If you're using APM 2.5, the best thing is to get the Power Module, which will not only provide clean power but will also let you datalog the battery voltage and current.
Are you saying that it sometime takes 100% throttle to hover? This would indicate that your quad doesn't have enough thrust. Could you please list out your motors, ESC, and battery info?
While I highly recommend the 3DR power module or a dedicated BEC for powering the APM, I really don't think that is your problem.
No, it just works on 2.5
Does it help if you lower your roll/pitch/yaw P and I gain parameter values some?
I agree, seems like the PIDs just need some tuning.
Tony: You should probably post what version of arducopter you are using and the PID values you have set.
Hey Tony, I am running a pretty simalar setup with the same weight. Turnigy Motors esc's etc. I am also using 9 inch turnigy slow fly propellers as the 8's would not give me enough lift. I have a set of 10's but I like the motor esc temps with the 9's better.
A 2200 mAh pack will give me 5 minutes before my low voltage alarm starts squaking. a 4000 goes almost 10. that is regular flying. Hovering around I get about 15 minutes on a 4000mAh pack or 7ish on the 2200
Your pids look like mine too. I am running a .090 on the angular rate control.
If I did not have the LVA alarm I would loosing altitude as the soft cut outs on the ESC's kicked in.
Oh and moving threadto flying tuning Multicopters for greater exposure.
That's interesting that you found the 8's did not have enough lift, I guess I need to get some different props and see how that changes things.
I have a LVA on mine too but the variable power problem seems to manifest way before the alarms start going.
I think I might pick up a 4000mah battery too and a good charger, what brand batteries do you run with? From what I've read Turnigy Nanotech batteries are not as frowed upon as much as their other components, but perhaps having something other than HK parts would help diagnosing?
I've also just added a gimbal to my copter and the beginning of this short video show the problem I am experiencing with inconsistent thrust. Don't know if it helps with understanding what component might be failing?