RF CO2 laser power supply types and principles

It has been more than 26 years since the first radio frequency (RF) waveguide laser was introduced in 1973. Initially, a coil was wound around the waveguide to achieve RF-excited waveguide laser emission. It was the first time that the superiority of low voltage excitation was demonstrated. At that time, there were still many imperfections. The main disadvantages were: uneven discharge, poor coupling efficiency, large coil inductance, which limited the increase in RF frequency and could only work below a few MHz.

Although it has been developed and used for more than 26 years, it is still in the stage of development and improvement. Its general development direction is to reduce costs, increase life, increase output power and efficiency, reduce size and weight, improve reliability, and improve various performance indicators to meet the needs of various uses.

The operating frequency of RFCO2 laser is 27～40MHz according to ISM regulations; its main classifications are as follows:

(1) By output method

1) Continuous output;

2) Pulse output - modulation frequency up to 1MHz;

3) Q-switch output - electro-optical Q-switching and acousto-optical Q-switching.

(2) According to the working conditions of the resonant cavity

1) Waveguide cavity - aperture D = 1 ~ 3mm;

2) Free space cavity - aperture D = 4 ~ 6mm.

(3) According to the polarity of excitation

1) Single phase;

2) Inversion.

(4) According to cavity structure

1) Single cavity;

2) Multiple cavities;

(a) Folding cavity: V-type - 2 folds; Z-type - 3 folds; X-type - 4 folds.

(b) Array cavity: short shoulder array; staggered array.

(c) Building block type: parallel connection - 2 cavities; triangle connection - 3 cavities.

3) Large area discharge

(a) Flat plate type, (b) Concentric ring type.

(5) Classification by uniform constant inductance distribution

1) Quasi-inductive resonance technology - used for low-capacitance laser heads;

2) Parallel distributed inductance resonance technology - for high capacitance lasers

head.

(6) Classification by resonant cavity material

1) Ceramic-metal hybrid type; 2) All-ceramic type; 3) All-metal type.

(7) According to cooling method

1) Air cooling; 2) Water cooling.

(8) By packaging method

1) Sealed type; 2) Flow type.

The materials of the resonant cavity are generally: metal-A1. Ceramics-BeO, BN, AIN, Al2O3, etc.

2 Technical conditions

Typical working gases for CO2 lasers contain: carbon dioxide, nitrogen, helium and xenon. Nitrogen and helium are good for the uniformity of the discharge. Xenon has a positive effect on the power and efficiency of RF discharge lasers. Adding 5% xenon can increase the power by 24%.

Usually for single-mode structure devices, the continuous (RF) excitation power injected per unit length is limited to 2-6 W/cm, and the conversion efficiency is 10%-20%. The general requirements for laser power supply are: stable and reliable, easy maintenance, high efficiency, small size and low cost.

The specific technical conditions are as follows:

Input parameters: AC input voltage 220V; AC input current 1.5A; switching power supply output DC voltage 30V; switching power supply output current 8A.

RF power supply: input power 160W; operating frequency 40MHz; output waveform sine wave; bandwidth △f±3MHz; efficiency 70%.

Modulator: Pulse modulation frequency 0~100kHz square wave continuously adjustable; duty cycle continuously adjustable; amplitude modulation degree 100%.

3 Principle Circuit

The RFCO2 laser power supply consists of five parts [1], as shown in Figure 1.

Figure 1 RF CO2 laser power supply block diagram

Part 1 in Figure 1 is a rectifier filter circuit, which uses full-wave bridge rectification and capacitor filtering to convert 220V AC into 311V smooth DC. Part 2 is a switching power supply, which converts 311V DC into 100kHz pulse current, and then converts it into 30V, 8A DC after filtering by capacitors and inductors. Part 3 includes an RF circuit consisting of an oscillator and an amplifier (as shown in the dotted box). The input DC is converted into 40MHz, 6W RF through a crystal oscillator. After amplification by an amplifier with a gain of 14.28dB, the output is 40MHz, 160W. Part 4 is a pulse modulator. Part 5 is a matching network.

This article will focus on Part 3 and Part 4. The circuit is shown in Figure 2. The oscillator consists of transistor V2, inductor coil L1, capacitor C5, C7, resistor R11, R12, quartz crystal oscillator G, etc. The crystal oscillator circuit generates a 6W, 40MHz sinusoidal oscillation wave, which drives the push-pull power amplifier through a 3:1 transmission line transformer T. The push-pull power amplifier consists of transistors V3, V4, inductors L3, L2, resistors R13, R14, R15, capacitors C9, C10 and transformer T to form a class D current switch push-pull amplifier, and the two transistors are turned on in turn. In order to pursue miniaturization, improving efficiency is the key, so the use of a class D current switch push-pull amplifier is an inevitable result. This can be clearly seen from the analysis of the following working process.

When the transistor is turned on, the fundamental component of the C-pole current is the largest, the midpoint voltage of the loop is also equal to the maximum value Umax, and the average voltage at the center point is equal to the power supply voltage. Therefore (when UCC≈30V),

It follows that:

Umax=(π/2)(UCC－UCS)+UCS(2)

The peak value of the AC voltage across the C pole loop is:

UCmax=2(Umax－UCS)=π(Ucc－Ucs)(3)

The amplitude of the fundamental component is: (2/π)ICC, so the amplitude of the fundamental frequency voltage generated by the circuit is:

UCmax = (2/π)ICCR (5)

Substituting (3) into (5) we get:

ICC=πUCmax/2R=(π2/2R)(UCC－UCS)(6)

Then the output power is:

P0=U2Cmax/2R=(π2/2R)(UCC－UCS)2(7)

DC output power:

PDC=ICC.UCC=(π2/2R)(UCC－UCS)UCC(8)

C pole dissipation power:

PC=PDC-P0=(π2/2R)(UCC-UCS)UCS(9)

The C-pole efficiency is obtained as follows:

ηC=P0/PDC=（UCC－UCS）/UCC（10）

It can be seen that the smaller the transistor saturation voltage drop UCS is, the higher the efficiency ηC is.

UCS→0, then ηC→100%. This is the advantage of the Class D current switch push-pull amplifier circuit. Therefore, when designing, you should pay attention to selecting power transistors with a low saturation voltage as much as possible.

The pulse operation is controlled by the modulator in Part 4 in Figure 1. The principle circuit of the modulator is shown in Figure 2. It is composed of IC1 and IC2 as the main body, forming an amplitude keying modulator, which is a linear modulator of digital signal amplitude modulation [3]. When working continuously, the S switch in Figure 2 is set to the OFF position. When working in pulse mode, the S switch is set to the ON position. The working process of pulse modulation is: the oscillation amplitude of the oscillator transistor V2 is controlled by a baseband signal of a rectangular pulse sequence. The purpose of modulation is achieved by controlling the start and stop of the oscillation circuit. The modulation frequency is controlled by potentiometer RP4, and the pulse width is controlled by RP7. Therefore, both the modulation frequency and the modulation pulse width can be continuously adjusted.

Figure 2 RF circuit schematic

Part 5 is the impedance matching network. The purpose of load impedance matching is to eliminate the reflection of the unmatched load. The method is to introduce reactive elements (capacitors, inductors or transmission lines). Artificially generate one or more reflected waves. Make it cancel each other out with the reflected wave generated by the original unmatched load. Make the input impedance of the laser and the output impedance of the RF power supply mutually conjugate complex. Matching networks are generally divided into two types. One is the lumped parameter matching network, and its main forms are L-type, T-type, π-type, etc. [3]. The main disadvantages of this matching network are: large insertion loss, high noise, and large volume. The other is a distributed parameter matching network, which is a 1/4 wavelength transmission line, which overcomes the shortcomings of the above-mentioned lumped parameter matching network. Its theoretical relationship is relatively simple. The voltage and current equations at any point on the transmission line can easily derive the following 1/4 wavelength (or odd multiples of 1/4 wavelength) impedance exchange method: Z0= (10)

Where Z1 is the impedance of the power supply output;

Z2——impedance of laser input;

Z0——1/4 characteristic impedance of the transmission line.

The 1/4 transmission line uses SYV-50-3 cable. One end of it is connected to the power supply and the other end is connected to the laser head. If the RF power supply is used as a building block structure, it can meet the needs of lasers such as 30W and 60W output lasers.

4 Conclusion

Finally, let's talk about the quasi-inductive resonance technology of the laser head. In order to make the voltage distribution of the input RF uniform along the length of the laser, a pair of inductors is added in parallel between the upper and lower electrodes of the resonant cavity. In this way, due to the compensation effect of the negative admittance of the inductor, the standing wave ratio along the length of the laser is greatly reduced, the mismatch angle is less than 9°, and the voltage unevenness is less than 3% according to theoretical calculation results.