AD5622YKSZ-1REEL7 output accuracy cannot be judged from its 12-bit resolution alone. Resolution defines the available code steps. The voltage you obtain also depends on the supply used as the reference, the converter's errors and the circuit around it. Start with the ideal code calculation, then build an error budget for the output range, load and conditions your application actually needs.
Keep Resolution and Voltage Accuracy Separate
The AD5622YKSZ-1REEL7 part page identifies the complete ordering code. Use that full code when matching a quotation to an approved design. A similar family name does not establish the same grade, package or interface option.
Analog Devices lists AD5622 as a 12-bit voltage-output DAC. Its reference comes from its power-supply input. This makes the actual supply voltage relevant to the output calculation; a rail labelled 3.3 V is not automatically an exact 3.300000 V reference.
Resolution answers how finely the digital setting is divided. Accuracy asks how close the resulting voltage is to the required value. A small ideal step does not remove supply error or the converter's own errors.
Calculate the Ideal Output First
The AD5602/AD5612/AD5622 data sheet, Revision F, gives the straight-binary transfer equation on page 15. For AD5622, the ideal relationship is:
Ideal output voltage = VDD × D / 4096
D is the decimal DAC code from 0 to 4095. VDD is the supply voltage at the device. The equation describes an ideal converter in normal output operation after settling. It does not describe start-up or power-down, a measured output or a complete error limit.
Assume an exact 3.3 V supply for this hypothetical calculation. One ideal step is 3.3 / 4096 V, or about 0.806 mV. The table shows three reference points.
DAC Code | Calculation at Exact 3.3 V | Ideal Output |
|---|---|---|
0 | 3.3 × 0 / 4096 | 0 V |
2048 | 3.3 × 2048 / 4096 | 1.650 V |
4095 | 3.3 × 4095 / 4096 | Approximately 3.29919 V |
The maximum code is one ideal step below the supply. Dividing by 4095 to force the maximum code to equal 3.3 V would describe a different calculation.
Check What a Supply Error Does
Now change only the assumed supply. Let it be 3.267 V, which is 1% below 3.3 V. Keep the DAC ideal and leave its code at 2048.
The calculated output becomes 3.267 × 2048 / 4096 = 1.6335 V. That is 16.5 mV below the original 1.650 V target. Relative to the nominal 0.805664 mV step, the difference is 20.48 steps.
This is an arithmetic example of reference scaling. It is not an AD5622 tolerance specification or a prediction for a supplied lot. It shows why checking the bit count alone cannot establish an absolute-voltage result.
Write down whether the application needs an absolute voltage or a ratio to the same supply. Those are different requirements. Do not relax an absolute-voltage limit merely because a circuit can be described as supply-referenced.
Build an Error Budget for the Required Conditions
Analog Devices' precision DAC error-budget tutorial distinguishes reference, gain, offset, linearity and noise contributions. They describe different effects. One headline number does not stand in for all of them.
Item to Check | Question for the Design Review | Record Before Approval |
|---|---|---|
Supply/reference | What voltage reaches the DAC under the intended conditions? | Allowed variation, noise and measurement point |
Offset and gain | How do the applicable limits affect the chosen output range? | Grade, limit type and any calibration assumption |
Linearity | Does the stated limit cover the codes you will use? | Definition, code range and test conditions |
Output load and timing | What is connected, and when is the voltage evaluated? | Load, settling allowance and acceptance interval |
Temperature and measurement | What changes across the required operating conditions? | Temperature range and instrument uncertainty |
Convert the relevant terms to consistent units before comparing them with the allowed output error. Keep maximum limits separate from typical curves. Explain any statistical combination and its assumptions. A typical estimate should not be presented as a guaranteed worst-case result.
Use Calibration With a Defined Scope
A calibration plan needs more detail than “adjust until correct.” State the calibration points, temperature, supply, load and measurement method. Keep the resulting correction tied to the tested board and configuration.
Analog Devices' open-loop DAC calibration explanation describes compensating gain and offset. It also notes that INL and DNL specifications do not include those errors. Do not assume that correcting gain and offset removes every code-dependent deviation.
For an AD5622-based design, check that any corrected code remains within the available code range. Confirm the required outputs after correction. Do not claim the entire operating range is validated from one successful room-temperature reading.
Turn the Review Into a Clear Procurement Requirement
Keep the approved full manufacturer part number in the purchasing record. Record the required grade and package separately from the board-level voltage requirement. A supplier's confirmation of a part number is not a measurement of your assembled circuit.
If a component test is required, specify the codes, supply, load, temperature and allowed error. The component inspection-request guide explains how to define the requested work without treating an available test as completed.
Use the ideal calculation to set expectations. Use the applicable limits and a documented review to judge the design. Keep procurement confirmation, component evidence and circuit validation as separate records.
Frequently Asked Questions
Does 12-Bit Resolution Mean the Output Is Accurate to One Step?
No. It describes the code resolution. Absolute output error also depends on the reference supply, converter errors and application conditions. Compare a complete, documented error budget with the allowed voltage error for your design.
Does Code 4095 Produce Exactly 3.3 V?
With an ideal 3.3 V supply, the data-sheet equation gives 3.3 × 4095 / 4096, or approximately 3.29919 V. Real output also includes the applicable device and system errors. The calculation is not an exact-output guarantee.
Why Can the Output Change When the Digital Code Stays the Same?
The reference is derived from the supply input. A change in that supply changes the ideal output for a fixed code. Review the supply at the DAC, along with load, temperature and measurement conditions, before assigning a cause.
Can a Two-Point Calibration Remove Every Output Error?
Do not assume that it can. Gain and offset correction has a defined scope. Review remaining linearity, noise and condition-dependent effects. Confirm the required code range and operating conditions using a documented measurement plan.
What Should a Buyer Confirm About an Offered AD5622YKSZ-1REEL7?
Confirm the complete ordering code, required grade, package and applicable documents. Describe any requested component testing separately. Neither an order-code match nor a component report alone confirms that the part meets the voltage accuracy required in your assembled design.