Pneumatic Instrumentation - Pilot Valves and Pneumatic Amplifying Relays

Self-balancing mechanisms such as the fictitious pneumatic laboratory scale in the previous section are most accurate when the imbalance detection mechanism is most sensitive. In other words, the more aggressively the baffle/nozzle mechanism responds to slight out-of-balance conditions, the more precise will be the relationship between measured variable (mass) and output signal (air pressure to the gauge).

A plain baffle/nozzle mechanism may be made extremely sensitive by reducing the size of the orifice. However, a problem caused by decreasing orifice size is a corresponding decrease in the nozzle’s ability to provide increasing backpressure to fill a bellows of significant volume. In other words, a smaller orifice will result in greater sensitivity to baffle motion, but it also limits the air flow rate available to fill the bellows, which makes the system slower to respond. Another disadvantage of smaller orifices is that they become more susceptible to plugging due to impurities in the compressed air.

An alternative technique to making the baffle/nozzle mechanism more sensitive is to amplify its output pressure using some other pneumatic device. This is analogous to increasing the sensitivity of a voltage-generating electrical detector by passing its output voltage signal through an electronic amplifier. Small changes in detector output become bigger changes in amplifier output which then causes our self-balancing system to be even more precise. What we need, then, is a pneumatic amplifier: a mechanism to amplify small changes in air pressure and convert them into larger changes in air pressure. In essence, we need to find a pneumatic equivalent of the electronic transistor: a device that lets one signal control another.

First, let us analyze the following pneumatic mechanism and its electrical analogue (as shown on the right):

As the control rod is moved up and down by an outside force, the distance between the plug and the seat changes. This changes the amount of resistance experienced by the escaping air, thus causing the pressure gauge to register varying amounts of pressure. There is little functional difference between this mechanism and a baffle/nozzle mechanism. Both work on the principle of one variable restriction and one fixed restriction (the orifice) “dividing” the pressure of the compressed air source to some lesser value.

The sensitivity of this pneumatic mechanism may be improved by extending the control rod and adding a second plug/seat assembly. The resulting mechanism, with dual plugs and seats, is known as a pneumatic pilot valve. An illustration of a pilot valve is shown here, along with its electrical analogue (on the right):

As the control rod is moved up and down, both variable restrictions change in complementary fashion. As one restriction opens up, the other pinches shut. The combination of two restrictions changing in opposite direction results in a much more aggressive change in output pressure as registered by the gauge.

A similar design of pilot valve reverses the directions of the two plugs and seats. The only operational difference between this pilot valve and the previous design is an inverse relationship between control rod motion and pressure:

At this point, all we’ve managed to accomplish is build a better baffle/nozzle mechanism. We still do not yet have a pneumatic equivalent of an electronic transistor. To accomplish that, we must have some way of allowing an air pressure signal to control the motion of a pilot valve’s control rod. This is possible with the addition of a diaphragm, as shown in this illustration:

The diaphragm is nothing more than a thin disk of sheet metal, upon which an incoming air pressure signal presses. Force on the diaphragm is a simple function of signal pressure (P) and diaphragm area (A), as described by the standard force-pressure-area equation:

                                                           F = PA

If the diaphragm is taut, the elasticity of the metal allows it to also function as a spring. This allows the force to translate into displacement (motion), forming a definite relationship between applied air pressure and control rod position. Thus, the applied air pressure input will exert control over the output pressure. The addition of an actuating mechanism to the pilot valve turns it into a pneumatic relay, which is the pneumatic equivalent of the electronic transistor we were looking for.

It is easy to see how the input air signal exerts control over the output air signal in these two illustrations:

Since there is a direct relationship between input pressure and output pressure in this pneumatic relay, we classify it as a direct-acting relay. If we were to add an actuating diaphragm to the first pilot valve design, we would have a reverse-acting relay as shown here:

The gain (A) of any pneumatic relay is defined just the same as the gain of any electronic amplifier circuit, the ratio of output change to input change:

For example, if an input pressure change of Δ2 PSI results in an output pressure change of Δ12 PSI, the gain of the pneumatic relay is 6.

Adding a pneumatic pressure-amplifying relay to a force-balance system such as our hypothetical laboratory scale improves the performance of that pneumatic system:

Since the relay amplifies the nozzle’s backpressure, the force-balancing bellows responds even more aggressively than before (without the relay) to any change in baffle position. This makes the scale more sensitive, better able to sense small changes in applied mass than without an amplifying relay.

The Foxboro corporation designed a great many of their pneumatic instruments to used a very sensitive amplifying relay:

The motion of the diaphragm actuated a pair of valves: one with a cone-shaped plug and the other with a metal ball for a plug. The ball-plug allowed supply air to go to the output port, while the cone-shaped “stem valve” plug vented excess air pressure to the vent port.

The Fisher corporation used a different style of amplifying relay in some of their pneumatic instruments:

The gain of this Fisher relay was much less than that of the Foxboro relay, since output pressure in the Fisher relay was allowed to act against input pressure by exerting force on a sizable diaphragm. The movable vent seat in the Fisher relay made this design a “non-bleeding” type, meaning it possessed the ability to close both supply and vent valves at the same time, allowing it to hold an output air pressure between saturation limits without bleeding a substantial amount of compressed air to atmosphere through the vent. The Foxboro relay design, by contrast, was a “bleeding type,” whose ball and stem valves could never close simultaneously, and thus would always bleed some compressed air to atmosphere so long as the output pressure remained somewhere between saturation limits.