Toomanyplanes

Forensics looking into crash. More later

New update. See below

Toomanyplanes

OK. It's been two weeks since the crash of my sailplane and I have investigated the damage closely enough along with the events as they transpired the day of the crash. I had been flying my Superhawk with my JR PCM 10XII on channel 23 for roughly 10 hours of total flight time with no problems whatsoever. I have owned and flown this system for the last ten years and never had ANY radio problems UNTIL I bought and installed a SPEKTRUM 2.4 Ghz module for this particular radio. The crash that occured was the very first flight with this "upgrade" module installed. The plane is a total loss.

To make a long story short, The radio was range tested per the manual before a winch launch and at home that same day. The battery was fully charged and in good shape. The servos are two new (no crashes until now) HiTec HS425BB standard Ball Bearing servos and are the same ones that were used since the plane was built (no problems at all). After the range test, the plane was launched and was climbing out normally until the apogee. During the dip down on the top of launch the elevator failed to respond to nose up command and the plane nosed down into an inside loop until inverted. Straight line distance was about 800-1000 feet and now the plane was inverted and heading back to me. I ran toward the plane to close the distance and hope to regain control but no response. I could clearly see the spoilers out in the preprogrammed fail safe mode so it was obvious something was wrong with the link between the Tx and Rx. As the sailplane passed overhead about 200 feet away I looked at my transmitter and saw 100% and two green LED's but I turned off the transmitter and turned it back on to try and re-bind the two with no luck. The sailplane took a steep dive inverted into the ground and destroyed the plane. As I walked up to the plane I wiggled the rudder and could see it moving. The stabs were pretzeled into odd shapes and the pushrod had broken the brass coupler, bent the metal end of the coupler and snapped off the mini-link from the bel crank impaling itself into the servo deck due to the force of the crash. The elevator servo was moving to stick inputs as well as the rudder. Working OK now.

There is very little carbon in the fuselage of this plane. Two 1/4" tow reinforcements down each side of the fuselage and the balance of the fuse is epoxy fiberglass. lots of open RF path.

I really hoped this new technology would work well but for me it was a complete disaster. Had this been a 1300 dollar moldie I would have been alot more PO'd.

I returned the Module and Rx yesterday and will not try another radio using this technology until more field use and testing is done. Until then its back to the Jurrasic reliable 72Mhz.

If your thinking about using one of these modules to update your radio I'd definitely think twice on it. I'm sure some folks have had no problems but I wasn't so fortunate.

Hope this will help you make a choice based on my experience. Maybe something good cane come out of it.

Thermals,

Tony

A10FLYR

My Feedback: (1)

Darn, that's too bad! I have two of the DX6 TX with 16 of the AR6000 RX and have only had trouble with install. That was in an electric powered bird. (Brushed) My guess is you had a very high current draw at the top of the launch causing RX to drop out and when you tried to "rebind" the RX was getting ready to do that on it's own so now it has taken twice the time to rebind. What RX were you using? It seems that every instance I've heard about low voltage could very well be the culpret.

Toomanyplanes

Yah... Well I wish I knew the answer. I wouldn't think a standard servo would make a huge draw on an 1800 Mah battery pack that was freshly charged but who knows. From the time I lost the bind to where I cycled the Tx on/off it was about 45 seconds. A Looooonngg time. The plane was flying inverted and did a couple of stall/recoveries. It seemed like an eternity until it crashed but it was about one minute.

The receiver was the AR9000 that came with it.

Now I'm really aprehensive about the new technology and probably won't try to use it anytime soon as it took a long time to build one plane. Just glad it did not hit someone or something during the mishap. That would have been even UGLIER.

da Rock

Waiting for others to prove new technology beyond doubt is a good and safe thing to do for lots of people.

And keep in mind that your one failure isn't a closed case.

You actually need to prove something about the system that caused it to not perform that one time in that situation to say the crash was the fault of that system. There was obviously something wrong, but what was it? Spektrums go out on hard launches? They don't. The launch was too high? Nope. What about that launch situation was something the technology couldn't handle? or didn't handle? And will repeat? even once more.

Sorry for the crash, but there was a cause, and just having different signal technology wasn't the reason. It could have certainly contributed, but how? Antenna blanked by the winch cable? Blanked by the CF tow strips?

BMatthews

I'm sort of thinking that there's actually something wrong with the Rx or Tx in this case as opposed to the technology. I've heard of lots of folks using DX7's without any range issues. Especially at this short a distance. So I'd be more inclined to suspect equipment failure instead of technology inadequecy.

Anyhow, none of this makes it any easier to accept the loss of an old companion. I know that I've got a few older models that I'd hate to loose to this sort of thing.

Toomanyplanes

Well I would hope to try and be logical about the failure as much as possible.

I would like to say it was not the radio as I like the concept of the 2.4 Ghz systems. I had flown this plane both on the slope and on several winch launches using the 72Mhz system to "speck-out" altitudes with NO glitches or problems. Total flight time on the plane was between 8-10 flight hours and at least that many flights. This mishap occured on the first flight using the Spektrum module upgrade in my JR radio. Book range test OK at 100 feet but at 800-1000 feet away in flight, failsafe engaged (visually verified) plane flew approx. 1 minute to 1:30 right overhead (within 200 feet) with no control response UNTIL after the crash.

Yes, anything could have happened and you folks are savy enough to draw your own conclusions. Would you give her another go at one of your expensive planes?

Returned the system to the dealer yesterday for a refund and explained this to the owner. They are going to return it to Horizon and they may or may not even look at it. I am washing my hands of it for now as I just don't feel like taking any more chances after this. I am just THANKFUL nobody was hurt or any property was damaged.

I can only hope that Spektrum will at least look at the system to find the cause of this mishap.

Post crash photos attached.

Thermals,

Tony

Toomanyplanes

Toomanyplanes

da Rock

Please resist the urge to curse, flame, degrade, insult or embarrass someone in your post.

It's better to talk about the topic than to talk about other posters. When you make comments about other posters, at the minimum, you're apt to embarrass.

da Rock

Toomanyplanes,
I don't see any antennas in your pictures. And I've reread your posts a couple of times and didn't see any mention of how you dealt with the dual RX installations and the antenna's for them.

The AR9000 has two receivers and each has two antennas. The primary has two that should project at right angles to each other. How did you get them to do that in that fuselage? Matter of fact, how did you arrange the two recievers, I bet that took some doing. Are all 4 antenna supposed to be pointing in different directions? It seems like that arrangement would be almost impossible in a tiny glider fuselage. How did you manage that?

Toomanyplanes

Being asked to "prove" the system failed seems like something a manufacturer rep would say. Kinda got me going a bit on that one... Didn't mean to "Curse or Inflame".

Toomanyplanes

Actually the fuselage is fairly large in x-sectional area as it is patterned after the original Hobie Hawk fuselage but 25% larger. The nose pod was big enough to support the old Kraft bricks back in the good ol'days. The receivers were mounted behind behind the deck in the roomiest part of the fuselage. The antennas were held in position by carving out the foam block cushion to accomidate the antennas. I suppose it's possible that the antennas shifted in a hard launch but when they were pulled out they were still in position.

Look, the radio took a dump for some reason like you said. I am reasonably cautious about trying it on another plane again.

I'll give it another try down the road but for now I'm a bit aprehensive.

BTW, Have you heard anything from any of the manufacturers of the new systems what is considered 2.4 Ghz friendly as far as fuselages are concerned??

Thanks,

Tony

da Rock

How did you manage to arrange the antennas at right angles to each other and keep them from being close to the CF strips?

If you place a wire close to a 1/4" strip, it can be shielded over a fairly wide range by what would seem to be only a 1/4" wide.

If those antenna are very sensitive to being placed close to shielding, it would be very worthwhile to know. See what I mean?

I'm actually quite sorry you lost your glider. I think losing gliders is about 10X worse than losing any other kind of RC, as a matter of fact. And a HobieHawk is such a beautiful thing in the air, it's gotta hurt even worse.

Toomanyplanes

The receivers were very close to the 1/4" CF strips. Hard to believe that this could shield the signal as long as it did, but seems to be a reasonable explaination since the radio was working post crash. I'd be willing to bet there was some sort of resonence as well depending on the length of the CF.

The Hobie Hawk is unique but this sailplane is a total re-design of the original Hobie Hawk with spoilers and about 950 in sq. wing (120 inch span) and fully winch capable. The ultimate RES machine! This was the prototype and was flying superb. I was very happy with it but now I can fine tune a new one even better.

I have the tooling for this plane so I can build another one.

I had posts earlier detailing the first flights several weeks ago.

In a way, it's good that this happened to me instead of someone who paid big $$ for it and had a problem similar to mine. I will eventually build one as 2.4 Ghz friendly. No CF anywhere near the receivers!! CF is a great material but it can be substituted by other composites. Will give that one a try when I get up the nerve again.

The prototype that crashed had a 1/2" CF wing joiner 12 " long (cracked on impact but did not snap, rather impressive actually...), two 32" long X 1/4" CF reinforcement strips in the fuselage from nose to tail, full length. The spar is endgrain Balsa with CF caps extending 28" into the root of each wing. The newer versions will use an aluminum wing joiner, S-Glass and Kevlar. No more Carbon!!

Cheers

da Rock

I've got a feeling that the antennas are going to be a problem area for these systems. The requirement that they be at right angles and their length are really outside the envelope for most gliders. Guys used to build antennas into a wing for other reasons than range, but I got a feeling we might see things like that for these critters.

BTW, it's also hard to believe the strips would have shielded the antenna, but strange things happen. I assumed the CF wouldn't have been between the RX and TX anyway, but it's best to find out every detail before assuming anything.

I think we do have something to learn about these things. I'm working with a student whose DX7 seems to have a problem bonding every once in awhile at startup for a flight. It's never faltered on ground checks. But every once in awhile, we've got to turn off both and try again. Not comforting at all. Trust me, we're trying our best to see some detail that repeats in these situations but haven't come up with anything. And his is the only one out of about 10 or so DX7s that're active at that field that has this problem. It's not the only one being buddy boxed either. I've got another student whose DX7 nails the startup almost instantly. Time will tell.

da Rock

BTW, what is that "wire" that comes through what looks like a CF tube toward the nose. It looks like it bends down to touch a quarter that's at the very front of the cockpit floor? I been looking at that eversince I saw the picture and puzzling.

thermal sniffer probe?

I figure it's a delivery tube for shifting ballast. You've got the balast compartments lined with heating elements that melt the ballast lead. The tube goes to a pump that kicks in after the lead is melted and transfers it from one compartment to the other. And the quarter is simply the cold lead surface with impurities. Close?

BMatthews

The four wire stubs on the DX 7 Rx are actually only two antenna. It's a little different because they are using dipole antennas instead of end loaded. Basically it's two 1/4 wave whips that work together to feed the one input of each module.

Shielding of antennas involves a LOT more than just blocking the signal. If you had the wires running very close to a carbon strip it will not so much shield the signal as detune the reciever with the close proximity conductor acting to direct or draw signal away from the actual antenna. But in your case if you did have the two antennas on the two modules running at 90 degrees to each other the second cross polarized unit should have brought it back in synch. Unless you had one arranged it so that one ran close and parallel and the other sat crossways but with the ends very close to the carbon strips then the carbon would or could act to detune BOTH modules. Also it's important to note that while the two units should be oriented at 90 to each other the two stubs sticking out of the modules should be kept as close to inline as practical. Orienting the dipole stubs of a given module at 90 to each other will hurt the pickup.

Why two RX modules with separate antenna you ask? That's because of the frequency. The same frequency use that gives us these delightfully short antennas is also the reason we need two Rx's. These higher frequencies are more easily blocked and more easily fooled by antenna orientation between the Rx and Tx. Having two "voting" Rx's with antennas at 90 to each other ensures that one will be working when the other is momentarily blanked or in an orientation to the Tx where it can't get a good signal.

This used to be pretty much standard knowledge in the "old" days where no one in their right mind would run an antenna along side a full length wire pushrod. That's a big reason why the old school models used built up pushrods with wood for most of the length. It's only in the past decade or two that radios got good enough that we were able to run the antenna inside and along side wire or carbon pushrods. But knowing what I know from being trained as a telecom tech I still cringe and cannot bring myself to commit such a travisty so my gliders and electric assist gliders that I fly typically have a plastic tube running out the wing to get the antenna as far from the "stuff" inside as practical. My old timers that also fly out at long range have the antenna inside the fuselage but in those I use the old style wood pushrods with short metal ends and run the antenna as far away from the metal as practical.

I saw a thread recently by someone that did a lot of testing of the DX7 system in various fuselages. The carbon in one made a very definite difference so it may well be that we can't get away with setups like we could with the 72 mhz systems.

wind junkie

My Feedback: (1)

I think you're referring to my thread which I started on RC Groups. There is a continuation here:

http://www.rcuniverse.com/forum/m_57...tm.htm#6142772

Sorry to hear about your misfortune Toomanyplanes. After reading this thread I was careful to do a range check before using the new spektrum modeule in my 9303. So far so good in my case.

ZebraOne

My Feedback: (8)

Ahhh...
2.4 is great for inhouse phones!!!!
CF is conductive.
Some of the CF threads MAY actually be the same physical length as the electronic frequency of the Xmitte, irrespective of it's frequency hopping capabilities.
.I have too many xmitters that work to try out the new stuff yet!(G)
Zebraone
AMA129570
KE6OKM

Toomanyplanes

OK guys. New update. The saga continues.

Bet ya want to know what happened on the day my Superhawk crashed with the brand spankin new Spektrum module installed in my JR 10SXII radio? Well read on...

It bugged me that so many guys install this new technology in their planes and have no problems whatsoever so I decided to rebuild the plane exactly as it was the day of the crash and try using the Spektrum equipment again and see what happens.

First off, the battery was not the problem. Used several capacity checks on my charging equipment and no issue there. The battery was fine.

Next was the servos. The servos were fairly new HiTec 425BB standard servos in the fuselage and Hitec HS55's in the wings for spoilers. All were used in the plane for approx 20 flight hours before the crash with no problems (using the old FM equipment). Checking the servos out post crash found all servos operating normally even after wiggling the wiring around on the servos to check for loose connections.

The plane was rebuilt in the exact same way, including Carbon fiber reinforcement and I recently purchased a new JR 9303 2.4 Ghz radio for the next installation (I also purchased a data logger to check signal performance). I had a tough time convincing myself to do this as the first try at using Spektrum equipment was a complete disaster... I had to find out what went wrong.

I installed the servos and wiring just as it was on the plane that crashed. I then did the range tests after re-binding the radio gear as described in the JR manual and setting the fail safe control positions with the spoilers up, making a loss of signal easy to identify on the range tests.

At home I did the range test at 30 paces and all seemed well, just as before, but I noticed the Frame losses on the data logger were about 100+ even for a short test. Frame losses are a measure of loss of bits of data in the signal. More on this test later.

I took the plane to the same location I had the crash last July and set the plane on the ground and lost signal at 50-60 feet as verified by the spoilers coming up into failsafe. Puzzled, I checked the data logger and it showed again 100+ Frame losses. Something was messed up as this was a whole different radio and it was obvious something was attenuating the signal. The first thing that popped into my mind was the Carbon fiber tow running down the sides of the fuselage. I suspected this on previous posts so I did not fly that day (wise move!) and took the radio gear out of the fuselage to figure out how to install the receiver antennas so they were as far as possible from the CF reinforcement.

I decided to mount the receiver on it's side with one antenna sticking up through the flight deck (via a small hole) and the other parallel to the fuselage bottom and held straight with a small length of poly tube slipped over it. This way both A and B receiver antennas are perfectly straight and at 90 degrees to each other. The remote receiver was mounted behind the flight deck on the inside upper surface of the fiberglass fuselage using double sided tape with the antennas 90 degrees to the A and B main receiver antennas but following the round inside contour of the fuselage. Hard to picture, but as recommended by Spektrum and followed to the letter.

I went outside at home and tried the range test again. 1 frame loss at 100 feet. I went 200 feet and intermittantly lost control on the range test but that was to be expected. I read that if you have a helper or stand the plane up against something (ie get it off the ground) during the range test, the result is a bit more consistant on tests. I placed the plane on my deck railing at 10-15' and tried the range test again. I could not walk far enough to loose the signal and still see the plane. I'm guessing that was 250' or so.
Looks like I found the problem. Carbon fiber is an attenuator to the 2.4 Ghz signal. Not only that, but it is not how much is in the airframe, but how long it is and where it is. There are certain combinations of conductive materials used to build our airplanes that definitely affect the ability of the equipment to function properly. When I say conductive, I mean metals as well as CF. Anything that has the ability to block or conduct the signal. You simply MUST experiment to find the optimal location and orientation for your receiver antennas especially if you fly planes with lots of metal or CF.

My antennas on the main receiver were right next to the CF reinforcing in the nose pod and it attenuated the signal so much, the receiver could not re-link even after almost one minute. All my antennas are now installed over fiberglass areas of the fuselage with no CF nearby.


I test flew the new Superhawk today and flew it for one flight lasting 1 hour and 15 minutes. Absolutely no problems with signal loss. I flew as high as 1000 feet and sent it as far away as I could comfortably see it. No glitches. Perfect.

I hope this helps those of you planning to use the new technology in your sailplanes or any plane for that matter.

Thermals,

Tony

h.eberbach

Tony,

Bruce explained it and you proved it! :-)))

Receiver diversity and anrenna configuration will be one, if not THE, crucial aspect when unsing 2.4 GHz - particularly here in Europe, where transmitter power will be much less than in the U.S.

Happy landings
Herbert