It's common. Check this out! http://www.dsm.org/how-tos/tttt/tttt3.html

welcom to DSMs its going to be the your base idel set screw its a little $hit. Sounds like on your car maybe as its gets colder your car might be doing it more often do you just need to mess around with it a bit i dont think your screw has come out otherwise your car would be idleing really weird all the time

christina

Does it do it all the time or just when the engine is cold? If it settles down as the car warms up, leave it alone... or read the thousands and thousands of posts - all complaining of idle surge/rpm'?/high idle/low idle/etc.... etc....etc... [Get the drift ! It's the DSM MEMBER's curse !]
There are a couple of things that you can do to tame the idle. Go to that site mentioned in post 2, and surf around the postings to read all the idle problems.... and the advice.

That happens to me to, everytime I idle it will go up 2000rpm, by itself,
It could be two, it could be your idle set screw or it maybe your O2 sensers.

It's not going to be your idle set screw OR your o2 sensor. Think about it... an idle set screw DOES NOT MOVE. If your idle goes down to normal... then the set screw is set properly. If you o2 reads improperly it may stumble or miss on acceleration but your problem is that the engine is getting too much air.
Your problem is more than likely the AIS motor. It is a small motor with a shaft and a pointed tip on the end of it. That tip, depending on the position of the shaft, presses in to a hole shutting off air flow that by-passes your throttle plate. This motor is designed to open and close this passage way when the engine requires more air while you're not pressing the gas. Example... a/c compressor kicks on at a stop light, your engine needs more air, you've not got your foot on the gas. If it didn't have that motor... the added load would kill the engine. With this motor... the computer sees the compressor is being commanded to come on, it bumps up the step count on the idle motor, opens the air path more, car then can maintain idle while the a/c compressor cycles off and on.
Your problem sounds like it's not seating properly in to the throttle body. The tip of it is pointed and may no longer be sealing the whole all the way around. The end result is that the computer sees more air coming in, it's telling the idle motor to shut but it can't. There for the computer hunts for the right idle.
Doug

yeah doug could be right with that one

IT IS A VERY COMMON PROBLEM, BUT EASY TO FIX. THE AICS IS OLD AND WORE OUT. YOU CAN BUY A NEW ONE OR YOU CAN UNPLUG THE WIRE TO IT AND IDLE IT DOWN.(MAY WORK OR MAYNOT WORK IT VERYS). SO BEST OF LUCK ON THE TRADE MARK SOUND.

If you unplug it, you'll have a check engine light in your face, and still have idle problems. While the motor will no longer hunt for a proper place to be... it will stick in what ever position it was unplugged at. Which means it will either constantly idle high, or constantly shut off every time you take your foot of the gas.
The information I've provided as well as that link above is sufficient reason why to NOT settle for rigging it to work and actually fix something, right, the first time.

Do you have access to an ohm meter? ... You do! Good. Then here is how to test the electrical part of the idle control motor. Pull the plug and you'll see several contact points. I don't know the pin-out so lets call the contacts A,B,C,D,E,F.... There are 2 sets of coils. Each uses 3 pins. Fish around to find the 2 pins that read about 60 ohms. Let's call them A & C. Then find pin B. It is a 'center tap' so from A to B & from B to C should both read about 30 ohms [that's half of the total ohms, get it?]. Now find the other set of 3 pins [D,E,F]. Do the same test. They should read the same way. If you find that all these points read correctly, then electrically, the motor should be working correctly and the fault could be the mechanical part of the device.
...You still with me?
The mechanical section consists of a metal 'sleeve' with a low friction bearing at each end. Why easy moving bearings? Because the motor is not very strong. Pulses from the ECU 'step' the cylinder [rotating it]. Inside the cylinder is a screw thread that mates with the thread of a 'plunger' [looks like a rod with a bullet for the head]. As the cylinder turns it advances or retracts the 'plunger'. This bullet thing moves tightly against a seat [when extended to max] and stops air flow. What air flow? The flow that it allows to bypass the throttle plate with at idle [99% closed], when the thing retracts. If the parts are gummed up [motor has trouble turning the cylinder] OR if the seat & plunger thing have crud on them, then the air passage doesn't get sealed off completely AND the result is a high idle. Sometimes one of the coils is intermittently open, so the motor stops & starts causing erratic idle. Sometimes the gummed up stuff lets the cylinder turn intermittently so again - spastic idle.
There's more.........
Another part in the throttle assembly that controls idle is the cold idle air valve. The antifreeze circulates thru this valve, and the valve opens when it feels cold and closes when it senses hot. That's why the idle starts out high from a cold start in the morning & gradually decreases within a few minutes. It can fail also, causing strange idle, but that's not as common a problem as the first 'epic' that I wrote.
Fheeewwww! That was a mouthful, but anyway - that's how the thing works! At least in my '91. A little birdie told me that a 2G uses 4 wires!!! Perhaps Doug99RS could look in a manual to find out what coils and pin configuration you have.

2gnt uses a four wire connector. Digging through car to find service manual now.
4 wires
Gray with a dark blue tracer comes from PCM pin 48
Orange from pcm pin 49
Gray from pcm pin 57
Yellow with a dark blue tracer comes from PCM pin 58

They are all listed as driver circuits. Hopefully this and the previous post before mine here can aid you in your diagnosis.