RRC3 Flight data questions

[Banzai88]

OK, flew my 2.6 fiberglass Tomach this weekend on an H120. Flight altitude was within 4% of OR prediction. Cross winds were 12-15mph at launch. Flight was BEAUTIFUL!!! Straight up, visible apogee lay over, smooth separation, nice initial fall with good airframe separation, and main deployment was text book. I FINALLY got the combo of harness length, separation charge, and shear pin set up working!

Data, not so much. I was using this battery: https://hobbyking.com/en_us/turnigy...-lipo-pack-fits-nine-eagles-solo-pro-180.html , which I've never had an issue with. The three previous flights on this RRC3 all presented text book data files. The only thing that's really different on this flight was the relatively higher winds below about 500ft.

As you can see, somewhere about 27 seconds into the flight, the voltage dipped significantly, which usually coincides with the main charge firing. This time, it precedes the main charge fire by about 1 second. Battery came right off the charger at 4.2v about 5 hours before the flight.

Somewhere about 300ft AGL during descent, data just stops, which I suspect is why it did not compute Main Rate. Upon recovery of the airframe on the ground, the altimeter was beeping out altitude as expected.

What happened to the data? Do you think that the high winds caused the funky data and failure to report below 300ft? Perhaps the low voltage spike should have coincided with the main charge deployment? Perhaps the low voltage spike somehow messed with the data logging?

 

[djkingsley]

Just for yucks check you lipo and see if it split, I have had three flights similar to yours and all three I discovered the lipo to have swelled to the point it split. I have gone back to 9v batteries until I figure out what is causing the problem.

 

[dixontj93060]

I have seen these dips quite often with the RRC3. In fact, so much so, that I did some bench testing a few months after their release and captured 5 or 6 of the worst traces. Saying that, I've never seen a failure, so didn't pursue it. But Jim Amos is always there if you want to ask him directly.

 

[Banzai88]

Funny thing is that when I use these 2S batteries, I have no issues: https://hobbyking.com/en_us/turnigy-nano-tech-300mah-2s-35-70c-lipo-pack.html

I had been flying 1S batteries in this rocket because of over-current concerns.

Next time that I fly it, I intend to use the 2S and see how the data compares.

 

[markkoelsch]

There is not much margin in a 1s. The 2s should give more consistent results. Another possibility is the old 9 volt- that will work.

 

[jderimig]

Batteries usually do not fluctuate like that. More likely you have a switch, connector or wiring issue on that flight.

 

[Buckeye]

Is this chart typical of RRC3 altitude data? It looks really messy and unsmooth, or maybe oversmoothed. What is the sampling rate?

Did your main chute take a long time to deploy? There is not much change in descent rate after t=28

 

[rharshberger]

I have seen similar ugly graphs from my nose cone mounted RRC3, so turbulent airflow due to higher winds could be a contributor to the messy look.

 

[MWC]

Tom -

The battery voltage is logged with a 1Hz snapshot with the exception being the time when a recovery event is initiated (drogue and main). In these cases, the voltage monitor switches into a 20Hz sampling mode in an effort to record the lowest voltage "sag" point during the 1 second event duration. The battery voltage recording is asynchronous to the event recording, thus when the voltage data is persisted into flash memory there's always going to be potential for skewing of the low voltage sag sample vs. the actual event marker.

With regard to the recording stopping early, this is driven by velocity. From the User Manual on Page 11:

Landing Detection Mode
Once the Main Event has occurred, the RRC3 starts monitoring for a Landing Event. The Landing Event occurs when current velocity > ‐3 ft. per second for a minimum of 2 seconds.

In your case there's a lot of what appears to be swinging, rotation, or gusting effects under canopy that manifests as velocity fluctuations. Normally velocity calcs are much less erratic, and the graph data extrapolation didn't gather enough data to calc a main descent rate either because of the early recording termination.

You mentioned it was a windy day, and how did the all the pieces behave when fully deployed?

 

[Buckeye]

With regard to the recording stopping early, this is driven by velocity. From the User Manual on Page 11:

Landing Detection Mode
Once the Main Event has occurred, the RRC3 starts monitoring for a Landing Event. The Landing Event occurs when current velocity > ‐3 ft. per second for a minimum of 2 seconds.

The only place I possibly see velocity > -3 fps for 2 seconds is highlighted in the red box. Maybe. Did this trigger the landing event?



Seems like the data can easily fool this criterion. Maybe a longer time limit is needed, like 10 seconds.

 

[Banzai88]

Wiring was tight and good, battery freshly charged, deployment nominal.

Due to the CRAZY gusting winds below 500ft, it looks like my elliptical parachute was developing some measurable lift, which I believe is clearly evidenced by the bouncing altitude and velocity track starting at T15/1200ft (under drogue). By T30/500ft (under main), there's clearly a LOT of bouncing up and down.

The rocket, on recovery, was clearly blowing down wind, and I didn't really observe much 'up and down', although it did appear to 'just hang there and blow away' on several occasions. Deployment was smooth and simple (finally) with no spinning or twirling. I think the data shows when that happened!

From the Main event until touch down, crazy cross winds and parachute lift simply fooled the RRC3 into thinking it was somewhere else.

Thanks everyone for your perspective and help figuring it out.

 

[Buckeye]

Do you have another dataset from this rocket/altimeter combo on a more "normal" fight? Velocity from baro altimeters is inherently noisy, and I am still puzzled by the landing criterion.

 

[jderimig]

I think a rotating cylinder with a single port in a wind will show an oscillating pressure inside. I remember a wind speed measurement technique similar to this a LONG time ago, but I could be wrong. Even a stationary cylinder in a breeze will be shedding vortices which can be measured in the port(s) are in the right place. I am with Jim, I do not thing you were lifting because the period of the pressure bumps seems to be consistent. Gusts don't work that way.

Was you rocket "swinging" at all under chute?

 

[Banzai88]

Do you have another dataset from this rocket/altimeter combo on a more "normal" fight? Velocity from baro altimeters is inherently noisy, and I am still puzzled by the landing criterion.

Yes, I have several. Complete 'textbook' profiles, with hardly any noise at all. That's what threw me off on this one.

I think a rotating cylinder with a single port in a wind will show an oscillating pressure inside. I remember a wind speed measurement technique similar to this a LONG time ago, but I could be wrong. Even a stationary cylinder in a breeze will be shedding vortices which can be measured in the port(s) are in the right place. I am with Jim, I do not thing you were lifting because the period of the pressure bumps seems to be consistent. Gusts don't work that way.

Was you rocket "swinging" at all under chute?

The whole string of the rocket was strung out straight as an arrow under chute, no swinging or swaying at all. There was some twist in the lower recovery harness and hardly any in the upper, but I put that down to the fin can twisting from apogee to main deployment.

So, is it your opinion that the data seems to indicate that twisting of the airframe in the high winds caused the cyclic nature of the spikes, thereby wrecking havoc with the software?

 

[jderimig]

So, is it your opinion that the data seems to indicate that twisting of the airframe in the high winds caused the cyclic nature of the spikes, thereby wrecking havoc with the software?

There was an indication of cyclic variation of the av-bay pressure. Its even there during ascent at about the same period. The cycle is pretty slow, period ~ 3secs. That's a slow rotation.

I am just guessing of course.

 

[Banzai88]

There was an indication of cyclic variation of the av-bay pressure. Its even there during ascent at about the same period. The cycle is pretty slow, period ~ 3secs. That's a slow rotation.

I am just guessing of course.

OK, I see that now. Thanks for the insight. This one has earned it's paint, so it'll get painted here soon and back in the sky. Paint should make it easier to observe and I'll see if it does anything funny.