Taranis X9D+ and X8R lock-out and plane crash

[Carbo]

Yes, a heartbeat signal in the log would be fine for some scenario.

[peterberetta]

Hi João,

Thank you for that interpretation. I think you are close to what happened in reality and I am glad that it was not the radio because intermittent radio failure is a much more difficult problem to identify. It is possible that I was causing the problem by giving too much elevator and stalling the surfaces. Perhaps this is why giving opposite aileron input did not cause the plane to change direction. I do have a suspicion that there could have been an elevator linkage problem.

[jhsa]

Well, in my opinion, the radio didn't log what the elevator itself was doing. It did log what YOU did to the Elevator stick, which was, full deflection to one side, presumably pulled the stick fully. So, it appears to me that the elevator deflection was induced by the pilot

My understanding of the problem is:
Pilot error caused the plane to lose speed and crash due to entering a stall, and aggravated by entering a spin due to the use of aileron..

I would do a very good test to the radio gear anyway, just to make sure all is ok.

João

[peterberetta]

Ok, quite possible, however, I was giving opposite aileron to the direction the plane was turning and it was not in a flat spin but a spiral dive. My reasoning in those few seconds was "the airspeed is high, it must respond". Pity I did not have a camera filming the event.

[Carbo]

I was flying on a slope in England (Ivinghoe Beacon) - wind was about 25knots, very strong lift. No motor or winch needed.

Another thought: Did you add some ballast? If so, is it possible that the ballast moved when you threw the glider and changed the CoG?

[peterberetta]

Hi Carbo,

The glider did have balast but it was still in the balast tube when I picked up the bits. The balast tube was still securely glued to the inside of the fuselage. It therefore could not have moved when I launched. The glider was still set up exactly the same as it was two days before when I had several flights in similar conditions. This is also a slope I know well.

See picture of crash site attached.

[jhsa]

If airspeed was high, the plane wouldn't be stalling, not in this case. There are also high speed stalls but it is a different story In my opinion, with a low airspeed, and giving opposite aileron, you caused the aileron on the inside wing to go down, increasing the angle of attack on that wing to increase, as you modified the profile. It also introduced more drag on that wing. At lower speeds all this becomes more important.

The wind being quite strong, and the plane moving fast, doesn't mean that it is actually flying fast. The indicated airspeed is what matters.
For example, imagine that your plane's stall speed is 10mph.
You have a 10mph tail wind.
So now it looks like your plane is flying at 20mph, so ground speed is 20mph.
But actually the airspeed of the plane related to the mass of air it is flying in is actually 20mph (ground speed) - 10mph (tail wind) = 10mph airspeed. That means that despite your plane appearing to be moving faster, is actually just about to stall

I can't express myself very well in English, sorry

João

[peterberetta]

Makes sense, thank you João.

Better to accept blame and learn from ones mistakes.

[jhsa]

Flying a model is actually more difficult than flying a real aircraft. You don't have the instruments to look at, specially airspeed indicator and angle of attack.
Sometimes it happens that pilots inside the aircraft, and with instruments stall it by mistake, so a model without any instruments that we fly just by looking at it would be much easier to stall by mistake..

So, it is one of those things in our hobby that we need to learn to deal with

João

[peterberetta]

Yes, that is true. I actually have an airspeed indicator, gps and vario that I often have in this plane. Such a pity I took them out. Will have to re-install them and try to simulate what happened without crashing (hopefully).

[peterberetta]

Well, one good thing, I have now learnt that Ersky exists - what not aware of this before. So will have some fun trying it out.

Does Ersky replace the LBT firmware? or is it just the software that runs on the TX for setting up models, telemetry etc?

[jhsa]

Ersky9x is another open source firmware for the radios just like openTX..

Something else. GPS IS NOT an airspeed indicator. It indicates the ground speed. No good for flying.. It doesn't take the wind speed into account.

João

[peterberetta]

Good point about the GPS. Thanks.
I guess that also means that when people use a radar gun to measure the speed they reach during dynamic soaring, it is ground speed that is being measured, not airspeed.

[Daedalus66]

Yes, because the radar operator is standing on the ground.

More accurately, it's one component of the ground speed. If the model is coming towards the operator or going away, the radar measures its ground speed pretty well, but if it's going across and the radar is looking at its side, the measured speed could well be zero or negative. With a radar gun, positioning is everything.

This doesn't even take into account wind speed.

[peterberetta]

I have purchased an external XJT module for my Taranis X9d+ and initial testing indicates an average RSSI of about 10dBi higher using the external XJT. Purchased an RX8R which I will try out with the X8R and two battery packs. Ordered the new hall effect M9 gimbals as well. All I need now is to upgrade my flying skills

I was wondering if it makes sense to use the RB10 and have both internal and external XJT modules active i.e. one bound to the primary X8R and the other bound to the secondary X8R.

Also changing my eneloop 4.8v battery for a 2/3AF NiMh that allows a higher current draw in case needed. I have two servos per aileron to avoid flutter at high speed (DS). Although my logs don't indicate much of a battery voltage drop, I have read that eneloop packs can cause a problem due to their low current discharge capacity vs standard NiMh packs.

Trying to figure out where to place the second receiver antennae to optimize reception if the first receiver loses signal or locks out.

Not sure if all these redundant configs are necessary but it is interesting to test them out.

Any advise welcome.

[peterberetta]

Managed to put my plane back together and flew it yesterday with the same X8R receiver. This time I turned up the volume so I could hear the RSSI reporting low signal. Previously I had my sound off. I noticed that as I flew near the area where I lost control of my plane, the RSSI dropped to below 35dBi and my radio was reporting signal critical. I brought the plane a bit closer until the signal strength returned and simulated the stall that occurred previously. I was easily able to recover which got me thinking that I probably lost connection to my plane when I crashed. The logs record input from the sticks on the TX, not what the receiver is acting on, so they can be deceptive when interpreting what commands the plane actually received. My failsafe had been set to hold current position which in my opinion now, is actually what happened why the plane did not respond to any inputs.

I did do a series of range checks before flying with the plane in different orientations. In the worst orientation I was only getting 20 paces. In the best orientation, about 50 paces.

I have changed my failsafe now to full crow breaking just so that it will be clearly visible if the model goes into failsafe mode.

I was flying with a new external XJT module which gives a slightly stronger signal, so I think the internal XJT must have been even worse. I suspect that either my RX antenna's are damaged or the RX is damaged. I never used to have such severe range issues previously and I am surprised that the range check is not worse given the poor range in normal mode.

[Carbo]

I did do a series of range checks before flying with the plane in different orientations. In the worst orientation I was only getting 20 paces. In the best orientation, about 50 paces.

Fine that you were able to fly the glider again! Your log showed a pretty good RSSI, when it simultaneously showed your effort on elevator input to recover the plane. From my pov there was no lost connection. The RSSI was far away from a RSSI warning in that situation. But i am only guessing because of my experiences flying my gliders.

My rangetests with X8R give always an RSSI of about 45-50 at 50m distance, failsafe at 200+ m (free LOS). There is someting wrong now with your setup. Attached is a flight with a glider (with X8R) in 400m altitude and 700m 3D distance and an additonal 'successful transmitted frames' index sensor developed by Tadango https://www.rcgroups.com/forums/showthr ... nk-quality
There is always a bullet proof connection.

I would continue with rangechecks and look for defective hardware and for a better RX antenna installation. From my pov redundant installations are much more complex with additional error possibilities compared to a clean single RX installation. Does your glider have a carbon fuselage and wings? But nevertheless, when it crashed, the RSSI was in an unsuspicious range. Hopefully you will find the reason for the crash and gain confidence again in your equipment.

[peterberetta]

Hi Carbo, thanks for your response.

My glider is 100% carbon (fuz and wings) and this is definitely a contributing factor. See photos of the antenna installation. I have a 60cm extension from the RX. (I would NOT normally tape the antenna like this but did not have clear tape available. I actually have a clear plastic case for the antenna but decided not to use it for this test)

Your RSSI looks a lot better than mine from yesterday (see attached). My RSSI drops as soon as the plane is launched. I probably did not exceed 200m for the entire flight.

I am in the process of reverting from the EU firmware back to the international version and using the latest RC16 build of 2.2 for some testing. Also going to swop RX.

[bob195558]

I do not like taping and would be better to have the antennas more away from your model.
I have been using the plastic wire-ties as a support for the antennas and Velcro to hold them to there proper place (90 degrees apart).
So when (if) a crash happens the antennas are then not ripped from there receivers.

Bob B.

[MikeB]

Looking at where you have the antennas, it seems you have them on the top of the model. To my mind, this means when it is flying they will be somewhat masked by the fuselage if it is carbon.

Mike.

[peterberetta]

I have swopped the receiver for a new RX8R and connected an XMP+ as backup. Also moved the antenna. The failover works well and the range test is is around 50m. Have yet to fly the plane like this but I am sure it will be far better given the range test results.