Low-noise Guitar Preamp
UPDATE (12/4/23): Note the schematic shows the opamps as “741s"; I used TL071s for reasons discussed in the text. The pinouts are the same.
I’m going to be posting some various projects and designs I have been working on lately, the first of which is a guitar preamp build I did for a friend of mine. It was easy to build, using readily-available components[1]. I built the circuit on a small perf-board, with copper holes, arranged in a grid-pattern.
This preamp is designed to boost the very low-amplitude signal that emanates from the magnetic pickups of guitars — typically from under 100 millivolts to about 300, peak-to-peak — up to line-level, about 2 volts, without adding noise or coloring the sound.
First the schematic, which is very straight-forward. It consists of a single gain stage, followed by a buffer stage:
With the exception of the input, which blocks any DC present — you never know what someone will plug into an input — there are no reactive components, i.e., capacitors in the signal path, which ensures a minimum of phase-shifts and distortion. The input impedance is relatively high, essentially the 220k resistor in parallel with the input impedance of the op-amp, which is hundreds of meg-ohms. This guarantees minimal loading of the magnetic guitar pickups, which would affect the timbre.
In any audiofrequency amplifier, it is always a good idea to roll off the gain at the low end for stability, and here, the input high-pass filter has a cutoff of about 20Hz, so this preamp will work well for bass (40–1kHz) as well as guitar (80–1.2kHz).
The gain and buffer stages use the TL071 op-amp, which has a delightfully high slew-rate (20v/uS) to ensure none of the faster transients of the guitar envelope are lost. The gain (1 to 20) of the stage is set by 20k, linear potentiometer R1; it can easily be raised by replacing it with a higher value.
The second stage is a unity-gain buffer and keeps the load from interacting with the source. This stage will drive just about any load, adjustable with the 10k, linear level control, R2. Output impedance is 10k or less.
With the Gain control at minimum and the Level control at maximum, the gain of the entire circuit is unity.
C2 is optional and suppresses any RF coming into the circuit from local, high-power radio stations. D2 prevents a wrong-polarity AC adapter from damaging anything.
The power supply is rather unique and deserves some explanation. The filtered DC, 12 to 30 volts from an AC adapter (pure AC will also theoretically work — because of the rectification action of D2, but has not been tested), is filtered by the 220uF capacitor and then regulated by 78L08 linear regulator to 8 volts. It is then further filtered by the 33uF capacitor.
But here things get a bit different: the 8 volts is then split in half by Rail Splitter IC, TLE2426, which is a three-terminal device, consisting of a precision voltage divider (set at the 50% point), and an op-amp, configured as a follower. The output of this op-amp, bypassed by the final 33uF capacitor is a very stiff midway point between the input and its ground, which provides a “virtual ground” for the op-amps. We treat this as the ground in our circuit, with +8 now being the +4 rail; the negative power input provides the -4 rail, which allows the op-amps to be powered with +/- 4 volts, and thereby allowing DC coupling throughout and the elimination of DC-blocking capacitors.
The LED indicates power is on, controlled by the power switch.
For best results, build this in a metal enclosure being certain to connect the virtual ground to it and NOT the negative rail of the power supply.
Below are pictures of the prototype I built. I used SwitchCraft jacks throughout.
ljg
[1] The TL071 in the PDIP package is becoming hard to find (in 2023); SMT devices seem to be readily available.
