Failsafe- Throttle down but what about the control surfaces?

[Simon Wright]

Another thought- How can you expect a model to self stabilise? unlike your self inflicted failsafe test where I assume you put the model a safe distance away in straight and level flight you have no idea when it will go into failsafe. half way through a loop? on a stall turn, low level beat up down the strip. a bit of up, and some rudder may just pull the model into the pilots box?
Remember Murphys law-
Prime- anything that can go wrong will.
2nd law- it will go wrong at the worst possible time
3rd law- it will go wong when you least expect it.


Discuss.

[Tony Collins 1073]

Simon , Apology accepted.
As far as I know, all Transmission losses take a minimium of about 1sec to reconnect.
Your intial question was about failsafe and where to position the control surfaces.
Whether it be 35 megs or 2.4 gigs Thel lockout as you call it means going into failsafe. Therefore The same answer applies to both frequencys. There are numerous schools of thought on the subject of how to position a failsafe and there is to my knowledge no definite setting laid down in black and white apart from the throttle having to come back to idle.
I think you will just have to make your mind up using sheer logic for your answer. I have told you my method which some are sure not to agree with but further than that I cannot help you, apart from saying that you should make 100% sure that if your aircraft were to go into failsafe that it would not be for anything that you had done wrong.

Regards Tony.

PS Funs Stuffs back on line.

[Simon Wright]

Thanks for the comments so far, we all agree to comply with CAP658 on the throttle but the operation of failsafe to preset neutral, preset spin it in or hold last command is not cast in stone.

From personal experience a lock out on 35mhz was not a transient thing, the failsafe kicked in and I did not regain any control until the model was buried and ddnt release until I found the guy who had ignored the pegboard and switched on my frequency and made him turn off. a classic shootdown where there was no escape or chance of reconnect. In retrospect the failsafe settings should have been pre set to dump it in a spin.

As I said in the o.p, I have not (yet) suffered (or noticed)a period of flight where I felt the failsafe had kicked in on my Spektrum systems. My 7kg + models use AR9000 or 9100 receivers (& JR Equiv) with extra remotes spread as far apart as possible to hopefully maitain the RF link in all attitudes and prevent shielding by the engine, mufflers and batteries etc.

My understanding is that my set up looks to all 3 or 4 receivers for the packets of digital RF data and processes the best packets applying holds where all the packets fail?

Before I can make any educated decision on the failsafe settings for the control surfaces I am trying to determine if, on Spektrum - or other 2.4 system the failsafe is likely to go into a momentary failsafe with immediate reconnect that I will hardly notice or a sustained failsafe where I can not expect to recover control.

Spektrum sell a flight log thingy http://www.spektrumrc.com/Products/Defa ... ID=SPM9540 which counts holds or fades or something. done a bit of research and think this is what it means to me and my failsafe settings.

Antenna fades – represent the loss of a bit of information on that specific antenna. Typically it’s normal to have as many as 50 – 100 antenna fades on any one of the antennas during a flight. If any single antenna experiences over 500 fades in a single flight, the antenna should be repositioned in the aircraft to optimize the RF link.

Frame Loss – represents simultaneous antenna fades on all attached receivers. If the RF link is performing optimally, frame losses per flight should be less than 20.

Hold – a hold occurs when 45 contiguous (one right after the other) frame losses occur. This takes about one second. If a hold occurs during flight, it’s important to re-evaluate the system, moving the antennas to different locations and/or checking to be sure the transmitter and receivers are working correctly.

After reading that over and over I think the Spektrum system as i have it needs all 3 or 4 receivers to fade and miss 45 sequential frames at the same time. this would put me in failsafe? unlike PCM there is no release time so the instant a full packet of information is received from any of the receivers the failsafe (Hold) is released.

Like I asked earlier, do we have an rf or spektrum or failsafe expert in the house

[Alan King]

I have experienced radio failure on two occasions with 2.4.

in the first instance i test flew a small trainer for a friend, i made the fundamental error of not checking what type of receiver he had on board, this was an IC model.
I had a successful test flight and landed safely, i then refueled and took off again and handed the transmitter to the owner, at the edge of his circuit and about 350m from us he experienced full control loss, the engine cut but the plane continued to roll and dive towards the ground, i grabbed the transmitter from him and attempted recovery to no avail. On inspection of the wreckage i discovered the shop owner who had sold him the model had fitted a park flyer receiver.

the fact is the engine cut but the plane continued with last input which was right aileron and slight down elevator but since this was from quite high the plane was scrap.

on my second failure i was flying an old Super stick of my sons and suffered full radio signal loss, the fail safe was set to cut the engine but this did not happen, i was concerned about a fly away so switched off the transmitter as i felt i had a battery issue, the engine suddenly cut and the plane impacted straight but slightly nose down in a field, closer inspection revealed black wire corrosion on the battery pack, all i can surmise is that the system locked out due to this and actually had a power failure, the situation was possibly caused by overload on the battery due to corrosion and insufficiant amperage, once transmitter was turned off i believe the fail safe kicked in as batt probably released a burst of power.

I have noted many folk still using 4.8 volt with 2.4 and as soon as power drops to 3.2/3.5 volt due to servo load or similar the receiver goes to failsafe and the plane impacts before reset can occur, I have though seen a failsafe kick in on electric model and then reset before the plane impactsand i am sure this has occured a few times on IC planes and is often described by the pilots i know as grey or brown out.

I personally feel that incorrect procedure and fitment often occurs and as in 35mhz where people mixed and matched different crystals people currently mix and match systems which appear ok on the surface but are far from it in reality.

the simple fact is some people seem to forget safety when cost becomes a factor.

[Alan King]

One final point
the only person i know who can answer all of what you want to know is a gentleman called Dave Armitage, he runs Radio control repair centre in pretoria South Africa but is originally from this fair Isle, Dave is well known for articles published in electronic mags a in some rc mags, send him your questions he has a website of the specified name and if he cannot answer them then there is no hope for a reasonable answer other than from the manufacturers.

[Jonathan Pelham]

With slightly more complex research aircraft our safety case and danger area is based on how quickly we can ditch the aircraft after giving a defined amount of time to re-establish the link.
The control logic is such that after the defined time period the engines are cut and the control surfaces are set to institute a nose up position and stall the aircraft and induce a flat spin to bring it down on as short a ballistic trajectory as possible rather than allowing it to glide outside of our danger area.

Donnithorne-Tait, D. (2012), "Current International Thinking on UAS Classification ", 27th Bristol International UAV Systems Conference, 2 and 3 April 2012, Bristol, Curran Associates Inc, .

Has a good figure which correlated aircraft mass and final velocity with the danger it represented to life on the ground.

That is the purpose of the fail safe. Minimizing risk to those on the ground.

[Bob Thompson1894]

Spin? Really? If that jet (which is what you are talking about) is travelling at speed (which they do) then your 'spin' would send it out of control and it could hit anywhere. Chances are with surfaces as Alan suggests it will impact on the field, causing no damage to third parties. The spin scenario is a recipe for disaster, as well as making the odd 'lockout and back in again' a certain crash, and not a brown trouser moment. No, not recommended.

[Bob Thompson1894]

With slightly more complex research aircraft our safety case and danger area is based on how quickly we can ditch the aircraft after giving a defined amount of time to re-establish the link.
The control logic is such that after the defined time period the engines are cut and the control surfaces are set to institute a nose up position and stall the aircraft and induce a flat spin to bring it down on as short a ballistic trajectory as possible rather than allowing it to glide outside of our danger area.

Donnithorne-Tait, D. (2012), "Current International Thinking on UAS Classification ", 27th Bristol International UAV Systems Conference, 2 and 3 April 2012, Bristol, Curran Associates Inc, .

Has a good figure which correlated aircraft mass and final velocity with the danger it represented to life on the ground.

That is the purpose of the fail safe. Minimizing risk to those on the ground.

No, it is to prevent a model going out of control and impacting a full size aircraft.

[Rob Buckley]

The purpose of the failsafe is to ensure the model 'lands' as quickly as possible within the defined flying area, and doesn't fly off and either interfere with full size a/c or crash on somebody's house.

That's why there should be a defined flying area, and models must never be flown over anything or anyone that could be damaged or hurt when the model 'lands'.

[Alan Cantwell 1131]

First time i have seen the fail safe function descrobed properly, Ta Rob , just as a matter of interest, there are LOTS of spectrum recivers out there, in all manner of aircraft, all the spekky units have a fail safe built in, how many are checked at the average club field before flight?(i know nothing of other systems)

in regards to the scenario above, it sounds lime the drone is a level flight unit, therefore, the fail safe function described is ok, but its deffo not ok for model aircraft

[Bob Thompson1894]

First time i have seen the fail safe function descrobed properly, Ta Rob , just as a matter of interest, there are LOTS of spectrum recivers out there, in all manner of aircraft, all the spekky units have a fail safe built in, how many are checked at the average club field before flight?(i know nothing of other systems)

in regards to the scenario above, it sounds lime the drone is a level flight unit, therefore, the fail safe function described is ok, but its deffo not ok for model aircraft

I assumed ALL 2.4 systems had failsafe built in, Alan? Hitec and Futaba certainly do. And on all surfaces.

[Alan Cantwell 1131]

I really dont know Bob, i just use Spekky, with a DX9, i wonder, in the main, if anyone at clubs actually checks they are set, for throttle closed, when you are down at Langar, just, out of interest, do a check on just how many throttles are set to failsafe idle on normal club models, bet you will be suprised at just how many go to full throttle, i was at our club,

[Chris Bradbury]

The purpose of the failsafe is to ensure the model 'lands' as quickly as possible within the defined flying area, and doesn't fly off and either interfere with full size a/c or crash on somebody's house.

That's why there should be a defined flying area, and models must never be flown over anything or anyone that could be damaged or hurt when the model 'lands'.

I agree with this, except for one word. It's not as quickly as possible, but safely as possible. Rob is spot on right though, the failsafe is to prevent the model carrying on from where you're flying and crashing in an area outside your safe zone. Anywhere you fly should be considered "safe", by which it's meant, not over property or people, so if a model spirals in to the ground vertically from anywhere you're flying it shouldn't hit anything. The issue in a fail safe scenario is that if control is lost as you are neutral and level, then how far does it fly. Consider your range on a full tank of fuel or lipo, or a glider from the slope.

Therefore throttle to zero is a minimum, as any plane becomes a glider and any helicopter stops being a blender and has limited range. As for control surfaces, consider the model style rather than a general rule. I always recommend that gliders for example should be put in to a gentle banked turn so there 'arrival' point will be back on the slope you're on, heck it may even come back in to range. A 3D plane can be set to full snap spin, this will lose all speed and create a slow floaty spinning descent in many models.

Essentially the main aim in the event of a lockout isn't to save the model, it's probably already a gonna, it's to minimise the arrival speed, minimise any spinning parts such as propellors and prevent it leaving your safe designated flight area, that's it.

An issue that we discussed at an SIG meeting for multi rotors (quads etc) was fail safe throttle to zero,as this guarantees a 2kg object drops like a stone. In many multi rotors, the stabilisation is separate from the radio and will continue to function without signal, so setting a low throttle instead would create a slow controlled descent, but this is still under negotiation at this time.

[Chris Bradbury]

I assumed ALL 2.4 systems had failsafe built in, Alan? Hitec and Futaba certainly do. And on all surfaces.

I've programmed Spektrum, Futaba, JR, Multiplex, Graupner and more and I've not come across any 2.4ghz systems that don't have failsafe facilities, though there are variations from brand to brand as to how they all work.

For example.

Spektrum defaults to throttle idle and control hold on binding, to program differently you power up a receiver with a bind plug in, then remove the plug before binding the transmitter with the controls held how you want them.

Futaba defaults to all hold, including throttle and is programmed via the transmitter, so you have to select throttle to idle manually every time. Another thing to watch here, as you set it, that you can set it the wrong way and set the failsafe to full throttle if you're not careful.

[Chris Bradbury]

Simons questions:

Duration of lock out part one:
When you do a fail safe test, e.g. Set model up, switch off transmitter and switch back on, you'll note on many transmitters a 2-3 second delay before regaining control, giving the impression of a 2-3 second lockout being what you can expect in the air, but it's not. That's the delay in the transmitter resetting itself. When you first switch on a 2.4ghz transmitter and this applies to all the brands I know, they do an initial boot up and frequency scan, that takes between 2-7 seconds depending on brand and model. In flight you won't have switched your transmitter off (at least we hope not), so the signal is there again the moment the receiver can detect it. With frame rates between 1024-2048, that's nano seconds, so you can't even blink fast enough to see it happen.

Duration of lock out part two:
This comes down to brand, system and situation. If you're on a system that frequency hops such as the new JR DMSS, Spektrum DSM-X or Futaba FASST, then any lock out due to frequency interference shouldn't last more than 1/80 or there abouts of a second, as that's how long it'll take to hop to a clear channel, so you won't notice it. Though that doesn't mean the receiver can't become shielded behind a carbon spar (carbon and 2.4ghz really don't play well together), hence many systems offer multiple aerials, so that if one is shielded the other still works. With non-hopping systems like Spektrum DSM2, treat them like smart 35mhz. The channel can be blocked the same as a 35mhz channel by something on the same frequency, but the radios have safety restrictions to try and stop it. These systems work by first scanning the environment for a free channel before operating, you're then protected in the air by the next one switching on seeing your usage and pickin another channel and so on, once all channels are full (takes 100's switched on at once) then the radios switching on should see the air as busy and set your model to failsafe to prevent you flying. This kind of interference lock out will last as long as if it were 35mhz.

That said, I've only seen a couple of incidents of 2.4ghz interference. I say interference as most of the 2.4ghz lockouts I've seen have been poor use, such as park flier receivers used past 300m of range, receivers buried in balls of wire (makes a great faraday cage) or poor position in the model. I always advise people range checking to rotate the model and check all the angles. How often do you see someone walk 90ft to just one side of a model and check it, but not the other. You've only tested one side!

Here are some classic interference examples:

1-Transmitter switched on inside a car. It only scans inside the car as the car acts as a faraday cage, then once the transmitter leaves the car it causes interference by being on a channel it scanned as clear inside the car.

2-FPV video feeds on 2.4ghz. A 2.4ghz signal is very narrow, but a video version is far wider, potentially crossing multiple channels. FPV gear is a fixed channel it doesn't scan to find a free one, so it can easily swamp a radio. The easy fix is to use 5.8ghz for the FPV or always make sure the FPV feed is switched on before all others, as the radios will see the signal and go somewhere else. NB: the new GoPro3 transmits on 2.4ghz, so you should have that deactivated in the settings before putting in a model.

3-Carbon models. Some gliders are carbon and Spektrum did a carbon friendly receiver which people assumed meant you could use it inside a carbon model safely, what they actually did was put longer aerials on a receiver so they could be guided outside of the carbon fuselage to clean air.

4-Ground objects. Compared to 35mhz, 2.4ghz does not penetrate objects as well, so just flying behind trees can lose your signal.

[Chris Bradbury]

Lastly, apologies, as I appear to have rambled and somewhat swamped the thread, I've just spent a lot of time playing with 2.4ghz systems as you might tell, hopefully nobody's offended by my waffle

[sean smith]

Thanks Chris, very informative.

Sean.

[Alan Cantwell 1131]

No waffle Chris, apart from the set the control surfaces over etc etc, i had a heated debate with Boddo about this, he thought as you, except he advocated setting HARD over, resulitng in a flick arrival, me, and MANY others, advocate control surfaces neutral, and closed to SLOW tickover, the windmilling prop acting as a very good brake, this is old OLD ground, and there will always be 2 camps, gliders are a difficult one, they have to have some sort of control surface movment, to get them down, but power is a very very differant ball game, model coming in, knife edge pass, lots of power on, goes into failsafe, flicks into the flightline, crowd, pitts, pick a scenario, yet model goes into failsafe, power comes back, model holds its knife edge input, model descends, in a straight line, and crashes relatively safe, its the old chestnut, and will be discussed many many times, lets just say, the model needs to descend, and if it aint got no mojo, it aint going anywhere, is it

[Bob Thompson1894]

Small models, as Boddo built mainly, would be fine with the flick into spin scenario, but a big model, especially a jet, would have the kinetic energy to hit anywhere if it went hard over with control surfaces. It could easily travel the 60 metres safety distance. And as your other question, Alan, I said earlier in the thread that we at Langar do failsafe spot checks on all models nowadays.

[Alan Cantwell 1131]

nice to know Bob, hope others do as well, we have the comments of, well, its not a large model, so i dont need to set it, i have even heard what fail safe? a lot dont know they even have one, and when the heated debate was taking place, it was a while ago, but we where in large model company, small stuff was not on the menu, the comments where aimed at show flying, got quite heated, but as always with Boddo, it ended with a smile all round,