DJI - WHERE ARE YOU?

RémyHeli · Jun 20, 2017

Thanks for feedback! These blades have fix pitch so logically high RPM does increase lift which is contradictory to reported fast descend speed. Also flight controller does limit descend speed to avoid VRS in normal conditions but not the case here.

Also excluding air density problem explaining motors to compensate assuming not flying over area with high temperature or high altitude.

Maybe one ESC hardware issue causing the three other motors to compensate which is very difficult to recover with quad configuration.

Just a guess seing high RPM you mention.
Have a nice day.

wa5tef · Jun 20, 2017

High RPM and excessive rate of descent? How can that happen unless the drone is inverted, like mine was? Our crashes sound similar in the fact that DJI never explained how it was possible that one of "their" drones messed up! At least in your case, DJI admitted it was not pilot error. In my case they never admitted that my crash even happened. So, was your drone inverted? This would explain how it was able to a rapid descent and high RPM at the same time?
Thanks,
Jim
WA5TEF

RémyHeli · Jun 20, 2017

Yes agee if it's FOD not mention so far.

wa5tef · Jun 20, 2017

What is FOD?
Jim

Jammer · Jun 20, 2017

As I mentioned the rotors were in vortex ring state.

When the blades are in a stalled condition the RPM increases because the aircraft sensors are trying to arrest the high descent rate.

You can add all the power you like but once in VRS power only makes it worse.

Jammer · Jun 20, 2017

wa5tef said:
What is FOD?
Jim

Foreign Object Damage

RémyHeli · Jun 20, 2017

Flip Of Death

RémyHeli · Jun 20, 2017

Ok so if craft was in VRS mode and he not flipped over, it mean that at some point he entered in non recoverable VRS state - which should not have happened as VRS managed by flight controller - this was maybe originated by abnormal command to descend and out of range.

Jammer · Jun 20, 2017

More on VRS

A vortex ring state can occur when you do a vertical descent. While descending, your propellers lose the ability to produce lift because they are traveling directly through the turbulent downwash they are designed for to create.
This might sound strange, propellers are designed to create this downwash, yet this downwash can harm the drone as well.

But there are ways to avoid it, and if it can’t be avoided, there are ways to attempt to recover from it.

First, the recovery:
To recover you need to get your drone back into clean (non-turbulent) air.
To get in clean air, you need to pitch forward or move sideways to get out of the propellers downwash.
It is also good to reduce throttle input, more power is more turbulent air, less power is less turbulent air. less turbulent air is more manageable to get out of.
Sometimes the air you’re descending through can be so turbulent that you will lose complete control. This near irreversible, and will almost always result in a crash. That’s why it is so important to not get in a VRS.

How to prevent:
One golden rule it to decent SLOW! a slow descent will cause little turbulent air, this will reduce the chance getting in a VRS.
Sometimes you need to descent fast because you are running out of battery, maybe try to prevent that the next time you fly. don’t try to overrule the automatic safety measure… IF you need to descend fast, do this in a forward/backward movement, or sideways, or a combination of that.. generally everything other than straight down.

REMEMBER: NEVER APPLY MORE POWER WHEN IN A VORTEX RING STATE, you will only make things worse!

RémyHeli · Jun 20, 2017

I have posted a vid in this forum with VRS demo. You can find it. With P4 you can apply full stick down and he will not descend quickly, he will control until little wobble prior VRS. To increase descend speed we apply some translation or doing circles.

Jammer · Jun 20, 2017

Under normal circumstances the P4P has a VRS limit in the firmware which prevents a high descent rate that would induce VRS. If I could have moved forward I could have manually escaped but as this post mentioned when I started, I didn't have control.

RémyHeli · Jun 20, 2017

Thank you Jammer. Fully agree.

wa5tef · Jun 20, 2017

That is indeed a strange sequence of events. I have made over 250 flights with this Phantom 4 and probably 150 with the Phantom 4 that powered upside down into the ground, and I have done many fast descents into landings and other fast descents just to lower altitude and never have I had the propellers get into a stall configuration or a vortex ring state. I do frequently see the drone wobbling somewhat due to the prop wash but the Phantom 4 is very stable and capable of almost immediately correcting itself should it ever get out of level even during a fast descent. I ran into all this stuff back in February when I was trying to prove to DJI that my old Phantom 4 did indeed flip over, upside down, and fly under power into the ground! DJI never, and I mean not even once, admitted that this was possible even though I had cell phone video made within 15 seconds of the crash. This video showed the Phantom lying on its propellers, or what was left of them, upside down, still powered on and by the audio interference still transmitting telemetry. What the video also clearly showed was a dent where the camera lens ring smashed against the bottom of the drone. DJI again never, never admitted that this happened, only that it must have been "pilot error" because a Phantom 4 can't perform that move!

I am not doubting that the drone fell out of the sky but like mine, no one else has ever seen that happen ?????

Thanks for your information,
Jim
WA5TEF

Jammer · Jun 20, 2017

wa5tef said:
That is indeed a strange sequence of events. I have made over 250 flights with this Phantom 4 and probably 150 with the Phantom 4 that powered upside down into the ground, and I have done many fast descents into landings and other fast descents just to lower altitude and never have I had the propellers get into a stall configuration or a vortex ring state. I do frequently see the drone wobbling somewhat due to the prop wash but the Phantom 4 is very stable and capable of almost immediately correcting itself should it ever get out of level even during a fast descent. I ran into all this stuff back in February when I was trying to prove to DJI that my old Phantom 4 did indeed flip over, upside down, and fly under power into the ground! DJI never, and I mean not even once, admitted that this was possible even though I had cell phone video made within 15 seconds of the crash. This video showed the Phantom lying on its propellers, or what was left of them, upside down, still powered on and by the audio interference still transmitting telemetry. What the video also clearly showed was a dent where the camera lens ring smashed against the bottom of the drone. DJI again never, never admitted that this happened, only that it must have been "pilot error" because a Phantom 4 can't perform that move!

I am not doubting that the drone fell out of the sky but like mine, no one else has ever seen that happen ?????

Thanks for your information,
Jim
WA5TEF

I'm sorry that you lost the aircraft under those rare circumstances. After experiencing a failure like that, every time you fly you keep wondering "what will happen this time". I am VERY careful and always on guard as one should be.

Interestingly, CASA (our federal regulator) has compiled data for the last 15 years on the failure points within "RPAS" (or drone) commercial and domestic operations and unlike CAT (Commercial Air Transport) the report signifies that the biggest failure point is the technology and NOT human factors or "pilot error".

This is new territory for them. Up until now they have focused on "operator" failure points but the data is showing a different story.

The whole tablet, apps, firmware, SDK is all very ad hoc in my opinion. Pilots can use an iPad in the cockpit (we all do) to supplement navigational tasks but it will always be a "Class B" navigation tool simply because of the firmware.
Apple firmware is pretty good on the IPad but it will never stand up to Class A avionics.
When we talk about DJI firmware stability we are in a completely different class. It is VERY shaky.

Just ask the programmers that wrote the Autopilot software... they will tell you (behind closed doors) that it's bloody awful.

J

wa5tef · Jun 21, 2017

Jammer, yeah I agree with you. DJI is not known for its reliability on firmware, especially the new releases to fix known bugs, that actually open up new unknown bugs. I switched to Litchi in November and have since not had an issue with the flying of the Phantom other than the crash. But, DJI sent me a new Phantom 4, brand new! But still would not tell me what had happened to the old one. Go figure!
Good luck, let's all get back to flying and have some fun. Yeah, I am aware it could happen again, even the next flight, but I keep programming 3.5 mile flights that take 22 minutes over unrecoverable terrain at time, but it is FUN!!!

Thanks,
Jim
WA5TEF

RémyHeli · Jun 21, 2017

I think I got why FOD is extremely rare or 'almost' impossible since DJI introduced P4.

DJI designed dual IMU hardware and more redundant processing to avoid flight controller 'give up' if unable to deal with unexpected inertial data received. That redundancy concept which is mandatory with real aircrafts.

Case where both IMU's would fail exactly at the same time is extremely low but not impossible on my views.

B25J · Jun 21, 2017

This thread developed into a very interesting learning experience for me. Thanks guys for carrying it further along with your personal experiences and logical thought.

B25J Out[emoji106]

pomonabill220 · Jun 22, 2017

It is sure nice to read some interesting and well informed posts as to what is going on in the Phantom processor.
Also, the great explanation of VRS and how to "get out of it" when it occurs.

It is sad though that DJI doesn't let us know what happened when a bird is repaired and what they found... seems like they are hiding something or don't want to tarnish their name by revealing a failure.
Even auto manufacturers put out TSB's, although it does take a while to hear what happened.

Thank you for your input and knowledge. Sharing information like this really helps me try and understand this technology!

I get so sick and tired of these "right fighters" that come here and just stir the pot.. doesn't add anything to the conversation that is useful!

B25J · Jun 22, 2017

pomonabill220 said:
It is sure nice to read some interesting and well informed posts as to what is going on in the Phantom processor.
Also, the great explanation of VRS and how to "get out of it" when it occurs.

It is sad though that DJI doesn't let us know what happened when a bird is repaired and what they found... seems like they are hiding something or don't want to tarnish their name by revealing a failure.
Even auto manufacturers put out TSB's, although it does take a while to hear what happened.

Thank you for your input and knowledge. Sharing information like this really helps me try and understand this technology!

I get so sick and tired of these "right fighters" that come here and just stir the pot.. doesn't add anything to the conversation that is useful!