The housing for the AAV also features a passage that bypasses the cylinder port.  Air can thus come off the bottom of the inlet connection, into the chamber at the bottom of the AAV cylinder, through a circle of six holes in the piston itself and/or through a relief in the side of the cylinder behind the piston to the top of the cylinder, and out the outlet connection to the engine.  The first portion of this route, the passage that goes from the inlet connection to the bottom of the cylinder, has a tapped hole into it arranged in such a way that a bolt threaded into this hole will obstruct the passage, depending on how far the bolt is threaded in or out.  This is the warm idle adjuster.

The AAV has two typical failure modes: the piston itself gets jammed in the cylinder, or the thermal bulb expires.  The piston can be jammed in any position, either causing a high idle when warm or a low idle when cold.  The thermal bulb failing, however, always causes the piston to remain closer to the cold position than it should, so the effect is always an idle that is too high.  Of course, the mechanic typically just adjusts the screw; this will work for a while, but eventually the thermal bulb will give out altogether and the idle will be too high with the screw threaded in as far as it will go.

Note: when the throttle pot is at idle on the Jaguar V12 with EFI and the engine RPM is above 1800 or so (for whatever reason), there is a feature of the EFI called "overrun cutoff" that kicks in.  The fuel is cut off totally, the injectors don't get any signal.  This feature is supposed to save fuel and reduce emissions when the driver takes his foot all the way off the accelerator when coasting.  The EFI system will bring the fuel back on when the RPM drops past 1000 or so, so the engine will not die as the car coasts to a stop, but rather will enter a smooth idle.  When the idle is too high when sitting still, however, the overrun cutoff can be engaged when the car isn't moving.  The result is an engine that quickly revs up to 1800 or so, then suddenly drops back to 1000, then back up to 1800, and so on.  This situation is very disconcerting to the owner, and may be to the mechanic as well if he's not familiar with it.  They may even instigate some expensive and ill-conceived "repairs" to attempt to deal with this cycling idle. All that is necessary is to get the idle down; the cycling will go away as soon as the idle is lowered below the overrun cutoff set point.

If you'd like to test whether or not your AAV is causing your high idle conditions, warm the engine up, shut it off, remove the left side air filter cover and element and restart the car.  With it idling too high, cover the opening leading to the AAV.  If the AAV is your problem, your idle should drop precipitously -- and if you had that cycling condition, that should cease as well.  Your engine should actually idle too low, although it will be sucking air like crazy through the distributor vent scheme to maintain idle.  If you cap off the filter on the end of the hose that serves as the intake for the distributor vent scheme, the idle should drop even more.  If these actions don't have a significant effect on reducing the idle, your problem is probably caused by vacuum leaks elsewhere -- and they may have become audible during this test, making a telltale hiss.

If your AAV is defective, the first thing you might attempt is to remove the hoses and squirt some carburetor cleaner into it in hopes of freeing a stuck piston.  Mike Morrin suggests, "Going along the non-destructive path, if you look down the bore, you will see some smallish holes drilled in the piston.  I managed to fit a long self-tapping screw into one of these, and used it to pull the piston up against the spring a few hundred times, while applying WD40 to wash out the crud.  It seemed to work, and I have had no problems in the following 12 years."

If that doesn't work, you might try removing the AAV and soaking the entire contraption in solvent.  It would appear that removing the AAV would involve draining the cooling system at least far enough that the fluid level is below it, but Craig Sawyers reports otherwise:  "Not a drop came out when I replaced mine.  In fact the ROM procedure doesn't mention draining prior to replacing the AAV."

If the piston moves smoothly enough, your problem might be the bulb itself.  Test it in a pot of water on the stove, making sure the bulb itself doesn't contact the bottom of the pot as the water is heated.  Well before the water reaches the boiling point -- actually by about 80ºC, if you have a thermometer -- the piston should have moved as far as its going to.  If it is still only part of the way towards closing the port, try this: let the AAV cool down to room temperature and then carefully crush the bulb at the bottom a little bit with a big pair of pliers (you'll note that it was already crushed a bit; perhaps some calibration procedure at the factory).  At room temperature, the inside bottom of the piston should sit at 70mm below the upper surface of the outlet hose fitting.  If it measures 71mm or more, carefully crush the bulb just until the piston comes up to 70mm and quit.  Then try the stovetop test again.

If all of that doesn't work, Jaguar expects you to buy a new AAV.  This is reasonable; the thermal bulbs in thermostats don't last forever, it is common practice to replace them periodically.  Why should the AAV be any different, since it uses a similar thermal bulb?

Because the AAV costs a lot more than a thermostat, that's why!  When you get off the floor after checking the prices of new AAV's, consider the possibility of rebuilding your AAV.  This site includes ideas on replacing the bulb in your AAV with one yanked out of a commonly-available thermostat.  This is not a perfect solution; read about the theory and implications before deciding what you want to do.  If you decide to go for it, there is a page on disassembling and reassembling the AAV, selecting a donor thermostat, examples of AAV baseplate rebuilding using a Robertshaw thermostat and a CarQuest thermostat, and some other ideas for repairs.

Note that the AAV seem to share another trait with thermostats: if the engine has been overheated, it's probably ruined.  The AAV contains a nitrile seal in the baseplate, so getting the engine too hot will make it brittle the same way it will for O-rings.  A leak here would cause coolant to enter the intake system.  There is also a sleeve within the thermal bulb itself, and it may also get hard, diminishing the response of the AAV.

Buying an AAV from a junkyard, of course, is just asking for trouble.  Would you buy a thermostat from a junkyard?  Of course not.  In this case, perhaps the only reasons to buy a used AAV are to get one with a different port configuration than yours (see the page on port configurations) or to get one to practice rebuilding before you tackle your own.