Okay, but if I unplug the O2 sensor I guess the frequency valve will stop working... am I wrong? Am I then supposed to check with O2 unplugged and freq valve not working?

Thanks John

John,

Yes, you are wrong. The FV must be pulsing at the default duty cycle for the fuel curve to be anywhere near correct for a 82 US 911 SC. That is the way it was designed for the fuel trim to operate and why it runs like crap when the relay or fuse fails. You cannot get near factory drivability or fuel economy without it. The correct CO spec for your engine is 0.4 to 0.8 %, in front of the cat. 3.4% CO is pig rich. My advice is to hook up the O2 sensor and read the duty cycle of the FV, open and closed loop, and determine if the lambda system is working as designed.

Fixed the O2 relay and engine is running amazingly. The only thing is that cold start is a bit more difficult than it was before (check my video 911scoppiettante - YouTube ): with no O2 relay I had an initial high revving (~1500 rpm) and perfect smoothness, then getting rough as revs went down and the well known spitting and misfiring, typical of broken O2 relay and non working FV. Now cold start is not as prompt and I won't get high revving at the very beginning, but it'll settle as low as ~900 rpm right away with only a little roughness in the first seconds/half minute, then it'll work already perfectly. Is it normal like so?

FV seems to be always on, at least idleing, when you can hear it vibrating standing close to the engine, but I may be wrong.

So I should hook up the voltmeter to the O2 sensor and read the values oscillating between around 0.2 and 0.8V when engine is idleing, will that be enough? What do you mean by reading the duty cycle of the FV?

Cheers

The FV is just like an EFI pulsed injector, it should be pulsing on and off for a varying amount when ever the engine is running. If your cold start rpm is too low, try unplugging and blocking the vacuum retard and resetting the idle to 950 rpm when hot. This gives a much stronger cold idle. The voltage test will tell you little. You need to hook up an analog dwell meter to the test connector under the left side cover in the engine compartment. It will tell you if the system is working as designed and once in closed loop, a proxy for making accurate mixture adjustments. If you do not understand this, try searching for "CIS lambda" and reading the posts.

Ok Paul, thanks again for your precious help!

Unplugging and blocking means what, taking away the blue hose and putting a cap on both sides? Should I cut a small piece of hose for both sides and tighten it with a clamp, if that's what you mean? Better solutions?

Take care