There are several different types of compressor topologies currently in the market, today. I will try to cover the basic topoliges, of each, and try to cover the basic "sound" of each of these units.

The first I'm going to cover is the "vari-mu" compressor. This has been recently resurrected by Groove Tubes (Glory Comp). The vari-mu compressor uses varying grid bias voltage (on the input tube) to change the operating point of the tube varying the tube from high conductance (high current) to low conductance (low current). Changing the operating point of the tube changes the tube conductance, or amplification factor (mu) of the gain stage. In the end, this varies the overall gain of the compressor.

There are other compressors, mainly guitar type compressors, that also work on this principle. The guitar compressors, rather than varying the grid voltage and operating point of a tube, are actually varying the base bias voltage on a transistor.

There are some inherent "problems" with vari-mu topology in that the low frequency notes tend to modulate the output. This causes intermodulation distortion and some other issues. The other "issue" that results, from this type of control, is that when a tube is run "hot" (high transconductance, or high output current) it typically sounds, to most people, as "warm", "full", etc. When the tube is "starved" by running it with low transconductance (low current) the sound typically is more "sterile", or "harsh". Thus, the varying operating point of the tube actually changes the sound of the program material as the compression is either applied, or removed. Whether this is good, or bad, is really up to the individual user. A lot of people really like hte vari-mu compressors.

Another type of compressor is the "current mirror" type. Many compressors, including the Alesis 3630, Peavey, some RANE compressors, Behringer, etc. use a "current mirror" chip made by THAT Corporation. Chips used in these devices range from the 2150, 2180 and 2181 series VCA's from THAT. All work the same way.

The input voltage is converted to current and input to the VCA chip. The output current is controlled by a voltage applied to the control voltage pin. The output current is fed into a current-to-voltage converter circuit. The VCA is used as part of the feedback loop of the current-to-voltage converter.

These units can be made to sound very good and, like their "opto" counter parts don't change the integrity of the original signal much. The "problem" wiht the Alesis 3630 compressor doesn't lie in the VCA itself, it's in the surrounding circuitry. Alesis used very low quality operational amplifiers and low-quality components feeding the VCA control lines. Coupled together, this causes the VCA to work differently than it should.

There are other issues, however, with these types of VCA's. Large quanties of low-frequency energy will cause the VCA to modulate the signal with the low-frequency energy. This can get ugly. Most compressor designers filter out at least some of the low-frequency energy to avoid this from happening.

Current mirror types of circuits tend to sound just a bit "dull" when compared to others. Compression pushes the signal level toward the noise floor which can cause masking of some of the subtle transient material. This dullness can be improved, significantly, by installing high quality, low-noise, operational amplifiers preceding the VCA and following the VCA. They will retain a "sparkle" and a "character" all their own, properly setup. Using low-noise, 1% metal film resistors will also go a long way in "cleaning up" the sound from these types of compressors.

The final type of compressor that we want to take a look at is the "opto-compressor". Opto-compressors come in MANY different formats. The venerable LA2A is one of those types.

In the LA2A the optical module is placed in parallel with the input gain control. When there is no compression going on, the resistance in the optical resistor is very high and all of the input signal goes into the amplifier circuitry.

When the input signal is high enough to trigger compression the resistance of the optical resistor begins to lower diverting more and more of the signal to ground, thus lowering the output of the end-result. In theory, this is the "cleanest" way to handle the signal, as the signal is not changed, at all, it is only varied in level, even when heavy compression is applied. The gain of an amplifier, or the output current of an amplifier is not changed as a result of the compression. The LA2A circuit acts as a variable volume control.

There are other optical compressors out there, that DO NOT work on this principle. They work more like the "current mirror" type of circuit. ART and Carvin are two examples of this that come to mind, immediately. The Pro VLA and the Carvin CG200 (if I remember correctly). In these circuits, the optical resistor, which controls compression, is located in the negative feedback circuit of the amplifier. In these devices, then, the signal level is not controlled, as it is in the LA2A, the gain of the amplifier is changed. This causes some artifacts to be produced, as a result of the differences in how the output signal is compressed. All optical compressors are not made the same.

The LA2A, and circuits that are made like it, have a sound that 's unique to them. Typically described as "clean", "clear" and "crisp" with "warmth" and "depth". They don't change the integrity of the signal, only the level of the signal, so all the transients, etc. go through the circuitry unchanged.


Those units, however, that vary the gain of an amplifier stage, using an optical resistor, will cause the transient distortion and "dull" the sound much more than something based on the LA2A.

In some cases, rather than using a light-based reistor to vary the amount of signal that reaches the first amplification stage, some designers used an FET - which acts like a variable resistor. Again, the placement of the FET varies by compressor topology. It can be in parallel with the input gain control (as in the LA2A), or it can be in a feedback loop varying the gain of an amplification stage somewhere in the compressor circuit. The end-result of the location changes the overall tonality of the device.


I hope that sort of helps ya'll understand a bit more about these monsters. Let me know if you have any questions. I'll be more than happy to try to answer them if I can.

Dar