Pspice are registered trademarks of Cadence.
All trademarks are the property of their respective owners.
The TPS7H4001-SP WCA model is a behavioral model intended for simulation of frequency response characteristics, such as phase margin, phase margin crossover frequency, and gain margin. Monte Carlo analysis can also be performed to observe the distribution of behavior across a specified sample size of devices. The following sources of variation are incorporated into the model:
Represented by data taken at 25°C before and after 1,000 hours of testing at 125°C, and meant to emulate 15 years operating at 65-95°C as defined by Group C specifications in MIL-PRF-38535
In the model, the specific device parameters influenced by the above variation include:
The following parameters are used within the model to define the internal characteristics, operating environment, and configuration of the device.
| Parameter | Description | Default Model Value |
|---|---|---|
| Mean_GMea | The mean value of the error amplifier transconductance | 1800 µS |
| Mean_GMps | The mean value of the power stage transconductance | 40 S |
| Mean_Vref | The mean value of the reference voltage | 0.605 V |
| Tol_GMea | The tolerance of the error amplifier transconductance [1] | 8.72 % |
| Tol_GMps | The tolerance of the power stage transconductance [1] | 5.22 % |
| Tol_Vref | The tolerance of the reference voltage [1] | 0.31 % |
| TID | The Total Ionizing Dose the device is subjected to, within the range of 0 to 100 krad | 0 krad |
| AGING |
Binary value indicating the presence of aging effects, where: 0 = No aging effects added 1 = Aging effects added |
0 |
| L | Output inductor value | 0.9 µF |
| FS | Switching frequency | 500 kHz |
| TEMP | Operating temperature, which is set within the simulation profile and takes the default value unless otherwise specified. | 27°C |
[1] For Gaussian distributions, tolerance is defined as the standard deviation (1-sigma) divided by the mean and is expressed as a percentage.
In addition to the listed model parameters, external component selection will also influence device behavior. The external components used in the default schematic take nominal values and, as such, users may see fit to add tolerances to them to model real-world variation.
Process variation for a given parameter is represented by its mean and tolerance values, which have been given default values based on statistical design simulations that are run at room temperature (27°C) and encompass variation caused by input voltage. If desired, the user may alter these values.
The influence of temperature and load on the listed device parameters is provided by nominal-case design simulations which exclude radiation and aging effects.
Radiation and aging effects are incorporated into the model in two ways. First, for the specified device parameters, the worst shifts observed between pre- and post-test measurements of approximately 50 units are determined for both TID testing and Life testing. The shifts associated with radiation and aging are applied independently to the mean parameter values in the model when it is set to simulate these conditions. If desired, users can incorporate similar shifts for non-modeled device parameters using Group C (Aging) and Group E (Radiation) reports. The model is also tuned so that the frequency response characteristics produced by simulations more closely match frequency response data collected using EVMs.
The EVM data used to tune the frequency response results of the TPS7H4001-SP WCA model was taken at two specific output voltages (1 V and 2.5 V) at 25°C using the default configuration seen on the 1-channel TPS7H4001-SP EVM, with the exception of changing the bottom feedback resistor to adjust the output voltage. Data collection was done using a total of 10 EVMs: three boards for High Dose Rate (HDR) TID testing, three boards for Low Dose Rate (LDR) TID testing, three boards for Life Test, and one control board.