This device displays the tempo of a bass drum in real time. I use it in jazz bands of about 7 pieces including a trap set with kick drum, but it can be used in any musical group that contains a bass drum.
The heart of the design is an inexpensive automotive panel-mounted LED tachometer that is made to receive one trigger pulse per revolution from a magnetic pickup mounted on the engine output shaft. In my device a microphone detects bass drum beats; the resulting signal is processed to be compatible with the tachometer. As such, the unmodified tachometer displays bass drum tempo as beats per minute. Musical tempo is customarily expressed as quarter notes or quarter note time equivalents per minute. If the bass drum is struck once per quarter note, then the tachometer will indicate tempo correctly. However, more commonly the bass drum is sounded on beats 1 and 3 in 4/4 time or on beat 1 in 3/4 time. In these situations the display would not indicate tempo correctly. Fortunately the tachometer can be modified to accommodate all three scenarios as described later.
The sound of a bass drum is essentially a damped sine-like wave having a frequency of about 20Hz to 100Hz depending on the drum; as amplitude decays, frequency drops slightly. The frequency and degree of damping depend on the individual drum, the softness of the striker, the drummer’s technique, and physical damping that the drummer may apply to the drum. The challenge is to electronically distinguish this signal from strong interfering signals from nearby cymbals and drums and from non-percussion instruments such as amplified bass guitar and bass violin. The design described here relies on a combination of frequency and spacial discrimination to separate the desired bass drum signal from the rest.
I modified a widely available off-the-shelf preamp-bias supply to provide most of the audio amplification. An internet search of max9814 electret microphone amplifier yields many suppliers. This product features programmable gain, microphone bias, and automatic gain control. The advertised low frequency response is 20Hz; the microphone supplied may go that low, but the electronics do not. To get a response low enough I changed 3 capacitors on the little circuit board. The circuit of page 11 of the Maxim MAX9814 data sheet is essentially the same as in the mic/amplifier package to be modified. Capacitor Cin should be changed to 1uf, Ccg to 22uf, and Cout should be shorted. I removed the microphone from the board for mounting elsewhere in the enclosure.
With the preamp programmed for 40db gain, the lowest available, it overloaded occassionally. The 300-ohm resistor shunted across the preamp input reduces the level from the microphone enough to eliminate overload; the 100uf cap prevents mic bias from being pulled down by the 300-ohm resistor. The 10uf cap rolls off frequencies above about 50Hz.
The preamp drives a 3-pole low-pass filter having 50Hz cutoff. This drives a rectifier, reservoir cap, transistor switch, and then a 555 monostable multivibrator, the output of which drives the tachometer. The monostable pulse width is set to the maximum possible for each tempo range; during this pulse the system is insensitive to incoming audio. This circuit may seem somewhat redundant, but I discovered that reliable exclusion of extraneous signal impulses, such as from a mallet bouncing on the drum head, depends on it.
The tachometer (search car 4 digit motor LED tachometer) contains a 11.05MHz crystal-controlled clock. The crystal is easily removed, and crystals of 1/2 and 1/3 this frequency are widely available. One of the three crystals is switched in depending on the bass drum rhythm being sampled.
A 555 astable multivibrator and rectifier provide a negative supply for the op-amps. I.C.s specifically designed for this purpose exist, and there are other techniques for single supply operation of opamps, but I used parts that were on hand.
The current drain of the tach is about 45ma with a superimposed periodic variation of frequencies 108Hz, 54Hz, or 36Hz depending on which rhythm is selected by S1, causing ripple on the 9 volt supply. As the battery becomes exhausted and its internal resistance rises, the ripple amplitude rises to several hundred millivolts p-p. The effect is to limit the lowest battery voltage at which the device will operate. This lower limit is about 6.5 volts, somewhat improved by the 2200uf cap across the battery. The life of an alkaline battery is about 6 hours, extended slightly by the presence of the smoothing 2200uf cap.
In use either the internal mic or an external mic should be as close as possible to the bass drum head; the enclosure can be mounted on or near the drum with a universal clamp. I get reliable operation with the mic a foot or less from the drum. The sensitivity control is set for reliable flashing of the LED for each drum beat; the setting is not critical. I tested 3 models of electret mics, each with an advertised lower frequency response limit of 20Hz; their performance in the device was essentially identical. The occasional drum having very high Q, that is, slow decay of the fundamental frequency, may trigger the tach prematurely between beats, giving an erroneously high tempo reading. A slight amount of physical damping of the drum cures this problem.