4. Line-Commutated and AC–AC Converters

Introduction and context

Line-commutated converters form the classical interface between an alternating-current utility and a direct-current load. Before the widespread adoption of self-commutated devices, mercury-arc valves and thyristors provided the principal means of converting fixed-frequency AC into controllable DC and, in reverse operation, DC into AC. Their importance remains substantial: high-power HVDC links, electrochemical plants, large drives, controlled excitation systems, and utility-scale rectifier supplies still employ thyristor technology because of its high voltage and current capability, overload tolerance, and comparatively low conduction loss.

AC–AC converters extend this function without an intermediate DC link. They regulate the RMS value, frequency, or waveform of an AC supply for heating, lighting, motor drives, traction, and power-quality applications. The principal families are AC voltage controllers, cycloconverters, and matrix converters. Their performance is governed not only by the switching devices but also by source impedance, commutation overlap, displacement angle, harmonics, reactive power, and the nature of the load.

This chapter develops the operating principles and analytical models of single-phase and three-phase diode and thyristor converters, then relates them to practical AC–AC conversion. Ideal equations are introduced first; nonideal effects are then used to explain why measured waveforms and power factor differ from elementary predictions.

Learning objectives

After studying this chapter, the reader should be able to:

1. Distinguish diode, semi-controlled, and fully controlled line-commutated converters.

2. Derive average output voltage for important single-phase and three-phase rectifiers.

3. Explain firing angle, extinction angle, overlap, inversion, and regeneration.

4. Evaluate the effects of source inductance on voltage, current, power factor, and commutation.

5. Explain harmonics and distinguish displacement power factor from distortion power factor.

6. Select and compare AC voltage controllers, cycloconverters, and matrix converters.

7. Analyze practical converter choices for industrial applications and identify their limitations.

The line-commutation principle

A thyristor becomes conducting when it is forward biased and receives a gate pulse, but it cannot normally be turned off by its gate. In a line-commutated circuit, the AC source naturally transfers current from one device to another. When the source voltage reverses the conducting device's current, that device turns off provided its turn-off time is respected. Thus the supply frequency determines the basic commutation sequence.

For a sufficiently large smoothing inductance, the DC-side current may be treated as constant. This assumption is particularly useful for controlled bridges. The firing angle, alpha, is measured from the natural commutation instant—the point at which the corresponding diode would begin conduction. Delaying firing reduces the average rectified voltage. For a fully controlled bridge with continuous current,

V_d = V_do cos alpha,

where V_do is the maximum average output voltage at alpha = 0. If alpha is less than 90 degrees, power normally flows from AC to DC. For alpha greater than 90 degrees, the average bridge voltage is negative; if the DC source or machine can maintain current and a suitable extinction margin exists, power flows from the DC side back to the AC system. This is inverter or regenerative operation.

A practical inverter cannot operate merely by setting alpha above 90 degrees. The outgoing thyristors require a finite interval of reverse bias. The extinction angle, gamma, measures the available turn-off margin. Commutation failure can occur when an AC voltage disturbance, excessive current, or insufficient gamma prevents the outgoing devices from recovering their blocking capability.

Diode rectifiers and the natural commutation sequence

A single-phase half-wave diode rectifier conducts during the positive half-cycle for a resistive load. With an ideal diode and resistive load,

V_d = V_m/pi,

where V_m is the source-voltage peak. Its ripple is large, its transformer utilization is poor, and the DC component in a transformer winding can be undesirable. It is therefore mainly a pedagogical or low-power circuit.

A single-phase full-wave midpoint rectifier uses a centre-tapped transformer, whereas the bridge rectifier uses four diodes and does not require a centre tap. For ideal continuous resistive or highly inductive operation, the bridge average voltage is

V_d = 2V_m/pi = 0.9V_s,

where V_s is the RMS source voltage. The bridge reverses the negative half-cycle so that load voltage remains positive. With an inductive load, current persists beyond the zero crossing of source voltage, and freewheeling paths or source commutation determine the detailed waveform.

For a three-phase, six-pulse diode bridge supplied by balanced line-to-line voltages,

V_d = (3sqrt(2)/pi)V_LL approximately 1.35V_LL.

At any instant, the upper device connected to the most positive phase and the lower device connected to the most negative phase conduct. Each device conducts for 120 electrical degrees, while each pair conducts for 60 degrees in the ideal continuous-current case. The output ripple has six pulses per supply cycle, making filtering easier than in a single-phase converter. Twelve-pulse arrangements, formed using phase-shifted transformer secondaries, reduce characteristic low-order harmonics and are widely used in high-power systems.

Controlled rectification

Replacing bridge diodes by thyristors permits the average output voltage to be regulated. In a single-phase fully controlled bridge with continuous current,

V_d = (2V_m/pi)cos alpha.

At alpha = 0, the circuit has the same average voltage as an ideal diode bridge. As alpha increases, the output voltage falls and the fundamental source current becomes increasingly displaced from the source voltage. In a semi-controlled bridge, two thyristors and two diodes are used. Its output voltage remains nonnegative for an ordinary load and, with continuous current, is approximately

V_d = (V_m/pi)(1 + cos alpha).

Consequently, a semi-converter supports rectification but not inversion. It generally has better freewheeling behavior and lower reactive-power demand than a full converter at comparable positive output voltage, but it cannot provide regenerative braking by reversal of average DC voltage.

For a three-phase fully controlled bridge,

V_d = 1.35V_LL cos alpha.

The six thyristors are fired in sequence at 60-degree intervals, with a firing pulse pattern arranged so that one upper and one lower device conduct. In continuous current, each thyristor conducts for 120 degrees. A three-phase semi-converter has the ideal relation

V_d = 0.675V_LL(1 + cos alpha),

under the usual continuous-current assumptions. Three-phase converters are preferred at medium and high power because their output ripple is smaller, their transformer utilization is better, and their AC current waveform is more favorable than that of single-phase circuits.

The above equations assume a stiff source, negligible device drops, no overlap, and continuous load current. Discontinuous current changes the conduction intervals and makes the average voltage dependent on the load impedance and extinction of current. Analytical treatment must therefore begin with the actual conduction mode rather than applying a continuous-current formula automatically.

Source inductance and commutation overlap

Utility transformers, cables, and supply reactors possess leakage and intentional inductance. During commutation, current cannot transfer instantaneously from one valve to the next. For an interval mu, both the outgoing and incoming devices conduct. This interval is called commutation overlap.

Overlap reduces the average DC voltage because part of the source voltage is consumed in changing current between phases. For a three-phase fully controlled bridge with continuous current and approximately equal source inductance per phase, a commonly used model is

V_d = 1.35V_LL cos alpha - (3omega L_s/pi)I_d,

where omega is angular supply frequency, L_s is the effective commutating inductance, and I_d is DC current. The exact coefficient depends on how source inductance is defined, but the physical conclusions are general: greater current and inductance produce greater voltage reduction and overlap.

Overlap also modifies the AC current waveform, increases distortion, and reduces the maximum attainable inverter margin. In a controlled rectifier, a smoothing reactor makes DC current more continuous but increases the commutation voltage drop. Design is therefore a compromise between ripple reduction and dynamic and commutation performance.

Power factor, harmonics, and reactive power

A converter's power factor is not determined solely by the cosine of the firing angle. The total power factor may be written conceptually as

PF = displacement factor multiplied by distortion factor.

For a strongly continuous-current, ideal six-pulse fully controlled bridge, the displacement factor is approximately cos alpha. The AC current is a stepped waveform containing characteristic harmonics of order h = 6k plus or minus 1, such as the fifth, seventh, eleventh, and thirteenth. These harmonics reduce the distortion factor. Overlap and commutation notches further deteriorate the waveform.

A passive AC reactor, multipulse transformer, tuned filter, or active front end may be used to limit harmonic current. Standards and grid-connection rules commonly impose limits on current distortion, unbalance, flicker, and displacement or reactive power. A converter that satisfies its DC-voltage requirement but violates these limits is not a satisfactory engineering solution.

The input power factor has operational consequences. Poor power factor increases RMS current for a specified real power, causing additional transformer and cable losses and reducing system capacity. In industrial plants, reactive-power compensation may be required, but capacitors must be coordinated carefully because converter harmonics can create resonance. Detuned filters or active power-conditioning equipment are often safer than an unexamined capacitor installation.

Regeneration and dual converters

A reversible DC drive commonly uses two fully controlled bridges connected in anti-parallel. One bridge operates as a rectifier and the other as an inverter, allowing reversal of DC voltage and power flow. In noncirculating-current operation, only one bridge is enabled at a time, with interlocking to prevent a short circuit. In circulating-current operation, both bridges may be gated simultaneously through reactors that limit the circulating current; response is faster, but conduction losses are higher.

For ideal dual-converter operation, the firing angles are coordinated approximately by alpha_1 + alpha_2 = 180 degrees. Real systems require allowance for overlap and extinction margin. Regeneration is valuable in rolling mills, elevators, test stands, mine hoists, and electric traction, where mechanical energy must be returned to the supply rather than dissipated in a braking resistor. It requires a DC source or motor capable of delivering energy and an AC system capable of accepting it.

AC voltage controllers

An AC voltage controller applies portions of each supply cycle to an AC load while maintaining the supply frequency. In a single-phase controller, anti-parallel thyristors or a triac are used. For a resistive load, firing at alpha and extinguishing naturally at the current zero produces an RMS output voltage

V_o,rms = V_s [1 - alpha/pi + sin(2alpha)/(2pi)]^1/2,

for alpha between zero and pi. The load receives a controllable fraction of the source power, but the current is discontinuous and rich in low-order harmonics.

With an inductive load, current lags voltage and may continue after the source voltage reverses. The extinction angle is then determined by the load circuit, and inappropriate gating can cause asymmetric operation or excessive current. A freewheeling diode may provide a path for load current in some configurations, improving waveform behavior but changing the power-transfer mechanism.

Three-phase controllers are used for resistance heating, induction-motor soft starting, lamp control, and transformer regulation. Integral-cycle or burst firing, in which complete cycles are connected and disconnected, is often preferable for thermal loads because it produces less high-frequency switching activity, although it can create low-frequency power fluctuations and flicker. Phase-angle control gives smoother short-term regulation but greater harmonic pollution.

AC voltage controllers do not change frequency. They should not be confused with variable-frequency drives, in which a rectifier and inverter synthesize a new frequency. For motor applications, a voltage controller is suitable mainly for soft starting or limited-speed-control duties; it is generally inferior to a properly designed variable-frequency converter for broad speed-range control.

Cycloconverters: direct frequency conversion

A cycloconverter synthesizes a lower-frequency AC output directly from an AC supply by selectively connecting portions of the input voltage to the load. A conventional three-phase-to-three-phase bridge cycloconverter uses positive and negative thyristor converter groups. The positive group constructs positive portions of the output voltage; the negative group constructs negative portions. By controlling their firing angles according to a low-frequency reference, the average output follows the desired waveform.

The output frequency is normally substantially below the input frequency because the converter is line-commutated and needs suitable source-voltage segments for commutation. The output waveform contains ripple at supply-related frequencies, and the input current contains substantial harmonics and reactive components. Circulating-current cycloconverters require reactors and permit both groups to conduct; noncirculating-current designs alternate the groups and reduce losses but demand careful control and current-zero coordination.

Cycloconverters were historically important for very large, low-speed drives such as rolling mills, cement mills, mine winders, and ship propulsion. Their principal strength is high power with direct bidirectional frequency conversion and regenerative capability. Their disadvantages include bulky low-frequency magnetics or filters, poor input power factor, waveform distortion, and a restricted maximum output frequency. Modern voltage-source inverters have displaced many cycloconverters, although cycloconverter principles remain relevant in high-power and specialized drive systems.

Matrix converters

A matrix converter uses an array of bidirectional switches to connect any input phase to any output phase, without a DC-link capacitor or inductor. A three-phase matrix converter has nine bidirectional switch cells. The switching strategy selects input-output connections so that the output voltage has the required amplitude and frequency while the input current is shaped toward a sinusoid with controllable displacement.

Because an input phase must not be short-circuited and an inductive output current must not be interrupted without a path, commutation must be carefully managed. Safe four-step or related commutation methods account for the signs of input voltage and output current. Space-vector modulation is commonly used to coordinate voltage synthesis and input-current control.

The matrix converter offers bidirectional power flow, sinusoidal input current, controllable input power factor, and potentially high power density because it eliminates bulky DC-link energy storage. Its limitations include a maximum voltage transfer ratio of approximately 0.866 in the basic sinusoidal modulation scheme, sensitivity to supply disturbances, complex protection, and demanding switch commutation. Sparse matrix converters reduce the number of bidirectional power paths at the cost of additional control or waveform compromises. Matrix technology is attractive where regeneration, compactness, and power-quality performance justify its control complexity.

Applied case: regenerative control of a mine-hoist drive

Background

Consider an illustrative 6.6-kV mine hoist driven by a large DC motor. The hoist must accelerate a loaded cage, hold it at constant speed, decelerate it accurately, and return energy to the electrical system during lowering or braking. The supply is relatively stiff but has non-negligible transformer leakage inductance.

Situation and analysis

A single uncontrolled rectifier could supply positive motor voltage but would offer no controlled regeneration. A single fully controlled bridge would provide voltage control, yet reversal of torque would require a suitable four-quadrant arrangement. During lowering, the motor acts as a generator and maintains DC current while the converter must operate in inversion. The firing angle must therefore exceed 90 degrees while preserving adequate extinction margin. Transformer impedance causes overlap, reducing the available inverter voltage and increasing commutation stress. The stepped input current also creates fifth- and seventh-order harmonics.

Approach

A dual six-pulse thyristor converter is selected, with interlocked noncirculating-current operation and a smoothing reactor in the DC circuit. A phase-shifting transformer can be used in a twelve-pulse arrangement to cancel important characteristic harmonics. The control system regulates motor current inside a speed-control loop, limits the rate of firing-angle change, monitors extinction margin, and blocks inversion when an AC voltage dip threatens commutation failure. Harmonic filters and reactive-power compensation are coordinated with the transformer and system impedance rather than sized independently.

Outcome

The arrangement provides motoring and regenerative braking without a large braking resistor. Its performance depends on maintaining continuous current, sufficient AC voltage, and reliable firing-pulse synchronization. If very low-speed operation, rapid torque reversal, or near-unity input power factor becomes dominant, a self-commutated voltage-source drive may be preferred despite higher semiconductor-switching and control complexity.

Learning points

The case illustrates that converter selection is a system decision. Average voltage equations alone do not reveal commutation-failure risk, harmonic compliance, cooling requirements, or the value of regenerative energy. The source impedance, load energy flow, operating envelope, and grid constraints must be considered together.

Engineering limitations, risks, and good practice

Common analytical errors include using a continuous-current equation when current is discontinuous, confusing phase voltage with line voltage, ignoring the firing-angle reference, and treating cos alpha as the complete power factor. Device voltage drops, transformer impedance, overlap, control delay, unbalance, and temperature can materially alter performance.

Protection must address semiconductor overcurrent, short circuit, overvoltage, dv/dt triggering, di/dt stress, loss of commutation, cooling failure, and gate-drive malfunction. Fuses, current limiting reactors, snubbers, surge arresters, fast detection, and redundant synchronization may be required. In regenerative systems, a loss of AC voltage must cause an orderly transition because stored mechanical or DC energy cannot disappear instantaneously.

Professional design should include electromagnetic-compatibility assessment, harmonic studies, thermal verification, insulation coordination, earthing, and compliance with applicable grid and equipment standards. Control algorithms must fail safely: a missing synchronization signal should inhibit firing rather than generate arbitrary pulses. Maintenance procedures should recognize that a converter cabinet may contain hazardous stored energy even after the AC isolator is open.

Current developments

High-power thyristor converters remain competitive where switching frequency is low, power ratings are very high, and the utility can tolerate or mitigate reactive power and harmonics. Twelve-pulse and higher-pulse systems, active filters, and hybrid arrangements improve grid performance. Digital controllers now provide synchronized firing, adaptive current regulation, event recording, and predictive diagnostics.

Self-commutated converters based on IGBTs and wide-band-gap devices increasingly replace line-commutated systems where fast dynamics, independent reactive-power control, low harmonic distortion, or operation with weak grids is required. Nevertheless, wide-band-gap devices do not eliminate system-level issues: insulation stress, common-mode current, electromagnetic compatibility, thermal cycling, and protection coordination remain important. The central engineering trend is therefore not the disappearance of line commutation, but the deliberate matching of converter topology to power level, energy-flow direction, grid strength, waveform requirements, and lifecycle cost.

Chapter Summary

Line-commutated converters use the AC source to transfer current between valves. Diode bridges provide fixed rectification, while thyristor bridges regulate average DC voltage through firing-angle control. Single-phase and three-phase six-pulse equations describe ideal continuous-current operation, but source inductance introduces overlap, voltage loss, harmonic distortion, and reduced inverter margin. Fully controlled bridges can regenerate; semi-converters cannot normally reverse average voltage. Converter power factor combines displacement and distortion effects and must be evaluated with grid harmonics and reactive power.

AC voltage controllers regulate RMS voltage without changing frequency. Cycloconverters directly synthesize low-frequency AC and are suited to high-power, low-speed drives, while matrix converters use bidirectional switches to provide compact, bidirectional AC–AC conversion with controllable input current. Practical selection requires consideration of commutation, harmonics, protection, cooling, control complexity, grid strength, and the required direction and dynamics of power flow.

Key Terms

Firing angle: Delay, measured from natural commutation, at which a thyristor is triggered.

Line commutation: Turn-off of a conducting valve caused by reversal of source conditions.

Natural commutation: Commutation occurring at the natural AC voltage/current transition without forced switching.

Fully controlled bridge: Converter bridge in which all main valves are controllable devices, normally thyristors.

Semi-converter: Bridge containing controlled and uncontrolled valves, normally two thyristors and two diodes.

Continuous current: Load current that does not fall to zero during a supply cycle.

Discontinuous current: Load current that reaches zero for part of a cycle.

Overlap angle: Interval during which outgoing and incoming valves conduct simultaneously.

Extinction angle: Available reverse-bias interval for a thyristor before it must block again.

Inversion: Operation in which average power flows from the DC side to the AC side.

Regeneration: Recovery and return of energy from a load or machine to the supply.

Six-pulse converter: Three-phase bridge whose ideal output contains six voltage pulses per input cycle.

Characteristic harmonics: Harmonics naturally produced by a balanced converter, commonly of order 6k plus or minus 1 for a six-pulse bridge.

Displacement factor: Cosine of the phase angle between fundamental voltage and current.

Distortion factor: Ratio describing the reduction in power factor caused by current waveform distortion.

AC voltage controller: Converter that regulates AC RMS voltage while retaining the supply frequency.

Cycloconverter: Direct AC–AC converter that synthesizes a lower-frequency output from input-voltage segments.

Matrix converter: Direct AC–AC converter using an array of bidirectional switches and no substantial DC-link energy store.

Commutation failure: Failure of an outgoing thyristor to recover blocking capability during commutation.

Dual converter: Two controlled bridges arranged to permit reversible DC voltage and power flow.

Review Questions

Short Answer Questions

1. What is meant by line commutation, and why is a thyristor's turn-off dependent on the AC source?

2. State the ideal average-voltage equation for a three-phase fully controlled bridge with continuous current.

3. Distinguish a semi-converter from a fully controlled converter.

4. Why does source inductance produce commutation overlap?

5. What is the difference between displacement factor and distortion factor?

Long Answer Questions

1. Explain the operation of a three-phase six-pulse diode bridge and derive its average output voltage.

2. Discuss the operation of a single-phase fully controlled bridge for rectification and inversion.

3. Explain overlap in a controlled converter and discuss its effects on voltage, harmonics, and inverter operation.

4. Compare AC voltage controllers, cycloconverters, and matrix converters in terms of topology, frequency capability, power flow, and applications.

5. Explain the operating requirements and protection concerns of a dual converter used for regenerative DC-motor control.

Analytical / Application Questions

1. A three-phase fully controlled bridge is supplied by a 400-V line-to-line source and operates at a firing angle of 30 degrees with continuous current. Calculate the ideal average DC voltage using the six-pulse relation and state the assumptions involved.

2. An industrial heater requires RMS-voltage control but must not change supply frequency. Should phase-angle control or integral-cycle control be selected? Discuss waveform, flicker, and thermal-response considerations.

3. A regenerative hoist drive experiences frequent AC voltage dips and commutation failures. Identify the likely mechanisms and propose topology, control, filtering, and protection measures to improve reliability.

22. Multiple Choice Questions with Answers

Question 1

For an ideal three-phase fully controlled bridge operating with continuous current, the average DC voltage is proportional to:

A. sin alpha

B. cos alpha

C. 1 plus cos alpha

D. 1 minus cos alpha

Correct Answer: B. cos alpha

Question 2

A semi-converter normally cannot provide regenerative inversion because:

A. Its output current is always discontinuous

B. It contains no source inductance

C. Its average output voltage is not driven negative in normal operation

D. Its diodes cannot conduct at line frequency

Correct Answer: C. Its average output voltage is not driven negative in normal operation

Question 3

The principal effect of source inductance during commutation is to:

A. Make current transfer instantaneous

B. Create overlap and reduce average DC voltage

C. Eliminate characteristic harmonics

D. Increase the maximum inverter margin

Correct Answer: B. Create overlap and reduce average DC voltage

Question 4

For a six-pulse converter, the characteristic AC current harmonics ideally include:

A. The second and fourth harmonics only

B. The third and ninth harmonics only

C. The fifth, seventh, eleventh, and thirteenth harmonics

D. Only even harmonics

Correct Answer: C. The fifth, seventh, eleventh, and thirteenth harmonics

Question 5

The total power factor of a controlled rectifier is reduced by:

A. Displacement only

B. Distortion only

C. Both displacement and current distortion

D. The DC load resistance only

Correct Answer: C. Both displacement and current distortion

Question 6

An AC voltage controller differs from a variable-frequency drive because it:

A. Changes RMS voltage while retaining the supply frequency

B. Always contains a large DC-link capacitor

C. Can operate only with DC loads

D. Cannot use thyristors

Correct Answer: A. Changes RMS voltage while retaining the supply frequency

Question 7

A conventional cycloconverter is normally restricted to output frequencies below the input frequency mainly because:

A. Its switches cannot conduct at low frequency

B. Line commutation requires suitable portions of the input waveform

C. Its output transformer blocks high frequencies

D. Its output current must always be discontinuous

Correct Answer: B. Line commutation requires suitable portions of the input waveform

Question 8

A major advantage of a matrix converter is that it:

A. Uses a large energy-storage DC link

B. Provides bidirectional AC–AC conversion without a substantial DC-link store

C. Operates only at unity voltage transfer ratio

D. Requires no commutation strategy

Correct Answer: B. Provides bidirectional AC–AC conversion without a substantial DC-link store

Question 9

In inverter operation of a line-commutated bridge, adequate extinction angle is required to:

A. Increase diode forward drop

B. Ensure outgoing thyristors recover blocking capability

C. Force the DC current to zero every cycle

D. Suppress all input harmonics

Correct Answer: B. Ensure outgoing thyristors recover blocking capability

Question 10

In an illustrative regenerative hoist, a twelve-pulse arrangement is useful primarily because it:

A. Removes the need for firing control

B. Eliminates all reactive power

C. Reduces important low-order input harmonics through phase shifting

D. Prevents power flow from the DC side

Correct Answer: C. Reduces important low-order input harmonics through phase shifting