Digital Pulse MIG Arc Welding Power Source Design Study Note

Literature Overview

This paper by Pang Qingle from Shandong University of Business and Management and Liaocheng University, published in the Transactions of the China Welding Institution (Vol. 31, No. 8, 2010, pp. 53–56), presents the design of a digital control system for pulse MIG welding power sources. The work was supported by the National Natural Science Foundation of China (Grant No. 50777040), the China Postdoctoral Science Foundation (Grant No. 20090461204), and the Shandong Province Postdoctoral Innovation Program (Grant No. 200903066). The paper addresses the limitations of analog and microcontroller-based control systems — poor flexibility, low control precision, and inadequate reliability — and proposes a digital control architecture based on a Digital Signal Controller (DSC) and a Microcontroller Unit (MCU).

Core Technical Concept

Pulse MIG welding is a highly effective process for thin-section welding, aluminum welding, and applications requiring low heat input with good penetration. The pulse waveform — characterized by peak current, base current, peak time, base time, and frequency — must be precisely controlled to achieve stable short-circuit-free droplet transfer. Traditional analog control systems use fixed hardware circuits to generate pulse waveforms, which limits flexibility and makes parameter adjustment cumbersome.

The proposed digital control system uses a two-tier architecture:

1. DSC (Digital Signal Controller): Handles real-time closed-loop control of arc voltage and welding current using a variable-parameter PI (Proportional-Integral) control method. The DSC provides the computational speed and precision needed for real-time control loops operating at kilohertz rates.
2. MCU (Microcontroller Unit): Runs an expert system that determines the PI control parameters based on the current welding state. The MCU also runs an embedded real-time operating system (μC/OS-II) with multiple tasks, including the expert system, parameter management, and communication interfaces.

System Architecture and Control Strategy

The control strategy is particularly interesting. The DSC executes the PI control loop at a high sampling rate (typically 10–20 kHz for welding power sources), ensuring fast dynamic response. The MCU, running at a lower rate, evaluates the welding state and adjusts the PI controller parameters accordingly. This hierarchical control architecture separates the concerns of fast dynamic response (handled by the DSC) and intelligent parameter optimization (handled by the MCU).

The variable-parameter PI control is a key innovation. Rather than using fixed PI gains, the controller gains are adjusted in real time based on the welding state. For example:

• During peak current periods, the PI gains may be adjusted to allow faster current rise.
• During base current periods, the gains may be adjusted for smoother current decay.
• During short-circuit events (if any), the gains may be adjusted to minimize spatter.

Pulse Parameter Optimization

The paper's approach to pulse parameter determination is noteworthy. In conventional pulse MIG welding, pulse parameters are set manually based on empirical charts and operator experience. The digital system described here enables automatic parameter optimization through the expert system, which can:

1. Identify the current welding state (peak, base, transition).
2. Determine the optimal PI control parameters for that state.
3. Adjust the pulse waveform in real time to maintain stable droplet transfer.

This approach has significant implications for welding quality. Stable droplet transfer is critical for weld quality — it prevents spatter, ensures uniform bead profile, and minimizes porosity. The digital control system's ability to maintain stable transfer across varying welding conditions (different joint geometries, different travel speeds, different material thicknesses) is a major advantage.

Engineering Implementation Considerations

Several practical considerations arise from this design:

The trimming and porting of μC/OS-II is a significant engineering task. The full operating system is too large for the MCU used in this application, so the kernel must be carefully trimmed to retain only the necessary tasks and services. This requires deep understanding of both the operating system internals and the welding control requirements.

Reflections and Insights

This paper represents an important milestone in the digitalization of welding power sources. The transition from analog to digital control is not merely a technological upgrade — it fundamentally changes what is possible in welding process control. Digital systems enable:

• Adaptive control that responds to changing welding conditions
• Automatic parameter optimization based on process feedback
• Data logging and traceability for quality documentation
• Remote monitoring and diagnostics

For pipe and fitting fabrication, digital pulse MIG welding is particularly valuable for:

• Thin-walled pipe welding where heat input must be minimized
• Dissimilar metal welding where precise heat control prevents cracking
• Welding of high-strength steels where HAZ properties are critical
• Production welding where consistency and repeatability are paramount

The hierarchical DSC-MCU architecture described here has become a standard approach in modern welding power source design. Contemporary welding systems employ similar architectures with even more powerful processors and more sophisticated control algorithms. However, the fundamental concept — separating fast real-time control from intelligent parameter management — remains sound and continues to guide the development of advanced welding power sources.

In conclusion, this literature documents a pioneering approach to digital welding power source design that has influenced the evolution of modern welding equipment, and its principles remain highly relevant for engineers working on advanced welding process development.