Volpiano (Italy)
December 13, 2024
Testing ADC/DAC Converter ICs: Ensuring Reliable Data Conversion
Today, a wide range of applications depends on Digital-to-Analog Converters (DACs) and Analog-to-Digital Converters (ADCs). These devices are crucial in signal processing, as they bridge the gap between digital and analog systems. By enabling the interaction of digital circuits with analog components, ADCs and DACs play a pivotal role in areas such as audio processing, telecommunications, data acquisition systems, and more.
The Role of Analog-to-Digital Converters and Digital-to-Analog Converters
Analog-to-Digital Converters (ADCs) are essential building blocks in modern data acquisition systems (DAQ or DAS systems). They convert conditioned analog signals into digital data streams, enabling data acquisition systems to process, display, store, and analyze this information. High precision in this conversion process is critical, as even slight inaccuracies can impact downstream analysis and processing.
On the other hand, Digital-to-Analog Converters (DACs) are integral to applications like digital audio storage, streaming, and transmission. DACs take digital data and convert it back to an analog signal, ensuring high fidelity in audio reproduction and other analog-output applications.
Both ADCs and DACs require testing to ensure optimal performance, with specific focus on parameters like maximum sample rate, bit resolution, total harmonic distortion (THD), noise, SNR, INL, DNL, ENOB, and jitter.
Key Testing Methods for ADCs
Testing ADC and DAC devices involves multiple techniques, with specific methodologies tailored to the unique functions of each converter.
The primary tests vary based on the target application, whether it’s video processing, imaging, telecommunications, control systems, or audio processing. The following table summarizes the most common tests for each application:
Static Testing for ADC
INL and DNL
ADC testing often involves measuring key performance metrics, such as:
- Offset error
- Gain error
- Differential Non-Linearity (DNL)
- Integral Non-Linearity (INL)
- Missing codes
To test DNL and INL, a signal is applied to the input of the ADC, and the output code occurences are analyzed. In contrast to DAC testing, where digital codes are applied and the corresponding analog output is measured using a precise voltmeter, ADC testing requires identifying “decision levels”—the exact input voltages at code boundaries.
Histogram Test with a Linear Ramp:
In this case, a linear ramp signal is applied, and the number of occurrences (or hits) for each output code is analyzed. Ideally, each code should appear equally. If a code occurs more frequently than others, it indicates a wider step and a positive DNL. Conversely, fewer occurrences of a code suggest a smaller step and negative DNL. Once the DNL is evaluated, the INL results from the cumulative sum of the DNL values.
Also known as the code density test, the linear ramp histogram test is the most used method for testing ADC static parameters.
Histogram Test with Sinusoidal Input
The histogram method uses a sinewave signal as input to an ADC. Compared to other signal forms like a linear ramp, generating a pure sinewave is often more straightforward. However, a sinewave has uneven voltage distribution, with more voltage steps concentrated near the lower and upper voltage ranges.
In this method, the output of the ADC is analyzed to assess the converter’s performance across various voltage levels. For DACs, high-precision digital channels are combined with a low-noise sinewave generator to assess performance across these voltage ranges, ensuring minimal distortion and noise.
Dynamic Testing for ADCs
Noise Sources in ADC Testing
Accurate ADC testing must account for various noise sources, as noise can significantly degrade the performance of data converters. The three primary noise sources include:
- Jitter on Digital Signals: Jitter introduces errors in the acquisition instant, leading to inaccuracies in the acquired signal. Minimizing jitter improves the signal-to-noise ratio (SNR).
- Waveform Generator Noise: The quality of the signal generated for testing purposes directly affects the results. Test equipment must provide high SNR, exceeding that of the device under test (DUT), to ensure reliable results.
- Noise in Voltage References and Power Supplies: The noise in an ADC’s voltage reference (Vref) or power supplies can translate into output noise and errors in offset and gain. For more accurate testing, external voltage references are recommended, and power supply noise must be carefully controlled through parameters such as power supply rejection ratio (PSRR).
Testing DACs: Simpler, Yet Precise
Compared to ADC testing, DAC testing is generally less demanding. The process involves applying a series of digital codes to the DAC and using a high-precision voltmeter (DVM) to measure the corresponding analog output. This allows for straightforward DNL (Differential Non-Linearity) and INL (Integral Non-Linearity) measurements. However, while DAC testing is less complex, it still requires a highly accurate digitizer to ensure precise results.
In cases where the digitizer’s accuracy is insufficient, additional testing strategies can be employed to improve measurement precision:
- Pedestal Test: This method involves subtracting a known pedestal voltage from the DAC output, improving the accuracy of small signal measurements.
- Bucking Source Differential Amplifier: This technique uses a differential amplifier to cancel out noise or unwanted signal components, further enhancing the accuracy of DNL and INL measurements.
These methods allow for more precise measurements, ensuring that even minimal inaccuracies in DAC performance are detected and corrected.
Pedestal Test
Pedestal Test can enhance precision in DAC testing when higher accuracy is required. Instead of relying solely on the digitizer, the ramp generator of the digital processing instrument can be used as a pedestal for the digitizer via an internal connection. The ramp is first pre-characterized by a highly accurate system voltmeter, ensuring that its properties are well understood. This method allows the digitizer to operate within a smaller range, significantly improving its measurement resolution and enabling more precise testing of the DAC output.
Bucking Source Differential Amplifier
For cases where even the pedestal configuration does not provide the required precision, the Bucking Source Differential Amplifier technique can be employed. In this method, a high-stability voltage source acts as a “bucking” voltage applied against the DAC under test. The output points of the bucking source are pre-characterized by a highly accurate system voltmeter, such as the HP3458A, to ensure precise reference values.
The bucking source and the DAC form the differential inputs to a high-stability, low-drift Programmable Gain Amplifier (PGA). The PGA amplifies the difference between the DAC’s output and the bucking source. This approach isolates the small differences between the expected and actual output, enabling highly accurate measurement of the DAC’s performance. Point-by-point differences are used to calculate DNL (Differential Non-Linearity) by measuring the LSB step sizes, and these results are then integrated to determine INL (Integral Non-Linearity).
SPEA’s Testing Solutions for ADCs and DACs
Testing ADC and DAC devices requires sophisticated equipment capable of providing both high-quality digital signals and precise analog sources. SPEA’s DOT testers offer the perfect blend of accuracy and performance. The combination of high-accuracy digital channels, high signal-to-noise ratio, and low total harmonic distortion makes SPEA equipment ideal for comprehensive testing of both ADCs and DACs.
For DAC testing, SPEA instrumentation provides highly accurate analog measurements, enabling precise performance assessment. For ADCs, SPEA’s testers offer high-quality digital signals with low jitter, combined with analog sources that feature high spectral purity and SNR, crucial for conducting linearity tests and noise analysis.
