Two similar instruments

The Ultrasonic Dialog Flaw Detectors USD 15 SX and USD 15 SQ do not only differ in user-defined functions, such as wall thickness data logger and tolerance monitor in the SQ version, but also in transmission technology: the USD 15 SX has, as with nearly all KRAUTKRAMER ultrasonic flaw detectors, a field-proven Needle Pulser as opposed to the SQ version which is equipped with a Square Wave Pulser.

The US Air Force

The story behind the changed pulser technique was started by an invitation to tender issued by the US Air Force who required a universal instrument for special applications in the field of testing and maintenance of aircraft which, amongst other things, was obliged to have a square wave pulser. This instrument is used in the many US air force bases around the world and is now available as a universal instrument for all applications. In addition to the new functions, which have been included especially for the thickness measurement, we will only be dealing with the differences in pulser design in detail here because we think that this could be a knowledge gap for ultrasonic practicians. In addition to a technical understanding of the various methods of excitation, we will also discuss possible consequences for the practical field, i.e. special applications. Although the electronics used in the instruments is complicated, we will keep the explanations simple and use practical examples when going into detail.

Generation of ultrasonic pulses

Short electrical pulses of a few hundred volts are required in order to generate the ultrasonic signals in the probe elements (piezoceramics). Each probe element has a characteristic oscillation behaviour which is determined by different parameters. The most important are:
• Element material
• Thickness
• Mechanical damping a Electrical damping and filtering
• Adhesive bonding of the delay block

Needle pulse

Fig. 1: Ideal circuit for oscillation generation by capacitor discharge

The shape and amplitude of the sound pulse transmitted from the probe, Fig. 3, are dependent on the electrical excitation's parameters. With most of the ultrasonic instruments the electrical pulser pulses are generated by capacitor discharge, Fig. 1.

When the switch is closed, there are a few hundred volts on the crystal element which is directly excited into mechanical self-oscillation (resonance), Fig. 3. At the same time, the potential quickly decreases via the damping resistor connected in parallel, Fig. 2.

The discharge curve's time and therefore the period of the mechanical excitation of the crystal element is mainly dependent on the charging capacitor's capacity C and the damping resistor R_d. The characteristic shape of the discharge curve is referred to as a needle pulse.

In order to optimally use the many probes, C and R_d are switchable in some ultrasonic instruments. However, without this adjustment capability, each ultrasonic probe can be excited with a needle pulse. The user is therefore not burdened with additional tasks.

Fig. 2: Electrical signal at the crystal element

Fig. 3: Typical crystal element oscillation

Excitation and oscillation amplitude

The elements are always excited with the same amount of energy, irrespective of which probe is connected to an instrument with needle pulse. This means that under certain conditions, elements with large masses (large diameters and low frequencies) may not be sufficiently excited. This would be the same as hitting a large church bell with a drum stick: the bell will give a weak sound. The bell will oscillate strongly (loudest) when the mass of the hammer is matched to the bell mass and the excitation period is exactly half a wavelength of the bell's resonant frequency. This fact can be quickly proven with a simple, practical test using a pendulum.

The square wave pulser

With regard to our system this means that the duration to excitation should be set to the resonance frequency of the element so that the element oscillates optimally. This can be made electronically with the so-called square wave pulser, Fig. 4.

Fig. 4 : Square wave pulser (ideal state)

Fig. 5: Square wave pulse

The complete pulser voltage is fed to the element via an adjustable time. This pulse length now determines the oscillation characteristics of the element.

• Pulse length too short: The element does not oscillate at maximum amplitude; the excitation is too weak
• Correct pulse length: The element is exactly excited at resonance (optimum)
• Pulse length too long: Element oscillation is distorted and extends in time.

Advantages

The advantages of the square wave pulser can be clearly seen at low ultrasonic frequencies (less than 2 MHz): now we are able to achieve a measureable increase in sensitivity by specially set pulse lengths. However, it should be said that the operator has one more important function which he must correctly set when calibrating the instrument. A wrong adjustment at this point will influence the test sensitivity, resolution power and possibly also the gain linearity.

At higher frequencies there is hardly any difference between the needle pulse and the square wave pulse because normal excitation of the element with a needle pulse excites this into resonance.

Probe matching

The parameters shown in the table below are available for optimum probe matching to an ultrasonic flaw detector.

The main differences between both modes of excitation are shaded in a grey background. With the needle pulse pulser the excitation period can only be set in coarse steps using the charging capacity, whereas with the square wave pulser it is possible to achieve exact matching to the probe's resonance frequency. in addition to this, the excitation energy can be changed with the pulse strength. All settings at the ultrasonic flaw detector are better made in the high frequency mode of display. By doing this, the effect on the pulse shape can be assessed much better.

Function	Range	Description	Needle pulse	Square wave pulse
Damping	Pulser	Electrical resistance parallel to the element	Yes	Yes
Charging capacity (intensity)	Pulser	Determines the discharge curve	Yes	No
Pulse strength	Pulser	Preselection of the pulser voltage	No	Yes
Pulse lenght	Pulser	Duration of excitation	No	Yes
Frequency	Receiver	Filter for support of probe frequency	Yes	Yes

Parameters for probe matching with needle and square wave pulse

Fig. 7 a: Pulse distortion caused by wrong damping	Fig. 7 b: Optimum setting for damping
Fig. 8 a: Pulse distortion caused by wrong damping of a wide band probe	Fig. 8 b: Optimum setting for damping, narrow echo = highest resolution
Fig. 9 a: Optimum pulse length	Fig. 9 b: Completely wrong setting

Both systems react the same with regard to damping, Fig. 7 (narrow band probes) and Fig. 8 (wide band probes). A wrong setting (Fig. 7a) decreases the sensitivity (by about 14 dB in this case) and distorts the pulse (ringing). In Fig. 8a the pulse is unnecessarily extended so that the measurement resolution becomes worse.

An instrument having a square wave pulser is able to adjust the pulse shape very accurately using the pulser pulse width function. Fig. 9a shows the optimum setting. The pulse length has been set too large in Fig. 9b: the sensitivity is about 7 dB smaller and the pulse is unnecessarily extended.

The effect of a wrong setting is more extreme with wide band probes:

Fig. 10 a: Optimum pulse length	Fig. 10 b: "Double pulse" caused by a pulser pulse which is too wide

In addition to experimental optimization of the pulse length, a simple formula can be used for calculating this length with an instrument having a square wave pulser:

IB =	500			with: IB = Pulse width in ns         f = Probe nominal frequency in MHz
	-----
	f

Conclusions for the practical field

With a square wave pulser, the influence of better probe matching is favourable especially with applications using lower frequencies (< 4 MHz): a direct comparison of the gain reserve is shown below:

Test frequency