LE-1 Flight
(Dec 2013)

Altitude: 55,198 ft

Flight: 1 hours, 45 min

Distance: 52 miles

Well, we launched our first balloon yesterday! We even got the darned thing back, though that seemed unlikely for quite a while. In this post we will discuss the highs and lows we experienced, and detail the ups and downs of the launch (and if those bad puns didn’t faze you, then read on).

This was mission number LE-1, a small 200 gram balloon from Edmund (http://www.scientificsonline.com/professional-weather-balloon.html) so we could do a reasonably-low-altitude flight to test things out and determine whether we could actually make this magic happen. We had unproven electronics, minimal radio experience, no ballooning experience, an old fart and two teens. Huh. The flight plan was to launch south of Maricopa, AZ, fly (well, be pushed) in an easterly direction, burst at an altitude of about 45,000 feet, and descend to a nice orderly pick up spot in a farmer’s field near the east side of town. Yeah, that was the plan.

As we were driving out of town, and I was calling the FAA to file a NOTAM (Notice to Airmen) report (alerting pilots about our upcoming balloon launch), we drove right under a hot-air balloon that had just launched. This balloon was quite cool, er, hot, and super close to the road. We thought it to be a good omen for the day!

It took longer to get out of the house than anticipated. The driving to south Maricopa took longer than anticipated. The payload prep and balloon filling took longer than anticipated. We powered up the payload, waited for GPS lock, monitored our datalogging (which is echoed to an external serial port for a laptop to see), verified camera start/stop, closed up the payload, and tied it to the balloon. We had an “If-Found” label on the cover, but realized later that it was facing inside the payload, not outside for someone to see. Our wonderful Adafruit (http://www.adafruit.com/) GPS module locked quickly, even when upside-down. Our Byonics (http://www.byonics.com/) MicroTrak-400 APRS transmitter was awesome right from the start. The LiPo power system from Sparkfun (https://www.sparkfun.com/) provided plenty of juice.

The crazy little framework that holds our styrofoam payload box does indeed serve a purpose. Our radio tracking system transmits on 144.390 MHz to amateur radio APRS ground stations. The antenna is a dipole that is over 3 feet long. 2m APRS ground stations use vertically-polarized antennas, but a vertical dipole (or quarter-wave, j-pole…) on a balloon would have a null facing down towards Earth. It is said that the null from a vertical is not a problem at altitude as there are many ground stations laterally, but we wanted to “hear” the balloon by pointing a yagi up at it, and an null underneath would make that more problematic. Many folks seem to use horizontal dipoles on balloons, so there is no downward-facing null, but they take a nearly-30dB loss in signal reception at low altitudes (critical in descent). Or, they hang a vertical dipole, vertical quarterwave, or vertical j-pole “zeppelin” antenna and live with the null facing down at altitude. So we thought, “why not a dipole at 45-degrees or so to split the horiz/vert-polarization issue?” From there, it evolved into a “jack” design (as in the ball-and-jacks kids game), with six legs, so that regardless of landing orientation, a dipole antenna on one axis would be at an angle, and not dropped flat into the mud with no propagation. But would the legs smash apart on landing? Hmm. We finally got our baby off the ground, and as the balloon was floating away, we were receiving radio position reports via our directional-yagi-antenna/radio-receiver/soundcard-modem/laptop, gave high-fives all around, packed up our crap, and called Prescott FSS to update our NOTAM (as our launch time was an hour late). The weather is nice, but a little bit cloudy. Outside air temperature is 55F (13C), air pressure is 13.7 psi, and Relative Humidity (RH) is 24%.

Since this camera runs very hot when taking video, we configured it to take pics every two seconds. This pic is from the balloon camera just one minute after it lifted off. The reported altitude is from the GPS (Global Positioning System) at 2355 feet MSL (Mean-Sea-Level) altitude, but, since the launch site is 1715 ft MSL, the balloon is actually 640 feet AGL (Above Ground Level). That is a lot of mnemonics, sheesh (ALM,S). As we headed east on I-8 to Casa Grande, we could not get position reports on the cell phone as we had hoped (slow cell data service I suppose), but a call to mom at her computer told us that our little balloon was indeed getting tracking data into the APRS amateur radio system, and the whole world (and a few folks on the ISS) could see our path on a map at http://aprs.fi/. We also got calls from Jack Crabtree, who (as we recently found out), runs the Arizona Near-Space Research ballooning organization (http://ansr.org/) here in Arizona. So it seemed that, even though our balloon was lingering a bit longer than expected over the launch site, it was indeed ascending and sending position data. Yeah!

As we drove to Casa Grande, the balloon reached 12,562 feet MSL (about 20 minutes after launch), and snapped the above pic. The diagonal line directly in front of the balloon is, I believe, a decaying contrail from an airliner that departed Sky Harbor airport. Outside air temperature has already dropped to 37F (3C), air pressure is down to 9.5 psi, and RH is 12%.

We continued to drive east, having no clue what the balloon was seeing (as these pictures were stored in the payload camera card), and the balloon hit 14,576 MSL (about 24 minutes after launch) and took the above pic — do you see the triangle formed by the roads, and the little light-brown patch of dirt in the middle? That was our launch site. I love this pic.

Here we are at 21,632 feet (36 minutes after launch) looking north, and the town of Maricopa is below. In the distance is Phoenix, and beyond that is Canada and Santa’s workshop. You need to squint.

We are still driving east, and the balloon is hitting 25,350 feet MSL in the above pic, at about 41 minutes after launch. It is starting to hit the clouds, which are now much thicker that we saw in the morning. Outside air temperature is -2F (-19C), air pressure is 5.6 psi, and RH is 6%. Since our humidity sensor is light-sensitive for some reason, we put it inside the payload, so the humidity readings are probably not very accurate.

Here the LE-1 has reached 30,637 feet MSL, and is looking down through the clouds at about 48 minutes into the flight. If you look very closely near the bottom of the picture, you can see us waving at the balloon. Look for my red shirt.

Now the balloon is at 35,749 feet (about 55 minutes in), just breaking through the cloud cover. Outside air temperature is a brisk -29F (-34C), air pressure was 3.6 psi, and RH is still showing 6%.

In the above pic, the balloon has reached a sunny 36,727 feet MSL (at about 56 minutes), and is just above the clouds. Down on the ground, we have just found a Starbucks in Casa Grande, and are waiting in a sluggish line to get recharged. We watch aprs.fi and see the balloon getting higher. It passes the expected burst altitude of 45K. Even though the balloon just farted around for a while at low altitudes, it seems to be moving faster than expected now that it is higher. This is indeed a bummer, since we don’t want it to drift into the untamed raw desert (spoiler alert: it eventually does).

It has now hit 49,634 feet (at about 73 minutes), and LE-1 is just having a great party up there. Meanwhile, below in Whoville, the natives are nervous. The higher is goes, the further it will drift, and the nastier the recovery terrain.

So my frap is almost empty, and the balloon is now at 53,718 feet (79 minutes in). The sucker still has not popped. It will now likely drift into nasty land. It is interesting how thin the blue atmosphere looks here, and how deep-black space looks. While there is indeed a noticeable curvature to the Earth at these heights, this camera lens is a bit wide-angle (2.8mm), and distorts lines (so ignore any curves you see in these pics). We have longer focal length lenses for this camera, and will substitute a more “normal” lens next launch (something around 5 or 6mm is close to normal on this camera), which should show curves only when there really are curves to be seen.

This pic is at about 54,879 feet (80 minutes after launch), just before it bursts at a maximum altitude of about 55,198 feet, as recorded by the internal datalogger. There were only a few pics after this before the camera stopped recording. Why? I dunno. A great flailing of the payload that caused a power or SD card intermittent? The fact that it was about 3 degrees F (-16C) inside the payload box, and -17 degrees F (-27C) outside the box? That’s just nasty cold. Maybe that fact that the air pressure had dropped to 1.4 psi (one tenth of that at sea level)? This means that there is very little conductive heat extraction mechanism, so a device that runs hot, like the camera, may overheat, even though the “air” temp is very low (thanks to James Ewen for pointing this subtle detail out). Crazy extremes up there.

Jack Crabtree called again and gave us a pretty good location of the final resting place of our payload. Sadly, we dropped out of cell phone range by the time we got to the target location, so we had no cell service, no cell-gps, no online maps, no angry birds.

After we reached the approximate location, we tried to use the directional yagi antenna with the radio receiver to RDF (Radio-Direction-Find) the angle in which to search. Unfortunately, we could not hear the payload beacon in any direction. We connected an omnidirectional antenna and we WERE able to hear the beacon! So it was pretty close. We connected the radio receiver to the soundcard-modem/laptop to get packet decoding and got a latitude/longitude. Sadly we had no portable gps, the cell phone gps was out-of-service-range, and we had no detailed map of the area. We searched as best we could, but finally gave up, and headed 20-minutes back up to where we had cell-phone range again.

We were almost ready to give up when we got a couple of pictures on the phone from mom, with close-up satellite pics of the estimated landing area. We got optimistic again and headed back there before dark, hiked in a bit further than the area we first searched, and finally found the balloon payload!

The first thing I noticed is that the framework of 1/4-inch dowels had not been smashed on landing, thanks to our 36″ rocketman chute (http://www.the-rocketman.com/chutes.html). So the 2m dipole antenna (transmitting latitude/longitude data now on 146.565 MHz at 300-baud as an RDF beacon), was indeed OFF the ground and gave us excellent range of over a mile (at 400 mW). If only I knew how to RDF on it, had an autonomous GPS, or just a decent paper map (since it gave me coordinates within 50 feet of actual location).

Here are the interesting stats:
– This trip was expected to to reach 45,242 feet to burst, and travel 36 miles in 79 minutes.
– This trip actually reached 55,198 feet to burst, and traveled 52 miles in 105 minutes.
– The balloon must have been filled with less helium than planned (we used a circumference string) since
it went higher than expected (hence drifted longer)
– The final transmitted APRS position was at 4022 MSL, which is about 1746 AGL.
– At about 3164 MSL (~888 AGL), our payload was programmed to switch to RDF mode on a different beacon frequency.
– The lat/long coordinates we received from the landed rdf beacon were LESS THAN 200 ft from the last APRS position.
– The actual payload was LESS THAN 50 feet from the rdf beacon coordinates! Wow.
– If we knew how to actually RDF we would have found it reasonably-quickly.
– If we had a non-cell-phone GPS we would have found it quite quickly.
– The stress of this project was expected to take 1.2 years off my life, but has has taken 3.5, +/- 1 year.

Our next launch (LE-2) will use a much larger 600g balloon, and is expected to reach over 90,000 feet. Stay tuned…
Ahh. What a day!

The balloon was expected to to reach 45,242 feet to the burst altitude, and travel 36 miles in 79 minutes, as the above pre-flight prediction from http://predict.habhub.org shows. I don’t know how long that site saves data, but the simulation run is at:

http://predict.habhub.org/#!/uuid=c78ca23a9b51aaceb21c6cd6352ea9e3f88ad274

Above is the actual balloon tracking from the APRS radio transmissions to ground stations, on the http://aprs.fi/ web site. Due to less helium in the balloon that we planned, the flight actually reached 55,198 feet to burst, and traveled 52 miles in 105 minutes. To watch this flight, you could have gone to aprs.fi, and searched for AF7EZ-11 (our tracker id). The data points were updated in near-real-time.

Here we are zooming in on the landing zone, and changing the aprs.fi map to satellite mode.

Zooming in even further, we can see how close things were at the end. The final transmitted APRS position was at 4022 MSL, (~1746 AGL). At about 3164 MSL (~888 AGL), our payload switched to 146.565.MHz for RDF beaconing — we could not find it with a directional antenna, but with an omni antenna we received and decoded the lat/long coordinates of the payload on the ground, which was LESS THAN 200 ft from the last APRS position. It was a good thing that we got an APRS position report this close to the ground, and also lucky that surface winds were low, so the payload would not drift too far from that last report.

When we found the payload, it was LESS THAN 50 feet from the rdf beacon coordinates! That darned GPS stuff is dandy.

Above is a map plot of a different theoretical path (calculated using the NWS winds-aloft forecast) and the experimental (actual) path.

And above are the theoretical vs. experimental paths plotted in Google Earth.

So, we did quite a few things right, had some sheer luck involved, and learned some hard lessons:

– Start earlier in the morning

– Don’t assemble the payload at the launch site or tie the rigging then either (will be routing power wires outside the box next time, so the payload is all ready for power-up by simple external cable connections).

– Have a better handle on the volume of helium. Flowmeters are pricey, but a calibrated weight for neck lift is easy.

– It is great to talk to someone sitting at a computer and watching the tracking, especially while driving.

– Before leaving civilization, find a wifi spot and look at the area of the landing site closely.

– Have detailed maps that extend well beyond your expected landing area.

– Have a gps/map system that does not need cell service to function.

– Bring extra water and some snacks.

– The payload went quiet for about six minutes in the middle of descent as it crossed Hwy 79 (see the lack of red dots on the aprs.fi map, above) — it was rebooting several times. Luckily it came back to life, or we would not have had gps data with which to find it.

– The SD datalogger card popped out of the socket upon landing. The beacon kept transmitting, but our datalog ended (we had all the log data we needed, but we will tape this card in place next time).

– The camera stopped functioning just after burst altitude, perhaps due to the payload being bounced about (which is a bummer, as the landing pics would have been interesting).

But all-in-all, we can’t complain about the LE-1 mission.