FT Kraken Won't Taxi Straight

[ethicalhacker]

I scratch built an FT Kraken, but I cannot get it to taxi right. It turns so sharply on the ground that there’s no way it could sustain flight for more than three seconds before crashing. Yes, I know that from experience. I’ve rebuilt this thing four times now without a single flight lasting longer.

I’m running dual SunnySky X2820 920kv motors with counter rotating 11x6 APC props on an 8000mA 3S battery. I’m also using dual HobbyKing 40A YEP ESCs. I’ve setup a differential thrust configuration through an EzUHF 8 channel Lite receiver and a Taranis Plus transmitter as described in the build videos, and everything seems to work in my living room floor.

I had to make modifications to the stock landing gear design to support the weight of these motors. The original design flexed side-to-side 45 degrees when taxing (which induced undesirable ground steer). I went with a heavier gauge steel rod to stiffen things up. Likewise, the tail fins required modification. The ground chewed them up way too quickly, so I put some curved, ridged steel wire on the underside of the tail fins to slow the process down. It travels over pavement better now.

My current problem is that even after setting the endpoints on both ESCs and monkeying with the rudder trim, I still cannot get the plane to taxi anywhere close to straight on asphalt. It will taxi in a 10’ diameter circle even with the trim set all the way in either direction.

I thought that maybe one of my motors or speed controllers was out of speck, so I ordered two more of each. Then I measured the thrust of each changing one variable at a time (motors, ESCs, and props). In the most extreme case, there was a 5% difference. I would think trim could overcome something as small as that. I did notice that the ESCs were different from each other at fixed points along the throttle curve, but again, it didn’t seem like any of them was off by a lot. Putting a new prop on one side while leaving a beat up prop that’s missing a 1/16” from both tips on the other side would make more difference than swapping a motor or an ESC.

I also modified my trim via the Taranis such that it would not raise one motor’s RPM while reducing another. It will now only reduce the RPM of one motor at a time. I don’t expect this to work since trimming didn’t work before and the ESCs are not perfectly matched along the throttle curve.

Does anyone have any ideas? I’m willing to post my Taranis config. I can also post a diagram of my prop directions and motor rotation directions if that would help.

 

[makattack]

Some times, it's the obvious stuff that gets us. You mentioned counter-rotating props, but do you actually have the correct props mounted on the motors? CCW and CW with the numbers facing forward / concave side towards the rear?

 

[Crazy Goats]

Put a bigger motor on the side it turns to.

 

[ethicalhacker]

Some times, it's the obvious stuff that gets us. You mentioned counter-rotating props, but do you actually have the correct props mounted on the motors? CCW and CW with the numbers facing forward / concave side towards the rear?

The FR Kraken is configured as follows. Note that in the image, "From Pilot's Perspective" means from the perspective of a miniature person sitting in the cockpit. Motor rotation is demonstrated by red arrows.



The following image shows how the props are installed. The sides of the props showing would be pointed in the direction the plane was going (towards air; not the plane). Again, the rotation direction is called out. Please note that the pusher prop's text does not face in the same direction as the normal prop. The leading edge of the prop should be clear in the photo.



Does this eliminate the obvious problems from the list of possibilities?

 

[ethicalhacker]

I should add that the plane turns hard left (from the pilot's perspective) despite the left motor's RPMs being higher than the right.

The motors are individually putting out about 1.25kg of thrust on the top end (as measured by my makeshift contraption that points a motor downward [air blows upward pushing the support structure down onto a zeroed scale accurate to one gram]). That puts total thrust at roughly 5.5lb (mixed units, I know; it just might be more useful to some people)

With rough props, thrust per motor drops to about 1.15kg.

 

[Bricks]

Something is not right with the prop on the left it looks like it is a left handed prop but yet the center looks like it should have the insert in it to fit the motor shaft. Is it a true left handed prop not one just turned around?

 

[Craftydan]

EH,

I agree, something isn't quite the same with that second prop.

yes, I know, it's a pusher not a tractor prop, but overall, due to the inconsistency between the props, I don't think they're performing the same at WOT, as air flows through the props. My recommendation is stop worrying about counter-torque. Particularly with ROG takeoff, you won't leave the ground until there's lift under the wings, so torque-roll won't be the murderous beast you think it is. Pick a prop, and put two of the same on the airframe, spinning the same direction.

If the full-scale Ford Tri-motor could fly well with three props spinning the same direction, your Kraken can tolerate two.

 

[ethicalhacker]

Is it a true left handed prop not one just turned around?

The one that appears backwards is a pusher prop. The other one is a normal prop. In order to prove this point, I've attached the following photo. In this photo, I have flipped the pusher prop over so that you can see the writing on it (and the normal prop as before). You will notice that they are indeed different props.

The top blades of both props in this photo have their leading edges toward the left. The pusher prop's top blade has its leading edge low on countertop. The normal prop's top blade has its leading edge high off the countertop.

 

[ethicalhacker]

Pick a prop, and put two of the same on the airframe, spinning the same direction.

At this point, I'm willing to try just about anything. I have several of these props, and this would be easy enough to try. Thanks for the suggestion.

 

[CrashRecovery]

You have two options as I see it. Same rotating props or find props for multirotors in the size you need.

 

[makattack]

Hard to tell in a 2D image, but it looks to my eyes like those props just have different shapes, and the pusher prop is a weird design, but from the looks of it, with the numbers facing the camera as you have it, they are now properly oriented for counter-rotating operation... With that perspective, the pusher spins CW, and the prop on the right spins CCW. Maybe it's a visual trick with the shine/light playing tricks on what I see. Yeah, I recommend as everyone else says, to keep it simple and use the same identical prop for both motors, and have them spin the same direction.