Handbook for Sound Engineers

Magnetic Recording and Playback 1091

under all conditions of tape pack and speed. Induction

motors, which do not run at synchronous speed, will

always yield a moving pattern.

Crystal-referenced servos may falsely appear to vary

in speed when tested with a strobe light if the frequency

of the ac mains driving the strobe varies. An oscillo-

scope and frequency counter are required to properly

verify correct servo operation.

28.6.2 Flutter

Speed drift represents only the very lowest frequency
components of the spectrum of speed errors. Measure-
ment of the higher-frequency flutter components
requires a specialized frequency demodulating instru-
ment called a flutter meter. As seen in Fig. 28-60, the
flutter meter may resemble the phase-lock servo of Fig.
28-4. The reference signal from the crystal clock must
pass through the record/playback process of a tape
recorder before being applied to one of the phase
comparator inputs. The low-pass filter and
voltage-controlled oscillator simulate a large flywheel
that stores the average value of the playback frequency.
By applying the average value to the second phase

comparator input, the phase comparator output will

consist of only the short-term variations from the

average speed. These variations are divided into various

frequency bands for further analysis. The metering

circuit provides a convenient quantitative measurement

of the speed variations.

Just as the sampling rate of a digital audio system

determines the highest possible audio frequency that

can be encoded, the frequency of the test tone deter-

mines the range of flutter components that can be

measured by any frequency demodulator. The typical

upper frequency is about 0.4 times the test frequency.

Due to the nature of the sidebands that are required to

operate the demodulator, a typical 18 kHz audio band-

width can support a 12.5 kHz test tone and a flutter

bandwidth of 5 kHz. This measurement technique,

referred to as high-band flutter measurement, is

supported by Audio Precision.

Unfortunately, most flutter meters use a

low-frequency test tone of 3150 Hz and cut off all

flutter components above 250 Hz, ignoring many flutter

components caused by modern-day servo systems and

virtually all scrape components due to the elastic vibra-

tion of the tape. To make matters worse, most flutter

specifications are made through a flutter weighting filter

that only measures flutter components near 4 Hz. Proper

maintenance requires that a broader spectrum test be

implemented to check for any possible problem.

Two methods of specifying flutter performance are

commonly encountered. If a flutter-free test tape is

available, the flutter reading obtained in the playback

mode can be reported. Most professional recorders,

however, have flutter levels that are equal to or better

than any available test tapes. In this case, recording and

Figure 28-59. Strobe light for speed testing.

120 Vac

47 k 7 1 / 2 W

1N4004 NE 2H

Neon bulb

Note: Package the components inside a discarded

plastic pen housing with the tip of the bulb protruding.

Figure 28-60. Flutter meter block diagram.

12.5 kHz

Reference

Input signal oscillator

Recorder under test

Flutter

phase

comparator

Low-

pass

filter

Voltage-

controlled

oscillator

Flutter

Analysis

filters Meter

90°

Phase

shifter

AM Phase

comparator AM Monitor output

Output signal