highaltitude.log.20070406

[00:01] Laurenceb (n=laurence@host86-136-160-29.range86-136.btcentralplus.com) left irc:
[01:16] flowolf (n=flowolf@host133-220-dynamic.21-87-r.retail.telecomitalia.it) left irc: "Leaving"
[02:03] icez (n=icez@ip68-3-56-121.ph.ph.cox.net) joined #highaltitude.
[03:11] Superkuh (n=synthase@c-24-245-49-197.hsd1.mn.comcast.net) left irc: Read error: 54 (Connection reset by peer)
[07:15] icez (n=icez@ip68-3-56-121.ph.ph.cox.net) left irc: Remote closed the connection
[08:11] jcoxon (n=jcoxon@81-178-247-220.dsl.pipex.com) joined #highaltitude.
[09:30] flowolf (n=flowolf@host133-220-dynamic.21-87-r.retail.telecomitalia.it) joined #highaltitude.
[09:37] flowolf (n=flowolf@host133-220-dynamic.21-87-r.retail.telecomitalia.it) left irc: "Leaving"
[09:57] jcoxon (n=jcoxon@81-178-247-220.dsl.pipex.com) left irc: "Leaving"
[10:36] flowolf (n=flowolf@host133-220-dynamic.21-87-r.retail.telecomitalia.it) joined #highaltitude.
[12:07] jcoxon (n=jcoxon@81-178-247-220.dsl.pipex.com) joined #highaltitude.
[12:24] flowolf (n=flowolf@host133-220-dynamic.21-87-r.retail.telecomitalia.it) left irc: "Leaving"
[13:27] jcoxon (n=jcoxon@81-178-247-220.dsl.pipex.com) left irc: "This computer has gone to sleep"
[14:02] flowolf (n=flowolf@host133-220-dynamic.21-87-r.retail.telecomitalia.it) joined #highaltitude.
[14:49] jcoxon (n=jcoxon@81-178-247-220.dsl.pipex.com) joined #highaltitude.
[15:02] icez (n=icez@ip68-3-56-121.ph.ph.cox.net) joined #highaltitude.
[15:03] icez_ (n=icez@ip68-3-56-121.ph.ph.cox.net) joined #highaltitude.
[15:06] Ebola (n=Ebola@host86-136-130-202.range86-136.btcentralplus.com) joined #highaltitude.
[15:08] jcoxon (n=jcoxon@81-178-247-220.dsl.pipex.com) got netsplit.
[15:08] jcoxon (n=jcoxon@81-178-247-220.dsl.pipex.com) returned to #highaltitude.
[15:09] icez (n=icez@ip68-3-56-121.ph.ph.cox.net) left irc: Read error: 104 (Connection reset by peer)
[15:38] Ebola (n=Ebola@host86-136-130-202.range86-136.btcentralplus.com) left irc: "the glas is half empty."
[15:45] Professor (n=bobd@x-division.com) joined #highaltitude.
[15:45] <Professor> hello all
[15:46] Professor (n=bobd@x-division.com) left irc: SendQ exceeded
[17:11] Laurenceb (n=Laurence@host81-157-248-159.range81-157.btcentralplus.com) joined #highaltitude.
[17:51] Ebola (n=Ebola@host86-136-130-202.range86-136.btcentralplus.com) joined #highaltitude.
[18:14] flowolf (n=flowolf@host133-220-dynamic.21-87-r.retail.telecomitalia.it) left irc: Read error: 104 (Connection reset by peer)
[18:14] Laurenceb (n=Laurence@host81-157-248-159.range81-157.btcentralplus.com) left irc:
[18:33] flowolf (n=flowolf@host133-220-dynamic.21-87-r.retail.telecomitalia.it) joined #highaltitude.
[21:36] rocketboy (n=steve@host86-132-36-253.range86-132.btcentralplus.com) joined #highaltitude.
[21:37] snoopl3s (n=Snoopl3s@209.105.206.140) left irc: Remote closed the connection
[21:38] snoopl3s_ (n=Snoopl3s@209.105.206.140) left irc: Remote closed the connection
[21:38] <jcoxon> evening all
[21:38] <rocketboy> hi jc - any video from N5 yet?
[21:38] <jcoxon> no nothing
[21:39] snoopl3s (n=Snoopl3s@209.105.206.140) joined #highaltitude.
[21:40] snoopl3s (n=Snoopl3s@209.105.206.140) left irc: Remote closed the connection
[21:43] <rocketboy> N5 did land near bedford - But I don't know how it got there all the winds were going the other way
[21:43] <jcoxon> rocket :-D
[21:43] <jcoxon> i've put togther that list for you - of launches
[21:44] <rocketboy> ah spiffing
[21:46] <rocketboy> thanks - i'll add some stuff to it
[21:46] <icez_> the magic of ballooning, you'll never know where it ends up :D
[21:51] <jcoxon> grr - i need one more connector to crimp
[21:51] <jcoxon> thought i'd worked it out exactly
[21:52] <rocketboy> the magic of electronics, .....
[21:55] <jcoxon> damn need for ground
[21:58] <jcoxon> http://www.pegasushabproject.org.uk/parawing/P1010052.MOV
[22:00] <jcoxon> thats the only video i've got from testing
[22:00] <jcoxon> its of the first drop
[22:01] <icez_> cool
[22:01] <rocketboy> yeah - its a shame we didn't video them all - especially the last 2
[22:03] Ebola (n=Ebola@host86-136-130-202.range86-136.btcentralplus.com) left irc: ""Don't let worry kill you -- let the church help.""
[22:33] <jcoxon> rocketboy, what was on XABEN 2?
[22:34] Ebola (n=Ebola@host86-136-130-202.range86-136.btcentralplus.com) joined #highaltitude.
[22:34] <rocketboy> sort of - there was a "payload" - that ended up in banbury
[22:34] <jcoxon> right
[22:34] <rocketboy> its comming back in the post (hopfully)
[22:39] <rocketboy> any idea what sort of accuracy you can get from a typical RC servo - I was thinking of usinf 8 bit acuuract for full end to end movement - thats about 0.5 degree accuracy
[22:40] <rocketboy> usinf = using
[22:40] <jcoxon> no idea to tell the truth
[22:40] <jcoxon> ed will know
[22:40] <jcoxon> he did lots of servo stuff in his gap year for his 6 legged walker robot
[22:41] <rocketboy> ah ok - ill drop him a line
[22:42] Laurenceb (n=laurence@host86-141-238-140.range86-141.btcentralplus.com) joined #highaltitude.
[22:43] Ebola (n=Ebola@host86-136-130-202.range86-136.btcentralplus.com) left irc: ""Don't let worry kill you -- let the church help.""
[22:43] <Laurenceb> Hi all
[22:43] <Laurenceb> I'm working on getting avr-dos working
[22:43] <rocketboy> lo lb
[22:44] <Laurenceb> havent got a sd card connector so I've soldered wire onto the card
[22:45] <Laurenceb> I was talking to some people over at #electronics and they suggested a 3 axis accelerometer, 3 axis gyro and 3 axis magnetometer
[22:45] <Laurenceb> and a kalman filter, very pricy
[22:46] <rocketboy> tathat would do it
[22:46] <rocketboy> lb any idea what sort of accuracy you can get from a typical RC servo - I was thinking of usinf 8 bit acuuract for full end to end movement - thats about 0.5 degree accuracy
[22:46] <Laurenceb> I think its excessive, especially for something thats stable like a parawing
[22:47] Ebola (n=Ebola@host86-136-130-202.range86-136.btcentralplus.com) joined #highaltitude.
[22:47] <Laurenceb> hmm, yes Id say thats the best youd get with a good servo
[22:47] <rocketboy> what accuracy did you use
[22:47] <Laurenceb> I use 16 bit timer
[22:48] <Laurenceb> 20ms is 40,000
[22:48] <Laurenceb> so 1ms=180 degrees is 2000
[22:49] <rocketboy> ta
[22:49] <Laurenceb> there is a null range in the middle of the servo, poor quality servos have +-5degrees
[22:49] <Laurenceb> very good ones +-2 degrees
[22:49] <Laurenceb> digital servos <+-1 degree
[22:50] <icez_> somewhat related to high-altitude balloons: http://imgs.xkcd.com/comics/blagofaire.png
[22:50] <rocketboy> OK - I'll go for 8 bit
[22:50] <Laurenceb> I know a guy who runs a model shop in oxford, he also makes uav helicopters, he gave me a demostration of different servos
[22:52] <jcoxon> icez_, hehe
[22:52] <jcoxon> i saw that on boingboing
[22:53] <jcoxon> but it only had one frame
[22:53] <jcoxon> it makes a lot more sense now
[22:53] <Laurenceb> he was making a uav helicopter with video onboard for shipment to the russian embassy
[22:53] <Laurenceb> rocketboy, how do ou plan to make servos pulses? hardware pwm?
[23:03] <Laurenceb> I wasnt using hardware pwm, but it looks promising as it avoids interrupts. Only problem with hardware pwm seems to be that the pulse repetition rate is fixed
[23:03] <jcoxon> night all
[23:03] jcoxon (n=jcoxon@81-178-247-220.dsl.pipex.com) left irc: "Leaving"
[23:03] <rocketboy> actually I was thinking of a dedicated PIC and doing it all in software - that way I can do
[23:04] <rocketboy> about 10 chammels at once
[23:04] <rocketboy> chammels = channels
[23:04] <rocketboy> I'm going to use a serial interface at 9600 baud
[23:05] <Laurenceb> hmm how will you do it in sw? a timer interrupt that happens very often
[23:05] <Laurenceb> and then decides which servo outputs to turn on/off?
[23:06] <rocketboy> yeah a 1ms interrupt - and than a simple instruction counter to iplemnet the PWM
[23:07] <rocketboy> since the pulses are 1 to 2ms every 20ms I can get 10 lots in a 20ms period
[23:07] <Laurenceb> oh so servo pulses come one after another?
[23:07] <rocketboy> actually I could do 20 channels - but I run out of pins on the PIC
[23:07] <rocketboy> yep - why not
[23:07] <Laurenceb> clever
[23:08] <rocketboy> its just a stagger of a few ms
[23:08] <rocketboy> - doubt that it will make any appreciable difference to response
[23:08] <Laurenceb> I'm wondering what would be best for controlling a parawing, 5hz gps or 1hz gps+rate gyro
[23:09] mc- (n=mfcastle@cpc2-glfd1-0-0-cust995.glfd.cable.ntl.com) joined #highaltitude.
[23:09] <Laurenceb> Hi mc
[23:09] <rocketboy> the latter I imagine
[23:10] <Laurenceb> it would need temperature calibration
[23:10] <mc-> hi all
[23:10] <rocketboy> does tempreture affect gain or drift?
[23:10] <mc-> rocketboy, I've got some servo code if that helps you.
[23:10] <rocketboy> hiya mc
[23:10] <Laurenceb> I dont know how linear rate gyros are, ie in their temperature response
[23:11] <rocketboy> that would be interesting mc
[23:11] <Laurenceb> well, emperature effects gain and drift
[23:11] <Laurenceb> from the analogue devisces datasheet it appears to be a linear relationship for both
[23:12] <mc-> some hints for servos, it doesn't matter what the repetition rate is, the pulse width matters.
[23:12] <Laurenceb> hopefully I could do a calibration by measuring the response at two different temperatures
[23:12] <mc-> I did the pulse width, by a timer in C/assembler, no need for interrupts
[23:13] <Laurenceb> you mean hardware pwm?
[23:13] <mc-> I did a s/w pwm on a pic
[23:13] <Laurenceb> with my avr setup, I can only get 33ms repetition
[23:14] <Laurenceb> using hardware pwm
[23:14] <mc-> mine works on a 12C508
[23:14] <mc-> 33ms is fine
[23:15] <Laurenceb> from what I've observed of 4hz gps's, they are not as good as they might appear, as the response is still slow
[23:16] <mc-> all gps will have some delayed response because of the kalman filter
[23:16] <Laurenceb> I was reading the antaris datasheet, and it seems the output is delayed due to the kalman filters
[23:16] <Laurenceb> yes
[23:17] <Laurenceb> the sirf2 actually seemd to be better than the antaris in that respect
[23:17] <Laurenceb> even though the antaris is 4hz
[23:18] <mc-> yes, ublox do tons of filtering to get a smooth output
[23:18] <Laurenceb> my antaris cant seem to get a lock now
[23:18] <Laurenceb> not sure if I broke it
[23:19] <Laurenceb> I shortened the antenna, and it appeared to be bad at getting a lock, so I reflashed it with the latest firmware, and it worked, but not any more
[23:20] <mc-> try lenthening the antenna again
[23:20] <Laurenceb> but that makes a mess in the payload area:( how does that change things?
[23:21] <Laurenceb> seem to remember that was a problem with you r payload
[23:22] <mc-> yes, I think so
[23:22] <Laurenceb> what difference does it make?
[23:22] <mc-> just emailed you both my servo code
[23:22] <Laurenceb> both = me and rocketboy?
[23:23] <mc-> yes
[23:23] <Laurenceb> thanx, will check it out
[23:23] <mc-> I don't really know, longer antenna seems to stop input saturating.
[23:24] <mc-> got on a tip on the AD gyros, glue them down upside down and wire to the pads
[23:25] <Laurenceb> It didnt make any difference on a sirf2
[23:25] <Laurenceb> the pads are really small
[23:26] <rocketboy> ta just looking at the code now
[23:26] <rocketboy> humm - minimal
[23:27] <Laurenceb> If I fixed the AD servo to the top of a servo, it would be possible to calibrate it nicely
[23:28] <mc-> AD=analog devices...
[23:28] <rocketboy> so you would go for 0.5ms to 2.5ms? rather than 1.0 to 2.0 ?
[23:28] <Laurenceb> I have a broken minifridge, could make a gyro cooler with the thermoelectric cooling things
[23:28] <Laurenceb> sorry, AD servo=AD gyro
[23:29] <mc-> yes, I found a servo I tried was 0.5 to 2.5ms
[23:29] <rocketboy> Peltier effect device?
[23:29] <Laurenceb> yes
[23:29] <Laurenceb> think the power supply got wet and fried, peltier cooler should still work
[23:30] <mc-> how about putting in a heater in the payload rather than cooler?
[23:30] <Laurenceb> I had it on my balcony at college
[23:30] <rocketboy> Ah OK - 20ms / 2.5ms allows me to do 8 channels - cool
[23:30] <Laurenceb> very power hungry, was thinking of using the cooler just for calibration
[23:31] <Laurenceb> if the temperature drift is linear, the datasheet doesnt make it clear, will email AD and ask
[23:31] <rocketboy> still gives better than 1 degree resolution
[23:31] <rocketboy> with 8 bits
[23:31] <Laurenceb> the gyro has a temp sensor built in, so guess thats how your meant to use it
[23:32] <mc-> I've heard the AD gyro drifts more because of vibration in other planes, than because of temperature
[23:32] <rocketboy> bbl
[23:33] <mc-> I'm off as well, night
[23:33] mc- (n=mfcastle@cpc2-glfd1-0-0-cust995.glfd.cable.ntl.com) left irc:
[23:33] <Laurenceb> looks quite vibration resistant according to datasheet
[23:34] <Laurenceb> cya all then
[23:35] Laurenceb (n=laurence@host86-141-238-140.range86-141.btcentralplus.com) left irc:
[23:55] Ebola (n=Ebola@host86-136-130-202.range86-136.btcentralplus.com) left irc: ""Don't let worry kill you -- let the church help.""
[00:00] --- Sat Apr 7 2007