Confused about failsafes

[dvogonen]

As far as I can understand the original second version of the Flysky 2.4 gHz system is pretty decent. Independent tests on the internet indicate that they have implemented a good channel hopping system. Besides that, I have not noticed any glitches, and I do not want to fly further away than I have walked for range tests. Besides telemitry, the main reason to upgrade to e.g. FrSky is that there is suposedly no failsafe, which is a big deal. Especially with fast and heavy stuff.

I received a couple of HobbyKing branded 3 and 6 channel receivers last week which I installed. The 3-channel receivers are intended for ground use (cars and boats), but they seem to work well enough in my parkflyers. You really, really need failsafe for nitro cars and these 3 channel receivers do have failsafe with an assigned button for locking default values. Tried it and it works. Great!

Much to my surprise the 6-channel receivers also seems to have a failsafe. When I turned off the radio with the motor running, it kept going for a couple of seconds and was then softly turned off. The other channels (besides throttel on ch3) seemed to remain in their last known position. I have found no way of controlling the behaviour of the failsafe, but I guess that the default behaviour is what you would normally want.

Now I'm a bit confused. I have read numerous times that the FlySky air receivers do not have a failsafe function. But there seems to be one anyhow. How come?
Might it be the controller card that somehow understands that the receiver has lost contact with the TX and decides to shut down the motor?
Or has Flysky changed the firmware in the receivers?
Or has there always been a failsafe in the 6-ch RX?

Come to think of it, I think I will connect a servo to channel 3 and see if it behaves like the motor or if it remains fixed in position.

[gohsthb]

The original V1 receivers had no failsafe at all. They would just keep whatever value they had last. The newer V2 will at least shut off the throttle, and I think that is locked to one channel. I didn't know the 3-channel ones where programmable. Maybe they have been improving things without telling anyone, and so your newer receivers have an update that allows the failsafe to work. I think what people like about Frsky is that all channels are failsafe. And you can program what they all go to.
-Gohst

[ReSt]

This normally is the behaviour of the ESC to shut down when the signal gets lost.
As far as I remember, the behaviour of the six channels is different, some do hold the last value, others not. So it may be important to put the ESC on a specific channel.

Reinhard

[gohsthb]

Correct an ESC should shut down if the signal is lost. I remember looking at the receivers with a logic analyzer to see what happens when the receiver loses the signal. And the V1s just kept outputting the same signal. So the ESC would never know the receiver lost signal.
-Gohst

[ShowMaster]

A real good test for failsafe is to power the rx with a BEC or esc, no motor, in a aux channel and plug in a servo in each channel you will be using. Move the servos to some position other than dead center and turn the tx off and see what the servos do.
If there is any form of real failsafe the servos will move to some position from where you left them. In the case of throttle failsafe the servo in that channel should go to one end when the TX is turned off. If it dose not then you don't have failsafe, or it's not setup.
As for the esc shutting he motor down when the tx is turned off, all modern good esc's, will kill the motor when the input pulses are removed. This will happen when the tx is turned off or the rx looses contact with the TX. This is not a rx failsafe so don't be confused.
As for the servos holding their last position making you think you have failsafe. All modern RX's usually just stop all pulses from going to the servos with loss of signal and only the gear train friction keros them there, again not a true failsafe but much better than in the past where they may all go to one end with loss of signal to the RX. Not a true failsafe because in a gas plane the engine state at what ever setting it was at but much better than a lawn dart straight in action we used to suffer.
A true rx failsafe will keep pulses going to the rx from internal memory and hold at full strength the last servo position or set failsafe position until signal is restored. I understand that some 9x RX's may output the last pulse but this is not failsafe if your using a gas engine.
So the real failsafe test is with servos in all you used channels turn the tx off and see if you can easily move the servo arms or you feel them being held there and are hard to move without breaking a gear.
If set Frsky has true failsafe and the servos are being driven and held with full voltage to their FS position. I think you'll find this is not he case in the stock 9x rx unless you use the original PCM rx that's not 2.4.
As stated, and throttle failsafe with the original rx I think you'll find it the safety feature built into your esc and you rx is simply not doing anything useful with loss of signal except stopping sending anything to your servos.
I'm told a lot day at our field about the stock Turnigy 9x failsafe being on he menu and how it kills the motor. WRONG!
There are many forms of signal loss options but a true failsafe will allow you if desired to make the throttle channel go to a position of your choice with signal loss no matter what the other channels do.
The lesser priced Spektrum RX's only allow throttle set failsafe, the higher end RX's do all channels as I remember.
SM

[Kilrah]

If there is any form of real failsafe the servos will move to some position from where you left them

That's not really what product marketing advertises, which is what often causes confusion.

There are 3 basic types of receiver behavior when signal is lost:

1. No Failsafe. Receiver stops sending signals to servos. Analog servos go dead, like if there was no power, so you can turn them by hand.
2. Hold-type failsafe (yes, it is still called a failsafe by most manufacturers out there). Receiver keeps sending the last good received commands, so the servo still "fights" to stay where it is. Even if this is full deflection because you lost signal when you had the stick in the corner.
3. Pre-programmed failsafe: On signal loss, receiver first holds the last good commands for some time (usually around 1 second), then moves all servos to positions you've programmed.

A basic Mhz PPM receiver would do 1) only.
More advanced Mhz PPM receivers, as well as all PCM ones would let you choose between 2) or 3).
I haven't yet encountered a 2.4GHz system that would do 1). Would be interesting if the Flysky / HK stuff does... I don't have a working set to test.
Most will do at least 2), sometimes with variations like offering 3) but only on the throttle channel, so you can program your throttle to be killed but the other channels will hold. Low-end Spektrum, Futaba in 7CH mode are examples.
FrSky lets you choose either 2) or 3), for all channels. Same for MULTI (14CH mode) Futaba.

As manufacturers say 2) is a failsafe, and some people like SM say it's not, you get "Confused about failsafe" users like dvogonen

Note that as 1) is really bad (even more so on MHz bands as the receiver would sometimes pick up random noise and send just a single erratic pulse at the servos) and you really want to avoid it, some digital servos implement 2) themselves. So if the receiver stops sending pulses, or just sends a random seemingly invalid one sometimes, the servo will still hold the last good position.
So if you want to test your TX/RX system to see how it behaves, take a plain old analog servo to do it

And then there are endless debates about what to set pre-programmed failsafes to, so maybe it's better not to go there too much
The only real solution - AVOID signal losses Fortunately it's a lot easier these days. You really need to have bad luck, an awfully noisy installation or push it like crazy to get severe drops on 2.4GHz.

[dvogonen]

I tested Showmaster's assumptions by switching the throttle and elevon channels. Thrustmaster is 100% correct. There is no proper failsafe. Regardless which channel the elevon was connected to, it stayed at it's last position at channel loss. The throttle was also choked regardless of which channel it was connected to.

So there are good and bad conclusions. The bad being that there is no RX failsafe and the good; that even my lowly and not modern at all GWS BEC is able to understand when the Flysky RX has lost connection with the transmitter and cut the throttle.
This actually works for me. I will not start flying nitro planes again any time soon (or ever). I will keep my Flysky modules and receivers for the time being.

As long as you live, you learn

[ReSt]

I just verified what happens wit TX power off:

3 channel EURGLE receiver:
Channel 2 can be set to predefined failsafe position, ch1 and 3 go dead

8 channel TURNIGY receiver:
Channel 4 and 5 hold position, all others go dead

So it is important which channels to use


Reinhard

[msev]

ReSt can you measure also the 6 channel receiver?

[ReSt]

This is what I verified when the transmitter gets switched off:
2 HK receivers: Ch4 and Ch5 hold their position
1 Turborix receiver: no channel holds it's position

But there maybe receivers available with a different behaviour. As it was reported that the older ones had a ppm sum signal on ch6 (?) available, while newer ones would not have it.


Reinhard