Issue with flap servos and DX6i

[97tj-neil]

I built a FT Bushwacker over the weekend and went with the aileron and flap option. All appeared well when I tested my servos on the bench with a servo tester, so I went ahead and glued up the wing. After it was all built and I was adjusting throws, I started having glitches with the flap servos. I had to cut some access holes in the bottom of the wing to remove the servo leads to troubleshoot the issue. Here is the diagnosis.

1. The problem is intermittent (lead me to believe it was a bad servo wire, bad Y-cable, or bad connector). I tried multiple Y-cables with no change. Moving or wiggling the wires had no impact. All connections appear to be fully seated and no damaged terminals after visual inspection.
2. Sometimes the flap servos will move on their own, with no input from the radio.
3. Sometimes, they will not work at all. When this happens, the other servos connected to a servo tester stop working as well.
3.Eliminating the Y-cable fixes the problem, even though the same Y-cable works just fine on the ailerons. This requires using a separate channel for each flap servo.
4. With each servo on a different channel, if I use a servo extension cable, the same problems return.


So now, I have each flap servo plugged into a separate channel on my receiver (gear and flaps). I mixed the channels on my radio so that the gear switch drops both flaps. This works, and I learned a few things about programming my radio in the process. However, not having any extension cable makes it a challenge to get everything connected and the wing strapped down.

Do any of the gurus on here have some advice for me?

Just for fun, here is a pic of my plane.

 

[Craftydan]

Which ESC are you using? I suspect 6 servos on that UBEC might be getting a bit too much . . . and you've been lucky the brownout hit the servo first and not the RX.

Heating up the ESC in flight will make this worse.

You'll probably need to switch to an external BEC and drive a few of he servos off of it. not intending this to be a shameless plug, but you can use the adaptor I laid out in this article:

http://flitetest.com/articles/simplified-twin-control-harness-for-the-kraken

In your case, you'd replace the slave ESC with your external BEC (yes, you can wire it's power in with the ESC's battery harness -- no need for a separate battery in this case). then you could either:

- run the two flaps with a Y off the slave servo plug,
- repeat the pattern once more with one of the end-pin-cut headers (3 rows of 3) and move the outbound red wire from the first row (which the BEC shouldn't have any wire plugging into it) to the new row.

That should take the load of the ESC's UBEC and get rid of any low voltage servo issues.

 

[97tj-neil]

Dan,
Thanks for your reply.

I am using a Suppo 30A ESC.

I hadn't considered this because I had the same issue if only the flaps and ailerons were plugged into the RX on the bench with zero draw from the motor.

 

[Craftydan]

Motor draw won't effect the power available to the servos/RX, assuming the ESC isn't overheating, however 4 servos is a pretty typical limit on 3S. On 4S, they'll usually support only 3.

It might not be your issue, particularly if disconnecting all but the flaps causes the stutter, but the common denominator is increasing the servo line length.

That leads me to think it's a voltage issue and the direct connect is just the working side of the line. Adding just a touch of resistance pushes it over into "unpredictable behavior", region, at which point it jitters like mad. If that's a low voltage issue from an overloaded UBEC, removing the ailerons as well will correct the problem.

It is ENTIRELY possible this is a signal strength issue and your servo extension and Y is a bit more resistive than they let on . . . odds of two failing is slim, but it's possible. In one case you're splitting signal, in another you're running the signal down a longer wire . . .

Have you tried swapping any of the servo connections between the ailerons/flaps to see if the ailerons would start to jitter? Depending on whether the problem moved or not, that would eliminate either the RX/ESC or the extensions/servo from the list of suspects.

 

[Cdbarron727]

Awesome

That reply was awesome. You analyzed his issue to the "t". I'm curious to see if the issue was solved. Regardless of that, you've taught me something. Thanks CraftyDan