Thanks again!
I'll try to remember to ask about brands when the car's ready. But yes, I'm rather sceptical. They're not just doing it to wring more money out of me, surely? I'm not (I hope) the usual "Here, something's wrong, fix it" type, but (try to be) the "This and this are happening at these specific times. My research suggests this or this could be the cause. Please investigate" guy. In other words, I hope I don't come off as some crazy fool with more money than sense (ah, to be in that happy situation!) who swallows any line of bull they feed me. The main thing, of course, is whether this does in fact solve the issue.... (And I guess the car won't be hurt by replacing both O2 sensors....)

I just don't know why the revs would fumble and then the CHECK ENGINE thing come on if it was just a faulty O2 sensor. That wouldn't affect the engine fuel ratio or revs or anything, surely? It's just a passive measurement, right? I've never read anything about it sending a signal requiring more boost or anything. Mind you, it does apparently need 450 revs of dodgy readings before showing, so perhaps that rev fumble is the engine attempting to switch fuel mapping, the O2 sensor not registering it, and the engine switching back to the original mapping a second before the car throws up the warning (it seems to run fine afterwards. No significant drop in fuel economy, either--14.7 mpg (US) on the nearly 700 km road trip last weekend, which had a fair amount of urban/suburban driving).

I'll give those ignition wires the push you suggest when I can. I've visually checked them, but haven't ever tried removing and replacing them--especially the front ones. I've pushed the dizzy ends before, but never rotated them, so I'll give that a go too.

 

If there is a misfire, then there is extra oxygen the exhaust, which gets sensed as lean. The computer adds more fuel, which then doesn't burn because it's too rich, but the sensor sees the unburnt air as lean and adds more fuel, etc. It can be a vicious loop until you hit the limits of the O2 sensors' authority and then things usually straighten themselves out.

The O2 sensors are not passive measurements. Perhaps you have heard the terms mentioned here of "open loop" and closed loop"? Open loop means the O2 sensors are ignores and the ECU injects fuel based on the load map programmed into it. It's like a carburettor in some regards, because it puts in fuel based on the program and does not check to see if that is a correct amount of fuel or not. Older ECu's like in the XJS will typically run open look when cold, and at idle revs. There is typically a rev threshold where it will transition from open to closed loop. This is done to improve idle quality, prevent hunting and stalling.

Closed loop means the O2 sensors are measuring the amount of oxygen in the exhaust, and from that the ECU can tell if it is rich or lean based on how the engine is operating at that particular moment. There is thus a feedback (closed) loop between measuring the exhaust gas, changing the fuel parameters based on that and measuring again etc. This happens many times a second when driving. So a bad O2 sensor can cause some weirdness, or typically if the ECU gets readings outside preset limits it ignores the O2 sensor and defaults to running on table values.

The plug wire twist might accomplish nothing, or something, but it's easy to do! With the amount of work that was done replacing the ignition components, they should be in good shape, but that doesn't mean a wire could still have worked loose.

For nearly 700 km, that seems like heavy fuel use. In my 6.0 I can get about 11.5 L/100km at a steady 90-100 km/h cruise and pure city driving is about 17-18 L/100km. Unless you were quite biased toward the urban driving rather than freeway?

 

More thanks!
Yes, I've come across the two loop types, and it's what I was trying to get at in my last post when I mumbled about switching engine fuel mapping--it's that change from open to closed I was referring to (or trying to refer to). So if the engine goes into Open Loop at idle, I wonder if that's why the CHECK ENGINE warning comes on often when it's idling...

So in Closed Loop, the O2 sensor does tell the ECU to change the fuel richness? Okay, I think I understand now. I understand better, at any rate. I'm going to read it several times and try to create a mental picture of what's going on.

There was a fairly heavy bias towards surface streets. Of the 697.5 km travelled (trip computer reset before heading out), I've used Google Maps to calculate 258 km of that was on expressways. Roughly a third. Average speed was 44.7 km, total fuel consumed was 111.3 litres, for an average of 6.2 km/l or 16.13 L/100km. One section we hit evening rush hour, managing just 55 km in 1 hour 50 minutes. And that was with a decent first half before we hit the traffic.
I should write up a trip report actually, illustrated with photos of Japanese temples in autumn. Might be of interest.

 

I don't think so, it will ignore the O2 sensors under open loop. I think. Older ECU's did, later ones might not.
However, one thing I thought of from my X300 is the air injection check valves - assuming you do have air injection? I'll have to look at your photos again. The air pump runs and the ECU is expecting to see a response from the O2 sensors, when it doesn't, it put up the code. On my X300 the check valves had rusted and the ECU wasn't seeing the proper response. A replacement fixed the problem. I have done them on my 6.0, they are at the back center of the engine.

Correct. That's part of how the ECU maintains the fuel mixture where it should be. There are lookup tables programmed into the ECU and first it measures manifold vacuum and RPM and then references a Volumetric Efficiency table to determine how much air is flowing into the engine. Then it looks up the Air Fuel ratio table to determine how rich or lean the mixture should be for that part of the demand curve, and then calculates the amount of on item the fuel injectors should have. Injectors are simply solenoid valves, and their flow rate is known at a given pressure and also programmed into the ECU.

So we have an engine that is running on calculated air flow and calculated fuel flow, not actual measured quantities. The O2 sensor is a trim, so that it can measure the actual results and then modify the injector on time to bring the actual air fuel ratio to what air-fuel lookup table. It's also partially used for altitude correction, as the air gets thinner with altitude. There is also a barometric sensor to apply the major correction.

Please do, sounds pretty!


That was all a long winded way of saying I was surprised the check engine came on at idle, but the ECU could have been doing a function test at the time and failed. Hopefully the new sensor will cure it.

 

Do I have air injection? Good question. How could I tell? I do have an air pump that comes on when the engine's cold. Where would the air injection check valves be? Back centre... where? I'll take a look when the car gets back.

I think I follow your wonderfully detailed explanation of the role of the O2 sensor in Closed Loop mode.The problem's when the measurements aren't read and the engine has to fall back onto pre-calculated quantities. I wonder why it's less likely to happen when the engine is restarted warm than from cold.... When it's warm, I assume it's straight into Closed Loop, but then that should mean the sensors are being read and it's more, not less, likely to glitch, surely? I mean, if Open Loop means Sensor Ignored, then when the Check Engine shows, that would be the change from Open to Closed and a failed reading from the sensor, leading to incorrect fuel-air mix, leading to quick bogging and then recovery as it defaults back to Open, which is less efficient and not ideal for the engine.... Is this what's going on, or am I more confused than usual?

I'll get onto that Trip Report as soon as I can. In the meantime, here's Her Ladyship at a temple carpark....

 

You do have air injection, I can see the air rails in your photo of the air filter element. The check valves are at the very back of the engine, between the engine and the firewall if I remember correctly. There is just enough space to get the needed large spanners in there. They are a standard GM part, used on many cars in North America, not that that is of any use in Japan. You won't be able to tell anything by looking at them, but if they are original probably worth changing on general principals.
Part Number 9: https://www.jaguarclassicparts.com/u...alve-5-3-litre

The common size is 5/8" hose, and from some digging in the parts book and cross referencing, the valve is the same as this one: https://www.rockauto.com/en/moreinfo.php?pk=40788&cc=1037315&jsn=797
If you need an application, it's the same as a 1992 Chev 1/2 ton truck with the 5.7.

The air injection system puts extra air into the exhaust on startup. This is get the catalyst up to temperature faster. The air injection ports are right into the cylinder head, you can see the yellowish metal tube on each side of the valley of the engine, runs parallel and outboard of the fuel rail. The check valve is to prevent exhaust from going back into the air system.


Regarding the hot/cold tendency of the light to come on, I have no idea why. I reread the description from the manual, and it says (paraphrased) if the sensor is at the maximum rich or lean when at idle and in closed loop, the ECU then adds or removes (as appropriate) fuel from the main lookup table. If the sensor does not respond to the change in fuelling, the code is set and it defaults to the table values without O2 sensor correction.

It could possibly be that when hot there is no warmup enrichment applied, so the fuel values are very close to where they need to be? When starting cold (say 0C) the engine will need about 75% more fuel than it does when fully up to to temperature. This enrichment tapers off with coolant temperature. I don't know exactly how the ECU is programmed, but it used to be that the system didn't enter closed loop until a particular coolant temperature had been reached. I think modern cars do it as soon as possible for better emission control.

Beautiful photo of your car, I love the red trees.

 

Ah-ha. So they're the pipes that run parallel to the fuel rails, then? And there's a cross-over at the back with a black rubber connector. Is the check valve located correctly on this photo of my engine?



Thanks also for the detailed writeup on what goes on with the air, and the links to the parts. Weird how it shows different choices for different markets, yet it's all the same part number. The hose would also need changing, then? Probably wouldn't hurt, anyway....
For what it's worth, I agree with your hypothesis about no warmup enrichment being supplied, so fuel values are close to standard. I seem to remember reading something about coolant temperatures being the controlling factor.

 

Yes, that's the check valve. There is one on each rail, so under the green connector should be another. The Hose is NLA, but I'm sure something could be made up if the old one is fossilized. Try and save it if you can to make life easier.

Yes, the only variable in the warmup enrichment is the coolant temperature.

 

I'll see how things look once the car gets back.
Thanks for all your help!

 

My pleasure.

 

Just for completion's sake, I did some more research, and for imported cars registered before April 1, 1998, the emissions limits are 4.5% for CO and 1200 ppm for hydrocarbons. (For JDM cars, the cutoff is a couple of years earlier.) Pre-1972, there didn't appear to be any limits. So keeping old cars legal in Japan is pretty hard at times. Mind you, pre-1972, Japan had a few pollution issues....

 

1200 PPM for HC's is a lot, my 3.8 E Type did around 250 ppm with no emission controls at all. As well tuned car should have no trouble meeting those standards.

 

Nice to know in case I ever fulfil a dream and am able to buy a Mark IX....

 

UPDATE:
She's back from the garage, all fixed (so far). It took them a month and a half to track down the problem, trying all sorts of different things (I also provided them with as much information on the issue as I could find on the forums, translated into Japanese).
In the end, replacing both left and right O2 sensors with Bosch ones rather than the ones they were supplied when they ordered replacements seems to have fixed the issue. The boss said that while the initial replacements did have their voltage fluctuate correctly, it was a much slower movement than the Bosch, so (in the garage's estimation), the engine would have done 450 rpm before the voltage had moved enough, throwing up the FF44 (and FF45) warnings as per the details I found on the forum about the mechanism. [Assuming I followed their explanation correctly of course....]

What they tried:
Extra earths to the mufflers and intake manifolds = X
Removal of intake manifolds, replacement of 12 gaskets = X
Secondary air pipe cleaned, sticking valve cleaned = X
Fast idle valve L-shaped hose was splitting; replaced = X
Spark plugs all replaced (one had lost the ceramic around the tip) = X
Left O2 sensor replaced = X
Both sensors replaced with Bosch units = YAY!

All that was 61,080 yen, or roughly US$557 (including several freebies like the extra earths).
It might be just my imagination after 55 days of the kei-car loaner, but she seems to run even more smoothly than before. Like gliding on glass at times. My wife also said she seems quieter.

They also replaced the heater core, as I'd been sitting on it for months, and the question of whether I would have a free day that was sunny so I could tackle the job was increasingly remote, especially as it's winter now. That was easy enough, as they could go in through the glove box and not even need to take the aircon CPU out. And they only needed to drain 2.5 litres of coolant, it seems. So I probably could have handled it myself, and did want to try, but I basically ran out of time before I needed the heater, and they already had the car, and I already had the parts, so it would only about US$250 in labour costs.

So, while I'm still a little nervous about it suddenly coming on again, for the moment she's purring like a kitten. Time to think about my next road trip....
And thanks to everyone here for all your advice. It was definitely useful.

 

I think I linked it, Let me know if it doesn't work.

I had the same problem. Replace all the O2 sensors, checked the ground, etc. What solved the problem (though might be a fluke) was when I repaired an exhaust leak by the exhaust manifold.,