High-Speed Stroboscope for Accelerometer Calibration

  • Published on

  • View

  • Download


  • HighSpeed Stroboscope for Accelerometer CalibrationPeter G. Sulzer, Ernest R. Smith, and Seymour Edelman Citation: Review of Scientific Instruments 25, 837 (1954); doi: 10.1063/1.1771189 View online: http://dx.doi.org/10.1063/1.1771189 View Table of Contents: http://scitation.aip.org/content/aip/journal/rsi/25/8?ver=pdfcov Published by the AIP Publishing Articles you may be interested in Alternative calibration techniques for high-speed pyrometers in shock experiments Rev. Sci. Instrum. 76, 013106 (2005); 10.1063/1.1833331 HighSpeed Bolometer Rev. Sci. Instrum. 31, 115 (1960); 10.1063/1.1716904 Highspeed rotation Phys. Today 12, 20 (1959); 10.1063/1.3060885 HighSpeed Photography Phys. Today 9, 32 (1956); 10.1063/1.3059894 An Inexpensive Stroboscope for High Speed Photography Rev. Sci. Instrum. 14, 273 (1943); 10.1063/1.1770186

    This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions.

    Downloaded to IP: On: Sun, 21 Dec 2014 01:21:15



    This device is similar in some respects to a boiler sweeper patented by W. T. Levi, of Charleston, West Virginia, in 1886 (Patent No. 343 769).

    1 Redhead, LeCaine, and Henderson, Can. J. Research A28, 73 (1950). 2 O. Retzloff, Rev. Sci. Instr. 20, 324 (1949). * Machining tolerances-plus or minus 0.001 inch. Bronze spheres up to

    4 inches in diameter can be obtained from the Detroit Ball Bearing Com~ pany.

    'R. R. Wilson, Rev. Sci. Instr. 12,91 (1941).

    High-Speed Stroboscope for Accelerometer Calibration *

    PETER G. SULZER, ERNEST R. SMITH, AND SEYMOUR EDELMAN National Bureau of Standards, Washington, D. C.

    (Received March 12, 1954)

    A N accelerometer can be calibrated on an absolute basis by subjecting it to a simple harmonic motion of known ampli-tude and frequency, and measuring the resulting output voltage. A linear sinusoidal displacement is comparatively easy to obtain, but there is difficulty in measuring the magnitude of the displace-ment, which becomes very small with moderate accelerations at high frequencies.

    At low frequencies commercially available stroboscopic light sources are satisfactory to slow the apparent motion. It is then possible to set the movable hairline of a filar micrometer eyepiece to the extremes of the motion and determine the amplitude from a calibrated micrometer screw.

    VI 6SN7 -PtlASE


    40mId 1"50,

    47K IW

    At frequencies above about 800 cps this method becomes un-satisfactory because the duration of the flash is so long that the image seen in the microscope is fuzzy and the illumination is in-sufficient because a flash occurs only once for several excursions of the accelerometer.

    The system described in this paper removes the difficulties described in the previous paragraph by making the duration of the flash proportional to the length of the cycle and a very small part of the cycle and also by producing one flash. for each cycle. Ease of measurement is improved by stopping the apparent mo-tion so that the hairline can be set as carefully as desired.

    In using the system described here, the signal from the oscillator which drives the accelerometer is also sent through the circuit which flashes the light. The phase shifter is adjusted to flash the light just as the accelerometer reaches one extreme of its motion.

    At this point an observer at the microscope sees a well illumi-nated stationary image. The usual target is a galvanometer mirror which has small imperfections in its surface. The hairline of the micrometer eyepiece is carefully set to one of these imper-fections. The toggle switch is then used to shift the phase of tDe flash by 180 0 and the observer sees the whole image shift in the field of view by a distance which is proportional to the motion of the accelerometer. The hairline is again set carefully to the same imperfection as before. From the difference in the reading of the micrometer screw and the calibration of the micrometer scre~, the amplitude of the motion is calculated.

    The electronic part of the system consists of a phase shifter, pulse generator, and associated amplifiers and power supplies. Referring to the schematic diagram (Fig. 1), an input amplifier VIA drives a phase inverter VIR. A resistance-capacitance constant-am-plitude circuit provides a phase range of nearly 90 degrees, with a

    V4 ~


    Vs ~

    AMPLFr_E_R _____ -o + 1500,




    50K lOW

    470~ IW

    ~----~----~c-----_o+ISOO' 8mfd lOOK 1000. 20 W

    8 mId lOOK 1000, 20W

    FIG. 1.


    OR NE-2 1125 WATT

    Tl,n STANCOR P6012

    ~-O-35011' AT 90ma 6.3 y CENTER TAP AT 3.5A S. CENTER TAP AT 2A

    T2.13,T4 THORDARSON T21F03

    S. AT 3A SEC. INSULATION .1600.


    RECT. FEDERAL 1002. 65 m. LI. L2 HALLOORSON C 5013

    10 H AT 60 .... 27511 de

    This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions.

    Downloaded to IP: On: Sun, 21 Dec 2014 01:21:15


    double-pole double-throw switch to add 180 degrees. Capacitors are switched to permit coverage of the frequency range from 50 to 10000 cycles per second. Consequently it is possible to obtain almost complete phase control at all frequencies with but small gaps around +90 degrees and -90 degrees. To eliminate gaps a small RC phase shift may be added at the input.

    A voltage amplifier V2 and cathode follower Va provide suffi-cient excitation for an untuned class-C amplifier V 4A. This stage developes pulses whose duty cycle can be varied from 180 degrees to less than 30 degrees by changing the input voltage to VlA. An important feature of this circuit is that the duty cycle is independent of frequency, and therefore constant illumination is obtained at all frequencies below the upper limit of the lamp. The lamp is excited by a voltage developed across the plate load of Vs, which is normally biased to plate-current cutoff.

    The power supplies are shown for the benefit of those who might want to duplicate the equipment. A variac is used on the input of the high voltage supply to control its output. Commer-cially available helium gas spectrum tubes provide satisfactory illumination.

    In making the preliminary adjustments the source, which should supply one volt or more, should be set to the desired frequency. The voltage across Ra (which has been adjusted for zero voltage at the grid of V'B) can be observed with an oscilloscope as Rl is adjusted to provide the desired pulse duration or duty cycle. An excessively large input to the grid of VI must be avoided, as overloading may occur in V, or V . The lamp drive should then be increased by means of Ra until stable lamp firing is obtained.

    The apparatus has proved to operate quite successfully. The results of a series of calibrations on anum ber of accelerometers will be published soon.

    * This work was supported by Branch ORDTB of the Office of Chief of Ordnance.

    Simple Controller for a Platinum Furnace MYRON B. REYNOLDS

    General Electric Company, Knolls Atomic Power Laboratory,* Schenectady, New York

    (Received April 15, 1954)

    T HE use of the winding of a resistance furnace as a tempera-ture sensing element is not new-an on-off controller based on this principle having been described by White and Adams some years ago. l There are a number of experiments in which it is de-sired to maintain a small specimen of material at a constant tem-perature and the arrangement of the apparatus can be made such that the only heat loss is through the furnace walls. In this case the specimen temperature is dependent only upon the temperature of the furnace core and the advantage of having the temperature sensing element coincident with the heat source is obvious. This method of temperature control is of course limited to furnaces wound with wire or ribbon of a material having a high temperature

    '----E -------'

    FIG. 1. Temperature control bridge.

    coefficient of resistivity, such as tungsten, molybdenum, or one of the platinum group metals.

    The furnace winding is normally made one arm of a bridge as shown in Fig. 1 in which RF is the furnace winding, E is the applied voltage, and e is the error voltage available for the operation of the temperature control device. The bridge will be in balance only at a temperature determined by the relative values of RF, R l , R 2, and Ra. The magnitude of the error voltage will depend on the difference between the furnace temperature and the balance temperature while the phase or polarity of the error voltage will depend upon whether the furnace temperature is above or below the balance temperature. Resistors R l , R 2, and Ra must be of material having a negligible temperature coefficient. Rl must be

    LINE, 120. A C

    R4 5 ;:==t:===:J 17 5000

    R3 R2

    FIG. 2. Furnace controller, schematic.

    sufficiently rugged to carry the furnace current without over-heating, while the sum of the resistances or R2 and Ra must be great enough to keep the heat generation in these arms of the bridge negligibly small. The sensitivity of the bridge shown in Fig. 1 expressed in volts per degree per volt input is

    1 de Edt,



    in which at is the first thermal coefficient of resistivity for the winding material at the temperature t. It may be easily shown by differentiation of the above that maximum sensitivity is obtained when RF=Rl . This condition cannot be obtained in practice be-cause of the limitation on heat generation in R I Heat generation in Rl can be reduced only by making its resistance small in com-parison to RF, thereby sacrificing sensitivity. However, sufficient sensitivity may still be obtained to justify use of the method. For example, if Rl = O.OlRF at operating temperature, the sensitivity using a platinum wound furnace operating with 20 volts input at loooC will be 0.16 millivolt per degree centigrade as compared with 0.011 millivolt per degree for a platinum-platinum 10-percent rhodium thermocouple at the same temperature.

    How the error signal from such a temperature sensitive bridge is actually used to control furnace temperature is largely a matter

    This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions.

    Downloaded to IP: On: Sun, 21 Dec 2014 01:21:15


View more >