1、PHILOSOPHY OF VIBRATION TESTING OF RECEIVING TUBES FORMULATED BY JEDEC ELECTRON TUBE COUNCIL JEDEC PUBLICATION NO. 46 PRICE $30 EIA TEP14b b4 323L1b00 0008396 ELECTRONIC INDUSTRIES ASSOCIATION Engineering Department 11 WEST 42md S.“REET, NEW YORX 36, KY, Publirihed by EIA TEPL4b 64 m 3234b00 O008397
2、 5 m PHILOSOPHY OF VIBRATION TESTING OF RECEIVING TUEES During the past few years many new proposals concerning vibration testing of electronic and other components used in fixed and mobile gear have been issued. In some instances the payticular character- istic intended to be controlled by the prop
3、osed testing procedure has not been indicated nor can it be deduced from the test technique. In other instances the intent of the proposal has been indicated, but the test technique leaves some doubt concerning its adequacy for the intended purpose. The following comments are offered as a guide to a
4、ssist in the clarification of the purposes, and in the preparation, of test specifications for vibration testing. There are three prime purposes for which vibration testing is performed: To determine (1) the degree of looseness and/or flexibili-ty in a structure, (2) the frequency or frequencies at
5、which the unit and/or its component parts are resonant and the degree to which these frequencies are excited under the influence of a given applied force, and (3) the durability of a structure under a specified vibrational environment. These three main purposes are referenced in specifications as “V
6、ibration“, II Resonance“, and “Fatigue“ tests. Vibration testing to determine the looseness and/or flexibility a structure usually requires a driving mechanism and an aural, visual, or electrical readout system. It is important that the driving mechanism be capable of delivering the desired waveform
7、 with low distortion to the test specimen under the required conditions of frequency, force, and loading. Likewise, the specimen holding fixture shoul6 be designed to transmit the excitation to the unit under test with a minimum of distortion, attenuation, or amplification. The excitation delivered
8、to the test sample should be monitored at a point on or as near as possible to the test specimen. of Resonance search to determine that no part of the sample under test exhibits a resonance below a certain frequency or to determine at what frequencies the various parts of the sample exhibit resonanc
9、e phenomena requires the same basic equipment as the vibration test. However, resonance testing usually covers a wide frequency range and therefore places a premium on the elimination of distortion or u.ndesired excitations of the sample by the driving system and holding fixtures. Bollowing resonanc
10、e search a test may be performed to determine degradation in performance or durability during vibration at resonance. EIA TEP146 b4 - 3234b00 0008398 7 I -2- Fatigue testing is performed usually for extended periods of time to determine the incidence of malfunctions, catastrophic failure,s, or the d
11、-egradation rate of a product, mechanical and/or electrical, as a function of time. Such testing requires a driving mechanism but may or may not require a readout system as demanded by the test specification. To maintain a controlled process, the same freedom from distortion and undesired excitation
12、 should be maintained as for vibration testing. E3y using an electron tube as an example of a complex structure, the various aspects of the different vibration tests may be illustrated. If an electron tube is operated as an RC amplifier stage in a vibrational environment, various output signals may
13、be detected in the output circuit. These outputs are the result of relative motion between various elements of the mount causing changes in characteristtcs, such as Mu, Gm, Rp, and capacitances and if extensive enough, shorts and opens. The frequency components of the total vibration noise output ar
14、e deteymined by the mode and extent of motion of the various elements. The major frequency components and their causes are a,s Follows: 1, An output at the fundamental frequency or harmonics thereof of the exciting vibration may be generated by transducer action within the tube structure. Since the
15、separate elements of a tube have different masses and stiffnesses, their dis- placements under the combination of their own inertia and the driving force will be different, These differences in dis- placements give rise to relative motion between elements . thus changing the effective tube geometry
16、at a cyclic rate equal. to the driving frequency. At frequencies below the resonant frequency of the individual elements, the amplitude of this output is proportional to the clearance at the element suppo&s, Inversely proportional to the stiffness of the elements, and proportional to the exciting fo
17、rce, As the exciting force in “gs“ is increased, a level is reached at which the motion of the elements becomes equal to the clearance at the supports. When this occurs, any further Increase in drive causes shock excitation of the element each time it moves from one side of the support hole to the o
18、ther. These shocks wiJl excite the element into vibration at its own resonant frequency twice for each. cycle of the driving frequency, This generates the second major frequency components in the vibration noise output. It should be remembered that displacements of such elements as heaters and gette
19、rs may cuse no significant transducer action even at levels where rapid mechanical failure will occur. Therefore, electrical outputs alone may have to be supplemented by visual, aural or other criteria when durability is a factor under investigation. 1 i O O EIA TEP146 b4 3234600 0008399 9 -3- 2. Ou
20、tputs at the resonant frequencies and/or harmonics thereof of the various individual elements of the tube structure may be excited by primary shocks delivered to the structure, by internal shocks as indicated in Item 1 above, or by vibration of the structure at the resonant frequencies. It should be
21、 noted that the frequencies of the electrical outputs may be harmonics of the mechanical vibration of the individual elements depending on their mode of excitation, i.e., half-wave quarter-wave, etc. and the degree of concentricity of the scruc$ure. With shock excitation, the displacements f the res
22、onant elements are proportional to the exciting force to the level where motion limitations or permanent distortion of the elements occur. Immediately after the shock impulse, the displacements of the elements and therefore the electrical outputs start to fall at rates depending on the mechanical Q
23、and iamping factors of the various elements. With sustained vibration at the resonant frequencies, the amplitudes of the outputs quickly build up to a maximum level and remain there until something starts to fail. Sustained vibration at resonance usually resultsin extremely rapid degradation of elem
24、ent supports or the element itself, except in those structures which are intentionally designed to withstand vibration at resonance for extended periods of time. This rapid deterioration is caused by the extremely high forces resulting from the amplification of the applied driving force by the mecha
25、nical “&Is“ of the vibrating elements, The higher . the Q of an element, the easier it is to excite into resonance by an applzed vibration, but the range of frequencies which will excite it to the maximum is reduced proportionately. Il II II II 3. Random noise outputs a% random amplitudes covering a
26、 broad frequency band may be generated durng vibration tests. These outputs result from random closures of leakage paths, intermittent shorts, and discontinuities. The amplitude of I these outputs may or may not follow the vibration level applied. On normally good tubes, these outputs are relatively
27、 low. They are not limited to a specifc mode of excitation and may be excited by any vibration of sufficient amplitude to cause motion of the elements. A study of the signals available in the output of an electron tube when subjected to a vibration environment gives an indication of the test techniq
28、ues which have the best possibilities for indicating and controlling the various forms of vibration noise outputs. The degree of control for both amplitudes and frequencies is related to the application in which the tube is to be used- and by the basic design of the structure. For instance, the reso
29、nant. frequencies of a large power tube cannot be expected to fa11 in the same frequency range as those in a miniature or subminiature tube. - EIA TEPI4b 64 3234b00 0008i.100 I M -4- If a test is intended to show the degree of looseness or flexibility of the elements in a tube structure, it should b
30、e performed under conditions which cause the transducer action of the tube to predominate. This condition is obtained by vibration at a sine wave frequency well below the Lowest resonant frequency expected in the tube structure and at a “g“ level just below that which generates internal shocks. Sinc
31、e this g“ level is difficult to determine and varies considerably for the different elements in a structure, it is usually set at a level where some shocks are generated. Under this condition, the output noise consists of both that due to transducer action and those at the resonances which are easie
32、st to excite. beyond the level at which shock excited resonances begin to appear, the output signals due to resonances swamp those due to transducer action, and the test becomes a measure of the “Q“ and damping of the resonant elements rather than a measure of the looseness of the structure, When th
33、is occurs, control over how well the tube is constructed is partially lost since a very loose structure may have lower ttQtt and more highly damped resonances than one tightly assembled. Zf the excitation is increased Resonance searches to determine the resonant frequencies of the parts of a tube st
34、ructure are performed, generally, by sweep frequency vibration over a broad range of frequencies and continuous observation or recording of the output from the tube under test. 4 The frequency of resonance is determined by the physical constants (mass, length, etc.) of the element and explains why,
35、in general, the larger tubes may be expected to have lower resonant frequencies than the miniatures, Resonance In a tube may be defined as: which an element may be excited and sustained by a very low applied force of the same frequency or subharmonics thereof,“ The frequency of this vibration is usu
36、ally determined by the mechanical properties of the element and its mode of support. Another definition of a resonance may be stated based on the output of the tube. It is: “A resonance is the vibration of an element which generates an output at a fixed frequency one or more times as the frequency o
37、f the exciting force is swept across a band of frequencies“, rapid rise and fall in amplitude of the output as the driving excitation sweeps through the resonance frequency, “A state of vibration into Usually a resonance is indicated by a relatively To reduce the response of the tube by transducer a
38、ction and to reduce the excitation of resonant frequencies by high order harmonics of the driving frequency, the driving force should be held to a low G level and as nearly since wave as possible. If these conditions are maintained over the sweep frequency range, a higher percentage of the resonance
39、 outputs will occur as the driving fa?equ.ency sweeps by the fundamental resonance, and less will appear which are excited by a subharmonics of the true resonance frequency, II II a EIA TEPL4b 64 E 3234b00 000840L 3 m -5- The output of a tube under resonance test should be monitored or recorded by f
40、requency selective metering. This is necessary since many high “Q“ resonances may be excited by extremely low levels of energy such as in the harmonics of the driving sfgnal and give a resonance indication which is related to but not at the driving frequency. Thus, monitoring or recording the total
41、output of the tube versus the driving frequency as commonly done in sweep frequency vibration testing is not valid for use as a resonance test. Since it is impossible to eliminate all resonances from conventional tube structures but it is possible to shift their frequencies and somewhat control thei
42、r excitation sensitivities, the prime use of aresonance test is to insure no resonances below a specified frequency or within certain bands of frequencies which will be excited by a specified level of excitation. Most current specif$- cations do not place a “g“ level on the excitation for resonance
43、testing and therefore are almost meaningless for use. Under sweep frequency resonance testing conditions, the tube output at a resonance is seldom directly proportional to the driving foree. It is usually a function of the motion of the resonant element, providing the resulting motion does not tend
44、to exceed the clearance at the supports or does not cause a short to another element. If the resulting motion tends to exceed the clearance at the supports or causes a short to another element the output at resonance will exhibit saturation but additional outputs at other frequencies may be excited
45、by the internal shocks resulting fpom the excess motion. Since excessive forces may be generated within the structure a% resonance, the time of sweep should be of short duration, It should permit a stabilized output at the resonances but not dweI1 on any resonance for more time than is required for
46、stabilization of the output to prevent unwanted tube damage, Fatigue testing to determine the durability or degradation rate of a tube may be performed adequately under either fixed or variable frequency vibration. Under fixed frequency vibration, the frequency should be below the lowest resonance a
47、nd of such a force level that elements are driven to the point where internal shocks excite the major resonances. Under sweep frequency vibration, the driving force required is much lower since more damaging effects will be developed in the structure as the drive sweeps through the resonance frequen
48、cies and will generate these damaging forces for longer periods at the saturation levels than occurs under the fixed frequency condition. EIA TEPL4b b4 3234b00 0008402 5 -6- Y I Fatigue testing by vibration at a predetermined resonance frequency while used for some products is not considered to be a
49、 well-controlled process for receiving tubes. The lack of contano1 results from several conditions which arise within. the tube structure : (I 1. The application of a fixed “gl excitation is no measure of the destructive forces acting within the structure since the total force is related to the Q of the resonant element. II II 2, As soon as slight damage to an element or its support occursI both the “Q“ and frequency of the resonance change value, These changes may be great enough that the resonance is no longer excited by the applied driving frequency. 3. Vibration at reso
