highaltitude.log.20100817

[00:06] LazyLeopard (~irc-clien@chocky.demon.co.uk) left irc: Quit: Bye
[00:10] SpeedEvil (~user@tor/regular/SpeedEvil) left irc: Ping timeout: 255 seconds
[00:16] timbobel (~timboebl@5ED040EE.cable.ziggo.nl) left irc:
[00:22] ejcweb (~chatzilla@222.133.gr5.adsl.brightview.com) left irc: Quit: ChatZilla 0.9.86 [Firefox 3.6.8/20100723084720]
[00:25] SpeedEvil (~user@tor/regular/SpeedEvil) joined #highaltitude.
[01:08] GeekShadow (~Antoine@reactos/tester/GeekShadow) left irc: Quit: The cake is a lie !
[01:12] SpeedEvil (~user@tor/regular/SpeedEvil) left irc: Quit: Leaving.
[01:12] SpeedEvil (1000@tor/regular/SpeedEvil) joined #highaltitude.
[01:31] natrium42 (~natrium@CPE000625d867e2-CM0014045885be.cpe.net.cable.rogers.com) joined #highaltitude.
[01:31] juxta (juxta@ppp203-122-193-94.static.internode.on.net) joined #highaltitude.
[01:39] <juxta> morning all
[01:41] <natrium42> evening juxta
[01:41] <juxta> hey natrium42
[01:42] <juxta> do you have HF setup at your place?
[01:42] <natrium42> yeah
[01:42] <natrium42> but i can only listen in (no license)
[01:42] <juxta> oh right
[01:42] <juxta> what do you use for an antenna? just a wire?
[01:43] <natrium42> yep, 100 ft long wire
[01:43] <natrium42> probably not the best setup...
[01:44] <juxta> hehe - do you have a tuner on it? or just the wire straight out?
[01:45] <natrium42> yes, i have one
[01:45] <natrium42> it
[01:45] <natrium42> it's specially for long wire antennas
[01:46] <juxta> I'm considering stringing a wire up for HF but i'd need to get a tuner
[01:46] <juxta> had a chance to play with HF in a hams car yesterday - was pretty impressive, able to make contacts in the US
[01:47] <natrium42> hehe, it's fun to listen in to stations from all over the world
[01:47] <natrium42> need to get a license....
[01:48] <juxta> hehe, it's not too hard ;p
[01:48] <natrium42> i know :S
[01:49] <SpeedEvil> natrium42: yeah - how else would one talk to people in other countries for free :)
[01:49] <juxta> hehe
[01:49] <natrium42> phreaking
[01:51] <natrium42> SpeedEvil is oldschool, rite? :)
[01:51] TraumaPony (~TraumaPon@202.173.162.181) joined #highaltitude.
[01:53] <SpeedEvil> Well - oldskool enough that I diddn't start out with internet porn.
[02:26] alikins_ (~adrian@cpe-066-057-101-225.nc.res.rr.com) joined #highaltitude.
[04:51] <griffonbot> @spaceflight: RT @cuspaceflight: Release of the new web-based landing predictor: http://www.cuspaceflight.co.uk/predict #cusf #ukhas [http://twitter.com/spaceflight/status/21372507973]
[05:14] alikins_ (~adrian@cpe-066-057-101-225.nc.res.rr.com) left irc: Quit: alikins_
[06:12] natrium42 (~natrium@CPE000625d867e2-CM0014045885be.cpe.net.cable.rogers.com) left irc: Quit: Leaving
[06:28] Upu (~Upu@ubn.upuaut.net) joined #highaltitude.
[06:36] blomlet_ (quassel@unaffiliated/blomlet) left #highaltitude ("Parted").
[07:00] Simon-MPFH (~simon@phantom.mpfh.co.uk) joined #highaltitude.
[07:10] earthshine (~earthshin@cpc3-orpi1-0-0-cust867.2-3.cable.virginmedia.com) left irc: Ping timeout: 265 seconds
[07:11] Simon-MPFH (~simon@phantom.mpfh.co.uk) left irc: Ping timeout: 265 seconds
[07:11] Randomskk (~adam@paladin.randomskk.net) left irc: Ping timeout: 265 seconds
[07:11] astevens (~astevens@caosvm2.osuosl.org) joined #highaltitude.
[07:11] Simon-MPFH (~simon@phantom.mpfh.co.uk) joined #highaltitude.
[07:11] SpeedEvil (1000@tor/regular/SpeedEvil) left irc: Ping timeout: 265 seconds
[07:11] DanielRichman (daniel@gateway/shell/anapnea.net/x-dbniswiasicocgzl) left irc: Ping timeout: 265 seconds
[07:11] epictetus (pattm@static-71-174-73-53.bstnma.fios.verizon.net) left irc: Ping timeout: 265 seconds
[07:12] Randomskk (~adam@paladin.randomskk.net) joined #highaltitude.
[07:13] DanielRichman (daniel@gateway/shell/anapnea.net/x-fylhtcmokdgdrkkr) joined #highaltitude.
[07:13] astevens_ (~astevens@caosvm2.osuosl.org) left irc: Ping timeout: 264 seconds
[07:13] epictetu1 (pattm@static-71-174-73-53.bstnma.fios.verizon.net) joined #highaltitude.
[07:13] SpeedEvil (1000@tor/regular/SpeedEvil) joined #highaltitude.
[07:34] Jasperw1 (~jasperw@212-98-44-230.static.adslpremium.ch) left irc: Ping timeout: 265 seconds
[08:04] Jasperw (~jasperw@212-98-44-230.static.adslpremium.ch) joined #highaltitude.
[08:04] m1x10 (m1x10@ppp089210068033.dsl.hol.gr) joined #highaltitude.
[08:05] <m1x10> morning all
[08:09] <Upu> morning
[08:36] GeekShadow (~Antoine@reactos/tester/GeekShadow) joined #highaltitude.
[08:47] m1x10 (m1x10@ppp089210068033.dsl.hol.gr) left irc: Ping timeout: 240 seconds
[08:51] m1x10 (m1x10@ppp089210097096.dsl.hol.gr) joined #highaltitude.
[09:13] GW8RAK (~chatzilla@host86-153-49-32.range86-153.btcentralplus.com) joined #highaltitude.
[09:19] rharrison (~rharrison@gateway.hgf.com) joined #highaltitude.
[09:20] earthshine (~earthshin@cpc3-orpi1-0-0-cust867.2-3.cable.virginmedia.com) joined #highaltitude.
[09:21] SpeedEvil (1000@tor/regular/SpeedEvil) left irc: Ping timeout: 265 seconds
[09:22] GeekShadow (~Antoine@reactos/tester/GeekShadow) left irc: Ping timeout: 240 seconds
[09:24] GeekShadow (~Antoine@reactos/tester/GeekShadow) joined #highaltitude.
[09:24] <rharrison> Hi Upu
[09:25] <rharrison> Still waiting for my permission to come back for Elvington, York
[09:26] <m1x10> rharrison, is this a bot response ?
[09:26] <m1x10> lol
[09:26] <rharrison> hehe yep
[09:27] <m1x10> lol
[09:27] <m1x10> god zeus lol
[09:28] <m1x10> :p
[09:28] <rharrison> zeusbot, is tired :)
[09:28] <m1x10> heh
[09:35] SpeedEvil (~user@tor/regular/SpeedEvil) joined #highaltitude.
[09:55] <earthshine> morning
[09:56] <m1x10> yo :)
[10:13] <Upu> York really would be good :)
[10:14] <Upu> just going to put an AVR on a breadboard now then once I have all that working it's PCB making time
[10:14] <Upu> I have some mad plans in my head
[10:16] <Upu> think you could put a kinderegg inside a balloon before inflating it ?
[10:26] <rharrison> hehe probably
[10:26] <rharrison> It might hurt if it hit you on the hea
[10:26] <rharrison> d
[10:27] <rharrison> Perhaps a small para would be good
[10:28] <Upu> considering putting one in with a pressure sensor + temp sensor
[10:29] <Upu> speaks to flight comp via bluetooth
[10:29] <Upu> and can vent at a certain temperature
[10:30] <Upu> then some how I can heat the balloon up not sure how as I don't have mega watt maser lying about
[10:30] <Upu> alot of somehows in this lot :)
[10:30] <Upu> I'll setting with getting a camera up and back first
[10:32] <Upu> settle even
[10:58] <Upu> m1x10 Timobel was online last night
[11:02] <m1x10> Upu np, I was testing the bot :)
[11:35] SpeedEvil (~user@tor/regular/SpeedEvil) left irc: Quit: Leaving.
[11:35] SpeedEvil (1000@tor/regular/SpeedEvil) joined #highaltitude.
[11:57] m1x10 (m1x10@ppp089210097096.dsl.hol.gr) left irc: Quit: I eat electrons for breakfast and I vomit thunders
[13:00] ejcweb (~chatzilla@61.80.125.91.rb4.adsl.brightview.com) joined #highaltitude.
[13:17] Simon-MPFH (~simon@phantom.mpfh.co.uk) left irc: Read error: No route to host
[13:21] Simon-MPFH (~simon@phantom.mpfh.co.uk) joined #highaltitude.
[13:23] Simon-MPFH (~simon@phantom.mpfh.co.uk) left irc: Read error: No route to host
[13:27] Simon-MPFH (~simon@phantom.mpfh.co.uk) joined #highaltitude.
[13:57] <fsphil> Upu, put a tiny tracker in the egg -- be interesting to see how far it lands from the main payload :)
[14:01] alikins_ (~adrian@cpe-066-057-101-225.nc.res.rr.com) joined #highaltitude.
[14:02] Simon-MPFH (~simon@phantom.mpfh.co.uk) left irc: Quit: Leaving
[14:08] alikins__ (~adrian@cpe-066-057-101-225.nc.res.rr.com) joined #highaltitude.
[14:09] alikins__ (~adrian@cpe-066-057-101-225.nc.res.rr.com) left irc: Client Quit
[14:11] alikins_ (~adrian@cpe-066-057-101-225.nc.res.rr.com) left irc: Ping timeout: 240 seconds
[14:13] m1x10 (m1x10@ppp089210097096.dsl.hol.gr) joined #highaltitude.
[14:39] alikins (~alikins@nat/redhat/x-xyykssdxlvenydrg) joined #highaltitude.
[15:00] gonzo_ (~gonzo_@cpc1-pool8-0-0-cust154.pool.cable.virginmedia.com) left irc: Read error: Connection reset by peer
[15:01] ejcweb (~chatzilla@61.80.125.91.rb4.adsl.brightview.com) left irc: Ping timeout: 246 seconds
[15:02] Darkside (~Darkside@hpavc/darkside) left irc: Ping timeout: 240 seconds
[15:03] Darkside (~Darkside@hpavc/darkside) joined #highaltitude.
[15:03] gonzo_ (~gonzo_@cpc1-pool8-0-0-cust154.15-1.cable.virginmedia.com) joined #highaltitude.
[15:18] lovelace (lovelace@unaffiliated/lovelace) joined #highaltitude.
[15:21] ejcweb (~chatzilla@91.125.248.228) joined #highaltitude.
[15:23] GW8RAK (~chatzilla@host86-153-49-32.range86-153.btcentralplus.com) left irc: Quit: ChatZilla 0.9.86 [Firefox 3.6.8/20100722155716]
[15:44] rharrison_ (~rharrison@gateway.hgf.com) joined #highaltitude.
[16:11] rharrison_ (~rharrison@gateway.hgf.com) left irc: Quit: Leaving
[16:27] Jasperw (~jasperw@212-98-44-230.static.adslpremium.ch) left irc: Ping timeout: 265 seconds
[16:33] alikins_ (~adrian@nat/redhat/x-ejvueuxjaczqnqdp) joined #highaltitude.
[16:36] Hualon (~Hualon@static-71-111-254-156.rlghnc.dsl-w.verizon.net) joined #highaltitude.
[16:37] ejcweb (~chatzilla@91.125.248.228) left irc: Read error: Connection reset by peer
[16:48] Jasperw (~jasperw@212-98-44-230.static.adslpremium.ch) joined #highaltitude.
[17:11] ejcweb (~chatzilla@192.9.gr6.adsl.brightview.com) joined #highaltitude.
[17:14] Darkside (~Darkside@hpavc/darkside) left irc: Ping timeout: 240 seconds
[17:16] Darkside (~Darkside@hpavc/darkside) joined #highaltitude.
[17:18] Simon-MPFH (~simon@phantom.mpfh.co.uk) joined #highaltitude.
[17:20] <alikins> any best practices on setup of a chdk'ed camera for a high altitude baloon launch?
[17:24] Hualon (~Hualon@static-71-111-254-156.rlghnc.dsl-w.verizon.net) left irc: Quit: ChatZilla 0.9.86 [Firefox 3.6.8/20100722155716]
[17:29] alikins_ (~adrian@nat/redhat/x-ejvueuxjaczqnqdp) left irc: Quit: alikins_
[17:42] m1x10 (m1x10@ppp089210097096.dsl.hol.gr) left irc: Ping timeout: 252 seconds
[17:43] rharrison (~rharrison@gateway.hgf.com) left irc: Quit: Leaving
[17:45] timbobel (~timboebl@5ED040EE.cable.ziggo.nl) joined #highaltitude.
[17:46] m1x10 (m1x10@ppp089210143131.dsl.hol.gr) joined #highaltitude.
[17:47] Action: timbobel succesfully tested this schematic in which you can select 5v or 3v3 with 1 switch to produce 425 shift
[17:47] <timbobel> http://hollandshoogte.files.wordpress.com/2010/08/radscheme.jpg
[18:08] spenguin[work] (~penguin@59.162.86.164) left irc: Read error: Connection reset by peer
[18:13] spenguin[work] (~penguin@59.162.86.164) joined #highaltitude.
[18:24] juxta (juxta@ppp203-122-193-94.static.internode.on.net) left irc: Ping timeout: 240 seconds
[18:25] <m1x10> is it possible to go up to 40km alt?
[18:29] <timbobel> y
[18:29] <timbobel> though not with a ~<1500 gram balloon and ~>1kg payload. that'd be very unlikely
[18:31] <timbobel> i think the ukhas needs another launch :). there hasnt been one for ages
[18:37] Simon-MPFH (~simon@phantom.mpfh.co.uk) left irc: Quit: Leaving
[18:45] <fsphil> definitely
[18:57] <timbobel> are there even plans?
[18:58] ejcweb (~chatzilla@192.9.gr6.adsl.brightview.com) left irc: Read error: Connection reset by peer
[18:59] <fsphil> there's a few payloads being worked on, but no dates yet I think
[19:00] <timbobel> which ones are those?
[19:02] <fsphil> I believe DanielRichman and co are working on one
[19:02] <fsphil> and my own
[19:03] <timbobel> o rly
[19:03] <timbobel> that'd be great
[19:04] <m1x10> hey dont forget mixio !
[19:05] alikins_ (~adrian@nat/redhat/x-ppytgnhbbfxvjgvf) joined #highaltitude.
[19:08] <timbobel> m1x10, what's taking so long ;)?
[19:10] <m1x10> lol
[19:10] <m1x10> now working on gsm module
[19:11] <m1x10> aprs is almost ready. just some little things
[19:12] <m1x10> then need to decide about cameras
[19:12] <m1x10> and then to find a suitable styrofam
[19:12] <m1x10> I will not laucnh this year
[19:14] <timbobel> geeeezzzz really!??
[19:14] <timbobel> why not
[19:14] <m1x10> I got to go to army 1 year
[19:14] <timbobel> not one year in greece
[19:14] <timbobel> 9 months right
[19:14] <m1x10> yeah
[19:14] <timbobel> they call it leigbour
[19:14] <timbobel> hahaha
[19:14] <m1x10> lol
[19:15] <timbobel> i had greek friends that had the asme
[19:15] <timbobel> same
[19:15] <m1x10> it is planned to be 13months next year
[19:15] <timbobel> doesnt that conflict with EU rules?
[19:15] <timbobel> military mumbo jumbo
[19:16] ejcweb (~chatzilla@192.9.gr6.adsl.brightview.com) joined #highaltitude.
[19:16] <m1x10> dunno, I just understand that some neighhour countries get aggresiv, hence country protection must be increased
[19:16] <timbobel> that is true
[19:16] <m1x10> I pressume this with my small mind
[19:16] <timbobel> the EU would not allow that though, so need to get them all into military
[19:17] <timbobel> if anyone tries they're finished, come on, in europe :)
[19:17] <m1x10> no.. its not like that.
[19:18] <timbobel> national defense is one thing, 'drafting' is another
[19:19] <timbobel> but anyway.. too bad you wont launch in a while
[19:19] <m1x10> there is oil in aegean, and oil brings bad future
[19:41] <fsphil> the sooner someone invents simple and easy to use nuclear fusion, the better ;)
[19:42] ejcweb (~chatzilla@192.9.gr6.adsl.brightview.com) left irc: Read error: Connection reset by peer
[19:42] <m1x10> yeah
[19:43] <m1x10> i think back in USSR they did some work and I think nowdays some scientist are trying to prove some stuff.
[19:44] <m1x10> improve*
[19:48] <m1x10> timbobel can I come to Holland to work ?!?! :):)
[19:55] <m1x10> yeah! GSM responds to my sms!
[19:55] RocketBoy (~steverand@217.47.75.8) joined #highaltitude.
[19:58] <m1x10> SMS received
[19:58] <m1x10> Saved in: ME
[19:58] <m1x10> Index: 29
[19:58] <m1x10> XAXAXA
[19:58] <m1x10> :P
[19:58] ejcweb (~chatzilla@192.9.gr6.adsl.brightview.com) joined #highaltitude.
[19:58] <fsphil> nice
[19:59] <fsphil> so the new module is working fine?
[19:59] alikins_ (~adrian@nat/redhat/x-ppytgnhbbfxvjgvf) left irc: Quit: alikins_
[20:00] <m1x10> up to now yes
[20:02] philihp (~philihp@sas08103.nat.sas.com) joined #highaltitude.
[20:02] <philihp> omg there is a support group for this. <3
[20:03] <fsphil> hi philihp :)
[20:03] <fsphil> thinking of launching a balloon?
[20:03] <philihp> yes!
[20:04] <philihp> been dreaming of doing it for about 2 months now
[20:04] <m1x10> heh
[20:04] <philihp> main question i have is are there any out-of-the-box gps/tracking devices?
[20:05] <fsphil> hmm, there was a link posted here to something like that some time ago. I don't think my log goes back that far though
[20:05] <fsphil> most here make their own - it's fairly simple
[20:06] <SpeedEvil> philihp: where are you?
[20:06] <philihp> raleigh, nc
[20:06] <Randomskk> philihp: the making your own is like, loads of the fun
[20:06] <SpeedEvil> USA?
[20:06] ejcweb (~chatzilla@192.9.gr6.adsl.brightview.com) left irc: Remote host closed the connection
[20:06] <philihp> confirm
[20:06] <Randomskk> are you a licensed radio amateur?
[20:06] <Randomskk> (or are any of your friends?)
[20:06] <philihp> Randomskk, i'm sure it is, starting points would be greatly appreciated
[20:06] <Randomskk> that's really the easiest way to go in the states
[20:06] <Randomskk> you can get /loads/ of pre-made trackers that operate on amateur radio frequencies/protocols
[20:06] <philihp> i am not... but i am a member of the national association of rocketry
[20:07] <Randomskk> here in the UK we are not allowed to operate airborne so have to do things our own
[20:07] <Randomskk> you can do stuff entirely yourself and there are various unlicensed freqs you can use
[20:07] <SpeedEvil> http://www.ringolake.com/pic_proj/gps_trak/freetrak202.html
[20:07] <Randomskk> but perhaps the easiest route would be to get licensed or make friends with someone who is
[20:07] <SpeedEvil> APRS canbe used
[20:08] <alikins> philihp: are you on the techshop balloon team?
[20:08] <philihp> no, i have no idea what that is
[20:09] <alikins> philihp: http://www.techshoprdu.com/forum/9 techshoprdu is going to do a launch this weekend, we hope
[20:10] <philihp> <3, i will be there
[20:10] Action: alikins has to run
[20:10] Nick change: alikins -> alikins_away
[20:10] <fsphil> launches are good fun, and pretty educational too
[20:12] jcoxon (~jcoxon@nat67.mia.three.co.uk) joined #highaltitude.
[20:12] <jcoxon> evening all
[20:12] <fsphil> howdy jcoxon
[20:13] <Randomskk> yo jcoxon
[20:14] <jcoxon> hey, sorry for the absence - been working
[20:14] <Randomskk> how's that going?
[20:14] <jcoxon> long days
[20:15] <jcoxon> just got home
[20:15] <Randomskk> :( I know what that's like at least
[20:15] Action: Randomskk got home 15min ago
[20:15] <jcoxon> eek
[20:15] <rjharrison> boo
[20:15] <Randomskk> need to start leaving work earlier
[20:15] <jcoxon> (best not get into a competition :-p)
[20:16] <Randomskk> haha yes, definitely the wrong thing to compete about
[20:16] <rjharrison> hey jcoxon r u back to work now then
[20:16] <jcoxon> hey rjharrison, apologies for the lack of a response
[20:16] <jcoxon> rjharrison, indeed pulling my 12hr days
[20:16] <Randomskk> and worse, I blew up a FET on the thing I was working on, so now I don't know what to do
[20:16] <Randomskk> the casing separated, molten metal came out, the insides had totally fused, all the magic smoke was released
[20:17] <rjharrison> ooh nice
[20:17] <jcoxon> it'll be worth it when i get paid :-p
[20:17] <jcoxon> got my fsa03 breakout boards
[20:17] <jcoxon> of course as soon as earthshine stopped selling them
[20:17] <Randomskk> hah yes I tell myself the same thing
[20:17] <Randomskk> then my company went into administration and I started working for a new company the same day
[20:17] Action: Randomskk is yet to be paid
[20:18] <jcoxon> oh dear
[20:18] <jcoxon> bless the NHS monstrosity
[20:18] <Randomskk> yes, it is at least quite unlikely they will go into administration :P
[20:18] <jcoxon> anyone subscribed to the GPSL mailing list? watching the delightful argument about formatting of tables
[20:19] <jcoxon> Randomskk, i sort of think they are in administration perpetually
[20:19] <Randomskk> hehe
[20:27] <jcoxon> ping earthshine
[20:31] <RocketBoy> hey its the jcoxon
[20:32] <jcoxon> hey RocketBoy
[20:32] <RocketBoy> sounds like you have been working your socks off
[20:32] <jcoxon> yeah
[20:32] <jcoxon> busy bust
[20:32] <jcoxon> busy*
[20:32] <RocketBoy> yeah - me to grabbing the money while its there
[20:32] <rjharrison> I like the idea of busy busts
[20:33] <RocketBoy> this burst has lasted 9 months
[20:34] <RocketBoy> there is talk of another 6 :(
[20:34] <RocketBoy> jcoxon: I finally go myself a mac
[20:34] <jcoxon> RocketBoy, yay for macs
[20:34] <fsphil> noooo!
[20:34] <fsphil> ;-)
[20:34] <RocketBoy> yea for macs
[20:35] <RocketBoy> 2
[20:35] Action: fsphil hides his mac
[20:35] <jcoxon> so any launches?
[20:35] <RocketBoy> just got my 3G of email off the PC - bit of a job - had to transfer it to macmail via thunderbird
[20:35] <earthshine> hello
[20:36] <jcoxon> hey earthshine i see you've stopped stocking fsa03 :-(
[20:36] <RocketBoy> I keep planning a chute test flight
[20:36] <jcoxon> i just made a breakout board
[20:37] <RocketBoy> Has jonsnowman been about - I havn't seen him on in days
[20:37] <jcoxon> right supper time bbl
[20:37] <earthshine> jcoxon: Yeah I bought a whole load of them months ago and they were really slow to sell
[20:38] <Randomskk> jonsowman is on holiday
[20:38] <earthshine> even selling them cheap[er than ANYWHERE else they went as fast as a slug in treacle
[20:38] <jcoxon> earthshine, shame
[20:38] <Randomskk> Cyprus for the week. lucky so and so
[20:38] <earthshine> So it is not viable to keep them in stock
[20:38] <jcoxon> bbl
[20:38] jcoxon (~jcoxon@nat67.mia.three.co.uk) left irc: Remote host closed the connection
[20:40] <RocketBoy> Randomskk: thanks I'll wait till he gets back
[20:44] GeekShadow (~Antoine@reactos/tester/GeekShadow) left irc: Read error: Connection reset by peer
[20:45] GeekShadow (~Antoine@reactos/tester/GeekShadow) joined #highaltitude.
[21:06] SpeedEvil (1000@tor/regular/SpeedEvil) left irc: Ping timeout: 276 seconds
[21:13] SpeedEvil (1000@tor/regular/SpeedEvil) joined #highaltitude.
[21:17] Hualon (~Hualon@cpe-069-134-137-028.nc.res.rr.com) joined #highaltitude.
[21:21] timbobel (~timboebl@5ED040EE.cable.ziggo.nl) left irc:
[21:33] Laurenceb (~laurence@host86-145-126-39.range86-145.btcentralplus.com) joined #highaltitude.
[21:33] <Laurenceb> hi
[21:33] <Randomskk> yo
[21:33] <Randomskk> :( I blew up a fet on my quad's motor controller
[21:36] <Laurenceb> nooo
[21:36] Action: Laurenceb realised you need DC blocking caps for RF switches
[21:37] <Hualon> :-(
[21:37] <Laurenceb> however you can get SC-70 packages which are smaller than SOT6
[21:38] <Hualon> Randomskk: bummer!
[21:38] <Laurenceb> Randomskk: firmware bug?
[21:38] <Randomskk> Laurenceb: potentially
[21:39] <Randomskk> need to implement the code that checks the current consumption and kills everything if it's too high
[21:39] <Randomskk> thing is though
[21:39] <Randomskk> I unpowered the AVR
[21:39] <Randomskk> leaving the 12V plugged in
[21:39] <Randomskk> yet the FET was still open
[21:39] <Randomskk> might have stuck open
[21:39] <Laurenceb> you can get charge stuck on the gate
[21:39] <Randomskk> yea but it should be pulled shut
[21:39] <Randomskk> all the inputs are pulled appropriately
[21:39] <Laurenceb> dodgy resistor soldering?
[21:40] <Randomskk> don't think so
[21:40] <Randomskk> my best guess is either the FET failed open earlier and just kept getting hotter
[21:40] <Randomskk> or the AVR remained powered through some of its inputs
[21:40] <Randomskk> some of the ADC inputs were going to divided down 12V, which could power the thing
[21:41] <Randomskk> and in remaining powered, kept the fet open
[21:41] <Randomskk> dunno if anything else on the board is damaged, but I've taken the fet off and may just solder a new one in place and hope for the best
[21:42] <Randomskk> annoyingly these are all r4 boards so have a few hacked on resistors and things (though they don't appear to be related); wasting fets on it is annoying when I have r5 designed (but not fabbed)
[21:42] <Randomskk> might move to SC-70 transistors instead of SOT23 for the FET drivers though
[21:42] <Randomskk> trying to work out how to get the bom cost down even more
[21:43] <Randomskk> might drop the special opamp for current monitoring for a normal opamp, adds a few passives but probably cheaper at the end of th eday
[21:43] <Laurenceb> how big isd the board?
[21:45] <Randomskk> 1.7" by 0.9"
[21:45] <Randomskk> about 43mm by 23mm
[21:46] <Laurenceb> not bad
[21:46] <Randomskk> https://randomskk.net/quad/mc/motor_controller_r4_brd.png is the r4 I have made
[21:46] <Laurenceb> whats the max current?
[21:46] <Randomskk> eeeh
[21:46] <Randomskk> really, limited by the connectors on r4
[21:46] <Randomskk> r5 swaps to big ass holes you solder wires to
[21:46] <Randomskk> then I suspect limited by MOSFET heat dissipation and the PCB traces
[21:47] <Randomskk> I wouldn't want to put more than 15A or so through, but haven't tested. theoretically it could do more
[21:47] <Laurenceb> nice
[21:47] <Laurenceb> autopilot board is 40x60mm
[21:47] <Randomskk> the mosfets are good to like 23A apparently
[21:48] <Randomskk> 33A for the N channel and -23A for the P channel
[21:48] <Randomskk> need to match them better too. need to pick better mosfets all around, those are expensive
[21:48] <Randomskk> like £1 each and there are six of them
[21:49] <Randomskk> open-bldc has found that using many smaller fets is a much better idea and while I'm inclined to agree I can't fit them on the PCB without putting parts on both sides, which I am both loathe to do and also would really require 4 layers
[21:49] <Randomskk> they use three PCBs stacked on top of each other, it's quite small but not tiny. uses an stm32
[21:49] <Randomskk> they are not trying to optimise for cost though, rather functionality
[21:50] <Randomskk> proper mosfet drivers and buffers, stm32 doing field vector control
[21:50] <Laurenceb> for the motor controller?
[21:50] <Randomskk> http://open-bldc.org/wiki/V0.3
[21:50] <Randomskk> yea.
[21:50] <Laurenceb> crazy
[21:50] <Randomskk> yea
[21:50] <Laurenceb> thats pretty hardcore
[21:50] <Randomskk> I imagine it's quite cool and efficient too
[21:50] <Laurenceb> you could surely run an electric car with that :P
[21:50] <Randomskk> probably very nice to fly with, and all that
[21:51] <Randomskk> but it does seem like overkill for having four of them on a quadcopter
[21:51] <Laurenceb> just a little
[21:51] <Randomskk> I like it a lot, and for a super awesome personal quad I'd be tempted
[21:51] <Randomskk> but right now my goal is massively minimising cost
[21:51] <Laurenceb> looks more like what you'd find in the back of a tesla roadster
[21:52] <Randomskk> does rather. check out the arrays of fets http://open-bldc.org/wiki/File:Open-BLDC-V0_3-pstage_top.jpg
[21:52] <Laurenceb> crazy
[21:52] <Randomskk> mine's quite a lot smaller but there's definitely a lot of work to get it cheaper
[21:53] <Randomskk> I'm also starting to get concerned about how complicated it is to do BLDC control properly
[21:53] <Laurenceb> talking of which... it seems pretty practical to make a fold up electric scooter thingy
[21:53] <Laurenceb> its not _too_ hard
[21:53] <Randomskk> it is perhaps harder than I thought at first
[21:53] <Randomskk> I'm sure I can do it - I have (had) a motor spinning, but without commutation detection
[21:54] <Randomskk> bldc mention that they were sampling ADC once every PWM cycle to see the voltage but that only allowed for like 4k rpm apparently, which seems crazy to me (especially on an stm32)
[21:54] <Laurenceb> you can get crazy speed, acceleration and range - and as its so small the batteries dont cost a fortune
[21:54] <Laurenceb> read the links on rcgroups
[21:54] <Randomskk> but doing it reliably, all the time, without fault, with good fault handling, with good startup/recovery, with decent non-blocking communication interfaces, etc
[21:54] <Randomskk> and at the same time getting the per unit cost way down
[21:55] <Laurenceb> theres lots of very good threads with lots of links describing the theory well there
[21:55] <Randomskk> decided to use SPI as the only comms interface
[21:55] <Laurenceb> :(
[21:55] <Randomskk> since I can then use the ICSP header without any extras, and it's fast and relatively easy to do
[21:55] <Laurenceb> my board is really designed for spi
[21:55] <Randomskk> allows the stm32 to reflash the motor controllers too
[21:55] <Laurenceb> s/spi/i2c
[21:55] <Laurenceb> fardoos
[21:55] <Randomskk> no spi pins broken out?
[21:55] <Laurenceb> yes, but its a bit of a pain
[21:56] <Randomskk> r4 had i2c pins on the 8pin icsp header, but I forgot that the icsp programmer cable is shrouded so I have to bend the i2c pins out the way
[21:56] <Randomskk> (bottom two pins on that png I linked)
[21:56] <Laurenceb> i see
[21:56] <Randomskk> r5 just has those two taken out, so easy enough to add back in but then has the same problems
[21:56] <Laurenceb> oh well my board has a brushed esc built in anyway...
[21:57] <Randomskk> why?
[21:57] <Laurenceb> tho larger than a few amps starts to screw with the magno
[21:59] <Laurenceb> *shrug*
[21:59] <Laurenceb> it might be useful, also there was board space spare
[21:59] <Randomskk> fair enough
[21:59] <Laurenceb> also its useful as a ballooning cutdown
[22:00] <Randomskk> I'll see if I can't keep i²c on there somewhere
[22:00] <Randomskk> also want to consider swapping to an all n-channel design for the power stage
[22:00] <Randomskk> cheaper and higher efficiency plus removes a line from the bom
[22:00] <Randomskk> but requires funny voltages
[22:00] <Laurenceb> how do you get the gate voltage?
[22:00] <Randomskk> exactly, it's a bit more complicated. wouldn't want a smps
[22:00] <Laurenceb> guess you could make a charge pump running off the avr
[22:01] <Randomskk> potentially
[22:01] <Laurenceb> if you have a pwm channel spare
[22:01] <Randomskk> just about
[22:01] <Randomskk> hmm, maybe, maybe not
[22:01] <Laurenceb> or.... you could use the tower pro firmware hack from rcgroups
[22:01] <Randomskk> think I'm using all three PWM outs for fet driving
[22:01] <Randomskk> nah
[22:02] <Randomskk> specifically set out to not do that :P
[22:02] Action: Randomskk has some tower pros sitting around though
[22:02] <Randomskk> I need to be sure I can even use the three PWM outputs independently, at that
[22:02] <Randomskk> OC1A, OC1B, OC2
[22:02] <Randomskk> technically all three should be able to do independent pwm
[22:02] <Randomskk> on the other hand now that I finally have a scope I can at least test stuff
[22:02] <Laurenceb> yes, tho two run off the same timer
[22:02] <Laurenceb> also timer 2 is only 8 bit
[22:02] <Randomskk> yea
[22:03] <Randomskk> was hoping 8 bits would be enough speed control
[22:03] <Randomskk> if not I'm going to need to upgrade to a beefier micro I guess
[22:03] <Laurenceb> its prob enough
[22:03] <Randomskk> the two being on the same timer shouldn't be a massive issue
[22:03] <Randomskk> I only need one at a time
[22:03] <Randomskk> A+, B+ and C+ are connected to PWM outs
[22:03] <Laurenceb> my tower pro with a mega8 works very well
[22:04] <Randomskk> A-, B- and C- go to just on/off
[22:04] <Randomskk> so at any given point one side of the bridge is being PWMd and the other left open
[22:04] <Randomskk> but never two PWMs at once
[22:05] <Randomskk> I can't help but feel a lot of careful design and engineering went into the tower pro's use of the mega8
[22:05] <Randomskk> it does work extremely well
[22:05] <Laurenceb> _very_ small as well
[22:07] <Laurenceb> seems to use variable pwm clock rate and other tricks from the noises it makes
[22:07] <Laurenceb> or maybe they just wanted it to sound awesome
[22:07] <Randomskk> http://www.flickr.com/photos/randomskk/3952061394/ and http://www.flickr.com/photos/randomskk/3952064768/
[22:07] <Randomskk> you can start to guess some of it
[22:08] <Randomskk> also a few dirty hacks, like sticking two dpak vregs on top of each other
[22:08] <Randomskk> four layer board though (at least)
[22:08] <Laurenceb> I have a smaller one, with QFN atmega8
[22:08] <Randomskk> they seem to do it with almost no parts at all
[22:08] <Randomskk> blimey. mine must be out of date
[22:08] <Randomskk> but even so, check out how few parts that thing uses
[22:09] <Laurenceb> someone on rcgroups has a schematic
[22:09] <Randomskk> and the vregs are only because it self powers from the battery, too, mine draws power from the computer
[22:09] <Laurenceb> thats what they used to write the open firmware for it
[22:09] <Laurenceb> and add I2C support
[22:10] <Laurenceb> tower pro ship it with the fuses unset so you can reflash the micro
[22:12] <Randomskk> I won't deny it's a fun, easy and somewhat tempting option
[22:12] <Randomskk> but still :P
[22:12] <Randomskk> I set out to make a motor controller, and that's what I'm doing
[22:18] Nick change: alikins_away -> alikins
[22:18] <Laurenceb> muha got the DC block caps to fit
[22:18] <Randomskk> nice
[22:18] <Laurenceb> - swapped the RF switch for an SC70 package
[22:18] <Randomskk> nice
[22:19] <Randomskk> sc70 is tiny
[22:19] <Laurenceb> that everything done apart from adding a backup reg..
[22:19] <Randomskk> ugh. I wish they actually taught us something that would have been useful to this in first year
[22:19] <Laurenceb> not sure if i want to bother
[22:19] <Randomskk> backup reg in case what? the first fails?
[22:19] <Randomskk> reg as in voltage regulator I take it?
[22:19] <Laurenceb> battery backup
[22:19] <Laurenceb> yes
[22:20] <Laurenceb> so you turn the board off, but the power connector is left plugged in
[22:20] <Laurenceb> so my idea is stick a low quintesscent curretn 3v reg on in place of a cr1216 cell
[22:21] <Laurenceb> and you have a lighter simpler backup for the GPS and stm32 RTC if you want to use it
[22:21] <Randomskk> neat
[22:22] <Laurenceb> problem is you might want to swap over the main lipo cell or whatever is powering the board, but its probably a useful feature
[22:22] <Randomskk> hmm. there are to220 mosfets at the 26p price point (but no longer stocked, just what's left in inventory)
[22:22] <Randomskk> yea, you don't tend to swap the main lipo that much. you might disconnect to charge though?
[22:22] <Randomskk> even so
[22:22] <Randomskk> just having it save gps when not powered on but still connected to main battery seems worthwhile
[22:23] <Laurenceb> RC lipo batteries usually have a seperate charging connector
[22:24] Action: Laurenceb wonders if you could make lipo cells blow up well enough to power a rocket
[22:24] <Randomskk> really? all mine have a separate connector for getting at the cell voltages while charging, but only one main connector
[22:24] <Laurenceb> you charge through that connector
[22:24] <Laurenceb> if you are using a proper charger
[22:24] <Randomskk> then why does my charger require connecting both? >.>
[22:25] <Randomskk> (or maybe it doesn't)
[22:25] <Randomskk> (maybe the main one is only for if it didn't have the little one)
[22:25] <Randomskk> (...?)
[22:25] <Randomskk> (have i been doing it wrong all this time?)
[22:25] <Laurenceb> prob as its a hacked charger :P
[22:25] <Randomskk> what? it's a proper charger
[22:25] <Laurenceb> hmm odd
[22:26] <Randomskk> it may well charge just on the little connector
[22:26] <Randomskk> I've never tried
[22:26] <Randomskk> http://www.wonderlandmodels.com/products/thunder-tiger-imax-b6-lipo-balance/
[22:26] <Laurenceb> I use one from hobbyking.com that runs off 12V, then a wall power supply from argos
[22:26] <Randomskk> though I paid a lot less than £70 for it
[22:26] <Randomskk> yea
[22:26] <Randomskk> well in my case an old atx psu but I have also run it off a 12V wall psu
[22:27] <Laurenceb> nice, it probably uses the main connector so it can charge at a high rate
[22:27] <Laurenceb> without risking overheating
[22:27] <Randomskk> ah, quite probably
[22:27] <Randomskk> it does charge at like 3A ish
[22:27] <Laurenceb> mine is just 0.5C
[22:27] <Laurenceb> yeah
[22:27] <Randomskk> oho, a 52p mosfet that'l do 30V 80A
[22:28] <Randomskk> why have I been paying £1 all this time
[22:28] <Randomskk> all I need now is a decent p channel, but they are all sick expensive, :(
[22:28] <Randomskk> you can buy mosfet driver chips that generate the drive voltage entirely internally iirc
[22:28] <Randomskk> but they are ££ and take lots of space
[22:29] <Randomskk> http://uk.farnell.com/stmicroelectronics/stp85n3lh5/mosfet-n-ch-30v-80a-to-220/dp/1752172 actually looks great, 40A continuous, 80A peak, 30V, 4.6mohm Rds, automotive temp range
[22:29] <Randomskk> 10V to turn it fully on isn't much of a problem
[22:29] <Randomskk> gonna get that from a 3S even when almost empty
[22:32] <Randomskk> even at 3V Vgs it'l do close to 80A at 10V Vds, so
[22:32] <Randomskk> now if only I could do an entirely n channel power stage
[22:33] er1k757 (~erik@tornado.beebe.cc) left irc: Ping timeout: 240 seconds
[22:33] er1k757 (~erik@tornado.beebe.cc) joined #highaltitude.
[22:37] <SpeedEvil> Read closely.
[22:37] <SpeedEvil> Is that 3V at the worst case Vgs spread, and temperature?
[22:38] <Randomskk> probably not, but my Vgs is the same as my Vds and both will be at least 10V, below that the thing cuts itself out anyway as the lipos are dangerously low
[22:39] <SpeedEvil> yeah- Vgs(th) varies from 1-2.5, with 1.8 being typ
[22:39] <SpeedEvil> If the typ is the one being graphed - then the '3v' line in that graph is really 3.7
[22:39] <SpeedEvil> Fair enough
[22:40] <Laurenceb> bbl
[22:40] ContraSF (email@89-180-198-16.net.novis.pt) joined #highaltitude.
[22:46] <Randomskk> hmm open-bldc are doing all sorts of smart things it looks like
[22:48] <Randomskk> the mosfet drivers they're using are £2.73 a pop though D:
[22:49] <SpeedEvil> If you're doing a bldc controller, you've probbaly got enough channels that it's worrthwhile doing a third rail.
[22:50] <Randomskk> but how do I generate the third rail?
[22:50] <SpeedEvil> even something ghetto like a 555, cap, and a couple of diodes.
[22:51] <Randomskk> to buck or boost?
[22:51] LazyLeopard (~irc-clien@chocky.demon.co.uk) joined #highaltitude.
[22:51] <SpeedEvil> to double the supply
[22:52] <Randomskk> if I could generate 24V I could drop the P channels, would make a fair few things easier I guess
[22:52] <Randomskk> how do you double the supply without an inductor too?
[22:53] <SpeedEvil> 555. Cap on output.
[22:53] <SpeedEvil> Other side of cap to a diode that keeps the voltage on it >VCC
[22:53] <SpeedEvil> This then goes from 10-20V at whatever freq
[22:53] <SpeedEvil> you then have another cap.
[22:53] <SpeedEvil> As a resovoir
[22:55] <Randomskk> not sure I entirely follow but sounds good, how exactly does the diode keep the voltage >VCC?
[22:58] <Randomskk> (or does the concept have a name I could look up?)
[22:58] <SpeedEvil> yes
[22:59] <Randomskk> charge pump?
[23:00] <SpeedEvil> yes, sory, was getting link
[23:00] <SpeedEvil> http://electroschematics.com/648/555-voltage-doubler/
[23:00] <Randomskk> ah sweet, thanks
[23:00] <SpeedEvil> There are also teeny charge pump chips that may work too
[23:01] <Randomskk> teeny is good, but they also tend to be expensive
[23:01] <Randomskk> is the 555 just to make a pwm?
[23:01] <Randomskk> I could probably have the avr do that
[23:02] <SpeedEvil> 555 also has nice beefy 200mA drivers
[23:02] <Randomskk> guess that helps though I wouldn't expect too much current draw
[23:03] <Randomskk> does mean I don't have to actually do it with the avr though, since that would probably get annoying
[23:03] <Randomskk> but a 555 isn't the smallest chip in the world and it would need a few passives of its own...
[23:03] <Randomskk> could swap to using one of those absolutely tiny ceramic oscillators instead of the massive smt crystal, that'd save some space
[23:06] <SpeedEvil> you get SMT passives
[23:07] <SpeedEvil> SMT 555s even
[23:07] <SpeedEvil> And also charge pump ICs
[23:08] <Randomskk> yea, but even the smt 555s are like soic7
[23:08] <stilldavid> just got an email with this link from a customer:
[23:08] <stilldavid> http://www.youtube.com/watch?v=nSyWDqgNRmo
[23:08] <Randomskk> s/7/8
[23:08] <Randomskk> watching now :D
[23:09] <Randomskk> man, I wish we could just 'drive to nevada'
[23:10] Upu (~Upu@ubn.upuaut.net) left irc:
[23:11] <Randomskk> hah the crator for that rocket with the failed parachute
[23:11] <SpeedEvil> It is also possible to bootstrap.
[23:12] <Randomskk> bootstrap in this context? I've seen it referred to on mosfet drivers too
[23:12] <SpeedEvil> For example - you instead of using the 555 to generate the initial voltage, use one phase of the driver.
[23:12] <SpeedEvil> This gets you an initial voltage of 0v-vcc-3v
[23:12] <SpeedEvil> well - minus a couple of diode drops
[23:13] <SpeedEvil> but then this rapidly rises, and you activate the other couple of phases.
[23:13] <SpeedEvil> Where activate = use them as bridge drivers, not just all at the same trime
[23:14] <SpeedEvil> Problem with this is that spikes can make the voltage too high - so you've got to care about EMF
[23:14] <Randomskk> so you don't need the 555 or any pwm at all?
[23:14] <stilldavid> imagine going to recover your payload and having to bring a shovel :-/
[23:14] <SpeedEvil> but the software gets more complex, and you can lose the rail if things go wrong.
[23:15] <Randomskk> which wouldn't be great
[23:16] <Randomskk> hmm
[23:16] <SpeedEvil> You're turning on (well - differentially) at full power for just long enough to keep the motor 'locked' ?
[23:16] <SpeedEvil> Or PWMing a sine or ...
[23:17] <Randomskk> nothing concrete, but the plan was to just PWM one half of the bridge at various on:off ratio but high frequency relative to the commutation frequency
[23:17] <Randomskk> stilldavid: awesome video
[23:18] <SpeedEvil> Proportionally to lag?
[23:18] <stilldavid> Randomskk, the guy who sent it is pretty high up and asked if anyone here wanted to collab with the rocket dude to get some stuff into orbit
[23:19] <Randomskk> they get stuff into orbit?
[23:19] <Randomskk> SpeedEvil: sorry? I don't understand
[23:19] <stilldavid> no, but they're working on it
[23:20] <SpeedEvil> Randomskk: You want to have the current through the motor so that it is large enough, but not much larger than the current required to keep the motor turning with the given load.
[23:20] <stilldavid> "Next up we're aiming to go into orbit. If folks at Sparkfun have any interest in being part of the project, just let me know"
[23:20] <SpeedEvil> As a stupid example - drive it at 60RPM with no load, and it will fly.
[23:20] <SpeedEvil> Fry
[23:21] <SpeedEvil> You need some way to control the current - one way is to measure the difference between the commanded and target RPM
[23:22] <Randomskk> how does that work out with propellors for load? there's very little resistance to rotation
[23:24] <fsphil> nice video
[23:26] <fsphil> 200,000 feet -- they can actually get to 60km with these things?
[23:28] <SpeedEvil> Randomskk: You want also a ramp/up/down limit
[23:29] <SpeedEvil> Randomskk: So you might have a commutation system. This is the lowest level - when it detects something happening (be that coil voltages or halls), it switches to next phase.
[23:29] <Randomskk> yup
[23:29] <SpeedEvil> Randomskk: You then have a system that looks at desired and current speed.
[23:29] <SpeedEvil> If desired speed is under required speed, you give the motor more current.
[23:30] <SpeedEvil> But the rate of rise of the current is set by the ramp/up/down limits
[23:30] <Randomskk> yup
[23:30] <Randomskk> and you control the current by the amount of on time to off time in the PWM?
[23:30] <SpeedEvil> You can do it two ways.
[23:31] RocketBoy (steverand@217.47.75.8) left #highaltitude.
[23:31] <SpeedEvil> Either by only activating the winding for a very short time at the 'peak'
[23:31] <SpeedEvil> Or PWMing
[23:31] <SpeedEvil> PWMing will give you better battery life, and lower heat
[23:31] <Randomskk> which would you suggest? I've considered both, not sure which is better
[23:31] <Randomskk> right
[23:31] <Randomskk> but...
[23:32] <m1x10> stilldavid, I can report that the gsm shield works very well until now. It does fully cooperate.
[23:32] <Randomskk> actually, why either? surely you get less switching loss as the mosfets turn on/off by only turning on once per commutation?
[23:33] <Randomskk> hah, I get how the charge pump works now. neat.
[23:34] <SpeedEvil> Randomskk: yes, you get less switching loss, but.
[23:35] <SpeedEvil> The motor is going 100RPM, it's relatively unloaded, and its terminal voltage is 1V - internally generated.
[23:35] <SpeedEvil> It would need 3V to keep it going at this load.
[23:35] <SpeedEvil> If you apply 10V, 7V/resistance extra current flows, heating things up.
[23:35] <SpeedEvil> And wasting battery
[23:35] <Randomskk> right
[23:36] <Randomskk> while PWMing...
[23:36] <SpeedEvil> Hmm.
[23:36] <SpeedEvil> I'm not sure I haven't confused myself.
[23:37] <SpeedEvil> On second analysis, I think that - modulo Rds(on) losses in the FETs and losses magnetically in the motor, it should be almost as efficient to reduce the on-time of the drive waveform, as to PWM it.
[23:37] <Randomskk> I wonder which would work out better. the one big on peak could probably be done with no PWM at all
[23:37] <Randomskk> so I could just use digital I/O
[23:37] <Randomskk> leaving the PWM free for a charge pump, say
[23:38] <Randomskk> so I get an all n-channel power stage
[23:38] <Randomskk> dropping Rds(on) losses since Ps are worse, and reducing cost
[23:38] <Randomskk> also meaning I can ADC really fast without worrying about PWM when not on
[23:38] <Randomskk> otherwise I have to time the ADC read of the phase voltage to when the PWM's off
[23:38] <SpeedEvil> Argh.
[23:39] <SpeedEvil> I forgot the ^2 in P=I^2R
[23:39] <SpeedEvil> PWM is better
[23:39] <Randomskk> ... :( aww
[23:39] <Randomskk> why don't you suffer the same losses with PWM, plus switching losses?
[23:39] <SpeedEvil> because the windings are inductive.
[23:40] <SpeedEvil> The PWM frequency is sufficiently high that the windings largely average the applied current.
[23:41] <Randomskk> right, okay
[23:41] <fsphil> rockoon launch: http://www.youtube.com/watch?v=eFp2Gg3cyTo
[23:42] <Randomskk> while doing it with just one big peak means magnetic losses as the current rises?
[23:43] <Randomskk> wait, I mean
[23:44] <Randomskk> hm
[23:45] <SpeedEvil> If the above case.
[23:45] <SpeedEvil> you have 7V between the desired voltage and the battery voltage.
[23:45] <SpeedEvil> You have 1V EMF
[23:45] <SpeedEvil> and need to apply 3V to keep it rotating
[23:46] <Randomskk> so the rest of the voltage is burnt off as I²R?
[23:46] <SpeedEvil> If you apply 10V in one pulse, you get 9V extra after the EMF, or 4.5* the current of the former case.
[23:46] <SpeedEvil> yes
[23:46] <Randomskk> while when PWMing, the effective current due to the PWM is averaged so there's no long the voltage gap?
[23:46] <SpeedEvil> yes, exactly
[23:47] Laurenceb (~laurence@host86-145-126-39.range86-145.btcentralplus.com) left irc: Ping timeout: 252 seconds
[23:47] <Randomskk> effectively the PWMing drops the time average to that ~3V needed and not much more
[23:47] <Randomskk> ?
[23:47] <SpeedEvil> exactly.
[23:47] <Randomskk> right
[23:47] <SpeedEvil> If you look at the instantaneous current, it's a sawtooth.
[23:47] <SpeedEvil> At the PWM freq
[23:47] <Randomskk> but the one big spike per commutation doesn't average down with the rest of the downtime?
[23:48] <Randomskk> it'd be such a lower frequency
[23:48] <Randomskk> even at about 10kRPM or so
[23:48] <SpeedEvil> No - to put it in a switched mode analogy - which isn't inaccurate.
[23:48] <Randomskk> 6 commutations per cycle
[23:48] <Randomskk> 60kHz ish
[23:48] <SpeedEvil> A buck converter might work well with an inductor at 100KHz.
[23:49] <SpeedEvil> But if you reduce the operarting frequency down to 1KHz, it will saturate and just act like a resistor
[23:49] <Randomskk> yup
[23:49] <Randomskk> but even one big spike per commutation is going to be about 60kHz
[23:50] <Randomskk> admittedly that's at full power
[23:50] <SpeedEvil> That limits the RPM somewhat.
[23:50] <Randomskk> which is not going to happen very often
[23:50] <Randomskk> if ever
[23:50] <SpeedEvil> At full power, the inductance becomes important.
[23:50] <SpeedEvil> At full revs, rrther
[23:50] <SpeedEvil> err
[23:50] <SpeedEvil> also
[23:50] <SpeedEvil> 10krpm/60*6
[23:51] <Randomskk> oh, yes.
[23:51] <Randomskk> well
[23:51] <Randomskk> 10krpm*6 / 60 surely?
[23:51] <SpeedEvil> yes
[23:51] <Randomskk> I guess your expression is the same given usual order of evaluation
[23:51] <Randomskk> fair point then
[23:51] m1x10 (m1x10@ppp089210143131.dsl.hol.gr) left irc: Quit: I eat electrons for breakfast and I vomit thunders
[23:51] <Randomskk> so, PWM it is
[23:52] <SpeedEvil> Naah - nothing so logical - I'm justr rather tired ATM, and not up to maths.
[23:53] <SpeedEvil> However!
[23:53] <SpeedEvil> :)
[23:53] <SpeedEvil> With a prop, this can be less important.
[23:53] <SpeedEvil> It's only really bad at small RPMs and appreciable torques.
[23:53] <SpeedEvil> But torque on a prop is rpm^3
[23:54] <SpeedEvil> So, it's merely bad efficiency, not utterly terrible.
[23:54] <Randomskk> hmm
[23:54] <Randomskk> yay, even harder tradeoffs
[23:54] <Randomskk> at the moment I have it able to control the top half of each H bridge with PWM
[23:54] <Randomskk> so, fine
[23:54] <Randomskk> albeit 8bit PWM for one of them
[23:55] <Randomskk> but that means not being able to use a PWM for a charge pump, which would be much nicer than using a 555 in terms of space, cost, etc
[23:56] <Randomskk> there are some cute little integrated step up ICs
[23:56] <Randomskk> costly though
[23:58] SpeedEvil (1000@tor/regular/SpeedEvil) left irc: Ping timeout: 255 seconds
[23:59] n900evil (~Speedevil@tor/regular/SpeedEvil) joined #highaltitude.
[23:59] <n900evil> stupid dsl
[23:59] <Randomskk> ooh. atmel do some of their more awesome atmegas in DRQFN
[23:59] <Randomskk> no doubt the biggest pita to solder at home, but tinier
[00:00] --- Wed Aug 18 2010