/dist/images/branding/favicon

I got Jammed today by a portable construction marque

Discussion in 'Cinestar Misc' started by Shaun Stanton, Feb 28, 2013.

  1. Shaun Stanton

    Shaun Stanton Active Member

    Joined:
    Oct 27, 2012
    Messages:
    748
    Likes Received:
    127
    I thought I would share this since it could happen to you.

    I was out for a quick shoot right at the golden hour. Right across the street from my residence is the Old Santa Fe rail yards, a former steam locomotive depot MX center. This sight has appeared in several movies such as Avengers, Transformers, X-men and several others including some TV shows. We are going to be doing some aerial for the local studios for these. We went out to do a practice shoot to fly around an old stone building that served as the fire station.

    We went to an area that was adequate to launch from without obstructions. I did not account for a construction advisory sign as being any concern that was about 15 feet away. I launched the aircraft and climbed to about 20 feet or so when the aircraft started to hickup and the LED's which were off, started to blink. Realizing something was wrong was occurring, I had the Cam Op stow the Gimbal. My control of the craft was intermittent. Thinking that I landed it, I found out after the fact that it had touched down on it's own:eek: After it landed, I shut the motors down immediately. I restarted the motors again to see if I could recreate the event with the craft on the ground and found out that it did not. I did two iterations of this and no re-producible error. We decided to do the smart thing and not put it back in the air and try our luck. I decided to terminate flying and investigate the incident.

    What I found after analyzing the GPX file is that I had 16 RC Lost and Edit: 4 Emergency land annunciations. I went back to the area to see what could be causing the AC to loose comms to the DX8, then it dawned on me. There was that construction marquee advising road work that was going on a few blocks away. Anyone who's ever owned a radar detector knows that some of these signs have a drone radar gun designed to force drivers to slow down in these areas. I looked at the sign and sure enough it had one. What I deduced is that I flew into the side lobe radiation of the radar that could cause the right frequency harmonics to JAM the 2.4 GHZ DSMX satellite enough for the aircraft to think it was lost link. Luckily good old MK did as advertised and it landed. The landing was safe albeit a little firm. It managed to touch down in one spot with no damage to A/C or the 5DmkII. :)

    The lesson learned here is that what may appear as a least significant thing in the environment can pose significant threats and may be detrimental to the safe operation of flight. Luckily the GPS or magnetometer was not affected this time and the craft did as advertised, but there could be times when this may not be the case.
     
  2. Steve Maller

    Steve Maller UAV Grief Counselor

    Joined:
    Oct 30, 2012
    Messages:
    3,981
    Likes Received:
    807
    Wow, I could say you were lucky, but I know better.
    You're good.
    Thank you once again for your astute analysis and for sharing it here.
     
    Shaun Stanton likes this.
  3. Shaun Stanton

    Shaun Stanton Active Member

    Joined:
    Oct 27, 2012
    Messages:
    748
    Likes Received:
    127
    Thanks for complement,

    I certainly do not mind giving some credit to "Lady Luck" on this one;)
     
    Steve Maller likes this.
  4. Andy Johnson-Laird

    Andy Johnson-Laird Administrator
    Staff Member

    Joined:
    Jul 31, 2012
    Messages:
    10,383
    Likes Received:
    1,164
    Hi Shaun:

    I think you may well be right about flying into the side lobe. I think radar speed detectors typically operate in the up around 33.4 Ghz. -- obviously if you get close enough, you can swamp the front end of the 2.4 Ghz RC controller. How far from the radar Tx were you?

    I've often wondered whether it makes sense to use the "Range Check" feature that most RC transmitters have -- basically when set it reduces the radiated power to a fraction of normal levels -- the idea being that you can then walk 30 yards away and see if you still have RC control of the aircraft. If you do, then you can be pretty certain to have control at longer ranges.

    One other important question: What did you have the Emergency Gas set to, please?

    Thanks
    Andy.
     
  5. Shaun Stanton

    Shaun Stanton Active Member

    Joined:
    Oct 27, 2012
    Messages:
    748
    Likes Received:
    127
    That's what I figured as well because radars are typically in the K or KA range. I was about 15 to 20 feet laterally from the sign.

    Probably a good overall practice. I might go out there tomorrow and test it. I wonder if that would have made a difference in this case since it did not seem to be an issue with the aircraft on the ground. It was not until airborne that I experienced the phenomena. I am guessing that on the ground it was not in the LOS of the disruptive radio harmonic. But it could have been a weaker signal than optimal though and putting it in the air could have made it more exacerbated.

    [/quote]

    I had it on the default setting of 65 not on the Vario, probably not Ideal. It seems that it went directly emergency landing mode which I am guessing it figures out a descent rate. You being the Guru to all that is MK, I would not mind getting some better insight on the logic on how this system setting works. The MK manual is kind of vague on its operation. My understanding is that emergency gas setting is active when away from the home location. In my case it went directly to a land. I never got a descent rate of greater then 2ft/s it did not drop out of the sky it seemed like a controlled descent until touchdown. It appears that the touchdown was at around -1.44 to 2 ft/sec.
     
  6. Joe Azzarelli

    Joe Azzarelli Active Member

    Joined:
    Jul 13, 2012
    Messages:
    210
    Likes Received:
    26
    If you had your failsafe set then it is using the gps for the descent rate to the landing, in which case the emergency gas is not in play..
     
  7. Shaun Stanton

    Shaun Stanton Active Member

    Joined:
    Oct 27, 2012
    Messages:
    748
    Likes Received:
    127
    Thats true I did set that, just answering Andy's question and trying to get his insight though;)
     
  8. Andy Johnson-Laird

    Andy Johnson-Laird Administrator
    Staff Member

    Joined:
    Jul 31, 2012
    Messages:
    10,383
    Likes Received:
    1,164
    Joe: Can you expand on that a bit?

    For my research into MK Tool (for the latest DVD set) I had occasion to look at the MK source code and as far as I could tell, if you have the GPS and Navi, on Tx signal loss, the copter will do an Emergency Come Home and then descend at a fixed rate (see http://www.mikrokopter.de/ucwiki/en/FailSafe which uses the specific phrase: "If the signal is not restoring at that time the Kopter will fly back to the Homeposition and will switch there automatically into a fixed rate of descent and tries to land.").

    I was never able to find out what "a fixed rate" was or whether it was the Emergency Gas. At first I thought it was just a mistake, but then, if you scroll down below the diagram on that page, you see that there are specific references to EMERGENCY-Gas/Throttle so I infer that "fixed rate" is something separate from Emergency Gas.

    Also note the box below the diagram that says:
    IMPORTANT​
    If no "Failsafe CH time" has been entered in the settings, the Kopter will directly go down with the emergency-gas and emergency-time settings​

    But what I have NOT seen is the suggestion that the GPS is used to control the descent rate as you suggest.
    Can you point me to a citation that says this please -- I'm going to have to issue an Errata statement for the new DVD set to correct what I said.

    Thanks
    Andy.
     
  9. Shaun Stanton

    Shaun Stanton Active Member

    Joined:
    Oct 27, 2012
    Messages:
    748
    Likes Received:
    127
    I lied I did not have my Failsafe set:eek:

    So I deduce that since I had partial control I aided it in its descent so when it did land it probably it did just plop out of the sky luckily that plop was 4 feet.

    I tried to recreate the event tonight without flying. I did not start the motors or a camera attached the gimbal was off so any influence from my systems two video systems the camera, the HDMi converter and the Gimbal itself or motors were not present. I put the copter on the to of the roof of my old landcruiser to see if it would recreate the anomaly. It did not.

    From my data I deduce I had partial disruptions in the digital bitstream that may have been prevalent when other stuff is on. I probably flew right into an unlucky sweet spot that caused the anomaly.
     
  10. Andy Johnson-Laird

    Andy Johnson-Laird Administrator
    Staff Member

    Joined:
    Jul 31, 2012
    Messages:
    10,383
    Likes Received:
    1,164
    So it sounds like the copter did the signal loss emergency landing -- good to know that worked.
    The curious thing is that if the signal comes back you'd have control, but if it goes away again, the copter should wait five seconds in AH to see if the signal loss is transient before executing the emergency gas descent.

    Andy.
     
  11. Shaun Stanton

    Shaun Stanton Active Member

    Joined:
    Oct 27, 2012
    Messages:
    748
    Likes Received:
    127
    That's what it seems like I figure I'll share the file so you can see what the event looked like. It would be nice if MK provided more docs on this file. You will notice at one point I had a complete RC quality of 0.

    I am real lucky that it did what it did and not what happened to Sabastian.
     

    Attached Files:

  12. Shaun Stanton

    Shaun Stanton Active Member

    Joined:
    Oct 27, 2012
    Messages:
    748
    Likes Received:
    127
    I got lucky but I question this system I wish more information. My situation was a little different since I was pretty much over the home location. So it obviously did not need to climb to go home. It must have entered that portion of the logic since it was already home. I guess I am lucky that it did not do some wild out of control situation like Sebastian, when he shut off his transmitter. Now on this board there are two different results. One case copter goes out of control, in mine it seemed to do as advertised.

    To summarize the logic tree from the MK site:

    1. Kopter loses signal
    2 Position Hold engaged for 5 seconds Climbs to the CH altitude IF the Kopter is >25m from starting point
    3. After 5 seconds come home is engaged and it will continue to climb or descend to CH alt.
    4. If link is restored then you have control of the craft.
    5. If not, it will fly towards home, and attempt to land itself until the 60 second timer is up, you actually get only 55 seconds for it to travel.
    6. If 60 Seconds are up it reverts to E-gas and in essence just blindly falls out of the sky until it hits the ground. After 27 seconds the motors shut off, hopefully not too high off the ground:oops: Or I guess there goes 8000 dollars. and a expensive camera.

    It seems to me that this asinine way to handle lost link logic. Why cant it just fly the H8** home and land if you have a good battery? To me the E-gas setting should be for the OH-S moment that you now are at a critical state and you need get the thing in somewhat of a controlled manner to mitigate the damage.

    I guess moral the story, don't fly too high or too far or else you may be SOL. It would be nice to know what that magical descent rate is to use in your planning so that way you can account for that in your 55 second transient time back to home station. I just just proved that is possible to loose link only a few feet away from the aircraft, so there may be a time when this thing will need to land itself, it would nice to have a little bit of a comfort factor than 60 seconds. You can get 247 seconds if you pay $750 to have the commercial licence.
     
  13. Andy Johnson-Laird

    Andy Johnson-Laird Administrator
    Staff Member

    Joined:
    Jul 31, 2012
    Messages:
    10,383
    Likes Received:
    1,164
    Shaun:

    Just one minor correction, in Sebastian's case the problem was that the firmware allowed emergency come home procedure to start even though the copter was the ground -- then the second factor (low battery -- which was why it was on the ground) kicked in and then it went out of control. Holger and his team have changed this and in the next version of the firmware, once the copter has landed it will suppress emergency come home.

    Your summary of the transmitter signal lost logic matches mine. It might be a slight overstatement in step 6. to say "blindly falls out of the sky" -- perhaps "blindly executes a descent at a reasonable vertical descent rate as specified by the user" might be closer to the truth. :)

    I can't argue with your thought that more than 60 seconds would be nice...that said, I've not done the math to figure out what that is likely to be in terms of operational radius from Home. If we assume a max horizontal groundspeed of say, 20 mph (or 8.9 meters/second), that would allow the copter to travel 495 meters in 55 seconds (allow for the 5 second position hold).

    Does anyone know what the likely maximum horizontal groundspeed is likely to be? I just made up the 20 mph. Sorry about the mixed units -- it's the result of having lived in England, France, Canada and the USA. :)

    I do think the transmitter signal lost logic is predicated on the copter being some distance away and the loss being caused by the distance or some obstacle -- the radar interference could well be something that Holger and his team have not experienced or thought of.

    Once the FAA puts FARs into place for UAS (sorry about the alphabet soup), I can see that this transmitter signal lost logic is going to be subject to considerable scrutiny -- the FAA will not tolerate UAS wandering around the sky (or leaping up from the ground) under these circumstances.

    Andy.
     
  14. Shaun Stanton

    Shaun Stanton Active Member

    Joined:
    Oct 27, 2012
    Messages:
    748
    Likes Received:
    127
    Yeah that's true his battery was dead, I am still glad mine did not do that since I never tested this close until now;)

    A little dramatic statement yours is more correct :)

    I tested this before, not on purpose lets say I had close encounter of the CareFree/(careLESS) kind. Yeah did the rooky mistake, and had a switch in the wrong position on takeoff No OSD to warn me, once I realized it I had let the aircraft go way out in BFE about 3/4 of a mile. I will never let that happen again. In fact that mode is gone in this baby altogether. It came back at about 20 to 25mph so 20 is probably a good round number. Not sure how much effect winds had. Got to test the RTL for the first time. ;)
    They probably did not. Radars or interference or a terminal failure in the RC controller at the wrong time should be a consideration.

    No way in hell, is the FAA ever going certify these in their current form IMO. So we will all have to fly under the "radar" no pun intended. On a more professional tone. These aircraft as you know as a pilot, would fall into the experimental aircraft category for restricted use only. When the day comes, we will have to be certified to fly these things and will have to be certified to build and maintain them. There is no way that I can see, the FAA will certify an aircraft for commercial use especially an RPA (Air Force parlance for UAS, UAV or Drone) that has multiple critical single points of failure that can cause the A/C to crash such as:

    • A faulty battery with no separate redundant back up where the bad bat is taken offline
    • Single channel communication system w/o the ability to use a backup controller
    • Single source primary control module i.e. the FC board without a backup system in some emergency mode
    • Single altitude reporting and heading reference/geo-spatial system.
    • Lost link logic that is not predictable on where and how the aircraft will return. i.e No Emergency pre-programmed flight plan.
    I am guessing that the future systems will have to be vetted with extreme scrutiny. Critical systems will have to be redundant, and no single point of failure will be tolerated. If I were to guess flying these on the ISM band will not acceptable because of potential signal disruptions and packet losses. These are just some of the few things that even the big guys US gov and our cohorts, Lockheed, Northrup Boeing and General Atomics are dealing with right as we speak. In fact I would not surprised if our future FC boards will be a Boeing product, or least LM. Lockheed is the first Major to get in the multi-rotor game.
     
  15. Gary Haynes

    Gary Haynes Administrator
    Staff Member

    Joined:
    Jul 6, 2012
    Messages:
    5,211
    Likes Received:
    460
    I think the default CH speed is something like 15m/s. It is one of the adjustments that can be made in the GPS PID settings.

    If you want to change the speed in which for example waypoints or ComingHome it's flown that can be set up with changed values of GPS-P and GPS-d.

    Increase/Decrease
    Increase speed = increase GPS-P + decrease GPS-D.
    Decrease speed = decrease GPS-P + increase GPS-D.
     
  16. Andy Johnson-Laird

    Andy Johnson-Laird Administrator
    Staff Member

    Joined:
    Jul 31, 2012
    Messages:
    10,383
    Likes Received:
    1,164
    Shaun: Good insights into the future of drone avionics (where drone covers fixed wing, single and multi-rotor).

    If our copters are type certificated as Experimental (as my self-launching sailplane was), I was not allowed to fly more than 50 miles away from my home field. Then I had to ask for the rating to be changed for flights within Oregon. Then I added State by State until I had Oregon, Washington, Idaho, California and Nevada. Basically, I had to demonstrate the the a/c was safe to fly. It took two years before I had all of those States. The FAA person couldn't quite get that the a/c was in a trailer and I could tow it behind a car to an airfield in Nevada and launch from there. :)

    I guess we better all be studying up to get our A&P (airframe and powerplants) ratings so we can maintain our own copters -- I was allowed to do basic maintenance on the a/c and powerplant, gel coat repair, but nothing structural. But I still had to get the a/c signed off by a A&P mechanic every year -- which I had no problems with as it meant the aircraft would be that much safer.

    We're going to have to smarten up our act when the Sheriff comes to Dodge.....
    Andy
     
  17. Joe Azzarelli

    Joe Azzarelli Active Member

    Joined:
    Jul 13, 2012
    Messages:
    210
    Likes Received:
    26
    Perhaps I should have said that the Nav board controlling descent - as the pressure sensor may be what is being used. In any case, within the failsafe 60 second window the computer is controlling the fixed rate of descent. After 60 seconds if not on the ground, the fixed POWER setting (emergency gas setting) will go into effect. Although I am too much of a coward to actually test it...yet.
     
  18. Andy Johnson-Laird

    Andy Johnson-Laird Administrator
    Staff Member

    Joined:
    Jul 31, 2012
    Messages:
    10,383
    Likes Received:
    1,164
    Errr....now I'm really worried, Joe....the pressure sensor is on the Flight Control board, isn't it? :)

    But I'm with you on the careful application of cowardice towards testing it. Maybe when the CS8 becomes just "my other copter," then I might try it.

    Andy.
     
  19. Shaun Stanton

    Shaun Stanton Active Member

    Joined:
    Oct 27, 2012
    Messages:
    748
    Likes Received:
    127
    I have an extra PDB with a burnt out ESC but still works the other 7. I have been toying with the idea of putting the MK on my Cheap DJI CLONE HEX and going out into the desert and do a several tests in different configs to see what it does in scenario's, although not sure if I want to risc screwing up my MK stack by doing something potentially not ideal for electronics.
     
  20. Shaun Stanton

    Shaun Stanton Active Member

    Joined:
    Oct 27, 2012
    Messages:
    748
    Likes Received:
    127
    I speculate with my situation since I did not have the FS setup properly it was going in and out of E-gas which is why it did not just fall out of the sky at the setting I had it in.

    • 21 Feet Its signal gets disrupted and falls
    • I get momentary control back descent is arrested, I think I am flying still.
    • It responds to my control inputs in until the signal gets disrupted again LEDS go off it falls some more
    • I have control again I am attempting to bring it down It falls some more.
    • Control restored again I can arrest the descent a little.
    • Then LED's blink again then it falls one last time until landing firm
    I am guessing that if it was a complete loss of signal it would have just fell out of the sky with 65 on the GAS not in vario setting. For some reason I thought I had this set before. Apparently not I am usually good about logging every change I do to the system. At some point I got a little complacent. Long story I guess super lucky that this cat landed on all 3's.:D

    It would be nice to know what its logic tree is and how it performs in FS mode over the launch spot.
     

Share This Page