My 4L XJ16 XJS just failed the emissions test, (Canada). The shop changed 2 upstream oxygen sensors believing this to be the problem, but no improvement, and it failed again. 2 downstream not changed. both cats were changed 1500 miles ago, because it was felt they were the cause of an exhaust odour, (they weren’t). They now suspect the cars computer. Needless to say, it works out quite expensive to keep on replacing components at shop rates untill I get lucky. Can anyone offer an opinion on the likelihood it’s the computer, or suggest what else to look for before I go down that road? It’s a convertible, and an exhaust smell is still present when I stop at intersections etc. possibly unburned gas? No other serious issues with the car, no engine lights, the park idle does go to 1200rpm when the car is hot, but only 800 before it gets up to operating temp. I’m not concerned about the idle speed at this point, but mentioned it in case it may be connected to my emissions problem. I have obtained a pass/fail from the shop, so I am able to buy a sticker and drive the car, but since I will not be able to do that again, I need to resolve the issue. I used a gas additive claimed to help it pass, and ran 400 kilometers minutes prior to test. Any ideas?
Les
Kirbert
(Author of the Book, former owner of an '83 XJ-S H.E.)
2
Is this AJ6 or AJ16? Whichever, you might consider cc’ing this question over to the XJ40 or X300 lists, whichever involves the same engine. There might be more answers there.
I think that is the problem, acting on local advice I removed and cleaned the sensor terminals yesterday, Idle speed dropped to 7-800 rpm. I have another sensor on order, so untill I get it, I’m putting this on hold.
Many thanks for your help, Les
I found a new OEM Lucas coolant sensor on Ebay, and installed it a few days ago. Initial results were generally good, with idle in park remaining at 700 during several test runs, also, the smell of unburned gas disappeared, all in all, a great improvement, so I had the emissions tested again. It failed, with even higher numbers than before, but as an additional blow, the park idle was back up to 1200 along with the smell of unburned gas as I drove away from the testing center. I can’t help thinking the erratic idle is tied to the high emissions readings, any thoughts on why the idle speed varies intermittently, is it relevant that the idle speed, when stopped at lights,. and still in drive stays at around 7-800, and only goes higher when I put it in park?
Can’t find a problem with the coolant sensor connectors, but I’d like to stick with the idle speed problem for now, any more ideas? The idle speed only increases when the engine is hot, with a cool engine, it’s fine.
Your upstream oxygen sensors don’t do much, they’re only a check. The downstream sensors are the ones that tell the ECU how to adjust fueling ratios. If you had the upstream sensors, he didn’t really fix anything. A failing upstream sensor would have initiated a check engine light or throw a code. A failing, or lazy downstream sensor could cause more mystery problems without necessarily throwing a code or a CEL.
It’s interesting that the car runs well until it gets to operating temperature, or open loop. That helps (it probably eliminates vacuum leaks). Once you reach open loop, the following sensors affect fueling:
TPS
Downstream oxygen sensors
MAF sensor
Coolant Temp Sensor
Sounds like you’ve addressed #4.
The next free option you have is to swap upstream with downstream sensors since those were recently replaced. (Or you can just replace the downstream ones now, since it’s inevitable, especially if they’ve never been replaced before)
After that, buy some MAF cleaner spray and clean the MAF.
You can also take a voltage reading of the TPS. That might indicate if the TPS isn’t reading properly at idle? It should read at 0.5 volts at idle. It may have been reset at sometime, so a slightly different voltage may not be a good indicator, but it might be worth checking into.
You have them reversed; upstream oxygen sensors control fueling trim, and downstream oxygen sensors determine catalyst function. Both will set OBDII faults, but downstream would be catalyst efficiency only.
That is in direct disagreement with the service manual. They are backwards from other engines. Here the upstream are used to monitor catalysts, the downstream are used for fueling decisions by the ECM. Most cars do have the downstream sensors as the important ones, but according to the service manual, this is not the case with these cars. Please let me know where you found the service manual to be incorrect and I will correct it.
I suspect that OP’s mechanic also did not read the manual and is not aware of this peculiarity.
No way that the upstream can determine catalyst function; they are upstream of the catalyst. The downstream cannot be used for fuel trim, as catalyst will affect oxygen present in exhaust. All basic OBDII stuff.
Ok. I was just making sure you weren’t basing this information on some kind of document.
The service manual clearly states that the downstream makes the fueling adjustments. I’ve had plenty of discussions with people far more knowledgeable than me on this subject, when I’ve made a similar case to yours.
Without wanting to recreate that discussion, as the search feature is available to all, I would strongly suggest changing the downstream sensors.
There is documentation out there explaining how each of the sensors work, all you have to do is read the manual.
Likely a misprint; it is contrary to what is stated on page 36 of same publication. I do not wish to argue and will cease commenting, but not the first time I have seen errors in service publications.
I’m not arguing, I’m just trying to get to the bottom of this. I saw what you referenced at Page 38, however if you flip further to Page 47, you’ll read the following:
• The ECM uses the upstream catalyst oxygen sensor signals to provide closed loop injector control (except AJ16).
Don’t have the inline 6 engine in my XJS so can’t speak from knowledge. But following this discussion, and looking at the picture on page 81 of the linked manual, is it possible that the mechanic disconnected leads to both sensors (up and down), replaced the upstream sensors, and got the connections crossed when re-attaching?