Proper selection of power electronic devices such as thyristors and rectifiers is of great significance to ensure the reliability of intermediate frequency furnaces and reduce equipment costs. The selection of components should take into account factors such as the environment in which they are used, the cooling method, the type of the line, the nature of the load, etc., and take into account the economics under the condition that the parameters of the selected components have margins.

In the following, only the selection of the thyristor element in the rectifier circuit and the single-IF intermediate frequency inverter circuit will be described. 

1 , the rectifier circuit device selection 

Power frequency rectification is one of the most commonly used areas for thyristor components. Component selection mainly considers its rated voltage and rated current.

 (1) Positive and negative peak voltages V DRM and V RRM of thyristor devices : 

 It should be 2-3 times the maximum peak voltage UM of the component , ie V DRM / RRM = (2-3) U M . The U M values corresponding to various rectification lines are shown in Table 1 .

 (2) The rated on-state current I T(AV) of the thyristor device : 

The I T (AV) value of the thyristor refers to the power frequency sine half-wave average value, which corresponds to an effective I TRMS =1.57I T(AV) . In order for the component to not be damaged by overheating during operation, the actual rms value of the flow through component should be equal to 1.57I T(AV) after multiplying by a safety factor of 1.5-2 . Assuming that the average load current of the rectifier circuit is I d and the current RMS value flowing through each device is K Id , the rated on-state current of the selected device should be: I T(AV) = (1.5-2)K Id / 1.57 =K fd *I d K fd is the calculation coefficient. For the control angle α =0 O , the K fd values under various rectifier circuits are shown in Table 1 .
                                 

Table 1 : Maximum peak voltage U M of the rectifying device and on-state average current calculation coefficient K fd

Note: U 2 is the effective value of the secondary phase voltage of the main circuit transformer ; the single- phase half-wave inductive load circuit has a freewheeling diode. 

Selecting the component I T (AV) value should also consider the component cooling method. Under normal circumstances, the rated current of the same component is lower than that of the water-cooled component; in the case of natural cooling, the rated current of the component is reduced to one-third of the standard cooling condition. 

2 , the choice of inverter components for intermediate frequency furnace 

Usually 400HZ Under the above operating conditions, should be considered KK device; frequency 4KHz or more, may be considered KA device. Here we mainly introduce the selection of components in the parallel inverter circuit ( see Figure 1 ) . 

 (1) The forward and reverse peak voltages of the components V DRM , V RRM The forward and reverse peak voltages of the components should be 1.5-2 times the actual maximum and reverse peak voltages . Assume that the DC input voltage of the inverter is U d and the power factor is cos ψ: V DRM/RRM = (1.5-2) π U d / (2cos ψ )  
                 

 (2) Rated on-state current element I T (the AV) when considering the working element at a high frequency, switching loss is significant, should be rated on-state current actually flowing through the element effective value I is 2-3 Considering that I T(AV) =(2-3)I/1.57 assumes that the DC input current of the inverter is I d , then the selected device I T(AV) is I T(AV) = (2-3 ) × I d /(1.57 )  
                

                 

(3) Shutdown time t q In the parallel inverter circuit, the turn-off time selection of the KK component is determined according to the pre-embedding time t f and the commutation time t r . Generally take: t q =(t f -t r ) ×             

 ( When the power factor is 0.8 when tf one cycle is about, TR press element di / dt is less than or equal to IOOA / [mu] S determined ) at higher frequencies can be by reducing the commutation time T R & lt , and Appropriately sacrifice the power factor to increase t f to select the component with the appropriate t q value