Hi everyone,
I am observing unexpected scaling behavior when modeling a 5 kW, 400 V, 50 Hz synchronous machine using the Synchronous Machine Round Rotor block from Simscape Electrical (Fundamental Parameters mode).
When attempting to excite the rotor using a real-world DC voltage source connected directly to the field terminals (fd+ / fd-), the internal parameter scaling of the Park (d/q) transformation creates a massive discrepancy between the physical terminal values and the model’s internal reciprocal system.
Here is the setup and the observed problem:
Machine Target Specs: 5 kW, U_rated = 400, f = 50 Hz, p = 1, driven mechanically at nominal speed 3000 RPM
Rotor Setup: I_fd = 10A, no-load field current required for rated terminal voltage, R_fd = 4.8 Ohm
Expected Behavior: Applying a 48 V DC source to fd+/fd- should result in an I_fd of 10 A and induce the nominal line-to-neutral peak voltage of 326 V, approx. 230 V RMS at the open-circuit stator terminals.
Actual Behavior:
If a 48 V DC Voltage Source is connected to fd+/fd-, the resulting field current drops to a tiny fraction of an Ampere (approx 0.001 A), and no stator voltage is induced.
If a Controlled Current Source set to 10 A is used instead, the stator side behaves 100% correctly (approx 326 V_peak phase voltage). However, the voltage measured across the fd+/fd- terminals skyrockets to over 460 kV.
Checking the Base Values log of the block reveals the underlying transformation split:
Field circuit base values (Non-reciprocal system): V_base = 48 V, I_base = 10A (Physical values).
Rotor circuit base values (Reciprocal system): V_base = 7804.88 V, I_base = 0.800781A (Stator-referred dq-system).
Because the physical block pins (fd+/fd-) interface directly with the stator-referred reciprocal Park transformation, the effective input impedance seen at the terminals scales with the square of the turns ratio (ü^2). As a result:
An applied 48V source sees an effective internal resistance of R_fd = approx 46.8 kOhm, collapsing the current.
Forcing 10 A into the terminals forces U_fd = 10A * 46.8kOhm to approx 468 kV
Is there a recommended way within Simscape Electrical to parameterize the block so that physical field voltages 48V DC can be applied directly to fd+/fd- without having to manually pre-scale R_fd by the reciprocal base ratio (ü^2)?
Is the documentation stating "Field circuit voltage required to produce rated terminal voltage" intended only for signal-based Simulink library blocks rather than Simscape domain physical connections?
Any insight into the preferred modeling workflow for real-world exciter circuits driving this Simscape block would be greatly appreciated!Hi everyone,
I am observing unexpected scaling behavior when modeling a 5 kW, 400 V, 50 Hz synchronous machine using the Synchronous Machine Round Rotor block from Simscape Electrical (Fundamental Parameters mode).
When attempting to excite the rotor using a real-world DC voltage source connected directly to the field terminals (fd+ / fd-), the internal parameter scaling of the Park (d/q) transformation creates a massive discrepancy between the physical terminal values and the model’s internal reciprocal system.
Here is the setup and the observed problem:
Machine Target Specs: 5 kW, U_rated = 400, f = 50 Hz, p = 1, driven mechanically at nominal speed 3000 RPM
Rotor Setup: I_fd = 10A, no-load field current required for rated terminal voltage, R_fd = 4.8 Ohm
Expected Behavior: Applying a 48 V DC source to fd+/fd- should result in an I_fd of 10 A and induce the nominal line-to-neutral peak voltage of 326 V, approx. 230 V RMS at the open-circuit stator terminals.
Actual Behavior:
If a 48 V DC Voltage Source is connected to fd+/fd-, the resulting field current drops to a tiny fraction of an Ampere (approx 0.001 A), and no stator voltage is induced.
If a Controlled Current Source set to 10 A is used instead, the stator side behaves 100% correctly (approx 326 V_peak phase voltage). However, the voltage measured across the fd+/fd- terminals skyrockets to over 460 kV.
Checking the Base Values log of the block reveals the underlying transformation split:
Field circuit base values (Non-reciprocal system): V_base = 48 V, I_base = 10A (Physical values).
Rotor circuit base values (Reciprocal system): V_base = 7804.88 V, I_base = 0.800781A (Stator-referred dq-system).
Because the physical block pins (fd+/fd-) interface directly with the stator-referred reciprocal Park transformation, the effective input impedance seen at the terminals scales with the square of the turns ratio (ü^2). As a result:
An applied 48V source sees an effective internal resistance of R_fd = approx 46.8 kOhm, collapsing the current.
Forcing 10 A into the terminals forces U_fd = 10A * 46.8kOhm to approx 468 kV
Is there a recommended way within Simscape Electrical to parameterize the block so that physical field voltages 48V DC can be applied directly to fd+/fd- without having to manually pre-scale R_fd by the reciprocal base ratio (ü^2)?
Is the documentation stating "Field circuit voltage required to produce rated terminal voltage" intended only for signal-based Simulink library blocks rather than Simscape domain physical connections?
Any insight into the preferred modeling workflow for real-world exciter circuits driving this Simscape block would be greatly appreciated! Hi everyone,
I am observing unexpected scaling behavior when modeling a 5 kW, 400 V, 50 Hz synchronous machine using the Synchronous Machine Round Rotor block from Simscape Electrical (Fundamental Parameters mode).
When attempting to excite the rotor using a real-world DC voltage source connected directly to the field terminals (fd+ / fd-), the internal parameter scaling of the Park (d/q) transformation creates a massive discrepancy between the physical terminal values and the model’s internal reciprocal system.
Here is the setup and the observed problem:
Machine Target Specs: 5 kW, U_rated = 400, f = 50 Hz, p = 1, driven mechanically at nominal speed 3000 RPM
Rotor Setup: I_fd = 10A, no-load field current required for rated terminal voltage, R_fd = 4.8 Ohm
Expected Behavior: Applying a 48 V DC source to fd+/fd- should result in an I_fd of 10 A and induce the nominal line-to-neutral peak voltage of 326 V, approx. 230 V RMS at the open-circuit stator terminals.
Actual Behavior:
If a 48 V DC Voltage Source is connected to fd+/fd-, the resulting field current drops to a tiny fraction of an Ampere (approx 0.001 A), and no stator voltage is induced.
If a Controlled Current Source set to 10 A is used instead, the stator side behaves 100% correctly (approx 326 V_peak phase voltage). However, the voltage measured across the fd+/fd- terminals skyrockets to over 460 kV.
Checking the Base Values log of the block reveals the underlying transformation split:
Field circuit base values (Non-reciprocal system): V_base = 48 V, I_base = 10A (Physical values).
Rotor circuit base values (Reciprocal system): V_base = 7804.88 V, I_base = 0.800781A (Stator-referred dq-system).
Because the physical block pins (fd+/fd-) interface directly with the stator-referred reciprocal Park transformation, the effective input impedance seen at the terminals scales with the square of the turns ratio (ü^2). As a result:
An applied 48V source sees an effective internal resistance of R_fd = approx 46.8 kOhm, collapsing the current.
Forcing 10 A into the terminals forces U_fd = 10A * 46.8kOhm to approx 468 kV
Is there a recommended way within Simscape Electrical to parameterize the block so that physical field voltages 48V DC can be applied directly to fd+/fd- without having to manually pre-scale R_fd by the reciprocal base ratio (ü^2)?
Is the documentation stating "Field circuit voltage required to produce rated terminal voltage" intended only for signal-based Simulink library blocks rather than Simscape domain physical connections?
Any insight into the preferred modeling workflow for real-world exciter circuits driving this Simscape block would be greatly appreciated! simscape, simulink MATLAB Answers — New Questions