Launch: Success. Retrieval: Failure. Or, "A Lament for Balloon 1"

Farewell, Balloon 1. We hardly knew thee.

Here the camera looks back at us one last time, through its tiny hole in the side of the capsule.  Within moments it was out of view, off on a journey to the edge of space.

What did it see, peering down through the unforgiving cold?  Did the abyss above give it pause?  Did the fragility of the harbor below it really offer solace when faced with the awful reality of the universe?

Perhaps it never fell to earth at all.  Perhaps now, instead of hanging high in a tree amongst the green pines of Virginia, or standing silently in a lonely field, or even sailing solemnly out to sea, perhaps it floats above us still.  To confront so great a challenge as the heavens!  Ah!  This is the place for one so bold as Balloon 1!  We may meet again, friend, when the abyss comes for us as well.  We may meet again.

Launch! T+20 minutes!

We have liftoff!  The launch went without a hitch.  We are a bit worried about a slow ascent rate, but watching the balloon disappear was really satisfying. We are still tracking now.  There will be a tense 4 or 5 hours after it leaves range until we hope to get a new GPS location upon impact.  WOOOOOOO!!!!

Here are the actual GPS location data from the balloon after about 15 minutes.  Surprisingly it is following Route 29.

Here is the simulation data for 12 noon, with a ceiling of about 60000 feet.  Landing in Culpeper would be ideal!  You can see the simulation closely follows the real data for the first 15 minutes!

Excitement!  Thanks to Jonathan, Tyler and Stache for help with the build and purchasing components.  Also a shout out to Pete, who watched the lauch.

Update:  It's now 6:15pm, almost 6 hours since the launch.  Although our rate of ascent seemed a bit slow, we still expected the balloon to reach 60,000 in 5 hours, then fall for around an hour.  If the balloon's ceiling was closer to 24km, that would mean another hour of ascent, but it's hard to tell exactly.  As the sun sets, the likelihood of finding the payload becomes much smaller.  It may be that it landed in an area with no cell phone coverage, in which case we have to hope someone will find it and call us.


Here's the live GPS data; if we ever get an update it will show here:

GPS tracking powered by InstaMapper.com

Launch: April 3rd

This is quick update post.  I've bought the phone and set up the GPS tracking software.  Ginny and I went for a walk in the neighborhood, and the tracker worked fine.  Cell phone coverage in C-ville is very good, but it'll be a miracle if tomorrow we land somewhere with four bars.  Hopefully the decision against an external antenna won't bite us in the behind.

I also got the cooler, some string, some neon orange duct tape, some "Hothands" hand warmers.

The camera script works well.  I haven't made a final decision on shutter speed, and the fact that the LCD screen is broken doesn't help the testing.  The last big worry, other than landing in a skeet-shooting range, is battery life on the used camera.  It seems to be holding up, hopefully it will be able to take the 1600 shots we might need over the course of the flight.

See everyone at 10:30 tomorrow morning!

Update:  I've found the replacement battery I bought for the camera, and it has 900mAh, instead of the standard 700mAh.  I feel much better about the camera now.  If we can find this thing tomorrow afternoon, this will be epic.

Launch Soon?

This Saturday is a possible launch day.  It should be beautiful, but hope for low wind.  I've loaded CHDK onto our sacrificial Powershot SD400.  My next worry is battery life on this used camera, but I'm thinking about replacing the 4.3V rechargeable with 3 AA disposables.

Tomorrow I'll be testing the camera to see how the intervalometer works, and how long the battery will last.  I'll also play with some ISO and shutter settings to figure out the best way to get awesome shots.  If everything looks good, I'll buy the phone tomorrow.

Tyler has taken care of the parachute, and I'll grab a cooler, string and duct tape.  We'll be putting together the launch vehicle at the Physics building on Saturday morning.

The camera should hold 6000 shots on the highest pixel setting and lowest jpg setting, so we may be able to do a shot every 5 seconds.  Here's the script that I'll use in CHDK as an intervalometer:

rem Author - Keoeeit
rem Upgraded by Mika Tanninen
rem Time accuracy and shutdown for a710is by Viktoras Stanaitis
rem h-accuracy for delay, j-accuracy for interval
h=-1
j=-1755
@title Ultra Intervalometer
@param a Delay 1st Shot (Mins)
@default a 0
@param b Delay 1st Shot (Secs)
@default b 0
@param c Number of Shots (0 inf)
@default c 0
@param d Interval (Minutes)
@default d 0
@param e Interval (Seconds)
@default e 10
@param f Interval (10th Seconds)
@default f 0
n=0
t=(d*600+e*10+f)*100+j
if c<1 then let c=0
if t<100 then let t=100
g=(a*60)+b+h
if g<=0 then goto "interval"
for m=1 to g
 print "Intvl Begins:", (g-m)/60; "min", (g-m)%60; "sec"
 sleep 930
 next m
:interval
  n=n+1
  if c=0 then print "Shot", n else print "Shot", n, "of", c
  shoot
  if n=c then shut_down
  sleep t
  goto "interval"

Near-space Photo Update #1

The project is moving along.  I've ordered the balloons, and found a used Canon Powershot SD600 on Ebay.  The camera's gears are stripped, so the lens won't retract, but it apparently still takes photos and was $20.

Several people have expressed interest in helping, so there should be a pretty good bunch.  We're trying to decide where to launch from next.  It's tempting to try to drive somewhere that will allow a more likely landing zone within cell phone range, but I think we'll just launch from the Physics Building rooftop.  We'll have to watch the winds.  The picture below shows a launch from the Physics Building today at noon:

As you can see, today it would have followed 64 pretty closely.  This would be quite fortunate.

To maximize our chances of recovering the balloon, I'm considering making an antenna for the cell phone.  The antenna could hang down from the craft, and be tipped with a blinking LED for greater visibility.  This way, even if the craft ends up far from a cell phone tower, we'll have a better shot of at least some data reaching us as it falls.

We may also remove the battery pack from the camera and wire in our own.  I trust a set of lithium ion AAs more than a used NiCd rechargeable.

While a streamer seemed sufficient at first, I think a real parachute will be safer.  Now I've got to find or make one.  To minimize weight, I think we'll use a much smaller styrofoam body than the MIT students.

I just found a video that details how to make a directional cell phone antenna with coffee cans.  This might be something to consider.

Near-Space Photo Project

I loved model rocketry as a kid, and I always wanted one of those rockets with a camera.  This project looks to be much cooler.  Although the coolness factor of this project is reduced by the fervor surrounding the success of MIT students' near-space photos, this is still going to be lots of fun.

Essentially, I hope to take pictures with a camera attached to a weather balloon.  The students I linked above give nearly step by step instructions, so this should be obscenely easy.  All you need are a sounding balloon (already ordered, $30 for 2), a camera (Canon, to take advantage of the firmware needed to take pictures at a given time interval), a cell phone (to provide GPS tracking to locate the thing after it lands, $50), and a vehicle.  I think I'll just steal the styrofoam cooler idea for the vehicle, although I think I'll use a streamer instead of a parachute.  Then you need extra batteries for the cell phone (although maybe not), and some heat packs (also maybe not needed).

The camera firmware is CHDK, and the tracking will be done with InstaMapper.  I'll just buy the cell phone from Best Buy; a used camera will be the hard thing to find.

Cell phone coverage is a worry.  To get GPS data from the phone, it'll need reception.  The coverage map for boostmobile above will suggest launch location.


I'll use this Balloon Trajectory Forecast tool to try to figure where the thing will land.  Here's what the trajectory might have looked like today:

 

Here, the vehicle ended up almost 50 miles away, 30 miles northwest of Richmond, 10 miles from Kings Dominion.  This would be bad, as this is not quite in cell phone range, it seems.  Of course, this will change day-to-day.

More to come on preparations for launch!

Low Power Home Server Project

For too long I've left a computer running in my house to act as a little server, eating up electricity and heating up the living room. I general have used old machines that are useful for little else, but while these computers are essentially free, they use your standard 300 W power supply.

My goal in this project was to find a cheap server configuration that I wouldn't feel guilty about leaving on all the time. I went ahead and bought two 1 TB hard drives; these are cheap as free nowadays. But what to use for the mainframe?

I thought about the Eee Box, a lost cost, low power desktop. Running Linux, it would be as flexible as I could hope for, and I could plug in the TB drives with a USB enclosure, but the price, at $300, was a bit steep. I looked at all in one network storage solutions such as this one from Netgear, but the cost was still a bit more than I wanted, and the flexibility was essentially nil.

It was the Eee Box that reminded me of the Eee PC netbook I wasn't using. This Eee PC was an old Celeron 900MHz model, clocked down to 630MHz. I had always found it difficult to use; the screen is just a touch too small, the keyboard was painful, and it was just plain slow. But, as a server I wouldn't have to interact with locally, it became very attractive.

I found here that the Eee PC draws about 910mA when on but with the screen off. With a ~10V power adapter this means about 9W. Running a USB hard drive adds around 160mA when idle, or 500mA when busy and the heaviest CPU and SSD usage pulls another 250mA This means I would be pulling about 11W at idle, but only as much as 16.5W busy. That's like one of those compact fluorescent bulbs running all the time. Not too shabby.

I picked up a two drive USB enclosure from Newegg, so next I had to decide how to do backup. A TB is beyond my current needs, so mirroring the drives was my plan. But, should I set them up in RAID 1? This was the idea, but I decided this was a mistake. Faulty writes to a RAID 1 still result in data loss. The smart approach is to use one disk as the main data drive, and set the other up as a backup.

After I installed Ubuntu, I wrote a simple RSYNC script to do the backups for me, and added it to cron for thrice a week. Here's the script:

#!/bin/bash

rsync -avc --progress --log-file=/home/my_user_name/rsync_log/$(date +%Y%m%d)_rsync.log /media/data/ /media/backup/

Now, my data is backed up three times a week to the "backup" drive. I was worried about the speed of the Eee and the USB enclosure for serving movies, but I haven't seen any problem streaming movies to my set top box yet. There is no doubt that copying big files is slow, but it's managable.

I then set up LAMP for serving, UFW as a firewall (redundant, since my router is acting a firewall, too), and NX for remote access. After port forwarding on my cable modem and router, I can access this machine through ssh from anywhere. My next plans are to setup sharing with my XBox 360 and use a spiffy sharing tool like Tonido in case I start to hate Samba.

Overall, tons of fun, cheap and eco-friendly. How can you go wrong?