Crash ... help ... please ... thanks!

Spooky. I couldn't see anything obvious, except that there was a compass error in the 1st GPX file, and not in the 2nd. Have you calibrated both the compass and the ACC? I think the ACC has to be recalibrated after a firmware upgrade. But none of this makes sense because the 1st flight was (in your words) uneventful. The only thing I can think of is that maybe there's an issue with your wiring or one of the 2nd set of LiPos that's causing some interference which sometimes has an effect, and sometimes doesn't. Sorry, that's probably no help at all.

Good on ya for getting 'er down in (more or less) one piece.

 

I have a problem similar whit my Octo, suddenly the QP batteries start to give me 43 sec fly but the voltage start and finish is 16.7 and 16.3
maybe have no relation bur everything is possible.

http://forum.freeflysystems.com/index.php?threads/cinestar-8-fly-problem.2404/

 

I am not sure if the motor restart is causal. It looks like that occurred when the copter was on the ground. I did see motor 8 ramp down to 2.1 A for a second but then looked normal after that. The battery looks fine no unusual sag. The one thing I notice with this file that is similar to the other copter crash is that the variometer seems erratic compared to the vertical velocity. I am not sure of what to make of that if there is anything. The negative altimeter is not uncommon for a Baro if there were rapid pressure changes in the area. But with the new sensor I was under the impression that it was normalizing the altitude, unless it reverted back to the old sensor.

I don't see any erratic yaw movement just the normal nose hunting. It could have been a draft and simply got into its rotor wash.

One of things I did notice is that you did a good job kicking altitude hold, however I did not see any throttle movement. That would suggest that the copter was in a stalled in the rotor wash. Its always a good idea to go altitude hold off throttle as required until you see it climb.

Not implying this was the issue,

What were the winds like?

Shaun

 

I've found that the mx20's telemetry is prone to dropouts ("zero point zero volts"), but if it's >0, it's generally correct.

And I run dual 4S 8000 mAh batteries on my MK stack, and I generally attach one, then disconnect and attach the other, and make sure the voltage on both is 16.8 before I fly. [knocking wood] I've yet to have a LiPo go bad on me. But if/when I do, I will immediately relegate it to ground service.

 

I would do a test flight in a grassy area and launch with each battery by itself and see if one battery is sagging quickerthan the other. That would be a more important issue to figure out. It is possible to have one battery degrading at a higher rate then the other. Like Andy said the voltage of the pair look fine. If you are concerned about this see if one battery is sagging faster than the other. Like if it takes one battery 5 minutes to reach 14 V and another 5.5 min than you would know one Lipo does not have the same discharge rate anymore.

 

I know how to fix that. Disconnect and remove the GPS/compass stack and fly using only the MK board, you'll never have another issue like that

 

There are two issues here. One is related to the battery sagging over time.

1. The initial throttle required to hover at the beginning of the flight is not same as the battery normally depletes it voltage. In theory since you did not fly that long it would be fairly close. The copter requires the same power to hover say 1440 watts to fly. At takeoff if the battery sags to 16V that means you need 90 Amps. Now for simplicity the battery is at 14.4 Volts you need 100Amps to hover.

In Altitude hold the system does this for you in free fly you constantly regulate this with you throttle.

2. Now the other issue if you induced a small rotor stall called "Vortex Ring State," where the propeller is flying back into its thrust. This causes a disruption of the smooth air flow through propeller. If caught early the only part of the prop that is stalling is the area near the center hub. At this point you can throttle out of it you still enough lift in the rest of the prop. If caught too late and no corrective input is added then it gets exponentially worse the descent increases and more of the prop surface area gets disrupted and eventually will lead to the entire prop being stalled. This has caused a few more catastrophic mishaps on this forum where the pilot is bringing down and suddenly it drops aggressively out of the sky and hits the ground at a high vertical rate.

I can say from my own experience that I have had this happen once, very early on when I was testing how much descent rate I could do. I thought had enough forward motion in the wind to handle the descent rate but it dropped out of the sky from 40 feet in a snap. Two things saved me. One was I was doing this at an airpark with a nice flat surface. Luckilly I caught it in time partially flew out of it not entirely, I went max throttle and it still hit the ground pretty firm. Fortunately the gyroscopic stability of being at max power allowed it to hit all three prongs to hit the deck without tipping over. Once the rotor stall was depleted by the ground the copter bounced back in the air then I was able to go back to normal power stabilize then land and check for damage. The whole thing happened in about a second or two. Had I not done anything I would have probably would have been writing $3000 check to Qudrocopter so to speak and buying my cam op a new 7D. I honestly thought that was what I was going to pickup broken pieces in the seconds that I saw it drop the way it did.

Difference between my stall and yours is I was being aggressive and I got into a worse situation and almost no recoverable position and got lucky. Yours was just partially stalled. You moving the copter back towards you, probably helped prevent it from getting worse, because you were flying somewhat out of it, but not completely. But you needed to arrest the descent rate. A little bit of power increase probably would have been enough to arrest the descent and get it back into control then stabilize it for a controlled landing.

The best way to fly out of these is by moving into the wind to force the copter to fly away from its rotor wash. But you do need to add power at some point, especially to counter the thrust lost do to lateral pitch and roll.

I would go to a field with no obstacles and practice how much throttle is required to fly out of the events. You start noticing vortex ring state by the props sounding choppy and you will notice a slight "wobble,' as the thrust on each motor is dirty and the FC is having a harder time compensating. Noticing it early is the best way to avoid this from happening again.

 

I am currently conducting a formal crash investigation involving a Cinestar8 with Flight Controller Hardware Version 2.2, Software 2.0a.
There may be some similarities with your incident. We would need to discuss this off-line.

 

That's a good question for which I done have a good answer for. Some comparison data suggests that the narrower props are more efficient although there is no comparison data out there to correlate the effects of altitude and prop speed. I have some Xoars that I am going to use on my HL, but it will be in a coaxial configuration.

There are a lot of variables that factor this in. For instance the APC slow fly props have a wide cord so they produce significant of the lift using Bernoulli principle versus the Newton Principle. The Xoars that I have are narrow cord which means most of the lift is generated by the pitch angle of the blade versus the airfoil of the prop. It seems that the research I have done on Vortex Ring state suggests that the more pitch the blade has the more exacerbated the stall gets. Now most of that data is in reference to a traditional helicopter rotor that does not have a significant aerodynamic twist compared to an airplane propeller. The nice thing about these props are that they will stall at the hub first and if recognized in time there is enough lift available to throttle out of it. On the other hand the the Bernoutlli component produces a wider pressure differential.

Its possible that the narrower props based off of other peoples observations concentrate the air in a smaller surface area and a more confined volume that helps alleviate some of these effects. Others even have suggested this is why Coaxials descend better. Being that no formal tests have been done on the propeller dynamics with multi rotors to compare at what point to disruptive airflow's cause the prop to stall its hard to make a qualified conclusion.

This phenomenon can happen regardless what props you use. The best way is to =recognize and recover, which is a fundamental tenet in the aviation world. The nice thing is that higher winds speeds help alleviate this to a point. Gusty winds can aggravate it. But you did the correct thing by switching off autopilot AH mode. That would probably would have made it worse because the throttle response of the AP does not seem to react quick enough when these events occur. When you recognize it the best thing to do is smooth power application followed by speed into the wind if able.

Shaun