93 posts • joined 7 Dec 2011
Re: Chinese lanterns
Well, to put that in perspective, Leo flies 1 balloon every week or so whilst the UK met office send up about 2000 every year.
Re: BBC at it again
I've almost given up trying to train media types to actually get facts right - e.g. no it's not "space" no the balloon doesn't explode no it doesn't send the position down via GPS satellites etc etc.
Of the many journos I've dealt with, Lester is one of the small percentage who actually checks his facts.
Good on the school for a) doing it and b) actually recovering the flight. I've seen a few flights lately where people (not schools) completely failed at (b).
Re: You know.... an idle thought occured...
You could, but the end result is the same as putting less gas in the balloon in the first place.
Re: Is that the 'globe' of helium that remains for a few frames?
It is talc.
There was one pointing one way and 2 the other way, as can be seen in one of the stills. I believe that thrusts were, approximately, balanced.
Shax believes one of the motors went pop, and I'm inclined to agree.
Re: 808 is not in the same league
I'd not use an 808 for any near-space mission, as the image quality is poor and their EMI is worse.
There's no cost saving if a cheap device like this ruins the mission by stopping GPS recovers from working.
Re: Too expensive
And wrap them in foil first - the 808 cameras are better at disrupting GPS reception than they are at recording images.
Re: Julie / Heston
You're not the first to notice this!
He's actually holding a piece of bacon. It would have been a bacon muffin (standard HAB fayre of course) but he arrived an hour late and all he wanted was the cold bacon.
Jokes about bacon and baps have also already been made ...
We were going to use mine for the LOHAN test flights last summer. It failed horribly before we launched - wouldn't even start up, even with known fully charged battery. It was weeks later when I got it to work again, following a guide on the internet about 20 times till eventually it booted for long enough to load updated firmware. Since then it's been fine, but I can't ever trust it again.
We believe that it actually blew back in the last test in Spain, but the glass wasn't blackened (well not much - just a tad at the underside) so it wasn't obvious. Igniters can short circuit when the blow, plus the battery supply erred on the "let's make sure this goes" side, resulting in an igniter current about 3 times the fuse rating.
So, for next time, "check the damn fuse" gets added to the check list. Also, the SPEARS firmware has been updated to check the fuse and other items so it will report over the radio if there's an issue, and the receiving end will display a "DON'T LAUNCH" message in such a case.
Not many doing the "live" pix. Only a couple doing it regularly and a few more doing it sometimes.
You have been watching someone else's flight. Our imaging tracker was "RICHARD" and as no doubt noticed the images are from one called "POPEYE".
Re: Oxygen supply?
I'm sure Lester has plans to add an oxygen tank, complete with a stirrer ...
Re: Rubber band problem?
True, but it'll be nice and warm in there what with the Pi etc.
Actually, it was 1605.
Re: The title is too long.
The camera already draws approx 200mA at 3.3V so it heats itself, so if I were worried (which I'm not) I could just leave the camera on all the time.
I've flown these cameras with very little insulation, and sometimes with none. They've never fogged. It's not a problem and doesn't need a solution.
Re: Too late but still
No, Lester got it right. For high altitude work, windows tend to fog up inside. I know of several HAB flights where people have used windows, lenses or Go Pro covers and without exception unless they're heated they fog up for part of the flight.
Re: We assume it has been updated
The problem was actually that the met office didn't launch their normal quota of balloons that day, so the prediction was based on older data than usual. A weather front was predicted to come in during the evening, but actually came in several hours earlier, changing the flight path enormously and thus dooming our heroic playmonaut to an early, salty, wet grave.
Yes, all true. Graphics use the OpenVG C library so no X at all. Total CPU for the whole thing (GPS receiver, video streaming, chase car upload, display update) runs at around 10%; without the h/w acceleration it'd be 100%!
Re: Airframe comparison
> .. idiot(s) ... military grade
Coincidentally, those are the 3 words that came to mind when reading your post. Though not necessarily in the same order.
Re: From the RockBLOCK spec....
SBD was extensively tested in our workshop, which thankfully was large enough for the "messages" to dissipate.
Re: smart phone
A reader suggested a Bond-style trigger box, so the RPi was an obvious choice!
The message can't be sent using SMS, but is sent by POSTing a URL to the RockBLOCK server. So yes it could have been done using a simple program on a smartphone or a PC (in fact I tested it all on a PC before porting the code to the Pi).
We wanted to put the full week's worth of empties in front of the camera, but decided that the weight was far too much, and the camera would have had to be the other side of the street.
Re: What are the various circles on the map?
We had 2 transmitters in each payload (1 main 1 spare) but only 2 receivers on the ground. The backup trackers were ignored during the flight hence appeared to stay at the launch site.
All images are stored on the SD card, which we recovered after the mountain rescue expedition the following day.
That really is a most appalling piece of deduction. They cannot rule out contamination and they appear to have done very little to prevent it.
Re: From the photo showing the battery pack...
Er, yes, that's how I switch it off you see :-)
I contacted them before when I got the "highest images transmitted by an amateur over radio", but they turned me down. So .... stuff 'em.
I know! Sadly I didn't think of that till after I'd done the deed.
It cuts the payloads down from the balloon. I usually wrap the nylon cord around a resistor and heat that up with a few watts. I suspect this time though pyros will be used.
The cut is above the parachute for what I hope are obvious reasons :-)
Hand Operated Remote Release Over Rockblock
Indeed. It was hardly worth having a custom PCB made for a one-off, and it's a lot quicker too.
Re: LOHAN shouldn't suck wet caps!
Yeah, I remember at school accidentally blowing up a 5000uF cap. Which I was holding at the time. Fortunately the top of the cap was facing away from my hand, otherwise A&E would have been required! It took some time for the cloud of white dust to settle.
Re: LOHAN shouldn't suck wet caps!
It's a super-cap. It's there so the device can be run from USB. When transmitting it uses a lot more than the 500mA that USB can manage.
In our case power comes from some Lithium Energizer cells so the supercap isn't needed.
Re: Most importantly...
I like this :-)
Re: Live piccies!
Well, we already have the option of sending images over the radio link, and even though we won't have so many radio listeners as usual that will probably still be the best option.
A foam box landing by parachute at 10mph is going to have trouble killing anyone. Even an American-sized foam box like that one. In the UK our flights are typically much, much smaller.
In the UK there are around 50 amateur flights per year listed on the UKHAS site, and maybe the same again that don't appear there. The UK met office fly about 2000 per year, and they haven't managed to kill anyone either (worst they managed was breaking a greenhouse window).
Re: accelerometer readings from the 3-axis magnetometer
Those are from the accelerometer. There was also a megnetometer and gyro, but not shown in the article.
Re: Vaguely OT: Dara O'Briains science club ..
No chainsaw yet, however the tree in question proved no match for the muscle-power and a rather effective handsaw.
Re: Are you saying...
Anything dropped from weather balloon *must* descend by parachute.
No gliding, guided or otherwise.
Re: Very stable
Just to confirm my earlier suspicion, the glitchy video is down to a sub-par SD card.
Re: More helium wasted
Yes, it was hydrogen. More lift less money and as a bonus stops Lester from lighting up too often :-)
The plan was to release the glider before burst, with a tethered line to pull a large 'chute out of the back of the fuselage. The small chute was to help make sure that happened. We had to ensure that the glider fell by parachute and didn't actually fly.
In the end the release mechanism didn't release the glider, so the whole lot came down together. 1.3kg of balloon came down with it to, so the tree-landing was a tad faster than we aimed for. Not that we aimed for a tree either :-)
It had to have a chute - CAA won't allow flight from altitude whether the glider is guided or not.
The James May thing had to stay within sight of an operator at all times.
yes, some (not many) GPS units correctly operate this way, but many stop if the velocity OR altitude limit is exceeded. Generally we use ublox modules which are known to work at high ltitudes (up to 50km which is higher than we can get our balloons), provided that we put them into "flight mode" first.
Re: Very stable
Yes the Earth concave/convex thing is down to the lens on the GoPro Hero 3 that took that video, popular for their ability to cram everything into the field of view. The downside of course is the enormous distortion, which is kind of OK in some stills but hideous when recording video of what should be a slightly curved Earth.
As for the Pi, the video part was a late addition. A script took small images during the launch phase, then larger images during flight (from 3km up, then down to 2km), and finally it ran a single 10 minute video for the landing before reverting to the small images again. The video part was the standard current Pi Cam software with no special parameters. Guess #1 is the SD card wasn't fast enough. Guess #2 is the processor was a tad busy (but was generally around 20% on the telemetry etc., which I'm guessing isn't enough to cause a problem). The jumps were visible when playing the video back on a Windows PC, as well as in the converted-to-mp4 version uploaded to Youtube.