We have all heard the new stories and read the market reports—the Industrial Internet of Things is changing the industry as we know it. The move towards industrial systems becoming connected intelligent devices is happening all around us, and Adesto is enabling these transformations.
In connected factories, there are three critical requirements:
- Large bandwidth communication channels
- Fast data processing
- Accurate decision intelligence (at the edge)
Decision intelligence at the device level, also called processing at the edge or edge computing, is important in an industrial environment where fast decisions are required, and delays related to communication latency with the central system are not acceptable. Processing at the edge optimizes system responsiveness, communication overhead, and security. In previous blogs, we have spoken about the advantages of Adesto’s SmartEdge platform.
If we pull back the lid on the SmartEdge platform and focus on the Analog Front End, the AFE is the workhorse. The AFE takes the information from all the sensors placed around the factory and converts the analog signals to the digital world. Central to this processing function is the analog-to-digital converter (ADC).
In an industrial environment, the main requirements of the ADCs used in AFEs are:
- Low latency – minimum latency is required to enable fast action based on a decision resulting from the acquired data.
- Low power consumption– many industrial systems are operated by batteries, and low power consumption is critical to extending the device lifetime. Sensor nodes can also be situated in difficult or even impossible to access locations, so replacing batteries can be extremely costly.
- High precision/high accuracy – accurate digital representation of the analog signal optimizes the system responsiveness and overall efficiency.
Achieving all these requirements in a single device can be difficult. Depending on the ADC architecture used, the latency, accuracy, and resolution can vary. The process node and the supply voltage can also have an effect. With power consumption, supply voltage, and thus technology node are key considerations. Selecting the correct ADC architecture for your AFE design is therefore not a trivial task. There are numerous architectures available with various advantages and disadvantages depending on your application requirement.
| Architecture | Advantages | Disadvantages |
|---|---|---|
| Flash ADC | Very high speed Low latency |
Not efficient for power and area |
| Delta-Sigma ADC | High resolution Low noise |
Long latency Not suitable for channel multiplexing |
| Pipeline ADC | Fast sampling rate Moderate power consumption |
Long latency |
| SAR ADC | Power and area efficient Low latency Suitable for channel multiplexing |
Less power efficient for very high resolution |
A new whitepaper from Adesto details the different types of ADCs, including flash ADCs, Delta-Sigma ADCs, Pipeline ADCs and Successive Approximation Register (SAR) ADCs. In the paper, we look at the trade-offs of each type of ADC in terms of power, performance, latency, and other key selection criteria.
Regardless of the requirements of your system and AFE, Adesto has a large family of silicon-proven SAR ADCs and other architecture devices. These devices are particularly suited to meet the challenging needs of the IoT market. The ADCs can be licensed by customers for the development of their own systems but are also used internally by Adesto in the creation of custom ASICs for the Industrial Internet of Things.