Saturday, October 15, 2016

Flight Wisp1c_12

Flight Wisp1c_12

Planned Changes

  • Tracker is in the basic configuration as Flight Wisp1c_10..
  • Wisp1l code base.  Added modifications to have the interrupt service routines do a bit more of the processing for GPS and 5351 transmissions.  This is a baby-step toward moving to a non-blocking main driver routine.
  • Swapped out a resistor on the voltage divider, allowing the supercap to charge to 4.8v.

Preparation


Item Weight (g)
Tracker
13.9
Tape and String attaching balloons
0.8
Free Lift
5.0
Total
19.7



Day 1 - 10/19/2016

Launch

Launched at 12:58pm local time from Rhodes Hall, Cornell University.

Here is Harry holding the two balloons in preparation for launch.  You can see that the balloons are only partially filled, and that we're using two 36" mylar balloons, attached to one another.  A tether with counterpoise runs from the balloons to the tracker, which I am holding while taking this picture.


Here's the tracker.  The wire and string coming out of the top lead to the pair of balloons, which Harry is holding to my left.  The antenna comes out the bottom, and goes off to my right, where Mark is holding it.  We're holding everything apart to keep it from getting tangled in the wind.

You can see a blue light blinking on the left side of the tracker.  It's transmitting one of the initial packets.  It's a bit hard to tell from this picture, but it was very sunny.  The solar panels charged it right up, and it booted up and started beaconing.





Here's Mark holding the very end of the antenna, keeping it from getting tangled on anything prior to launch.



And it's away!


Day 1 results


We traveled 235 miles before the tracker ran out of daylight.  It averaged about 58 mph throughout the afternoon.  The yellow prediction line was calculated right after launch.  We followed it beautifully.



We were received quite well during the day, averaging 12 receiving stations per packet.  We were heard across the ocean in Germany quite early in the day.

The hysplit model seems to give us about a 50/50 chance of doing a big loop-de-loop in the Atlantic.  Here's hoping we follow the track going east in to Africa!

We'll see what tomorrow brings.


Day 2 - 10/20/2016


The tracker awoke on schedule, but pleasantly north of the projected path, which improved the likelihood that it would move east rather than south.  It had traveled 1,247 miles overnight.

The altitude climbed steadily through the morning and settled out at 10,777 meters, which is right on target.


WSPR was received very well throughout the day.  Lots of stations received on each transmit cycle.  The only lost packets happened at the beginning and end of the day.

Predictions look good going into tomorrow!

Day 3 - 10/21/2016



Day 3 went very well.  It averaged around 82mph all day, and the night before.  It held around 10,888m in altitude for most of the day.

Predictions look good for the coming days.

WSPR reception was incredible.  Australia even heard us!

Day 4 - 10/22/2016




The day went smoothly weather-wise.  We traveled about 81 mph over night and 71 mph throughout the day.  All seemed well, except that we had an exceptional number (6) of GPS resets today.  Not sure what's up with that.  Also, for the second time, we experienced a watchdog reset at boot time (or perhaps shutdown last night).  Altitude remained solid at 10,888 for most of the day.

WSPR reception was exceptional throughout the day.  We were overheard in Antarctica, Australia and Japan in the wee hours of the night!

The prediction continues to be "due east" for the coming days.  Awesome!

Day 5 - 10/23/2016



Flight day 5 complete this morning, and went quite smoothly.  The speeds are down a bit, averaging only about 39 mph during the daylight hours of the flight.  The altitude remains great at 10,888 meters through the day.


The hysplit predictions remain awesome.  The balloon is moving slowly, but steadily almost due east!


Propagation wasn't as amazing today as it was yesterday, still we had a few distant stations picking us up.  Europe was hearing it very well.

Flight Day 6 - 10/24/2016



Flight day 6 was a strong one.  We remained at around 10,888 meters for most of the day.  We were moving a little slowly the night before, but the winds picked up as the day wore on.

The flight day ended as we entered Tehran air space.  Marvelous! hehe!


Predictions still look excellent for the coming days.  Let's hope that dip in the Pacific straightens out as we get closer!


WSPR traffic was a bit spotty today.  We had a few distant stations, but on average only about 3 stations per cycle.  We had the most dropped packets of any day.  We're getting to a bad area for WSPR.


Flight Day 7 - 10/25/2016



Flight day 7 ended just barely over the border into China.  All systems look good!  Altitude was great at 10,888 meters.  Speeds remain about 65 mph for the most part.

We're getting more remote, and have many fewer receivers.  Still, we have enough to get the job done!

Flight Day 8 - 10/25/2016 - 10/26/2016



Flight Day 8 was a little nerve wracking!  It was supposed to wake up and beacon starting at about 9:30pm eastern.  It wasn't heard until after 5am.  The remote location, coupled with terrible band conditions due to a solar storm led to very poor reception.  Still, thankfully, when it finally popped up, it had covered serious ground!  The final few beacons of the day were the only ones heard, but they gave enough data to show we were still going!


The predictions updated with the latest location look totally awesome.  They show the tracker coming right back to New York.  Way cool!



As mentioned above, band conditions were miserable.  These were the only receiving stations for today's data.

Flight Day 9 - 10/26/2016 - 10/27/2016


I didn't think to get imagery for today.  We cleared Asia and made it to the Pacific!  Statistics below.

Flight Day 10 - 10/27/2016


We blazed across the Pacific today, averaging more than 150 mph.   The altitude came up a bit, to 10,555 through the day.


The weather ahead is a little scary.  We should make landfall about the time the tracker awakens in the morning.  Hopefully the weather ahead will continue to move with us.


Reception was good across the US.  With luck, we'll hear the tracker from Ithaca tomorrow!

Flight Day 11 - 10/28/2016



The balloon awoke on schedule today in Washington State.  I progressed quickly NorthEast into Canada, averaging speeds in excess of 100 mph.  The altitude was great, climbing to 10,666 meters.

The predictions for tomorrow look optimistic that the tracker will have already circumnavigated by the time it wakes up in morning


WSPR reception was solid throughout the day.

Flight Day 12 - 10/29/2016 - The Circumnavigation!!




I awoke on Saturday morning, waiting for the tracker to beacon.  It was expected at around 8am local time, and to be east of the original launch location in Ithaca, meaning it would have circumnavigated. The white box is pointing to about where I expected the balloon, in Vermont or New Hampshire.

Unfortunately, it didn't beacon.  Investigating the weather, I saw this new split in the prediction. Uh oh!  That usually means a front is in front of us.



Oh no!  There was a HUGE storm right in the area where the tracker was expected!  It probably went down due to weather.  I assumed the flight was lost.  I listened for a few days, just in case and gave up on Monday.  I shut down my WSPR -> APRS gateway.

This was an epic bummer, as I presumed I had circumnavigated, but had no evidence to prove it.

Lo and behold, just on a whim, I hit reran an SQL command to check the database on Thursday morning and discovered that my tracker had beaconed the night before! WOOT!   The telemetry indicated that the tracker was in grid square FN44be, at 333m in elevation.


A quick peek indicates that it was indeed in New Hampshire, pretty much exactly below where I suspected it had hit that storm.  Presumably it's in a tree, and it managed to catch enough sunlight late in the afternoon on Wednesday to beacon.

Given that the flight had flown east of the launch point, and had beaconed under its own power, it seems to me I achieved circumnavigation!  Woo hoo!


This is a closer look into the area.

Here's a Satellite view.  It's pretty hilly there. Poking around on Google Earth, it seems that the elevation ranges from 200m to 400+m, which jibes with my beacon report of 333m (accurate to 111m increments).



These are the only two stations that heard the tracker on Wednesday.  Presumably it was one bounce away in the atmosphere.  My home station, W2CXM, did not hear it.

The tracker may still be alive in a tree.  At this point, I'm planning a road trip to New Hampshire to see if I can foxhunt it!  It should be sunny on Monday.

The Foxhunt - 11/7/2016


My tracker hanging in a tree 9 hours away is just too tempting to leave alone.  It was be beyond cool to not only circumnavigate, but also recover the tracker myself.  To that end, I'm going to head out to New Hampshire and give it my best shot on Monday.   My chances are extremely slight, but the payoff of recovering my flight would be huge, so I need to try.

Hank Riley, N1LTV, has hiked in the region from which the tracker beaconed.  He was kind enough to provide a topo map, which he annotated based on the predicted elevation of the tracker.  The tracker beaconed an elevation of 333 meters.  It has a granularity of 111 meters.  So, theoretically, the tracker is at an elevation of 333-444 meters.  However, the telemetry packet indicated that I only had 5 satellites in the lock.  Padding a little for error, I presume the tracker is somewhere between about 300-475 meters.

Further, the tracker beaconed at 15:58, 16:00, and 16:08 local time.  At this time of year in the Northeast, the sun is quite low to the horizon.  Given that, it's not an unreasonable assumption that the tracker is on the west side of any hill that might be there.  Also, looking at the WSPR coverage above, you can see that the receiving stations were to the SouthWest of the tracker.  Again, that reinforces the idea that we're on the west side of a hill, rather than the east side.

Hank annotated the map.  The areas surrounded by the blue are too high in elevation to be considered.  The orange-ish section is centered around the 333 meters that was beaconed.  The pink dots denote the lower extremes in elevation in which we might find the tracker.
I've also made an APRS overlay, indicating the grid square, and gone through and cached maps. I'll have my mobile APRS setup as well as my mobile HF operating.  This will help me keep track of areas I've investigated as I search.

If I do get signals, APRS will also allow me to cast lines on the map for triangulation.


I've assembled some foxhunting gear, including a loop antenna and Kenwood TH-F6, and done a little practicing against a test tracker that I set up at home, and put in a constant beaconing mode. With the tracker laying on the ground, with the antenna draped over a tree branch about 10 feet high (emulating a poorly deployed antenna), I was able to hear the tracker from about 1/4 mile away with my mobile HF rig.  Watching the S-meter, I was able to home in on it pretty well.  Further, the TH-F6 with the loop antenna demonstrated pretty good directionality.  If I can't get too close by road, I should be able to triangulate a bit with the directional antenna.

Of course, this plan presumes that the tracker wakes up and beacons on Monday (which is forecasted to be a sunny day),  If it does not, Plan B is to use field glasses from the road, and work a circuit of the driveable area.  I am also packing an RC airplane upon which I can mount a camera.  The area is heavily wooded, but I might be able to fly over the woods with my camera pointing down and perhaps catch a glint of balloon.

Chances are slim, but I'm committed to spending all the daylight I can spend on Monday to try it!


Foxhunt results

Well, there was a hard frost the night before the hunt, but I got started around 9am.  I circled the area and traversed all the roads, with my HF radio on, listening on 20m, and running WSJT-X in my vehicle.   The tracker never woke up, and never beaconed.  The area was beautiful, but fairly heavily wooded.  I spent the day finding open spots and looking at treetops through my binoculars, but really didn't spot anything.

The balloon icon in this picture is not an accurate representation of the location, but it shows the grid square, and where I searched.


The day after the Foxhunt

I drove home, disappointed but not surprised.  Without a beacon it would be VERY difficult to find the tracker.  As if to spite me, the tracker woke up and beaconed again, for most of 3 hours!


As before, most receiving stations were to the west-southwest.


I'm making plans for a second trip out.


Statistics



Day Rcvd / Lost
Pkts
Pct
Rcvd
Late GPS locks GPS Resets Watch
dog Resets
Farthest Rcv Station (miles) Average stations / beacon
Max / Beacon
Start Grid / Stop Grid Dist covered Night / Day (miles)MPH Night  / DayAvg Float Altitude (m)
1
56/ 3
94.9
2
0
0
3,897
12 / 22
FN12sk / FN42bu
n/a / 235
n/a / 58
10,222
2
109 / 5
95.6
5
0
1
3,258
13.6 / 23
GN65nr / HN25nd
1247 / 478
93 / 73
10,777
3
127 / 5
94.5
4
0
0
9,952
12.2 / 22
HM75lj / IM12ow
849 / 807
83 / 86
10,888
4
106 / 1
99.1
12
6
1
8,960
24.9 / 43
JM08en / IM69ev
763 / 550
81 / 71 
10,888
5
111 / 6
94.9
14
2
0
2,368
10.7 / 23
KM27qi / KM55pj
802 / 548
54 / 39
10,888
6
107 / 15
87.7
0
0
0
4,721
3.1 / 10
LM04uq / LM55dt
591 / 789
44 / 89
10,888
7
21 / 81
20.6
1
0
0
3,480
0.2 / 2
MM36gm / MM86ks
907 / 573
65/ 66
10,888
8
10 / 110 (est)
8.3 (est)
??
0
0
5,334
1.6 / 3
ON22uj / ON32kn
1576 / 61
74 / 120
10,555
9
76 / 20
79.1 
3
0
0
3,889
1.6/ 4
QN10qv / RN16kw 
1871 / 1053 
152 / 130 
10,333
10
87 / 17
83.7 
2
0
0
3,125
2.4 / 6
BN22vn /  BM98ch
2101 / 775
161 / 144
10,555
11
88 / 8
91.6 
2
0
0
2,114
5.3 / 16
CN97do /  DO60ul
1144 / 725 
78 / 105
10,666
Final
?? / ?? 
??
?
0
0
922
1 / 1
FN44be /  n/a
1653 / n/a
?? / ??
??

7 comments:

  1. It was fun to help Mike. Quite cool!

    ReplyDelete
  2. Fantastic Hojo. I've really enjoyed following your flights! - Mike N2VR

    ReplyDelete
  3. Here comes the sun. Beacon please. We want to know where you are.

    ReplyDelete
  4. Quick question: Doesn't it transmit its Lat/Long in each packet?

    ReplyDelete
    Replies
    1. Due to the very low bandwidth of the WSPR protocol, we can't send a very specific Lat/Long. It's using Maidenhead grid squares. The tracker sends a 6-character grid square. In this case, FN44be, which is a 2x3 mile (roughly) area.

      Delete
    2. Thank you - I didn't know that. With 6 bytes I reckon you might be able to do a little better. Using 25 bits for your long. and 23 bits for your lat. should yield (worst case at equator) precision of 1.5 metres and 3 metres respectively. There might be a good reason why you wouldn't do it like that - but that's just shooting from my hip with a back of the napkin calculation

      Delete