Um... okay, but how do I find what the cg actually is? Like if I change my battery location, what's my new cg?

Shreeyak simple thing is whatever load you put on your copter should be distributed equally to all motors. As ashimda mentioned you should consider center of thrust more over center of gravity. You can try any method you like but concept is simple.

Any method to verify that my centre of thrust is where it's supposed to be?

simply

look the pics to find the cg

Oh! That's a simple technique!!! Thanks for sharing, very helpful! Thank you!

On my quad, I adjust my CG according to take off. When the CG is to far forward, on take off the quad will tilt forward, then adjust. When the quad lifts off perfectly level I know the CG is good.

you must calibrate your sensors to exact level and make sure your roll pitch sticks centered to work your method.

Calibrating sensors is the first job when setting up an FC.

actually if you really want to be accurate .. the ideal center would be at the point where all motors (assuming identical motor / prop combo )
are drawing equal current at hover. the only way to check this would be to have individual current telemetry for each motor.

as I said in a different post ... this hardly matters . the FC will compensate  because multis are inherently
unstable and inefficient.

technically it is possible to just hook up a recieiver and esc's and control a multi directly from your tx.
this would result in a dynamically unstable platform and you would need the reflexes of a jedi knight
to fly it.

Ys you are correct but we can simply do without the help of electronics is by finding center of thrust and hang it by holding @ that point using a rope, then arrange everything until it become level.

this is not like weighing potatoes .... once the aircraft starts flying how do you know
that all your motors are going to produce the exact same thrust ?? if you do not know then
what is the point of going through all this ?

what you are talking is static balance ... what is required is dynamic stability... again a moot point
with multirotors.

also you are only balancing on the pitch and roll axis ... what about yaw axis balance ?

not for arguing,
 if the thrust is different even if the motors and props are identical (as per you said in the previous post) then the efficiency also different and current drawing too then how can you calculate very accurately by checking current draw of each motors?

Exactly my point. This is why I find my method(#55) to work well. The FC will always compensate to keep the Quad level. That's why if CG is off, at take off it will tilt before the FC corrects. I keep adjusting CG until it lifts off level.  

Missing the point again - at take off if it tilts , it is not necessarily a CG issue
 it could be unequal thrust .  you have no way of knowing unless you monitor current draw.

@akhilzid - good point .  That is why it is no point balancing the multirotor when it is static. There
are far too many variables.

Cg is simple to balance . In a square symmetrical layout the intersection point of two diagonals is the cg. Mostly the centre only except for assymetrical motor layout. Tbs is different due to its motor layout.
Also cg for z axis - fc and prop on same height. With load on either side, top heavy if moving around, bottom heavy for stationary stability. multi rotors can not be judged of current consumption because there are many factors the fc is compensating all the time.

CG of Tbs is already shown in post #49 by ashimda, just arrange CG of everything to that point will statically (in X and Y dimensions) balance the quad , and the FC do the rest.

I arranged cg of tbs discovery more than 1 year back when I built mine.
I don't have issues balancing quads. I was saying z axis can be balanced too.
What the fc will do is dependant from fc to fc.

Edit- if I remember right, the tbs recommended cg for disco is little off.

That being said, if altering the CG to balance out unequal thrust, which in turn results in level take off and smooth flight, isn't that the goal ? 

Some interesting facts i found from past experience are.
- static balancing not necessary if you are flying LOS. YES FC takes care of slightly off CG BUT if CG way too off you might experience drift in particular direction and need constant sticks input, you can adjust trim BUT you must know FC giving more thrust constantly to that motor(s) to compensate gain. If you have onboard camera you will clearly see that. best example i have in this video.

Check this video, same quad same setup, props balanced with numbers on them, motors balanced as well.
1. from 35 sec to 1:57 sec segment taken in the morning. CG was slight (i mean 1-2mm) off and you can see some jello and vibes.
2. from 2:25 sec to 2:42 segment taken in the evening. CG was balanced (best i could do) and see the results. the video was completely jello and vibes free. i have raw footage of this video (1080p) observed hundreds of time to reach to this conclusion.



3. check this video (on board footage), no jello or vibes however taken at the night but still you can judge from it. CG balanced statically. no trims (all centered) .

Excellent illustration of the point being made.

Nice demonstration , CG balanced statically is do good job.
and the small variation in thrust is take care of FC,

if CG balanced statically and the thrust difference is too much then its difficult anyway, even in ashimda,s method
bacause the thrust rate is always differ along with throttle changes, if we correctly balance that @ throttle 1400
will not be balance @ 1550

 

In this fpv video i did not even took care to balance my final setup, however props were balanced but i just placed battery visually in the middle of the frame. But when i hooked 5.8ghz tx on the frame, CG moved off and see the results! It was like huge jello and lots of vibes.

nice one Dharmik   - I have to say the the point is well made.

Good Job Dharmik