In that last closed "crash" thread, I commented that I have seen the similar errant "failsafe" behavior in a Futaba 9ZAP using PCM (maybe I am on a lot of "ignore" lists!!).

Anyway, here is what happened. I was starting my plane, kneeling next to it. Suddenly, the gear go up (we have all done it), but I could swear I did not hit the retract switch.

Turns out another pilot, using a Futaba PCM transmitter on an adjacent frequency (!), overpowered my signal and could take control of at least SOME of the channels (probabaly all). At the time, it was not apparent which channels beyond the retracts were affected, as his neutrals turned out to be pretty close to mine, so everything else on the plane just stood still.

Later on I put my plane on a stand and watched my plane respond to his while he was flying. His neutrals and ATVs were very close to mine, although a couple of channels were reversed.

Now I know this is not the same scenario as David Shulman's, and might not have anything to do with the other crash, I just wanted to state that I have seen (with my own eyes), a case where an INTERFERING transmitter (on a different channel even!) could take at least partial control of a PCM radio (in this case a Futaba 9ZAP) WITHOUT causing a failsafe to occur.

At least in the Futaba case, I do not think this is a bug per se, but rather a statistically rare event.

For what it is worth,

Hey Matt; was the other Tx a fut? and which model? Sounds to me that his Tx module needs a tune up (or even yours). As you know, the output transmitting transistor gets old and weak and drifts a little bit with age. How about You try to see if YOUR Tx does the same to HIS plane!!!

Matt,

My guess would be that his transmitter was slightly mistuned and your RX was too, and his transmitter basically took over your RX (your RX locked onto his TX more likely). I do not think it would be possible for a second PCM TX to take over PARTIAL control of your aircraft - sending commands to some of the servos, but not all. That would take some pretty fancy synching up to replace/modify some of the bits in a series of packets, but not others. Seems highly unlikely...

Bob

The one thing I can see in the previous thread was the fact thatthe movement could have been extremely minute and only for a millisecond, and at the speed of which he was flying and altitude (really low) it made a difference. If he had been at 100 fee thigh on the back side of the circuit it would probably have looked like the wind bumped him or something. He would probably not even acknowledge it.
I do not see the fail safe as an absolute. Obviously there will have to be a small lapse in data for the fail safe to kick in. Perhaps that is what we were seeing with the apparent instaneous recovery, a combination of recovery to fail safe and or David's reflexes. We put too much emphasis on this fail safe feature for an item that comes along with a radio system that costs less than a thousand dollars to put together. I would hope that one properly set up would at least shut down my turbine prior to impact, if I am high enough, but at the speed and altitude of this crash it was pretty obvious that nothing much would have saved it. The discouraging part of this is the fact that no matter what precautions we take, some yahoo not paying attention could caused something much more signifcant than the loss of the plane, it could have taken someone's life. I do not blame the management of the field. I have been there and it is an awesome place and quite an experience, but it is just too damn big to control everything that is going on. Suppose some guy in a camper up there decides to reconfigure his trims on his plane before the next day and turns on his transmitter while a guy is down on the flight line flying 3D circuits with his large gas helicopter. If it goes into failsafe, I don't want to be around it when it goes wherever it is going. I know I sound like a "girlieman" here, but I think we are missing the forest for the trees. There is nothing we can do to prevent it using our most utilized technology at the present (72 Mhz), and there are issues with the new stuff as well (2.4). We are going to have to do a better job of being vigilant in and around the flying fields and events with radios and where we fly our aircraft. Over the past two years we have seen some pretty dramatic crashes and even some people killed, some of these by the noted best pilots in the world. Their machines were considered to be state of the art.
Tommy

I agree that partial takeover has got to be very close to statistically impossible for the reasons mentioned above.
A full takeover is very possible.

Was your antenna down at the time and his fully extended? If so his signal could easily be a lot stronger than yours at your RX.
Especially if your RX tuning may be a little off one way and his TX may be a little off the other way.
When you did the test on a stand, was your TX on? Antenna up?
He would have to have been flying a Futaba since none of the different brands of PCM will talk to another.

Hi Bob,

I agree, I do not believe that the interefering TX took over a partial set of channels (like I said in my post I think it took control of all of them), I just wanted to be clear when I FIRST saw the event (my gear retracting) I could not observe the effect on the other channels. They did not move.

Later testing revealed that his radio took over ALL of the channels. When my gear retracted, he might have had his tranmitter sitting on the bench or was not touching the sticks, so all that would APPEAR to happen was that his TX reversed my retract valve position. I am sure (but cannot prove) the radio had control of all of my channels.

Yes Ehab that TX module got fixed with a big hammer and a trip to the round file!

Not just that - it would have to not only modify certain bits in the primary data section of the frame, it would presumably also have to modify the EXACT correctly corresponding bits in the CRC or whatever error checking mechanism Futaba uses, would it not ?

Assuming even a half-decent CRC algorithm, that IS definitely a possibility in statistically rare cases, but I don't see how it would be reproducable - seems more likely that if you left both TXs and the RX on continuously and recorded every frame sent from both TXs and every frame received by the RX, you'd get maybe ONE bogus-yet-accepted frame every day, or similar large period rather than being able to cause it to happen at will.

Also, if it was a case of the stronger TX being able to take over as suggested above, then this is different than Dave's scenario as I understand it. I believe Dave indicated that the acceptance of bogus info occurred only when the 2nd TX was some distance away - as it came closer (i.e signal strength increasing), the regular failsafe kicked in as expected. If the 2nd TX was stronger and better tuned to the RX, then as it came closer it should have been even MORE able to take control. (Unless I'm misunderstanding the "closer' aspect, and the 2nd TX was only getting closer to the 1st TX, but simultaneously getting SIGNIFICANTLY further away from the model ?)

One last thought ... instead of considering this from the point of view of two signal sources "blending" to create one which the RX works on, is there any chance this is some artifact of TX 1 inadvertantly picking up the signal from TX2 and absorbing that into the data that it's passing to the CRC encoder ? I understand a reasonable amount of DSP, but am not an RF guy by any means, so please don't laugh TOO hard if this is ridiculous. Maybe the fact that the TX circuitry is 'tuned" for the freq in question makes it susceptible to incoming RF of the same freq, which is why only another TX on the same channel causes such issues but all the nearby TXs on other freq's don't ? If that were the case though, it would be susceptible to picking up reflected "bounces" of its own very recently transmitted signal though... so this seems not to be feasible. Sorry - just thinking out loud.

Gordon

And yes they were both Futabas, and it occured with my antenna partially down.

Again, I want to stress that my post was ONLY made to show that events can occur where one PCM radio can overpower and take over control of another without causing a failsafe.

It might be a rare event or require a strange set of circumstances, but it can occur. I have NO IDEA if this is what happened in David's case...none.

Sounds like the HARMONICS of the 2nd TX (on an adjacent channel) got into Matt's Rx....

I think that the answer to this post is going to be really interesting - if we get one:

http://www.rcuniverse.com/forum/fb.asp?m=5892122

I can see how Matt's situation took place (a full takeover by another transmitter), but I do not know how a partial take over or an interference induced glitch could occur on a PCM system... Not to say that it couldn't occur, I'm just very interested in the details of how...

Bob

Hi Bob,

I think by your post you are implying that David's crash suggests only a single channel was interfered with (not trying to put words in your mouth).

In my case, the interfering radio took over all of the channels, but only the retracts were visually affected (meaning neutrals of the other primary control were about the same and the controls just sat still when the interfering transmitter took over).

I can envision a scenario wherein the elevator (or any other channel) on the interfering transmitter had a neutral very different that the "interfered with" transmitter. This could cause a violent pitch change (including a dive into terra firma from 5 feet) while all other flight controls "appear" to be normal (even though the interfering transmitter has control of them).

Agreed - but why would this behaviour exist when the (stronger / closer tuned / whatever) TX was far away, then disappear when that TX came closer ?

I'm not trying to shoot your theory down ... just trying to get it all to add up in my head. (Which it may never do, but hey - at least I can TRY to be an edumacated user)

Gordon

Don't know.

In my case it was obvious why it happened. My antenna was partially down, and the offending TX was close by with the antenna up.

In the real world you would never expect this to happen. You would have to have a fade of some kind from the "interfered with" transmitter (bad transmitter antenna orientation, weak or out of tune tx module, etc.) and the interfering tx might have had an ideal antenna orientation, strong module, right on tune with the rx, etc.

To be honest, when I first heard about this (a few minutes after it happened), I thought it could be elevator servo failure. I guess David has ruled that out.

All,
Well, I have setup as close as I can, the scenario David has indicated might have caused his crash. Using two JR 10 channel transmitters, one older straight 10 channel, and the newer 10X. Those are the only JR 10 channel transmitters I have access to at this time.

In my setup the newer 10X transmitter & S-mode 950 receiver are matched to the airplane. In this case a Dave Patrick Ultimate biplane. Four servos one for each aileron, two servos for the elevators / elevator drives Aux 3, one rudder, one throttle... An older relatively simple setup! I use two channels for aileron control, aileron and flap with a simple mix. Each servo top & bottom left wing is connected to a Y-harness, which then is plugged into the receiver aileron channel. Each right wing servo top & bottom is connected to a Y-harness and the receiver flap channel.

In my basement I setup essentially the old swamp test with each transmitter on the same channel. The older 10 channel being the interfering transmitter. With the interfering transmitters antenna collapsed and the newer 10X transmitter antenna fully extended both transmitters ON... the interfering transmitter will not cause a lockout condition until it is closer to the receiver antenna by more than half the distance of the operative transmitter. With both antennas’ extended lockout occurs with almost any relative positions of either transmitter.

I can NOT EXACTLY duplicate David’s findings of elevator movement with this setup. The operative / controlling R/F link always goes to HOLD or pre-set fail-safes if selected, when the interfering transmitter gets relatively close enough to activate the fail-safe.

Anyway what I did find and can duplicate time and time again.. Is that when both transmitter antennas are extended, I can continually create a strange situation by positioning the interfering transmitters antenna slightly closer to the airplane.

One of servos...the flap / left wing servo...one only... drives completely full travel to the down position. Yep, one servo only out of 8 servos in the airplane! It looks to me like that servo is actually getting an R/F signal directly not through the receiver (later I find this might not be the case! The strange movement / signal is getting to the servo through the receiver I believe.) Remember the interfering transmitter in this scenario is closer than the operative transmitter to cause this servo action. The servo will drive full throw that same direction only when two JR transmitters are ON. One transmitter alone will not cause a full throw movement of the servo! If I put the operative transmitters antenna within inches of this servo it will jitter a little but will not travel full throw?? Fact is other servos will jitter very slightly with close transmitter antenna proximity… something I consider almost normal and no flight risk.

Now for the really interesting finding…. Same scenario as described above.. But now the interfering transmitter is my Futaba 14MZ. What do you think happens to the servos now? Well, the same flap servo now moves full travel UP not down! And now my elevator servos in some cases will both drive full down also, and in other relative transmitter positions will split.. Yep one UP one DOWN... and the aileron / flap UP.

Keep in mind this test and these observations were made within my concrete rebar enforced basement walls. For me to cause the operative system to do anything but lockup / HOLD the interfering transmitter had to be closer than the operative transmitter. But the reactions of the 10X / S-mode JR receiver # 950 to JR interference was different than Futaba interference. My Futaba transmitter seemed to cause more strange servo reactions from a relatively father distance away than the JR 10 channel transmitter??

Who knows? what a turn on shoot down situation might cause in an open field? In my basement I could not cause the strange servo action unless the interfering transmitter was in a closer proximity than the operative / controlling transmitter. Until I witnessed this strange servo reaction for the first time today…. I would have said this would have been impossible at a flying field or in my basement…. But now I’m not so sure about the open field situation.

I can tell you I am ready to jump head long into the 2.4 world…if for no other reason than eliminating possibility of this type of shoot down. Shoot downs have always been my greatest fear when flying R/C aircraft. With a little luck maybe those shoot down days will be behind us forever once SS gets fully debugged and up to speed.
Lee H. DeMary
AMA 36099

Very nice writeup Lee, thank you.

I think Dave said that his Rx was an R2000 (?). I could be wrong , but it's pretty likely that the firmware differences are minute between the two.

I am considering (re)joining a club in my area, but the last time I flew at that site my 10x freaked out more than I have ever seen it freak. There is a large metal building about 150 yards (maybe a tad more) out from the flightline and parallel to it as well. If memory serves me, I was getting the most glitches when the model was 60-75% out to that building. Granted, I got the most glitches with my PPM Rx's, but I was getting some strange drift then lock then clear reactions from my S-PCM Rx in my AJ330. Would the building amplify the original signal before bouncing it back at the plane, and thus reproduce your findings, but with only one Tx?

Whitt
497494

Radio waves can be reflected and deflected... generally the bigger (lower frequency) the wavelength the more reflection, the smaller (higher frequency) the more it can be deflected. You must understand I am no expert; my knowledge has been gained by hands on experience with very little classroom electronic education.

As I stated in my basement test post.. Until I witnessed this strange servo reaction for the first time today…. I would have said this would have been impossible at a flying field or in my basement…. But now I’m not so sure about the open field situation.

Color me very skeptical about a single channel / firmware / fail-safe error of this nature happening in open field with transmitter(s) and aircraft relative positions being so far apart.....but then again something sure did push David's airplane quickly into the ground… that we all are very sure happened. I am not convinced yet as to what exactly caused his aircraft to do what it did.

My first impression was an elevator servo / glitch occurred on that pass when David flipped his smoke switch ON…but David would have already confirmed switch activation if that were the case... and we would have seen some smoke possibly....
Lee H. DeMary
AMA 36099

It's interesting what you said about the wavelength and reflectivity. My 10x is on Ch17 (lowest I could get for that radio).