I would seriously look at the design of Huygens as the best analog for this project. It sounds like the camera system is designed to be akin to the Huygens camera system, with three cameras imaging from nadir to horizon. For mosaicking purposes, look at the design of Huygens and how it spun so that full 360 degree mosaics would be possible. of course, make sure it is designed to spin in the correct direction.

I was having similar thoughts, how controlled can we have the rotation (up the hill or down) - because looking at other similar projects, the imagery seems very very chaotic.

Doug

Weight distribution would also be very important for platform stability; you might have to use some soft lead weights to at least get a three-point balance, bottom heavy (thinking of 1/2 oz fishing sinkers...)

Spin's a little harder. I guess some fins to offer wind resistance might help, but wouldn't do much good at peak altitude...

The problem isn't a million miles from MER terminal decent. It's not chute-backshell-lander but it's envelope-chute-gondola.

I'm reminded of this : http://www.jpl.nasa.gov/news/features.cfm?feature=1449

I don't think controlling the rotation is going to be possible, certainly not to the level required to guarantee that we get 100% coverage.

iirc Huygens was designed to spin just to make sure all the images didn't point in the same direction, not to have a controlled sweep to pan the camera. Looking at images from other balloons it doesn't look like forcing the balloon/parachute to rotate will be a necessary, it looks like quite a wild ride.

Lets just concentrate on taking as many shots as possible, it looks to me that we won't have to be too lucky to get good 360 coverage and we have plenty of talent here to fill in any small gaps in the data.

James

Discovered the FM transmitter may be a no-no in the UK

OO - no - 144.800 is good for APRS in the UK Sweet.

Grrrrrrrr

Doug - I wanted to reply to your paragraph about options for future flights but when I click on 'reply' to your post that part seems to be missed off the quote. Anyhow, you know what I'm going to say - "What, no stereo? " I think this is a significant, attainable, genuinely scientific objective given that weather phenomena are complex four-dimensional things. It could have practical applications not just on Titan but for Venus and the giant planets too. I can see why the first flight should be as simple as possible though.

On the cost issue - would there be some way for us 'spectators' to chip in a little, perhaps anonymously?

For stereo to work - you either have to have a very long baseline within one payload ( which is a near impossibility given the limitations of size and mass ) - or two simultaneous balloons pointing in the same place, at the same time, at the same altitude ( a near impossibility given the limitations of chaos).

Clouds are usually fuzzy, roughly defined, dynamic objects - I can't imagine that getting stereo imagery of them would be particularly valuable anyway. They're not like a rock. If someone can explain to me how it could be done and why it should be done, then why not....but I struggle to see a means or a purpose at the moment.

You are right however - the first launch ( and indeed I would have thought the first few launches ) are intentionally simple and as far as possible, with redundant systems. i.e VHF GPS, and GPRS/SMS GPS, GMC and onboard cameras etc etc.

And I've got to get a Radio license

Doug

OK this is getting off topic and I won't keep on about it here, beyond this one post. Understanding the detailed dynamics of atmospheres is surely a major target for planetary exploration. Clouds may be fuzzier than rocks but they can sometimes take quite crisp forms, on scales of just a few metres in some cases. They act as visible tracers of air movements and markers of humidity levels as well as possible compositional gradients of gaseous and suspended liquids and solids. In short they are most worthy imaging targets. I envisage possibly a cluster of mini-balloons descending through a planetary atmosphere each taking pictures in perhaps three different directions. Even with random separate orientations and spins there could be enough dual 'hits' to generate a 3D movie of the atmospheric goings-on. I think this would be an exciting thing to do and could be done with simple 'Ellison' balloons. Getting even one lucky stereo hit with just two of the things in Earth's atmosphere would demonstrate the potential and would perhaps be a technical first.

all this aside I wish you every success with the exciting stuff you are planning.

Thirty Thousand Metres Via A Plastic Bag

The above is a Flash tool for helping develop high altitude balloons. Briefly: the Met balloons are made of some rubbery compound which stretches under reduced pressure until the point at which they burst. With a given quantity of lifting gas in them, the approximate altitude for the burst can be calculated beforehand. Much more Romantic, and Golden-Age-Of-Ballooningy, are the light plastic/Mylar balloons that are part-filled with lifting gas, and gain sphericality at altitude. These are truly bouyant - with the right parameters, they'll find a level of no-lift, no-sink - which makes them ideal for higher altitudes and longer durations. Naturally UMSF engineers could employ a timer to cut loose the Earth-Return Package from such a balloon and achieve higher altitudes than otherwise.

On to the tool...

There's four yellow sliders, one for altitude, three for the balloon. Set the desired balloon parameters first. Radius is the desired balloon radius, Density is envelope density (970kg/m3 seems to be ok for most commercial polyethylenes) and Thickness governs the quality of the envelope, measured in gauges, as shown below:

Gauge Sort-of-Thing
70 Light Duty Rubbish bags - (those bags which literally are rubbish when they burst after putting hardly anything in)
150 Heavy Duty Rubbish bags
200 Refuse Sacks
800 Heavy Construction Film

Setting the three balloon sliders produces a number of results for the balloon. Most important are the volume and envelope mass.

Now move the altitude slider - as the height increases, the figures for Hydrogen and Helium's excess lift drops. Once the excess lift, for a chosen gas, falls below the payload mass, that's the maximum altitude achievable. At this point, the Hydrogen and Helium required figures make sense - these are the mass and volume of the chosen lifting gas required at sea level.

Example: Radius - 5m, Density - 970kg/m3, Thickness - 70 gauge. Max Helium altitude with 2kg payload = 30500m, 31100m with Hydrogen (so not worth the bother for the added risk?)

Max altitude for same balloon with a 100kg Doug payload (generously including thermal gear and some oxygen) ~14000m.

Enjoy!

Andy

The envelope itself is arguably the easy part. Buy it - fill it - launch it. They're a commercially available part that doesn't need that much thinking about.

The stats are here : http://www.kaymont.com/pages/sounding-balloons.cfm : you over fill compared to those stats slightly - and get a better lifting capacity - but you reduce your ceiling as a result.

Oh - and the chances of flying me on something like that to 14km.....slim....unlike me

Doug

OK a few simulations of camera coverage.

Two cameras, one landscape pointed horizontally, one portrait, pointed 45 degrees down both normal 35mm field of view. Take one image with each and here is the coverage on a 360x180 degree equirectangular projection.

Click to view attachment

Rotate and swing to the payload, 30 degrees looks about right from a lot of images I've seen from other attempts.
Take 25 images per camera (one every 15 seconds for ~6 minutes say) and you'd be unlucky not to get a 360 and most terrain should be covered twice.

Click to view attachment Click to view attachment

Coverage tends to improve (less likely to get that jagged line of missing data) if you don't have the cameras going off simultaneously (which should also be easier to implement )

Click to view attachmentClick to view attachment

Now that's cunning - and even though the data would suggest otherwise - I'm still tempted to use a simultaneous release just because we then know we've got n 2 or 3 frame panoramas. What do your sims say for two cameras in portrait, two in landscape, three in portrait, three in landscape with say, 10% FOV overlap?

I've had another idea for a self monitoring camera which could be cool....got to wait till I have the gondola boxes and a few cheap digi cams

Doug

Your wish is my command...

In each of these I've used the same orientations of the gondola when each picture is taken to make them somewhat comparable. In each case there are 15 images from each camera. Also included is the coverage after one shot from each camera and two percentages indicating the total coverage and the coverage in the lower hemisphere (as the upper one should be mostly black sky ). Of course this is very biased as a lot of pixels near nadir represent very little solid angle in this projection but hopefully it tells us something. The cameras are arranged so they have about the same amount of overlap between the images (5 degrees or so).

1) Similar to before two cameras, upper in landscape, lower in portrait and pointed down at 40 degrees. 61% 86%

Click to view attachmentClick to view attachment

2) Both cameras in portrait one pointed down at 45 degrees. 65% 91%

Click to view attachmentClick to view attachment

3) Both cameras in landscape one pointed down at 35 degrees. 54% 71%

Click to view attachmentClick to view attachment

Sims using three cameras to follow.

[...]

4) Three cameras all in landscape with elevations +15, -20 and -55 degrees. 76% 99%

Click to view attachmentClick to view attachment

5) Three cameras all in portrait, two side by side (35 degrees) with one underneath (-45 degrees) 72% 96%

Click to view attachmentClick to view attachment

So in conclusion, if we can stretch to three cameras I'd go for option 4 (even with just 10 images per camera I still get 94% coverage of the ground hemisphere in this simulation). This one also has the added benefit of being a bit Huygens like.

James

Venus Titan and Neptune, on the other hand would be less than ideal. Nobody was suggesting that you need balloons to study terrestrial cloudforms. Indeed a movie or sufficient still shots from any fast aircraft or satellite would achieve excellent results without the need for mountaineers. My idea was to use Earth's conveniently accessible and largely transparent atmosphere and Doug's simple-as-possible balloons to try out a technique that might prove practical and cost-effective in more awkward locations elsewhere in the solar system.

Well - I've just ordered three very VERY cheap 3 megapixel cameras. Cheap enough that I don't care if they get...

1. Broken by me stripping them out to work with a timer circuit
2. Frozen to death
3. Broken on landing via drowning, crushing etc
4. Lost for all eternity - last seen at 95,000ft over Norfolk.

And also - if they're any good at all - I''ll get another half dozen just so we've got 'stock' as it were. Once I know roughly how large their images are, I'll get SD cards to suit the requirements of >3 hrs at 20 second intervals.

For the price of my 400D (which isn't going within 10 miles of a polystyrene fish storage box hung under a giant condom full of helium some idiot from Leicester's going to let go) - I could by 24 of these 3 megapixel jobs. Once I have them (this weekend - I'm away giving a talk tomorrow) - I'll 'calibrate' them to find their actual FOV and work on some sort of internal framework (balsa and/or foam) to mount them at appropriate angles to one another in landscape (if the vertical FOV is >45 degrees) or portrait (if the vertical FOV is <45 degrees)

Before that - I want to figure out how to power all three from one big-ass central power supply, how long that will last, and if I can wire them up into one relay - or if I'll have to power three relays with the one timer circuit relay...to trigger them. Once I've got my mirrors - I'll be figuring out if CatCam can do the GMC campaign on its own - or if I want to tie that in with the same circuit (I think I do)

Meanwhile - got to convince the photographer that she wants to let catcam get trashed....could be difficult.

Stunning work with those coverage sims James.

The coverage question and whether we should try for three cameras leads me back to some research that we need to close out on. We know that some of the mid range Digicams (Some of the Canon PowerShots (S3is, S5is , 570) and some Casios have an internal time-lapse\intervalometer mode that is hugely useful. What we really need is to find a camera that meets the following:

Must have features:
Lightweight: <200g ? If we want to fly three cameras then we really should be looking for <133g
Time-Lapse mode \ Intervalometer _or_ a lightweight trigger circuit
Capability to be powered by external DC source
Fully automatic exposure control (this is a given in a digicam)

Desirable Features
Low cost: The cameras are very likely to be consumables in this exercise so it would be smarter to select a cheap one.
Wide Angle.
High resolution
Good optics

Any other suggestions?

Doug - My Fujitsu Siemens Loox-T is not responding to intensive CPR so I've given it to someone better qualiied to try and recover it. In the meantime I'm trying to scam a replacement out of my various contacts. If all else fails I'll buy sufficient kit for the cause. I reckon I need about a week to get the code to send GPRS and SMS GPS coordinates working, fortunately I don't need a working GPS enabled device to develop and test that so we should have a high level of confidence in that system fairly soon even if I haven't nailed down the hardware.

Well - I think with these cheap cameras I've got coming - I'll tick the low-cost, wide angle boxes and bolt on a simple timer circuit. I need to establish what, if anything, is happening when the shutter is pressed ( is their current or voltage involved that might feed back from one camera to another etc - and does that even matter if they're all identical anyway )

With multiple cameras, the independent timer circuit is going to be a better bet than relying on the intervalometers of whatever type of camera we have - as I imagine they would drift out of sync with one another Power is still a 'hmm - what to do' issue - because the FM APRS Tx needs a power source as well - although it will happily run of a standalone 9V for a week or more. Amazingly - there are some C and D recharable cells that run to 5000Ahrs and 11000 Ahrs respectively. A pair of those for each camera would give us all the juice we would need (5000Ahrs is twice the capacity of an excellent AA)

I was worried about the cold on the GMC - but going for the most simple possible solution - http://www.bargainboards.co.uk/P/MyCoal-Fo...mers(1252).aspx - would make the problem go away entirely.

Once I have chute, box, cameras, FM APRS Tx, GPS - I can think about the mass budget and figure out how many Whr's we can get into it. If it's looking desperate then in order..things would get culled like this

3rd camera
2nd camera
GMC
IBS
Replace 1st camera with GMC fitted internally.

The bare minimum is one camera, and the FM APRS Tx & GPS. Then it's that + the IBS etc etc.

If your Loox is properly dead - and ignoring the beggars can't be choosers factor - it'd be brilliant if you could scrounge a similar model so that the dead one can become a mass model.

If Helvick can't come up with another unit, I have a fully functioning Palm Treo 700 that I'd be willing to part with for this project.

[...]

Someone else can drown their cameras intentionally - I'm only aiming to do it unintentionally There's a real engineering challenge in 'depth' for the payload protection - far more so than altititude - but the 'transport' would be so much easier ( a line, and some rope) interesting concept though.

Top marks for the heating option - though I would be somewhat wary of enclosing one of these in a thermally insulated sealed box. Could get toasty! Any idea of their wattage?

And I'm sold on the idea of flying a thermometer - Maplins do a 100k Bead Thermistor (for 90p!) that works down to -55C. But if a thermometer is to go - then a manometer's a must, too, in order to check altitude.

Andy

Edit: Rs Components offer a suitable manometer - but it's £40! Surely they can be found for less than that?

A 90p thermistor's great...BUT....you need something to read it, log it, and then spit it back out to you on the ground - it becomes a big problem.

There are commercially available stand alone thermometers that log - but they're not cheap. More sense for a first flight might be a simple standalone max/min system like this http://www.maplin.co.uk/Module.aspx?Module...er&doy=3m10 Something like that will also let me test the GMC sub-ass in the freezer with the lead hanging out.

We wont have a full trace- but we will have a record of the 'worst' conditions both internally and externally. For understanding the 'system' - that's a usefull point.

I want to pack this thing full of as many simple, stand alone, independent systems to get as much as we can, as easily as we can, as cheap as we can - to learn how it all works - so we can say "Yes - we know how to fly a balloon to >80,000 ft and get it back safely"

THEN....we can start thinking about some onboard intelligence that can measure and log...

Lat
Long
Alt
Pressure
Multiple temperatures ( internal battery - internal camera - external - external GMC )
Voltages
Acceleration in 3 axis
Orientation in 3 axis

PLUS - scientific study such as UV etc etc.

But that's a whole realm of 'how the hell....' beyond my knowledge - and I want to figure out how to fly a balloon and get it back before figuring out how to make it a genuine platform for investigation.

Monitoring various sensors sounds exactly like a mission for a Basic Stamp or similar...light, affordable and programmable. I wouldn't rule it out of the equation even for a first test.

Andy

[...]

Well, really we did. The mission was approved and in many official documents was referred to as Mariner Jupiter-Saturn. It simply underwent a name change.

I find this an interesting possibility. I'm surprise to not see any response to it.
One of the instruments could be a compass.

Someone suggested donating a Treo. There might be some Palm software that could run the exposures for the camera system or store information from a weather station. Would require a serial connection, IIRC.

Re Treo's and other potential offerings. They are great end user devices but trust me on this cos I do this for a living - you will encounter pain beyond belief if you try to develop any network oriented software worth attempting on any hand held platform that isn't based on Windows Mobile. I am a major, major fan of free (as in beer and as in speech) software and open source \ collaborative \ what not software but when it comes to hand held systems the only open platform worth working on today is Microsoft's. It seems mad to say it but their development environment is the most open and developer friendly system that you will find today. I know I can develop a GPS aware application (using either a built in GPS or one connected via Serial\IR\Bluetooth) that periodically (every 15-30 seconds) posts updated position data via a GPRS\UMTS\EVDO data connection, or via SMS, to a remote server of some sort using totally freely available tools for development and testing (via emulators for the platform and for the GPS). I've tried to do the same with Palm, Symbian (Nokia) and Linux (Motorola) and have been cut off at the knees every time.

Just my 2c.

It is sad because I refuse to use a Windows Mobile device, to the point that I am still nursing my 4 year old Sony NX80v.

It is a Windows based Treo. Sorry I didn't make that clear. http://www.palm.com/us/products/smartphones/treo700w/

FWIW I used it for two or three months but hated some of its features and went with a Blackberry instead. I can probably sell it for $200-$250 on eBay, but I'm still willing to donate it if it will work. I'll even include the 2G SD card.

Will it work in the UK do you think? Quad/Tri band is the requirement I think. I can buy a pay-as-you-go sim card over here and drop it in.

You got me on that one. Maybe Helvick has some idea. I had assumed it was all global now, subject to service arrangements.

Got my three very very cheap cameras and the polystyrene enclosures ( perhaps a bit big - but even if I chop a chunk out of them and glue it back up it'll be a better job than a DIY effort ). Bloody postal strikes mean lots of other cool bits won't be arriving for a while.

Doug

Well - my idea of a front-door peep hole lens to get fish-eye for the GMC isn't going to work. On both the old catcam, and the new 3 MP cameras - they're FOV is too large to crop into the round FOV from the lens. So it's a WAM technique that'll have to work.

The normal cameras come apart fairly easily and I was able to hard-solder them to a housing and run them off 2 C cells instead of the two AA's that fit internally. I was also able to strip the flash and it's capacitor off the circuitboard with no ill effect - and I've been able to remove and re-attache the ribbon that drives the LCD with no problems as well. The shutter is actually on a tiny PCB of its own and that will be fine in terms of wiring something up. The trigger circuit that I thought would do the trick may not be up to it - it actually takes a 12v supply which seems excessive.

Current plan for imaging.

• Find a new triggering mechanism
• Wire up cat-cam as it is for C-Cell operation (it will happily do image after image at whatever interval I ask for as long as I want all on it's own...I'm not bothered about getting GMC and the other cameras in sync) and do some life-testing of a 4000 mAh C-cell with it - pointing at trees and clouds (appropriate target)
• Shout at the postal service because I don't have my wide angle mirrors, parachute, radio or half my order of electronic bits and pieces.
• Carve up these new polystyrene gondola's to make them an appropriate size, glue with a hot glue gun and then find some alu tape and possibly further insulation.
• Calibrate GMC and normal cameras. (let's call them FOC - Flight Observation Cameras - FOC A, B, C...from top to bottom) Find their FOV's and figure out a means to mount them. The way the l see it - A can mount in a side wall, C in the floor - but B is more difficult without cutting a lot of polystyrene or putting the camera almost outside.

Gizmo has a thing for Elysium and Chryse... she reached up like a Mearcat and pulled the pin I put in at the V1 landing site

Doug

Grin. I'm glad we built something you liked. I'm in MS Research now, but ten years ago I was on the original Windows CE team, from which Windows Mobile is descended, and I have the "Ship-It" award to prove it.

If it would help, I could donate some MS software towards this effort. Or do you already have everything you need?

--Greg

I have nothing to offer this project except encouragement!

I must admit I lurk in this thread in to bask in the glow of the great DIY spirit.

Oh, and for the occasional inside peeks at UMSF's "secret shedquarters" and Gizmo.

These crap 3 megapixel cameras will happily record a 15 fps 320 x 240 movie for 2 hrs none stop (just tried it - <200 meg video - 2 hrs 2 mins long). No sound...but I'm tempted to go with two cameras for photos, and one with a WAM for video. It works great.

Doug

I have been following this thread pretty closely. I though the following links from makezine.com might be interesing:

Video podcast part 1 and part 2 about lanching a balloon with 4 canon sd cameras set up to take a panoramic picture every 7 seconds.

The second is a successful following the make guidelines using a video camera.

seeing the gondolas used is very interesting.

My 2cents, make sure you stick your name, email address and telephone number on the outside of the box and on the individual components inside.

Their wiki doesn't have many details on the timing circuit - that's exactly what I need. PICAXE might do the job - looks like the sort of controller that a moron like me can code, but will do what we need.

I have a LOT of that foam around - it's awesome.

[...]

We're talking about a digital camera bought for less than £10.

Doug

This sounds pretty exciting In my mind this can be something with a lot of potential. First starting out with simple designs just to learn how to build, launch, take data and finally make a safe landing. The next step would be to fly real scientific instruments. The final step would be to fly on a commercial suborbital rocket. With companies like Masten Space aiming for $250 per kilogram to 100 kilometers altitude, I really think this is something we could pull off.

Before we start to do any real design work we first have to think about how to run a project like this. What software tools do we use to design the hardware ? How do we make sure parts designed by different teams are compatible with each other ? How do we communicate and share designs ? Perhaps the first step would be to look at how open source software projects are run. To put it as we say in .nl: "a good start is half the work".

The second thing that came to mind is that to achieve mission success, we need to keep things simple and cost effective. One way to achieve this would be to use off the shelf components, and thinking of creative ways to use them. One crazy idea that popped up in my head was to use MIDI for timing and synchronization.

Last night my mind kept racing and I couldn't stop myself from thinking about technical issues as well I mainly thought of ways how to decide when to open the chute:

• Open the chute as soon as the payload is released from the balloon
• Open the chute after x seconds
• Open the chute at x meters per second
• Open the chute at x meters altitude
• Open the chute at x g-load.
• Open the chute at x meters altitude and y meters per second

I guess the first option would be the easiest to trigger, but also has the highest risk of the chute not opening properly. With the last option you can calculate the dynamic pressure in real time and open the chute based on that. I think this would have the highest chance of the chute opening properly, but also has the highest risk of the triggering going wrong.

Can't wait until first ATLO

Balloon
|
|
|
|
|
|
|
|
|
Chute
|
|
|
|
Gondola


Balloon bursts - chute opens automatically - couldn't be simpler. It's worked for hundreds of amateur HAB's.

Doug

Really! What kind of sensor is used to detect the balloon burst? Only things I can think of are some sort of tether tension cell (like a maritime winch) or an indirect acceleration measurement from an onboard nav system. Would be interested to know the actual method, because both of these techniques seem pretty awkward.

A sudden acceleration of -9.8 m/s/s is the triggering signal. When that force is applied on a (now) untethered gondola there will be (relative to the gondola) an upward movement of air that will inflate the parachute.

I tried to take out a patent on it in mine and Doug's names (using Doug's diagram) but apparently its been done before.

[...]

The 144.8 Mhz FM APRS kit arrived. I need to get a 10mw radio module for it ( pin compatable swop out coming soon ) - but it WORKS.

This is the module just sat on my desk, the gps rec on the window - and my Icom IC-R5 listening in with an audio cable into my PC using 'Packet Engine Pro' and AGW Tracker. Couldn't BELIEVE It worked first time! Screenshot of the first lat-long coming out of it attached.

Just turned it on again - slightly updated Lat Long this time - and plugged the GPS Coords into Google Earth. It's REALLY working. That point is the corner of this garden - <10 metres.

Doug