AC TIG Welding- Output Inverter Design Basics an-1045b.pdf

A PPLICATION N OTE

AN-1045

International Rectifier

233 Kansas Street El Segundo CA 90245 USA

AC TIG Welding: Output Inverter Design Basics

By A. Roccaro, R. Filippo, M. Salato

Topics Covered

Introduction

Balancing mechanisms

Application on TIG welding

Paralleling IR standard speed IGBTs

Full Bridge Inverter

Power circuit

Half-Bridge Inverter

Output inverter stages

Transient voltage and output rectifier

Multi-process welding

Advantages of paralleling IGBTs

Mounting instructions

Current and temperature unbalance

ESD and correct handling

Thermal runaway

Discrete devices approach

Conclusions

1. Introduction

A common use of IR Standard Speed IGBTs is in the output inverter stage of the AC TIG 1 welding

machines. IR has designed application specific modules and the aim of this document is to provide

information on how using them. Considerations and guidelines to connect several devices in parallel

are also provided for very high current applications.

IR IGBT Type Diode type V CES

(V)

I C @25ºC

(A)

I C @100ºC

(A)

Configuration

Part number

Package

Single switch without

freewheeling diode

GA200SA60S

Standard Speed

600

200

100

SOT-227

Half bridge without

freewheeling diode

GA200HS60S

Standard Speed

600

380

250

IAP

Half bridge with

freewheeling diode

GA100TS60SQ Standard Speed Fast QuietIR 600

220

200

IAP

Table 1 . IR recommended products for switching output stage of AC TIG welding machines

1 TIG, Tungsten Inert Gas, also called GTAW (Gas Tungsten Arc Welding)

www.irf.com

AN1045

2. Application on TIG welding

Unlike other welding machines, TIG ones are not suited to work with magnetic-type power supplies.

These deliver an alternating current of line frequency (50Hz or 60Hz depending on geographic area)

with slow reversals that hamper reignition of the arc at next half wave; auxiliary means can provide high

frequency ionizing voltage (HF), but often the instantaneous current is too low [3]. Nowadays, this

problem is avoided by using an output inverter stage connected to a regulated dc power supply.

The output inverter stage provides an ac square-wave (rapid zero crossover) that improves ac perfor-

mance enhancing arc reignition (deionization does not occur) to the extent that HF systems are

unnecessary [3]. As HF generates abnormal high electromagnetic emission, its use could cause

interference especially in electronic equipment as radio or television (EMI).

The frequency of this ac-wave is kept low (few hundreds Hz) for intrinsic application requirements.

Two commonly used configurations of an output inverter are discussed in the following sections.

2.1 Full Bridge Inverter

Consider the full-bridge inverter output stage

shown in figure 1. Every IGBT symbol in the

picture is equivalent to one or more IGBTs con-

nected in parallel. The inverter consists of two

legs: X 1 , X 2 form leg A, while X 3 , X 4 form leg B.

Commonly, the IGBTs are arranged to switch

in pairs, (X 1 , X 4 ) and (X 2 , X 3 ); the IGBTs in each

pair are turned off and on simultaneously. Also,

the pairs are switched in such a way that when

one of them is in its ON state, the other is

OFF: when X 1 and X 4 are ON, X 2 and X 3 are

OFF and vice versa (although in practice this

is not always true, and it will be addressed

later). In figure 1 are also included anti-parallel protection diodes connected to the IGBTs (D 1 , D 2 , D 3 ,

D 4 ). The use of these diodes is highly recommended since they provide a path to the reverse current to

ensure a low, safe V EC .

Figure 1 . Simplified circuit of the Full-Bridge output

Inverter.

The reverse current can be named a “reactive” current, because it would not be present with a purely

resistive load (the arc). Under some conditions the reverse current is neither present nor harmful, but

in other circumstances it is; since many times the operating conditions are totally unpredictable,

protection diodes should always be used.

www.irf.com

AN1045

The arc just resists to the current flow and has low reac-

tive components. The real threat is the inductive compo-

nent introduced by the wires connecting the inverter out-

put to the electrode and the work piece (inductance is

proportional to cable length). Hence, an inductor in se-

ries with the arc composes a more accurate model of

the load (figure 2b). When a pair of IGBTs is ON and the

current I L flows through the load, energy is stored in the

stray inductance of the wires. When the inverter switches,

the current I L changes direction, but the inductance re-

jects this sudden change and pumps the reverse cur-

rent. If protection diodes were not present, this current

would cause undesirable under stress operation of the

IGBTs and might even lead to exceed the maximum allowed V EC voltage damaging the device.

Figure 2. Models of the load: (a) basic and

(b) more accurate

Another consideration arises. All the IGBTs should not be OFF at once during switching, since the

power source feeding the inverter continuously supplies current. It follows that a pair of IGBTs should

be turned ON a time t

depends on turn-on and turn-off times

associated to the devices (normally is some hundreds of nanoseconds long).

before turning OFF the other pair; time t

As a result, for a finite short time, a cross-conduction of current takes place through the legs and the

IGBTs must withstand it. During this interval of time, the only harmless path for the reverse current is

through a “free-wheeling” diode and an IGBT; thus a further reason for using the diodes.

To reduce reverse current and cross conduction effects, most designers use the approach of “shaping”

the current waveform through an appropriate control of the primary inverter. It consists in reducing the

magnitude of the load current to few amperes before changing its polarity (switching the IGBTs). In this

case, relatively small clamps avoid unsafe voltage spikes during intervals in which all the IGBTs are

kept OFF. Current shaping also allows to build very accurate current profiles to optimize the welding

process.

When selecting the protection diodes, it is important to consider parameters such as the maximum

peak, average and RMS forward current, the dissipated power, the breakdown voltage, and the speed.

In most cases, the discrete diode 40EPF06 can be a good choice: it has 40A of average forward

current, its maximum peak reverse voltage is 600V, it has an ultra soft recovery, and it is optimized for

short reverse recovery time and low forward voltage. It also ensures stable and reliable operation in

severe temperature and power cycling conditions.

www.irf.com

AN1045

Another valid option can be the rectifier bridge SA55BA60 2

that contains, in one single device, all four freewheeling di-

odes arranged to allow a direct connection to “H-Bridge” in-

verters (figure 3). It has 55A of average forward current, its

maximum peak reverse voltage is 600V and it has fast recov-

ery time. Also, the SOT-227 package with an electrically iso-

lated base plate and its bridge configuration allow common

heat sinks usage, simplified mechanical designs, and com-

pact and rapid assemblies.

Where the “current-shaping” technique results in very low

switching current, even smaller diodes can be used, such as

the 10ETF06, which has the same characteristics of the

40EPF06 except for a lower average forward current of 10A.

Figure 3 . Application and pinout of

SA55BA60 in “H-Bridge”

inverters (load not shown)

2.2 Half-Bridge Inverter

Consider now the inverter circuit of figure 4. The number of diodes in the output rectifier is doubled, in

respect to the “Full-Bridge” configuration, but the number of equivalent IGBTs is now a half. This circuit

usually leads to system simplifications and cost savings.

The IGBTs are switched in such a way that when one of them is in ON, the other is OFF. During the

positive half wave, X 1 is ON and X 2 is OFF thus

D A and D B work as output rectifier while D C and

D D are OFF. Vice versa, during the negative

half wave, X 2 is ON and X 1 is OFF, thus D C and

D D work as output rectifier while D A and D B are

OFF.

The following considerations arise:

In this circuit, cross conduction must be

accurately avoided. X 1 and X 2 must never

be ON at the same time. This is accom-

plished by switching X 2 ON a time t

∆ after

X 1 is switched OFF. Again, this blanking

time depends on turn-on and turn-off times

associated to the devices.

Figure 4 . Simplified circuit of the Half-Bridge output

inverter

2 Now available on request as S1223

www.irf.com

AN1045

There is no freewheeling current. D 1 and D 2 are protection diodes which ensure that V EC is always well

below the safety limit, in particular during switching transients. Very low current diodes can therefore

be used, provided they are fast enough. The 8ETH06 or the 10ETF06 suit well to the purpose.

During the blanking time, the energy stored in the load and in the filter inductor L F has to be

somehow dissipated prior of the current reversal. RC-Diode clamps are often used. Sometimes,

the IGBT which is switching OFF is used in linearity as part of the clamp. Caution and accurate

design verification must be performed in such a case, since the IGBTs are designed for switching

operation and are not intended for linear use. The energy to be dissipated in the clamp can be

made very small by implementing the current shaping technique, resulting in additional cost sav-

ing.

2.3 Transient voltage and output rectifier

In most cases open circuit output voltages are around 80V, which is considerably higher than the arc

voltage. However, the transient voltage peaks might represent an issue, and for this reason the IGBTs

used must have a high enough collector-emitter breakdown voltage.

When an IGBT is turned off, it dissipates the stored energy in the circuit stray inductance, causing a

voltage overshoot across the device. The magnitude of this transient voltage is mostly determined by

the gate drive circuit, and is proportional to the stray inductance, the magnitude of the switched current

and its rate of fall at turn-off. Hence, performing the shaping of the load current reduces overshoots.

Semiconductor devices having blocking voltages of at least 400V are normally used. For the output

rectification, IR offers a wide series of 400V Ultrafast Recovery Epitaxial Diodes which provide a safe

margin against transient voltages. In particular, for a full modular approach, IR has developed the

UFB200FA40 which provides two independent, insulated diodes in SOT-227 package. More modules

can be paralleled together to reach higher current. The IRUD360CW40 (containing two common-

cathode diodes in non-insulated TO-244 package) is another interesting choice for very high current

application whenever the insulation can be easily provided by the designer.

IR IGBT modules for output inverters can instead withstand at least 600V of collector-emitter voltage

(V CES ) while in the OFF state.

3. Advantages of paralleling IGBTs

The main reason to parallel IGBTs is to increment the driven current. Sometimes, the only way to

achieve the desired current level is by using similar devices in parallel.

Paralleling helps to reduce the conduction losses and the junction to ambient thermal resistance.

www.irf.com

Plik z chomika:

Inne pliki z tego folderu:

Inne foldery tego chomika: