Ask the Doctors: Gain Control, Part Two
By Si Moorehead
|
Part Two--LA-2A Gain Control
Much like the 1176LN, the gain control element of an LA-2A is a device that changes its resistance when a control signal is applied. It sits in the circuit in roughly the same place and shunts the audio signal to ground at a level determined by the sidechain’s control signal. There are a lot of differences however. The largest is that the gain reduction element is a light-sensitive resistor. The brighter the light hitting the sensor, the lower the resistance becomes. As the resistance lowers, more and more of the audio signal is shunted to ground, resulting in a reduced audio signal level. Photocells can be found in all kinds of applications, like turning lights on and off automatically, industrial control, and of course audio.
The other main part of the gain control element (a.k.a. T-4) is the EL (electro-luminescence) panel. This is basically little sheet of material that glows when a voltage is applied. These are used as nightlights (the kind that are really flat and glow a sort of weird green color) and as backlights for instrument panels. The magic happens when you put these two not-designed-for-audio pieces of electronics together. It just so happens that the photocells with the right chemical makeup have built-in attack and release settings that are great for audio.
There are no attack and release adjustments on the LA-2A because there is no way to control it. Unlike an 1176, the control voltage in the LA-2A is not a rectified version of the audio. It is the audio signal amplified to between 60 and 100 VAC and applied directly to the EL panel. Simple as it gets. The reason for the detail on the EL panel and photo resistor (T-4) is because it really is the heart of the LA-2A, and the 2-LA-2. Because the attack and release settings are purely controlled by the T-4, it is the T-4 that needs to track and be perfectly matched to have proper stereo operation.
As a side note, there are other ways to implement optical gain control elements. Using LEDs as the light source opens up the option of adding attack and release because you use rectified DC as the control voltage. These newer ways of going about it just aren’t the same, and don’t have that great program dependency that the LA-2A is famous for.
To make a stereo unit built around the T-4 requires a few important things. The audio path and T-4 driver circuit (sidechain) must be matched. We accomplish this using all 1% metal film resistors for solid matching and consistency over time. Most importantly, the T-4 components must be really well matched. They must have the same resistance level when the same signal is applied. Beyond that, they must have the same attack and release settings because there is no way to adjust this. It is a really long process to get a matched set of cells. It requires multiple passes through our purpose-built test fixture, and then testing in the unit itself. Our vendor supplies us with prescreened cells, most of which will work in an LA-2A. Out of these we often process hundreds of cells to get a few matched sets.
Also necessary for a robust product is a way to make sure that as the T-4 fades over time (and they all do, just a fact of T-4 life) there is a way to compensate for it. This is built in to the 2-LA-2 in the form of a trim pot. Although it shouldn't need to be touched for years, eventually it will. The final piece of the process, much like the 2-1176, is a complete stereo test and calibration routine using the Audio Precision test station. Like the 2-1176 test process, its importance cannot be over emphasized.
Like all UA gear, this stuff is built to last. We plan on them being around for many years, just like their mono brother.