Guitar Vibrato/ Tremolo Unit

The original vibrato circuit was shown in Project 49, and while it's still potentially useful, this one is better all round.  Vibrato is used to obtain a variation in pitch (as opposed to tremolo, which varies the amplitude).  The unit was inspired by the Vox AC-30 guitar amp, but the resemblance stops there.  The 'Effect' control lets you select 'pure' vibrato, tremolo (with a very small amount of vibrato still present), or a mix between the two.  Rather than using JFETs, this version uses LED/ LDR (light dependent resistor) optocouplers, giving improved linearity.

Where the original version was limited to an input level of around 20-50mV (depending on the JFETs used), the design shown here can accept up to 5V RMS with little distortion.  LED/ LDR optocouplers are (by the very nature of LDRs) fairly slow, so any speed above 10Hz diminishes the effect quite dramatically.  However, this isn't a limitation for normal usage, because most vibrato is fairly slow (3-7Hz would be typical).

Unfortunately, there's a trade-off.  FETs are fast, and can operate over their full range at any frequency, but LDRs are slow.  If you need a vast amount of vibrato, then you'll be disappointed, because you won't get it.  You will get high linearity and the ability to work at 1V RMS or more without distortion, and it's up to the individual to experiment.  While it's shown as a complete circuit, you can change the capacitance to vary the amount of phase shift at different frequencies, and tinker with the LED current in the optocouplers.  Depending on the ones you use, they may need more or less current to provide the maximum result (and they will vary one from the next).

The circuit of the unit is fairly simple, and is not especially critical to set up.  If built exactly as shown, it will work straight away, although you may need to adjust the LED current a little for best effect.  Figure 1 shows the circuit for the vibrato system, and consists of an input buffer and two phase shift (all pass filter) networks.  As the phase is varied, so is the frequency.  The phase shift networks are designed for a centre frequency of 86Hz, although this is (relatively) unimportant in the way the circuit operates.  Using a lower than expected frequency for the phase shift networks provides more frequency shift where it's needed.

The phase shifter is a standard opamp circuit, and has been used for this sort of application many times.  After experimenting, I decided that the LED/ LDR was the best choice.  JFETs are fairly critical to set up, and have severe linearity problems at high levels.  Most vibrato circuits use only one stage, but the effect is not as good (especially at low rates), and the 'Effect' control is completely useless with a single stage.

Figure 1 - The Vibrato Circuit

The circuit is straightforward, except for the 'Effect' control.  With this, you can select the clean signal direct from the buffer stage, the fully phase (and hence frequency) modulated signal from the output of U2A, or a mixture of the two.  With the pot centred, there may be a loss of bass, but there's a very strong tremolo effect with an interesting tonal change.  You can vary the depth of tremolo with the pot, in lieu of the 'conventional' depth control.  The 100 Ohms resistor prevents the opamps from oscillating with long guitar leads.  Note that U2B is used in the oscillator section.

The LDRs are used as a variable resistance, and they introduce very little distortion at any level.  Figure 2 shows the modulator oscillator, which is a conventional opamp feedback circuit.  The modulation signal is taken from the capacitor, and is amplified so it's just clipping by U3B.  This also buffers the signal to prevent loading of the oscillator.  Closing the switch disables the oscillator, and stops the vibrato effect - any tonal variation obtained by the 'Effect' control remains.  To eliminate this effect, a complete bypass is required - see Figure 4 for an example.

Figure 2 - Modulator Circuit

The switch Sw1 is used to disable the oscillator.  If connected remotely using J3, this must be wired with a shielded lead to prevent extraneous noise disturbing the oscillator circuit.  The 'Speed' control changes the rate from about 3Hz (S) to 13Hz (F).  This can be extended, but below 3Hz the effect is not very great, and above 13Hz it becomes pretty much useless because of the limited speed of the LDRs.  I used VTL5C3 Vactrol^® optocouplers, but they may be hard to get and/ or expensive.  Project 200 describes DIY versions which will usually be just as good.

Although the opamps I have specified are fairly basic, they are more than adequate for the job.  They are common in commercial guitar amplifiers and effects pedals.  If you want to, substitute TL072 or OPA2134 FET input opamps.  The latter are quieter and a far better opamp, but will give a marginal improvement (if any) to sound quality in this application.  If lowest noise is an issue, then I suggest that the OPA2134 be used for U1, as this is more critical for low-level, high impedance circuits.  There are quieter and better opamps, but I shall leave this to the reader to decide.  Personally, I wouldn't bother.  Most dual opamps use the same pinouts, so the circuit is unchanged.

When building the circuit, make sure that there are no signal leads anywhere near the output of U3A, R15, R16, R17 or VR2.  The signal here is a square wave, and will make clicking noises in the audio signal.  Better (faster) opamps will make this much worse, so are not recommended.

Basically, there is (almost) no setup needed, but it depends on the optocouplers used.  The voltage divider (VR4, R20) following U3B provides a quiescent current of about 3mA through the optocoupler's diodes, and this is modulated via the capacitor (C4).  The LED current varies from zero to a maximum of 6mA.  This allows the maximum variation consistent with minimal current consumption.  You may need to change R20, depending on the optocoupler performance.  The value suggested (4.7kΩ) is a good starting place.  VR4 is a trimpot, and is adjusted to get the maximum effect.  Ideally, this would be set with the oscillator stopped, and the audio signal level at U1B.6 and U2A.2 adjusted for half the input level.  It can be done by ear as well, with the oscillator running at about 5Hz.

You will need to note the polarity across C5, because it may need to be reversed if the voltage at TP1 (the junction of VR4 and R20) is less than zero volts.  Normally, VR4 will be set at roughly half resistance, so the polarity shown will be correct.  However, this depends on the characteristics of the optocouplers you use, and it is possible that the voltage may be slightly negative.

There's a big difference between the ease of getting this version working compared to its predecessor (P49), as the only thing that you will need to adjust is the diode current.  VR4 and R20 set this.  If VR4 is reduced that will provide more LED current (and vice versa of course).  Increased current may be offset by the slower return to high resistance if the LEDs illuminate the LDRs too strongly (they have a 'memory' effect, and more illumination means they take longer to recover to a high-resistance state).  It might be necessary to change the value of R20 if VR4 can't be set for maximum modulation.

Unlike the P48 version, this unit can be used in the effects loop of an guitar amplifier, as it has the ability to operate at much higher levels without distortion.  It can be permanently wired inside the amplifier if you are building your own with the Project 27 boards.

The supply needs to be regulated, but typically a simple zener regulator is sufficient.  Figure 3 shows a suitable power supply ,and uses a 16V plug-pack type transformer - this gives the necessary voltages and ensures electrical safety.  An alternative power supply is the P05-Mini, which uses 3-terminal regulators.  While the supply shown here is adequate, P05-Mini is a better proposition for minimum noise.

Figure 3 - Power Supply Circuit

The power supply circuit can be mounted inside the main case, which is most easily a floor mounted unit, but the entire vibrato unit can be installed in the amplifier chassis if space (and front panel space!) permits.  In this case, use the existing preamp supply - as long as it is regulated and can supply the extra current.

D1 and D2 should be 1N4004 or similar, the zeners must be rated at 1W.  C1 and C2 need to be 35V caps, but C3 and C4 can be 16V or 25V units.  R1 and R2 should be 1W to ensure cooler operation.

In some instances, you might want to bypass the vibrato completely.  Depending on the setting of the 'Effect' control, there may be some change in tone that are undesirable.  Figure 4 shows how this can be done, with a SPDT switch.  A complete bypass (removing all circuitry from the signal path) is less desirable, because of changes in the impedance presented to the guitar when the vibrato in or out of circuit.

Figure 4 - Bypass Switching

The extra capacitors and resistors are to prevent clicks as the unit is switched in and out of circuit.  You might be able to get away without them, but for a few cents, it's not worth it.  The point marked U1.1 goes to pin 1 of U1, R3 is disconnected from the jack, and connected to the point marked R3, and the output goes to the jack.  The switch is shown in the 'Normal' position.

Main Index

Projects Index