SLPS764 Data sheet

SLPS764A September 2024 – October 2025 RES60A-Q1

ADVANCE INFORMATION

Package Options

Refer to the PDF data sheet for device specific package drawings

Mechanical Data (Package|Pins)

DWV|8

Thermal pad, mechanical data (Package|Pins)

Orderable Information

slps764a_oa

6.3.1 Absolute and Ratiometric Tolerances

The resistors of the RES60A-Q1 are described by the following equations:

Equation 1.

R_{HV} = R_{HVnom} \times (1± t_{absRHV}) = R_{HVnom} \times (1± t_{RHV}) \times (1± t_{SiCr})

Equation 2.

R_{LV} = R_{LVnom} \times (1± t_{absRLV}) = R_{LVnom} \times (1± t_{RLV}) \times (1± t_{SiCr})

R_HVnom and R_LVnom are the nominal values of each resistor. The parameter t_abs is an error term that describes the absolute tolerance of the RES60A-Q1 resistor in question, such that |t_abs| ≤ 15%. For example, a nominally 12.5MΩ resistor with t_abs = 5% actually measures 13.125MΩ. This error is analogous to the specified absolute tolerance of most single-element resistors, or the end-to-end tolerance of more specialized resistor dividers.

Note: The RES60A-Q1 is not a laser-trimmed device. Each ratio of the RES60A-Q1 features a unique die specifically optimized for that ratio, providing the precise matching and consistent thermal characteristics necessary to achieve extremely low drift.

The absolute tolerance is dominated by the variation in the SiCr resistivity, t_SiCr. The two resistors of a given RES60A-Q1 are interdigitated and come from the same area of the wafer; therefore, t_SiCr is effectively the same for both of the two resistors, although t_SiCr varies on a part-to-part basis.

The following examples show that when the divider is considered in ratiometric terms, the t_SiCr error terms drop out. Parameter t_Rx is a residual error term that describes the remaining effective tolerance of each resistor of the given RES60A-Q1 device, after accounting for the universal t_SiCr.

Equation 3.

\frac{R_{HV}}{R_{LV}} = \frac{R_{HVnom} \times (1± t_{RHV}) \times (1± t_{SiCr})}{R_{LVnom} \times (1± t_{RLV}) \times (1± t_{SiCr})} = \frac{R_{HVnom} \times (1± t_{RHV})}{R_{LVnom} \times (1± t_{RLV})} = G_{nom} \times \frac{(1± t_{RHV})}{(1± t_{RLV})} = G

Equation 4.

\frac{R_{HV}}{R_{LV} + R_{HV}} = \frac{R_{HVnom} \times (1± t_{RHV}) \times (1± t_{SiCr})}{R_{LVnom} \times (1± t_{RLV}) \times (1± t_{SiCr}) + R_{HVnom} \times (1± t_{RHV}) \times (1± t_{SiCr})} = \frac{R_{HVnom} \times (1± t_{RHV})}{R_{LVnom} \times (1± t_{RLV}) + R_{HVnom} \times (1± t_{RHV})}

The individual values of t_RHV and t_RLV describe the tolerance of each individual resistor, but are not independent variables in a Gaussian sense. Rather, the matching of these values to each other (by design) is used to achieve highly stable ratiometric relationships between the resistors, giving an effective ratio with an extremely low error.

The RES60A-Q1 is specified with a maximum initial divider ratio tolerance of 0.1%, meaning that the relationship between the actual divider ratio, G, and the nominal ratio, G_nom, of a given divider is described by the following:

Equation 5.

G = \frac{R_{HV}}{R_{LV}} = G_{nom} \times (1± t_{D})

such that t_D ≤ 0.1%. The limits of t_D for the RES60A-Q1 are enforced by precise parametric testing in production, with the divider voltage swept from V_D = 250V to V_D = 1000V. Single-element resistors do not have an equivalent to t_D, because no part-to-part matching is considered other than the gradeout limit. In other divider data sheets, the equivalent of t_D is often called ratio tolerance.

Because any devices that do not meet these criteria are screened out at final test, this equation can be used with the previous equations to prove the effective bounds of t_RHV and t_RLV for a given ratio. Therefore, despite the wide absolute tolerance bounds of ±15%, the worst-case initial absolute error tolerances of t_RHV and t_RLV are within approximately ±0.115% of each other.