Angulation Theory PDF
Angulation Theory PDF
One of the common functions found in all CT scanners is a mechanism that measures the (mechanical)
focal spot position and signals this information to control angular sampling and rotor position.
1.1.1.Angular Sampling:
The Data Measuring System (DMS) needs to take data samples (views) on predefined and accurate Rotor
Angle positions. This angle, together with the measured data of the detector channels, is provided to the
reconstruction system, so that the reconstruction system is able to backproject the data in the correct way.
1.1.2.Positioning:
The Main Drive Processor (MDP) needs feedback to be able to position the Rotor on the right angle in case
of a surview (PA, AP or Lateral).
Rev: 10/22/02 Full Philips Medical Systems Confidential and Proprietary Statement Page 1 of 10
1.2. RESOLUTION AND ACCURACY
The specification for our Angulation System of the Mx8000exp has 2 basic components:
1.2.1.Resolution
The resolution is the size of the small intervals, filling a whole measure (like rulers marked with different
divisions). The Mx8000 exp requires a resolution of 4640 parts per whole circle exactly.
1.2.2.Accuracy
The accuracy is the amount we believe in the measurement, in our case, tolerance to uncertainty measuring
angle. The Mx8000exp spec is +/- 15 parts out of 1,296,000 of a circle or 0.004 degrees. This is needed to
support high spatial resolution of 24 linepairs / cm.
1.3. RESOLUTION
In the Mx8000exp, one detector channel has a width of 1.408mm. The focal spot to detector distance is
1040mm. This means that we have (1040*2ð)/1.408 = 4640 imaginary detector channels (see Figure 1).
This means we need to have 4640 trigger samples (resolution of 4640) to be able to sample directly on
detector angle of motion boundaries, as necessary for Quarter Detector Offset (QDO) and Dual Focal Spot
(DFS).
1.4. ACCURACY
The common method used to measure the angle is a shaft encoder, or simply encoder. Angular position is
somehow 'encoded' into electronic signals of different format. To reach the accuracy in the Mx8000exp, the
form of encoder used is unique in the industry. It is called an electromagnetic resolver, or resolver for short.
2.
Figure 2: Angulation system
The basis for measurement is a variable coupling between a primary coil (stator) and a rotating coil (rotor)
that provides a variable output according to mechanical angle. Practical resolvers use sinusoidal coupling
and provide a pair of secondaries, a sine and a cosine. The resolver is actually nothing more then a
transformer with a single primary and dual rotating secondaries.
Figure 3: Resolver
The excitation signal used is an 8kHz sin-wave signal. The Sine and Cosine signals on the secondary side
will be a sine and cosine on an 8KHz carrier (see Figure 4), because of the rotating secondary side.
The resolver is designed in such a way that we will have 145 sin and cosine cycles per rotation
(a so called 145 cycle resolver).
A means must be provided to convert the resolver signals (analog sine and cosine) to digital values, used
later for trigger sampling by the DMS (Angular Sampling) and provide position feedback for the Main Drive
system (Positioning). This function, including signal monitoring, is done on the R2D board (resolver to digital
board).
To be able to do this, we need the carrier (the 8kHz excitation signal) on the secondary side.
Because there is no free track on the slipring to bring this carrier to the secondary side, a
solution has been made to make the carrier on the secondary side (excref) and transfer this to the
primary side (using the control link between DMS and Ghost) and use it as excitation signal.
Step 2:
The second step is to convert the resolver signals (analog sine and cosine) to digital values in
such a way that we can use it for positioning and accurate angular sampling.
We know the Rotor angle position, if we know:
To achieve this, each sine and cosine cycle is divided (by the R2D logic) into 256 intervals.
As already mentioned we have 145 sin and cosine cycles per rotation (145 resolver).
This means that we have in total 145*256= 37,120 intervals per revolutio n.
The logic part on the R2D will generate these positions as Quad-A and Quad-B signals (which
are 90 degrees out of phase), together with 1 Northmark signal for each cycle, and forward them
to the DMC.
As already mentioned in chapter 1.1, the Data Measuring System (DMS) needs to take data samples
(views) on predefined and accurate Rotor Angle positions. The Data Measurement Card (DMC) receives the
Quad-A and Quad-B from the R2D board. Quad-A and Quad-B are named AP0 and AP90 on the DMC.
The edges of both signals come at a rate of 148480 pulses (dataclock) per revolution (see Figure 6). This
big number results in the necessary accuracy of the angular samples to be able to support high spatial
resolution of 24 linepairs / cm.
The DMC also needs to know were to start with sample 0. This is reached by a zero flag signal (sensor on
rotating part + flag on frame) together with the northmark signal (from the R2D board), which will generate
the signal called Index.
The PhaseLockLoop (PLL) circuit and the Filter on the MDP will remove the jitter from the signal. Also a
delay circuit is used to get the excitation signal in phase with the excitation signal on the secondary. Before it
is used as an excitation drive signal, the Auto Gain Control (AGC) will set it to the right drive voltage. This
voltage should be set with the gain potentiometer in such way that the Sine and Cosine on the R2D are 1.9V
(true rms) See R2D adjustment procedure. The MDP also receives the AP0, AP90 and Index signal to use it
during the positioning of the Rotor (positioning loop).