Developer Comment by John Arne Birkeland on June 22, 2012 at 12:25am

A agree with all your points Matthew. My point is that in the context of this thread, I don't see how a second autopilot would have realistically helped in any of those situations.

Comment by Lorentz on June 22, 2012 at 2:14am

@Matthew

"The greatest risks are electrical connector faults, radio failure and mostly my own flying capabilities"
I totally agree

At the risk of going off topic:

- although unpleasant, accidents and near misses are the best source of knowledge for increasing safety and reliability
- proper design and dimensioning of linkages, spars, required servo forces, strength is bare minimum except for foamies (perhaps)
- reliability is paramount, reliable components are believed to be so if inspected, maintained and replaced before failure (think MTBF)
- AFAIK connectors are one of the biggest source of failures

Possible countermeasures :

#1 inspection ? manitenance ?
#2 proper dimensioning ?
#3 proper dimensioning ?
#4 diverse radio link ? redundancy wouldn't have helped: same radios, same SW error
    diverse backup system (autopilot) ? provided autopilot SW is reliable enough
#5 training ?

Still, I'd put the effort to the reduce the riskiest hazards, considering risk as probability of hazard X severity of damage.

Comment by Veikko Vierola on June 22, 2012 at 3:55am

I was just thinking that total dual system would have a backup for almost every problem. If the plane is not staying on its flight path the 2nd network could still be able to do that. For example if APM(A) network has a lost elevator connector output, the APM(B) network still connects to the servos normally. Of course in this case you'll have a bit more wires going to the servos. But maybe I could reduce the amount of wires by using direct cppm connection to APM from RX?

Comment by Lorentz on June 22, 2012 at 5:05am

@Veikko

 

I was just thinking that total dual system would have a backup for almost every problem.

That's true, except for common cause / common mode failures and systematic failures.

If the plane is not staying on its flight path the 2nd network could still be able to do that.

Yes, we only need to tell the failed / wrong network from the good one and a mean to switch the controls from the bad network to the good one.

For example if APM(A) network has a lost elevator connector output, the APM(B) network still connects to the servos normally.

Good point, even better with redundant servos that default to central position when PWM signal is missing.

Still not sure if the risk of loose connectors is worth a second APM instead of replacing connectors with soldering.

Of course in this case you'll have a bit more wires going to the servos.

But maybe I could reduce the amount of wires by using direct cppm connection to APM from RX ?

I saw somewhere here on diydrones a post regarding servos with daisy-chain serial connection that would reduce wires and, more important, connectors. Also direct PPM connection from RX to APM may help reducing wires / connectors.

Comment by Veikko Vierola on June 22, 2012 at 6:08am

@Lorentz

Do you mean Futaba S-Bus system?

Comment by Lorentz on June 22, 2012 at 8:11am

yes, that one !

Comment by Veikko Vierola on June 22, 2012 at 12:21pm

Isn't the Futaba S-Bus system just a normal bus that feeds cppm data and the servos are capable of decoding the cppm individually?

Comment by Zachary Eldridge on June 22, 2012 at 12:32pm

Think about this servo on the right gives out in flight your system is going to be flaw. I would be put dual servos with that setup. Over two APMS are still over killed. Two servos no so much since they are running on the same signal and one servo fails the other will handle the slack. APM is pretty smart device if that goes down it goes to RTL and failsafe mode automatically.

Comment by james sowell on June 22, 2012 at 6:32pm

http://www.diydrones.com/profiles/blogs/apm2-and-rx-all-in-one

Comment by Lorentz on June 23, 2012 at 12:51am

@ Veikko

I admit that I don't know the details of S-Bus protocol, but it could be as simple as forwarding the CPPM signal to all servos and having them decoding.

Being bare CPPM signal sensitive to disturbance, I suspect they implemented a serial protocol with error detecting mechanisms.

Another nice solution could be I2C protocol as implemented in OpenServo

Comment

• ‹ Previous
• 1
• 2
• 3
• 4
• Next ›
• Page