Precision 4.096V reference for ADC/DAC systems
The Maxim Integrated MAX6126B41+ is a precision series voltage reference delivering a fixed 4.096V output with ±0.1% initial tolerance and a 12ppm/°C temperature coefficient. This output voltage aligns directly with the common full-scale reference for 12- to 16-bit successive-approximation ADCs and multiplying DACs, simplifying the system voltage-divider math. The series topology means the part sources current to the load rather than sinking it, which suits it to driving the reference input of an ADC without an external buffer in most cases. Low-frequency noise is specified at 2.4µVp-p over the 0.1Hz to 10Hz band, a critical parameter for high-resolution measurements where flicker noise would otherwise limit the effective number of bits.
The ±0.1% initial tolerance sets the absolute accuracy floor for the reference voltage. In a 16-bit ADC system with a 4.096V reference, this 4.1mV worst-case error at 25°C consumes roughly 0.1% of the full-scale range — acceptable for most industrial measurements but worth noting if the design targets better than 14-bit accuracy without a calibration step. The 12ppm/°C temperature coefficient means the output voltage drifts by about 49µV per degree Celsius. If the system must hold 0.05% accuracy across temperature, a lower-tempco grade (e.g., the MAX6126A41+ with a tighter spec) would be the pick. Output current capability of 10mA is generous for a precision reference — enough to drive the reference input of multiple ADCs or a small resistive divider network. The supply current draw of 725µA is the quiescent current into the VIN pin, not the output load; total input current will be 725µA plus the output current delivered to the load.
The MAX6126B41+ is offered in an 8-uMAX/uSOP package (also described as 8-TSSOP / 8-MSOP with 0.118-inch body width, 3.00mm width). This is a surface-mount package with a 0.65mm lead pitch, suitable for standard PCB assembly processes. The ±0.1% tolerance is the B-grade selection; the A-grade variant offers tighter tolerance for higher-accuracy applications.