Sky View Autopilot

[RICHAWK]

I have configured my two axis AP by using seperate power breakers for each servo. I have done this to allow flight testing of one axis at a time. However, I found today that with the roll axis servo power turned off, the pitch axis servo will not function despite the display indicating it is engaged. Movement of the stick for and aft shows 'slip' but there is no torque applied by the servo. Restoring power to the roll axis restores torque to the pitch axis. Very puzzling !

Richard

 

[dynonsupport]

We'd recommend doing that sort of test with both servos connected to power, with just one active at the same time. Though SkyView network is designed to gracefully cope with failed (power disconnected) components, and though it can deal with components being powered on and off endlessly, we wouldn't consider it a normal operating condition.

We test for exactly the scenario that you're seeing and aren't able to reproduce your problem here. The remaining servo should remain engaged. Does it behave like this when you disconnect the pitch axis turned off as well?

The first thought that comes to mind is that perhaps some power is getting to the servo, but just not enough to fully drive it, so therefore it still acts as if it's engaged.

You might give us a call for some more personal support on this one.

 

[jakej]

Just a thought - are you using the power wire in the network cable or are the servos powered independently ?

Jake J

 

[RICHAWK]

Jake,

Each servo has its own independantly wired and breakered supply.

Dynon,

I'll get back to some more ground testing tomorrow and see if I can come up with some more info.

Richard

 

[dynonsupport]

Also, just to be clear here, you can engage the pitch and roll in SkyView totally independent of one another so you don't need to kill power to a servo to test the other axis.

 

[RICHAWK]

To give you some background on how this has come about. I chose to put a separate switch breaker power feed to each servo so that if either one one misbehaves I have a positive method of isolation that will leave the other servo/axis functioning. This is the result of a bad experience in the past with another make of AP.
Today I can confirm that the sytem is wired up as to drawing , except the two separately switched power feeds. I can also confirm that if either servo power is removed - both servos cease to provide torque.
The worrying aspect is that if roll servo power is removed the screen shows 'roll err' (as it should) but the pitch indicates, with a green arrow that its still in 'alt' mode although the servo no longer functions. In this state the pitch indicates slip/trim if the controls are moved. Exactly the same corresponding condition occurs in the other axis if the power on/off is reversed.

Richard

 

[dynonsupport]

We've actually reproduced your problem. It stems from having the disconnect wires tied together upstream of the switch (as directed by us), which causes the remaining servo's disconnect switch to be affected by the one that's connected to the one who's had his power pulled. We're going to take a deeper look into this behavior and how it might be improved both in the field and on our end. But for now, if you want your scenario to work, replace your disconnect switch with one that's single throw dual pole (still normally open with a momentary ground), where each disconnect wire is individually switched to ground. Basically, you don't want the disconnect wires connected to each other on the servo side of the disconnect switch.

 

[jakej]

Or ... you could use a 1N4001 diode on each line ??

Jake J

 

[RICHAWK]

Problem now is if the diode scheme doesnt work Infinity dont list any double pole PB switches for their stick grip. Is anyone familiar with the part number of the switches so I can search the manufacturer for a DP version ?

Richard

 

[dynonsupport]

Now that we can recreate this, we will look into solutions and have an answer soon. I expect a resistor in each wire will be an easy answer.

This issue is related to the fact that the servos have a fail safe circuit in them that reacts to the disconnect switch a bit earlier than the rest of the circuit. In a sense, the servos are acting very much as designed, where they fail to the off position. This fail safe circuit is seeing the load the failed servo is putting on the disconnect line and shutting off the servo motor drive.

99% of our customers wire both servos to the same circuit breaker which is why it has taken us 2,000+ AP installs to run into this.