Mine (97) does this, too. It's less noticeable when the air conditioner is on (maybe the idle is raised slightly).

Your vibration may be different than mine, but I recently replaced the front O2 sensors and it had no affect on the rough idle issue. YMMV

 

I believe on of the earlier posters may have hot the nail on teh head here. Left to right bank timing issues. The earlier cars had the 964 type sprocket which used a steel "peg" to lock the timing wheel. From 1996? this want the case and talking with Colin, Steve and others it appears the engine can slip timing slightly with the newer hole less sprocket. Much has been written on these boards. This could explain why later Vario ram engines more often exhibit these symptoms...

 

Yes, but the 1996 and newer engines effectively allows infinite number of timing positions due to the friction-fit technique. Which, if I understand it correctly, allows the possibility of a finer tuned engine for left and right banks vs 964 cam sprockets. I'm not convinced this is the source of our rough idle for 993s. Perhaps others with more engine building experience can chime in on this.

 

If the idle valve is clean, the caps and rotors are less than 2 years old and if your car has over 70k on it - you have replaced the ignition wires and plugs at some point I would follow the old adage: “accept the things I should not fix, have courage to fix the things I must and have the wisdom to know the difference”. The 993 has fixed valve timing, all old school 911's shake a little at idle. If it was a material misfire issue your OBD would throw a code.
Andy

 

I have had for many years similar problems running the engine at idle, after a sleepless week, I was able to measure the lambda value in the exhaust ( there are two fittings in the middle of the exhaust, you are able to attach the sond easily with an adapter ). At idle the lambda was far to high, that means there was too much oxygen for ignition and that causes the rough running ( the lambda was at about 18 , it should be at 14.7 ), while driving everything was perfect close to 14.7. I observed that the connector to the air mass flow meter was not reliable, while I was shaking by hand the connector. Cleaning the connector solved the problem. Hope this hint helps and now my engine is running smoothly.

 

Can someone explain how to test for a vacuum leak? I just changed the cap and rotors on my '97 C2 with 48,000 miles. After doing that it idles more roughly than ever before (which previously was barely evident) but is not throwing any codes. So based on this thread, testing for a vacuum leak seems like a good start in diagnosing the rough idle. I haven't changed the plug wires or cleaned the ISV valve. Any suggestions would be appreciated.

 

Deleted duplicate post

 

Start with the cheapest easiest possible fix first. Remove the isv and clean it thoroughly with carburetor cleaner (Not brake or electrical contact cleaner)

To find vacuum leaks, the car should be smoke tested... a good independent air-cooled specialist should be able to perform this for you.

Next, possible lazy/old O2 sensors, and/or plug wires. I recommend OEM Beru plug wires.

Old/failing dual mass flywheel.