by M. J. SALVATI
This article describes the construction and operation of a low cost device capable of measuring turntable or recorder speed accuracy and wow and flutter. Cost is kept as low as possible by utilizing the audiophile's own d.c. and a.c. voltmeters to read out speed accuracy and wow and flutter, respectively. However, instructions are given for internal metering, if desired. While this device does not have the same accuracy and features as commercial instruments selling for $500 to $1,000, it certainly will check the specs of medium- and low cost recorders and will indicate serious deficiencies in top-of-the-line machines. Full specifications for the meter are given in Table I.
Theory of Operation
The output of a turntable or tape machine is fed to a frequency discriminator whose output contains a.c. and d.c. error-signal components representing wow and flutter and speed error, respectively. These error signals are individually processed by the two signal channels, and read out by means of a.c. and d.c. voltmeters connected to the unit. The schematic of the basic device is shown in Fig. 1.
Frequency Discriminator. Op-amp IC1 is a unity-gain buffer that provides a high input impedance for the unit. It drives the input of phase-locked loop IC2. As long as the input signal is above 10 mV and within ± 30% of the 3,150- or 3,000-Hz test frequency, the IC's d.c. output level is directly proportional to the frequency of the input signal. As the input frequency shifts (due to recorder speed variations), the d.c. output voltage of IC2 also shifts. Capacitor C5 and resistors R3 and R4 determine the center frequency of the discriminator.
Speed Channel. The error-signal output voltage at pin 7 of IC2 is referenced to its pin 6 voltage, so a differential amplifier (IC3B) with a voltage gain of 75 x amplifies the error signal and converts it to a ground-referenced signal. A filter with a very long time constant (resistor R21 and capacitor C13) removes the rapid variations (wow and flutter), so the d.c. voltage appearing at the output of buffer IC5 represents only the (relatively) constant speed error. A d.c. voltmeter connected to jacks J3A and J3B will there fore indicate speed accuracy (or "drift," as it is sometimes called). Be cause of the drive capability of IC5, even a low-impedance voltmeter can be used for readout.
Wow and Flutter Channel. The error-signal voltage at pin 7 of IC2 is also applied to a 200-Hz low-pass filter.
This second filter and amplifier chain removes the test frequency, sets the upper frequency limit of the wow and flutter signal passband, and amplifies the signal to the extent necessary to establish the 1 V per 1% scale factor.
Op-amp IC3A uses a small amount of gain (1.7 x) to produce a sharp corner despite the use of only two filter sections (R7, R8, C6 and C7). The following op-amp (IC4) provides a voltage gain of 44 x , so the total voltage gain for this channel is 75 x . Since this gain is the same as that of the speed channel, it means that when the "Speed" output is calibrated (via pot R17 across the output of IC2), the "W & F" output is calibrated simultaneously.
---------
Table I--Specifications
Operating Frequency: 3,000 or
3,150 Hz.
Input Level: 10 mV to 20 V.
Input Impedance: 100 kilohms.
Speed Readout Ratio: 1 V d.c.
per 1% speed error.
Speed Accuracy Ranges: Depends on d.c. voltmeter used;
±5% for internal metering.
Wow and Flutter Frequency
Range: 0.5 to 200 Hz.
Wow and Flutter Readout: 1 V
a.c. per 1% wow and flutter.
Wow and Flutter Ranges: Depends on a.c. voltmeter used;
0.2%, 1%, 5% for internal metering.
---------
The input coupling to op-amp IC4 is capacitive, to remove the d.c. component from this channel. The output signal of IC4 is applied to the "W & F" jacks (J2A and J2B) via another low pass filter (R13 and C11), which pro vides an additional 10 dB reduction of the 3,150/3,000-Hz test frequency.
Internal Metering. Meter M I (see Fig. 2A) has a zero-center scale, so it is not necessary to switch polarity as tape speed drifts from faster to slower than normal. For wow and flutter measurements, a high-pass filter and precision rectifier are combined to drive meter M2. Op-amp IC8 is configured as a maximally-flat, two-pole filter whose corner frequency is 0.5 Hz. This establishes the low-frequency limit of the 0.5 to 200-Hz standard wow and flutter passband, and also provides some amplification before the wow and flutter signal is applied to the precision rectifier (IC9), which converts it to a unipolar voltage. The "W & F" percent age switch (S4) selects the resistors that establish the scale factors for meter M2.
Construction Notes
There is nothing especially critical about layout beyond the common sense precaution of not routing the power-switch leads near the "Input" connector and its switch.
With the exception of the power transformer, rectifiers and filter capacitors, nearly all of the basic-unit components are mounted on a single piece of Veroboard or equivalent perfboard.
Furthermore, if you obtain compact filter capacitors, there is no reason why all power-supply components other than the transformer cannot be mounted on the same board. However, if you opt for internal wow and flutter metering (Fig. 2B), it might be more convenient to build the additional circuitry on a separate board.
Choice of Parts. There is consider able latitude in the ratings for many of the parts listed in Tables II and Where space permits, the alternate choices are given in the parts lists, with the preferred part listed first. For example, the low-power (100 mA, TO-92 case) versions of the three-terminal plastic voltage regulators will do nicely for IC6 and IC7, although any of the equivalent higher-power versions listed work just as well. Similarly, any capacitors having a voltage rating of over 20 V can be used (except for C15 and C16). The most important consideration with capacitors is to locate the ceramic disc bypass capacitors physically close to the associated ICs, and capacitor C16 close to IC7. Where no type of capacitor is specified, you can use paper, polyester, mylar, or ceramic.
================
Table II-Basic Unit Parts List
IC1, IC4, IC5--LF351 or LF13741 FET op-amps.
IC2--NE565 or LM565 PLL.
IC3--µA747 or LM747 in 14-pin package.
IC6--78L15, 7815, or LM340T-15 positive regulator.
IC7--79L15, 7915, or LM320T-15 negative regulator.
J1--Phono jack.
J2, J3--Pairs (red/black) banana jacks.
S1--SPST toggle or slide switch.
S2--SPDT toggle or slide switch.
S3--DPDT toggle or slide switch.
T1--16 to 18-V, 0.25-A power transformer (Mouser 41FJ300 or Digi-Key T102).
D1, D2-6.0-V, 500-mW zener diodes.
D3--LED.
D4, D5--100-uV, 1-A silicon rectifier diodes.
C1, C2, C8, C9, C11, C12, C14--0.022-µF, 25-V ceramic disc capacitors.
C3--0.1-uF, 25-V capacitor.
C4--0.001-µF, 25-V capacitor.
C5, C6, C7--0.01-µF, ± 10% styrene or silver-mica capacitors.
C10, C13, 017--0.22-uF, 25-V capacitors.
C15, C16--220-g, 35- or 50-V electrolytic capacitors.
C18--2.2-µF, 25-V tantalum electrolytic capacitor.
R1--100-kilohm, 1/4-W carbon-film resistor.
R2, R13--4.7-kilohm, 1/4-W carbon-film resistors.
R3--10--kilohm, 15-turn trim pot (Weston 830P or Spectrol 43P).
R4--5.6-kilohm, 1/4-W, 5% carbon-film resistor.
R5--560-ohm, 1/2M resistor.
R6--13-kilohm, 1/4-W, 5% carbon-film resistor.
R7, R8--120-kilohm, '1/4-W, 5% carbon-film resistors.
R9--430-ohm, 1/2-W resistor.
R10--9.1-kilohm, 1/4-W, 5% carbon-film resistor.
R11, R21--4.7-megohm, 1/4-W carbon-film resistors.
R12, R20--10-kilohm, single-turn trim pots.
R14--43-kilohm, 1/4-W, 5% carbon-film resistor.
R15--1-kilohm, 1/4-W, 5% carbon-film resistor.
R16, R18--10-kilohm, 1/4-W, 5% carbon film resistors.
R17--47-kilohm, single-turn trim pot.
R19, R22--750-kilohm, 1/4-W, 5% carbon-film resistors.
7 x 5 x 3-inch aluminum box (Mouser LMB TF-782).
Fig. 1--Schematic diagram of basic unit.
================
The input and output connectors specified are the most typical for their application. They can (and should) be changed to other types more suited to your own tastes and/or test equipment. (I certainly did this in my prototype!)
Options and Modifications. The circuit shown in Fig. 1 was designed to interface with as wide a variety of test equipment and tape machines as possible. In most cases, you can save a little time, effort and money by eliminating the features not necessary to your setup.
If all your tape machines have at least one output jack apiece whose normal output level is in the range of 10 mV to 1 V rms, you can eliminate S2 and the 4.7-kilohm resistor. Connect the "Input" jack (J1) directly to pin 3 of IC1, and resistor R1 from there to ground.
If you are using an external d.c. volt meter to read out speed, and that volt meter has its own polarity (" + /-") switch, you can eliminate S3. Simply connect the red ( + ) "Speed" jack to pin 6 of IC5, and the black (-) jack to ground.
If you plan to use internal metering for speed readout, eliminate the "Speed" jacks and switch S3, and add the parts shown in Fig. 2A. Resistor R23 then connects directly to pin 6 of IC5, and the meter "-" terminal connects to ground.
Using your a.c. voltmeter to read out wow and flutter is both convenient and cost efficient. However, many a.c. volt meters lack full response down to the 0.5 Hz commonly used as the lower limit measured by commercial wow and flutter meters. Therefore, you may want to build a low-frequency a.c. volt meter into this device. A suggested circuit, which consists of a 0.5-Hz high pass filter, precision full-wave rectifier and metering, is shown in Fig. 2B. This circuit provides full-scale wow and flutter ranges of 0.2%, 1%, and 5%. Since multi-scale meters are rarely available as stock items, you will have to use a 0 50 scale microammeter and add the 0-0.2 and 0-1 scales with dry-transfer lettering. The decade zeros of the existing scale graduations can either be erased or painted over with white enamel.
A list of additional and alternate parts needed for internal metering is given in Table III. Note that some of these parts (mostly capacitors) are used in place of their equivalents in the Basic Unit Parts List. After you have decided what (if any) options or modifications you want in your unit, carefully go through both parts lists to select only those parts you will actually use.
Packaging. This depends on the degree of complexity selected. The unit I built is housed in a custom-made box.
The recommended standard-size aluminum case for the basic unit is 7 x 5 x 3 inches, with a 5 x 3 end used as the front panel. A suggested layout is shown in Fig. 3A. If you use internal metering, a larger case is necessary to accommodate the meters. For both internal speed and wow and flutter metering using the recommended 4-inch meters, a 10 x 6 x 3 1/2-inch case is ideal. The suggested layout shown in Fig. 3B uses one 10 x 6 face for the front panel.
Parts Availability. All of the ICs, resistors, and capacitors are available from consumer-oriented mail-order houses such as Digi-Key and Jameco Electronics. The meters, power transformer, case, and many of the other parts are available from Mouser Electronics.
Order their catalogs first to check prices and availability. (Digi-Key, Box 677, Thief River Falls, Minn. 56701; Jameco Electronics, 1355 Shoreway Rd., Belmont, Cal. 94002, $10 minimum order; and Mouser Electronics, 11433 Woodside Ave., Santee, Cal. 92071, $20 minimum.)
Fig. 2--Schematic diagram of optional, internal metering circuitry for speed
accuracy (A) and wow and flutter (B).
==================
Table III-Options Parts List
IC8--LF351 or LF13741 FET op-amp.
IC9--LF353, LM1458, RC4558, TL072, or TL082 dual op-amp.
S4--1-pole, 3-position rotary switch.
M1--50-0-50 µA microammeter (Mouser 39LK416).
M2--0-50 uA microammeter (Mouser 39LK414).
D6, D7--1N914 or 1N4148 silicon signal diodes.
C19, C22--0.022-uF, 25-V ceramic disc capacitors.
C20, C21--0.22-uF, ± 10% capacitors (Panasonic M1224).
C23--100-pF ceramic disc or mica capacitor.
R23--100-kilohm, 1/4-W, 5% carbon-film resistor.
R24-13-kilohm, 1/4-W, 5% carbon-film resistor.
R25, R26-1.5-megohm, 1/4-W, 5% carbon-film resistors.
R27--9.1-kilohm, 1/4-W, 5% carbon-film resistor.
R28, R29--10-kilohm, 1/4-W, 5% carbon-film resistors.
R30--4.7-kilohm, 1/4-W carbon-film resistor.
R31--5.3-kilohm (5.1 kilohms and 200 ohms), 1/4-W, 5% resistor.
R32--33-kilohm, 1/4-W, 5% resistor.
R33--168.5-kilohm (160 kilohms and 8.2 kilohms), 1/4-W, 5% resistor.
10 x 6 x 3 1/2-inch aluminum case (Mouser LMB TF-784).
===================
Adjustments
To adjust this device, an audio oscillator, a.c. voltmeter, d.c. voltmeter, and frequency counter are needed. Before energizing the circuit, set all pots (R12, R17 and R20) to mid-rotation. Proceed as follows:
1. Plug in the power cord and flip the power switch on. The LED should illuminate.
2. Check the output voltages at IC6 and IC7. They should be + 15 and-15 V respectively.
3. Check the voltages at pins 7 and 4 of IC1. They should be +6 and-6 V respectively.
4. Check the voltage at pin 6 of IC2. It should be approximately +4.5 V.
After making these checks, go on to the calibration adjustments, but wait until the wow and flutter meter and test equipment have been warmed up for at least 20 minutes. Do the calibration procedures in later adjustments depend on the pre ceding ones.
D.C. Balance. To adjust the d.c. balance of IC4, connect a d.c. voltmeter across the "W & F" jacks (J2A and J2B). Then, carefully adjust pot R12 for an output voltage of zero, ±01 V.
To adjust the d.c. balance of the speed channel, proceed as follows:
1. Connect the d.c. voltmeter across the "Speed" jacks (J3A and J3B).
2. Either connect a jumper from pin 6 to pin 7 of IC2 (best way) or rotate pot R17 so it has zero resistance.
3. Carefully adjust pot R20 for an indication of zero, ± 0.1 V.
4. Remove the jumper or restore pot R17 to mid-rotation.
Calibration. The procedure given here will adjust for a scale factor of 1 V d.c. per 1% frequency error at the "Speed" jacks, and a 1-V a.c. output per 1% wow and flutter at the "W & F" jacks, based on the newer standard test frequency of 3,150 Hz. However, the 3,000-Hz figures will be given in parentheses, in case your test tapes use this frequency. After determining which frequency applies, proceed as follows:
1. Connect an oscillator whose frequency can be accurately set (either by dial calibration or by a counter) to "Input" jack J1. Set the oscillator output to around 100 mV frequency to 3,150 (3,000) Hz.
2. Set the range switch of the d.c. volt meter (connected to the "Speed" jacks) to 5 V and its polarity switch (or S3) to positive. If the prior adjustments were properly made, this voltmeter should indicate 0 V with exactly 3,150-Hz (3,000-Hz) input.
3. Change the oscillator frequency to exactly 3,276 (3,120) Hz, as indicated on the frequency counter or dial.
Adjust trim pot R17 for a d.c. volt meter indication of exactly 4.0 V.
Center-Frequency Adjustment. Pot R3 allows the center frequency to be set to either of the standard test frequencies (3,150 or 3,000 Hz). Again, the procedures given here are for 3,150 Hz, with figures for 3,000 Hz in parentheses. To adjust R3, proceed as follows:
1. Connect a frequency-stable oscillator to "Input" jack J1. Set the oscillator output level to around 100 mV rms.
2. Connect a d.c. voltmeter to "Speed" jacks J3A and J3B.
3. Carefully adjust the oscillator frequency (with a frequency counter) to 3,150 (3,000) Hz,-± 2 Hz. Then adjust trimmer R3 for zero d.c. out put at the "Speed" jacks.
4. When you have done this, recheck the frequency counter display to make sure that the input frequency is exactly 3,150 (3,000) Hz. Read just the oscillator frequency if necessary, and readjust trim pot R3 for exactly zero d.c. output. The accuracy of your speed-error measurements depends on how carefully this adjustment is made.
Note: If you do not have a frequency counter, pot R3 can be set well enough for wow and flutter measurements by using the dial calibrations of the oscillator. However, this alternative is not accurate enough for reliable speed measurements.
Fig. 3-Suggested panel layouts of basic unit (A) and unit with full internal metering (B).
Operation
There are two ways to use this de vice to check tape machines. The simplest method is with your own oscillator as the signal source. This permits an overall check of wow and flutter; re cord and playback performance is measured as a whole. The second method uses a prerecorded test tape only for wow and flutter and speed accuracy. Turntables use this second method with a test disc such as the CBS STR-151. For an overall measurement:
1. Connect an a.c. voltmeter with good low-frequency response to the "W & F" jacks.
2. Connect a stable audio oscillator to the tape machine's input connector.
Set the oscillator output to whatever level the tape machine requires for recording. Set the oscillator frequency to the same test frequency (3,150 or 3,000 Hz) as that for which your device was previously calibrated.
3. Allow the equipment to warm up for a few minutes, then record 10 to 15 minutes of test signal on a blank tape. Do this near the middle of the tape reel or cassette for best results.
4. Disconnect the oscillator from the tape machine. Connect the tape machine's output to "Input" jack J1.
5. Play back the recorded test signal, and set "Input" level switch S1 to match the tape machine's output level.
The a.c. voltmeter will indicate total wow and flutter. A fairly steady meter indication means that the output is mainly high-frequency components (flutter). A widely varying indication means that low-frequency components (wow) predominate. However, this is just a generalization, since the damping and ballistics of the a.c. voltmeter affect the amount of pointer movement for any given amount of wow.
Overall wow and flutter measurements will vary considerably from the playback-only wow and flutter specifications listed by most manufacturers.
When the tape is recorded and played back on the same machine, the same cyclic speed variations occur in play back and recording, and can either add or subtract, depending on their phase. With test tapes recorded on a different machine of lower wow and flutter, this is less of a problem. In my experience, multiplying the overall measurements by 0.6 or so yields a figure that is roughly comparable to playback-only specifications; some testers recommend a figure of 0.7. It also helps, even with commercial test tapes, to make multiple readings and average them.
Wow and flutter test tapes are avail able from several sources. Standard Tape Laboratory ( 26120 Eden Landing Rd. #5, Hayward, Cal. 94545) has open-reel and cassette tapes at $40 to $50 for home formats, higher prices for half-inch, 1-inch or higher-speed tapes.
LC Engineering Laboratories ( 9451 North Kostner Ave., Skokie, Ill. 60076) has open-reel, cassette and microcassette wow and flutter tapes, at $19 to $34.
To check speed accuracy and/or playback-only wow and flutter, proceed as follows:
1. Connect an a.c. voltmeter with good low-frequency response to the "W & F" jacks and a d.c. volt-meter to the "Speed" jacks.
2. Turn on the device, the voltmeters, and tape machine. Allow everything to warm up for 20 minutes.
3. Connect the "Input" jack to the tape machine's output jack. Set the "Input" level switch to match the tape machine's output level.
4. Play the test tape in the tape machine. The d.c. voltmeter will indicate speed error at the rate of 1 V (meter indication) per 1% speed error. Positive voltage means the tape machine is fast; negative voltage means the machine is slow. Note: if wow is severe, the d.c. readings will also fluctuate slowly.
The a.c. voltmeter will indicate wow and flutter of the playback section alone.
Again, a steady meter reading indicates mainly high-frequency speed variation (flutter); if the reading fluctuates, low-frequency components (wow) predominate. The damping and ballistics of the particular a.c. voltmeter used will affect the speed and degree of observed variation for any given amount of wow.
(Audio magazine, Jun. 1983)
Also see:
A New Idea for Measuring Wow and Flutter (Jun. 1973)
Cassette Deck Transport Problems (Sept. 1974)
= = = =