SBVS395 July 2022 TPS7A57
PRODUCTION DATA
Output voltage margining is a technique that allows a circuit to be evaluated for how well changes are tolerated in the power supply. This test is typically performed by adjusting the supply voltage to a fixed percentage above and below its nominal output voltage.
This section discusses the implementation of a voltage mode margining application using the TPS7A57. A margining target of ±5% is used to demonstrate the chosen implementation.
Figure 8-26 shows a simplified visualization of the TPS7A57 REF pin with a voltage DAC.
Table 8-7 summarizes the design requirements.
PARAMETER | DESIGN VALUES |
---|---|
VIN | 2.5 V |
VOUT | 1.8 V nominal with ±5% margining |
CNR/SS | 4.7 μF |
RREF | 36 kΩ |
DAC VOUT range | 1.432 V to 2.108 V |
In this example, the output voltage is set to a nominal 1.8-V using a 36-kΩ resistor at the REF pin to GND. Equation 12 calculates the value for the RREF resistor.
The DAC63204, a 4-channel, 12-bit voltage and current output DAC with I2C, was selected and programmed into the voltage-output mode with an output range set between 1.432 V and 2.108 V. In conjunction with the 12-bit voltage DAC resolution, this output range allows a minimum step size (or LSB) of approximately 1.22 mV or 122 μA when the voltage-to-input (V2I) conversion or RV2I (100 kΩ) is taken into consideration. Into the 36-kΩ resistor, this LSB translates into a 0.44-mV voltage resolution or approximately 0.025% of the nominal 1.8-V targeted voltage. To achieve the full ±5% swing around the nominal voltage, the DAC63204 must source 3.1 μA or sink 3.7 μA.
The current flowing through RREF changes to 53.1 μA and 46.3 μA, thus adjusting the output voltage to 1.88 V and 1.7 V respectively.
Section 8.1.17 and Figure 8-28 show the voltage margining results.
When implementing voltage margining with this LDO, there is a time constant associated with its response. This RC time constant originates from the parallel combination of RREF and CNR/SS. Section 8.1.17 and Figure 8-28 show this RC effect.
Equation 13 calculates the time constant for this implementation:
where: