Ok, I've been running compassmots on both my quads tonight, and while my fear of spinning up to 100% on the ground is much reduced...the results coming out of compassmot give cause for concern.
If I follow the instructions to the letter, I get 48% on quad 1 (heavy), and 38% on quad 2 (small). However, only quad 2 actually loiters correctly. Quad 1 still toilet bowled.
I've been advised to run compassmot to 100% throttle instead.So I did.
Following the same basic routine (low for a few seconds then held at max for 5) I get Q1: 113% Q2: 123%
Modify this, so after a few seconds on low, you move to mid stick for 5, then 5 seconds at multiple levels until 100%: Q1: 138% Q2: 143%.
Concerning is that if I repeat the last one 3 time each, the end result varied by as much as 30%!
So my questions:
- Should I worry about such big variances?
- what causes the variances? Too long at a certain throttle? Not spooled down fast enough?
- 100%+ result seems like and awful lot. How much is too much?
- What are the optimum ranges?
- what will the APM being doing (compass-wise) when mag is nearly 150%, and what degredation can I expect?
- Both quads loitered beautifully on 2.9.1...why am I having so much hassle with this now? If it's compass interference, why did it not affect loiter or auto on 2.9.1?
Sorry guys...been enjoying my APM until now, but the issues I'm having with 3.0.0 (I'm also having arming issues, for no reason that I can see, and there's no way to verify if a failsafe is in effect) are denting my faith in the platform. I'm told 3.0 is worth sticking with, but at the moment, I need to get RTL etc all working, so I'm seriously thinking about going back to 2.9.1 so I can actually fly again.
Replies
Quad 1: 148% compassmot...loiters ok. It's a bit "fidgety" in the air, even in calm conditions. There is a noticable attitude change when engaging loiter. It takes a minute or so to completely "settle" into to position. At first I thought it was still toilet bowling very slightly, but it went away. Moving it via the sticks also takes a minute to re-lock-in, and it's final position was not where I let go of the sticks. loiter P set to 0.8, I is 0.4, D is 0.0.
Quad 2:132% compassmot...loiters ok. It also takes about 30 seconds or so to completely lock-in. Loiter at 1.0 seemed too aggressive, so loiter P set to 0.8, I is 0.4, D is 0.0.
Additional:
1. At 100%, Q1 was beginning to squirm in protest due to vibe dampers between the legs and the frame (but stayed in the same spot); will that throw off the results? It was fine at 75% throttle - will that be enough?
2. My 3DR power module is/was set to Volts & Amps. However, it does not power the motors - that's done by a 6S pack. Will it affect the results if compassmot is using the wrong Amps data?