• Provides automatic shutdown of unwanted noise during 'pause' conditions
• Compander technique eliminates 'signal snapping'
• User adjustable characteristics for high or low level network insertion
• Allows the use of otherwise 'too noisy to use' effects units
• Can effectively cancel crosstalk in multi-microphone set-ups
• Can be used in multi-instrument layouts for instant unit shutdown on changeovers
• Will eliminate 'beehiving' in older type 'spaghetti' wired organs
• No circuit trimming required or tight specification devices used
• Can be used in its own right as an effect to create soft attack bowing characteristic
• Uses only two low cost and readily available ICs
• Self contained, jack-in-jack-out unit allows instant in-line connection

PARTS COST GUIDE £14

In these modern times of multidevice usage, noise or crosstalk can often become such a nuisance that desired sounds often have to be forfeited. Although this unit can do nothing for the possibly distracting noises emanating from audiences, producers etc., it can considerably reduce, or even eliminate, pause noise from devices in which it is otherwise impossible to improve on signal to noise ratios.

Several types of noise gate are available for this kind of noise elimination - 'snap-off' units - programmable types - externally controlled units - low level expansion devices etc. In order that the unit may have as wide a range of applications as possible, the low level expansion or dynamic gate has been selected. This type tends to be less critical in set-up and general use, giving a more musically acceptable sound entrance and exit than the 'sudden shutdown' units.

Having a wide range of user adjustable characteristics, it should find many a useful working place with, for example, guitar / organ / keyboards / mic levels to mixer desk / P.A. / recording levels. Not only can it be used for its main purpose, that of closing down noise or unwanted signals below a selected level, but as an effect in its own right, creating soft attack, bowing type characteristics.

Noise gates have been in use for considerably longer than most people would imagine. In fact, the first application of these devices was in the 1930's, when they were used to reduce unwanted crackles and pops from soundtrack film that had become dusty or scratched.

The most popular use for the dynamic noise gate is that of removing undesirable noise during pauses in live or recorded performance. To achieve this, the threshold would normally be set just under the required signal level.

The unit is easily constructed and does not require any 'setting up'. As a self contained unit, it can be used in different applications at will, although being very compact it may be just as easily panel mounted for 'one-off' noise conditioning.

Circuit

Block Format

1. Jacking input connects the -ve battery line to the common earth line by the use of a stereo socket. (A mono jack will short ring and earth contacts on the socket to effect power up).

2. A high input impedance stage is used to prevent input device loading. The output of this stage serves also as the straight driving unit for pre-post comparisons.

3. A switched gain stage provides adjustable sensitivity and allows flexible device usage.

4. The first stage of the dynamic noise gate's active circuitry consists of a fixed 2:1 compression network.

5. The second stage of this active circuitry consists of a 1:2 expansion network with a simple resistive control element (potentiometer) for adjustable low level mistracking. This action is responsible for the total characteristics of the unit.

6. A passive attenuator is switched in conjunction with (3) such that an overall 1:1 input/output level is maintained. Included in the switch position is a straight-through route which allows comparison tests.

For those interested in such details, phase inversion between input and output does not occur.

Electronic Circuitry

Signals enter the +ve input of IC1a via the decoupling capacitor C1. IC1a is biased by R1 and R2 via R3 (R3 sets the input impedance level plus R1/R2). The voltage set by R1 and R2 ensures that outputs from IC1a and b will, at maximum levels, be evenly clipped. (At lower than normal V_cc levels maximum +ve excursions will be greater than available -ve excursions due to internal circuitry). R4, connected in the negative feedback line, equalises input biasing (R3=R4) for minimum output offset in IC1a.

IC1b is used in the non-inverting mode and has gain levels of 0dB, +6dB and +10dB, selectable via S1a, which introduces feedback reducing resistors R7 and R8.

Entry to the Compander section is now effected via C3, with compression achieved through IC2a and expansion by IC2b. All components in this section have been selected for the best overall performance in terms of frequency, speed of operation and distortion when bearing its 'musical' application in mind.

In the expansion section, RV1 is the resistive control element and operates by giving an increasingly false representation of lower level. Rectified signal voltages appear on C10, as RV1 resistance is decreased, causing an overall increase in attenuation rate.

This increasing attenuation characteristic creates low level expansion and is used to reduce the dynamic range of any signal within its domain with consequent drop in noise level. At higher signal levels RV1 becomes less effective in its role and allows a return towards original signal dynamics.

Returning to normality, S1c selects the resistors R15-R17 that are required to achieve an overall device gain of unity. S1a in conjunction with S1c also selects the output of IC1a, enabling pre and post 'gate' comparisons to be made.

Construction

The construction details relate to a cased unit, for panel mounting all that is necessary is to 'fly-lead' the input, output and RV1 wiring. Circuit board fixing is effected by the rotary switch on which the board is mounted. PCB assembly should begin with the two link wires.

Resistors, capacitors and ICs may now be fitted. Double-check polarised capacitor and IC orientation. Pins, or tinned copperwire should be soldered on to the board to hold RV1 in position, remembering that since RV1 is mounted above the board, components beneath it should lay flush to the board.

Solder the battery connector with the positive lead connected to the correct point on the board, and the negative lead to the centre tag of the stereo socket.

Fit the rotary switch next. This switch is normally obtained with solder type eyelet tags which need simple modifications for PCB fitting. Cut these eyelets off, as close to the solder hole as possible, then using snips, taper the ends to assist in aligning and fitting to the board. If the use of fixed pointer, push-on knobs is envisaged, remember to position the switch correctly bearing in mind the shaft 'flat' orientation.

Using tinned copper wire, earth, input and output leads can be connected to the board (using sleeving on the leads or slightly bending the wiring to avoid board shorts). After making sure all components have been fitted correctly, a few quick checks with a meter will verify correct basic operation. Switch to 100mA DC current range and connect the negative lead to battery and positive lead to earth. By touching the battery on to the connector a reading will be registered. A quick 'blip' and a reading under 10mA shows all is well. Voltage checks on outputs of ICs will confirm this - pins 1 and 7 on IC2 should be approximately 5V and pins 7 and 12 of IC2 about 3.6V. (These voltages are related to internal references that remain constant, and therefore allow for battery voltage reduction.)

All that remains now is to mount the unit in a suitable case. The metal case suggested in the parts list is ideal since it is small, easily workable, and provides good screening. Two appropriate holes in both the top and the side are all that are needed so that the whole assembly can just be fitted straight into the main case body and secured with the switch and jack sockets. An earth wire soldered to the edge of RV1 will ensure case earthing. If a plastic case is used, use adhesive backed metal foil tape (as used in pipe cladding) or glue household metal foil for internal screening.

Operation and Application

To obtain maximum usage of the unit, its functional characteristics should be fully understood. This can be achieved mainly by studying the response curves. The 1:1 gain slope reference allows visualisation of the deviation from normal input/output characteristics. The curve closest to the 1:1 gain slope shows input/output of the device when set in any gain position with RV1 set for minimum effect (clockwise). Note that just below -60dBm input levels, the output deviates more rapidly towards -85dBm. This is the operating region of the unit and any signal or noise below -60dBm will be attenuated, reducing its effective level. Any signal above -60dBm will have virtually normal dynamic range. Studying the curves of maximum effect characteristics shows that the noise gate can expand signals from even -15dBm down and completely shut down below -38dBm. (This setting will shut off most extraneous noises.) Since the compander section uses rectified signal levels in its operation, speed of recognition of these levels becomes a compromise between several factors, one of which is the loading of the circuitry by RV1. It should, therefore, be remembered that the unit will take a finite time to attack and decay and that these times bear a direct relationship to the threshold level (that level selected at which deviation from normal characteristics occurs), and change in amplitude of the input signal. For example, with the noise gate set at 0dB gain with input levels gated from infinity to 0dBm, attack times will vary from typically 2ms to 0ms for minimum to maximum threshold settings.

Before continuing with typical applications, it should be remembered that this unit is designed for 'pause' noise reduction, and has no magic ingredient for reducing any noise present in actual signals (see E&MM May 1981 for a noise reduction project).

Wherever noise exists, in-line connection of the noise gate should discriminate and further separate it from the required signal. Probably the lowest levels encountered will be from microphones and low output guitars (low Z mics will have to be transformed or the unit modified to suit). In the case of microphones, crosstalk (pick-up from sources other than that intended) will be the problem to cure. The +10dBm setting will allow maximum dynamic range with fast attack to be achieved when dealing with low level crosstalk (drum kit miking, awkward placement instrument miking, outside recording in windy conditions excluded).

Guitarists working in cramped conditions may encounter induced hum from closely situated amps, especially via single coil pick-ups. This often embarrassing situation can be alleviated by using the noise gate between guitar and amp., adjusting the threshold control for the best overall effect. Another interesting use of the noise gate is for changing the attack characteristics of an instrument. If the threshold level is taken to extremes, soft attack bowing type sounds can be created.

Similar treatment can be applied to special effects units (Echo, Chorus, Phasing, Distortion etc.), by using the noise gate between the last unit and its main amplification. The +6dB mode will be the norm in this application, since typical peak levels may introduce clipping.

Multi-instrument set-ups can often make the background noise unacceptable, considering that all units probably remain set at the required output level when only one or two instruments are actually being used at any time. Fitting individual noise gates to the noticeably 'noisy' instruments will automatically shut off their outputs when not in use.

Extreme levels of noise, or higher input levels, can be coped with in the 0dB mode, making this setting suitable for most line and mixer desk levels.

In the studio, the noise gate is often used with drum kit multi-miking, where as many as 14 microphones may be allocated to the kit in an isolation booth. A rather woolly sounding bass drum can be tightened up effectively by using a fast attack setting (further improved by coupling with a slow attacking limiter) that recovers some of the basic square wave response. It's even worth trying this method using an LFO signal source to create a synthesised drum sound.

'Dynamic reversal' effects can be obtained using the gate followed by a fast attack limiter. With slow recovery times the signal applied is thus strongly overlimited and during the limiter's slow recovery is gradually attenuated by the gate on a slow release setting. The resulting effect sounds like a 'backward' snare drum, tom-tom or cymbal.

It is also possible to reduce fixed delay times of instruments or reverberation treatments by suitable setting of the noise gate control parameters. One further application useful as a treatment in electronic music, is the 'keyed' or programmable gate where an external control voltage switches the signal being processed (used by Irmin Schmidt on his 'Last Train to Eternity' featured last month). But that's another project...