Eggfinder Tracking Problems??

[wsume99]

I had something strange occur this weekend and I'm looking for some feedback on what could have caused it.

Here's the scenario:
I launched a rocket carrying an eggtimer TRS to ~6500ft. I was not able to maintain visual lock on the rocket throughout the entire flight however I was able to track it during ascent, saw the ejection and some of the initial descent. The prevailing winds that day were SW and that matched the initial trajectory of the rocket under descent.

Once I lost sight of the rocket I watched the flight status screen on my eggfinder LCD. The last position update I received from the TRS was at 84ft AGL. Based solely on the partial visual track I anticipated needing to drive to recover my rocket as it was at least a 1/2 mile away in a field on the other side of the nearby road. As I walked towards my car I switched over to the heading screen on the eggfinder LCD and once I had moved enough for it to compute a valid track I noticed the heading it was displaying was due east. I was a surprised by this at first because it didn't match the NE track I saw during descent and the prevailing SW winds would not have pushed the rocket further south so that I ended up with an east track. I should note that I have flown this hardware in that rocket linked to this LCD multiple times and it has always performed flawlessly. At this point I turned around and started walking east because there was nothing but open farm field to the east.

When I got the the location of the last transmission the rocket was nowhere to be found. This happened to be in the middle of a relatively flat farm field with no obstacles within at least 500' in any direction. Based on the terrain I should have been able to see the rocket. At this point I was pretty confused. I plotted the GPS coordinates into Google Maps and according to it I was in the right location. That seemed to confirm the receiver had taken me to the right position. At this point I decided to abandon electronic tracking and walk north towards the direction of the visual track I had observed. The nearby road I mentioned earlier was approximately 1000ft north of the originally reported rocket location. When I got to the road the LCD received a position update from the rocket. I followed the heading and it took me right to the rocket. The rocket was more than 1/4 but less than 1/2 mile N of the reported position at 84' AGL. I was glad to find the rocket and not lose the motor case, electronics and chute release I had onboard. What I am most confused about is this, clearly the TRS reported its position accurately because once I re-established a connection at the road I tracked right to the rocket. Ok, so why then was the position at 84ft AGL so far off?? There's no way the rocket could have drifted that far in it's final 84ft of descent.

I feel pretty confident in assuming that as long as the TRS and the LCD are getting a good GPS signal they are going to be pretty consistent in terms of positional accuracy. I don't see how it could be accurate to within tens of feet only to be off by 2000 ft 30 minutes earlier unless there was a satellite connection issue. The rocket was reporting connection to 7 or 8 satellites from power up until I lost connection and the orange GPS LED on the LCD was on the entire time so I'm assuming I had good GPS. Perhaps that is a poor assumption and someone can educate me. Beyond that I am left with only two other possibilities - the position difference was caused by user error (I don't see what I did wrong) or someone moved my rocket from it's original landing spot. Both locations were not visible from the launch location so it is possible however I saw no indication of that at all. When I found the rocket it looked like a completely normal landing and the rocket had not been disturbed in any way.

Trying to solve this mystery. Maybe this is not abnormal but I've never had this kind of issue before. Appreciate any advice or feedback.

 

[NTP2]

That is odd!

 

[Buckeye]

My friend's EggFinder did the same thing one time. The actual landing spot was nowhere near the last reported coordinates, like 1/4 mile away. We started walking in the more logical direction, reacquired the signal, the coordinates refreshed and were now correct. I chalked it up to my friend screwing up his GPS handheld, but now I am not so sure. This was in flat farmland, no mountains or valleys.

If a GPS receiver has line of sight and a 1Hz reporting rate, the actual landing should be like 20 yards max away from the last reported coordinates. That's been my experience over the years.

 

[Rocketclar]

I fly a lot of Egg stuff. I've had cases where the accuracy was poor due to low number of satellites. Perhaps it lost lock along the way and only had 3 or 4 satellites when it reported out the last position.

 

[user 30299]

Was there possibly another person at the launch with an Eggfinder? Same frequency as yours? We have had an incident or
two at launches where university teams where testing their Eggfinder units, and one of our regular member's base unit
locked onto a team's rocket transmitter. The team and the member had not changed the frequency the Eggfinder was
set on when shipped to them.

 

[Handeman]

I had almost the same issue at the last launch. The last 8 to 10 reported positions as the rocket came down on main was 4900 ft from the pad. The actual position was 3100 ft from the pad and about 25 degrees different. It was 2100 ft. between the last reported position and it's actual position. The next launch was also off by 1200 ft. in the same manner.
I called and talked to Cris. He was of the opinion that there was interference with another transmitter. Mine was at the default freq. and ID. There was another person there using a EggFinder. I haven't heard back from them if they were also on the default freq. and ID. I suspect they were. I have since updated the firmware and changed the freq. and ID number. We'll see if that works this weekend.

The only other issue I have is that when it does connect and update, the distance from the transmitter to the reported position is off anywhere from 40 to 300 ft. I use rocket locator app because we have good cell coverage and I like the satellite maps. The blue dot for my position is very accurate. There are enough tree lines, fence lines, roads, etc. that I can tell the self location on the locator app is pretty much ±10 ft. The map pin for the rocket locations is from the reported lat. and long. that comes from the LCD receiver which of course is coming from the GPS TX unit. That isn't as accurate. Of course, the phone is also triangulating off cell towers to increase the position accuracy and the transmitter doesn't have that luxury.

 

[wsume99]

If a GPS receiver has line of sight and a 1Hz reporting rate, the actual landing should be like 20 yards max away from the last reported coordinates. That's been my experience over the years.

That's been my experience as well which is why this surprised me so much. Never had that happen before. I had two other flights that day with eggtimer trackers in them and they performed exactly like that.

I fly a lot of Egg stuff. I've had cases where the accuracy was poor due to low number of satellites. Perhaps it lost lock along the way and only had 3 or 4 satellites when it reported out the last position.

I never saw the flight display show less than 7 satellites in view at any point from startup to when I lost signal. Suppose it's possible to lose connectivity and have it restored quickly without that being visible on the flight status display.

Was there possibly another person at the launch with an Eggfinder? Same frequency as yours? We have had an incident or
two at launches where university teams where testing their Eggfinder units, and one of our regular member's base unit
locked onto a team's rocket transmitter. The team and the member had not changed the frequency the Eggfinder was
set on when shipped to them.

I had not thought of that but I did not see anyone else using one while I was flying. Additionally, there were no other flights in that direction or altitude range when I flew my rocket.

I had almost the same issue at the last launch. The last 8 to 10 reported positions as the rocket came down on main was 4900 ft from the pad. The actual position was 3100 ft from the pad and about 25 degrees different. It was 2100 ft. between the last reported position and it's actual position. The next launch was also off by 1200 ft. in the same manner.
I called and talked to Cris. He was of the opinion that there was interference with another transmitter. Mine was at the default freq. and ID. There was another person there using a EggFinder. I haven't heard back from them if they were also on the default freq. and ID. I suspect they were. I have since updated the firmware and changed the freq. and ID number. We'll see if that works this weekend.

The only other issue I have is that when it does connect and update, the distance from the transmitter to the reported position is off anywhere from 40 to 300 ft. I use rocket locator app because we have good cell coverage and I like the satellite maps. The blue dot for my position is very accurate. There are enough tree lines, fence lines, roads, etc. that I can tell the self location on the locator app is pretty much ±10 ft. The map pin for the rocket locations is from the reported lat. and long. that comes from the LCD receiver which of course is coming from the GPS TX unit. That isn't as accurate. Of course, the phone is also triangulating off cell towers to increase the position accuracy and the transmitter doesn't have that luxury.

I'm definitely going to make sure all my units have the lasest FW and I will switch them all to something other than the default freq & ID. My preference is to have all my stuff on the same freq and ID, right now they are all using the default freq and ID. Any issues with them all being the same? If I was flying two at the same time obviously that would be a problem but I don't see that ever happening.

 

[Rocketclar]

I have all of my Eggfinder transmitters set to one freq and code that's not the default. I did the same with my Eggtimer telemetry modules.... Very convenient to do so in my opinion. For some regional launches I've attended, I've notified the group about the frequencies and codes I'm using so we can try to not have the same setups.
One time at a regional launch, I was heading to my rocket via the coordinates, and it was moving. Turns out someone had found it and were driving it back to the flight line!

 

[Tractionengines]

Any issues with them all being the same?

Only if you want to have multiple transmitters powered up. ie. so one is getting a lock and ready for flight, before you go and recover the previously flown one...

Mine are all on the same freq. (Makes it easier to check around and see if anyone else is on the channel.)

 

[cerving]

Yes, you NEED to change the frequency/ID from the default after you've built them and confirmed that they're working. Almost every frequency conflict issue that I've ever seen has been because the users didn't change them from the default.

And yes, I've had the "moving rocket" thing too... somebody grabbed it and drove it back to the flight line. In one case, I tracked it down to their truck... I politely mentioned to them that they had my rocket.

 

[user 30299]

I had not thought of that but I did not see anyone else using one while I was flying. Additionally, there were no other flights in that direction or altitude range when I flew my rocket.

You never know who's got what for a tracker, and some people turn their trackers on early (in the parking lot)
so that it can start getting a lock on as many satellites as possible before heading out to the rail.

I only have one Eggfinder and it's the basic unit - so I have the least experience here.

I'll relay an interesting bit of info regarding the Missile Works trackers, which I use extensively.
If you turn on two transmitters, that are coded to the same base unit, the base unit does not
favor either unit. It jumps between the transmitters and you never get a full lock on one unit.
And so the GPS info is basically garbled.

 

[Charles_McG]

Wouldn't co-channel interference just lead you to someone else's rocket?

Since I've started logging the eggfinder data, I've seen large jumps and errors after high accelerations - like launch. The older GPS receiver seems more prone to it than the model on recent eggfinders.

But 84' would be way low, even for a dual deploy jolt.

 

[wsume99]

Wouldn't co-channel interference just lead you to someone else's rocket?

That's what I'm thinking as well. In the case of multiple transmitters on the same freq & ID then I'd just be receiving messages from multiple sources, right? The receiver is receiving messages from multiple sources but you can't tell which one is yours. Maybe whichever Tx is stronger will wash out the weaker signal?

If my issue was caused by a duplicate Tx on the same freq & ID problem there would have needed to be another transmitter in that location during my launch and recovery window. There was nobody else in that area all day beside me. I was the only idiot flying high enough in the windy conditions to land there. Regardless, I'm still going to switch all my gear to something other than the default.

After I re-acquired the signal and saw the rocket in the distance my initial thought was that someone had moved it. Perhaps a nearby landowner was unhappy. However the rocket was strung out on the ground exactly like I'd expect it to be if it had landed there.

 

[Handeman]

While I was having my issues, I found out that the only person I knew running another EggFinder was on different freq. and ID than I was using so I don't think they had anything to do with the issues I've had. Our launch this weekend has been cancelled because of a fire ban. Apparently most of the firefighters are working in the next county on a wildfire along the Blue Ridge so they are short of resources. I'll test again the next launch and make sure all EggFinder users are on different freq and ID.

 

[waltr]

I have had very weird location from Eggfinder Mini a few times. Once the end track in Spain (yes all the way across the pond) when importing the GPS file into Google Earth.

Other times hundreds of feet away from where it really was.
Most of the flight the GPS was perfectly dead on the real location.

Could it be a corrupted data in the NMEA sequence?
Does the Eggfinder LCD Receiver compare CheckSums to ensure of data integrity?

If not then these way out locations could be a bad bit in the data...

 

[KenECoyote]

I got my first Eggfinders TXs years ago when they first came out with fixed frequencies.

First one was 913.

Next one was worse... 915.

@cerving Cris - Is there a way to reprogram the old units to new frequencies?

 

[cerving]

I have had very weird location from Eggfinder Mini a few times. Once the end track in Spain (yes all the way across the pond) when importing the GPS file into Google Earth.

Other times hundreds of feet away from where it really was.
Most of the flight the GPS was perfectly dead on the real location.

Could it be a corrupted data in the NMEA sequence?
Does the Eggfinder LCD Receiver compare CheckSums to ensure of data integrity?

If not then these way out locations could be a bad bit in the data...

Yes, the receiver does the checksum verification.

Don't know why you had it go so far off... what kind of terrain was that? In a relatively "busy" area, you can get something call "multi path interference", when the signals bounce off other objects (such as trees, buildings, etc.) and cause either null signals or delayed signals. GPS's really don't like delayed signals... it messes up the timing.

 

[waltr]

Seems that when Trees are near the GPS accuracy is worse. Never have any 'Hard objects' around but lots of trees.
On another flight the rocket landed in the woods in the top of a tall tree. GPS first located the rocket outside of the woods and over the next hour moved slowly in a wide circle around where the rocket actually was located. GPD location was never closer than about 200 feet from the rocket.
I have the NMEA file saved (Open Logger board in the LCD receiver).

 

[cerving]

Any chance it could have landed with the little square GPS antenna pointing at the ground?

 

[enkindler]

While I don't know what the problem was here was specificly, it is important to remember that objects like trees are difficult for even survey grade GNSS.

Ceramic patch antennas have high directivity twords the azimuth, mixed in with the incredible -1 65 dBm sensitivity, anything that is a radio reflector at 1.6Ghz can lead to multipath errors.

Shading due to antenna position plus a reflective surface in the direction of antenna sensitivity can pick up signals that have traveled significantly further being louder to the receiver than the shortest path.

while this error is typically less than 10 meters in ideal cases, it can become much worse under some conditions.

while I have no information to conclusively state that is what happened in this case, it is important to remember that even your smart phone uses accelerometers and pressure sensors to try and compensate for this and in the case of rocketry those methods are not available.

SO while still rare, occasional large errors in reported location should be expected with patch antennas laying close to the ground with suboptimal antenna positioning.

it would be nice if it was possible to retrieve the last successful packet received while the rocket was still in decent but I don't know if that is possible with hardware constraints and possible UI confusion.

 

[Handeman]

it would be nice if it was possible to retrieve the last successful packet received while the rocket was still in decent but I don't know if that is possible with hardware constraints and possible UI confusion.

This is where my concern is at. I've had several flight where there were 10+ packets received that marked out an arch as the rocket was under chute and mapped out a trace of the location before the packets ended as the rocket landed, and ALL of the packets were reporting a location +2000 ft away from the rockets actual location. That lead me to go to that location and not getting any other packets to update the location or give me any possible location of where the rocket really was at. I only found the rocket because I started moving closer to the pad where I thought it was actually at and picked up a single packet. I used that in google maps to get a location and started to pick up packets as I got closer to the location. When I stood over the TX, it was still reporting a location 80 ft. away from where it actually was.

 

[Charles_McG]

You guys are reminding me that I should get back to eggStation testing this winter.

 

[pwrwgnwalt]

I checked on GPS.gov to see if there were any reported or scheduled outages earlier this month; none found (but reports can be delayed a couple of months). Experience appears to be that unexplained GPS anomalies do occur - whether intentional jamming, or more likely unintentional. Of course, a hardware anomaly cannot be overruled either (power fade, connection, etc.).