This is a project from an older (early 1980's) Japanese projects book. As a design it is pleasantly simple with but two transistors in the audio signal path, and also sufficiently flexible in its' envelope/side-chain extraction that it can be customized a bit. Unlike the other noise gate project at this site, this one will work fine off a 9v battery and fit comfortably in a Hammond 1590BB chassis.
PARTS: The 2SC1000 transistor is high-gain low-noise and can probably be easily replaced with a 2N5089. The LM324 quad op-amp is not part of the audio path so don't bother replacing it with anything to "improve" functioning. If all you have is a TL076/84/64, that will probably work fine. The 2SK30A-Y FET is common enough, but can probably be replaced with the usual suspects (2N5457, 2N3819, MPF-102). Just keep an eye out for pin equivalents. The zener diode (RD5A) would appear to be any 5.1v zener, and the 1S1555 can be replaced with any standard silicon signal diode (1N914/1N4148).
MODS: It is often helpful to be able to adjust the degree of gating/attenuation so that the beginning and tail end of notes doesn't get shut off too hard. To add this function on the project, insert a 100k-250k variable resistor/pot between the source of the FET and ground so that you can adjust the amount of resistance posed by the FET/pot combination in parallel with the 470k resistor. As the pot value/setting is increased, making the FET turn on will still provide only moderate volume reduction. (The FET remains "on" until a note occurrence turns it "off")
The attack time can also probably be altered by varying the value of the 1k resistor after the envelope follower. It is currently set to produce a 5msec attack time, which is fine for most uses.
The 33k/10k-pot combination at the bottom provides a reference voltage for the comparator built around that op-amp. If your guitar provides a signal at the point where you insert this gate that does not let you use the Threshold pot's rotation easily/precisely enough, you can consider changing the value of the 33k resistor so that the reference voltage seen at pin 6 is higher or lower in range. A good strategy is probably to replace the 33k resistor with a 24k-27k fixed resistor and a 10k trimpot. The trimpot's wiper goes to pin 6and provides a bit more or less linear resistance to each leg of the voltage divider currently formed by the 33k/10k values.
SIZE: The scan appears to be about 5% or so smaller than scale, so be careful to rescale the PC layout beforehand if you are going to photocopy to Press-n-peel.
|