MF1 Sync Unit
Control your sequencers and drum machines from tape.
How many times have you dreamt that you could control your sequencer or drum machine from a pre-recorded tape? Well, if you've got a sync input, here is something to plug into it!
There are three basic situations in which the MF1 Sync Unit can help:
* As an independent clock generator.
When a number of effects have to run in synchronisation, one is far too often faced with the fact that although many effects have a sync input, very few have an internal clock input. The only way out is to use an external clock generator and this can be our sync unit. Its normal and inverted outputs (active high and active low respectively) will happily control a large number of sync inputs of both standards, provided their 'master clock' rates are compatible.
* Another direct application of the unit is for the pre-programming of a sequence of tempos for a live performance.
In this case, an ordinary cassette recorder can be used to 'memorise' the sync pulses (available at the 'record' output) whose rate can be changed at any time. When we replay the recorded sequence into the 'play' input of the MF1, it will again generate a clean set of pulses.
* The most important application of the unit is in multitracking situations.
The problem arises from the fact that often different instruments like drum machines, sequencers, etc., have to be recorded at different times, yet with perfect synchronisation, with respect to each other. The only solution is to use a multi-track recorder, reserving one of the tracks for synchronisation purposes (as a 'click track'). We can then use exactly the same technique as in the previous example, controlling a different instrument at every replay and recording the output of that instrument onto another track. The net result is that every time, the new instrument will play exactly in synchronisation with the previous ones.
* As a sound trigger.
Here's an incidental use of the unit with the high sensitivity of the 'play' input and the internal wave shaper being used to shape the outputs of a transducer, e.g. microphone, in order to use it as a synchronisation or pulse trigger source.
The unit has been used with various Roland sequencers and drum machines and has proved totally reliable in operation. In fact, to avoid any spurious breakthrough of the control pulse recorded on a tape track to other adjacent channels (important with the new Fostex A8 and Teac A3340S), a very low level signal only is required. The new Roland series (TR808, CSQ600 etc.) all need the value of C1 to be 10nF.
Whether you have a 4 or 8 track, this unit gives you the opportunity to put down e.g., a standard 4 in a bar rhythm throughout a piece from which you can synchronise your other tracks, and then later modify the drum or sequence tracks as you wish - a great advantage.
The block diagram of Figure 1 shows the arrangement of the various stages of which the unit is comprised. A variable frequency astable multivibrator generates the clock signal which is fed via an attenuator to a buffer amplifier. Some of the output from the buffer stage is fed to an output level control and then to the output socket by way of an amplifier. When switched to the 'record' mode, some of the output from the buffer stage is fed into the circuitry used to interface the recorder to the controlled instrument. This gives a monitor signal at the antiphase clock outputs.
In the 'playback' mode the clock oscillator is disconnected from the pulse shaping circuits, and these are instead fed with the input signal from the cassette tape recorder.
The input signal is fed to the pulse shaping and processing circuits by way of a variable gain amplifier which ensures these circuits receive a suitable signal level. Although the output from the tape recorder may well have serious waveform distortion due to the limited frequency response of an inexpensive cassette recorder, the processing circuits ensure that the output from the clock output sockets will always be good quality squarewaves of opposite phase.
Figure 2 shows the complete circuit diagram of the MF1 Sync. Unit. IC1a and IC1b are used in a conventional CMOS astable multivibrator circuit which has a stable operating frequency that is not significantly affected by variations in the supply voltage. RV2 enables the operating frequency to be varied from just a few Hertz to several kiloHertz, and this is the 'tempo' control. RV1 is a preset control which sets the upper frequency limit of the clock oscillator.
R2 and R3 attenuate the output of the clock oscillator by a little over 20dB before it is applied to a simple buffer stage based on operational amplifier IC2a. The signal is then applied to output level control RV4, and then to the 'record' output socket via another buffer stage. The output is sufficiently strong to drive the 'aux' input of a tape recorder or some other high level input.
With S1 in the 'record' position, some of the output of IC2a is fed through RV3 to the signal shaping and processing circuits. With S1 in the 'playback' position these circuits derive their input signal from the output of the tape recorder and a variable gain amplifier using IC3b enables a suitable signal level to be obtained. RV5 is the gain control and gives less than unity gain when at near minimum resistance, and over 40dB of gain at maximum resistance. This enables the unit to function properly with a very wide range of input levels. The first stage of the signal processing circuitry is a simple buffer amplifier employing IC3a.
The signal at the output of IC3a may be a squarewave of reasonable quality, but it is likely that the rise time will be rather low due to limited high frequency response of a cassette recorder (or virtually any other type of recorder for audio frequency use). If the clock frequency is very low, say in the region of 20 Hertz or less, the low frequency response of a cassette recorder, or any normal audio frequency recorder, will produce severe waveform distortion. In a severe case alternate positive and negative pulses will be obtained at the output instead of each squarewave cycle. The positive pulses are produced by the rising edges of the input signal, and the negative pulses are produced by the trailing edges.
C4, D1, D2, and the surrounding components are used to process the input signal so that brief antiphase pulses are produced regardless of whether the input is a reasonable squarewave or has undergone serious waveform distortion. These pulses are amplified separately by IC4a and IC4b and used to set and reset an RS flip-flop formed by IC1c and IC1d. This gives a good quality squarewave output from each of the antiphase outputs of the flip-flop, and any deficiencies on the input signal are not present on the output signals.
Power is obtained from two PP3 batteries and the current consumption is only about 6 to 7mA from each of these.
An instrument case measuring about 150mm by 120mm by 50mm makes a good housing for the unit, but note that this is about the minimum size that will accommodate all the components comfortably. As can be seen from reference to the photographs, RV4 and the four sockets are mounted on the rear panel, and the four main controls are fitted on the front panel.
The two batteries fit on opposite sides of the case, and all the other components are mounted on the printed circuit board. Details of the printed circuit board are shown in Figure 3, and there should be no difficulties when fitting the components into place provided the specified printed circuit mounting capacitors are employed. In addition to the components there are five short wire links to be soldered in place. IC1 is a CMOS device and in order to minimise the risk of damage by static charges it should be the last component to be soldered into place. As the 4001 is a very inexpensive device it is probably not worthwhile using a socket for this device, but use a soldering iron having an earthed bit when connecting it, and handle this component as little as possible once it has been removed from its protective packaging.
The printed circuit board is mounted centrally on the base panel of the case using four 6mm stand-offs, or four 6BA bolts and 6mm spacers. As there is not much excess space inside the case it is probably best to fit the board into a suitable position on the base panel and then use it as a template when marking the positions of the four mounting holes. After temporarily removing the controls and sockets from the case the mounting holes are then drilled. This method ensures that the controls, batteries, or sockets do not obstruct the printed circuit board and prevent it from being fitted into place.
Either the component panel must be wired to the other components before it is finally installed in the case, or pins must be fitted at the points where connections to off-board components will be made, so that this wiring can be completed with the board mounted in the case. Figure 3 illustrates the point-to-point style wiring of the unit. Be sure to use screened leads where indicated in the wiring diagram.
There are only two preset controls to set up, and it is not necessary to have any test equipment in order to give these the correct settings. Initially, RV1 should be set at midway, and RV3 should be adjusted fully clockwise. Set the tempo control (RV2) to maximum, S1 to 'record', and connect the clock input of the unit which is to be controlled by the MF1 to one of the clock output sockets (JK3 or JK4). Set the controlled equipment for use with an external clock signal and then switch on both units. Nothing should happen at this stage, but if RV3 is slowly adjusted in an anticlockwise direction at some point the controlled equipment should start operating. This is the correct setting for RV3.
Next RV1 is adjusted to give the fastest tempo that will be required, and it should be borne in mind that setting RV1 for a higher tempo than this will make RV2 cover a wider frequency range and thus make it more difficult to adjust to precisely the required tempo. It may be found that the range of clock frequencies available is too low for satisfactory results with some items of equipment, and this can be corrected by reducing the value of C1 to 0.01uF (10nF). This boosts output frequencies by a factor of ten.
The output signal from JK1 should drive the 'aux.' input or other high level input of a tape recorder without any problems. It should also be possible to use the microphone input of a recorder if a high level input is not provided, but output level control RV4 would need to be adjusted almost fully anticlockwise to prevent the recorder from being overloaded. If the unit is used into a microphone input it would be advisable to reduce the maximum output level by reducing R6 from 56k to 5k6.
As the MF1 will operate with an input level of anything from a few millivolts RMS to several volts RMS it will operate properly with any normal cassette recorder or deck, and the position of the recorders volume control or output level control will not be critical. Gain control RV5 is adjusted just beyond the point at which the unit has sufficient gain to produce a clock output signal.