'Actual consumption' values...

[bazzbazz]

That's a good price for a 'correct' sensor. Doesn't mean I can afford two of them just now with other large outgoings looming, unless I put it on the dreaded plastic...

Yes, "being about as financial as a Greek Bank!" is something I too am all too familiar with. 

[Citroënbender]

A universal Bosch or NTK sensor will work fine with one caveat - the wiring connections must be 100%...

Also, I am not convinced it will give meaningful data. Changing one sensor is highly unlikely to provide a change in fuel consumption that can be simply correlated to a notional "before" number.

[johnl]

A universal Bosch or NTK sensor will work fine with one caveat - the wiring connections must be 100%...

Of course. So very clean wires / 'crimp' connectors, and, make sure that air can get in between the wire insulation and the wire metal, so no solder that could conceivably block airflow or tight fitting shrink wrap covering the point where the insulation ends.

Also, I am not convinced it will give meaningful data. Changing one sensor is highly unlikely to provide a change in fuel consumption that can be simply correlated to a notional "before" number.

I'm not sure I follow what you mean. If an O2 sensor is misreporting then the ECU has no way to 'know' this (if ECUs have a god, it is the O2 sensor, which they tend to believe 'religiously'). I've had an O2 sensor failure (on the Saab) so bad that the plugs would foul very quickly and the engine run obviously very rich and rough, with extremely poor fuel economy and low power, but still no engine check light. My experience suggests that ECUs seem fairly blind to O2 sensor faults, with faulty sensors generally reporting lean, so the ECU inappropriately and (possibly) enthusiastically enriches, with substantial impact on economy. Changing a bad sensor for a good sensor would surely improve this (certainly did with the Saab, dramatically).

In this case (my 147) I'm suspecting the sensors are not so much reporting lean, but reporting very slowly, so failing to report a lean condition quickly enough when the throttle is opened (airflow increases, so AFR goes lean until ECU corrects it, lazily), and then failing to report a rich condition quickly enough when the throttle opening is lessened (airflow decreases, so AFR remains rich until ECU corrects, eventually...). The ECU can only correct AFR based on 'data in' from the O2 sensors. Slow reporting of 'lean' seems not as slow as glacial reporting of 'rich'. In either case I think the sensors probably eventually get near enough to the correct voltage output, just way too slowly.

As should be clear from what I've been saying, this suspicion is not directly informed by data that might be available if I could interrogate the ECU, I'm just making inferences based on the behaviour of the 'actual consumption' readout (which I'm fairly sure is letting me know that something is not as it should be, and from there I'm having a stab...). This doesn't preclude the possibility that I'm barking up the wrong tree, and that the issue could instead be the MAF rather than the O2 sensors, or a combination of both MAF and O2 sensor...

Regards,
John.

[Citroënbender]

It's going to be at the earliest Saturday night before I can respond to this in more detail. But the thrust of my comments is as stated prior; it's about conservative spending for concrete gains vs punting.

[Citroënbender]

I've had a brief skim of your prior posts and note:

Thermostat fitted 2016-ish - this should still be OK.
Two injectors replaced due to being "stuck open".  This could predicate bore wear, surely? 
Can't find your compression test results (hot, cold, wet, dry) from the last few years.

So maybe if we could start with some data there, using a screw-in type gauge?

[johnl]

Two injectors replaced due to being "stuck open".  This could predicate bore wear, surely? 
Can't find your compression test results (hot, cold, wet, dry) from the last few years.

So maybe if we could start with some data there, using a screw-in type gauge?

Hi CB,
I doubt that the two injector failures would have had time to cause appreciable bore wear. The failures (appeared to be) very sudden, one day the engine was running fine, the next day black smoke and very rough running. In fact so sudden that the two journeys in which both failures occurred started with the engine running well, but ended with an engine that abruptly became sick near the end of the trip. It would not have been many kms driving with the bores being 'washed' of oil, just a handful. Note that each failure was in a different cylinder, so no cumulative KMs for any cylinder.

Having said that, a compression test isn't a bad idea, not that I have any strong reason to suspect poor compresion as the engine runs very well when it's running well (usually in colder ambient temperatures, it's on hot days that the performance tends to become a bit 'doughy'). Now, I have owned at least two compression testers over the years, the first one died, the second one just disappeared at some stage. I do need to buy another one, sometime...

My engine does use oil (big surprise there...). I suspect it's likely to be the valve stem seals, since the only time oil smoke is ever evident is at start up (and then only occasionally). Oil consumption is rather sporadic, the engine might go a week or two without needing a top up, then might lose half a litre in a couple of days. Whatever, this does mean that some oil will be getting into the exhaust system, and thus onto the O2 sensors. I have two new O2 sensors on the way from the UK.

I'm considering changing the oil to a different oil, because the 'fully synthetic" Penrite oil I've been using has a significant zinc content (very good for lubricating sliding surfaces, such as cam lobes, the most highly loaded / stressed surfaces in an engine). It's enough that you can smell it in the exhaust smoke (when smoke is actually evident at start up). The smoke smells distinctively like differential oil, which has a very high zinc content to protect the 'hypoid' crown and pinion gears as found in most RWD diferentials (hypoid gears slide substantially against each other as they rotate).

My thinking is that zinc is known to be bad for O2 sensors, so an engine that burns a significant amount of zinc laden oil may well be at added risk of damaging it's O2 sensors....

Regards,
John.

[bazzbazz]

The smoke smells distinctively like differential oil

Ah-ha. Pull the upper two O2 sensors and have a peek down the holes and take a look at the Catalytic Converter cores. If you can't see, borrow a inspection camera or just use something appropriate to insert down the O2 sensor hole to see how far it goes. You may be surprised to see just how far it goes down, as in like "It should hit the core by now . . . . .it REALLY should have hit the core by now??"   

[johnl]

Hi Bazz,
Why would some issue with the pre-cats cause burning engine oil to smell like zincy differential oil?

I did just find an oil film on one of the header pipes, where a drop of oil had obviously dripped onto the pipe (and spread out). This might explain some of the zincy oil aroma I have occasionally smelt (though I haven't previously seen evidence of oil on the pipe, I assume because it has burnt off). This is annoying because the cam cover gasket must be leaking, and it's only fairly recently been replaced...

Anyway, as per your suggestion I pulled the O2 sensors and peeked in. What I found was, nothing suspicious. The core matrixes (matrixi ?) both look clean, unobstructed, unmelted, uneroded and otherwise intact, with a mid to darkish grey colour. I couldn't see the entirety of the matix 'faces', only what the hole permitted (I don't have a techy little camera...), so this might not be a 'definitive peek', but I'm pretty sure what I couldn't see would probably look the same.

Something I did note, that I also saw the last time I pulled the sensors out (i.e. seen it more than one time, so it seems likely to be a fairly consistent thing...);

One of the probes (right side as viewed from the front of the car) is a fairly uniform light grey colour over most if its' surface (other than the very tip, which is a significantly lighter grey). The other probe is more or less the same, but for a quite distinct / defined band of dry black sootiness about 5 or 6mm in width around the entire circumference of the cylindrical part of the probe (next to the threads). Both sensors are quite dry, no sign of oiliness.

This at least suggests that something in some way somewhat different is happening to the AFR in cylinders 1 and 4 as opposed to 2 and 3. Since one sensor is associated with cylinders 1 and 4, and the other associated with 2 and 3, is it a stretch to suspect that this might not be related to some difference between the way in which one sensor is behaving (voltage to ECU) relative to the other sensor...? I tend to think not...

Regards,
John.

[johnl]

This seemingly persistent band of black soot on only one sensor suggests (to me) that at some point this sensor is probably having soot deposited over the entire probe surface, which then gets burnt off depending on how hot any given part of the probe may get.

The part of the probe directly adjacent to the wall of the exhaust (pre cat canister) would be kept cooler due to it's proximity to the relatively cool wall (which will be radiating substantial heat away, 'pulling' heat from the probe). Exhaust gasses will also tend to be hotter toward the middle of the exhaust flow, and cooler adjacent to the wall. Soot deposited on the probe near the wall would be less likely to burn off. The parts of the probe that are further from the wall, and more directly in the strong flow of the uncooled hot exhaust gasses, would remain hotter, so carbon deposits would burn off much more readily from those parts of the probe.

Questions might be; due to a possibly richer AFR in cylinders 1 / 4 relative to 2 / 3, is one probe being deposited with unburnt carbon and the other not? Or; are both probes having carbon deposited on them, but one is getting hotter than the other due to a difference in AFR (and thus exhaust gas temperature) between cylinders 1 / 4 and cylinders 2 / 3? I suppose either is possible.

Can't wait for my new sensors. But I'll be disappointed if they make zero difference...

Regards,
John.

[bazzbazz]

Have you pulled spark plug #1 recently and checked it?

If you pull plug #1 and it is oily and all the rest are fine you will find that you have a failed inlet manifold gasket around the oil gallery area for the Variator Solenoid.

This is an old problem in these engines. Read post #4

http://www.alfaowner.com/Forum/alfa-147-156-andamp-gt/196687-156-ts-oil-contamination-in-inlet-manifold.html

[johnl]

Have you pulled spark plug #1 recently and checked it?

Yes, a few weeks ago. All plugs looked OK and similar. They are relatively new plugs, well under the 100,000 km interval. Wouldn't hurt to have another look at them.

Regards,
John.

[johnl]

Have you pulled spark plug #1 recently and checked it?

If you pull plug #1 and it is oily and all the rest are fine you will find that you have a failed inlet manifold gasket around the oil gallery area for the Variator Solenoid.

So I checked the spark plugs, which all looked OK. (I also took the cam cover off, cleaned everything, and refitted the gasket smeared with a liberal amount of anaerobic 'gasket maker', which has fixed the annoying oil leak).

I received the new O2 sensors some days ago (RTG brand 'universal' sensors, that needed to be fitted to the old plugs with crimp connectors).  After fitting them I zeroed the 'average consumption' readout.

Afterward, at first I was seeing 'average consumption' readouts quite a bit lower than before the sensor change, around 8.3L/100km give or take (as opposed to a consistent 8.8L/100km with the old sensors). But, after a few days and a few hundred km of driving the 'average consumption' readout has risen to a consistent 8.5L/100km (still a significant if not huge improvement). This included one trip towing a trailer laden with maybe 200kg (or so, thereabouts, best guess) of firewood, largely uphill...

With similar light throttle openings I'm now seeing 'actual consumption' numbers that are significantly lower than I was seeing with the old sensors, but the readout still seems slow to react to reduced throttle (not much change to rate of response to reduced throttle, though the readout does tend to fall to a lower number in whatever time it takes). On very light throttle openings I'm now often seeing 'actual consumption' readouts around and below 4L/100km, whereas before I almost never saw less than 6L/100km (ignoring closed throttle when the display is and was 2L/100km).

I can't entirely rule out a placebo effect, but since I changed the sensors the engine feels to be running quite significantly better. I feel like I'm needing to use lesser throttle openings to cruise and climb inclines / hills, and the engine feels happier to do so in a higher gear. Response to opening the throttle feels better, I'm not as often feeling the need to change down a gear for the engine to do what I want it to do. It feels more responsive and a bit less 'fluffy', 'crisper', and generally more 'enthusiastic' to move down the road (the car feels as if it were a bit lighter in weight...).

I might be comparing an apple to an orange though. The engine used to feel a bit 'doughy' and otherwise less than great on hot days (and better on cold days), but since fitting the new sensors we haven't had any of those, winter has finally come.

I suspect that because the engine now feels a bit 'happier' and more 'willing', I'm tending to drive it just a bit harder than I was doing before the sensor change. If so then the apparent improvement in 'average consumption' could be a bit misleading, i.e. if my driving habit were to be identical to previously, then the 'new' economy might possibly be a bit better than the readout numbers suggest.

Regards,
John.

[bazzbazz]

John, is your car still producing that horrible burnt diff oil smell from the exhaust?

If so you may like to check this out -

https://www.alfaowner.com/Forum/alfa-147-156-andamp-gt/1164765-smelly-jts.html#post17272173

[johnl]

John, is your car still producing that horrible burnt diff oil smell from the exhaust?

If so you may like to check this out -

https://www.alfaowner.com/Forum/alfa-147-156-andamp-gt/1164765-smelly-jts.html#post17272173

Baz,
If you look carefully you'll see a contributor to that topic named 'johnlear', who is me, my secret identity. You might be excused for not twigging since the disguise is so cunning...

My car has ceased occasionally smelling like burnt differential oil, because I've fixed the cam cover oil leak. Due to this leak, and because I'm running the engine sans exhaust manifold heat shield, every now and then a drop of oil would fall onto one of the hot naked header tubes, which would instantly vapourise and create a short lived pong that would reach into the cabin.

The (other) smoke emanating from the end of the exhaust only briefly occurs (as far as I am aware) at start up after the engine has been shut down for some while, and then only on the odd occasion. I suspect leaky valve stem seals, the symptom is classic for that issue. This smoke does smell (like burnt diff oil), but the smoke is rare and at the rear of the car, so I almost never actually smell it, only see it briefly in the rear view mirror as I drive off, or as I reverse into its' haze.

I'm sure it's not related to sensors, cats, AFR. It's just a very minor issue, that over time will inexorably become a bigger problem...

Regards,
John.

[bazzbazz]

Ahh yes, I did twig, my direction to the post did steep in a little sarcasm. 

And yes, your quite right about the stem seals.