Tuning Forks

Application Note for Fox Tuning Forks
by Louis Bradshaw

The 32.768kHz watch crystal is one of the most popular crystal units sold worldwide today. This crystal, in its varied packages, is manufactured by the millions and yet remains one of the least understood of the commercial crystal units.

In our experience, there are two (2) aspects of this crystal unit that are the least understood.

1. Drive level (current) supplied to the crystal unit
2. Frequency deviation over a given temperature range

Drive Level:

Drive level is defined as the current available at the electrodes of the crystal unit. For the sake of simplicity, this value is most often specified in terms of the amount of power dissipated by the operating crystal unit. As such, the values range from several hundreds mW (milliWatts) down to the µW (microWatts) range and even, in certain cases, the nW (nanoWatts) range. In the case of the 32.768kHz watch crystal, the most commonly specified value of drive level is 1.0µW maximum. The obvious meaning of this is that the current available at the electrodes of the crystal unit must be limited to a value that will result in the crystal unit dissipating no more than 1.0µW of power while in its operating state. Exceeding this value can, and will, result in permanent, irreversible changes in frequency, erratic behavior of the crystal unit versus changes in temperature, and even in catastrophic failure of the crystal unit due to breakage of the quartz plate.

The design of the applicable oscillator circuit and the selection of the components of that circuit are of prime importance in achieving the recommended drive level. In the usual case, the circuit used is built around an inverter and is commonly called a “parallel resonant” circuit. Such a circuit is illustrated in Figure 1.0 below.

Figure 1.0

In the circuit depicted by Figure 1.0, R1 is a resistor whose function is to force the IC into its linear mode of operation. R2 is a resistor whose function is to limit the current available at the electrodes of the crystal unit. As such, the value of R2 must be carefully calculated and specified. Due to the plethora of inverter chips, each with its own internal design and subsequent operating characteristics, Fox Electronics can assign no hard and fast value for either R1 or R2.

We have found, however, that a good starting point for an oscillator circuit using the watch crystal is that R1 initially be set to 20 Megohms and R2 to 0.500 Megohms. In addition, we have found that setting capacitors C1 and C2 at some value between 10 and 20pf each provides a good starting point.

We do, however, as do all crystal vendors, specify a value for the maximum resistance of the crystal unit. Due to the type of vibration exhibited by the watch crystal, the specified maximum resistance value is quite high, typically 35 to 50 Kilohms. In the usual case, the actual resistance will be approximately 75% of the specified maximum, thus providing a starting point for the calculation of the value of R2.

Once the calculation has been made, and the circuit is operational, it is important that the actual current to the crystal be measured.

In the case of the 32.768 kHz watch crystal, the current to the crystal is at a very low value, typically 1.0uA or less. As this level is too small to permit the use of an oscilloscope, the use of a current probe in conjunction with a selectable level meter is recommended. A typical set-up of this nature is illustrated in Figure 2.0.

Figure 2.0

It is strongly recommended that the current be measured with a leaded crystal, even if an SMD will be used in production. Given the characteristics of the watch crystal, no appreciable differences in operating parameters will be encountered from one package to another. Figure 3.0 illustrates the attachment of the current probe to the crystal lead.

Figure 3.0

If a current probe is not available, an alternative method is to temporarily install another resistor in series with the crystal. This resistor should be selected to have a value equal to approximately 10% of the specified maximum resistance of the crystal unit. The voltage drop across this resistor may be measured by conventional means and, by means of Ohm’s Law, the current through the resistor calculated. Once the current is known, the value of the power dissipation may be calculated.

Negative Resistance:

Another approach which may be taken, the the absence of suitable current measuring equipment is the “oscillation allowance” or “negative resistance” method.

Basically, this method consists of placing a variable resistor, with a value of between five to ten times the specified maximum series resistance of the crystal, in series with the crystal. With the resistor set to its maximum value, the oscillator circuit is energized and the output waveform monitored with an oscilloscope. By varying the value of the resistor, a point will be reached at which the oscillator will start, as evidenced by the existence of a waveform. The value of the variable resistor in conjunction with the value of R2 from Figure 1.0, at the initial oscillation point, represents the maximum resistance and/or lowest current at which the crystal will start and oscillate reliably. The value of the resistance at this point is measured and a resistor of the same, or nearly so, value is used to replace R2.

Frequency Deviation Over Temperature:

Like all crystal vendors, Fox Electronics specifies a value of frequency deviation over a temperature range. The temperature range is typically one that experience has taught us will satisfy the requirements of the majority of our customers. Even so, we are often asked what the results will be if that temperature range is modified.

A plot of the frequency deviation versus changes in temperature for the 32.768kHz watch crystal exhibits a parabolic (inverted “U”) form. At a certain temperature, usually around 25ºC ± 5ºC, the frequency deviation versus temperature becomes zero, or very nearly so. The temperature at which that occurs is called the “turnover” temperature. The frequency deviation on either side of the turnover temperature is negative with respect to the frequency at the turnover point. A graph of this type of frequency deviation versus temperature is shown as Figure 4.0 below.

Figure 4.0

The exact temperature at which the turnover point occurs is a function of the angle at which the specific quartz plate was cut from the original stone. Therefore, it is theoretically possible to alter the turnover temperature and we are sometimes asked to do this. Practically, it is well to remember that these watch crystals are manufactured by the millions, using as many automated processes as can be devised. Even so, manufacturers of these devices are hard pressed to keep up with demand. The probability of inducing a manufacturer to interrupt his manufacturing sequence for a limited run of a non-standard part is virtually zero.

The frequency deviations in the above chart are calculated by:

f = k* (TTo)2 Where f is the frequency deviation in PPM, k is a constant having a value of -0.04 max., T is any temperature and T0 is the turnover temperature, equal to25ºC ± 5ºC.

It is considered that the specific crystal unit under consideration will oscillate at temperatures outside the range specified but with ever increasing deviation from the nominal of 32.768kHz. However, these crystals should not be used outside the temperature specification without consulting the factory because of other possible effects of temperature.

If additional technical information is needed, contact Fox Technical Support at:contact-us

Request Datasheet

Request a datasheet for 

FOXSelect™ is loading...