Mavic Pro Motor Mod

[beanbubba]

I recharged the new battery with the charger cybernate recommended that will over pack to 4.4v.

I re-ran the identical hover test with same results except it was just shy of 40 minutes.

BTW - once voltage hit 3.0 the bird was coming down and not in a pretty fashion after about 15 seconds of up stick. Chazz might call it a hard landing

EDIT: the red alert from the factory setting to Non-DJI/4s is spot on (at least for this mod setup). Once the red alert comes up on the battery, you are running on fumes and need to find a landing spot ASAP. You have about 1 minute of controllable flight left at that point.




Notes taken during the flight:

 

[beanbubba]

I had several brand new MS4000/3s/LiHv and tore them down and made some 4S's. I think I have the right combination of batteries to test this mod incrementally to the failure limit.

 

[lolo780]

For reference a Thunderpower 4500HV 3s1p weighs 225g without heatshrink or labels so a 4s1p should be a bit over 300g. That's a significant weight savings and makes 18Ah 4s4p possible for under 1250g.

 

[beanbubba]

For reference a Thunderpower 4500HV 3s1p weighs 225g without heatshrink or labels so a 4s1p should be a bit over 300g. That's a significant weight savings and makes 18Ah 4s4p possible for under 1250g.

I have a couple more batteries in the mail coming. It appears multistar LiHv individual cells come in 2600 and 2000 mAh variants.

I have two last battery combo's to make:

10400/LiHv/4s4p

8000/LiHv/4s4p

 

[lolo780]

I have a couple more batteries in the mail coming. It appears multistar LiHv individual cells come in 2600 and 2000 mAh variants.

I have two last battery combo's to make:

10400/LiHv/4s4p

8000/LiHv/4s4p

It looks like the 2600 cells are the ones to use based on your pack weights.

 

[beanbubba]

It looks like the 2600 cells are the ones to use based on your pack weights.

Agreed!

Well, I am side lined for at least 14 hours of 3D printing

Took a chance and had the batteries not fully balanced left-to-right and crashed hard 10 seconds after takeoff.

Live and learn...

 

[lolo780]

Agreed!

Well, I am side lined for at least 14 hours of 3D printing

Took a chance and had the batteries not fully balanced left-to-right and crashed hard 10 seconds after takeoff.

Live and learn...

View attachment 39651

How were the batteries positioned?

As long as you're not flying in strong crosswinds I think the best battery layout is bottom mount and not much wider than the Mavic bodywork. That keeps the batteries out of the prop wash for aerodynamic efficiency and the CG low and centered for stability.

 

[beanbubba]

How were the batteries positioned?

As long as you're not flying in strong crosswinds I think the best battery layout is bottom mount and not much wider than the Mavic bodywork. That keeps the batteries out of the prop wash for aerodynamic efficiency and the CG low and centered for stability.

Now you tell me

 

[lolo780]

I never did fly this triple 3s 4500 setup but this gives you an idea of the layout. 3s2p 9000 pack strapped to the bottom cover. I removed the plastic rear feet and space the pack down to leave an air gap.

Also I mentioned this before in another thread but in assistant you can tell the a3 flight controller where the GPS and IMU are located with xyz coordinates. Take some measurements and see if plugging those values in helps stability.

 

[beanbubba]

Hey Lolo. since I have to reprint all my rotor arms, did you ever come to a conclusion if longer is better? My theory is the longer ones have the prop wash clear the main body, but I have no clue if this creates any negative consequences.

 

[beanbubba]

I love when unintended consequences (the good kind) happen.

Chazz had me use the banana connectors on all the motors. When I just now removed all the busted rotor arms, I realized that the soldering effort is greatly diminished on the next go around

 

[lolo780]

How about making each end modular and bond it to carbon tube like an Inspire. Those usually break the carbon and are easy to repair.
For testing you could even use set screws on aluminum tube until you find the ideal setup.

 

[beanbubba]

How about making each end modular and bond it to carbon tube like an Inspire. Those usually break the carbon and are easy to repair.
For testing you could even use set screws on aluminum tube until you find the ideal setup.

I need a picture, not exactly following your description.

 

[beanbubba]

For reference a Thunderpower 4500HV 3s1p weighs 225g without heatshrink or labels so a 4s1p should be a bit over 300g. That's a significant weight savings and makes 18Ah 4s4p possible for under 1250g.

I came up with a big goose egg searching for that thunder power battery. Are you sure they still make it?

 

[lolo780]

Thunder Power RC TP4500-4SHV

All you need is a small Mavic frame to round tube 3d print and the rest can be had off the shelf. Standardized motor mount pattern and prop mounting means an optimized power system for the weight. CF tube will be light and strong plus easy to fix on the spot. (But hopefully you won't need to!)

 

[chazz]

I didn't realize that you were talking about the .DATs that I had looked at earlier. The glitch isn't associated with the concatenation of the two .DATs. Using the highest sampling rate it looks like
View attachment 39463
Seems that the FC thought it was necessary to correct a (small) pitch down, roll left situation. It's not clear, at least to me, why a combination of 100% PWM on the low corner and 10% PWM on the high corner was needed. These messages occurred throughout the eventLog stream
1489.107 : 93594 [L-FMU/LED]action changed. aircraft install warning!:yaw install err(1)
1494.207 : 93849 [L-FMU/LED]action changed. aircraft install warning!:mass center install err(2)
1519.707 : 95124 [L-FMU/LED]action changed. aircraft install warning!:yaw install err(1)
1526.507 : 95464 [L-FMU/LED]action changed. aircraft install warning!:mass center install err(2)
1550.307 : 96654 [L-FMU/LED]action changed. aircraft install warning!:yaw install err(1)
1552.007 : 96739 [L-FMU/LED]action changed. aircraft install warning!:mass center install err(2)
1572.407 : 97759 [L-FMU/LED]action changed. aircraft install warning!:yaw install err(1)
1577.507 : 98014 [L-FMU/LED]action changed. aircraft install warning!:mass center install err(2)
1589.407 : 98609 [L-FMU/LED]action changed. aircraft install warning!:yaw install err(1)
1591.107 : 98694 [L-FMU/LED]action changed. aircraft install warning!:mass center install err(2)
1601.307 : 99204 [L-FMU/LED]action changed. aircraft install warning!:yaw install err(1)

Maybe the Mavic is unbalanced causing the FC to occasionally take seemingly inappropriate actions?

How did you get to the error messages? Can you access these in CsvView?

 

[chazz]

The first piece is @lolo780 flight with factory motors and the bottom is mine with P3 motors. I call the top piece "Swan In Flight" and the bottom "Duck Out Of Water".

 

[chazz]

beanbubba i see what you mean with solder not sticking to the P3 motor leads. Two of the new motors i got are like that. I scraped them down with an Xacto and turned the temp up to 800. Hopefully the joint holds..

 

[BudWalker]

How did you get to the error messages? Can you access these in CsvView?

You'll need to use DatCon to see the eventLog stream.



And then press the View It button after DatCon processes the .DAT

 

[chazz]

You'll need to use DatCon to see the eventLog stream.

View attachment 39773

And then press the View It button after DatCon processes the .DAT

Thank You! Great tool by the way. Good Job!