Changes with R1625 introduced an error to MS561101 code

[EOSBandi]

Hi !
It seems that changes made to MS561101 calculations introduced an error. Under 20 degree Celsius, the altitude value will go haywire. Apparently changing to float from int64_t introduced an overflow somewhere (BaroTemp value will be nonsense).
As a quick fix I reverted these changes in my navi trunk, but since baro code does not belong to me, I would rather not touch it in the _shared.

[Y.Mita]

Same at my copter.
Sensor is Drotek 10DOF MS.
Before version up, stable under +-0.2m. After, +-3.0m.
Check twice on two configuration.

[EOSBandi]

Same at my copter.
Sensor is Drotek 10DOF MS.
Before version up, stable under +-0.2m. After, +-3.0m.
Check twice on two configuration.

It even worse, since it will gain or loose altitude constantly (at least the baro will think...)

[Alexinparis]

Sorry for this.

I think I found the bug (tested on the bench with some ice ).
and corrected in r1649

[Y.Mita]

Sorry for this.

I think I found the bug (tested on the bench with some ice ).
and corrected in r1649

Thank you!
I tried it on desktop, and seems more stable than before!
Tomorrow, I'll try it at field with MWNavi.

[Y.Mita]

Sorry for this.

I think I found the bug (tested on the bench with some ice ).
and corrected in r1649

Thank you!
I tried it on desktop, and seems more stable than before!
Tomorrow, I'll try it at field with MWNavi.

Yesterday, I went field and test flight my copter.
No problem, and stable than "MultiWii_2.3-navi-b4-xmas".
I replaced Sensor.cpp of "MultiWi2_3-navi-b5".

Thank you!