Phantom 2 Catastrophic Failure after Firmware 3.06

[Jaybee]

Interesting and confusing. Thanks for taking the time to attach it.. I noticed the P2 start losing altitude at approx. 15-20 feet up. At first it did respond to throttle but very very little response. It kind of went in stages from then on up until it was about 3 feet off the water when it dropped. I forgot to mention earlier that when I realized that there was a problem, probably around 10 feet. I began to turn towards Home to bring it back.

It is very interesting isn't it. From reading about the Fresnel zone, it looks like there is a strong probability of false / out of phase signals over water. If you consider that the 2.4GHz waves are essentially light, it's not unreasonable to expect some reflection will happen from a water surface. This sounds particularly relevant if you're fairly close to the surface:

...the prop wash creates a chaotic wave pattern which generates a myriad of false-signals.

Other factors would be the angle of the TX signal to the RX. I would assume the higher the craft is above water the less likely the signal will be out of phase as the strongest signal will be in the center of the Fresnel zone, the edges of which would be further from the water, especially if you are at, or around, sea level.

I feel your pain anyway man, gutting loosing your kit like this. For your next setup, I'd definitely recommend getting yourself a DVR and also switch off that GoPro OSD, it's just unnecessary confusing clutter on the screen. I grabbed a very inexpensive DVR a month or so ago, purely so I'd always have a record of the telemetry info. This is handy for fault diagnosis and also recording operational distances etc. Also, if the RPA went rouge, you'd have an instant replay-able record of it's last location. Great for added peace of mind.

 

[Navman]

battery voltage is difficult to read on a FPV screen with iosd mini and the Gopro recording. The Red Recording indicator get's overlaid right over the Battery Voltage reading.

Wish they would fix this as it's meant to fly with the GoPro

 

[Navman]

Amen to that! Very frustrating for those of us that don't think in metric!

And sorry about losing your machine!

Briefly off topic but this is a prime example of why I want my iosd to show me my flight measurements in Imperial and not Metric.

For me, the only reason to have an OSD is to have your pertinent flight data in front of you. I have flown aircraft and have been a Yacht Captain on 60'-100' Vessels. Without your instruments and Nav info in front of you so you can routinely scan them and know what is going on, you are basically running blind.
Pretty much all instruments and Nav equip show imperial units and the option is there to show Metric.

My Brain was trained in Imperial units of measurement. When the **** hits the fan. Like it did. The last thing I need is to be converting Meters to Feet so I can visualize what is going on.I know what 10 feet or 15 feet looks like. I know what mph looks/feels like.

Sorry for the off topic rant but it might have given me a quicker awareness of what was happening.

 

[Jaybee]

battery voltage is difficult to read on a FPV screen with iosd mini and the Gopro recording. The Red Recording indicator get's overlaid right over the Battery Voltage reading.

Wish they would fix this as it's meant to fly with the GoPro

Who would fix what? Just turn off the GoPro OSD - it's in the GoPro settings.

 

[Navman]

battery voltage is difficult to read on a FPV screen with iosd mini and the Gopro recording. The Red Recording indicator get's overlaid right over the Battery Voltage reading.

Wish they would fix this as it's meant to fly with the GoPro

Who would fix what? Just turn off the GoPro OSD - it's in the GoPro settings.

It's tells me if I forgot to turn on recording rather than spend 10 minutes getting the perfect footage only to land and find out I took a stupid pill first thing in the morning

 

[Jaybee]

It's tells me if I forgot to turn on recording rather than spend 10 minutes getting the perfect footage only to land and find out I took a stupid pill first thing in the morning

We've all been there ; ) Just make sure the red light light on top of the GoPro is flashing before you take off.

 

[ToddSmi]

8mins in the air and was still at 78%? that doesnt make much sense... Also what was your voltage when it started to descend? Did you use the old default gains or the new ones?

8Mins was what I was guessing from the time I put the battery in and took off. That was a rough guess.
78% Battery was a definite.. As it started to descend I immediately checked the numbers. Battery Level was 78%, GPS was 12 Sats. but battery voltage is difficult to read on a FPV screen with iosd mini and the Gopro recording. The Red Recording indicator get's overlaid right over the Battery Voltage reading.

I was using the Default 3.06 Gains. Whenever I have a Firmware Update I click Default in the Gains Tab.

If it was a battery related problem (which seems very likely IMHO), voltage is the key. The P2 has a hidden third low battery threshold in addition to the 1st and 2nd level warnings (has always had this - unrelated to the firmware update). This uses 10.65V as its threshold. Both this voltage threshold and the 2nd Level Low Battery Warning will trigger auto-landing. Altitude theoretically can be maintained if necessary by pushing up on the throttle stick, but this is dependent on the health of the battery. In your case, it seems as though the first two battery warnings were never hit because it just went from typical voltage reading to a low voltage reading really fast (not uncommon with a bad battery cell). At first you were able to fight the altitude loss somewhat by using full throttle, but due to the defective battery it eventually could no longer provide enough power to the motors and "dropped" as you described it. If any of us were in your shoes that day, our model would have experienced the same fate. I don't know if it will help, but if I were you I would petition DJI to see if they will step up for you given all the evidence points to this being a result of a bad battery.

 

[ianwood]

Auto land at 10.65v wouldn't explain why xrover couldn't override the descent. Hitting 10.65v at 8 minutes or 78% would indicate a really bad battery (i.e. bad cell).

RF interference due to water has been discussed before. It fits with the disproportionate number of water related incidents. But that wouldn't cause an auto land either unless the Phantom was set to auto land during failsafe. I know many successful flights have been done over water but it could be that several conditions need to be in place for the issue to manifest itself fully.

There is I suppose the possibility that a bad cell triggered auto-land and once it reached a certain height, it was too low over the water to overcome water based RF oddities.

xrover, sorry this happened. DVR for the OSD makes for good insurance in these cases. I know that doesn't help now.

Do we have any other verifiable cases of uncontrolled descents (not auto land due to low V) on 3.06?

 

[N017RW]

Auto land at 10.65v wouldn't explain why xrover couldn't override the descent. Hitting 10.65v at 8 minutes or 78% would indicate a really bad battery (i.e. bad cell).

RF interference due to water has been discussed before. It fits with the disproportionate number of water related incidents. But that wouldn't cause an auto land either unless the Phantom was set to auto land during failsafe. I know many successful flights have been done over water but it could be that several conditions need to be in place for the issue to manifest itself fully.

There is I suppose the possibility that a bad cell triggered auto-land and once it reached a certain height, it was too low over the water to overcome water based RF oddities.

xrover, sorry this happened. DVR for the OSD makes for good insurance in these cases. I know that doesn't help now.

Do we have any other verifiable cases of uncontrolled descents (not auto land due to low V) on 3.06?

Not sure I can even consider water an issue with regards to being causal.
I have flown seaplanes and raced r/c boats on water for years (Spektrum) and never had any issues.

 

[ianwood]

Not sure I can even consider water an issue with regards to being causal.
I have flown seaplanes and raced r/c boats on water for years (Spektrum) and never had any issues.

Different frequencies and frequency technologies will be a factor. Atmospheric conditions. Salinity. Wind effect on water surface. RF reflection/interference over water is not a made up theory. How it works is very complicated, way beyond my depth. But I know it takes several factors for it to happen. All the holes in the cheese need to line up. And it's only a suspect at this point.

However, the disproportionate number of loss of control events over water seems to bear evidence of something in that environment being unique compared to others.

 

[thongbong]

I'm sorry for you guys who have a J hook problem, it really blows that the only way to fix it is with new firmware. I've been flying Phantoms ever since DJI made them. Was one of the first, and I can tell you one thing, everytime they introduce a new firmware to fix a problem, they introduce a new one. Dropping straight out of the sky was never a problem with the Phantom v1. (that's 1, not 1.1.1), but fly away were (thats flying horizontal with the horizon without control until it hits something or runs out of batteries.) Now with the new Phantoms it seems to be dropping out of the sky. This happened to my buddies PV2+, over land. It was not VRS. And he's a **** good CP heli pilot which is way tougher than flying a Phantom in Manual mode. I've had firmware problems in the past and DJI KNOWS their firmware is buggy, that's why there's a back door to downgrade. There are too many stories of this happening to brush it off as environmental. It's like when flyaways were first reported, there were a bunch of pilots who denied it as user error until it happened to them.

 

[ToddSmi]

Auto land at 10.65v wouldn't explain why xrover couldn't override the descent. Hitting 10.65v at 8 minutes or 78% would indicate a really bad battery (i.e. bad cell).

Not being able to override descent is the most mysterious part of this in my view, but with a bad cell, plenty of strange behavior has been reported. Full throttle in particular calls for the heaviest load on the battery. What makes you think this could not have been entirely the result of a bad battery? Seems more plausible to me than a combination of low probability events. I would revise my hypothesis of course, if there are many other similar incidents that soon get reported, in which case the firmware update would become the primary suspect.

 

[hsestes]

Has anyone seen any strange behavior with their 3-D gimbal after upgrading to firmware 3.06? It was working fine before the upgrade and now after the upgrade it almost appears as if the motors are in a bind?

 

[Xrover]

8mins in the air and was still at 78%? that doesnt make much sense... Also what was your voltage when it started to descend? Did you use the old default gains or the new ones?

8Mins was what I was guessing from the time I put the battery in and took off. That was a rough guess.
78% Battery was a definite.. As it started to descend I immediately checked the numbers. Battery Level was 78%, GPS was 12 Sats. but battery voltage is difficult to read on a FPV screen with iosd mini and the Gopro recording. The Red Recording indicator get's overlaid right over the Battery Voltage reading.

I was using the Default 3.06 Gains. Whenever I have a Firmware Update I click Default in the Gains Tab.

If it was a battery related problem (which seems very likely IMHO), voltage is the key. The P2 has a hidden third low battery threshold in addition to the 1st and 2nd level warnings (has always had this - unrelated to the firmware update). This uses 10.65V as its threshold. Both this voltage threshold and the 2nd Level Low Battery Warning will trigger auto-landing. Altitude theoretically can be maintained if necessary by pushing up on the throttle stick, but this is dependent on the health of the battery. In your case, it seems as though the first two battery warnings were never hit because it just went from typical voltage reading to a low voltage reading really fast (not uncommon with a bad battery cell). At first you were able to fight the altitude loss somewhat by using full throttle, but due to the defective battery it eventually could no longer provide enough power to the motors and "dropped" as you described it. If any of us were in your shoes that day, our model would have experienced the same fate. I don't know if it will help, but if I were you I would petition DJI to see if they will step up for you given all the evidence points to this being a result of a bad battery.

I just sent a letter to DJI Director of Sales..

 

[wkf94025]

Wish they would fix this as it's meant to fly with the GoPro

It's tells me if I forgot to turn on recording rather than spend 10 minutes getting the perfect footage only to land and find out I took a stupid pill first thing in the morning

I agree that having the option of red recording circle on the screen, not in the way of other important iOSD fields, would have avoided my experiencing the stupid pill loss you described. It WAS great footage, trust me.

 

[Xrover]

I think an important thing to note after installing 3.06 was the lack of throttle response I noticed when taking off. I had to push the throttle stick up further then on prior FW versions (Beta Included) in order to take off. As I mentioned initially, I noticed it my first day testing out 3.06 and again on the shoot day. It used to pop off the ground with just a little bit of throttle up. Now it required about 20% more throttle. Not enough to cause me concern. At that time.

Can anybody else confirm this on 3.06?

I think when the P2 non Vis goes into Autoland or FS it will display it on the iosd mini. Nothing changed on the iosd. It still read GPS and my FPV signal was clean and the GoPro Red Recording mode indicator Light was still on. Everything appeared normal except Throttle Response diminished very quickly even after i was able to get it to climb a little. I was still on the stick when it rose up a bit and remaining on the stick it then began to make a slow descent and then just fully drop.

 

[ToddSmi]

I think when the P2 non Vis goes into Autoland or FS it will display it on the iosd mini. Nothing changed on the iosd. It still read GPS and my FPV signal was clean and the GoPro Red Recording mode indicator Light was still on. Everything appeared normal except Throttle Response diminished very quickly even after i was able to get it to climb a little. I was still on the stick when it rose up a bit and remaining on the stick it then began to make a slow descent and then just fully drop.

I don't believe I've triggered the 3rd level warning before, so maybe someone that has can chime in...but it may not send any message...perhaps why DJI calls it the "hidden" level. It would be nice if there was more detail, but the section at the bottom of this page is all we can ascertain from the manual:

http://wiki.dji.com/en/index.php/Phantom_2-Flight

 

[wkf94025]

My P2 has hit the "hidden voltage" many times on multiple batteries. No warning or change in OSD status, just red tail lights and auto-land at about 0.6m/sec.

Kelly

 

[Walter]

http://appleinsider.com/articles/14/08/ ... les-iphone

I bet a few million people read this.

 

[EMCSQUAR]

If it was a battery related problem (which seems very likely IMHO), voltage is the key. The P2 has a hidden third low battery threshold in addition to the 1st and 2nd level warnings (has always had this - unrelated to the firmware update). This uses 10.65V as its threshold. Both this voltage threshold and the 2nd Level Low Battery Warning will trigger auto-landing. Altitude theoretically can be maintained if necessary by pushing up on the throttle stick, but this is dependent on the health of the battery. In your case, it seems as though the first two battery warnings were never hit because it just went from typical voltage reading to a low voltage reading really fast (not uncommon with a bad battery cell). At first you were able to fight the altitude loss somewhat by using full throttle, but due to the defective battery it eventually could no longer provide enough power to the motors and "dropped" as you described it. If any of us were in your shoes that day, our model would have experienced the same fate. I don't know if it will help, but if I were you I would petition DJI to see if they will step up for you given all the evidence points to this being a result of a bad battery.

I'd be interested to where you obtained this "hidden 3rd low battery threshold" information. I've done alot of testing and have flown well below 10.65. Aa a matter of fact, my 1st level on my current P2 is 10.6, 2nd level is 10.2 and have flown it with warnings off to 9.9. Could you provide where you obtained this information?