Identify reference amplitude and digital code as separate signal inputs
The DAC8811 separates the digital choice of current path from the analog quantity being scaled. Its R-2R architecture steers ladder branches according to the applied code. The external reference establishes the current scale. These are two inputs to the multiplication function: code determines the steering relationship, while reference amplitude supplies the analog magnitude on which that relationship acts.
The reference can vary in polarity within the named device's documented context. Its role consequently extends beyond a fixed calibration label attached to a digital converter. Changing the analog reference changes the signal presented to the multiplying path, even when the code is unchanged. Changing code changes that path's scaling relationship even when the reference is unchanged. Keeping the two roles visible explains the multiplying description.
The digital supply and serial interface serve different functions. The supply supports device operation and the interface delivers code; neither supplies the external analog reference function merely by being present. Likewise, successful code transfer does not describe the conversion of output current into the required voltage. The analog signal path continues through the ladder, its current-output interface and the external conversion stage after the digital value has been established.
Separate the reference load from the code-dependent output-node load
The ladder's load on the reference and the impedance at IOUT have different code dependencies. TI describes a code-independent load presented to the external reference. At the current-output terminal, however, the impedance looking into the DAC changes with code. Both statements can hold in the same R-2R converter because they concern different points in the signal path.
That distinction gives an amplifier imperfection a code-related consequence. The external current-to-voltage amplifier has an input-offset contribution. As the DAC's IOUT impedance changes, the amplifier's noise gain also changes, allowing that offset to be modulated with code. TI identifies resulting transfer nonlinearity as a concern. An error that is associated with the amplifier can therefore interact with the converter's changing network rather than remaining one fixed output displacement.
A code-independent reference-load description cannot remove this output-node relationship. It describes what the analog reference drives, while the code-dependent impedance describes the interface seen by the conversion amplifier. Reading the two together explains why reference drive and amplifier offset require different attention. This is a static transfer mechanism tied to code-dependent impedance, separate from the brief disturbances associated with digital feedthrough or a code-transition glitch.
Preserve the matched feedback role in current-to-voltage conversion
The current-output DAC requires a current-to-voltage conversion role when the desired result is a voltage. TI describes an external amplifier in a transimpedance configuration and an on-chip feedback resistor available as RFB. That resistor is sized to match the ratios in the DAC's R-2R ladder. Its purpose is therefore tied to the converter's internal resistor relationship rather than being an arbitrary external voltage-gain element.
The matched element and external amplifier act together in the first conversion stage. The DAC supplies the steered current, the amplifier performs the conversion, and the matched feedback role preserves the ladder-related transfer basis. Describing only an amplifier gain or a permitted current-output voltage would omit how this particular converter and its feedback element participate in producing the voltage signal.
The application discussion then identifies a separate gain-and-offset stage. Its external resistor matching and amplifier properties belong to that later stage, rather than being automatically covered by the first stage's on-chip matching. Keeping the stages distinct makes the error allocation understandable: ladder-related feedback matching, conversion-amplifier behavior and the added summing network are connected responsibilities with different origins. A matched RFB element does not supply matching for every external resistor in the complete application.
Distinguish two-quadrant multiplication from a bipolar output stage
TI identifies the base DAC8811 as a two-quadrant multiplying DAC. The analog-reference polarity and the current-output polarity have a documented relationship. Changing the digital code changes the current magnitude through that multiplying relationship; it does not by itself establish a voltage output spanning both sides of a chosen center. The base converter's function and a complete bipolar voltage-output function therefore describe different architectures.
The named bipolar example adds an external summing amplifier and an offset contribution. The application prose explains a current-to-voltage stage followed by another amplifier that applies the offset. The additional stage changes how the converted signal is positioned and scaled. Its presence is the substantive reason that the four-quadrant application should be distinguished from the converter's base two-quadrant description.
This extension also introduces its own accuracy context. TI identifies external resistor mismatch as an error source in the bipolar arrangement, and the added amplifier retains separate offset and dynamic requirements. Those properties do not follow from the DAC's digital resolution or the availability of matched RFB. The multiplying capability, initial conversion and quadrant extension must all remain identifiable when describing the analog function that an application actually requires.
Sources and further reading
DAC8811 SLAS411D bounded feature and application prose on physical pages 11–13 and 16. Page 14 was read but its current-source design is excluded. No formula, component value, circuit recipe or achieved linearity is presented.