Having more or less continuously had an average fuel consumption display reading of between 20 and 22 mpg in my beloved XJR, it has now dropped down to around 16mpg. I noticed this on a predominantly motorway journey which I often make when the fuel gauge dropped from full to three quarters after around 65 miles. I initially concluded that I couldnt have topped the tank properly, but on the return journey, another quarter tank was consumed. On checking the trip computer, I reset it, and went for a brief MPG friendly drive on a dual carriageway: average 16.2mpg.
The car is running like a charm (on 98 octane) and pulling really well. Idle perfect at 675 rpm, no warning lights, oil and temp readings normal. No leaks or smell of fuel.
I put her on the scanner today, and no codes showing. I ran her up to temperature and looked at the live data. Coolant and IAT readings realistic, and she fairly quickly moves into closed loop. Long term fuel trims both show -11.7%, and short term both move to +24.2% when in closed loop and stay there. My scanner doesnt register any MAF reading (it never has done) but shows O2 sensor 1 with 0 volts, and sensor 2 oscillating between 0 and 0.005 volts indicated. The air cleaner is relatively new (paper) and removing it makes no difference. The exhaust looks and sounds ok.
My current theory is that O2 sensor 1 has given up, and in the absence of a signal the fuel trims have defaulted to fully rich. The absence of a code and the fact she goes into closed loop are undermining my confidence in this theory though.
Any ideas anyone?

 

Your theory on the O2 sensors would be my prime suspect also. I don't think working O2 sensors are a requirement for Closed Loop so I wouldn't let that put you off.

However, my understanding of how the the sensors and voltages work is that High Voltage is Lean and Low Voltage is Rich. So, if your voltage is Low (Rich) you'd expect STFT to be negative value. The following is from the Jag Training manual for the AJ16 Engine Management system

Whenever the fuel / air mixture ratio passes 14.7:1 (Lambda = 1), the sensor delivers a voltage swing:

Air / fuel mixture leaner than 14.7:1 – reference voltage high (maximum 4.89 V)

Air / fuel mixture richer than 14.7:1 – reference voltage low (minimum 0.015 V)


There's a "but" here. Some years ago when doing some testing on sensors, I discovered that ELM327 based readers report low voltage when actual signal voltage is high. I tested this on 3 different readers and all did the same thing. If you're using ELM327 type reader that would probably explain the discrepancy.

Here's a picture of a test where a Lean condition was induced on the right of the graph. You can see on the phone graph that O2 Voltage is reported as 0 and STFT is max'd positive on both banks. You can also see on the multimeter that the actual voltage being read directly from the sensor is high.









If you want to try some diagnostic part swapping, I have 2 working O2 sensors that you can borrow. I replaced these a few months back only as preventative maintenance (I think they're original sensors and are 26 years old, but I know they were functional). PM me if you want to do that.

 

Brendan,

Many thanks for taking the trouble to respond in detail, and for your very kind offer to lend me some good O2 sensors.

Since writing the original post, but before reading your reply, I came across a youtube video which purported to a general tutorial on O2 sensors, and which suggested that low was lean. As this suited my hypothesis, I ordered two new sensors, figuring, like you, that they wont last forever,( 25 years and 170000 miles isnt a bad innings! ) and replacing them wouldnt do any harm regardless of whether they are the cause of the current issue. We shall see.

The excerpt from the training manual has me wondering though. My scanner is badged as an AUTEL , which I am given to understand is a reputable make, but although it wasnt inexpensive, it was sourced directly from China, and so might be a "lookey-likey" AUTEL...... The possibility that it could be reading high as low would be less than helpful. Probably the easiest way to find out at this stage will be to see what it makes of the new sensors.

I will report back hopefully in the next few days, but in the meantime, thank you again for your help, and most particularly for your kind offer of a loan of your good sensors.

 

I’d say that oxygen sensor “wonkiness” would (almost?) never be a reason to keep an engine in open loop. If that was the case, then a you could only rely on a code to indicate a bad sensor, AND the car would have to remain in open loop indefinitely. That would be your only clue that might lead you to figuring out an oxygen sensor was bad.

The previous XJS OBD2 system is very stingy with codes. I’m not sure if that changed for the x305.

As you said, the sensors you have now owe you nothing. Even if it isn’t the cause, and I believe it will be, those things have given you everything they had.

 

Happy days!
Replaced both O2 sensors, and short term fuel trims immediately shifted from being maxed out at +24.2% to the low single digits negative at idle. Took her for a drive, and average fuel consumption restored to just over 21 mpg.
Case closed!
Thank you again for your kind offer Brendan

 

Good news.

For your future reference (and ours) it would be interesting to see what you now get on your reader by inducing a lean condition.