labroides@yahoo Posted at 2016-3-31 00:33
"The stats prove it. The fact people have CSCed their bird in the air with such a high frequency s ...

I am guessing there is a systemic problem of people accidentally hitting CSC or DJI would not have changed the CSC sequence on the P4.  The old adage "if it ain't broke don't fix it" applies.  

Nink Posted at 2016-4-1 00:22
I am guessing there is a systemic problem of people accidentally hitting CSC or DJI would not have ...

Or since there wasn't a problem to fix at all, it was DJI caving in to the crowd that has been squealing every time CSC is mentioned.

labroides@yahoo Posted at 2016-3-31 09:25
Or since there wasn't a problem to fix at all, it was DJI caving in to the crowd that has been squ ...

Well either DJI has a quality control issue or people keep doing CSC but drones keep falling out of the sky. You decide?

It looks like it was more prevalent on the P2's as there was lots of ways to CSC (any down combo) but they appear to reduce it on the P3 to 2 and changed to RTH on P4

Nink Posted at 2016-3-30 22:01
I think there are 3 but maybe there is 4 not sure.  Guess I should fire up tutorial and test them  ...

Dear lord.  Everything now fires off the CSC.  What in the bloody @##$%!!

Morph1 Posted at 2016-4-1 01:24
would you agree though that changing the execution method of this command would eliminate the chanc ...

> ... I was using both sticks to maneuver and I realized that I was pretty close to CSC as I was descending and panning at the same time flying diagonal ...

Except you would need to be
1) descending at maximum rate AS WELL AS
2) flying at maximum speed to the left AND
3) flying at maximum speed to the rear  AS WELL AS
4) rotating at maximum yaw to the right

ALL AT THE SAME TIME and holding this maneuver for a couple of seconds.  I'm sure that is long enough to do at least one complete rotation.

It would look like you are falling out of the sky in a death spiral if you were to DELIBERATELY perform that combo.  Who in their right mind would deliberately want to fly like that?  With such a violent maneuver you couldn't even tell the motors had stopped, it would already look like it was out of control.  If that was going to ruin your fabulous artistic shot and you insist on flying that way, I'm sure you could just release and CSC to restart and take complete control again.  It must be completely stable twisting like a tumble dryer in the air.  No harm done then, all good.

Suggesting that the aircraft should be immune from a 10 year old performing violent aerobatics to the limits of the aircraft on their first flight is just silly.  Accidental or not, they are going to crash if they fly it like that.  Much the same as if I turn the steering wheel of my car full left while at speed on the freeway and smash into the wall -- who could have seen that coming?  What were the designers of the car thinking including such a dangerous feature in the steering wheel, who do I sue???

microcyb Posted at 2016-3-31 22:58
Dear lord.  Everything now fires off the CSC.  What in the bloody @##$%!!

To clarify, CSC is the same as it's always been (both sticks bottom inward or outward as it states in the manual)
BUT the P4 is the ONLY aircraft that has the new CSC command of left stick bottom right and RTH button at the same time (mode 2).
And the P4 will ONLY CSC with that command, (the P3 CSC commands will NOT CSC the P4).
And the same as the P3, it will NOT CSC if the P4 CSC command is given.

Nink Posted at 2016-4-1 01:14
Well either DJI has a quality control issue or people keep doing CSC but drones keep falling out of ...

In the P3 Only A and C are CSC.  B is a left-stick-down motor-off command when it is sitting on the ground.  The left-stick-left and right-stick-back-and left are irelevant and coincidental positions, the motor will turn off SOLELY because the aircraft is not descending (assumed landed) and the left stick is held down.  This WILL NOT happen in flight.  Similar explanation for D

DJI-Ken Posted at 2016-3-31 11:28
To clarify, CSC is the same as it's always been (both sticks bottom inward or outward as it states ...

Is there a plan to standardize this Ken ?  Seems illogical to have different systems.  And as the official  DJI video morph1 posted states the #1 cause of pilot error crash is CSC

DJI-Ken Posted at 2016-3-31 11:28
To clarify, CSC is the same as it's always been (both sticks bottom inward or outward as it states ...

Thankfully, I learned that before my first flight.  I really do feel that the CSC should be an option to be turned off in the advanced section within the DJI Go menu.

Nink Posted at 2016-3-31 23:34
Is there a plan to standardize this Ken ?  Seems illogical to have different systems.  And as the  ...

That is why they changed it on the P4 because it is brand new.
I do not know if it will be implemented into the P3.
If they ever did, it would be in a firmware update and you would probably get a message in the app about the new process.

microcyb Posted at 2016-3-31 23:40
Thankfully, I learned that before my first flight.  I really do feel that the CSC should be an opt ...

It's put there as a security measure, I don't think there will ever be a way to disable it.

Nink Posted at 2016-4-1 01:14
Well either DJI has a quality control issue or people keep doing CSC but drones keep falling out of ...

None of those videos show any evidence of CSC.  They are typically smooth even flight, or slow descent, followed by a sudden violent fall from the sky.  If there were a CSC involved, there would first be fast descent, with fast rotation, with fast backward and sideways movement for a couple of seconds BEFORE CSC kicks in.  All these videos show something else.

labroides@yahoo Posted at 2016-3-31 21:25
Or since there wasn't a problem to fix at all, it was DJI caving in to the crowd that has been squ ...

This. So much this.

In all likelihood, there was no problem with the CSC except among a tiny minority of very vocal people on the forums. So now not only do we have a confusing set of different CSCs for different models, it's barely documented which works on which, people have been tossing mis-information about which one works on which, and it will now probably be changed even more in the future as threads like this one show up.

Good job guys. Good job.

Morph1 Posted at 2016-4-1 00:12
it's simple,  honestly when would you even want to do this as a controlled drop to destroy your air ...

100% incorrect. If the aircraft loses connection then it goes into RTH mode. And a lot of those are crashes due to pilot error of improper RTH settings.

Morph1 Posted at 2016-4-1 03:22
how would you know that if you never seen it happen ?
what other controlled descend would you see  ...

> how would you know that if you never seen it happen ?
what other controlled descend would you see happen when you shut down the motors in mid flight lol

Try and read and follow what I have written in the previous replies. I saw in the videos posted that they were flying smoothly, followed by a sudden out-of-control situation.  IF you input a CSC, what do you expect the aircraft to do during the second or two while it is waiting to execute the motor-off instruction?  You are holding the left stick down -- I expect to see some downward movement.  You are holding the left stick right -- I would expect to see some rotation to the right.  You are holding the right stick back and to the left -- I would expect to see the aircraft move diagonally to the rear and left.   I SAW NONE OF THIS PRIOR TO THE AIRCRAFT LOSING CONTROL.  Feel free to step through the videos and find any such movement before loss of control.  There isn't any.   CONCLUSION: It isn't reacting to a CSC.


> Seriously you are now what an FAA investigator you can tell how the craft would behave in mid air with engines off lol.

I never professed to predict how it would behave AFTER the motors were stopped.  I am talking about the seconds immediately before the motors stopped.  The time it takes from receiving a CSC and the motors stopping.  The aircraft is still receiving directional input (assuming it is able to receive the CSC command continuously for 2 seconds or so).  It will still act on them and change altitude direction and heading during that time.

You don't know anything about me or my credentials.  Don't make assumptions about other people because you will be shown up to be stupid.    You know what they say about assumptions and being an ass...  As a matter of fact I was a crime scene examiner for 33 years and investigated many crash scenes, and attended my fair share of air accidents too.  I know the things to look out for and how to look at the evidence objectively and what questions to ask.  Any other comments?

Morph1 Posted at 2016-4-1 03:12
it's simple,  honestly when would you even want to do this as a controlled drop to destroy your air ...

I have intentionally engaged CSC on two separate occasions while in flight.  The first was when I flew too close to steel girders, which just acts like a giant magnet and draws the aircraft towards it at great speed, as the internal compass is screwed up.  Rather than risk an impact with steel, I elected to drop it to the ground onto soft grass.  Fortunately it was only a meter or two.

The second time I was checking out the guttering on a pre-cast concrete building.  Again the compass screwed up , likely due to cyclone fencing and rebar in the concrete structure.  Again the aircraft headed for the immovable and inflexible wall at speed and wasn't responding to opposite direction movements or decrease altiude commands, so I dropped it onto a (soft) plastic corrugated verandah roof rather than strike concrete at speed.  Fortunately on both occasions there was ZERO damage, but there would have been had I let it continue into a hard object.

I've seen youtube videos of large birds of prey snatching aircraft out of the sky.  If I see a large bird attacking, I'd prefer to drop the aircraft and at least recover the spare parts and SD card, rather than have it lost forever in a birds nest somewhere.  There's also the possibility of a police or rescue helicopter popping up unexpectedly, and I'd gladly sacrifice my aircraft to avoid a fatality.  10m/s or so is not fast enough to escape or avoid some hazards.

If you're going to quote the case of... "that time I heard one Phantom flew away, or flew up into space, or dropped from the sky", it would help if you actually provided a reference and a link to some EVIDENCE that supports it, so it can be examined and refuted if necessary.  Much like saying certain videos show CSC gone wrong, where they actually DON'T support that view.  It goes a long way to your credibility.

endotherm Posted at 2016-3-31 11:50
None of those videos show any evidence of CSC.  They are typically smooth even flight, or slow des ...

Here is a CSC. Where is your "fast descent with fast rotation with fast backward and sideways movement for a couple of seconds BEFORE THE CSC kicks in"  ?

I don't think the sticks needs to be fully 100% left and right stick down and inwards.  100%. I think once it reaches a certain threahpoint (<100 percent say 90%) it still counts. But yeah. Csc does happen more often then ppl say or believes .

Morph1 Posted at 2016-4-1 01:07
Ken I agree with you, but that's not the case with the p4 that just climbed to 10,000 ft and disap ...

Obviously there was an issue there. And since it was crashed into a tree earlier, I'm sure there was some  internal damage that caused it not to RTH. We will review the logs to find out exactly what the problem was.

Nink Posted at 2016-4-1 04:40
Here is a CSC. Where is your "fast descent with fast rotation with fast backward and sideways mov ...

That's an interesting video. There the CSC seems to be acted upon almost immediately.  My aircraft does not, and takes a moment (a count of two or three) before it shuts down.  Certainly 3 seconds for a left-stick motor shutdown.  Is that a P2 in the video? Certainly a different controller, not sure if that has something to do with it?  I know the P2 and P1 had different characteristics in their firmware and perhaps different features (I think I read Intelligent vs immediate mode somewhere).  That could explain a longer shutdown time in my later model.

Anyway my point is still valid.  If you watch when the sticks are moved into position, you will see the aircraft react to the input and start to roll CCW and tip in response to the move-left command.  After the motors slow to a stop it tips the other way and falls.  This behaviour is confirmed in the subsequent test.  In this example the shutdown is so immediate the aircraft doesn't have a lot of time to react to the yaw and descend inputs before falling out of the sky.  In my model it would certainly have more time and further movements would be a lot more obvious before the motors stopped.  The author of the video does concede it would be pretty stupid to fly with those inputs.

My comment about fast rotation etc. was more in response to Morph1 saying that he was "already pretty close to CSC", in which case his aircraft would have been already spiraling to the edge of loss of control.  Had he gone further, it would merely have added to the spin before cutting the motors.   The linked videos don't "prove" a CSC occurred.  The last one, as per the title, lost all power and fell from the sky at the same time the recording stopped.  Still no evidence of any stick input.  The one with the overview of the soccer field shows a lone pilot establishing his shot before testing a second aircraft.  It's safe to assume he put the controller down on the ground (as seen shortly after impact) at the time it fell from the sky.  Again no evidence of CSC being applied.  Just a series of aircraft falling from the sky for some unknown reason.  I still believe you will see some evidence of the aircraft trying to follow the stick inputs before a CSC event, even if it is just a twitch or roll to the left.  The amount of reaction will depend on the delay between recognizing and reacting to a CSC.  Perhaps in the P2 and earlier it only needed the sticks in that position for a millisecond before it reacted.  Can you say that the engineers haven't increased that to needing say 500ms in that position before it reacts?  I can say with certainty my aircraft takes much longer than the videos to CSC and that's plenty of time to react to turn/roll/descend inputs.  Show me P3 tests with the same results and I will concede.  I haven't seen any videos of P3's falling from the sky, perhaps DJI has already addressed the CSC sensitivity issue if that was indeed the problem?

Morph1 Posted at 2016-4-1 05:45
Holy Crikey,
If you now tell me that you also traveled to the space station and spent 2 weeks servi ...

I'm not an astronaut, never claimed to be.  Clearly there is no way to prove to some random 12 year old idiot on the internet that you actually have expertise in a field.  I obviously don't exist because you said so.

You are wrong.  DJI don't claim it is the #1 reason, it is one reason amongst a list of 10 in no particular order.  Have a look at this thread.

In my experience the response is not instant.  Just because you feel that way or others believe something, it doesn't necessarily make it so.  You keep citing all this knowledge and experience (which seems to be better that everyone elses), and your research has found some holy grail of DJI statistics showing it happens with great regularity, but you can't share a link to this smoking gun of proof.  A cynical person would call bullshit on this, and accuse you of making it up to support your position to win an internet argument.  Real mature.

Call me stubborn if you like, I have my own opinions and won't be bullied by someone with a different one.  You may be surprised to know I am open to opposing or conflicting views, and I can be swayed with actual, you know, proof.  I can't find evidence that it is such a dire condition that occurs with monotonous regularity, and you haven't provided any.

endotherm Posted at 2016-4-1 06:37
That's an interesting video. There the CSC seems to be acted upon almost immediately.  My aircraft ...

G'day endotherm ... I agree with everything you've said except that CSC is instant.  No 2 or 3 seconds.
Try it on the ground with no props.

Morph1 Posted at 2016-4-1 01:24
would you agree though that changing the execution method of this command would eliminate the chanc ...

" this one straight from DJI lol
Top 10 Most Common Pilot Errors;
Number 1, quote - " New Pilots often crash their drones by pressing a combination ..... blah blah blah...."
this comes from DJI "
"It is documented with DJI statistics that this happens fairly often among new pilots, what other proof do you need "

And it's very wrong and not at all backed up by any evidence.
It's been debunked many times as DJI's incompetence in communications.
http://forum.dji.com/forum.php?m ... 8&fromuid=10818

So Ken says that the evidence isn't there in the crash repairs.
The evidence isn't here in the reports of flyers accidentally hitting CSC while flying.

labroides@yahoo Posted at 2016-3-31 16:35
" this one straight from DJI lol
Top 10 Most Common Pilot Errors;
Number 1, quote - " New Pilots o ...

Then why did Did DJI bother to put out a video saying it and put on the list number 1 CSC  pilot error are one of the main reason for crashes based on examining black boxes. If it is not a problem they would have left it off the list in the video   Also if it is not a problem why did DJI change the CSC command from P2 (4 different dual joystic commands) to P3 (2 different dual joystick commands)  to P4 (1 Joystic and RTH command).  

Finally why on earth does Endotherm just keep making stuff up like his Phantom takes 3 seconds to do a CSC and when it does it first will do all sorts of crazy maneuvers that would be extremely dangerous in an emergency when you actually would want to use it.   

I give up.  

Nink Posted at 2016-4-1 07:58
Then why did Did DJI bother to put out a video saying it and put on the list number 1 CSC  pilot er ...

Because they are incompetent.
They make great machines but their communications are sloppy.
http://forum.dji.com/forum.php?m ... 8&fromuid=10818
If accidental CSC was the number one cause etc, there would be almost daily reports of it happening.
As it is, the only reports of it are very infrequent and on investigation they usually aren't normal flight at all.
To be flying in a way that you are at risk of accidental CSC, you are descendinging in a max speed backward-sideways spiral at the same time as travelling backwards-sideways at max speed.
And that isn't a normal flight move.
If you are flying in a way that you risk that, you are flying like a monkey on meth and you'll soon find a way to crash CSC or not.

labroides@yahoo Posted at 2016-4-1 07:17
G'day endotherm ... I agree with everything you've said except that CSC is instant.  No 2 or 3 sec ...

OK, I'll wear that.   After testing it, I'll agree it is not 2 or 3 seconds although it is nearly the same time that a left stick down command takes (I thought I read hold it down for 2-3 seconds for that somewhere, so I guess that's why that figure stuck in my head). My left-stick-stop is around 0.75 seconds

It definitely isn't instantaneous though -- I can't activate it by rotating the sticks in slow circles and making momentary contact in the inner/outer positions.  I can count out loud one-two-three, with it stopping on three.  That's got to be half a second at least, and I'm sure one of the sites that analyse flight records could report on motor RPM after a CSC and I could provide an exact time.  It would save having to set up a tachometer and try to coordinate and measure things. If I spin up the motors hard before the CSC it will run fractionally longer before stopping, and will resume running if I release it on a count of "3".

Listening to the pitch of the motors at idle, then moving to CSC shows an increase of speed, supporting my claim of the motors are acting on left/right stick input.  Again this can be proved with flight record analysis.

So I definitely have at least half a second of control input before the CSC kicks in and the motors shut down.  This is sufficient time to measurably alter course/heading/altitude.  (I apologise if I exaggerated the duration and extent).  However, my contention stands that if CSC is responsible for shutting down an aircraft in flight, there will be a telltale change of course/heading/altitude immediately before the shutdown.  In all fairness the test should be repeated in the air in case CSC behaviour on the ground isn't the same as in flight and it is solely acting on the left-stick-down stop.

I should point out my unit is crying out for an IMU calibration as it is taking over a minute to "warm up" prior to letting me use it.  I don't know if that would affect the timings at all.

P.S. If I compare my aircraft to the P2 in the freefall video, mine doesn't shut down until the P2 has fallen halfway to the ground.

Bet most accidental csc is during flying ioc.  No one in their right mind would fly in a way that would induce accidental csc . But I'm willing to bet most accidental csc is when flying using homelock/poi/course lock etc rather when flying solely on the sticks

Nink Posted at 2016-4-1 07:58
Then why did Did DJI bother to put out a video saying it and put on the list number 1 CSC  pilot er ...

> Finally why on earth does Endotherm just keep making stuff up like his  Phantom takes 3 seconds to do a CSC and when it does it first will do  all sorts of crazy maneuvers that would be extremely dangerous in an  emergency when you actually would want to use it.   

Sorry, I never said mine takes 3 seconds.  I did make references to one to two seconds.  I did admit further down that I overestimated a bit.  The 3 seconds was mentioned for the left-stick-stop only.  In fact that was closer to 3/4 of a second.  CSC is closer to half a second in most cases.  (Sorry, I've been up all night arguing on the internet.)

In response to the "crazy maneuvers", tell me why it wouldn't do these things. You are pushing the sticks fully left/right and down.  What happens when you perform ONE of these actions in flight? The aircraft performs the maneuver. Now stack 4 on top of each other.  Result = crazy maneuver followed by a motor stop.  That probably strengthens the argument not to have CSC on a flight control if anything. I agree it probably is dangerous in an emergency too.

After testing, the minimum input would be about half a second of crazy commands.  From an idle position, such as the drop video, there's not much to see, but you do see a response as it twitches and tips CCW before falling the other way.  If the CSC took longer, the result would be worse and more noticeable.  Fortunately it isn't.  But stack that onto a flight where the pilot says they are already close to a CSC in normal flight (imagine lurching and twisting aerobatics already), then add half a second of full on crazy, it will be bucking like a bronco as it goes into CSC and probably fall to earth in a flat spin.  Hopefully I don't come across as too crazy now

In the tests I did yesterday, any VALID CSC command, that is both sticks down toward the centre or down towards the outside, always resulted in a momentary speeding up of some, but not all, the motors. This would tend to suggest that the aircraft sees this as a valid command sequence BEFORE it recognises it is a CSC. This would then account for a couple of effects, the first being that the aircraft goes into a tumble at the point of CSC, and not simply drop straight down, as the motors tip it over. The second effect would be when the aircraft tips over on the ground when a CSC command is used to stop the motors rather than the simpler left stick down command.

I think that DJI should remove the CSC method of stopping the motors on the ground from the manuals, as it leads to damaged aircraft, and strongly emphasize the use of the left stick down method.

And this:

it's simple,  honestly when would you even want to do this as a controlled drop to destroy your aircraft flying along boeing 747 ?

There are actually a number of control stick movements you could perform in a B747 that would screw the aircraft into the ground. And that goes for almost every type of aircraft that is not purely fly-by-wire. Fortunately, most real pilots are trained in how to recover from those situations. Many Phantom pilots however, do not have a clue, and when confronted by an apparent 'runaway', rattle the sticks around in frustration. Inevitably that can lead to a CSC, and in those cases it is always the fault of DJI.

Morph1 Posted at 2016-3-31 02:32
sorry but where were you during the killswitch thread where DJI determined that a drone was droppe ...

Its futile arguing with a mindless drone. Try convincing an apple fanboy that apple products are overpriced.

lido_bmt Posted at 2016-3-31 11:57
This. So much this.

In all likelihood, there was no problem with the CSC except among a tiny mino ...

Nah.

DJI instruction video teaches you its a problem.

When I first got my phantom, I remember what stood out to me was the method of CSC using the joysticks. That just screamed bad idea from the get go. This is a design/ergonomic/engineering issue. They can fix this. It shouldn't be like this because it is illogical and non-intuitive.

Its like selling you a vehicle and telling you to floor and hold the gas peddle for a few seconds to activate the breaks on the car. That is just stupid all around.

Joysticks are for motion. This must be left completely separate from commands to turn the thing off. Even if in your ivory tower, you can't possibly see people doing those joystick commands in a real world maneuver, the fact is some people might want to do that for acrobatic reasons (fast spiral down), or may do the commands inadvertently trying to move around.

There should be no ambiguity when shutting off motors.

quaddron3 Posted at 2016-4-1 15:10
Its futile arguing with a mindless drone. Try convincing an apple fanboy that apple products are o ...

'Its futile arguing with a mindless drone. Try convincing an apple fanboy that apple products are overpriced.'

In general this is a forum for discussion, do you think you could enter a concversation without leading with your chin? Omly people with weak arguments resort to labelling people and insults.

quaddron3 Posted at 2016-4-1 15:14
Nah.

DJI instruction video teaches you its a problem.

You aren't paying attention.
That stupid DJI video is way off and Ken said as much.
If accidental  CSC is so common, why is it almost never reported and why does Ken say that they hardly ever see it in Phantoms sent for repair?

Your whole argument hangs on a myth that can't be backed up with any facts.
It's the same bleating that gets wheeled out every couple  of weeks by newbies that imagine accidental CSC is a real risk.
Experienced flyers know it isn't but beginners like you will  keep going on and on with this crap while DJI's incompetent communications aren't smart enough to fix their stupid and baseless top 10 video.

Until Dji gives us something concrete then how can anyone assume if csc is TOP 10 newbie mistakes or not .  It's not like they made a video NOT saying csc is a major drone crasher.    Dji ken works the forums not in actual quality control / customer service . You can best bet that they keep ken and Amy and Tim and the rest of the Dji customer service separated and ill informed just like how they do their customers .so much is kept secret and swept under the rug who knows what's really going on .     Hell the csc wasn't even listed on the change log for the p4 nor p3 when it did indeed updated csc during firmware upgrades .

mal6514 Posted at 2016-4-1 16:02
Until Dji gives us something concrete then how can anyone assume if csc is TOP 10 newbie mistakes or ...

"how can anyone assume if csc is TOP 10 newbie mistakes or not "
By reading forum reports of flyer troubles you get a pretty good idea of what is responsible for crashed Phantoms.
There are  loads of disoriented owners flying into trees and buildings, there are plenty that fly downwind in strong wind conditions, there are loads that fly behind obstacles etc,  etc. etc and accidental CSC is so rare that it doesn't even rate as a cause for crashed Phantoms.
And that's because you have to yank the sticks around like a monkey on meth to accidentally CSC - it isn't something that happens in normal flight.