PartQuest Explore Promote

designs

groups

Add a bio to your profile to share information about yourself with other SystemVision users.

NCV891930 Evaluation Board Test Bench

RobertDavis
Designer256

Member for

9 years 6 months

designs

groups

Add a bio to your profile to share information about yourself with other SystemVision users.

Description

This version of the NCV891930 "Virtual Evaluation Board" simulates &gt; 4x faster than if full MOSFET switch models are used. This model represents most of the features of the part. By changing various input stimulus and loading conditions, you can test these capabilities in the context of their intended usage.For the nominal configuration, the input voltage is set to 12V, the synci source is delayed for 10 msec (well beyond the simulation end-time), and the load resistance is fixed at 1 Ohm until 500 us, then it jumps to 10 kOhm for 100 us, then back to 1 Ohm. The soft-start operates normally, but during the high resistance operation of the load transient, a single pulse-skip interval is observed. You can zoom in on the PWM signal and see that only two 500 kHz pulses were applied during the “burst” interval, but it will generate 3 pulses under other operating conditions. You can also see that the PWM frequency is spread-spectrum after the soft-start is complete, by measuring the clock frequency (as a waveform) in the main waveform viewer.To test the “GH continuous ON” behavior, you can select the “DC” option in the load resistance function generator to set the value to a fixed 1 Ohm, and set the input voltage to 5.2V. In the simulation you'll observe several long periods of "continuous GH" operation, with 200 ns boost recharge windows every 16 us, after soft-start has finished.To test SYNCI operation, set the input voltage to 12V and the load resistance to “Pulse” (i.e. back to their original values), and then change the synci source delay to 10 ns (instead of 10 ms). Then the device will operate in forced PWM mode, with negative inductor currents allowed during the high load resistance period (i.e. not diode-emulation), and no pulse-skip. Other features can also be tested (high voltage frequency fold-back, output voltage select, LDO bypass jumper removed, etc.). In addition, you can force abnormal operating states using the two fixed logic sources available in the lower left corner of the schematic. For example, if the controller signal “d_spread_spectrum” is disconnected right at the input to the oscillator, and then a new net is connected from “d_set0” to that open input port, then the oscillator's spread-spectrum function will be disabled regardless of the normal controller commanded behavior.

NCV891930MW00 Model

RobertDavis
Designer256

RobertDavis

Member for

9 years 6 months

designs

groups

Add a bio to your profile to share information about yourself with other SystemVision users.

Description

This version of the NCV891930 "Virtual Evaluation Board" represents most of the features of the part. By changing various input stimulus and loading conditions, you can test these capabilities in the context of their intended usage.Note: This version is different from Version 7 because it now includes the Renlo internal load resistance that is activated when the output voltage exceeds the target voltage by 7.5%. It will also now function in continuous-on mode at low input voltage when using either the local rosc (which is in foldover mode) or synci if it has a valid synci input signal. For the nominal configuration, the input voltage is set to 12V, the synci source is delayed for 10 msec (well beyond the simulation end-time), and the load resistance is fixed at 1 Ohm until 500 us, then it jumps to 10 kOhm for 100 us, then back to 1 Ohm. The soft-start operates normally, but during the high resistance operation of the load transient, a single pulse-skip interval is observed. You can zoom in on the PWM signal and see that only two 500 kHz pulses were applied during the “burst” interval, but it will generate 3 pulses under other operating conditions. You can also see that the PWM frequency is spread-spectrum after the soft-start is complete, by measuring the clock frequency (as a waveform) in the main waveform viewer.To test the “GH continuous ON” behavior, you can select the “DC” option in the load resistance function generator to set the value to a fixed 1 Ohm, and set the input voltage to 5.5V. In the simulation you'll observe several long periods of "continuous GH" operation, with 200 ns boost recharge windows every 16 us, after soft-start has finished.To test SYNCI operation, set the input voltage to 12V and the load resistance to “Pulse” (i.e. back to their original values), and then change the synci source delay to 10 ns (instead of 10 ms). Then the device will operate in forced PWM mode, with negative inductor currents allowed during the high load resistance period (i.e. not diode-emulation), and no pulse-skip. Other features can also be tested (high voltage frequency fold-back, output voltage select, LDO bypass jumper removed, etc.). In addition, you can force abnormal operating states using the two fixed logic sources available in the lower left corner of the schematic. For example, if the controller signal “d_spread_spectrum” is disconnected right at the input to the oscillator, and then a new net is connected from “d_set0” to that open input port, then the oscillator's spread-spectrum function will be disabled regardless of the normal controller commanded behavior.

NCV891930 Symbol

RobertDavis
Designer256

RobertDavis

Member for

9 years 6 months

designs

groups

Add a bio to your profile to share information about yourself with other SystemVision users.

Description

This version of the NCV891930 "Virtual Evaluation Board" simulates &gt; 4x faster than if full MOSFET switch models are used. This model represents most of the features of the part. By changing various input stimulus and loading conditions, you can test these capabilities in the context of their intended usage.For the nominal configuration, the input voltage is set to 12V, the synci source is delayed for 10 msec (well beyond the simulation end-time), and the load resistance is fixed at 1 Ohm until 500 us, then it jumps to 10 kOhm for 100 us, then back to 1 Ohm. The soft-start operates normally, but during the high resistance operation of the load transient, a single pulse-skip interval is observed. You can zoom in on the PWM signal and see that only two 500 kHz pulses were applied during the “burst” interval, but it will generate 3 pulses under other operating conditions. You can also see that the PWM frequency is spread-spectrum after the soft-start is complete, by measuring the clock frequency (as a waveform) in the main waveform viewer.To test the “GH continuous ON” behavior, you can select the “DC” option in the load resistance function generator to set the value to a fixed 1 Ohm, and set the input voltage to 5.2V. In the simulation you'll observe several long periods of "continuous GH" operation, with 200 ns boost recharge windows every 16 us, after soft-start has finished.To test SYNCI operation, set the input voltage to 12V and the load resistance to “Pulse” (i.e. back to their original values), and then change the synci source delay to 10 ns (instead of 10 ms). Then the device will operate in forced PWM mode, with negative inductor currents allowed during the high load resistance period (i.e. not diode-emulation), and no pulse-skip. Other features can also be tested (high voltage frequency fold-back, output voltage select, LDO bypass jumper removed, etc.). In addition, you can force abnormal operating states using the two fixed logic sources available in the lower left corner of the schematic. For example, if the controller signal “d_spread_spectrum” is disconnected right at the input to the oscillator, and then a new net is connected from “d_set0” to that open input port, then the oscillator's spread-spectrum function will be disabled regardless of the normal controller commanded behavior.

Fast (Ideal Switches) Version 8 of NCV891930 Virtual Eval Board

RobertDavis
Designer256

RobertDavis

Member for

9 years 6 months

designs

groups

Add a bio to your profile to share information about yourself with other SystemVision users.

Description

This version of the NCV891930 "Virtual Evaluation Board" simulates &gt; 4x faster than if full MOSFET switch models are used. This model represents most of the features of the part. By changing various input stimulus and loading conditions, you can test these capabilities in the context of their intended usage.For the nominal configuration, the input voltage is set to 12V, the synci source is delayed for 10 msec (well beyond the simulation end-time), and the load resistance is fixed at 1 Ohm until 500 us, then it jumps to 10 kOhm for 100 us, then back to 1 Ohm. The soft-start operates normally, but during the high resistance operation of the load transient, a single pulse-skip interval is observed. You can zoom in on the PWM signal and see that only two 500 kHz pulses were applied during the “burst” interval, but it will generate 3 pulses under other operating conditions. You can also see that the PWM frequency is spread-spectrum after the soft-start is complete, by measuring the clock frequency (as a waveform) in the main waveform viewer.To test the “GH continuous ON” behavior, you can select the “DC” option in the load resistance function generator to set the value to a fixed 1 Ohm, and set the input voltage to 5.2V. In the simulation you'll observe several long periods of "continuous GH" operation, with 200 ns boost recharge windows every 16 us, after soft-start has finished.To test SYNCI operation, set the input voltage to 12V and the load resistance to “Pulse” (i.e. back to their original values), and then change the synci source delay to 10 ns (instead of 10 ms). Then the device will operate in forced PWM mode, with negative inductor currents allowed during the high load resistance period (i.e. not diode-emulation), and no pulse-skip. Other features can also be tested (high voltage frequency fold-back, output voltage select, LDO bypass jumper removed, etc.). In addition, you can force abnormal operating states using the two fixed logic sources available in the lower left corner of the schematic. For example, if the controller signal “d_spread_spectrum” is disconnected right at the input to the oscillator, and then a new net is connected from “d_set0” to that open input port, then the oscillator's spread-spectrum function will be disabled regardless of the normal controller commanded behavior.

NCV6356_Demo_Board_fastsim_Phil

Bob Davis
Designer40261

Bob Davis

Member for

8 years 7 months

76

designs

groups

Add a bio to your profile to share information about yourself with other SystemVision users.

Managing PWM Waveforms - Buck Converter

Dagi
Designer190236

Dagi

Member for

6 years 8 months

designs

groups

Add a bio to your profile to share information about yourself with other SystemVision users.

Description

To demonstrate data reduction in a PWM design I've copied and modified Mike Donnelly's Buck DC to DC Converter at https://www.systemvision.com/design/buck-dc-dc-converter-switchingI've added a simplified softstart to simulate startup. Sofstart is implemented by ramping up internal voltages over a 5ms period after a 0.5ms delay for a total startup time of 5.5ms. Mike's load transient is timed to activate after steady state. A line transient can be simulated by adding a pulsed voltage source in series with v_line.The total time to simulate startup, steady state and the load transient is set to 10.5ms of which only the final 4ms involve the response to the transient disturbance. Downloaded .csv files of the output voltage (v_out_buck), inductor current (l1inductor/i), or any other signal cover over 84,000 time points.For a 40us switching cycle, a 1us resolution samples each switching cycle 40 times and requires only 4,000 time points to cover the steady state and load transient response. A voltage source is included in the design to mark off this sampling resolution with a sawtooth signal that starts at 6.5ms. The sawtooth appears at the net named x_smpl. Downloading x_smpl along with v_out_buck to Excel enables edits that reduce the original 84,000 time points to 4,000.To demonstrate the data reduction technique for large .csv downloads,t he design includes a dedicated table voltage source which has been preloaded with a set of 4,000 samples from an earlier simulation. The voltage signal vout_reduced displays those 4,000 samples. The table voltage source can be edited to insert to store reduced data records of any transient response generated by this buck converter.Copies of the table voltage source can be placed into this or any other design in order to display the transient response or apply it to another simulation. This mechanism can be used to test designs that are driven by the buck converter output for their ability to withstand the transient disturbance. Other simulations can be used to compare the simulated response to measured data or to analyze the simulated transient response.Watch for an accompanying blog that will explain how the table voltage source works.