I do not agree that you need to change motors on a F550 setup that carries a Sony NEX-5N. What you do need to do, however, is replace the stock 10" props with 10x5 Graupners. My F550 is a beast at 8lb 4.9oz/3766g, with our new TP550-NX 2-axis gimbal, the integrated landing gear, a NEX-5N/16mm lens combo and two 4s-3300 packs. I can fly for 10 minutes with this setup, and the motors barely get above ambient, which today was in the low 80s. With the stock props, the motors run hot as hell, with the weight at only 6lbs. This setup has plenty of power too, and will hover right around 50% in Manual mode.
We've been testing this gimbal for the last couple months, and it is very close to being released. We tested virtually every high-end servo under the sun, and have tested various belt drive ratios from 2:1 up to 2.8:1. What we are settling into as the most optimum configuration is a 2.5:1 belt drive ratio, in each axis, and MKS HBL950 brushless digital servos. With this setup, we can achieve a similar "Zenmuse-like" platform independence type of performance, with the right gimbal controller. Forget using the WK-M, or Naza gimbal outputs, as they have issues (90 degree limitation, slow 50Hz servo PWM rate and a very noisy WK-M GPS integration that causes twitches in the gimbal outputs around the neutral point...). Anyway, I'm currently using a v1 XA FC/AHRS combo as a standalone gimbal controller because it lets you set the PWM rate to 500Hz/2ms, open up the throws to a full 180 and because it doesn't have any noise in the outputs. My only issue right now is that the XA configuration software only lets you set the servo gains up to 255 and with our current 2.5:1 ratio, in order to get the gains dialed in perfectly, it would need to be 257 or 258. Being off this very small amount is enough to cause some small movements in the video below.
This first one is the raw video, right out of the NEX-5N:
As I said, with the gains being slightly off, there's a little movement in the video, while the WK-M struggles to hold position in gusty winds, but not much. I get pretty anal about these things, so it bugs me. :highly_amused: I then did a slightly stabilized version, with the settings at their minimums in the FCP stab function. You can see a small lens smudge in the left-center part of the picture. In the first video, it stays put. In the version below, you can see the smudge move around slightly, which is due to the corrections need by the gains being off a little. Anyway, this shows that the stab function is not having to work very hard.
I think we are done now, with the tuning and testing cycle on the gimbal itself. We will still need to do is finish testing our standalone gimbal controller, which is basically one of time cop's nifty little Naze32 board, with some special "tweaks". I actually have it working now as good, or better, than the XA FC/AHRS, but only on the ground.

We have a weird twitch that occurs during flight, while hovering. We need to add a moving average filter to the raw accelerometer data, and we think that will take care of the twitches. At that point, we will be done with everything, and can concentrate on production, etc.
-- Gary