Phantom 4 Advanced flyaway

[Meta4]

Perhaps I don’t need to, as the system reported it as a fault...or did you forget already...you seem to like to argue with no point. It’s a reported failure by the system, and if you were honest you can compare the values throughout the flight and see it went over spec several times...or do you prefer mindless argument?

The system did not report it as a fault.
The actual data only gives cell voltages.
Msinger who set up the Phantomhelp log viewer highlighted the colours so they stand out but no error was logged regarding voltages.

 

[sar104]

With all your log snooping what do you see as a typical cell deviation for a seemingly good battery? I get suspicious of a pack that shows frequent 70mv+ deviations hiwever I have one pack that can be 120mv out on occasion that hasn’t failed me yet.

The deviations typically depend on current draw, since the voltage drop is primarily due to the internal resistance of the cells. Mostly they are less than 100 mV. 70 mV is common but higher values don't necessarily mean anything significant, especially if they are transient excursions from otherwise normal values. You can see from the graph above that the maxima in the deviation plot coincide with voltage drops on all the cells, but reflect cell 4 being consistently lower than the others. It looks like a slight cell balance issue, that's all. It is most unlikely that it had anything to do with the event in my opinion.

 

[sar104]

So are you saying Phantomhelp based it’s reporting on a whim? You’re not even logical.

No - I'm saying that their software has a set threshold (100 mV) above which it reports cell deviation in red, presumably to flag it as possibly indicative of a problem. Blindly assuming that it means the aircraft is about to crash is what is illogical.

 

[Meta4]

So they sSTAND OUT BASED ON WHAT? In life you will find people don’t do something for NO REASON, THEY HAVE MUCH BETTER THINGS TO DO THAN RANDOMLY DECIDE ILL MAKE THIS NUMBR RED..FOR NO REASON...COME ON..I don’t know you, but honestly, I can tell you really enjoy arguing ..

I have no idea except to make it easier to spot ... how about we ask @msinger himself.
Mike ... why are cell voltage deviations >0.1V shown in red?
Are you aware of this being a serious issue?

 

[Slickvolt]

The deviations typically depend on current draw, since the voltage drop is primarily due to the internal resistance of the cells. Mostly they are less than 100 mV. 70 mV is common but higher values don't necessarily mean anything significant, especially if they are transient excursions from otherwise normal values. You can see from the graph above that the maxima in the deviation plot coincide with voltage drops on all the cells, but reflect cell 4 being consistently lower than the others. It looks like a slight cell balance issue, that's all. It is most unlikely that it had anything to do with the event in my opinion.

But that’s precisely where you’re jumping the gun, on small scale microelectronics millivolts fluctuations are HUGE, because as yourself pointed out it’s reflective of an uptic in current demand...and it’s time sensitive demand...some thing is screaming..I need a lot of power right now. Now as an engineer you never design a power structure that does not fully support max power draw specs..period. It’s like putting a ten amp fuse on a 15 amp circuit..bad things result. This isn’t even an argument for me, this is wasting my time. It’s that obvious..you have wild swing in cell deviation...over 130% increase at random times...this is one of two things...either a poorly designed power support structure ( which I doubt), or something is not right on the drone requiring more power than it should...I should add one more thing which we haven’t really spoken about..but these intelligent batteries are also software driven animals...there could be a problem there to, I must concede that...we really haven’t gone down that path.

 

[Meta4]

I waited almost 20 years to buy a drone, thinking in that time all these issues would be solved...nope. Drones really are a danger to aircraft, we still lose them, because the technology is still flawed...IMUs have magnetometers and e-compasses. The military ruled out any IMUs having these for one very good reason..they lose calibration from magnetic waves...that’s why warnings about flying close to large metal structures, especially iron or cast iron...huge magnetic fields. Instead the gold standard IMU uses real gyros...but costs near $50,000 -$70000 and fits in a breadbox...hence why we have these cheap IMUs. We are going to lose drones guys..hate to be the one who tells you. One day a lost drone will take down an aircraft you mark my words...it will happen.

We see lost drones every week and almost all are due to pilot error rather than hardware issues.
But lost drones aren't really a serious risk to other aircraft (unless they get lost at high altitudes where they could have been just as much of a risk whether lost or piloted).
Lost aircraft simply hover and RTH.
They don't go wandering off on their own and invading airspace.
The DJI Phantoms are very reliable, unfortunately the users aren't always.

Now what's this about flying close to large steel buildings?
There's nothing to indicate the pilot flew close to any steel structures, but if they did that doesn't do any damage to the compass.

 

[With The Birds]

But that’s precisely where you’re jumping the gun, on small scale microelectronics millivolts fluctuations are HUGE, because as yourself pointed out it’s reflective of an uptic in current demand...and it’s time sensitive demand...some thing is screaming..I need a lot of power right now. Now as an engineer you never design a power structure that does not fully support max power draw specs..period. It’s like putting a ten amp fuse on a 15 amp circuit..bad things result. This isn’t even an argument for me, this is wasting my time. It’s that obvious..you have wild swing in cell deviation...over 130% increase at random times...this is one of two things...either a poorly designed power support structure ( which I doubt), or something is not right on the drone requiring more power than it should...

If power demands were actually or calculated as likely to exceed apec we would expect to see a propulsion output limited warning. SAR posted a voltage graph earlier (can’t see it now it’s a page or so back) but from memory lowest depicted cell voltage was 3.5V and any dips were short duration. The charted voltage is seemingly more important is arriving at a determination of a problem than the deviations reported.

 

[sar104]

But that’s precisely where you’re jumping the gun, on small scale microelectronics millivolts fluctuations are HUGE, because as yourself pointed out it’s reflective of an uptic in current demand...and it’s time sensitive demand...some thing is screaming..I need a lot of power right now. Now as an engineer you never design a power structure that does not fully support max power draw specs..period. It’s like putting a ten amp fuse on a 15 amp circuit..bad things result. This isn’t even an argument for me, this is wasting my time. It’s that obvious..you have wild swing in cell deviation...over 130% increase at random times...this is one of two things...either a poorly designed power support structure ( which I doubt), or something is not right on the drone requiring more power than it should...

The current draw fluctuations are not a mystery at all if you look at the pitch and roll data. This aircraft was working very hard flying into a strong headwind. And we are not dealing with "small scale mircoelectronics" - these systems are pulling up to 20 A or so. Look at the data - this is just a slight battery cell imbalance. There is no reason to believe that it caused the event. You are just making random assertions here - not even reasoned hypotheses.

 

[Slickvolt]

No - I'm saying that their software has a set threshold (100 mV) above which it reports cell deviation in red, presumably to flag it as possibly indicative of a problem. Blindly assuming that it means the aircraft is about to crash is what is illogical.

I’ve never stated a sole problem, I stated four...the point being there is a lot in the logs pointing to real hw or sw problems..I’m factually correct on this. It was a few overzealous yahoos who were crucifying the operator for breaking regulations etc, ad nauseous...and I pointed out there was an abundance of evidence to indicate system failures..and we have in fact been speaking of those..proving my point implicitly...now go back and see my last line of my original post states, can’t know for sure the cause. That is the mystery. We have a lot of pieces in the logs in spite of how vacuous of telemetry data it was...and there were several issues with the drone as reported in the logs..that is undeniable.

The very fact the drone flew off and is missing supports that..wow

 

[Slickvolt]

The current draw fluctuations are not a mystery at all if you look at the pitch and roll data. This aircraft was working very hard flying into a strong headwind. And we are not dealing with "small scale mircoelectronics" - these systems are pulling up to 20 A or so. Look at the data - this is just a slight battery cell imbalance. There is no reason to believe that it caused the event. You are just making random assertions here - not even reasoned hypotheses.

Omg it’s totally supported by the logs..wow guys, my last recommendation..don’t ever work in electronics, you don’t seem to get it. Good night, I schooled every one of you and you won by wasting my time.
Y’all lost serious credibility with me tonite, serious credibility...I literally did this for a living for like 15 years...wow, I’d fire everyone of you if you had worked for me. Ignoring data on the logs to blame the pilot...wow

 

[sar104]

I’ve never stated a sole problem, I stated four...the point being there is a lot in the logs pointing to real hw or sw problems..I’m factually correct on this. It was a few overzealous yahoos who were crucifying the operator for breaking regulations etc, ad nauseous...and I pointed out there was an abundance of evidence to indicate system failures..and we have in fact been speaking of those..proving my point implicitly...now go back and see my last line of my original post states, can’t know for sure the cause. That is the mystery. We have a lot of pieces in the logs in spite of how vacuous of telemetry data it was...and there were several issues with the drone as reported in the logs..that is undeniable.

The very fact the drone flew off and is missing supports that..wow

No one is arguing that there was not some kind of failure. What we have been arguing with you about is your fixation on specific issues for which there is little or no evidence. First it was the compass, which you apparently abandoned after misunderstanding how the FC uses the magnetometer data, then it was some abundance of other, unspecifed clear evidence of hardware failure, and then it was the battery cell deviations, all because they were in red on the PhantomHelp page.

If you want to post and argue a hypothesis that's fine, but there's no point getting offended when it gets critiqued by others, especially if you are unable to support it with anything tangible.

The FC only reported one kind of error - the compass error. And that does not explain the sudden end of the log record mid-flight. Nor do small, transient cell deviations under high current load. If those had been getting larger through the flight then they might have been significant, but they didn't do that.

 

[sar104]

Omg it’s totally supported by the logs..wow guys, my last recommendation..don’t ever work in electronics, you don’t seem to get it. Good night, I schooled every one of you and you won by wasting my time.
Y’all lost serious credibility with me tonite, serious credibility...I literally did this for a living for like 15 years...wow, I’d fire everyone of you if you had worked for me. Ignoring data on the logs to blame the pilot...wow

Well that was kind of inevitable. Unable to make a cogent argument, reply to any technical questions, or post any supporting data at all, he has a tantrum and storms off. I have to say, after that, that if I did still have any credibility with him then I would be seriously worried. But it was fun being schooled.

 

[Meta4]

The very fact the drone flew off and is missing supports that..wow

Here's an example of reading between the lines.
There's nothing at all to show that the Phantom flew off.
We have data showing the drone was flown to where the recorded data stopped but absolutely nothing to say what happened after that.

Omg it’s totally supported by the logs..
Y’all lost serious credibility with me tonite, serious credibility...I literally did this for a living for like 15 years...wow, I’d fire everyone of you if you had worked for me. Ignoring data on the logs to blame the pilot...wow

And I'd hope that someone claiming so much experience would look at what the data says rather than making up things that might have happened and suggesting they are clearly shown in the data.
That's how we work with data ... you need data to actually show something happened before deciding it did happen.

 

[Neon Euc]

I'm thinking a tad outside the box here that may cause a few issues

*battery was loose playing havoc with the data causing it to lose power and drop

*bird strike

*firmware may have been the one where RTH gets reset miles away (happened a few times on here but very rarely)

*tall structure that was up and not shown on Google maps as it's not updated daily

This is just a few *out of the box" ideas guys so don't shoot me for this. Unfortunately I can't read detailed data like the one posted (airdata is simple for me to Understand despite it not showing as much data as the flight log) but hope we can figure out what happened for the OP. sadly this post turned out to be pretty negative with arguments and shooting the OP down [emoji20][emoji20]

 

[With The Birds]

I'm thinking a tad outside the box here that may cause a few issues

*battery was loose playing havoc with the data causing it to lose power and drop

*bird strike

*firmware may have been the one where RTH gets reset miles away (happened a few times on here but very rarely)

*tall structure that was up and not shown on Google maps as it's not updated daily

This is just a few *out of the box" ideas guys so don't shoot me for this. Unfortunately I can't read detailed data like the one posted (airdata is simple for me to Understand despite it not showing as much data as the flight log) but hope we can figure out what happened for the OP. sadly this post turned out to be pretty negative with arguments and shooting the OP down [emoji20][emoji20]

Maybe the OP friends AC was shot down. It’s another possibility.

 

[Slickvolt]

No one is arguing that there was not some kind of failure. What we have been arguing with you about is your fixation on specific issues for which there is little or no evidence. First it was the compass, which you apparently abandoned after misunderstanding how the FC uses the magnetometer data, then it was some abundance of other, unspecifed clear evidence of hardware failure, and then it was the battery cell deviations, all because they were in red on the PhantomHelp page.

If you want to post and argue a hypothesis that's fine, but there's no point getting offended when it gets critiqued by others, especially if you are unable to support it with anything tangible.

The FC only reported one kind of error - the compass error. And that does not explain the sudden end of the log record mid-flight. Nor do small, transient cell deviations under high current load. If those had been getting larger through the flight then they might have been significant, but they didn't do that.

Wrong Again, 1. there are four different anomalies in the log...four, not one. I'm not going over them all again. It's a fact. 2. The drone is gone, it's not suppose to do that PERIOD. Another Fact 3. It DID NOT return to home...a failure in itself...Another Fact...this is really getting rather pathetic. 4. The flight path shown while stable in altitude shows erratic course changes...that may be only appearance, as not enough user information was given, other than a "long time experienced pilot", which isn't a fact, every bit of info on the pilot we have is called heresay, not Facts...yet these yahoo's blame the operator?

The bottom line given the FACTS, there is ample evidence to support a system failure.. on the human side we have NO FACTS in the pilot, yet the very first inclination of the knuckleheads in here was to torque off on and blame the pilot...that to me is just plain despicable behavior.

 

[sar104]

Wrong Again, 1. there are four different anomalies in the log...four, not one. I'm not going over them all again. It's a fact. 2. The drone is gone, it's not suppose to do that PERIOD. Another Fact 3. It DID NOT return to home...a failure in itself...Another Fact...this is really getting rather pathetic. 4. The flight path shown while stable in altitude shows erratic course changes...that may be only appearance, as not enough user information was given, other than a "long time experienced pilot", which isn't a fact, every bit of info on the pilot we have is called heresay, not Facts...yet these yahoo's blame the operator?

The bottom line given the FACTS, there is ample evidence to support a system failure.. on the human side we have NO FACTS in the pilot, yet the very first inclination of the knuckleheads in here was to torque off on and blame the pilot...that to me is just plain despicable behavior.

Back again? Of course you are not going to cover the anomalies again. You didn't cover them in the first place. The detailed discussion that you tried to participate in last night was specifically about what the failure might have been - it was not blaming the operator. It's hard to have a meaningful discussion with someone who doesn't read or understand what is posted.

I'm going to have to point out that you use a lot of technical terms either very loosely or completely incorrectly - that makes me very suspicious of any of the credentials that you claim or that you have any clue what you are talking about. And your resort to ad hominem attacks when questioned, as you did last night and again on at least one other thread this morning, is not going to work well around here.

 

[sar104]

I'm thinking a tad outside the box here that may cause a few issues

*battery was loose playing havoc with the data causing it to lose power and drop

*bird strike

*firmware may have been the one where RTH gets reset miles away (happened a few times on here but very rarely)

*tall structure that was up and not shown on Google maps as it's not updated daily

This is just a few *out of the box" ideas guys so don't shoot me for this. Unfortunately I can't read detailed data like the one posted (airdata is simple for me to Understand despite it not showing as much data as the flight log) but hope we can figure out what happened for the OP. sadly this post turned out to be pretty negative with arguments and shooting the OP down [emoji20][emoji20]

The battery may have disconnected without warning - it's a single point failure consistent with the log, but it would have put the Phantom in the parking lot below, which the OP said was searched.

Catastrophic failure due to a bird strike is also possible. Don't hear of that much, if ever though. And would likely have had the same outcome.

RTH reset on disconnect is always associated with an attempted RTH reset of some kind I think. No evidence of that in this case.

I doubt that they built a 700 ft structure in the car plant parking lot in the approach pattern to a moderately large airport.

The pilot (by proxy) was initially criticized for by a few of us for the flight, but if you read the rest of the thread you will find a lengthy discussion of various possibilities, although mostly anchored with the observation that the data, themselves, only shed light on the problem by what they don't say.

 

[Slickvolt]

Back again? Of course you are not going to cover the anomalies again. You didn't cover them in the first place. The detailed discussion that you tried to participate in last night was specifically about what the failure might have been - it was not blaming the operator. It's hard to have a meaningful discussion with someone who doesn't read or understand what is posted.

I'm going to have to point out that you use a lot of technical terms either very loosely or completely incorrectly - that makes me very suspicious of any of the credentials that you claim or that you have any clue what you are talking about. And your resort to ad hominem attacks when questioned, as you did last night and again on at least one other thread this morning, is not going to work well around here.

Hey I have news for you, I'm rapidly learning , as seen real other people pointed out yesterday, that this forum is NOT the place to come for help whatso e er...I do t think that was the goal of this foeum...but that is what it became...the nerve of the ecperts to attack a person who literally was looking for help...the nerve, nor a shred of evidence the user was negligent or careless, not one FACT. I pointed out FACTS in the data. You guys really need to check your egos at the door, people are already telling you...youre of no help.

In fact I would not be surprised to find out, there maybe some DJI employees on this board, as thats what the above behavior is indicative of...blame the user, ignore that the drone didn't perform as advertised...These are FACTS

 

[Slickvolt]

We see lost drones every week and almost all are due to pilot error rather than hardware issues.
But lost drones aren't really a serious risk to other aircraft (unless they get lost at high altitudes where they could have been just as much of a risk whether lost or piloted).
Lost aircraft simply hover and RTH.
They don't go wandering off on their own and invading airspace.
The DJI Phantoms are very reliable, unfortunately the users aren't always.

Now what's this about flying close to large steel buildings?
There's nothing to indicate the pilot flew close to any steel structures, but if they did that doesn't do any damage to the compass.

Never stated pilot flew close to bridges or large steel buildings...reread post...post said those were objects during testing that were shown to impact magnetometer performance and compass calibration.among other things.