Motor_Position_MDD

Module --

High-Level Description

This module calculates the cumulative and corrected motor positions. It also outputs the sine and cosine position signals used for diagnostics and motor electrical and mechanical position.

Figures

Component Diagram

Variable Data Dictionary

Module Internal Variables

User defined typedef definition/declaration

• *In FDD 06B, Ver 6, range for SinHarTbl and CosHarTbl is given as -1 to 1 which is incorrect as it’s a 8bit signal. Informed to FDD Owner and will be updated accordingly.

Constant Data Dictionary

Calibration Constants

Program(fixed) Constants

Embedded Constants

Local

Global

This section lists the global constants used by the module. For details on global constants, refer to the Data Dictionary for the application.

Module specific Lookup Tables Constants

Functions/Macros used by the Sub-Modules

Library Functions / Macros

The library and functions / Macros that are called by the various sub modules are identified below,

1. _atan2_asm_

2. Limit_m

3. Sign_s16_m

Data Hiding Functions

1. Adc2_GetSinTheta1_Cnt_u16_m

2. Adc2_GetCosTheta1_Cnt_u16_m

3. Rte_Pim_MtrPosSnsr_EOLData

Global Functions/Macros Defined by this Module

1. MtrPos_Per1

Local Functions/Macros Used by this MDD only

Signal Correction

Design Rationale

This function is responsible for scaling and correcting the motor sense board sine and cosine signals based upon End Of Line calibrations. The correction uses the sensor calibrations to subtract the DC offset and normalize each signal to a nominal amplitude of +/- 1.

Description

Quadrature Correction

Design Rationale

This function is responsible for adjusting the cosine signal for any quadrature error between the sine and cosine signals. Ideally, the sine and cosine signals are exactly 90 degrees out of phase of eachother. This function will adjust the cosine signal based upon sensor calibrations to correct for any error from the ideal 90 degree phase difference.

The Quadrature correction should follow the following algorithm per requirements…

However, “δ” becomes a small enough number that the cos(δ) term becomes close enough to “1” to have a negligible effect on the signal. Therefore, in order to reduce throughput of this algorithm, the application of 1/cos(δ) is omitted.

Description

Round and Shift by 13 Bits

Design Rationale

This function will perform a rounding before applying a shift of 13 bits on the variable passed to the function. The function will use the sign bit of the input to help determine the value to be added to the signal before performing the shift operation. For a positive value, the value to be added is simply half of 2^[#bits to shift]… for this function, 13 bits are shifted off so 2^12 should be added to the signal before the shift is performed. However, for negative numbers, the value to be added is 2^[#bits to shift] -1. Therefore, the sign bit of the input is subtracted from 2^[#bits to shift], and this will result in the correct value for both positive and negative numbers.

Note that this function is designed to produce identical results to the FPM_FixWithRound_m that will cause a shift of 13 bits, but has a restricted input range as shown above.

Description

return (((x)+(D_HALFPREC13_CNT_S32-((x)>>D_SIGNBITSHIFT32_CNT_U16)))>>D_SHIFT13_CNT_S16)

Get Sine Voltage Macro Function

Design Rationale

The actual implementation of this function is in macro form. When called, the function (macro) will return the ADC reading for the corresponding analog input.

Get Cosine Voltage Macro Function

Design Rationale

The actual implementation of this function is in macro form. When called, the function (macro) will return the ADC reading for the corresponding analog input.

Software Module Implementation

Runtime Environment (RTE) Initial Values

Initialization Functions

Init: _Init1

Design Rationale

None

Module Internal and Outputs

Periodic Functions

Per: _Per1

Design Rationale

The computation of AlignedCumMtrPosMRF_Rev_M_s15p16 is optimized to take advantage of the fact that the relative position is sized to be a modulo 65536 value (i.e. rolls over at the16 bit value boundary). The operation specified to the compiler is to:

• Perform an unsigned subtraction of the z-1 and current position

• Explicitly mask the subtraction result to 16bits to ensure, that in the event that the compiler determines that a 32 bit subtraction is appropriate, the result is masked to only provide the lower 16 bits of the result.

• Cast the unsigned result of the mask as a signed 16 bit value so that rollovers occurring in the subtraction correctly indicate the integer amount of change.

• Finally cast the expression up to this point to a signed 32 bit number to promote the signed 16 integer and sign extend it to allow it to be added properly to the 32 bit accumulator, AlignedCumMtrPosMRF_Rev_M_s15p16

The above method eliminates any need for a conditional branch expression to handle the rollover points, thus avoiding the branching penalty.

Program Flow Start

N/A

Store Module Inputs to Local copies

Signal Processing

Store Local copy of outputs into Module Outputs

Program Flow End

N/A

Per: _Per2

Design Rationale

This function accesses the AlignedCumMtrPosMRF_Rev_M_s15p16 signal produced by Per1 which is running in the MtrCtrl ISR. Ther is no concern for data consistency in this case because the signal is 32 bits wide and it is assumed that the target processor supports atomic 32 data accesses (i.e as is the case for the TMS570).

This function should be run before MtrPos3_Per1, as the comparisons (in both the main and diverse paths) on the generated signals assume that this order will be followed.

Program Flow Start

Rte_Call_MtrPos2_Per1_CP0_CheckpointReached()

Store Module Inputs to Local copies

AssistAsmPolarity_T_f32 = Rte_IRead_MtrPos_Per2_AssistAssemblyPolarity_Cnt_s08()

Processing of function

Store Local copy of outputs into Module Outputs

Rte_IWrite_MtrPos_Per2_AlignedCumMtrPosMRF_Deg_f32(AlignedCumMtrPosMRF_Deg_T_f32)

Rte_IWrite_MtrPos_Per2_CumMtrPosMRF_Deg_f32(CumMtrPosMRF_Deg_T_f32)

Rte_IWrite_MtrPos_Per2_SysCCumMtrPosMRF_Deg_f32(CumMtrPosMRF_Deg_T_f32)

Rte_IWrite_MtrPos_Per2_CumMtrPosCRF_Deg_f32(CumMtrPosMRF_Deg_T_f32 * AssistAsmPolarity_T_f32)

Rte_IWrite_MtrPos_Per2_SysCCumMtrPosCRF_Deg_f32(CumMtrPosMRF_Deg_T_f32 * AssistAsmPolarity_T_f32)

Program Flow End

N/A

Fault Recovery Functions

None

Shutdown Functions

None

Interrupt Functions

None

Serial Communication Functions

SCom: _SCom_ReadEOLMtrCals

Design Rationale

None

Program Flow Start

N/A

Store Module Inputs to Local copies

None

Processing of function

SCom: _SCom_SetEOLMtrCals

Design Rationale

Provide exclusive area around EOL Data update to ensure the data stays consistent. If the BEMF R/C check, for example, pre-empts this update while the BEMF and R_BEMF values are inconsistent, a false failure will be detected. For simplicity, this implementation of the exclusive area protection does not provide a configurable type of protection, but rather always uses Suspend/Resume interrupts. If the need arises in the future this design can be updated.

Program Flow Start

N/A

Store Module Inputs to Local copies

None

Processing of function

Store Local copy of outputs into Module Outputs

N/A

Program Flow End

N/A

Execution Requirements

Execution Rates for sub-modules called by the Scheduler

Execution Requirements for Serial Communication Functions

Memory Map Definition Requirements

Sub Modules (Functions)

Local Functions

This table identifies the software segments for local functions identified in this module.

Known Issues / Limitations With Design

1. INLINE functions in GlobalMacro.h are not unit tested

Revision Control Log