mr_matt

My Feedback: (10)

Yeah like a cattle prod to the arse is "interesting"!!

So JR PCM failsafe basically does not work and Weatronics is no better?? What is going on here?

MiragePilot

My Feedback: (6)

Has anyone tried the Multiplex IPD receivers to see if they exhibit this "down elevator failsafe" phenomenon?

Peter

Pyolet

I am a little confused by all this "testing" you guys are doing, other than it supporting the claims that another JR transmitter was involved in the crash process.

What difference does it make whether the elevator channel programs down 14ms when the other radio takes priority control or it stays neutral? The pilot is not in control! Even if failsafe trips, you've been shot down.

The cause of David's Bandit crash was that some other moron was flying it!!! All this bashing JR and 20 year old technology is pointless to me. Please place blame where it firmly belongs. The fault here lies solely with the idiot who turned on his "dial a crash", unimpounded 72mHz radio, whether it was for just a nano second to drop his gear, or to intentionally shoot down the Bandit.

David said repeatedly that if he'd been using 2.4 there would have not been a crash. 'Nuff said.

Woody.

causeitflies

If it would have gone to hold for a nano second instead of down he MAY NOT have crashed.

David Gladwin

I agree with Pyolet entirely but in answer to Gordons request I did some more testing using JR synth Rxs and Tx modules. .

First test was again with both Txs on same frequency and similar programmes, results as posted earlier. There was NO intermediate failsafe position, either one Tx had control or the other had SOME control.

Then tried with Tx 2 on Z pcm. No intermediate failsafe position. Either the main tx had full control or the second, Z Pcm had, limited and very erratic control.

Then tried PPM. When second, ppm, tx got very close, Rx went into failsafe.

Conclusions: JR makes very good radio equipment , the impossible it can almost manage (I mangaged to safely land a model when someone arrived at our field in the UK and switched on his Tx on my frequency AND extended his aerial) but miracles (same frequency transmitters transmitting virtually identical PCM codes) is beyond it. JR can't bend the laws of physics however much we would like it to. 2.4 might be the creater of the miracle.

Since using the new JR synth Rxs and Txs I have not had a single failsafe recorded by my JetCat ECUS which are checked after EVERY flight and other models all have glitch recorders which have always shown a "clean" flight.

The JR manual actually states that failsafe works "in the event of loss of signal" not the reception of a slightly corrupted signal. There must be thousands of 10xs out there giving superb service, so stop worrying and get out flying BUT with positive Tx control, the system at NALL is critically flawed, in my view.

I repeat, David Shulman's Ultra was shot down, my tests confirm his, and Dave, I certainly regret your sad loss.

Regards,

David Gladwin.

rcdriver22

My Feedback: (1)

Tom
Tend to agree, is it usual in the USA for are all TXs to be tested by the CD initially for drift into adjacent channels? Its fairly common in the UK. As 2.4 becomes established people are safe to turn on randomly anywhere, anytime. As you may also shortly swop TX modules from 2,4 to a 35meg or 72 meg, there will be the momentary I'll just check the TX voltage quickly to see if the TX needs charging (its a long days flying). Switch on about five seconds to check (with a 72-35 meg online not 2.4) and David or someone else is in the dirt.

I mention this having lost two models, as the guy thought he was on a given frequency and took that peg, but was really on my frequency, result each time he switched on mine bit the dust!! Only found out as he was flying first one day, I arrived took my peg , switched on and craaamnmppp, his went in big style. The safety officer went to view the debris and inside his model was marked my peg number and sure enough he was transmitting on my number. **** happens, especially after the show crowds have left, people trade or sell gear from the car. In the UK shows many years ago show Traders sold whole sets of TX gear, and buyers sat in their cars and switched on their lovely news sets, result large complex expensive show models bit the dirt. Now crystals are removed and sealed till the buyers gets home. My residual concern is not JR or Futaba gear going AWOL but the novice turning up unannounced to cause a little havock.

Paul

Gordon Mc

My Feedback: (11)

It makes an ENORMOUS difference, and it sounds like you totally misunderstand the point of failsafe - it's not there to prevent your model from crashing, it's there to allow you to make informed decisions about how you can minimize risk to life & limb by reducing kinetic energy, providing a reasonably predictable path that the modle takes since it is not 'glitching' all over the place, and (in the case of our turbines) reducing fire risk by shutting the turbine down to let it cool before impact (if there's time).

1) If a failsafe kicks in and operates in the way I (and most others I believe) expected then the engine will shut down once the threshold (programmable, and no more than 2 sec) has been reached. In the "someone else is now in control of my model" situation, then that expectation goes right out the window.

The above means that our efforts to minimize the risk of fire / explosion on impact by setting our failsafe to kill the engine are being sabotaged. Now, quite apart form the increased risk to people and property that this represents - which of you wants to explain to the AMA that we are complying the best we can with the intent of that rule, but that the technology that is supposed to implement it doesn't work ?


2) Even if a failsafe did trip 1/4 second after the interference started, but instead of the expected "hold" for that 1/4 second it shoved in gobs of down elevator, that woud STILL be terrible. Despite the vertical maneuvers that we do, in my experience / opinion most of the jet guys that I know spend the majority of their flight right-side up and relatively level. In that case, if the failsafe operates in the "expected" way, for the first 1/4 sec the RX holds the controls where they are (i.e relatively neutral) and after that they go to the programmed position (usually still a hold, i.e still relatively neutral), then the engine gets shut down. In most cases that means that for a few brief seconds, the aircraft continues on a relatively predictable path until impact, and every moment that it stays in the air before impact constitutes a considerable reduction in fire risk (as well as a reduction in kinetic energy due to drag).

If failsafe kicks in, but only AFTER gobs of down elevator have been applied, then the "hold last" holds the down elevator and now the aircraft goes in much sooner than it ought to have - with more energy and with a hotter engine (read "increased fire / explosion risk" ) resulting.


The whole point of failsafe is to allow risk reduction w.r.t the people and/or property that a shot-down model may encounter, so having failsafe not work is therefore a risk increment instead.

Gordon

[Edit: typos]

Gordon Mc

My Feedback: (11)

No need to change the laws of physics, just the behaviour of the equipment used. If the system is seeing only one TX or the other, rather than an amalgalm that triggers failsafe, then embed a UUID in the frame. If the RX'd frame contains a UUID from a different TX, kick the failsafe into action.

That would have an issue of old & new equipment not being compatible, but I'm just pointing out that it's not physics that's stopping us - just the design / designer.

Gordon


Edit - sorry - just realised I used 'jargon' that many may not understand. UUID stands for "Universally Unique IDentifier". It's a common way of having one program entity identify itself to another, so that the second party knows who it is comunicating with. Put simply, it's like sending your fingerprint with each message ; the receiver sees the fact that the fingerprint it just received is NOT the same as the fingerprint of teh TX it was previously listening too, and can therefore differentiate between a validly constructed frame from one source, and a validly constructed frame from a different source. That difference could be used to trigger failsafe.

GrayUK

Im glad i fly Futaba.

rhklenke

My Feedback: (24)

David,

It is clear that you are a very experienced, knowledgable guy with lots of good information to share. However, you also tend to make broad sweeping statements some times that are more based on your opinion that on fact. The two statements above are examples of such.

First, as I've said several times, the incident happened AFTER the official flyin had shut down. Therefore, there was no "system" in place other than the normal "honor" system used at 99% of the fields around the world. Clearly there are issues with that methodology. When the "normal" frequency control was in effect at the Joe Nall flyin, NOTHING HAPPENED. Guys got their transmitters, got in line, got the pin when on deck, had their transmitters checked for proper channel, and THEN turned on and went flying. It worked perfectly. You have never been there and seen the process in action, so to criticize it, and the hard work done by those who have instituted it after 15+ years of hosting the largest flyin in the US, and probably the world, is not fair.

Second, your tests DID NOT confirm that David was shot down. In fact, you DID NOT suceed in replicating the effect that David said his plane exibited. I do not doubt that what he has said was true, but from what you have posted from your tests, you've simply shown that one JR transmitter can take over the RX when it overpowers the first signal. That effect is to be expected.

As Gordon stated above, that is not the issue. The issue is the apparent ability of the JR RX/TX in use by David to momentarily output erroneous commands to some of the control surfaces BEFORE the RX goes into failsafe or gets taken over by the other TX. I DOUBT that the second TX that was turned on succeded in completely "taking over" Davids Ultra. His antenna was fully extended and oriented such that the model was in one of the major sidelobes, so I'm sure that the signal strength at the RX was good. Most likely the second TX was on the ground or close to it, with its antenna collapsed and this its signal strength should have only been enough to "interfere" with David's and thus should have only caused a momentary "hold" or perhaps a short "failsafe." The fact that the JR TX/RX combo put in an uncommanded down elevator output is what caused the model to impact the ground at that point. If the hold/failsafe had worked properly, the model still may have crashed if David did not regain control, but it most likely would have done so further down the flight line and with the engine shutdown.

The JR system has proven itself VERY reliable over MANY users for many flights over many years, so I don't think anyone should be concerned, or abandon it. However, it would be useful to understand exactly what happened and why, so that failure mode is not duplicated in future equipment. If I was a designer for JR, I'd be very interested in duplicating this result and understanding why it happened for just that reason...

Bob

c/f

My Feedback: (1)

I did my testing as best to the scenario David described. My modules where Fixed. R945S Rx

A 10X anttena does not collapse fully so you get plenty of range on first section alone. You must unscrew it from the ball swivel for ground range testing to shorten the distance. So with that.

I set Master TX near Kingcat anttena collapsed first section, and operated controls normally. I then took second offending TX with antenna OFF and walked out of range of MASTER (too lessen range to observation of model) set it down and returned to Master to insure Master was under full control. I then went and picked up the offender TX and walked within range of the model with the offender, as soon as the offending TX was within range to cause a reaction, the elevator went to almost half down before returning nuetral prior to interferrence and the model was NOT in control on either TX and in full failsafe.

With two JR TX operating within range of model I never got any TX to control the model, only failsafe, If you turn them on within the operating range of TX or model, that IMO is not what David described. If I turned one on within range of each other the elevator blip does not happen. IMO where talking milliseconds of processor coding and on initial power up I'm sure its a differant algorityhm prior to online broadcasting.


Davids explaination and my testing explains to me WHY no scanner results were seen, My thoughts are someone flew on his channel previuosly, collapsed their antenna which leaves one length extended (10X), returned to their pit area, got busy with their model, and forgot to turn OFF their TX, This would be out of range of the scanner and Davids model and setup this Failsafe quirk as David entered into the limited broadcast zone of the offending TX.

The 9303 as Master or Offender did the same thing, The 8103DT did not, IMHO Jr engineering team who once had it right, went wrong and did not cover this base in sofware as previuosly designed in the beginning. My son has my old 10SXII so I will get that one and see if it works like old JR failsafe.

I'm also pretty certain the list of offending JR TX is small within the zone of the blip, who flew prior to David... And they KNOW who they are and must be freaking out with that giant skidmark on the field show center, the rest of the weekend. So what would be the right thing to do in this instance knowing you just trashed a $10K+ model and have no means to replace it, and this amount of time and Monday quarterbacking has passed???

L Turner

My Feedback: (3)

Gordon,

I think you hit the nail on the head. Although I do not completely understand how PCM works at the rf level, mfrs always advertized greater numbers of bits (2048) as being better than fewer (1024). Things could have been better if they had sacrificed 3 or 4 'bits' of resolution to include a specific UUID (universal unique IDentifier) linking one specific Rx to the one specific Tx intended by the pilot. Never understood why they did not go that direction.

Then with Rx failsafe logic designed to require the correct UUID for each frame, the ideal of failsafe 'reducing risks' by instituting preset servo movements and throttle changes (kill the turbine after x.x secs of either the wrong Tx signal or loss of the unique signal) would work better.

Perhaps this is the type of 'less primitive' failsafe design that we would want for the next generation of jet Tx-Rxs. Would some such logic be beneficial in 2.4 MHz systems?

S_Ellzey

This is the very baises of Spread Spectrum. Each radio has a unique code that is used to control the modulation of the transmitter, which, if I understand correctly, is what spreads the frequency out. If your reciever does not have the code it can not demodulate the signal, and like wise any other transmitter can not create a signal that your reciever can recieve. Your signal just looks like a small increase in the noise to another radio. Again if I understand it correctly the transmitter sends out a sync signal, which I guess can be used to let the other radios know that the frequency is in use.

Steven

GJr.

My Feedback: (3)

I have some questions to clarify some items for the guys doing some testing.

1) Are the uncommanded elevator, and other inputs due to the fact that the servo and channel programming in the interfering TX is different than the original TX programming and this is the signal being read by the RX? Or, is it a byproduct of the failsafe chacteristics?

2) Does it happen with any JR RX used, be it ZPCM, SPCM, synthesized SPCM, or PPM?

4) Does it occur exactly the same with the Weatronics and Multiplex IPD RXs, as they handle failsafe in a different maaner?

5) Is it exclusive to the JR 10X and 9303 TXs and new SPCM RXs? The previous series JR 10, 10S, 10SX, and 10SXII had the ability to program failsafe time settings. Was that feature only available when used with the ZPCM RX?

6) Is it possible to "reprogram" the JR 10X and 9303 TX to correct this occurrence if JR finds a problem?

7) Has anybody been able to recreate a similar reaction using either the Spetrum systems, the JR/Spektrum module updates, or the XPS modules? Meaning, is it a PPM /PCM issue and not a PPM/2.4 problem?

Thanks.

David Gladwin

Bob, Please accept my apologies, for the apparently sweeping statement, but I stated what is IS my firm opinion and conclusion base on the tests I did with both the JR and Weatronics Rxs and the information I have read so far. I don't know enough ( very little) about PCM to debate the issue, I'll leave that to the experts such as Gordon.

In view of what happened at Nall I will move to 2.4 , JR and Weatronic, in the near future, but only after I see more experience gained with the JR 2.4. I look forward to seeing DS and DB flying 2.4 in NI and discussing it with them.

I still think frequency control at Nall is flawed because we do it another way at events (and at my club field,) in the UK and Australia, no pin, no Tx . Guys in possesion of Txs to leave the field could well be asked to remove the RF module. Because there have been no incidents (except DS's) so far doesn't mean the system cannot be further improved.

What I am ever more convinced is that having a listen to my frequency before flying on my Icom 5 receiver IS good practice !

Again, my apologies, now I'm going flying and will leave this debate to others. !

Regards,

David Gladwin.

johnls

I was flying my Flash on its sixth flight last week. Tx was a JR PCM9 and RX JR10DS fed through a sensor switch to twin Li-Po's. Flying straight and level at 400 yd range the model suddenly pitched up recovered to level , and while I was thinking what the hell was that, it pitched hard down. Engine appeared to go to idle or off and I had no control over the model which was completely destroyed. Fail safe was set to engine off and controls to last position. I have now lost three expensive models in three months at this same flying site to radio issues, and in two cases the model pitched down.
Needless to say I am not flying there again.

John

Silver182

My Feedback: (2)

If he knows who he is...The right thing to do would be to whip out the checkbook...

I'll tell a few stories of shoot downs I'm familiar with. I said if he knows who he is because...I've got a friend whose radio locked up and hit a fence just after landing.. Because he forgot and left his own second transmitter (same frequency) ON) Yep, his own transmitter… and he got excited for a while looking around for who might have tuned on… ) )

I also shot myself down one-day years ago not long after getting my then new JR 10 channel. I was in the habit of starting the timer on screen just after takeoff, gear up, and a slight crosswind turn (the old pattern days remember em). Well this day I did just that, but I had just been using the trim offset screen.. to center up trims.. That was a simple neat feature on our older JR 10 channels before digital trims came about. Anyway you might have guess it yep... I hit the timer at my normal time.... but I'd forgotten to switch back to the flight screen, and as luck would have it the button for the timer in the flight screen was located in exactly the same place as the Store trim offset button was located. And it just so happens that when you store trim offsets for what I'm sure must be a good programming reason….. yep all input control to all channels is locked up.. Lost nada.. Well I just watched as my beautiful new JOKER went sailing off into the sunset... boom(

Another time many years ago, I got shot down by the owner of the local hobby shop! But he and I had made an agreement several weeks earlier because he was in the habit of coming out to the field setting up and switching on without even thinking about looking around to see if someone else was on the same frequency as he was. He did this all the time.. It was a bad habit he had developed and you might know it that was in the days when out here in the west we didn't have official frequency control at our flying site. That was in the 27mHz days. Remember Green, Yellow, Brown, ....etc. flags. Yep, the agreement was which every one of us shot the other down we'd whip out the old checkbook and write a check on the spot... he did. As I recall the amount was about $300 bucks.

I’m so happy / glad new days are ahead with the 2.4 stuff, just keep in mind there is more than one way to crash, and as the saying goes and it’s true, so very true if it flies it also will crash.

I liked David’s attitude in reflection during one of his last posts about the BIG crash he stated it’s now time for the popcorn…
Lee H. DeMary
AMA 36099

GJr.

My Feedback: (3)

Anybody have some responses to post #39?
Thanks.