3DR Iris Screwed Me Again!

Well I don't know whether to laugh or cry. I had a horrific crash a couple of months ago, so I've been laying low. Then last week I started to piece everything together and get back into it VERY SLOWLY and carefully.

Some of you (on facebook) helped me look at my logs from 3 simple flights just to foresee any problems, everything looked ok: http://diydrones.com/group/iris/forum/topics/please-check-my-logs

So yesterday I assemble a brand new Tarot 2D gimbal, and also pieced together a couple of broken GoPros and made 1 that would record. Just scraping up the pieces to get back into it slowly, ya know?

I took it to the park tonight, a different park and I made sure to STAY OVER THE GRASS AT ALL TIMES, and no higher than 10 meters for the most part. I flew several auto missions I just drew up tonight in DroidPlanner2. Just quick flights, relatively low to the ground, and I brought 3 batteries.

I was testing different flight patterns and waypoint functions in DroidPlanner, nothing too crazy. A pause here for 10 seconds, a 10 meter circle around this area, point the camera here, no big deal, right???

I'm in my 4th flight and using my 3rd and final battery for the night. The audio comes on and says "Battery at 90%" so I'm like ok, cool. I'm just letting it run it's auto mission that probably would have lasted a grand total of 3 minutes, and the son of a bitch drops out of the F#^%ing sky AGAIN like a rock, NO WARNING. I break one of the blue arms, and my BRAND NEW F$*%ing gimbal snaps off of the Iris.

Is this a joke?

So anyways, here I go again, I'll put this one in to 3DR support with almost ZERO confidence, let's see if they'll offer 30 bucks off of something again and find a way to suggest it was all my fault.

Here are the logs, and better yet I've got the video. What else can I provide you fine folks? I'll just share the Dropbox folder with all 4 logs from tonight, and of course the crash is the latest log. I would absolutely love some analysis on what went wrong.

https://www.dropbox.com/sh/rn0h0q8msmbw87w/AAD3R6iA_sBKkO0SDxDphz00a?dl=0

And the video. I can share the other videos from the earlier flights tonight, but they are uneventful, just going to a dozen waypoints, looking at points of interest, pausing, and landing.

The crash happens at 2:37, and I think it's just absolutely adorable how the motors continue to spin EVERY TIME it crashes even if it's upside down, as if there is no setting on the compass that can turn those damned motors off.

I'm not sure what ESC's you are using, or what motors or voltage, but It seems to me like the ESC's get pretty hot, especially if your running ESC's that are right at the requirements of the motors. Ive noticed that If I just pay a couple extra bucks per ESC and get the next one up they will stay allot cooler. Having 10 amps of overhead is my rule of thumb.

I'm just throwing a wild guess out there on this one, but it's possible that running 3 flights in a row could have gotten the ESC's nice and toasty, and then it shut down for protection. I haven't read anything from 3DR about running consecutive drone flights or a cool down period (of course I haven't really looked either) But I'm guessing that only a small percentage of users runs theirs birds through 3 full batteries.Either way its someplace to start. Plus I think 3DR only runs 20 AMP ESC's on the IRIS. (again I don't know that for sure)

You could easily replicate the conditions by putting something that weighs about the same as the GoPro and gimbal on it and then either tether it over something soft or just reverse the props (either way I would keep the props on since their drag contributes to the load on the electronics) and run it through 3 batteries with a camera pointed on it and see what happens. Also I dont know if you have a laser thermometer, (IF not you can one at harbor freight for $20) but getting the temperature of all 4 ESC's during this test may flush out one that is overheating early. I have built more then a few drones and I always order at least 1 Extra ESC per drone. It seems like about 1 out every 12-15 or so will fry prematurely.

Replies

- Justin M > Scott July 9, 2015 at 11:06am
  
  Sorry, just getting back to this. Do you not have access to these files? They should be public.
  
  How about Google Drive, would that work?
  
  And sorry, I get confused on which files to share. There are 6 different file extensions for each flight, it seems. bin, gpx, tlog, txt.....
Christoph July 8, 2015 at 6:00am

Does anybody knows how to interpret the fact that the RCout to the ESC went down?

Will the log show such behaviour in case the ESC overheats? My best guess would be if the RCout signal goes down, it's coming from pixhawk, thus from the software.
Will Snodgrass July 7, 2015 at 1:31pm

I'm not sure what ESC's you are using, or what motors or voltage, but It seems to me like the ESC's get pretty hot, especially if your running ESC's that are right at the requirements of the motors. Ive noticed that If I just pay a couple extra bucks per ESC and get the next one up they will stay allot cooler. Having 10 amps of overhead is my rule of thumb.

I'm just throwing a wild guess out there on this one, but it's possible that running 3 flights in a row could have gotten the ESC's nice and toasty, and then it shut down for protection. I haven't read anything from 3DR about running consecutive drone flights or a cool down period (of course I haven't really looked either) But I'm guessing that only a small percentage of users runs theirs birds through 3 full batteries.Either way its someplace to start. Plus I think 3DR only runs 20 AMP ESC's on the IRIS. (again I don't know that for sure)

You could easily replicate the conditions by putting something that weighs about the same as the GoPro and gimbal on it and then either tether it over something soft or just reverse the props (either way I would keep the props on since their drag contributes to the load on the electronics) and run it through 3 batteries with a camera pointed on it and see what happens. Also I dont know if you have a laser thermometer, (IF not you can one at harbor freight for $20) but getting the temperature of all 4 ESC's during this test may flush out one that is overheating early. I have built more then a few drones and I always order at least 1 Extra ESC per drone. It seems like about 1 out every 12-15 or so will fry prematurely.
- Justin M > Will Snodgrass July 9, 2015 at 11:04am
  
  I'm just using the factory ESC that came with the IRIS+, The only modifications I've done to the craft is adding the gimbal and Liveview board and antenna, which wasn't present during this crash.
  
  Other than this, I have absolutely zero idea of what model everything is or how much voltage, amps, or temp it takes.
- DavidFolts > Will Snodgrass July 7, 2015 at 6:19pm
  
  Will - The Iris+ uses a V-VLSI ESC system. Remove the bottom cover and its all crammed in there with little venting available for bleeding heat out of the fuse. 1 BIG hunk of silicon... kinda slick.
  
  I've developed a theory that the Iris overheats in direct sunlight and its Pixhawk, mounted above the ESC system is sensitive to heat (possibly causing gyro issues.... its what killed my first unit)
  
  I'm working on placing a micro cooling fan in the forward end of the fuse deck to blow air thru a slit in the "bow" and then out the "stern" of the fuse.... might help, might not. Gotta try at least.
- hotelzululima > DavidFolts July 7, 2015 at 9:40pm
  
  forget the fan simply make your IRIS+ preggers!
  
  ie pregnant IRIS with a up to 10mm air gap for bigger lipos and better air circ.
  
  btw there IS a rather thick LIPO in between the ESC and the pixhawk..
  
  http://ardupilot.com/forum/viewtopic.php?f=48&t=7272
  
  hzl
- DavidFolts > hotelzululima July 8, 2015 at 9:08am
  
  And that lipo is also producing an amount of heat.
  
  To Christoph's point, I too fly in a hot, humid and sun intense environment (Houston area). And the sun at 29 degrees latitude heats-up a black object to 200 degrees...making the GPS module under the top cover hotter as well.
  
  My idea was to cut a 2-3 mm slit in both top and bottom covers on the forward end to let air in... and use a micro-fan to force air over the ESC & pixhawk especially whilst in hover.
- Will Snodgrass > DavidFolts July 7, 2015 at 8:52pm
  
  Oh wow, I didn't know they had gotten quite so proprietary. But yeah I think a push pull system would go along ways towards cooling everything down.
Christoph July 3, 2015 at 8:48am

Had a look at your logs again. Still I am not an expert yet ;)

My interpretation is something went wrong with the ESC.

When your bird fell out of the sky the current drawn from your battery is going against zero, which confirms that your motors all shut down. At the same time the arducopter is raising throttle to full because it's realizing that your IRIS is losing height but it should not, because nobody gave the command to do so.

What I don't understand yet, why the RCoutput to the ESC goes down. So to me it looks like the signal to the ESC went down and shut them off, but why?
Mhct July 3, 2015 at 4:01am

I checked your logs but couldn't find anything special (Though I am not an expert yet in reading such logs).

I had a, apparently, very similar issue with my Iris+, by coincidence, one day after you hard yours...

If you want to take a look at the footage I've made, and the logs... perhaps we can find some similarities in the crash...

I posted it here: http://diydrones.com/group/iris/forum/topics/iris-sudden-fall

of 6

This reply was deleted.