lean angle/ throttle position TC

[e-shock]

The last generation of traction control system consider the lean angle as input to the system.
This is fundamental to correctly tune the system and having the best performance.

E-LEAN is currently the unique sensor that on-line measures the lean angle of the motorcycle with a method that has been specifically designed for two-wheeled vehicle.
It does not use 3D magnetomers (they are not needed) but is is based on 3D accelerometers, 3D gyroscopes and speed signal.

When an IMU is used on a motorcycle, it is very important to be pay attention on the mounting position avoiding to place the sensor where the effect of the engine's vibration is to high.

Ivo

[e-shock]

The last generation of traction control systems consider the lean angle as input to the system.
This is fundamental to correctly tune the system and having the best performance.

E-LEAN by e-Shock is currently the unique sensor that on-line measures the lean angle of the motorcycle with a method that has been specifically designed for two-wheeled vehicle.
It does not use 3D magnetomers (they are not needed) but is is based on 3D accelerometers, 3D gyroscopes and speed signal.

When an IMU is used on a motorcycle, it is very important to be pay attention on the mounting position avoiding to place the sensor where the effect of the engine's vibration is to high.

Ivo

[Holmz]

The problem with some generic IMU is they use 3D magnetometers to get additional input for the sensor fusion algorithm. They measure earth magnetic field, that's why alternator magnets or high current cables (coil charge current is high) could cause problems. On other solutions, they use GPS speed and heading.
...

Any magnets are a problem but the "hard iron" calibration can take the majority of that out.

If the wiring is run as twisted pairs with a separate earth twisted to (for instance) the coils V+ should help.
But most of those magnetic fields will be high frequency... I suppose if they are always firing then there is a low frequency bias to the magnetic field.
It does help to look at the raw data as you will only see the magnetic field effects that way, and the bike only needs to be running, not moving.

Ivo's e-loan/Dolfin is probably the way to go as I doubt that the maths can be done in a datalogger or ECU.

[Polux RSV]

The math involved for acc and gyro sensor fusion are matrix multiplications, additions, etc...At high rate, about 50 to 200Hz to get good quality. My prototype is ready, and only need real ride to confirm it is precise enough for TC, and to drive a gyrocam

Angelo

[Holmz]

The math involved for acc and gyro sensor fusion are matrix multiplications, additions, etc...At high rate, about 50 to 200Hz to get good quality. My prototype is ready, and only need real ride to confirm it is precise enough for TC, and to drive a gyrocam

Angelo

Nice!
Are you doing that in the ECU?
Or where?

[Polux RSV]

The ecu and dash are not able to do such math.
I built my own system, which could send CAN messages. I used small sensors board and algorithm used in the radio controlled models world. The algorithm is called DCM, direct cosine matrix. I absolutely don't understand anything about the math but I used it "as is" and it works.

Angelo

[e-shock]

You can directly PLUG the E-LEAN sensor on the MOTEC ECU.
You don't need to buy DOLPHIN.

Angelo, the DCM algorithm has been designed for planes.
Using this kind of algorithm with a motorcycle you will get a lean angle estimation that has the right trend, but that has not a sufficient precision to be used in a TC system.

[Polux RSV]

Yes, I already know the limitations of DCM, and I am working on this. I am also trying to use other algorithms .
DCM as no correction for "change of speed" in the forward direction(acceleration and braking phase). And will not be correct under wheelie or stopie.
But it is a hobby project. And my multiple logs didn't told me I need the TC

Angelo