Props abruptly stopped and IRIS fell from the sky from 40 ft

Well, it was bound to happen. But, what I can’t figure out is why. It wasn’t very long into the flight (4 mins?) when the props just stopped spinning. I am attaching the log here… I’m still trying to figure out how to interpret the logs. But basically, here is what happened.

I took off and started flying around in STAB mode then I realized that I hadn’t turned on the GoPro, So I landed. I then turned on the GoPro and made a couple of full-forward stick deflection passes. On the one before it fell out of the sky, I was turning it around for another pass and then it just turned off. I remember that just as the props stopped spinning, the RC transmitter started to beep - it had never done that before (then again, I’ve never had my IRIS fall out of the sky). I don’t know if that was the telemetry radio telling the transmitter that the props stopped or what. Anyway, I ran over to retrieve the pieces in the grass. The IRIS still had a green flashing LED and it was just beeping slowly. I picked up all the pieces and went back to my takeoff area. The RC radio was still beeping. The radio reads 10.8 VDC on the little screen. At first I thought that was the problem, but I thought the IRIS was supposed to initiate a landing if it loses the RC connection. In any case, I know all of this can be read in the logs, so I really appreciate the help. If it was a low battery condition, then that is something that baffles me because in all my previous flights, the IRIS properly initiated a landing.

The damage was isolated to one arm, for which I have a replacement, ALL the props, and I am full of props. The really big sucky part of all this is that the tarot gimbal platform/mount is badly broken. I couldn’t even find all the parts in the grass. The gimbal PCB is perfectly fine and even the GoPro is fine. I could use some help finding out where I can buy this part. I have the part breakout from the little manual that came with it, but no clue as to where to buy it.

Sorry to hear about your crash. It really sucks when you break things. Hope you’ll get the replacement parts you’re looking for.

Looking at your logs: You might be out of luck here. It seems like you facing the same brownout issues that others here are reporting. Your logs stop at approx 2 minutes 45 seconds with the IRIS showing to be about 25m (75ft) up in the air (See attached graphic).
Is it possible that the crash happened earlier than you wrote and at a different altitude? Where on the baseball field did you crash? That might help find out if the logs stopped way before the crash or with the crash. The logs end when IRIS was at the red X that I drew in the map below.

Up until the logs stop, I can’t see anything wrong:

  • BarAlt and GPSAlt match
  • No mechanical failure
  • Vibrations are OK
  • GPS coverage isn’t great initially but gets better after 1-2 min
  • Battery voltage at end of logs was still at 10.75V
    Did you measure the LiPo with a voltmeter after the crash?

kangajump I sure appreciate you being in here and doing log analysis for people. Thanks!!

Thanks for getting back to me. The battery read 11.33VDC after I disconnected it and inspected it for damage.

I am attaching the map back to with the X as to where the crash was. I have reviewed the flight video too, and the crash was from much higher than I originally estimated. The IRIS was pretty much even with the tops of those light posts when it suddenly stopped. I have drawn an arrow and an x from your X. It was moving along at a pretty good clip. Like I said, I had the right stick fully deflected forward, and it was going with the wind.

Will the flight video help?

I have been reading about the Brown Out’s but I was having so much luck with so many flight without fault and then this happens. Is this my fault? How can I keep it from happening again? Especially now that it looks like I might be in for some hefty repair costs, I don’t know. Now I’m going to be flying this thing around like it’s my grandma’s old Model T - and that’s not why I got this.

Here is the crash footage

youtu.be/lL8Q3vqx0WI

Guys are you reading other forums about the brown outs or just here ??

One of the reasons I went with Iris instead of a build is that I wanted something balanced and engineered so things like this could be avoided.

it would be good if 3DR would comment on this…

Our IRIS arrived monday, good thing I have been ill the past 2 days and have been able to examine all the contents of the shipping package thoroughly with discrepancies reported to 3DR.

Beyond the very common multicolor LED stretched cable complaint(needs 2 cm more) and some issues about the quality of the hull mounted microUSB connector everything else seems fine.(besides 2 missing 10mm cap screws and 1 offsize prop nut) (QC inspectors seem to be needed)(or a robotic cam to detect these external shortcomings).

The 433 telemetry came up immediately once all was started in the proper sequence.(yes i do hold a technician license) and we are moving to 915ISM radios after the initial checkout.

And then I saw the brown out video above (and I was planning on adding an 8 amp sbec(NOT!) to drive a 155 led RGB NeoPixel strip). I have seen concerns that the brownout may be connected to the 3DR PDB. Or too much current draw on the pixhawk.

As everyone I have seen reported having this syndrome seems to be running the stock batteries, I have grounded the IRIS(before a test flight even) until new higher C rating branded LiPo batteries and a different brand/batch can be obtained and the IRIS discrepancies resolved.

I suspect the current© and capacity rating of the LiPos supplied is optimistic or the defect rate abnormally high.(I have seen this kind of issue before on optimistically rated batteries(these seem to lack even a brand)).

I am probably going to add another smaller battery JUST to power the pixhawk(if this is possible) alone so that even if a brownout occurs it will be logged in its entirety.

Its a new consumer/prosumer product. there are bound to be hiccups during launch the real test is how 3DR resolves this as a company.

Customers showing up in forums with broken pieces instead of the IRIS are NOT good for business and the fact of onboard video and logging should make warranty claims arising out of such easy to resolve.

Our present 30C “zippy” batteries are just barely fitting in that compartment and probably not the best alternative.(Deans plug instead of XT60 connectors also.

As our IRIS was purchased for research in to algorithms for detecting HuangLongBin/Citrus greening plague via minature multispectral camera , we are NOT in a hurry to reduce it to a collection of broken pieces by moving too quickly.

Nevermind! :slight_smile:

Let’s stay honest here:

Yes, it really sucks for everyone that has a crash and especially if afterwards something is damaged. Even if it’s only a prop. But something like the gimbal certainly hurts. I don’t wish that for anyone.

Yes, some of us are seeing some anomalies around truncated or corrupted log files. At least the truncated log files can be caused by brownouts. But as Craig stated here (viewtopic.php?f=48&t=6590#p12625) the log file issues could be caused by something very different. Again, brownouts are a cause for truncated log files. But that doesn’t mean that it’s the only cause. We might not be facing brownouts at all. Wit that it’s very much possible that we can find usable log files on the IRIS itself. I haven’t tried it myself, so I can’t tell.

I also have to admit that I’ve been so far very careful with my IRIS, not taking here higher than 15ft / 5 m in the air. And I also have made my fair share of rookie mistakes like attempting to configure IRIS via Mission Planner, even though it was already setup ready to fly. And I’ve also set the wrong frame type doing so, seeing IRIS wobble like cray during the first flights that didn’t even last 10-15 secs.

But even with the anomalies that we are seeing I’ll still take her out next weekend. Only that way I’ll be able to test out things and learn more.
And while my last 6 flights all show truncated log files, I didn’t experience any issue during the flights. In fact IRIS performed wonderfully and I was very happy with what I saw.

So let’s not overreact. Just be careful and mindful while you learn to handle your new “toy” and push the limits slowly. That’s at least what I’ll be doing.

Chris

@KFayal:

Can you try what Craig recommended here (viewtopic.php?f=48&t=6590#p12625) and attempt to download the log files directly from the SD card?
That would help to determine whether we actually have complete logs to look at. Right now my fear is that we are missing the interesting pieces.

Also could you stop the exact time from takeoff to crash based on your video?
This would also help to determine whether the log file you posted so far is complete.
But even if you find another log file with more data on the SD card, it will help to look at the right spot.

I don’t suppose that GoPro was a WiFi one per chance? With the WiFi on?

@kangajump:

I attached the log file from the flight right off of the SD card. The funny thing is, the log file from Mission Planner has 55697 rows, but the .bin log off of the SD card has only 52268 rows. And yet it appears as though the .bin log off of the IRIS Pixhawk has more data - it even shows the fact that I landed, and took off again, as well as longer GPS tracks.

@Gary
The GoPro Hero3 Black for which the Tarot GoPro Gibmal is fashioned, has a WiFi feature. But I don’t use because of the chance of interference of the RC radio on the same frequency.

Thanks for attaching the logs that you directly pulled off IRIS.

First of all it confirms what Craig wrote earlier: The reason we are getting truncated log files, seems to be caused by Mission Planner and not due to a brownout. So that’s kind of a good thing.

The bad thing with the complete logs and I only had a very brief look at you logs and will look closer at them again: You can see that the voltage dropped to 10.5V and below towards the end of the flight (See attached graphic). That should have triggered the Battery Failsafe. For some reason it didn’t.

To me that’s the smoking gun: You just run out of juice. But why did IRIS plunge out of the air instead of doing a RTL or landing operation? Let’s find out.

I sent these logs in to 3DRobotics support as well. Looking at the flight duration, with a full battery to start, I have never been getting less than 7-8 minutes with the Tarot gimbal and GoPro attached - even when flying aggressively. Hmm…

Looks like this flight duration was somewhere around the 5 minute mark.

Here is a log of a prior flight that went fine. Notice the flying time and even somewhat aggressive nature of the flight. I do know that there is one DISTINCT difference than all the other perfectly fine flights that I had compared to this one where the IRIS crashed.

Upon arriving my IRIS had a broken antenna for the Telemetry radio, so I immediately removed it so that I could ship it back to 3DR. While I was waiting for that radio, I flew at least 19 flights WITHOUT the telemetry radio attached with no problems at all.

The flight I’m reporting here in this thread was the very FIRST flight that I had with the new telemetry radio attached. I don’t know if that is just a coincidence or not, but I do find it suspicious.

Looking at the doomed flight logs, at line 51239, the CURR Volt log first dropped below 10.5vdc. I read in the APM documentation that the Failsafe is supposed to trigger when the vehicle stays below the trigger voltage for more than 10 seconds consistently, which it never did - maybe why the failsafe never triggered. The very last CURR Volt log was 10.58vdc. I guess what I’m saying is that even on the past logs where the flight was successful, and the battery level failsafe triggered properly, the IRIS stayed in the air way past the 10.5vdc trigger. In addition, after the 11.33vdc reading I took from the battery (after removed from the load) would not indicate a battery issue. In fact the log I just submitted in my previous reply shows that the flight ended at 8.41vdc, and the props were still spinning, and more importantly, the PX4 was still logging.

It is because of this that I still believe that a “Brown Out” condition occurred.

@KFayal

Yes, you are right. After having had a closer look at the logs I saw that you only shortly crossed the 10.5V line, but never longer than 10sec. And therefore the Battery Failsafe never kicked in.

[attachment=3]Flight17_Battery.png[/attachment]
Before we go into the logs: Your log files from flight #17 are concerning (See graphic above). Your voltage drops below 3.0V per cell (9V total for the stock 3 cell pack). That’s beyond the lowest safe voltage per cell for a LiPo. See here fore more info on this: diydrones.com/forum/topics/lowes … ll-on-lipo
It’s possible that your LiPo became bad after flight #17. Do you know if you used the same LiPo for both flights?

The other thing I notice in flight #20’s logs but your flight #17 confirmed that: As Battery Failsafe action you have RTL set. The IRIS default params show that this should be “Landing” instead (See here: copter.ardupilot.com/wiki/arduco … ATT_ENABLE). Looking at flight #17, that setting would have probably saved your LiPo.

Back to the log files of flight #20: The last seconds of the logs don’t match what I could see on the video. In the log files I can’t see you turning IRIS to the left, which is what you wrote on the video:

[attachment=2]Flight20_Pitch_Roll.PNG[/attachment]
Looking at the Desired Pitch, Pitch, Desired Roll and Roll of the last seconds of your flight (See graphic above), I can only see that you put IRIS all the way foward via Desired Pitch and it tried to do that as good as possible (Marker 1). But there is nothing to be see on the Desired Roll or Roll. It stays flat, which means you didn’t initiate a left or right curve while the logs were still running.
[attachment=1]Flight20_Ch_IN.PNG[/attachment]
Similar picture, looking at the Channel IN (See above graphic): Channel 2 (Pitch) shows movement, Channel 1 (Roll) does not.

[attachment=0]Flight20_Curr.PNG[/attachment]
Towards the end of that flight, the currency was a bit higher with about 25C (See above graphic). Together with the potential issue that you might have let the LiPo go below 9V could have caused a brownout here.
An indication for that would be that your logs also stop again midair (See here: copter.ardupilot.com/wiki/common … wnOuts_etc).

But I’ve also seen other reports about people complaining about potential poor performance of the stock 3DR IRIS battery. See here: viewtopic.php?f=48&t=6400.

Hope that helps.

@kangajump
I have 4 of the 3dr stock batteries that I have serialized and on that flight 17, I don’t remember which battery was used. But I do know which one was used in flight 20. I wish there was a way to check for the battery being bad other than the voltage check because since then I have recharged the battery and it reads 12.55 volts.

In fact the other three have been unused after a recharge since March 8 and they all read

Battery 1: 12.53
Battery 2: 12.50
Battery 3: 12.48
And Battery 4 used in the last flight and most recently charged is 12.55

On my turns unlike RC planes, I turn mainly with Yaw first. After watching this video over and over I am now wondering if the turn shown in the video isn’t just a coincidental turn to the left because of the sudden stop to the motors because when I watch the video in slow motion the motors stop and THEN it yaws to the left suddenly.