A. GENERAL DESCRIPTION

The HeNe laser is essentially an optical cavity consisting of a capillary tube made of glass with an inner diameter of about 1 mm, with a mirror at each end. Outside the capillary tube, a housing of glass or special metal is constructed to contain the working gases. The assembly is pumped to a high vacuum and small amounts of the named gases are let in at precisely measured pressures. One of the mirrors reflects virtually 100% of the desired wavelength and the other, about 99%. Thus, the output coupler will pass about 1% of any light generated within the cavity at the desired wavelength. This has two immediate implications: first, that the light exiting the cavity has made an average of 100 passes through the tube and second, that the light inside the cavity is about 100 times as bright as the exiting beam.

From this, it is seen that the mirrors must be precisely aligned. In fact, any sideways strain on either of the mirrors will likely adversely affect the laser output. In practice, just a push sideways with a pencil eraser will produce a visible change in output, and may actually stop the lasing action. This means that the laser must always be mounted so that there is no strain on the mirrors, even from the ballast resistor.

The fact that the light between the mirrors is 100 times as bright as that outside is sometimes used to advantage. Some lasers are built with no output mirror, but just a special window called a Brewster window. This is a glass window mounted at a precise angle called the Brewster angle, where the light passes through, now linearly polarized, with virtually no reflection, and thus practically no loss of power. The output coupler (mirror) is then positioned outside, some distance from the laser, and adjusted with three screws. Now, when the laser operates, the very powerful "inside" portion of the beam is outside, and may be used for such jobs as observing tiny particles of dust, since they are now clearly visible in the intense light.

Normally, the longer the laser cavity, the higher the output power, and the less the beam will diverge over distance.

The output of most HeNe's is randomly polarized, but if desired they may be ordered with a polarizer built in.

 In very simple terms and attempting not to muddy the waters with technical detail, singlemode (TEM 00) lasers produce a single, round spot with a Gaussian profile, brightest in the center and becoming less bright toward the edges. Multimode lasers produce what appears to be multiple beams arranged in a sort of structured pattern. For example, a TEM 01 beam will appear to be divided down the center into two separate beams; a TEM 11 beam will appear to be divided into four beams.

Multimode lasers normally produce more power than singlemode of the same length, but with the complex beam structure described above; the beam diameter is also larger.

Thus, if you need a Gaussian beam which will collimate or focus to the smallest spot, singlemode will do it. If you need maximum power and a larger spot is acceptable, multimode may be what you need.

The optical power output of a HeNe laser will change with changes in input current, as one might expect. The changes, however, are normally not in proportion or even in the same direction. That is, an increase in current might produce a decrease in light output.

In most HeNe's, the output power will at first increase with increased current. Above a certain level, though, output will decrease with increased current. See fig. 1, "P - I CURVE".

While these curves vary among laser types and sizes, the effort here is to show something typical of the general type of laser: small lasers in the one-half to one milliwatt range, larger lasers in the two to seven milliwatt range, and multimode lasers.

It will be seen from fig. 1 that all types start producing light when conduction starts, produce increasing power as current increases to a certain point, then produce less power after that point until, when a certain upper current is reached, the laser stops lasing altogether. As shown in fig. 1, the smaller lasers have a steeper slope, both increasing and decreasing, than the larger lasers. Normally, the recommended operating current is chosen to coincide with the maximum output power. Thus, we see that increasing the current to the laser will not necessarily result in more light from the laser. Generally, an operating current setting different from that recommended by the laser manufacturer is counterproductive.

D.1:  Definition: 

Optical noise, for our purposes, will be defined as any variation in laser output power from any cause.

D.2:  Noise Sources:

 Noise has a number of causes, some from the laser and some from the power supply. First, the laser: 

D.2a:  Modesweeping: 

This is a very low frequency variation in output power caused by variation in the length of the cavity from thermal effects. It appears as repeated cycles of gradual brightening and dimming of the output as the laser warms up or cools as, for example, from an occasional draft of air. It will go through many cycles of this while warming up only a few degrees. These fluctuations in power occur at a much faster rate when the laser first starts, slowing markedly as it warms up. The effect is usually less observable in a longer laser, and more pronounced in a shorter laser. In fact, if we try to build a laser shorter than about five inches, the beam will actually come in and go out as the modes pass through. This quirk has actually been overcome to a large degree in recent years. There are now lasers as short as 3 1/2 inches that do not exhibit it. 

D.2b Single Frequency Oscillation:

 At certain current levels, generally at least one milliamp above recommmended operating current, a laser will strike up a low level oscilllation, at a frequency of roughly one to three megahertz. This might modulate the beam power at a depth of up to twenty or thirty percent peak-to-peak in an extreme case, but generally just a few percent. It also modulates the current to the laser. Raising or lowering the current will greatly affect the strength of this oscillation. In most applications, this noise is not a problem from any point of view.

D.2c: Broadband Noise:

 If the input current is raised to still higher levels, we come into a zone where the laser starts to generate something like white noise, which may come and go as current is increased. This noise looks quite disorganized on an oscilloscope and is normally stronger than the single frequency noise. This noise, too, is normally not a problem in most applications. If recommended operating current is used, it should never be encountered.

D.2d: Noise from the Power Supply: 

The power supply creates noise on the output beam too. This noise is all due to variations in current through the laser.

The most important source of variation, or current ripple, is from the switching action of the power supply. Most modern laser power supplies make use of high frequency power conversion. This occurs at frequencies between 20 KHz and 110 KHz. It is difficult or impossible to filter all of this out at the power supply output. As the current increases and decreases, the laser output follows naturally along, increasing and decreasing. This noise at the laser output, though, is not proportional to the power supply noise. By looking at fig. 1, we see that a given current variation will not usually produce the same variation in laser output. In fact, depending on the individual type, HeNe's attenuate this current ripple by from three to ten times. Thus, if power supply ripple is 10% peak-to-peak, laser power ripple may well be as little as 1% p-p.

A separate effect of power supply ripple is that, if current ripple is strong enough to go below the laser's dropout current, dropout will occur and the whole system will become unstable, repeatedly going through the start-dropout cycle.

A secondary, and not so important, source of power supply noise is the AC line where the power source is the line. This noise occurs at 120 Hz, the frequency of the line after full-wave rectification, and may be incompletely controlled by the power supply's regulation. Generally, this type of noise is less than 2% p-p and is not a major concern. However, power supplies are available which eliminate this effect.

D.3:  Measurement Techniques: 

•  Measuring laser beam noise is done by placing a photodiode in the beam's path and connecting it to an oscilloscope. This process, though, contains enough pitfalls and difficulties that it really is beyond the scope of this manual.
• Measuring the electrical ripple or noise to the laser is done by placing a small resistor, usually 100 ohms, in series with the cathode of the laser and connecting an oscilloscope across it. See fig. 3. When doing this, be sure all connections are secure.
• 
  

D.4:  Is Noise a Problem? 

Noise from the laser or the power supply is in no way harmful to the laser or the power supply, and if the application is such that it does not require a constant power level, noise and ripple should not be a concern. An example of a noise-intolerant application might be a scientific particle measuring instrument. Power supplies with extremely low output noise for these applications are readily available at slightly higher cost. 

E.1:  Operating Voltage: 

The HeNe laser operates much like an old-fashioned gas regulator tube; when you put a current through it, the voltage across it remains more or less constant, regardless of the current. In practice, voltage actually declines somewhat with increased current, thus it is seen by the power supply as a complex impedance (not a simple resistance) and requires a ballast resistor to prevent it from simply being an oscillator, turning on and off rapidly at a rate of several tens of hertz. With an adequate ballast resistor, the operating voltage will increase slightly with increased current. Thus, we see that the laser / ballast combination is strictly in control of the operating voltage, keeping it fairly constant. The power supply must be designed, then, to comply with the voltage needs of the laser / ballast combination being used.

Operating voltages vary with several factors, among them the diameter of the capillary, the gas fill pressures, and the length of the laser. Voltages, including the ballast, will vary from around 1,200 volts for a tube 5 inches long to around 3,500 volts for one which is two feet long.

A common pitfall to watch for is specifying only the laser operating voltage when ordering the power supply; the ballast must be figured in, too. The laser manufacturers commonly publish only the laser operating voltage; they do not include the ballast unless they are selling a laser head which includes the ballast resistor.

E.2   Operating Current:  

Operating current, like voltage, depends on several factors. First, there is a minimum current below which ionization cannot be sustained, and the laser "drops out" of operation. Above this, as current is increased, there is a range where the laser output increases with current. This increase, however, is not proportional to current increase. See fig. 1. At still higher currents, power will typically decrease and the laser will begin to intermittently generate noise.

Recommended operating current normally is specified by the laser manufacturer with the following considerations: At or near peak output power; safely above the "dropout current"; below the levels where the noises begin to be generated; and below levels which will cause excessive heat and resultant reduction of service life.

E.3  Ballast Resistor: 

Taken by themselves, these lasers will behave as relaxation oscillators. To stabilize this effect, a ballast resistor is placed in series with the laser. The value must be high enough to stop the oscillation and low enough to avoid wasting electrical power and generating heat unnecessarily. Wattage must be appropriate; normally a 75 K, five watt resistor is used, except for the smaller lasers running at less current. A one or two watt resistor may be appropriate for these.

The ballast must always be placed at the anode end to be effective; HeNe's have a comparatively large cathode, whose capacitance reduces the effect of the ballast. Some configurations, however, seem to benefit from a small amount of cathode ballast in combination with the anode ballast.

In the interest of stability, the ballast must always be mounted within an inch or two of the anode. Any small amount of added capacitance on the low side of the ballast may cause the system to become unstable and start oscillating, in and out of dropout.

Important: In choosing the ballast resistor, bear in mind the following: Before the laser fires, there may be as much as 11,000 volts across it, with no current flowing. When conduction begins, laser voltage drops instantly to, for example, 2,000 volts. This means that, for a few microseconds, as much as 9,000 volts is dropped across the ballast resistor, until the power supply's output capacitors drain to operating level. This kind of abuse takes a mighty resistor. Generally, wirewound is the only kind that will stand up. Some smaller lasers, using only 6,500 start volts, can sneak through with carbon composition or experimentally proven carbon film types. Another caution: Some offshore resistor manufacturers advertise " wirewound" which turn out, in fact, to be metal oxide and will not stand up to a single start transient.

E.4  Dropout Current:

 Below a certain level of current, ionization is not sustainable and the laser stops operating. This is called "dropout current"; it may be reduced somewhat by increasing the ballast, but eventually, around 150 K ohms or so, a limit is reached where the current just is not enough to sustain ionization and the laser stops conducting, allowing the power supply to build up to start voltage, fire the laser and start another cycle. This will appear as a form of oscillation at around 10-20 Hz. This behavior is not good for either the laser or the power supply and should be stopped as soon as it is recognized.