Vibration and Shock - Nature's Wrecking Crew
August 1966 Radio-Electronics

August 1966 Radio-Electronics [Table of Contents] Wax nostalgic about and learn from the history of early electronics. See articles from Radio-Electronics, published 1930-1988. All copyrights hereby acknowledged.

When I saw this 1966 Radio-Electronics magazine article entitled, "Vibration and Shock - Nature's Wrecking Crew," for some reason the first thing I thought of was "The Wrecking Crew," that anonymously played the music for a huge number of popular singers - mostly those without prominent bands of their own (Sonny & Cher, The Mammas and the Pappas, The 5th Dimension, Simon & Garfunkel, and others - see video) - during the 1960s and 1970s rock-and-roll era. ...but I digress. My introduction to the potential deleterious effects of vibration on electronics was in the 1970s, with airborne receivers and servos in my radio controlled model airplanes. Even though they were transistorized (vs. vacuum tube types from a couple decades earlier), vibration from glow fuel internal combustion engines (ICE) could wreak havoc with potentiometers in servos and solder joints everywhere, including battery packs. I remember seeing the control surfaces (rudder, elevator, ailerons) jitter as the servos driving them jittered at some engine RPM points. Usually, packing the receiver and battery pack with foam, and mounting the servos with rubber grommets under the mounting surfaces mitigated the issue. A lot of R/C airplanes have bit the dust due to vibration problems. The systems are much better nowadays, and a majority of models are powered by brushless motors, so most vibration sources are gone. My next memorable experience was with the mobile vacuum tube airport search and precision approach radars I worked on in the USAF. Nearly every time we packed up the trailers and convoyed them to a new location, upon setting back up and going operational we would find failed tubes and/or tubes that had vibrated partly out of their sockets. This article delves into methods and equipment used to investigate vibration and shock effects on electronics.

Vibration and Shock - Nature's Wrecking Crew

By William F. Kernin*

In controlled usage, vibration and shock are powerful and beneficial tools of nature. If they are uncontrolled - all havoc breaks loose. The random shock forces encountered by a ballistic missile, for example, exert tremendous loads on the electronic circuitry packed inside the "bird." Combined with ever-present vibration fields due to specific-impulse and frictional factors, these spurious effects do plenty to shake up - sometimes even destroy - the "black boxes."

To investigate these forces and design electronic equipment to survive them, the intensity and specific characteristics of the vibration and/or shock must be defined. That is where the vibration pickup and its associated electronic instrumentation play their roles. Let's examine typical equipment used to analyze these two very similar effects and determine how we can use it to de-fine and catalog our twin topics.

The most common device used to transform mechanical motion into an equivalent electrical signal is the piezoelectric crystal accelerometer. Fig. 1 shows two general-purpose devices suitable for many applications. Basically, a pickup of this type consists of a piezoelectric crystal which is mechanically preloaded, using a spring-and-mass arrangement. Any movement in the sensitive axis of the pickup causes a variation in its electrical output. The signal obtained from the pickup is fed to a companion amplifier by means of a low-noise, single-conductor shielded cable.

Various amplifiers with very high input impedance are available commercially for use with crystal accelerometers. One type uses a modified cathode-follower input stage to produce an input impedance variable from 100 to 1,000 megohms, depending on the circuit design and the user's selection of resistor values.

Fig. 2 shows a typical input stage and its associated amplifier circuitry. Because input grid resistor R_G is returned to ground through the cathode-load resistor R_L, the output-signal voltage is common to both input and output; thus a large degree of feedback is obtained. The feedback level is great enough to raise the input impedance from an expected value of 20 megohms, suggested by R_G, to a value of 200 megohms or higher. The actual impedance depends on the value of grid resistor R_G and the overall gain of the cathode follower - a value always less than 1. As gain approaches 1, the input impedance rises to a very high value. Very high input impedances are necessary because the crystal accelerometer is a high-impedance voltage-producing device. It must be lightly loaded to maintain its sensitivity and response to low frequencies.

A gain control and voltage amplifier follow the input stage, allowing low-level signals to be boosted to a usable value. The output of this circuit then feeds indicating instruments of somewhat lower impedances: an audio frequency vtvm, scope, or tape recorder, for example. Also, this output can drive an additional power amplifier that in turn drives a low-impedance recording galvanometer or other type of direct-writing recording device.

To determine the frequency response of a cathode-follower/amplifier and its crystal accelerometer - particularly at frequencies as low as 1 Hz or so - the setup shown in Fig. 3 can be used. In this circuit, a 10Ω resistor is connected in series with the pickup case. The combination feeds the input of the cathode-follower and amplifier, feeding a nominal load. A low-frequency oscillator is connected across the 10Ω resistor as shown, using a 510Ω series resistor. The dc scope is used first to set the input level across the Ion load, then to measure the relative output level for each frequency. Gain at 1 kHz is found first and used as a reference level. Other frequencies can then be checked as desired to obtain a complete response curve.

In a typical test setup used to measure environmental vibration levels, three accelerometers are used - one for each plane (horizontal, vertical and lateral).

The pickups can be stud-mounted or cemented directly to the component under test. In most cases, cement offers the best solution. Caulk Grip Cement - a dental product - is a good, general-purpose adhesive that sets sufficiently strong for vibration work in 30 minutes or so. For high-temperature work, as encountered on some points of a rocket engine, Armstrong type A-1 adhesive may be used. Fig. 4 shows a typical "black box" on which three accelerometers have been cemented prior to testing.

After mounting, each pickup is connected to its accelerometer amplifier through low-loss cable. A small rubber washer on the connector helps prevent shaking loose the cable. For high-humidity or altitude-chamber work, the connector is coated with a liquid silicone rubber (General Electric RTV-20) that cures at room temperature. When the test is over, the pliable rubber coating can be peeled off easily. As a final precaution, the miniature cable from the pickup is taped down close to the pickup and elsewhere along its length to prevent it from whipping around during testing.

Depending on what use is to be made of the test signal, the output of each amplifier can feed a scope, meter, tape-recorder channel, or recording oscillograph. Fig. 5 shows some equipment that may be used, singly or in combination.

Before a setup can provide meaningful data, the equipment must be calibrated, beginning by finding the pickup sensitivity. For most purposes, vibration and shock are measured in G's, one G being equivalent to normal gravitational force. The pickup to be checked is mounted on a small shake table which will provide known variable G forces. Its output is checked against that of a standard pickup - also mounted on the table - at different frequencies. The sensitivity of the test pickup is then expressed in millivolts rms per G peak.

Pickup sensitivity is affected by the characteristics of the cable connecting it to its amplifier, by the amplifier's input capacitance, and by the pickup's internal capacitance. Thus, complete calibration information may read like this for a typical unit:

E (sensitivity) = 20 mV rms/G pk

C (pickup) = 900 pF

C (cable) = 87 pF (3 ft of cable)

C (amp input) = 23 pF

If, when setting up a test, a different length of cable is used, the pickup sensitivity must be corrected using the following equation:

Thus, any convenient length of cable may be used and the calibration sensitivity corrected correspondingly, as long as the cable capacitance and amplifier input capacitance are known. The low-noise cables supplied with most pickups are normally tagged with their total capacitance values.

Continuing our hypothetical calibration: If for a particular test the expected level may reach ±10 G's, this suggests maximum possible signal of 10 x E_run (200 mV rms, using the sample calibration data). Using an audio oscillator, a signal of approximately 1 kHz is fed into the input of the accelerometer amplifier at the expected 10-G level (200 mV rms). The amplifier gain control is then adjusted for some convenient output value - 1 volt on a vtvm, 10 scale divisions on a scope, the maximum acceptable input for a tape recorder, or 2-in. deflection on the recording oscillograph. The system has then been calibrated so that a level of ±10 G's peak seen by the pickup will produce an easily read, known output. All pickup channels are calibrated this way, and the system is ready to go.

There are any number of applications for an accelerometer system in the measurement of vibration and shock levels. For example, the system can be used to obtain vibration information from apparatus under actual operating conditions, as in the recording of vibration levels on a rocket engine during hot-test firing. In actual tests of this type, multi-channel tape recorders record highly varied vibration data ranging from dc signals to high-frequency vibrations. A permanent record of the rocket test run is obtained on tape, and it can be reduced, analyzed or transformed into any form desired.

Vibration-Inducing Devices

After actual operating or environmental levels are determined, much vibration testing is concerned with applying G forces to components, black boxes, and systems at known levels to reveal any defects, or to quality the equipment.

A common instrument used for such testing is an electrodynamic vibration exciter. This type of shaker consists of a massive driver - a voice-coil-driven table-field-coil arrangement. It is powered by high-energy amplifiers or a controllable, variable-speed motor-generator set. Equipment to be tested is mounted on the shaker head. Standard pickups are then mounted on the device under test so that forces applied can be recorded and used later for reference or evaluation.

A more elaborate type of shaker is the random-vibration machine. Instead of steady frequencies, this instrument employs a band of frequencies with no definite repeatability characteristics-random noise - to drive a specially designed shaker head through powerful amplifiers. Failures often show up during random-noise tests that might never be discovered in steady frequency tests. (See "Big Noise", Radio-Electronics, Aug. 1963.)

Maintenance on vibration test equipment is basically a matter of common sense. The pickup should be kept clean to afford good mechanical contact; the connector must also be clean as possible. The miniature cables are tough, but no undue strains or sharp bends should be allowed, especially at the connector ends. Avoid excessive temperatures near the pickup. To shield a pickup from a nearby radiating heat source, several layers of shiny aluminum foil have proved quite helpful as a rough reflective cover. The cathode-follower/amplifier circuitry is not complex and can be serviced with standard signal-tracing techniques when trouble develops. To insure optimum operation and accuracy, the pickups and associated circuitry should be calibrated as often as possible; once a month is considered adequate.

While the typical industrial electronics technician may seldom see much vibration testing equipment, he should be aware of the basic principles and circuitry involved. Not only may this information come in handy when he does encounter such devices, but it may lead to further study and open new areas of new business through direct contact with those actively engaged in vibration testing.

* Instrumentation technician, Engineering Labs, Bell Aerosystems Co., Division of Bell Aerospace Corp., Buffalo, N.Y.