ReMOAR

Posted by JonM on 12 May 2008

Woke up with a start at 6 am. Up and out of the bed, eat breakfast, pack up, and head back out to base camp.

Like a well-oiled machine, we got camp setup, although the winds were twice as bad as yesterday!

We tore apart the non-functioning repeater, moving one transceiver to the ground ops station, and putting the other in my truck. This would give us much better distance communications in the field. The Internet link was relocated to a shallow rise on the other side of the road, where Verizon could hear us. We set the APRS ground station up with a three element Yagi, and relocated the balloon fill site further into the field, to get more wind shelter from the hill.

The wind was knocking down our shade shelters, so the soldering station was setup in the back of Chicken's pickup. Cozy, but functional.

With the base camp reconfigured, we got the MOAR launch under way. Batteries were soldered to the transmitter, I assembled another cutdown device, and everything was tucked into the eggshell. With a bright, longer set of streamers, we waited for confirmation from the ground crew that our beacons were going out correctly. No dice on the first go: The GPS unit's ground wire had broken loose. After fixing that, we still couldn't get a 3D lock; without altitude information we couldn't fly. We cut it open, and realized that the GPS was upside down; with the antenna resting on the Microtrak's ground shield, it wasn't getting any signal. We flipped it over, and promptly got 3D lock.

3ric got the final go-ahead from the FAA, so we cracked the valve on the helium cylinder, and filled the balloon. Even with 8 wranglers, it was all over the place, knocking people off their feet and popping off the fill tube. We persevered, however, and got the balloon filled and tied to the payload. I set the cutdown timer, this time for 210 minutes, and we let the balloon loose.

This time the wind took it straight east, rising about 400 feet per minute. We lost tracking again after a few minutes, and I was worried the cutdown timer fired prematurely. Adam and I loaded up and headed out to the field again. A few minutes later, we got word that tracking was picking up again, and it was headed steadily east.

Headed out on the open road, we loosely followed it through open farmland, getting position reports from the ground station, until we got out of range of our simplex radio communications. Fortunately, we were back in cell coverage, and were able to get continued position information from the base camp.

Thirty miles out, the base camp stopped picking up the MOAR beacons, and it still wasn't getting picked up on the APRS digipeaters. Adam and I were about 50 miles from base camp when we decided to try finding a repeater with coverage. The AK2O repeater covers most of eastern Washington, and I had gotten permission from them last year to use it, but we couldn't hit it with our handhelds, so I didn't ask them again this year. However, with the 50 watt mobile unit we could hit it just fine, and the base camp could to. We were a little worried about offending the AK2O folks by taking up their repeater, but they seemed interested in what we were doing.

3ric's latest predictions said MOAR was going to land in Idaho, about a hundred miles from our current position. With that much ambiguity, Adam and I decided to call it a wrap and head back to camp.

No less than two minutes after we turned around, we received a beacon from MOAR. It was at 85000 feet, and almost directly above us!

We turned around again, and headed off to the last position. There were only thirty minutes until the cutdown timer went off again, so we wanted to get more position reports before it dropped out of the sky. We drove around on all sorts of entertainingly named backroads (Bachelor Flats Road S), but didn't get any more beacons. Shortly after 2pm, 10 minutes after the cutdown would have triggered, we called it a wrap and headed back towards camp.

Since it was getting late in the day, and the FAA was growing increasingly impatient with us, base camp decided not to go ahead with the REHAB launch, and packed everything up. We converged back in Coulee City, to eat at Grandpa Joe's Cafe, which is run by the wife of the farmer whose land we used for our launch site.

They were very gracious and accommodating, seating nearly 20 rowdy hackers on Mother's Day, and the food was fantastic after a long day's work.

Properly sated, we began the return caravan to Seattle, during which there were several multicast jellybean transmissions. Packet loss was high, but acceptable.

Airborne!

Posted by JonM on 12 May 2008

Saturday started off pretty rough. We rolled into town well past 3:30 in the morning, and woke up again at 8. Frank and company wrangled an awesome breakfast, and then we hit the road. The weather was less than stellar, and we were perhaps a bit optimistic about our chances of launching, but we arrived at the launch site, performed a quick survey, and rapidly set about setting up camp.

There were boxes to be unloaded, shade structures to be set, and masts to be guyed. Everyone knew what they were supposed to be doing, and things got setup in good order. The only real hitch was that our Internet connection wasn't operational. It appears that the AT&T tower near the launch site fell over, and wasn't passing any data or voice. Verizon was unreachable from our hilltop location, as well.

The weather started clearing up, though, so we got the machinery rolling to get MOAR, our attempt at beating the amateur high altitude record, launched.

The clouds may have let up, but the wind sure didn't, and it was quite the challenge to get the balloon filled, but fill it we did. The payload was sealed shut, tied to the balloon, I set the cut down timer to go off in 4 hours, and we released the balloon

The wind took it far and fast, and pretty quickly we were up to 8000 feet, rising at about 600 feet per minute. Then we got a packet from only 5000 feet. Followed by radio silence. The last report was only a couple of miles away, so Adam, EricV and I got in the truck and headed out to search. The last beacon was right in the middle of a 1 mile grid square, so we hiked in and looked around a bit. We couldn't see it, but we could hear it on the radio, so we got back in the truck and drove in on a 4x4 road, hoping to pick up a beacon. Success! We hiked to the coordinate, and found it nestled in the sheared wheat rows, less than 100 feet from where we walked before. MOAR was largely unscathed, despite dropping over 5000 feet and hitting the ground.

A quick inspection revealed that the line was cleanly melted through, indicating that the cutdown timer had gone off prematurely.

We took our prize back to base camp, and packed up for the day. Since it was 5pm already, there was no way we could launch REHAB that day.

Over dinner (thanks again, Frank), we looked at the cutdown timer code again, and discussed options for the next day.

The code revealed a bit of logic I had forgotten about in the year since I first wrote the code: The cutdown timer is set by pressing a button which increments the timeout value in 10 minute increments. After 240 minutes (4 hours), the next press will wrap the timer around to 10 minutes. Apparently in the jostling of the wind, I pressed the button as I was arming the timer, and it cut the payload loose shortly after launch. This was somewhat relieving; at least we know that it's functioning as designed.

After dinner, we worked out a plan for Sunday: We would awake at the crack of dawn to relaunch MOAR, and then two hours later launch REHAB. REHAB was far from complete, so we pared it down to two video cameras, an APRS tracker, and a Digi Xtend modem which would broadcast our GPS coordinates. Three hours later, we had a nearly complete REHAB, including wired components and remolded airframe. We hit the sack shortly after midnight, with our alarms set for 6 in the morning.

Busy day at the lab

Posted by JonM on 27 April 2008

We had a lot going on at the lab yesterday. In the early afternoon, Brian and Sirus continued hacking on the ground software, while 3ric, Leigh and I got the thermal chamber going again. We tested the BeeLine failsafe beacon, and it performed marvelously, showing only a very slight frequency drift (5kHz or so) down at -60C.

With a successful round of thermal testing under our belts, 3ric and Leigh then worked on getting our ground-station Internet gateways configured and running, and I worked on getting our new glider assembled.

Phil assembled a rig for thermal testing our glider flight controller and it's data acquisition, which involved some fairly intricate soldering:

As I was getting ready to pack it in for the night, there was continued thermal testing. Phil's thermal tests of the flight controller and data acquisition were quite successful, while Dan took a much more hands-on approach to thermal testing the science payload controllers:

All in all, a good day at the lab. It's good to see things coming together, and have everyone working so hard on getting things built and working together.

Radio range testing

Posted by JonM on 12 April 2008

Dan and I went out today to do some radio range testing. We decided to start small and stay within Seattle city limits. We had two sets of sites, the first about four miles, from Magnolia Park to Alki Beach Park. I went up to Magnolia Park and set up a nice little testing station:

Test station at Alki

First and foremost was testing the Digi XTend RF modem, which will be our primary data link from the ground station to the payloads. Dan set one up on the Alki beach running medium power, and they immediately paired with a nice, strong, signal. Depending on the antenna orientation, we got from -62 dBm down to -74 dBm, which matches our calculations nicely. Further testing will be needed to verify the path fade, but I'm pretty sure we'll be able to get up to 100 miles of range with no problems.

The surprise star of the day was the BeeLine homing beacon. That little green guy was heard 4 miles away:

Pretty impressive for a self contained package that weighs less than half an ounce and
will run for 48 hours on a single charge!

After that success, we decamped and headed over to the second pair of test sites. I set up at the out of service boat ramp at Sand Point:

I got the call from Dan that we was set up and...nothing. It turns out that Dan had misinterpreted the south site (a boat ramp near Seward Park), and had setup in Seward Park proper. Our clear 8 mile line of site was now a clear 6 miles with 2 miles of Mercer Island in between. Ah well, at least the weather was nice.

Popular

Tag Cloud