Futaba SBUS & Brain Fail-safe

[smeggy]

Hi guys.

I can usually figure this stuff out, but this one has stumped me.

I'm trying to set up the fail-safe for my new 700E.

Details: T8FG Super, R6208SB, connected to MSH Brain via S-BUS, driving BL700H/BL750H servos.
Rx, Brain and servos at 7.4V.

I set up the F/S and POS for the channels in the Tx failsafe menu as normal, wait a while and turn of the Tx. According to the Brain diagnostic menu, the inputs from the Rx don't change - and of course the heli servos/throttle don't go to the desired positions.

However, when I plug other servos directly into the servo ports of the Rx and turn off the Tx, those servos DO go to the fail-safe positions.

It would seem the SBUS or Brain is not acting on the fail-safe - which ironically isn't very safe.

Could anyone shed any light?

[kingmeow]

Trying posting the same on the MSH Brain forum. You'll probably get an answer there.

[ahahn]

I don't really understand why it isn't working, but just one comment.

It is the receiver that is doing your failsafe. The Brain has no clue you turned the transmitter off since the receiver is still sending out valid servo signals--I think it always has some failsafe values, even if not what you want.

So my feeling is that something in the receiver isn't setup/working correctly.

[smeggy]

Thanks guys. I also think this issue is related to the rx. I might buy a cheap sbus servo and confirm. It seems i have to rewire the choppie to use the conventional outputs instead of sbus

Edit: i found the sbus protocol details, and i am handy with a scope, so i will directly check what the rx is pumping out. Interestingly, there is a failsafe bit......

[ahahn]

Originally Posted by smeggy

Thanks guys. I also think this issue is related to the rx. I might buy a cheap sbus servo and confirm. It seems i have to rewire the choppie to use the conventional outputs instead of sbus

Edit: i found the sbus protocol details, and i am handy with a scope, so i will directly check what the rx is pumping out. Interestingly, there is a failsafe bit......

I don't have an IKON, so it isn't clear to me if you can set FS values in it.

If SBus does send out a failsafe bit in its protocol, then a receiving module could in principle decode it and pull up its own failsafe values, even though the sbus receiver is actually sending out valid receiver values (its own failsafe numbers).

I don't see any "official" comments on this either in the Futaba Sbus or the IKON/Brain literature.

Makes it tough to figure out what is really going on, that's for sure.

[Dmaxison]

Brain/iKon has it's own failsafe. What's happening is the delay timing programmed in the Brain is a shorter interval than what is set in the Futaba rx. For the sake of conversation here's what's happening. When the failsafe condition is triggered either by loss of signal or hardware failure the Brain transitions to it's internal failsafe after 1000ms.
Now the issue is Futaba waits for about 2000ms. Because of this delay, programmed failsafes are null. I've asked MSH to implement some control in future software already. All we need is a way to adjust the timing. Seems like a fairly simple request.

[Dmaxison]

Just to clarify, the Brain/iKon does not currently have any user settings for failsafe.

[ahahn]

Originally Posted by Dmaxison

Brain/iKon has it's own failsafe. What's happening is the delay timing programmed in the Brain is a shorter interval than what is set in the Futaba rx. For the sake of conversation here's what's happening. When the failsafe condition is triggered either by loss of signal or hardware failure the Brain transitions to it's internal failsafe after 1000ms.
Now the issue is Futaba waits for about 2000ms. Because of this delay, programmed failsafes are null. I've asked MSH to implement some control in future software already. All we need is a way to adjust the timing. Seems like a fairly simple request.

What I was curious about is how the loss of signal gets noticed by the Brain. If I understand what you are saying, the Futaba receiver won't give any servo updates (of even the previous value) for ~2s?, and that the Brain/IKON notices this lack of servo signal loss after 1s?

That's new info for me since the manufacturer's don't really say too much about it. I'd always assumed the receiver would continuously update the servos, even with the failsafe values, basically "instantaneously".

Learn something new everyday!

[Dmaxison]

I'm guessing there is a link confirmation between the tx, rx, and flight controller. This is what would trigger the delay counter in the Brain.

So upon loss of link... the Brain starts an internal timer before going into failsafe. There needs to be a delay that way there is no interaction during slight brownouts or signal loss. So like I said the delay here is approximately 1000ms.

Now Futaba acts the same way. However it waits a little bit longer (say 2000ms) before it sends out the failsafe commands. Since no link exists between the tx and rx during failsafe, all commands are ignored by the Brain until the link is reestablished.

So with correct timing failsafe commands could be sent to the Brain where it would hold those last known values till the link was reestablished. I would love to see this available that way Self Level could be used during a fail safe condition.

[davehour]

This is something that needs the attention from the Ikon guys.

Because I don't use the Ikon governor, I connect the throttle signal directly to the Rx. This one of the beauties of the S-bus protocol. At least, I'm sure I'm using my preferred fail safe signal in the throttle channel.

[smeggy]

Now I know exactly what S-BUS does

Checking with a scope, I can see the desired "failsafe" value (set up from the Tx) is indeed taking effect on the appropriate channels, at just over 1 second after the Tx is switched off.

Good!

I can also see the "failsafe" bit ('flags' byte) is going from 1 to 0 (inverted) at exactly the same time (just after 1 second after the Tx is switched off)

Reasonable!

However,

I can also see there is a "Frame Lost" bit ('flags' byte), which is going from 1 to 0 almost immediately after the Tx is switched off (in sympathy with the red LED on the Rx). The delay between the "Frame Lost" and the "failsafe" bits changing is seemingly 1.0 seconds.


So I would guess the Brain is seeing the "Frame Lost" bit and then simply 'hold's the previously live channel values, hence ignoring the subsequent failsafe setup of the radio.
I think that process is flawed. I will join the Brain forum (wherever that is) and see what they think.