~rmp@host-2-96-116-62.as13285.net) joined #highaltitude.
[08:33] lz1dev (~lz1dev@46.47.80.192) joined #highaltitude.
[08:35] gartt (~gart@ip68-0-205-248.ri.ri.cox.net) left irc: Ping timeout: 244 seconds
[08:36] gartt (~gart@ip68-0-205-248.ri.ri.cox.net) joined #highaltitude.
[08:43] Laurenceb_ (~Laurence@host86-143-4-133.range86-143.btcentralplus.com) joined #highaltitude.
[08:45] <griffonbot> Received email: James Coxon "[UKHAS] Pico launches sat 15th"
[08:50] Armand (~martin@cpc17-haye16-2-0-cust427.haye.cable.virginmedia.com) joined #highaltitude.
[08:51] G8KNN-Jon (~Jon@cpc1-cmbg10-0-0-cust144.5-4.cable.virginmedia.com) left irc: Ping timeout: 252 seconds
[08:52] <griffonbot> Received email: Anthony Stirk "Re: [UKHAS] Pico launches sat 15th"
[08:59] Penfold (~mike@194.75.60.69) joined #highaltitude.
[09:01] rmp (~rmp@host-2-96-116-62.as13285.net) left irc: Quit: rmp
[09:02] Laurenceb_ (~Laurence@host86-143-4-133.range86-143.btcentralplus.com) left irc: Ping timeout: 240 seconds
[09:09] daveake (~androirc@daveake.plus.com) left irc: Quit: AndroIRC - Android IRC Client ( http://www.androirc.com )
[09:16] <griffonbot> Received email: John Underwood "Re: [UKHAS] Pico launches sat 15th"
[09:16] daveake (~daveake@daveake.plus.com) joined #highaltitude.
[09:19] G8KNN-Jon (~Jon@cpc1-cmbg10-0-0-cust144.5-4.cable.virginmedia.com) joined #highaltitude.
[09:20] jonquark (jonl@nat/ibm/x-lnriogvuopaxfqnw) joined #highaltitude.
[09:44] hyte (5e000397@gateway/web/freenode/ip.94.0.3.151) left irc: Ping timeout: 245 seconds
[09:47] nosebleedkt (~kostas@213.140.128.74) joined #highaltitude.
[09:47] nosebleedkt (~kostas@213.140.128.74) left irc: Client Quit
[09:47] nosebleedkt (~kostas@213.140.128.74) joined #highaltitude.
[09:48] rmp (~rmp@74.163.17.46.bridgep.com) joined #highaltitude.
[10:08] Nabobalis (~Nabster@cpc11-shef11-2-0-cust26.barn.cable.virginmedia.com) left irc: Ping timeout: 240 seconds
[10:08] Cadair (~Cadair@unaffiliated/cadair) left irc: Ping timeout: 252 seconds
[10:13] Cadair (~Cadair@unaffiliated/cadair) joined #highaltitude.
[10:13] Nabobalis (~Nabster@cpc11-shef11-2-0-cust26.barn.cable.virginmedia.com) joined #highaltitude.
[10:14] <eroomde> morning all
[10:14] <daveake> morning one
[10:14] <zyp> morning none
[10:14] miscat (59f909e2@gateway/web/freenode/ip.89.249.9.226) joined #highaltitude.
[10:14] <Armand> Morning .o/
[10:15] <daveake> Morning 
[10:15] <eroomde> morning {}
[10:16] <daveake> It is indeed a bracing morning
[10:16] <Armand> Sorry, daveake wins. :P
[10:17] miscat (59f909e2@gateway/web/freenode/ip.89.249.9.226) left irc: Client Quit
[10:18] <fsphil> /* I refuse to comment */
[10:18] <Armand> lol
[10:19] <Armand> Commenting on your refusal is a comment in itself... so, fail to not fail there. :P
[10:23] G8KNN-Jon (~Jon@cpc1-cmbg10-0-0-cust144.5-4.cable.virginmedia.com) left irc: Ping timeout: 252 seconds
[10:28] dharnke (4e92e827@gateway/web/freenode/ip.78.146.232.39) joined #highaltitude.
[10:40] danf747 (3ab25693@gateway/web/freenode/ip.58.178.86.147) joined #highaltitude.
[10:40] <danf747> Hello everyone
[10:41] chris_99 (~chris_99@unaffiliated/chris-99/x-3062929) joined #highaltitude.
[10:42] <Armand> Morning, danf747
[10:42] <danf747> morning.
[10:42] <danf747> I just wanted to ask a few questions about HAB. Im completely new to this
[10:43] <Armand> I'll have to leave that to others..
[10:45] <danf747> can any one help me choose my balloon hardware?
[10:47] <UpuWork> ask away danf747
[10:50] <danf747> i was looking into which gps module to use and what transmitter to pick (i found that radiometrix NTX2 was good except for a frequency shift?)
[10:51] <daveake> NTX2 is good. Gues syou mean the frequency drift with temperature. That's not a problem with good insulation
[10:51] <UpuWork> Well all transmitters shift just insulate it well
[10:51] <UpuWork> GPS wise most people use either the older Trimble Lassens or the new Ublox MAX6 modules
[10:52] <UpuWork> what microcontroller are you using ?
[10:52] <daveake> For GPS get a UBLox here - http://ava.upuaut.net/store/index.php?route=product/category&path=59_60
[10:52] G8KNN-Jon (~Jon@cpc1-cmbg10-0-0-cust144.5-4.cable.virginmedia.com) joined #highaltitude.
[10:52] <UpuWork> yeah good shop highly recommended
[10:53] <danf747> arduino most probably. anything that would be significantly better to use?
[10:53] <UpuWork> Well Arduino is fine
[10:53] <UpuWork> most of them run at 5V though
[10:53] <UpuWork> which means you need to ensure the GPS module you get can handle this
[10:53] <UpuWork> http://ava.upuaut.net/store/index.php?route=product/product&path=59_60&product_id=68
[10:54] <UpuWork> Have a read of www.ukhas.org.uk/guides:gps_modules
[10:54] <danf747> That seems like the best as it has the back up battery
[10:54] <UpuWork> you won't need the battery really
[10:54] <UpuWork> it gets a lock in 30 sec from cold under an open sky anyway
[10:55] <UpuWork> are you in the UK ?
[10:55] <danf747> australia
[10:55] <UpuWork> Ah ok
[10:55] <UpuWork> might want to consider APRS if you have a radio license
[10:56] <Darkside> australia eh?
[10:56] <Darkside> sydney
[10:57] signaleleven (81f71fe0@gateway/web/freenode/ip.129.247.31.224) joined #highaltitude.
[10:57] signaleleven (81f71fe0@gateway/web/freenode/ip.129.247.31.224) left #highaltitude.
[10:57] <danf747> yeah
[10:57] <Darkside> tip: don't launch from sydney
[10:57] <fsphil> yea not the best launch site :)
[10:58] <Darkside> best to go far, far inland
[10:58] <fsphil> or landing site
[10:58] <Darkside> like, griffith way
[10:58] <Darkside> or parkes area
[10:58] <danf747> yeah yeah... i was planning on dirivng like 7 hrs before launch
[10:58] <Darkside> basically, get at least 100km the other side of the great dividing range, or you are very likely goign to have a splash landing
[10:58] <Darkside> started talking to CASA yet?
[11:00] <danf747> yeah ive been emailing... very slow process. im not planning a launch till january though so no rush
[11:00] SamSilver (2985f4d6@gateway/web/freenode/ip.41.133.244.214) left irc: Quit: Page closed
[11:00] <Darkside> hah
[11:00] <Darkside> good luck
[11:00] <Darkside> it'll take a few months
[11:00] <Darkside> and yeah, slow process is right
[11:01] <danf747> :( i dont mind. a lot of time for testing then. What about a portable radio system for the car
[11:01] <fsphil> are they doing something that takes time, or is it like here where they just leave it to the last minute possible?
[11:01] <Darkside> anyway, if you have a standard or advanced amateur radio licence, APRS can work pretty well, though out that side of the dividing range you will have a distinct lack of aprs igates
[11:01] <dharnke> hey upu, i still dont get the freq shift, i checked the resistors using a multimeter and it seems ok
[11:02] <UpuWork> something not rigth dharnke
[11:02] <danf747> i can get hold of a icom ic 9100 however they are really big
[11:02] <Darkside> danf747: yeah you want somethin g abit smaller lol
[11:02] <UpuWork> until you get that voltage difference between the high and the low tones you're not getting RTTY
[11:02] <dharnke> could it be the component
[11:03] <daveake> Doubtful
[11:03] <UpuWork> well just check the actual voltages
[11:03] <Darkside> danf747: we use IC-706's in our chase cars, but you don't need anything that expensive
[11:03] <UpuWork> should differ by about 0.2v
[11:04] <dharnke> i got 2.58v
[11:04] <dharnke> across the 4k7
[11:04] <danf747> suggest anything simpler?
[11:05] <UpuWork> dharnke the issue has to be with the resistor network
[11:05] <UpuWork> you have a carrier
[11:05] <UpuWork> and you can see the wobble of the carrier in time with the changes
[11:05] <Darkside> danf747: a rtlsdr dongle can work well
[11:05] <Darkside> and is a pretty cheap option
[11:05] <UpuWork> but the difference between the tones should me much larger which means the voltage isn't changing enough
[11:05] <UpuWork> which means ... dodgy wiring
[11:06] <dharnke> ok, ill try and rewire from fresh
[11:06] lz1dev (~lz1dev@46.47.80.192) left irc: Ping timeout: 255 seconds
[11:07] <dharnke> also i had another question re: ublox
[11:07] <eroomde> i wouldn't wire from fresh as that's a slightly stochastic approach and you might not learn anything. much better to properly look at what you have done now and work out why it is not working as you intend
[11:08] <dharnke> is there any app to get the data on a map
[11:08] <danf747> That dongle would really work?
[11:08] <dharnke> e.g. google map
[11:08] <Darkside> danf747: yeah, its sensitive enough
[11:08] <Darkside> obviously not with the stock antenna
[11:08] <Darkside> but with something else, they can work quite well
[11:08] <danf747> can you attach a simple yagi antenna to something like that?
[11:09] <Darkside> yeah, you can
[11:09] <UpuWork> dharnke yes
[11:09] <Darkside> but you don't need a yago if you'r efollowingthe balloon
[11:09] <Darkside> ack
[11:09] <Darkside> danf747: a simple 1/4 wave antenna works well for tracking while moving
[11:09] <Darkside> once the payload is in the air, it's pretty easy to receive
[11:09] <danf747> a lot simpler and pointing then
[11:09] <Darkside> yup
[11:10] <Darkside> though if it lands and you're not close, a yagi can be very useful
[11:10] <danf747> could you suggest any uhf cb radios?
[11:10] <Darkside> no
[11:10] <Darkside> UHF CB is like mos eisely
[11:10] <Darkside> eisley*
[11:10] <eroomde> sandy?
[11:10] <Darkside> a wretched hive of scum and depravity
[11:11] <eroomde> (a wretched hive.... i know)
[11:11] <Darkside> villiany*
[11:11] <Darkside> whatever
[11:11] <danf747> ahahah explain?
[11:11] <Darkside> danf747: i'm an amateur radio operator, i'm terribly biased
[11:12] <Darkside> i used to use CB, then i got my amateur licence and stopped using it
[11:12] <eroomde> you 'ascended', to use the parlance
[11:12] <Darkside> to a higher plane of existence
[11:12] <Darkside> yes
[11:14] <danf747> would something like a TH-9000U be useful
[11:14] <Darkside> (googling)
[11:14] <Darkside> oh lol
[11:14] <Darkside> not for receiving the balloon
[11:14] <Darkside> those kind of radios make nice cheap amateur radio rigs though
[11:15] <Darkside> and i guess they can do UHF CB too, but they are in breach of the UHF CB class licence
[11:15] <dharnke> upu, whats the ideal antenna for tx with ntx2,
[11:15] <Darkside> danf747: i'm from Adelaide by the way, one of the Project Horus team
[11:16] <danf747> Ive seen your websites. amazing stuff
[11:16] <Darkside> we've been doing it for a while, it's good fun
[11:16] <Darkside> expensive hobby tho
[11:17] <danf747> i can imagine. and time consuming
[11:17] <Darkside> yup
[11:17] <Darkside> getting out launch systems to the point it is at now took quite a while
[11:17] <eroomde> there are more expensive hobbies
[11:17] <eroomde> try rockets
[11:17] <Darkside> unfortunately here in australia, we can't always rely on having 3g access wherever we go
[11:18] <Darkside> so the spacenear.us tracker, while still awesome, isn't always usable
[11:18] <eroomde> our n2o bill is quite scary now
[11:18] <Darkside> so we spent quite a while building an offline tracking system
[11:18] <danf747> yeah... telstra is the best option it works most places ive been to
[11:18] <danf747> around parkes and further
[11:18] <Darkside> yeah, nextg is pretty damn good
[11:18] <Darkside> with an external high gain antenna, we usually have some kind of coverage
[11:18] <UpuWork> dharnke just a 1/4 wave one http://ava.upuaut.net/files/2010-09-03%2019-26-16_0003.jpg
[11:18] <Darkside> but we don't rely on it
[11:19] <Darkside> we have had quite a few chases where there was no coverage at the landing site
[11:19] <Darkside> the mildura launches being one of the most notables ones for that
[11:19] <danf747> offline system would be amazing...
[11:19] <eroomde> UpuWork: sorry that's given me an aneurism
[11:19] <eroomde> how can the vertical height of the ground bit be 164mm
[11:19] <Darkside> danf747: go code your own, you're not getting outs :P
[11:19] <Darkside> ours*
[11:20] <eroomde> but also the non vertical radial that seems to span the same vertical distance also be 164mm?
[11:20] <eroomde> oh i see, it's probably just not constrained on the bottom
[11:20] <UpuWork> it was my original sketching, I need to redo that and put it on the wiki
[11:20] <UpuWork> but basically 164mm wires
[11:20] <danf747> ahahah. nah ill just copy and paste into google earth
[11:20] <Darkside> oziexplorer is good
[11:20] <UpuWork> that image isn't clear and some of it is coax
[11:20] <Darkside> if you can get hold of the topo maps for the area
[11:20] <Darkside> http://projecthorus.org/wp-content/uploads/2012/10/DSC_5702.jpg
[11:20] <eroomde> $ oziexplorer
[11:21] <eroomde> >> There's nothing here Bruce
[11:21] <Darkside> danf747: we parse the data from dl-fldigi and plot it on oziexplorer
[11:21] <eroomde> that's how i imagine a program like oziexplorer would be implemented
[11:21] <Darkside> Bruce here is in charge of the sheep dip
[11:21] <Darkside> at the university of woolomoloo
[11:21] <Darkside> i'm pretty sure there is no university in woolomoloo
[11:21] <danf747> that looks like a good program.
[11:21] <Darkside> but it is a real place
[11:21] <danf747> ahahah
[11:21] <UpuWork> use Apple Maps ?
[11:22] <Darkside> danf747: yeah, plotting points into ozi is hoe we started
[11:22] <Darkside> manually, i mean
[11:22] <eroomde> you can start your hoes on your own time Darkside
[11:22] <danf747> i think ill go for the manual way for starters
[11:22] <Darkside> thanks eroomde
[11:22] <Darkside> danf747: but yeah, ozi is a good mapping program to use
[11:22] <Darkside> runs in wine too!
[11:22] <dharnke> i assumed the 1/4 wave, but doesnt it lose alot of its power tx-ing everywhere, what if you use a circularly polarised antenna?
[11:23] <danf747> wine for mac?
[11:23] <Darkside> danf747: yeah
[11:23] <eroomde> you could use a cp antenna, but it's just a bit more faff for what you need to receive
[11:23] <Darkside> dharnke: its quite complex to do a circ polarised antenna on a payload
[11:23] <Darkside> the phasing harness is a pain
[11:23] <eroomde> as long as the ground antenna is also vertical, you should be fine
[11:23] <Darkside> they are usefull on the ground though
[11:23] <Darkside> especially if the payload is directly above you
[11:23] <eroomde> it's nice to have random hams be able to pick it up well
[11:23] <dharnke> ok
[11:24] <eroomde> rather than just amsat boys
[11:24] <Darkside> we use them for chase car antennas
[11:24] <Darkside> but htey are a bit of a pain to get right
[11:24] <danf747> ok so ill have http://ava.upuaut.net/ a ntx2. i just need a radio to receive with now
[11:24] <danf747> the gps from ava.net
[11:25] <Darkside> danf747: Upu also sells the rtlsdr dongles :P
[11:25] cuddykid (~acudworth@cpc2-basf8-2-0-cust23.12-3.cable.virginmedia.com) joined #highaltitude.
[11:25] <UpuWork> btw www.habsupplies.com also works now
[11:25] <Darkside> and the HABAmps, which are a good frontend and amplifier for them
[11:25] <dharnke> it would be efficient, plus if you use an omnidirectional antenna upon receiving the signal
[11:26] <Darkside> dharnke: a circ polarised antenna on the balloon is just more annoying
[11:26] <Darkside> well, annoying to make and fit to a payload
[11:26] <dharnke> i see
[11:26] <Darkside> easier to use a 1/4 wave on the payload, and do the fancy stuff on the ground
[11:26] <danf747> is it simple interfacing with the dongle?
[11:26] <Darkside> the only reason we use a circ polarised antenna for receive is it has a hemispherical radiation pattern
[11:26] <Darkside> danf747: theres stuff on the ukhas wiki about using it
[11:26] <Darkside> though thats for windows
[11:27] <Darkside> not much harder using it on linux or mac tho
[11:27] <dharnke> if you use a 1/4 antenna, is it easier to receive it using a yagi or an omni-directional antenna
[11:27] <eroomde> heh, i'm using a 208 pin lqfp chip
[11:27] <eroomde> it's beastingly big
[11:27] <Darkside> dharnke: well, when moving, a yagi is difficuly
[11:27] <eroomde> like 3cm to each side
[11:27] <Darkside> so you use an omni :P
[11:27] <eroomde> didn;t really realise until i opened up the layout editor
[11:27] <Darkside> but if stationary, sure, pul out the yagi for extra gain
[11:27] <Darkside> danf747: if you want simpicity, then get a separate radio for RX
[11:28] <Darkside> it'll use less CPU power too
[11:28] <Darkside> depends how much you want to pay :-)
[11:28] <mattbrejza> eroomde: DSP IC?
[11:29] <mattbrejza> FPGA?
[11:30] <danf747> yeah i think ill probably get a separate radio. although would using a cb radio be sufficient... They seem to be pretty cheap to get
[11:30] <eroomde> mattbrejza: fpga
[11:30] <mattbrejza> well at least it has legs
[11:30] <eroomde> the price of not wanting to go near BGA packages
[11:30] <Darkside> danf747: no
[11:31] <Darkside> danf747: you need a SSB receiver
[11:31] <Darkside> none of the cheap radios do that, they are only FM receivers
[11:32] dharnke (4e92e827@gateway/web/freenode/ip.78.146.232.39) left irc: Quit: Page closed
[11:32] <zyp> eroomde, coarse-pitch bga shouldn't be too bad
[11:33] <eroomde> zyp: maybe, but still harder
[11:33] <zyp> and with a FPGA you are pretty free to pick the pins that are easiest to route out
[11:34] <eroomde> sure. but still harder
[11:34] <eroomde> i'm not space constrained for this one
[11:34] <eroomde> so see no reason to make things harder
[11:34] <zyp> fair enough
[11:34] <eroomde> there is an evil masterplan to use some fpgas which are 5mm x 5mm, 0.4mm bga
[11:35] <zyp> :)
[11:35] <eroomde> which will definitely need 4 layer boards with 0.1mm drill holes and 4mil fanout traces
[11:35] <eroomde> but that can wait :)
[11:35] <eroomde> actel igloo nano for interest
[11:35] <zyp> I just finished a design with an 1mm pitch LBGA256
[11:36] <eroomde> 4 layers for the fanout?
[11:36] <zyp> I'm looking forward to see how hard it will be to solder it properly
[11:36] <zyp> fanout only on outer layers, inner layers are strictly gnd and vcc planes
[11:36] <eroomde> that's the other thing i don;t really want to risk
[11:37] <eroomde> although actually we haven't really had problems on the couple of bga-ish stuff we've done before
[11:37] <eroomde> but it's still a poco risky
[11:37] <zyp> I'm doing this board strictly for the challenge of it, to gain experience with BGA
[11:37] <mattbrejza> eroomde: just deadbug instead: http://www.chiaki.cc/Pyxis2010/images/pyxis2010-fpgasol2.jpg
[11:38] <zyp> http://bin.jvnv.net/f/eaiEM.png http://bin.jvnv.net/f/Uby5B.png
[11:38] <mattbrejza> source: http://www.chiaki.cc/Pyxis2010/index.htm
[11:38] <eroomde> jesus wept
[11:39] <mattbrejza> thats a $600 fpga
[11:39] <mattbrejza> wouldnt want to solder that like that
[11:39] <eroomde> zyp: what size are the vias for that pitch?
[11:39] <eroomde> can you get away with 0.3mm drills?
[11:39] <zyp> 10 mil drills
[11:39] <eroomde> ok
[11:39] <Darkside> what is that in real units
[11:39] <eroomde> so about 0.25mm
[11:40] <zyp> yeah
[11:40] <Darkside> ah
[11:40] <eroomde> 0.3mm is the lower limit for most of the chinese fab places, i've noticed
[11:40] <eroomde> without going special order
[11:40] <Darkside> i thought mitch could go smaller?
[11:40] <eroomde> 0.3mm on the standard spec stuff
[11:40] <Darkside> or does it cost more
[11:40] <eroomde> might be able to go smaller as a special, but 0.3mm is the lowest in his guide to pcb
[11:40] <Darkside> ahh ok
[11:40] <eroomde> he can do 4mil traces and spacing as a special certainly
[11:40] <zyp> the guy that's doing the pcb for me said 10mil were standard spec at the place he's using, so I designed to that
[11:41] <eroomde> which i was looking at for a 0.4mm pitch bga
[11:41] <Darkside> hmm it says pluged via diameter of 0.25mm
[11:41] <Darkside> plugged*
[11:41] <eroomde> oh
[11:41] <eroomde> interesting
[11:41] G8KNN-Jon (~Jon@cpc1-cmbg10-0-0-cust144.5-4.cable.virginmedia.com) left irc: Ping timeout: 252 seconds
[11:42] <Darkside> i haven't had a need to go smaller than 0.4mm/0.6mmm so far
[11:46] <zyp> I tend to always go for the smallest vias according to spec, and rather use several for high current stuff
[11:46] <eroomde> are you another lucky altiumer?
[11:46] <UpuWork> I use 0.3mm alot
[11:47] <Darkside> eroomde: that doesnt look like an altium render to me
[11:47] <Darkside> but i could be wrong
[11:47] <Darkside> oh wait, the text looks a bit like altiums stroke text
[11:49] <Darkside> zyp: is it altium?
[11:49] <Darkside> or somethign else
[11:52] <zyp> no, it's eagle
[11:54] <eroomde> gerv rendering?
[11:54] <eroomde> gerbv*
[11:54] <zyp> yes
[11:54] <eroomde> ah
[11:57] <zyp> http://bin.jvnv.net/f/G5Pt4.png <- looks a bit better than eagle's own :p
[11:58] BrainDamage (~BrainDama@dynamic-adsl-78-12-250-165.clienti.tiscali.it) joined #highaltitude.
[12:09] G8KNN-Jon (~Jon@cpc1-cmbg10-0-0-cust144.5-4.cable.virginmedia.com) joined #highaltitude.
[12:13] lz1dev (~lz1dev@46.47.80.192) joined #highaltitude.
[12:16] <fsphil> awww, I missed it again: http://www.bbc.co.uk/news/uk-northern-ireland-17817085
[12:16] <gonzo_> any experts on 36" foil balloons in the chan??
[12:19] danf747 (3ab25693@gateway/web/freenode/ip.58.178.86.147) left irc: Ping timeout: 245 seconds
[12:21] <craag> gonzo_: I'm not an expert, but what's the question?
[12:21] <craag> Are you still planning to launch this wkend?
[12:29] Nick change: Armand -> Armand|AFK
[12:30] Grumbleist (~Grumbleis@cpc8-camd13-2-0-cust233.hari.cable.virginmedia.com) joined #highaltitude.
[12:40] <gonzo_> question is on the fill and lift
[12:41] <gonzo_> yes, planning a non tracket launch. Basically a wireless cctv cam only
[12:42] <gonzo_> not my favourite option, but the rest of the guys want to see pictures and are not that interested in the engineering. So flying a cam just to see what pictures we get.
[12:43] <gonzo_> but trying to calculate the ascent rate based on the overall free lift (if that is the correct term?) the lift with balloons and payliad attached
[12:44] <daveake> Some flight data including neck lift (for recent flights) here - http://ukhas.org.uk/projects:microballoons:data
[12:45] <griffonbot> Received email: Philip Crump "Re: [UKHAS] Pico launches sat 15th"
[12:45] <gonzo_> thanks dave. Looked for data on the site for about half an hour
[12:45] <gonzo_> it's well hidden then
[12:46] <daveake> Yeah took me ages to find it to update it :)
[12:46] <daveake> Googled for "ukhas foil lift" which worked :)
[12:46] <craag> Extrapolated from my own 2 data points, I've found that ascent rate (m/s) ~= free lift(g)/10.
[12:46] <craag> But I haven't checked anyone else's data against that.
[12:47] <craag> (For a single foil balloon)
[12:49] <gonzo_> thanks that's a useable rule of thumb
[12:49] <gonzo_> what's the definition of free lift?
[12:49] <craag> necklift - payload weight
[12:49] <gonzo_> Overall lift with all balloon+payload mass?
[12:49] <craag> Yeah.
[12:49] <gonzo_> rr, ok
[12:50] <craag> How heavy is your wireless camera?
[12:50] <daveake> What I do is tie it to a mug and put the mug on scales. tare the scales, then undo the balloon. Weight shown is the lift.
[12:51] <gonzo_> I did the same to measure soem small 'balloons' a while ago
[12:52] <gonzo_> the one I have to fly is just under 60g, and that is with a 9v batt (with the casing removed)
[12:52] <daveake> 1 92cm foil balloon?
[12:53] <gonzo_> well, fully inflated we could get away with 1. As we do not care if it bursts early this time
[12:53] <gonzo_> and a carrier bag chute, if it's a single balloon
[12:54] <daveake> OK, yeah 60g is about your limit I think
[12:54] <gonzo_> but yep, a 36" foil
[12:55] <craag> Just remember if you're inflating at home, *don't* put the heating on the car!
[12:55] <craag> *on in
[12:55] <gonzo_> that was my thoughts. so it's a toss up between 1 or 2
[12:55] <gonzo_> good point!
[12:56] <gonzo_> (and we are using h2 so don'tb want a burst in the car
[12:56] <gonzo_> though that is prob not an ideal way of doing it anyway!)
[12:59] G8KNN-Jon (~Jon@cpc1-cmbg10-0-0-cust144.5-4.cable.virginmedia.com) left irc: Ping timeout: 252 seconds
[13:11] Nick change: Armand|AFK -> Armand
[13:13] SamSilver (2985f4d6@gateway/web/freenode/ip.41.133.244.214) joined #highaltitude.
[13:20] <craag> Using 2 with hydrogen doesn't sounds like a great idea, as one could come down still inflated.
[13:21] signaleleven (81f71fe0@gateway/web/freenode/ip.129.247.31.224) joined #highaltitude.
[13:21] <craag> There's still a risk of that even with 1 I think. Especially as you have no tracking so I assume you aren't planning to retrieve.
[13:23] <fsphil> any gsm modules small enough to fit on a pico?
[13:25] <UpuWork> not a pico
[13:25] <UpuWork> well
[13:25] <UpuWork> it would be border line
[13:25] <UpuWork> ping KF7FER
[13:25] <fsphil> the one I have is a bit heavy
[13:25] <fsphil> but those usb dongles are nice and light
[13:25] <fsphil> the flaw being the usb part
[13:25] <daveake> indeed
[13:26] G8KNN-Jon (~Jon@cpc1-cmbg10-0-0-cust144.5-4.cable.virginmedia.com) joined #highaltitude.
[13:27] <fsphil> might open the dodgey one I have, see if there's a uart exposed anywhere
[13:27] <craag> I saw an article about hooking a vodafone usb dongle up to a small mbed board. Add a GPS and you'd be good to go.
[13:27] <fsphil> http://hackaday.com/2012/11/02/vodafone-usb-3g-modem-driver-from-mbed/
[13:28] <craag> Thats the one.
[13:29] <fsphil> I believe they appear to the host as a serial device
[13:29] <fsphil> or a pair of serial devices
[13:30] <fsphil> I remember getting /dev/ttyUSB0 and ttyUSB1 appearing when I was using it on the laptop
[13:30] <daveake> Yep. Some also appear as a drive, for software installation, but the modem size is serial
[13:30] <daveake> side*
[13:30] <fsphil> if it's a uart bridge then there might be hope
[13:32] <daveake> This one does http://www.circuitsathome.com/mcu/interfacing-arduino-to-usb-gprs-modem
[13:33] <daveake> Is there anything useful that they don't sell? :)
[13:40] <Randomskk> hey eroomde, I actually have a changepoint thing running as a decoder now
[13:40] <Randomskk> it's _horrifically_ slow
[13:41] <Randomskk> with synthetic 5kHz sample rate and 20 baud data, it's just over a second per bit
[13:41] <Randomskk> a lot of which is the peak finding code
[13:41] <Randomskk> I mean this is totally unoptimised and can probably be sped up a ton, especially the peak finding
[13:41] <Randomskk> and because it's so slow, generating decent numbers for performance in noise is a pain, because I'm only running like 100 bits per SNR point
[13:43] <eroomde> Randomskk: really?
[13:43] <eroomde> how implemented?
[13:44] <Randomskk> peak detection is using wavelets
[13:44] <eroomde> are you using a forgetting window?
[13:44] <Randomskk> no
[13:44] <eroomde> ah, i think i just did thresholding and assumed it was noisy
[13:44] <Randomskk> how did you work out what to threshold against?
[13:44] <Randomskk> or for that matter work with a noisy assumption?
[13:45] <Randomskk> (also I'm still running two changepoint detections, one for mark-to-space and one for space-to-mark)
[13:45] <Randomskk> the peak detection is basically just scipy's peak finder
[13:45] <eroomde> i did it a bit by eye :)
[13:45] <Randomskk> oh, right
[13:45] <Randomskk> yea if I just spit out graphs of the posteriors it's easy to do by hand
[13:45] <eroomde> to pick a threshold value that is
[13:45] <eroomde> but you could just do average statistics
[13:45] <Randomskk> hmm
[13:45] <Randomskk> need to smooth too
[13:46] <eroomde> because the peaks are so passive in comparison to the noise, usually
[13:46] <Randomskk> I don't know if that continues to be the case when you get into heavier noise
[13:46] <eroomde> also i did try laplaces method for fitting a guassian to the noisy gaussian and taking rhe mean of that
[13:46] <eroomde> but it was sort of not adding much to the party
[13:46] <Randomskk> but it definitely feels like there should be a much better way to go from changepoint posteriors to a recovered bitstream
[13:47] <eroomde> well, this is why i wanted to change it to tracking frequency in the fourier space
[13:47] <eroomde> cp to find jumps and bayesian fitting to estimate the value. which sort of fell out of the maths
[13:48] <Randomskk> sounds like it'd be easier to marginalise over the actual frequencies then too, if you wanted
[13:48] <Randomskk> but hm
[13:48] <Randomskk> isn't that kinda messy?
[13:48] <Randomskk> as in, how do you recover time information?
[13:48] <Randomskk> lots of moving window FFTs?
[13:48] <Randomskk> and then find the one with the transition?
[13:48] <eroomde> let me show you a sexy graph of something i implemented
[13:49] <eroomde> http://i.imgur.com/5owJC.png
[13:49] <eroomde> in freq space
[13:49] <eroomde> this was how i was going to make it cope computationally with MFSK for arbitrary M
[13:50] <eroomde> this is as far as i got before employment at The Other Place
[13:50] <Randomskk> what're the lines?
[13:50] <eroomde> which lines?
[13:51] <eroomde> also it occurs to me now that a particle filter would work really well for the tracking...
[13:51] <eroomde> hmmm
[13:51] <eroomde> this has given me an idea
[13:51] <Randomskk> well, what are the traces
[13:51] <eroomde> on that graph, grey is peak detect on a spectrogram
[13:51] <fsphil> daveake: interesting IC that, the MAX3421E. Provides an SPI interface to the USB bus
[13:51] <eroomde> dots are varience in one of my tracking filters
[13:51] <fsphil> should work for cameras too
[13:51] <eroomde> black line is estimate of the state
[13:52] <Randomskk> got it
[13:52] <daveake> fsphil I see the twinkle of a plan in your eyes
[13:52] <eroomde> this was implemented with a lot of reference to mackay
[13:52] <Randomskk> and then what, you can just take the DC level as a threshold and see if it's above/below that?
[13:52] <Randomskk> or something a little bit smarter to account for the data not being happy
[13:53] <Randomskk> it seems like you suddenly have clock recovery issues all over again?
[13:53] <eroomde> if you plot posterior probability against time it sort of makes sense too
[13:53] <eroomde> it looks like this:
[13:53] <eroomde> ////
[13:53] <eroomde> i.e. a bunch of posterior probs rising with time
[13:53] <eroomde> then dropping on the changepoints
[13:54] <Randomskk> right, yea
[13:54] <fsphil> sawtooth wave
[13:55] <eroomde> yeah
[13:55] <eroomde> so how big is your window?
[13:55] <Randomskk> for peak detection or the changepoint detection?
[13:55] <eroomde> cp detection
[13:56] <Randomskk> at the moment it's as long as the data and doesn't move
[13:56] <eroomde> ok
[13:56] <Randomskk> i.e. it's finding a single changepoint in a stream of data. so the posterior just gets multiple peaks
[13:56] <eroomde> that can be horrible from a numerical precision pov
[13:56] <eroomde> you end up doing stuff to the power -1e5+
[13:56] <eroomde> granted log space
[13:56] <Randomskk> yea. the very first attempt I was just finding posteriors and it didn't stand a chance at all
[13:57] <Randomskk> moved to log liklihoods and it worked much better
[13:57] <eroomde> i don;t think machine learning would be possible as a field without log liklihoods :)
[13:57] <Randomskk> indeed
[13:57] <Randomskk> though for some curious reason
[13:58] <Randomskk> using slogdet in numpy/scipy/whatever was working much worse / not working at all compared to using log(det())
[13:58] <eroomde> odd
[13:58] <eroomde> but the latter is also more readable
[13:58] <eroomde> to me
[13:58] <Randomskk> yea. not hugely fussed.
[13:59] <Randomskk> ugh yes my results are still way too noisy. 100 bits doesn't seem to be enough for good BER estimation :P
[13:59] <eroomde> my online approach is not necessarily sensible of course
[13:59] <Randomskk> maybe I'll try one accurate run at a single snr
[13:59] <eroomde> because you don't really need real time telem on a hab
[14:00] <eroomde> it's more a general thing, because i was potentially doing a phd in fault tolerant control, i.e. self learning control, and wanted an efficient way to spot discrete changes in the dynamics model
[14:00] <eroomde> eg a wing being shot off
[14:02] <eroomde> i still would like to do some research in this area
[14:02] LazyLeopard (~irc-clien@chocky.demon.co.uk) joined #highaltitude.
[14:02] <mattbrejza> so what limit of SNR do you reckon your reciever will get a decentish ber?
[14:02] <Randomskk> mm. it'd still be neater to have it working online
[14:02] <eroomde> it will yep
[14:02] <Randomskk> mattbrejza: define decentish...
[14:02] <mattbrejza> 10^-3
[14:02] <eroomde> depends on your fec :)
[14:02] <mattbrejza> uncoded
[14:02] <Randomskk> hm. I'm gonna adapt this to run on a compute cluster and do a load of runs in parallel
[14:03] <Randomskk> this is way too slow to be fun on my pc
[14:03] <Randomskk> or maybe profile it...
[14:03] <Randomskk> but the department has plenty of servers sitting unused
[14:03] <eroomde> profile it def
[14:03] <mattbrejza> this is when you find itll run 10x faster if you do coloum accesses rather than row or whatever
[14:03] <eroomde> you might have done something silly like i did
[14:03] <eroomde> eg sum() instead of numpy.sum()
[14:04] <eroomde> which took the gps tracking algorithm from 30s per sec to about 1 sec per sec
[14:08] <eroomde> %prun in ipythob, as i'm sure you already know
[14:08] <Randomskk> yup
[14:12] <eroomde> you can print it to a file with some flag too, as sometimes the output is massive
[14:12] <eroomde> not founding anything particularly convincing that is graphical
[14:12] <Randomskk> I used gprof2dot before which just makes a dot graph for you
[14:12] <Randomskk> I've heard of but not used runsnakerun too
[14:14] <eroomde> http://www.vrplumber.com/programming/runsnakerun/screenshot-2.0.png
[14:14] <eroomde> i suspect an ordered list might be a bit more useful :)
[14:14] <Randomskk> haha well that's nice. 49% of the time spent doing the bayesian changepoint and 51% of the time spent doing peak detection
[14:15] <eroomde> and what are the big hogs within those two?
[14:15] <Randomskk> which when it boils down to it is 48% of the time spent doing the bottom level C code for the wavelet stuff
[14:15] <Randomskk> and the rest is a bit messier
[14:15] <Randomskk> a range of various linalg bits
[14:15] <Randomskk> oh
[14:15] <Randomskk> 30% is spent in numpy.dot
[14:15] <Randomskk> lol
[14:16] <eroomde> not sure if that uses SSE by default
[14:16] <Randomskk> http://randomskk.net/u/cp_prof.png
[14:16] <eroomde> but there is an experimental module
[14:16] <eroomde> pyexpr or something
[14:17] <eroomde> what is this tool?
[14:17] <eroomde> i want it
[14:17] <Randomskk> aforementioned gprof2dot
[14:17] <Randomskk> you save a pstats file from cProfile
[14:17] G8KNN-Jon (~Jon@cpc1-cmbg10-0-0-cust144.5-4.cable.virginmedia.com) left irc: Ping timeout: 252 seconds
[14:17] <eroomde> to gen the graph
[14:17] <Randomskk> and it turns the pstats into a dot graph
[14:17] <Randomskk> which dot can then render
[14:17] <Randomskk> dot is part of graphviz
[14:17] <Randomskk> if you've not come across graphviz before, well, you should check it out :P
[14:17] <eroomde> ah
[14:17] <eroomde> so the dot is just a class to blas
[14:18] <eroomde> unlikely there are many savings to be had there
[14:18] <Randomskk> yes it looks that way
[14:18] <Randomskk> indeed
[14:18] <Randomskk> I think windowing the data would give me one speedup
[14:18] <Randomskk> and a more efficient peak finding another
[14:18] <eroomde> maybe put the wavelet stuff on a gpu?
[14:18] <eroomde> that's n^2 operations per n reads, right
[14:18] <Randomskk> that'd be nice but probably I can just do less wavelet stuff
[14:18] <eroomde> the convolution, i mean
[14:19] <Randomskk> yea
[14:19] <Randomskk> a lot of this would be lovely on a gpu
[14:19] <Randomskk> but writing it is a faff
[14:19] <eroomde> tell me about it
[14:19] <Randomskk> I think I could probably just have less convolutions
[14:19] <eroomde> though i shall have to face it like a man at some point soon
[14:19] <eroomde> fewer
[14:19] <Randomskk> my 4th year project involves opencl stuff
[14:19] <Randomskk> fewer indeed
[14:19] <Randomskk> anyway
[14:19] <Randomskk> at the moment it does something like 30 for each posterior
[14:20] <Randomskk> just decimating the posteriors would help too
[14:20] <eroomde> i had an awesome idea at lunch that will be a computational nightmare
[14:20] Action: Laurenceb is currently failing with extended kalman filters
[14:20] <Laurenceb> i have a linear actuator that is applying pressure to an object
[14:20] <Randomskk> maybe decimate the posteriors and drop the range of wavelet sizes it tries
[14:20] <Laurenceb> i want to try to solve for object position and spring constant
[14:20] <Randomskk> but the whole wavelet thing is still a mess
[14:20] <Randomskk> a moving window would help on the posteriors side of things
[14:20] <Laurenceb> from pressure versus position
[14:21] <Laurenceb> does this sound reasonable?
[14:21] <eroomde> wavelets do seem like a bit of a heavyweight solution for basically peak detection
[14:21] <Laurenceb> i cant make it stable
[14:21] <Randomskk> well I mean it's basically http://docs.scipy.org/doc/scipy/reference/generated/scipy.signal.find_peaks_cwt.html
[14:21] <Randomskk> it's quite a reliable and noise-free solution
[14:21] <eroomde> true
[14:21] <Randomskk> but it's not exactly computationally cheap
[14:22] <eroomde> it will make nick kingsbury smile at project presentation day though
[14:22] <Randomskk> sadly he's not involved with my presentations
[14:22] <eroomde> as he basically invented discrete complex wavelets
[14:22] <Randomskk> and also this isn't my 4th year project
[14:22] <eroomde> oh
[14:22] <eroomde> this is just fo bill?
[14:22] <Randomskk> yea
[14:22] <eroomde> i know how that feels
[14:23] <Randomskk> well. with a view that it could make a fun start to a phd
[14:23] <eroomde> you want to please jesus
[14:23] <eroomde> sure, though be careful with bill
[14:23] <eroomde> talk to rachel
[14:23] <Randomskk> just using multiprocessing on python and then running it on a couple of the CUED servers gives me a 32x speedup
[14:23] <eroomde> i won;t chat publically on irc
[14:23] <Randomskk> which might be the easiest approach for now
[14:23] <Randomskk> heh
[14:23] <Randomskk> I've been chatting to someone who's doing a 4th year project under them
[14:24] <eroomde> i went through the same process you're going through now at almost exactly this time too
[14:24] <eroomde> i.e. cp detection spinning into phd with bill
[14:24] <eroomde> but
[14:24] <eroomde> well, chat to rachel
[14:25] <Randomskk> very mysterious :P
[14:25] <Laurenceb> eroomde: can i beg for assistance?
[14:25] <Laurenceb> pleasey pleasy please please
[14:25] <Laurenceb> my kalman filter blew up :(
[14:26] <Laurenceb> now im not sure if this idea is even sane...
[14:27] <Laurenceb> im try to calculate object position and reformability from applied pressure versus position of an actuator
[14:27] <Laurenceb> *de formability
[14:27] <eroomde> what is reformability please?
[14:27] <eroomde> oh
[14:27] <eroomde> ok
[14:27] <Laurenceb> lol
[14:27] <Laurenceb> ill paste my code
[14:28] <Laurenceb> http://pastebin.com/RtJEJ5CS
[14:28] <Laurenceb> pressure=State[0]*(Position-State[1])
[14:28] <Laurenceb> simple model
[14:28] <Laurenceb> but only one observable
[14:29] <Laurenceb> im currently pissing about wondering why it explodes as soon as it hits the object
[14:30] <Laurenceb> where it==arm, object==cardboard box
[14:30] <eroomde> your unit of force is 'psi'
[14:30] <eroomde> which is upsetting
[14:30] <eroomde> and confusing
[14:30] <Laurenceb> lol sorry
[14:31] <Laurenceb> its a pneumatic bag based sensor
[14:31] <Laurenceb> anyway, ive replicated the issue with some simple sim code
[14:31] <eroomde> and that also explodes?
[14:32] <Laurenceb> yes
[14:32] <Laurenceb> seems the covar becomes unstable
[14:32] <Laurenceb> covar[0][1] and [1][0] are always zero btw
[14:33] <eroomde> that seems suspicious
[14:33] <Laurenceb> http://pastebin.com/DLWEdkSs
[14:33] <Laurenceb> yes
[14:33] <Laurenceb> ACTUATOR_LENGTH=30
[14:34] <eroomde> so, quick fire questions
[14:34] <eroomde> what is p_
[14:34] <Laurenceb> partial derivative
[14:34] <eroomde> what is it physically
[14:34] Armand (~martin@cpc17-haye16-2-0-cust427.haye.cable.virginmedia.com) left irc: Quit: Leaving
[14:35] <Laurenceb> error between kalman model and position
[14:35] <eroomde> right
[14:35] <eroomde> and s_
[14:35] <Laurenceb> the model "springyness"
[14:36] <eroomde> so the state vector is what - [springiness; position]?
[14:36] <Laurenceb> yes
[14:36] <eroomde> antispreinginess
[14:36] <eroomde> why is it negative of state[0]?
[14:37] <Laurenceb> as its the partial wrt State[1]
[14:37] <Laurenceb> and State[1] is negative
[14:38] <eroomde> hat's the physical model relating p and s?
[14:38] <Laurenceb> its the H vecotr
[14:38] <Laurenceb> yes
[14:38] <eroomde> what is*
[14:38] <Laurenceb> relating p and s?
[14:38] <eroomde> yeah
[14:38] <Laurenceb> there is no relation
[14:38] <Laurenceb> position is unrelated to springyness
[14:39] <Laurenceb> i want to solve for position and springyness
[14:39] <Laurenceb> observation model is pressure=State[0]*(Position-State[1])
[14:39] <Laurenceb> if thats what you mean
[14:39] <eroomde> hmm, why are you using a state space model then if you state vector contains unrelated things?
[14:39] <Laurenceb> ooh good question
[14:39] zamabe (~zamabe@unaffiliated/zamabe) left irc: Quit: The privilege of any man is the ability to deny limitation. Don't let gravity hold you down.
[14:40] <Laurenceb> thats why my Covar is always diagonal
[14:40] <eroomde> mmm
[14:40] <eroomde> there is no physical link between them in the state transition matrix, so they can't know about each other
[14:41] <Laurenceb> so, maybe i could do this with two decoupled filters?
[14:41] <eroomde> yeah
[14:41] <eroomde> i'd have thought so
[14:41] <Laurenceb> ill try this
[14:41] <eroomde> but note that for tracking position, tracking velocity is a good idea
[14:42] <Laurenceb> im using s stepper motor
[14:42] <eroomde> aaaaargh non linear woo
[14:42] <eroomde> cannot help you there
[14:42] <Laurenceb> so i know all this fairly accurately
[14:42] <eroomde> ok fine
[14:42] <Laurenceb> ive written an optimal controller
[14:42] <Laurenceb> it kind of works
[14:42] <Laurenceb> this bit is giving me the issues
[14:42] <eroomde> so you really just want a kalman filter to estimate deformability?
[14:43] <Laurenceb> yes
[14:43] <eroomde> that dounds like an easier problem to solve, especially if it's just modellable as a spring
[14:43] <Laurenceb> with covar diagonal, its almost decoupled already?
[14:43] <Laurenceb> oh wait
[14:43] <Laurenceb> s = p_ * ( Covar[0][0]*p_ + s_*Covar[0][1] ) + s_ * ( Covar[1][0]*p_ + s_*Covar[1][1] ) + Measurement_Covar;
[14:43] <Laurenceb> that line "couples" the elements
[14:43] <Laurenceb> ill rewrite it
[14:44] G8KNN-Jon (~Jon@cpc1-cmbg10-0-0-cust144.5-4.cable.virginmedia.com) joined #highaltitude.
[14:49] <eroomde> oh god this fpga stuff. i have a 500 page datasheet. this is basically the summary. i can having to refer to one of three other 500 page datasheets
[14:49] <eroomde> i am seriously not in kansas anymore
[14:49] <Randomskk> enjoy that
[14:50] <Randomskk> reminds me of the falling sensation when you finally accumulate all the relevant datasheets and programming manuals for an stm32, including the cm3 programming guide too
[14:50] <Randomskk> probably worse
[14:50] <Randomskk> though at least with the fpga you have somewhat fewer onboard peripherals
[14:50] <eroomde> it is a bit worse. this is really just electrical characteristics
[14:50] <eroomde> yes, fewer peripherals certainly :)
[14:51] <eroomde> it is nice though, i have a whole bunch of builtin nice things like hardware multipliers and 18k banks of ram
[14:51] <eroomde> dual ported
[14:51] <eroomde> so that provides a nice interface between one clock domain and another
[14:51] <Randomskk> on a 640 bit sample, at 0dB SNR, I get 0.019 BER, compared to 0.16 BER for a theoretical coherent matched filter decoder
[14:51] <Randomskk> but I have no idea what the variance is on my 0.019 figure.
[14:51] <eroomde> eg i can stick the gps data coming in with its own generated clock signal into a ram buffer
[14:52] <Randomskk> indeed. they are very happy for that sort of thing
[14:52] <eroomde> and then read it and put it into flash from the other port, on a different clock domain that i'm using for the flash interface and imu interface
[14:52] <eroomde> which is gonna be 32MHz i think
[14:52] <eroomde> these things are fabuslously capable though
[14:53] <eroomde> i'm barely taking it out of idle for this
[14:56] vladimirek (~vladimire@95.105.250.72) left irc: Remote host closed the connection
[14:57] <eroomde> Randomskk: so on this board i have 5V, 3.3V, 2.5V, and 1.2V
[14:57] <eroomde> a happy little family of LDOs
[14:57] bertrik (~quassel@rockbox/developer/bertrik) joined #highaltitude.
[15:04] <Randomskk> nice.
[15:04] <Randomskk> the sparkfun fpga board I have also has a chunky load of ldos
[15:05] <eroomde> yeah i need to pick a happy ldo family
[15:07] <Randomskk> entertainingly high currents too, peak
[15:07] <Laurenceb> grrr
[15:07] <Randomskk> except "peak" can mean "continuous" if you're working the fpga hard enough
[15:07] <Laurenceb> its still not behaving
[15:07] <Laurenceb> maybe i need to just bite the bullet and try it in matlab
[15:08] <Laurenceb> now as soon as it makes contact with target the covar jumps to zero
[15:08] <Laurenceb> for both of the decoupled filters
[15:08] <eroomde> Laurenceb: yes totally
[15:08] <eroomde> it's so hard to do these backwars
[15:08] <eroomde> get the maths working first
[15:09] <eroomde> then fight the coding bugs
[15:09] <Laurenceb> here goes another days/weeks work
[15:09] <eroomde> if there is one lesson with any of this lin alg stuff, it's get it working in matlab first
[15:09] <Laurenceb> been trying to get this think working for 3 months now :(
[15:09] <Laurenceb> i need to apply and release pressure to someones arm
[15:09] <Laurenceb> apply 3psi in ~5ms
[15:10] <Laurenceb> leave for ~1s with <0.5% error, then release over ~5ms
[15:10] <Laurenceb> so its stupidly hard
[15:10] <Laurenceb> arms do not play nicely
[15:11] <eroomde> you need their programming reference manual
[15:11] <Laurenceb> lol
[15:12] <Laurenceb> unconscious muscle fluctuations and stuff are enough to mess up the results
[15:13] <Randomskk> still 0.02 BER. I wonder if my code is wrong anywhere.
[15:13] <Randomskk> if not that is looking promising for the concept.
[15:13] <eroomde> :)
[15:13] <Randomskk> but it'd be so easy for there to be a mistake in the calculation of how much noise to add
[15:14] <eroomde> who makes good LDOs in happy packages, anyone? fixed value stuff
[15:14] <Randomskk> and I'm a little concerned that it doesn't seem to take into account the shift in use
[15:14] <Randomskk> everyone does
[15:14] <eroomde> i used micrel last time but have no emotional attachment to anything
[15:15] <Laurenceb> Ti make good stuff
[15:15] <Laurenceb> but expensive
[15:15] <eroomde> might just go for lm317
[15:15] <eroomde> old skool
[15:15] <eroomde> in honour of bob pease
[15:15] <eroomde> 'my favourite programming language? solder.'
[15:17] <gonzo_> I've always liked LT stiff. Becasue of the amusing app notes
[15:17] <gonzo_> stuff
[15:19] nosebleedkt (~kostas@213.140.128.74) left irc: Read error: Connection reset by peer
[15:20] <eroomde> yeah, the jim williams stuff is a gift to the world
[15:20] <eroomde> aswell as being funny
[15:20] <eroomde> there are all these funny little gems in them that you can miss
[15:21] <eroomde> like in the acknoweledgments section of one of his papers i was reading recently, it said something like 'the author would like to acknowledge the extensive commentry provided by Jerry Smithfield, some of which was helpful'
[15:21] <gonzo_> one app notes book had all the notes rated by numbers of baby bottles, as they had just had a kid and were rating the notes by the numbers of baby feeds that took place whilst writing them
[15:22] <jarod> Darkside you here?
[15:23] <gonzo_> could be a bit early in the morning for him
[15:25] Randomskk (~adam@paladin.randomskk.net) got netsplit.
[15:25] jonsowman (~jonsowman@kryten.hexoc.com) got netsplit.
[15:26] Randomskk (~adam@paladin.randomskk.net) returned to #highaltitude.
[15:26] jonsowman (~jonsowman@kryten.hexoc.com) returned to #highaltitude.
[15:30] cuddykid (~acudworth@cpc2-basf8-2-0-cust23.12-3.cable.virginmedia.com) left irc: Quit: cuddykid
[15:30] Randomskk (~adam@paladin.randomskk.net) got netsplit.
[15:30] jonsowman (~jonsowman@kryten.hexoc.com) got netsplit.
[15:31] <griffonbot> Received email: navrac "[UKHAS] Re: Pico launches sat 15th"
[15:32] Randomskk (~adam@paladin.randomskk.net) returned to #highaltitude.
[15:32] jonsowman (~jonsowman@kryten.hexoc.com) returned to #highaltitude.
[15:32] <Laurenceb> eroomde: ok so i fail at c
[15:32] <Laurenceb> http://i.imgur.com/2SR3D.png
[15:32] <Laurenceb> same thing in matlab
[15:32] <Randomskk> oh yuck. the eng servers have python 2.6 with a really old scipy
[15:32] <Randomskk> which doesn't have the wavelets peak finding thing
[15:32] <Randomskk> cry cry
[15:32] <Laurenceb> stabilises to the correct values soon as it makes contact
[15:33] <Laurenceb> exact same code...
[15:33] <Laurenceb> i must have typoed somewhere
[15:33] <Laurenceb> red=position, blue=spring constant
[15:33] Navrac_Work (545c0e05@gateway/web/freenode/ip.84.92.14.5) joined #highaltitude.
[15:33] <Laurenceb> values were 22.5 and 8
[15:33] <eroomde> Randomskk: ping them an email
[15:34] <eroomde> Laurenceb: glad it's working
[15:34] <Randomskk> they're running CentOS 6.3
[15:34] <Laurenceb> thanks for the help
[15:34] <eroomde> so you've kept the coupling in the model but fixed the C?
[15:34] <Randomskk> I don't think there's much hope :P
[15:34] <Laurenceb> so can anyone port matlab to my stm32f4 ?
[15:34] G8KNN-Jon (~Jon@cpc1-cmbg10-0-0-cust144.5-4.cable.virginmedia.com) left irc: Ping timeout: 252 seconds
[15:34] <Laurenceb> :P
[15:37] <Laurenceb> http://i.imgur.com/vOTyz.png
[15:37] <Laurenceb> thats actually pretty impressive
[15:39] bertrik (~quassel@rockbox/developer/bertrik) left irc: Quit: No Ping reply in 180 seconds.
[15:39] bertrik (~quassel@rockbox/developer/bertrik) joined #highaltitude.
[15:41] <Randomskk> maybe I'll be lucky and they'll have the things I need installed to build my own scipy
[15:42] <mattbrejza> any ideas for BER? still reckon ull break shannon?
[15:42] <Randomskk> I think you can safely assume I won't break shannon
[15:42] <mattbrejza> lol good
[15:43] <mattbrejza> that would have been a right pita
[15:43] <Randomskk> the simulation I just ran has about 0.02 BER for 0dB SNR
[15:43] <Randomskk> compared to 0.2 ish theoretical for a coherent matched filter
[15:43] <Randomskk> so like one order of magnitude better
[15:43] <mattbrejza> that seems better than your -14dB SNR prediction :P
[15:43] <Randomskk> that was just finding a single changepoint in a lot of noise
[15:43] <Randomskk> it's very good at that
[15:44] <mattbrejza> yea
[15:44] <mattbrejza> i would have thought you could get a paper out of what you are doing though
[15:44] <Randomskk> that would be fun
[15:44] <Randomskk> if I get it working on the department's 16 CPU servers it'l get a bit of a speedup
[15:46] <mattbrejza> at least you dont have to do BER with a coding scheme
[15:46] <Randomskk> I have multiprocessing working fairly trivially
[15:47] <Randomskk> well. for some aspects of the computation
[15:47] <Randomskk> enough to saturate 16 CPUs though
[15:47] <mattbrejza> could be a nice GPU or fpga application
[15:47] <Randomskk> yes
[15:47] <Randomskk> ugh ffs
[15:47] <Randomskk> BLAS isn't installed
[15:53] iamdanw (uid459@gateway/web/irccloud.com/x-bisxsuzivsnluijz) left irc: Read error: Connection reset by peer
[15:53] russss (uid30@gateway/web/irccloud.com/x-sjnutvlmngcqydby) left irc: Remote host closed the connection
[15:53] noonian338_ (uid6172@gateway/web/irccloud.com/x-diwcmtcrwjtfzben) left irc: Remote host closed the connection
[15:56] <Laurenceb> eroomde: P = 3.9666e-06 1.0367e-05
[15:56] <Laurenceb> 1.0367e-05 5.2823e-05
[15:56] <Laurenceb> ^interestingly there are off diagonal elements
[15:56] <jarod> anyone here knows a proper groundplane antenna calculator?
[15:56] <Laurenceb> seems to be where i messed up previously - the two state components end up being related via P
[16:02] G8KNN-Jon (~Jon@cpc1-cmbg10-0-0-cust144.5-4.cable.virginmedia.com) joined #highaltitude.
[16:14] russss (uid30@gateway/web/irccloud.com/x-ebstjpsactfxjtck) joined #highaltitude.
[16:16] <Randomskk> right. built my own ATLAS and LAPACK on the servers
[16:16] <Randomskk> now to build numpy
[16:19] <Randomskk> ugh what
[16:19] iamdanw (uid459@gateway/web/irccloud.com/x-owadqedsynebshvr) joined #highaltitude.
[16:24] <Laurenceb> <Randomskk> now to build numpy
[16:24] <Laurenceb> <Randomskk> ugh what
[16:24] <Laurenceb> i lolled
[16:24] <Randomskk> ATLAS failed to build PIC apparently
[16:24] <Randomskk> so numpy couldn't li nk
[16:24] <Randomskk> can't believe they don't just have a modern scipy and numpy on the system
[16:24] <Randomskk> almost tempted to just rent a beefy EC2 box
[16:24] <Randomskk> but no
[16:26] <griffonbot> Received email: MikeB "[UKHAS] Re: Pico launches sat 15th"
[16:32] Penfold (~mike@194.75.60.69) left irc: Ping timeout: 240 seconds
[16:39] num (4d6b9d0a@gateway/web/freenode/ip.77.107.157.10) joined #highaltitude.
[16:40] noonian338_ (uid6172@gateway/web/irccloud.com/x-qkhxxtsktqojoddm) joined #highaltitude.
[16:44] <griffonbot> Received email: navrac "[UKHAS] Re: Pico launches sat 15th"
[16:44] Nick change: num -> number10
[16:44] <Navrac_Work> this jt65 modulation is rather impressive
[16:44] <Randomskk> it's certainly interesting
[16:45] <Randomskk> my favourite part is the huge, huge rants a lot of hams get into about it
[16:45] <Randomskk> and how it's not "really" comunicating, and that people shouldn't be allowed QSOs or records for EME with jt65
[16:45] <Navrac_Work> just copied an australian 5watt station - despite not being able to hear it or see it on the waterfall
[16:45] <Randomskk> but when you say "copied"
[16:45] <Randomskk> there is an interesting question as to what you actually received
[16:45] <Navrac_Work> well it makes tweets and texts look 'wordy'
[16:46] <Randomskk> it makes for really amusing reading when hams who are unhappy with the idea of priors realise how it actually works
[16:46] <Navrac_Work> since pretty much all you get is a callsign and maybe your callsign and a signal report and the occasional 73's it doesnt really count as a conversation
[16:46] <gonzo_> it's just grumpy old men moaning
[16:47] <Randomskk> more to the point, most of the data is transmitted over the internet
[16:47] <craag> It's more the deep-search option, that looks for common words/callsign formats/layout and guesses what the missing characters might be.
[16:47] <Randomskk> craag: aiui deep search does a lot of trying to correlate against data it knows is being transmitted via the internet
[16:47] <gonzo_> infinite number of monkies
[16:47] <craag> Ah ok, hadn't even heard about that bit of it!
[16:48] <craag> It's an awesome mode for beacons imo. But I was surprised when I heard people use it for QSOs.
[16:48] <gonzo_> on eme, a contact is almost copying a single letter
[16:49] <Navrac_Work> 13 chars<> qso!
[16:49] <gonzo_> that's more than many people write in an SMS !
[16:49] Grumbleist (~Grumbleis@cpc8-camd13-2-0-cust233.hari.cable.virginmedia.com) left irc: Quit: Grumbleist
[16:50] <Navrac_Work> k
[16:51] G8KNN-Jon (~Jon@cpc1-cmbg10-0-0-cust144.5-4.cable.virginmedia.com) left irc: Ping timeout: 252 seconds
[16:54] SamSilver (2985f4d6@gateway/web/freenode/ip.41.133.244.214) left irc: Quit: Page closed
[16:55] DrLuke (~quassel@v120420003125117.hostingparadise.de) left irc: Quit: No Ping reply in 180 seconds.
[16:56] DrLuke (~quassel@v120420003125117.hostingparadise.de) joined #highaltitude.
[17:01] <jarod> http://x264.nl/dump/ads-b-antenna-groundplane-proper.jpg
[17:01] <jarod> http://x264.nl/dump/virtual-radar-server-adsb-working.jpg
[17:01] <jarod> :-)
[17:03] cuddykid (~acudworth@cpc2-basf8-2-0-cust23.12-3.cable.virginmedia.com) joined #highaltitude.
[17:05] <cuddykid> anyone know who's behind the "BBCSkyBalloon" - another UKHAS creation?
[17:08] <fsphil> where was that?
[17:08] <cuddykid> all over twitter - from the looks of it they have some knowledge but not that much
[17:09] <cuddykid> one sec
[17:09] <cuddykid> http://twitter.com/BBCSkyBalloon
[17:09] <cuddykid> why they have just got the balloon out the wrapper is beyond me lol
[17:10] <cuddykid> it will be interesting how they're planning to send down live tweets in the UK with all our restrictions
[17:10] <daveake> https://twitter.com/i/#!/BBCSkyBalloon/media/slideshow?url=pic.twitter.com%2FPyCNHSo4
[17:10] <daveake> Oh dear
[17:11] <russss> that is suboptimal
[17:11] <cuddykid> daveake: yep!
[17:11] <cuddykid> madness
[17:12] <Randomskk> haha oh dear
[17:12] <cuddykid> at least they've got the right radio
[17:13] <daveake> fsphil They're launching that from NI :p
[17:13] <griffonbot> Received email: steamfire@gmail.com "Re: [UKHAS] Re: Pico launches sat 15th"
[17:13] <daveake> Enniskillen
[17:14] <fsphil> I just spotted the BBC NI logo
[17:14] <daveake> https://maps.google.co.uk/maps?q=bt92+1ew&hl=en&ll=54.253894,-7.793255&spn=0.096372,0.239124&sll=52.8382,-2.327815&sspn=12.771239,30.60791&hnear=BT92+1EW,+United+Kingdom&t=m&z=13
[17:14] <fsphil> I'll give them a shout, cause they're probably going to land in the sea
[17:14] <cuddykid> wonder if they've spoke to david miller yet& haha
[17:14] <daveake> The postcode is visible in the pic of the CAA application
[17:14] radim_OM2AMR (~radimmuti@2a02:dd8:8aef:46:5117:bdae:dd7:ed7f) left irc: Ping timeout: 240 seconds
[17:14] <daveake> https://twitter.com/i/#!/BBCSkyBalloon/media/slideshow?url=pic.twitter.com%2F2ZKSsqIT
[17:15] <cuddykid> well spotted daveake
[17:15] G8KNN-Jon (~Jon@cpc1-cmbg10-0-0-cust144.5-4.cable.virginmedia.com) joined #highaltitude.
[17:15] <arko> morning
[17:16] <fsphil> that launch site is pretty good actually
[17:16] <fsphil> I'll still give them a shout
[17:16] <mattbrejza> http://www.bbc.co.uk/news/uk-northern-ireland-20671527
[17:16] <daveake> Aww, Adam, you've offered to help :)
[17:18] <cuddykid> fsphil: yeah, get involved!
[17:18] <bdale> fyi, those of you here who haven't heard about it already, keithp's MicroPeak design may be interesting as a tiny/light independent data point regarding peak altitude reached up to around 31km .. http://altusmetrum.org/MicroPeak
[17:18] <cuddykid> daveake: ha, I don't think I would be able to go along anyway - got too much on with exams atm
[17:19] rmp_ (~rmp@74.163.17.46.bridgep.com) joined #highaltitude.
[17:19] rmp_ (~rmp@74.163.17.46.bridgep.com) left irc: Client Quit
[17:19] <daveake> Well I'd be happy to help, but Phil is far better placed :)
[17:19] heathkid (~heathkid@unaffiliated/heathkid) left irc: Ping timeout: 244 seconds
[17:20] <fsphil> c'mon. we have bacon :)
[17:20] <fsphil> heading home. bbl
[17:21] <daveake> bacon butties later?
[17:22] mclane (~mclane@p5DD16FEB.dip.t-dialin.net) joined #highaltitude.
[17:23] rmp (~rmp@74.163.17.46.bridgep.com) left irc: Ping timeout: 248 seconds
[17:23] SamSilver (2985f4d6@gateway/web/freenode/ip.41.133.244.214) joined #highaltitude.
[17:24] heathkid (~heathkid@unaffiliated/heathkid) joined #highaltitude.
[17:24] <nigelvh> Speaking as a non english person, what are "bacon butties".
[17:25] <Randomskk> oh boy
[17:25] <mattbrejza> well bacon is meat from a big,
[17:25] <nigelvh> Don't get me wrong, I love bacon.
[17:25] <mattbrejza> pig
[17:25] <Randomskk> https://upload.wikimedia.org/wikipedia/commons/b/b1/Half_a_bacon_sandwich.jpg
[17:25] <Randomskk> +variation, to include sauces, butter, different types of bread or roll, etc
[17:26] <nigelvh> So it's bacon on an english muffin?
[17:26] <Randomskk> english muffin? no, bread
[17:26] <Randomskk> sometimes a bap
[17:26] <Randomskk> or a bun
[17:26] <Randomskk> usually warm
[17:26] <bdale> bacon is good. bread is optional.
[17:26] bertrik (~quassel@rockbox/developer/bertrik) left irc: Ping timeout: 252 seconds
[17:26] <nigelvh> That picture looks like an english muffin to me.
[17:26] <Randomskk> it's not
[17:26] <nigelvh> Ok, but bread of some sort
[17:27] <Randomskk> sure
[17:27] <Randomskk> you can just s/buttie/sandwich
[17:27] <mattbrejza> i am wondering what bacon + muffin would taste like though
[17:27] <Randomskk> it's bacon and bread. and butter and perhaps ketchup. and hot.
[17:27] <nigelvh> So bread + bacon + optional stuff. Alright
[17:27] <nigelvh> Also perhaps we have different ideas of english muffins.
[17:27] <Randomskk> perhaps
[17:28] <Randomskk> but I don't think so
[17:28] <nigelvh> http://www.foodsubs.com/Photos/englishmuffin.jpg
[17:28] <Randomskk> yea, that's what I was thinking of
[17:28] <Randomskk> you could maybe put bacon in there but it'd be weird
[17:28] <nigelvh> I don't see why.
[17:29] <Randomskk> I would typically have english muffins as part of eggs benedict or something
[17:29] <nigelvh> It's common here to do like eggs and cheese and bacon on an english muffin.
[17:29] <Randomskk> mmmm eggs benedict
[17:29] <Randomskk> sooo good
[17:29] <Randomskk> poached eggs are just great
[17:30] <daveake> mmm eggs benedict indeed
[17:30] <nigelvh> I'm generally not so into eggs alone, but as part of other stuff they're good.
[17:30] <nigelvh> Like french toast.
[17:30] <Randomskk> french toast is good
[17:30] <Randomskk> but I mean, eggs are good
[17:31] <nigelvh> You know what's also good? Buiscuits and Gravy
[17:31] <nigelvh> mmmmmmmm
[17:31] <Randomskk> runny fried egg with tabasco and maybe some bacon inside a warm mini baguette
[17:31] <Randomskk> glorious
[17:31] <Randomskk> nigelvh: ewww.
[17:31] <nigelvh> What's wrong with buscuits and gravy?
[17:31] <Randomskk> do you put gravy on your chips? (by chips I mean french fries)
[17:31] <nigelvh> No
[17:32] <Randomskk> https://en.wikipedia.org/wiki/File:Biscuits-and-gravy.jpg
[17:32] <Randomskk> that looks like someone sicked up last night's dinner all over a very unappetising breakfast
[17:32] <nigelvh> It's some buscuits, with meaty gravy, and hashbrowns. Delicious.
[17:33] <Randomskk> super gross
[17:33] <Randomskk> fwiw both biscuits and gravy typically conjure up very different imagery here
[17:33] <nigelvh> Disclaimer: My family is from the south, so growing up we ate a lot of southern foods despite living in the northwest.
[17:33] <nigelvh> Chicken and dumplings was another good one.
[17:34] <nigelvh> Chicken fried steak
[17:34] <nigelvh> meatloaf
[17:34] <nigelvh> In summary, meat tended to be involved. Generally some starches too.
[17:35] <Randomskk> hehe
[17:36] <nigelvh> This is closer to what i imagine a buscuits and gravy breakfast to be. http://www.chefseattle.com/images/review/151-biscuits-gravy.jpg
[17:36] <nigelvh> But yes, the idea is the same
[17:37] <Randomskk> http://jerjohns.com/wp-content/uploads/2012/08/uk-biscuit.jpg + http://static.guim.co.uk/sys-images/Guardian/Pix/pictures/2008/08/25/gravy460.jpg
[17:37] <Randomskk> ew ew ew
[17:37] <daveake> Biscuits, jelly, chips, crisps ... any other foody words you'd like to misuse over there? :p
[17:37] <nigelvh> Yes, you've got like turkey gravy there
[17:38] <nigelvh> and the buscuits are cookies
[17:38] <nigelvh> What's your jelly?
[17:38] <Randomskk> looking at it, it's dark enough to be beef gravy maybe
[17:38] <daveake> same as your jello
[17:38] <Randomskk> our jelly is like your jello, and our jam is your jelly
[17:38] <nigelvh> Ah, yeah, we use Jam and Jelly for preserves
[17:39] Armand (~martin@cpc17-haye16-2-0-cust427.haye.cable.virginmedia.com) joined #highaltitude.
[17:39] <Randomskk> those aren't cookies, they're biscuits. cookies are like http://upload.wikimedia.org/wikipedia/commons/9/9a/Choco_chip_cookie.jpg
[17:39] <nigelvh> Though, we differentiate Jam and Jelly based on how they're made
[17:40] <nigelvh> And yes Randomskk, I see the difference, but they're both called cookies here.
[17:40] <Randomskk> <:/
[17:40] <Randomskk> that's mildly upsetting
[17:41] <nigelvh> What are bangers and mash?
[17:41] <Randomskk> sausages and mashed potato
[17:41] <nigelvh> Ah
[17:41] <daveake> Proper sausages not the ones you have :)
[17:41] <Randomskk> yea, actual real sausages
[17:42] <Randomskk> not your gross reconstituted things
[17:42] bbjunkie (~bbjunkie@95.151.7.184) joined #highaltitude.
[17:42] <nigelvh> Yes, the ones you're thinking of are generally known as "breakfast sausages"
[17:42] <Randomskk> though our sausages are pretty gross too
[17:42] <Randomskk> when you stop and think about it
[17:42] <daveake> depends
[17:43] <nigelvh> You can move into "real" sausage territory when you start looking at hot dogs.
[17:43] <Randomskk> hot dogs aren't real sausages...
[17:43] <nigelvh> For example bratwurst is a common one for use in hot dogs
[17:43] <nigelvh> But yes, there are also cheap "hot dogs" made of mystery meat
[17:44] <daveake> Reminds me of http://www.youtube.com/watch?v=3EHsbIcV-6I
[17:44] <x-f> all this topic of food, i shall make some pancakes with jam now
[17:45] <nigelvh> Yep
[17:45] <nigelvh> A buddy of mine's dad used to work at a meat plant and told us all what went into the cheap "hot dogs"
[17:46] <nigelvh> It was less than appetizing.
[17:47] <Randomskk> seriously.
[17:47] <Armand> Whenever someone says "hotdogs" my automatic assumption is "mystery meat" :P
[17:47] <Randomskk> in my head it's more like
[17:48] <Randomskk> «mystery "meat"»
[17:48] <nigelvh> Yes
[17:48] <Armand> Solent sausages ? :P
[17:48] <Randomskk> lovely
[17:48] <Randomskk> omg ATLAS hurry up and compile!
[17:48] <Armand> spelling? :P
[17:48] <Randomskk> how can it still be going
[17:48] <nigelvh> Soylent
[17:48] <Armand> Ahh.. close. ^_^
[17:48] <Armand> I won't eat peas, so... meh.
[17:49] jonquark (jonl@nat/ibm/x-lnriogvuopaxfqnw) left irc: Remote host closed the connection
[17:52] Grumbleist (~Grumbleis@cpc8-camd13-2-0-cust233.hari.cable.virginmedia.com) joined #highaltitude.
[17:53] mclane (~mclane@p5DD16FEB.dip.t-dialin.net) left irc: Quit: Verlassend
[17:53] <nigelvh> Unrelatedly, I got my boards sent off to the fab, so hopefully soon I'll have a new version of my transmitter.
[17:54] Grumbleist (~Grumbleis@cpc8-camd13-2-0-cust233.hari.cable.virginmedia.com) left irc: Client Quit
[17:54] <nigelvh> http://uploads.oshpark.com/uploads/project/top_image/eOnj8xbq/i.png
[17:54] <Randomskk> cool
[17:55] <nigelvh> It's a bit smaller, and fixes some stuff, and adds a boost converter, so I'm pleased with the changes.
[17:55] <Randomskk> be fun seeing how well the boost converter plays with radios :P
[17:55] <Randomskk> this is moving off the adf7012?
[17:55] <Randomskk> or still using that for the pll bit?
[17:55] <nigelvh> Still the same ADF7012
[17:55] <nigelvh> It's been working well for me.
[17:56] <Randomskk> cool
[17:56] <nigelvh> So, with the ADF7012 and a little PA, I get about 250-300mW out
[17:56] <Randomskk> neat.
[17:56] <nigelvh> Works pretty well for APRS
[17:57] Lunar_Lander (83ad09ed@gateway/web/freenode/ip.131.173.9.237) joined #highaltitude.
[17:57] <Lunar_Lander> hello
[17:57] <nigelvh> Also got my packet code working pretty well. (not based on trackuino). So I'm pleased with that.
[17:57] <Lunar_Lander> I'm still at the lab
[17:57] <Lunar_Lander> we didn't turn the GPS up yet as we want to find 90° headers tomorrow
[17:58] <Lunar_Lander> but I added the coax cable to the NTX2 and then lambda/4 antennas
[17:58] <Lunar_Lander> seems to work OK
[17:59] <nigelvh> Good deal lunar
[17:59] <Lunar_Lander> yea
[18:00] <Lunar_Lander> also I got GPS lock with 5-6 satellites
[18:00] <Lunar_Lander> but that took considerable time
[18:00] <Lunar_Lander> so I think the GPS really should be moved
[18:01] <Lunar_Lander> OK
[18:01] <Lunar_Lander> cu!
[18:02] Lunar_Lander (83ad09ed@gateway/web/freenode/ip.131.173.9.237) left irc: Quit: Page closed
[18:02] <nigelvh> That was quick
[18:11] SamSilver (2985f4d6@gateway/web/freenode/ip.41.133.244.214) left irc: Quit: Page closed
[18:16] <jarod> http://x264.nl/dump/virtual-radar-server-adsb-working.jpg seems the home made antenna has good results :)
[18:19] signaleleven_ (~signalele@p5DE4734E.dip0.t-ipconnect.de) joined #highaltitude.
[18:20] signaleleven (81f71fe0@gateway/web/freenode/ip.129.247.31.224) left irc:
[18:22] johnboiles (~Adium@bastion.sfo1.yelpcorp.com) joined #highaltitude.
[18:23] Nick change: signaleleven_ -> signaleleven
[18:29] <costyn> so any tips on getting the EZCAP to work under Mac OS X with gqrx?
[18:29] <costyn> I think I have the right settings, but there's nothing happening on my screen
[18:31] <costyn> nevermind
[18:31] <Gadget-Mac> jarod: How high ?
[18:31] <costyn> found it :)
[18:31] radim_OM2AMR (~radimmuti@213.215.84.62) joined #highaltitude.
[18:32] johnboiles (Adium@bastion.sfo1.yelpcorp.com) left #highaltitude.
[18:33] <jarod> Gadget-Mac high enough :P
[18:37] <craag> jarod: Out of interest though, how high have you got the antenna. At that frequency I'd think clear line of sight matters more than the antenna itself.
[18:38] <jarod> yup, but having a properly made groundplane makes a huge difference as well
[18:38] <craag> It's a well-made antenna, I'll give that!
[18:39] <jarod> soldered it myself... i solder about 1-2x a year
[18:39] <jarod> :P
[18:46] <jarod> http://x264.nl/dump/adsb-london-hanover.jpg :)
[19:02] <costyn> morgen wordt 't 2012-12-12 12:12:12
[19:03] <costyn> erm
[19:03] <costyn> sorry, wrong channel, in any case, tommorow afternoon there's a moment of a lotta 12's :)
[19:03] daveake (~daveake@daveake.plus.com) left irc: Quit: Leaving
[19:07] <costyn> I should be able to receive FM on EZCAP with a 70cm 1/4 antenna right? I seem to remember it'll receive those with pretty much any antenna?
[19:07] <jarod> yup easy
[19:07] Hiena (~boreger@81.93.195.181.datatrans.hu) joined #highaltitude.
[19:08] <costyn> I'm getting nothing but static
[19:09] <costyn> running gqrx on mac os x
[19:10] <jarod> sure the rtlsdr is selected?
[19:11] <costyn> device string is label='ezcap USB 2.0 DVB-T/DAB/FM dongle',rtl=0
[19:11] <costyn> sounds about right
[19:13] <jarod> tuned a proper freq?
[19:13] <costyn> yea tried 102.1 FM
[19:13] <costyn> no peaks
[19:13] daveake (~Dave@daveake.plus.com) joined #highaltitude.
[19:14] <jarod> weird
[19:27] bbjunkie (~bbjunkie@95.151.7.184) left irc:
[19:30] <costyn> just tried a PMR walkietalkie
[19:30] <costyn> still get no peaks
[19:32] <radim_OM2AMR> ping fsphil
[19:34] <g0hww> costyn, have you fiddled with the gain?
[19:34] <g0hww> rf gain that is
[19:34] <costyn> g0hww: yea I think so
[19:35] <costyn> gqrx has only 1 gain slider it looks like
[19:35] <costyn> wait
[19:35] <costyn> somethings happening
[19:36] <g0hww> well, there's audio gain on the audio tab
[19:36] <costyn> I turned on auto gain on the ipnut controls
[19:36] <g0hww> and rf gain on the input controls tab
[19:36] <g0hww> it works now?
[19:36] <costyn> getting a tiny peak
[19:37] <costyn> don't hear my payload signal yet tho
[19:38] <fsphil> pongs radim_OM2AMR
[19:38] <radim_OM2AMR> fsphil http://i.imgur.com/WaPzt.png
[19:38] <radim_OM2AMR> :-D many thanks !!!
[19:38] <fsphil> oooh success!
[19:38] <costyn> g0hww: you also use gqrx?
[19:38] <fsphil> is it suppose to be purple? :)
[19:38] <radim_OM2AMR> just first packet, I will work on it :-)
[19:38] <fsphil> lol
[19:39] <fsphil> ah you're retransmitting the same packet
[19:39] <radim_OM2AMR> yes, for now (testing)
[19:40] <g0hww> yep
[19:40] <g0hww> with a USRP mostly
[19:40] <g0hww> the rtl dongle is in the loft purely for ADS-B stuff
[19:41] <radim_OM2AMR> fsphil, now I will check filesize and send whole image regarding it's size, thank to you and daveake for inspiration guys !
[19:41] <fsphil> np :)
[19:41] <costyn> g0hww: well there's a guy here who has got it working, but he's taking his sweet time with dinner
[19:41] <costyn> (here = hackerspace)
[19:41] <fsphil> you can never rush dinner :)
[19:42] <daveake> radim_OM2AMR well done :-)
[19:42] <daveake> 640 x 480 eh? Luxury :)
[19:42] <fsphil> lol
[19:42] <fsphil> that's like SSDV-HD
[19:42] <radim_OM2AMR> daveake, just for tests, I will follow your "conference" suggestions
[19:42] <daveake> hah
[19:43] <daveake> Well, have a play with resolution and compression to see what you think is best
[19:43] cuddykid (~acudworth@cpc2-basf8-2-0-cust23.12-3.cable.virginmedia.com) left irc: Quit: cuddykid
[19:43] <fsphil> I need to add the quality option
[19:44] <daveake> Add a greyscale mode for that early-Apollo feel :)
[19:44] <radim_OM2AMR> yes, I have to change usb cam as well, that's random one :-)
[19:44] <costyn> fsphil: actually he's doing the dishes. heeh
[19:44] <fsphil> oh that you can rush
[19:45] <radim_OM2AMR> ok, guys, back to gcc, nice evening to all :-D
[19:46] <daveake> I've been trying to get a USB 3G dongle running on the Pi. It appears in usb-devices as 2 storage devices. Been trying to hit it with usb_modeswitch but it won't play :(
[19:48] <fsphil> that's odd
[19:48] <fsphil> no serial interfaces at all?
[19:48] <radim_OM2AMR> hmm, I'm just beginner, so I can't help you unfortunately :-(
[19:49] <daveake> yes, the modeswitch disables the storage ones but doesn't seem to switch them to serial
[19:50] <daveake> Or, it may be that Linux isn't then installing them as serial devices.
[19:50] <daveake> I'll have another go tomorrow.
[19:51] johnboiles1 (~Adium@bastion.sfo1.yelpcorp.com) joined #highaltitude.
[20:00] <Upu> ping KF7FER
[20:01] johnboiles1 (Adium@bastion.sfo1.yelpcorp.com) left #highaltitude.
[20:06] <Upu> PAVA4 turned up in Calais :)
[20:06] <costyn> g0hww: we tried another EZCAP under windows and it just works. trying to get it working under virtualbox now
[20:07] <Upu> Well Wissant
[20:07] Dan-K2VOL (~Dan-K2VOL@145.sub-70-192-140.myvzw.com) joined #highaltitude.
[20:08] <daveake> Great news :)
[20:08] Action: g0hww is amazed to here of windows working
[20:08] <g0hww> hear, even
[20:08] <Navrac_Work> wow nice one upu
[20:09] <Upu> got a text message
[20:09] <arko> eroomde: guess who was in line behind me at Chipotle?
[20:09] <daveake> Does that mean you now have a 100% recovery record Upu?
[20:10] <Upu> well not got it back yet :)
[20:10] <daveake> Well, after you've paid the ransom :)
[20:10] <fsphil> how long's that been in the water?
[20:10] <Upu> 8 days
[20:10] <daveake> minus 4?
[20:10] NickSF (027bc91d@gateway/web/freenode/ip.2.123.201.29) joined #highaltitude.
[20:10] <daveake> They found it 4 days ago you said
[20:12] <Upu> yeah Sat
[20:12] <Upu> was launched on the 31st
[20:12] <Upu> 30th
[20:12] <fsphil> be interesting to see what conditions it's in
[20:12] <Upu> 8 days in the sea
[20:12] <daveake> Oh OK you already did the maths :)
[20:12] <Upu> Yes :)
[20:12] <Upu> Well asked for a photo see what comes back
[20:14] <Upu> anyway evening Navrac_Work
[20:14] <Upu> did you get my mail ?
[20:14] <Navrac_Work> oh havent ooked for a while its been manic at work - 1sec
[20:14] <Upu> no worries
[20:16] <Navrac_Work> upu yep I'm about
[20:16] <Upu> lol
[20:16] <Upu> busy ?
[20:16] <Upu> or go 10 mins for a PM ?
[20:17] <Upu> got
[20:18] <Navrac_Work> yep just firing up irc on the local machine rather than vncing intio thisone
[20:18] <Upu> ok
[20:19] navrac (navrac@84.92.14.5) joined #highaltitude.
[20:19] <navrac> right I'm here
[20:23] <costyn> Upu: power LED would be nice on the HABamp :)
[20:23] rmp (~rmp@host-2-96-116-62.as13285.net) joined #highaltitude.
[20:24] <fsphil> nah
[20:24] <fsphil> but it does need some text
[20:24] Laurenceb_ (~Laurence@host86-178-5-207.range86-178.btcentralplus.com) joined #highaltitude.
[20:39] cuddykid (~acudworth@cpc2-basf8-2-0-cust23.12-3.cable.virginmedia.com) joined #highaltitude.
[20:45] G8KNN-Jon (~Jon@cpc1-cmbg10-0-0-cust144.5-4.cable.virginmedia.com) left irc: Ping timeout: 252 seconds
[20:48] <eroomde> observe a genuine xkcd drawing from 2012
[20:48] <eroomde> http://what-if.xkcd.com/imgs/a/24/model_space.png
[20:48] <eroomde> now observe a drawing i did in the style of xkcd for a talk in 2008
[20:48] <eroomde> http://i.imgur.com/4VKYS.png
[20:48] <arko> haha
[20:49] Dan-K2VOL (~Dan-K2VOL@145.sub-70-192-140.myvzw.com) left irc: Quit: Leaving.
[20:52] <Upu> lol eroomde
[20:54] <daveake> nice :)
[21:04] <navrac> eroomde - with gps repeaters if you move around within the repeaters coverage area does the gps actuaslly work accurately? I cant see how it can
[21:05] <jarod> eroomde seen my new antenna? :)
[21:05] UpuWork (~Upu@2a02:b80:12:1:d4c0:c599:3103:f78d) left irc: Ping timeout: 260 seconds
[21:05] <eroomde> correct - they basically just tell you where the repeater is
[21:05] <eroomde> usualled used for things like parachute jumpers so they have a lock before jumping out
[21:05] <eroomde> otherwise it spends 30s trying to get a lock by which time you've already landed
[21:05] <eroomde> jarod: nope?
[21:06] <eroomde> linky
[21:06] <jarod> http://x264.nl/dump/ads-b-antenna-groundplane-proper.jpg
[21:06] <nigelvh> We've got one in a lab at the university to get GPS signals inside.
[21:06] <jarod> result: http://x264.nl/dump/virtual-radar-server-adsb-working.jpg
[21:06] <eroomde> v nice work jarod
[21:07] <eroomde> nigelvh: i guess that will just tell you where the repeater's antenna is
[21:07] <eroomde> good enough to get the time and check lock though i guess
[21:07] <nigelvh> Yes, the key there is just for time and functioning.
[21:09] UpuWork (~Upu@2a02:b80:12:1::4) joined #highaltitude.
[21:09] <navrac> might just build a preamp and couple it to a small aerial at either endfor testing in my office
[21:09] <nigelvh> That would work.
[21:09] <eroomde> yeah
[21:10] <eroomde> i'd do saw/lna/saw if poss
[21:10] <eroomde> and active ant at either end
[21:11] <eroomde> the signal is just so poo otherwise
[21:12] Grumbleist (~Grumbleis@cpc8-camd13-2-0-cust233.hari.cable.virginmedia.com) joined #highaltitude.
[21:12] <navrac> yeah was just looking upsawsasyou saidthat
[21:13] G8KNN-Jon (~Jon@cpc1-cmbg10-0-0-cust144.5-4.cable.virginmedia.com) joined #highaltitude.
[21:15] <eroomde> navrac: indeed, it's a good idea
[21:16] <eroomde> Randomskk found a nice combined saw/lna chip with power supply
[21:16] <eroomde> pathological footprint though
[21:16] <nigelvh> Aren't they always?
[21:17] <Randomskk> still my favourite footprint name
[21:17] <Randomskk> SUPERNEMO
[21:17] <eroomde> this especially so
[21:17] <eroomde> https://randomskk.net/u/supernemo.png
[21:17] <eroomde> that's about 3mm long
[21:18] <Upu> ewww
[21:18] <eroomde> i made it in eagle too
[21:18] UpuWork (~Upu@2a02:b80:12:1::4) left irc: Ping timeout: 264 seconds
[21:18] UpuWork (~Upu@2a02:b80:12:1::4) joined #highaltitude.
[21:18] <eroomde> then didnt use it when i saw the current consumption
[21:19] <arko> sup eroomde
[21:19] <eroomde> saw the current consumption
[21:19] <eroomde> geddit?
[21:19] <eroomde> lol
[21:19] Grumbleist (~Grumbleis@cpc8-camd13-2-0-cust233.hari.cable.virginmedia.com) left irc: Quit: Grumbleist
[21:19] <eroomde> yo arko
[21:19] <eroomde> my having happy fun time
[21:19] <eroomde> doing an fpga based pcb
[21:19] <arko> :) nice!
[21:19] <eroomde> learning curve
[21:19] <arko> heh
[21:19] <arko> bga?
[21:19] <eroomde> lots of datahseets to read
[21:19] <nigelvh> SUPERNanoExtraMicroOmgicantsolderthis
[21:19] <eroomde> nope, lqfp208
[21:20] <eroomde> 3cm to a side
[21:20] <arko> oh ok
[21:20] <eroomde> chunky monkey
[21:20] <arko> yeah thats huge
[21:20] daveake (~Dave@daveake.plus.com) left irc: Quit: Leaving
[21:21] <eroomde> v interesting though
[21:22] <eroomde> also check out the actel igloo nano range
[21:22] <arko> actel has nice software
[21:22] <Randomskk> eroomde: hmmm. the matched filter decoder is actually working a whole ton better.
[21:22] <eroomde> spartan 3 gate count in 5mm x 5mm packages
[21:22] <arko> i hate xilinx ISE
[21:22] <eroomde> 0.4mm bga
[21:22] <arko> bga rocks
[21:23] <arko> just make sure you have a heat gun or hot plate and a microscope ;)
[21:23] <eroomde> rocks for necessitating expensive pcbs
[21:23] <eroomde> 4mm traces and 0.1, maybe 0.2mm drill holes
[21:23] <arko> when footprint matters
[21:23] <eroomde> 4mil*
[21:24] <fsphil> good luck
[21:24] <arko> hah, i was like thats a huge trace
[21:24] <eroomde> yeah, i have a cunning plan for using them one day
[21:24] <Upu> yeah 4mm is like Duplo lego trackes
[21:24] <Upu> traces
[21:24] <eroomde> combined gps/ins with 1khz 18dof samples the size of a postage stamp
[21:25] <arko> 18dof?
[21:25] <eroomde> 18 degrees of freedom
[21:25] <arko> yeah i know
[21:25] <eroomde> that's the size of the state vector it could produce
[21:26] <eroomde> with the post processing software
[21:26] <arko> ohhh
[21:26] <eroomde> so, xyz position, velocity, accel = 9
[21:26] <eroomde> the the same for attitude
[21:26] <arko> makes much more sense now
[21:26] UpuWork (~Upu@2a02:b80:12:1::4) left irc: Ping timeout: 264 seconds
[21:26] <arko> i have only seen 9dof
[21:26] <arko> actually
[21:26] <arko> 12
[21:27] <arko> mag x,y,z
[21:27] UpuWork (~Upu@2a02:b80:12:1:d4c0:c599:3103:f78d) joined #highaltitude.
[21:27] <eroomde> 9dof usually includes mag
[21:27] <eroomde> usually
[21:27] <arko> right
[21:27] <eroomde> 3 x gyro, 3 x accel, 3 x mag
[21:27] <arko> velocity i assume is generated by integrating
[21:27] <arko> or derivative, but that would noisy
[21:28] <eroomde> but using kalman filters (or particle filters or whatever) provides a nice way of looking at these problems
[21:28] <eroomde> because you have an 18 dof state vector
[21:28] NickSF (027bc91d@gateway/web/freenode/ip.2.123.201.29) left irc: Ping timeout: 245 seconds
[21:28] <eroomde> you have a huge great state transition matrix where various of the things are functions of each other
[21:28] <eroomde> and you have this v large measurement matrix of all your sensors
[21:28] <eroomde> and lots of the sensors can be used to infer more than one thing in the state vector
[21:29] <eroomde> so you can treat it as a big bayesian probelm of correlated sensors and just fart a state vector out the end
[21:29] <eroomde> this is not exactly a real time thing computationally, mind :)
[21:31] <eroomde> so to answer your q, you can get velocities or accels by doing nicer things that just differentiation or integration of the output of a single sensor
[21:32] <eroomde> eg the classic example of a kalman filter infering velocity from position measurements in a much less noisy and more accurate way than differentiationg the position measurements
[21:32] <eroomde> it makes me quite excited just to think about it
[21:38] <nigelvh> eroomde is a smart dude.
[21:38] <nigelvh> I just call it math.
[21:39] <nigelvh> Then again I don't have a degree.
[21:39] <eroomde> it's just bayes rule
[21:39] <eroomde> i think one could explain a kalman filter to anyone who can multiply 2 numbers in an hour or so
[21:39] <eroomde> and this is notthing more complicated than that, just quite a big model
[21:40] <nigelvh> I'm sure I could figure it out, I've just never done so.
[21:40] <eroomde> but same principle as a small model one could explain quite simply
[21:41] Hiena (~boreger@81.93.195.181.datatrans.hu) left irc: Remote host closed the connection
[21:42] <griffonbot> Received email: Anthony Stirk "[UKHAS] PAVA/QAVA"
[21:42] <nigelvh> Yeah, I got through calc 1 in college and didn't really go any farther in math.
[21:43] <fsphil> idea for a video series. "Ed Explains"
[21:43] <Laurenceb_> eroomde: my EKF code, guess what was wrong
[21:43] <Laurenceb_> i was updating the covar in place
[21:43] <Laurenceb_> and using updated values in the next line
[21:44] <eroomde> oh god
[21:44] Last message repeated 1 time(s).
[21:44] <eroomde> i have done that
[21:44] <Laurenceb_> hehe
[21:44] <eroomde> there are 2 hard problems in computer science
[21:44] <eroomde> naming variables, cache coherency and off-by-one erros
[21:45] <eroomde> and this is an example of the first
[21:45] <eroomde> PREDICTED!!!_Cov
[21:45] <eroomde> should be the variable name
[21:45] <Laurenceb_> yes
[21:45] <Laurenceb_> thanx for the help
[21:45] <Laurenceb_> im quite impressed by the performance now
[21:47] <eroomde> they are a bit witchcraft aren;t they
[21:47] <eroomde> in a sort of make-you-smile sort of way
[21:48] <Randomskk> really good inference is magical
[21:48] <Laurenceb_> if spring constant is changing at the same time as position it can find incorrect values
[21:48] <Laurenceb_> but its still sane enough that the controller can manage
[21:48] <eroomde> you track the biases and see the estimate stay well on the truth and the tracked bias wondering off
[21:48] <Laurenceb_> and tbh if both variables are changing at the same time its impossible to estimate accurately
[21:48] <eroomde> and it's just
[21:48] <eroomde> http://cdn.memegenerator.net/instances/400x/26865815.jpg
[21:48] <Laurenceb_> hehe
[21:49] <arko> haha
[21:50] <eroomde> i would seriously get bayes rule as a tattoo if i ever had to get a tattoo
[21:52] <arko> haha
[21:52] <eroomde> good stuff UpuWork
[21:52] <arko> i couldnt do tattoos
[21:53] johnboiles (~Adium@bastion.sfo1.yelpcorp.com) joined #highaltitude.
[21:53] <Randomskk> eroomde: right. this implementation of the changepoint decoder is actually pretty terrible
[21:53] <Randomskk> and additionally I don't think I'm adding enough noise for a given SNR
[21:54] <Randomskk> because the hugely simple matched filter is beating everything hands down
[21:54] <eroomde> hey it's optimal when it's optimal
[21:55] <eroomde> it's when the poo hits the fan snr-wise that cp based thing should help i think
[21:55] hyte (5e000397@gateway/web/freenode/ip.94.0.3.151) joined #highaltitude.
[21:55] <Randomskk> if I knew that I was adding the right level of noise for a given SNR I'd be more confident in saying that the matched filter is still doing better :P
[21:56] <eroomde> is albery fabregas still at cued?
[21:56] <eroomde> albert*
[21:56] <Randomskk> the name rings a bell
[21:56] <Randomskk> yes
[21:56] <eroomde> he would be a very good person to chat to about it
[21:56] <eroomde> he used to take 4F6
[21:57] <Randomskk> 4f6 wasn't run this year :( :(
[21:57] <Randomskk> hmm. the CUED servers are taking about 2x as long to run a computation here
[21:57] <Randomskk> and the matched filter is still running in about ten milliseconds vs about one/two minutes for bayes
[21:57] <Randomskk> heh
[21:58] <eroomde> :)
[22:03] G8KNN-Jon (~Jon@cpc1-cmbg10-0-0-cust144.5-4.cable.virginmedia.com) left irc: Ping timeout: 252 seconds
[22:19] MNSP (~Mitul@cpc22-lutn10-2-0-cust252.9-3.cable.virginmedia.com) joined #highaltitude.
[22:19] <MNSP> Hello All :-)
[22:19] <fsphil> hiya MNSP
[22:19] zamabe (~zamabe@unaffiliated/zamabe) joined #highaltitude.
[22:20] <MNSP> Alright fsphil, hows hings?
[22:20] <MNSP> *things
[22:21] <fsphil> nice and quiet here, you?
[22:21] <MNSP> all good thanks
[22:22] <fsphil> my room has been quite tidy for nearly two weeks now. think this is a record
[22:23] <MNSP> well done sir! mine has yet to see that state of grace
[22:27] Lunar_Lander (~gd-compto@p54882F6D.dip.t-dialin.net) joined #highaltitude.
[22:27] <Lunar_Lander> hello
[22:30] <MNSP> hi lunar
[22:30] G8KNN-Jon (~Jon@cpc1-cmbg10-0-0-cust144.5-4.cable.virginmedia.com) joined #highaltitude.
[22:36] <joph> fsphil, sounds like magic
[22:36] <griffonbot> Received email: Steve Daniels "RE: [UKHAS] PAVA/QAVA"
[22:36] <griffonbot> Received email: Ali al-Azzawi "Re: [UKHAS] Pico launches sat 15th"
[22:38] MNSP (~Mitul@cpc22-lutn10-2-0-cust252.9-3.cable.virginmedia.com) left irc: Quit: Leaving
[22:38] number10 (4d6b9d0a@gateway/web/freenode/ip.77.107.157.10) left irc: Quit: Page closed
[22:38] johnboiles1 (~Adium@bastion.sfo1.yelpcorp.com) joined #highaltitude.
[22:42] johnboiles (~Adium@bastion.sfo1.yelpcorp.com) left irc: Ping timeout: 255 seconds
[22:45] <jarod> a question .... a groundplane is vertical right? is it possible to make a horizontal groundplane?
[22:46] <fsphil> that would make it a dipole
[22:46] <jarod> a dipole receives vertical signals as well
[22:46] <jarod> i am using this for 1090mhz now: http://x264.nl/dump/ads-b-antenna-groundplane-proper.jpg
[22:47] <jarod> i can make a similar for 860mhz .... but that band... we use it horizontal
[22:47] <jarod> so i use this: http://x264.nl/dump/konig-ant-uhf70-kn.jpg
[22:47] johnboiles1 (Adium@bastion.sfo1.yelpcorp.com) left #highaltitude.
[22:47] <jarod> but can you make a ground plane horizontal?
[22:48] <fsphil> you mean a quarter-wave antenna?
[22:48] <eroomde> yes
[22:48] <jarod> yes
[22:48] <eroomde> you can
[22:48] <fsphil> indeed
[22:48] <jarod> not its receiving vertical.... can it be made for horizontal?
[22:49] <jarod> *now
[22:49] <eroomde> oh hang on
[22:49] <eroomde> i don;t understand the q
[22:49] <eroomde> you want an omnidirectional antenna (in the N/E/S/W sense) that is horizontally polarised?
[22:49] <jarod> yes, is that possible?
[22:49] <eroomde> if so, have a look at something like a big wheel
[22:49] <eroomde> yep it is
[22:49] <eroomde> ^
[22:50] <eroomde> http://www.hamtv.com/wheel.html
[22:50] <jarod> ah
[22:50] <jarod> still looks directional :P
[22:50] <eroomde> it is
[22:51] <eroomde> every single antenna is
[22:51] <jarod> horizontal omnidirectional is possible or not?
[22:52] <eroomde> big wheel is prob as close as you're gonna get
[22:52] <eroomde> it's not as bad as you might think
[22:52] <eroomde> especially if made well
[22:52] <jarod> why would i think its bad?
[22:52] <eroomde> http://wesselva.home.xs4all.nl/bigwheel/big_wheel_4m.jpg
[22:52] <eroomde> because you said it doesn't look very omni
[22:53] <jarod> ah try... you mean not as bad.. directional wise
[22:53] <eroomde> sorry yes
[22:53] <jarod> ok ok
[22:53] <jarod> well the ground plane truely works amazing when properly soldered and cut to length
[22:54] <eroomde> skew planar wheel antenna
[22:54] <eroomde> for the truly omni
[22:54] <jarod> cool design ;)
[22:54] <eroomde> or quadrifilar helical
[22:54] <eroomde> which is a great antenna design
[22:55] <eroomde> that's probably the shizzle for your aeroplaneidge
[22:55] <eroomde> although actually intake that back as i don;t know exactly how they are transmitting
[22:56] <jarod> its all amazing stuff
[22:56] <jarod> in Amsterdam i saw a plane on the map in the middle of germany :)
[23:00] <Lunar_Lander> hi jarod
[23:00] <Lunar_Lander> you are from NL?
[23:02] <jarod> ya
[23:03] <Lunar_Lander> cool
[23:03] <Lunar_Lander> I'm from germany
[23:05] <__red__> We won't hold that against you :-)
[23:05] <Lunar_Lander> xD
[23:06] bertrik (~quassel@rockbox/developer/bertrik) joined #highaltitude.
[23:12] chris_99 (~chris_99@unaffiliated/chris-99/x-3062929) left irc: Quit: Leaving
[23:12] <Lunar_Lander> btw
[23:12] <Lunar_Lander> this is the first ever picture of earth from space http://s.gullipics.com/image/y/a/c/5yvo9j-jhcf8k-twby/explorer6earth.jpeg
[23:12] <Lunar_Lander> 1959, taken by Explorer VI
[23:14] hyte (5e000397@gateway/web/freenode/ip.94.0.3.151) left irc: Ping timeout: 245 seconds
[23:18] rmp (~rmp@host-2-96-116-62.as13285.net) left irc: Quit: rmp
[23:22] G8KNN-Jon (~Jon@cpc1-cmbg10-0-0-cust144.5-4.cable.virginmedia.com) left irc: Ping timeout: 252 seconds
[23:22] hyte (5e000397@gateway/web/freenode/ip.94.0.3.151) joined #highaltitude.
[23:22] radim_OM2AMR (~radimmuti@213.215.84.62) left irc: Quit: radim_OM2AMR
[23:23] Arbition (Arbition@compsci.adl/treasurer/arbition) left irc: Changing host
[23:23] Arbition (Arbition@unaffiliated/arbition) joined #highaltitude.
[23:27] johnboiles (~Adium@bastion.sfo1.yelpcorp.com) joined #highaltitude.
[23:29] bertrik (~quassel@rockbox/developer/bertrik) left irc: Ping timeout: 252 seconds
[23:30] Laurenceb_ (~Laurence@host86-178-5-207.range86-178.btcentralplus.com) left irc: Ping timeout: 264 seconds
[23:30] <SpeedEvil> from orbit
[23:31] Nabobalis (~Nabster@cpc11-shef11-2-0-cust26.barn.cable.virginmedia.com) left irc: Ping timeout: 240 seconds
[23:31] <arko> go home space ship , you're drunk
[23:31] Cadair (~Cadair@unaffiliated/cadair) left irc: Ping timeout: 252 seconds
[23:36] Nabobalis (~Nabster@cpc11-shef11-2-0-cust26.barn.cable.virginmedia.com) joined #highaltitude.
[23:37] Cadair (~Cadair@unaffiliated/cadair) joined #highaltitude.
[23:46] Armand (~martin@cpc17-haye16-2-0-cust427.haye.cable.virginmedia.com) left irc: Quit: Leaving
[23:48] johnboiles (Adium@bastion.sfo1.yelpcorp.com) left #highaltitude.
[23:49] cuddykid (~acudworth@cpc2-basf8-2-0-cust23.12-3.cable.virginmedia.com) left irc: Quit: cuddykid
[23:49] G8KNN-Jon (~Jon@cpc1-cmbg10-0-0-cust144.5-4.cable.virginmedia.com) joined #highaltitude.
[23:52] phuzion (~phuzion@ipv6.irc.teh-server.com) left irc: Ping timeout: 260 seconds
[23:53] <meatmanek> wow there are a lot of different types of artifacts there
[23:53] <meatmanek> this is why changing formats is bad
[23:54] <meatmanek> presumably they took a film photo, developed, and scanned it on the craft?
[23:54] <meatmanek> then printed that onto film on the ground
[23:54] <meatmanek> which then got dusty and was scanned later
[23:54] <meatmanek> plus scanner noise plus jpeg compression
[23:55] <lz1dev> wat
[23:56] phuzion (~phuzion@ipv6.irc.teh-server.com) joined #highaltitude.
[23:56] <lz1dev> they could aways take a photo with a phone
[23:56] <lz1dev> and then send it through fax
[23:57] <nigelvh> Then send that fax through SSTV
[23:59] <Lunar_Lander> ah as far as I know the camera was similar to the one on the ABLE-1 Pioneers
[23:59] <Lunar_Lander> so that one strip was scanned during one rotation of the satellite
[00:00] --- Wed Dec 12 2012