Fuel grades; Sooty Exhaust Tail pipes And Post - Cat Lambda Faults.

[Ascari32]

Hi,

I was looking for the post of a few days ago when I commented my Q4 started on the button after I filled with RON 95. The response to that was 95 has a lower self ignite level - made sense and it jogged my memory about the issue.

Found it! Quote BazzBazz "If memory serves me correct 95 Octane has a slightly lower Auto-ignition Temperature, which may well help starting."


However, again I filled up with 95, simply because it was the max grade at a garage I have never used before. No issues before given I am doing a fair number of urban/low speed driving, so I didn't foresee any problems.

However, what was totally unexpected was:-

I had a standing post - cat lambda alarm for weeks which I was going to investigate once my new cat arrived, so it was ignored. Also, I too have sooty tail pipes - they didn't take long to foul given these are my new "Noisy" boxes. This too I expected to have sorted when the exhaust system is competed and I have the AFR checked - I think she is running rich.

The drive home from filling up with 95 was quite short and everything seemed to be OK, albeit with the above conditions prevailing.

Today however, we went for a little jaunt to a local Marina, some 25 miles away. Soon after leaving home and before the car really had much chance to fully warm up, the dash alarms cleared. "Get Engine Checked" cleared and I now no longer have any engine management alarms - after weeks and weeks of them being continuously displayed.

I have not reset the Average consumption trip for about 600 miles~ so that any improvement in economy causes a very slow improvement in the dash figure.

The roads to the Marina are small, through villages, twisty and undulating. Average consumption moves from 25.4 to 25.7 mpg. Not much amiss there I say to myself. However, on parking up at the Marina, on checking the tail pipes, they are devoid of soot. Starting to get a "Brownish" tinge to the deposits.

Now I remember some time ago watching a YouTube video about using high grade fuel - the drift of it being it is a waste and does not benefit either the engine or economy. The chap was not saying use low grade stuff, just the appropriate grade for the engine.

So, have my sooty exhausts be a consequence of incomplete burn due to additive in the 98/99 ron I hitherto have been using? I believe Alfa state a minimum of 97 for the 3.2 JTS. But that is probably appropriate for high speed motorway use and long journeys. 95 seems to suit my engine, despite the heads being skimmed, The block too being skimmed/dressed and therefore the compression ratio a little higher than spec!

Can anyone throw further light on this phenomenon?   

[bazzbazz]

Face it, your car is bit of a wino and prefers cheaper plonk.   

[Ascari32]

Brilliant - no repost for that one!!!!!!!!!!!!!!!

But one Italian Tuning House and a UK specialist, both asked about what grade of oil I was using. This was in response to my engine temperatures dropping after the Manifold Cats were removed and Colombo Bariani camshafts fitted. My oil temperature gauge had dropped like a stone - now ~ 70 degree. The engineering company which rebuilt her said that was where they like to see the temperature as it is measured in the sump. Bearing face temperatures can be as much as 45 deg. higher on load.

This is fundamentally why I believe the Man - Cats destroy the oil prematurely and the temperature is so high at the bearing faces, on load it is difficult to see how metal to metal contact can be avoided. And oil flow and oil pressure across the engine - in standard guise, is appalling.

I am using the same oil, but without getting a probe into the sump I can only rely upon the dash indication for the time being. So it is difficult to qualify the absolute figure.

I have just checked my handbook and it quotes a minimum of Ron 95. From that the implications are, anything above is beneficial. But the new valve timing and the Autodelta manifolds carry a lot more heat away from the cylinder heads. Prior to modification, the temperature of the cylinder heads virtually demand pre-ignition inhibitors, by way of additives. Now that is not the case as lower cylinder head temperatures appear sufficient for Ron 95 to remain stable. Plus, modest loads on the engine in urban areas do not fully exploit the compression ratio of the engine.

So my conclusions are, the additives in fuel grades of 97/98/99 are inhibiting the spark ignition to the extent that the burn is incomplete - still a high content of HC's at blowdown. The engine remains really silky with Ron 95. I will have to pursue this line a bit further and get a couple of long journeys under my belt. But it is encouraging. 

[GTVeloce]

This is interesting. I have an example that is slightly different but I think aligns with your discovery. My 75 TS I normally run on 98 (it has a fuel quality plug that I have removed which denotes 98) however, occasionally I have run it on 91 without replacing plug. No pinging I can hear but invariably I feel the car has a significant improvement in throttle response when I first change to the lower octane fuel. Slowy the ECU learns the new fuel and the improvement diminishes and so I put 98 back in. I never could work out why this would happen but perhaps it runs too rich on 98 and the leaner mixture cleans the plugs a little and hey presto.

[Ascari32]

We have just had great day out, driving through some of the loveliest villages, though on atrociously difficult rural roads. Continually up and down the gear box at speeds never above 40 - 45 mph. S bends, undulations and numerous cross roads and junctions. The return was via the city, in rush hour 146 miles in all. The journey was more suited to a mini or 2CV. But the average consumption improved to 26.6 mpg and again the brownish tinge to the tail pipes is becoming better established.

And still the dash remains free of any fault display. Bit of a bummer really, given my new lambdas for the sports cat arrived today.

I can see high upper cylinder temps would need fuel which has a high auto - ignite figure. But with lower temperatures, it seems they are enough to deter proper combustion and the residual HC's foul the tail pipes and the post cat lambdas.

However, in essence, it is not running rich. It is simply the additives are inhibiting a proper burn which creates the kind of tail pipe deposits that suggest it is running rich. Well that's the way I interpret the symptoms.

[Ascari32]

This is interesting. I have an example that is slightly different but I think aligns with your discovery. My 75 TS I normally run on 98 (it has a fuel quality plug that I have removed which denotes 98) however, occasionally I have run it on 91 without replacing plug. No pinging I can hear but invariably I feel the car has a significant improvement in throttle response when I first change to the lower octane fuel. Slowy the ECU learns the new fuel and the improvement diminishes and so I put 98 back in. I never could work out why this would happen but perhaps it runs too rich on 98 and the leaner mixture cleans the plugs a little and hey presto.

The TS isn't that high a compression is it? And it doesn't run that hot either. So could the problem be ignition timing? Does the I.T. change with octane rating? Meaning I.T. is correct for the lower grade, but the Software adjusts it away from optimum - because the setting for 98 is not in itself optimised!

Or am I off the mark?

[Ascari32]

This is interesting. I have an example that is slightly different but I think aligns with your discovery. My 75 TS I normally run on 98 (it has a fuel quality plug that I have removed which denotes 98) however, occasionally I have run it on 91 without replacing plug. No pinging I can hear but invariably I feel the car has a significant improvement in throttle response when I first change to the lower octane fuel. Slowy the ECU learns the new fuel and the improvement diminishes and so I put 98 back in. I never could work out why this would happen but perhaps it runs too rich on 98 and the leaner mixture cleans the plugs a little and hey presto.

I'm not grasping something here! What happens when you leave the 98 plug in, but run 91?

[GTVeloce]

Ok, so the plug that came with my car sets the car up to 91 RON octane fuel by adjusting IT. If I remove the plug (no plug at all) it defaults to 98 octane fuel by again adjusting IT. There is a noticeable improvement in power and torque when the engine runs on 98 (no plug and 98 fuel). However, if I fill up with 91 fuel (but no plug), I get an improved throttle response for a short while. I haven't been able to understand why but perhaps you are on to something. Perhaps it isn't running rich (on 98 fuel) as I have thought (I get sooty exhausts too), but the additives aren't being burnt properly.

The TS is not a particularly HC engine (10:1 from memory) and it runs quite cool.

[Ascari32]

I'm still not there yet, but don't worry, its just me. Is the plug a resistor?

From what I have read, it is the anti - knock detectors that adjust the timing, but there may be more to it than that. When I was having my original Alfa back boxes refitted, I mentioned to the engineer what was going on - he was the one who re-installed my V6 and he has checked my lambda error codes before when they infrequently arose.

I explained again my motives for ridding the engine of the man - cats and changing the camshafts, essentially to get the engine temperature down. He was aware of that but he has a lot of cars which pass through the garage daily.

Before I had even finished he simply said, "So now you've got your temperature down, the engine doesn't need the higher grade fuel. Boom! - straight in.

It is; as I understand it, a question of the speed of the flame and local hot spots causing auto - ignition before the spark ignites the fuel and in a localised and uncontrolled fashion. So looking at it from the other end, temperature is now that much lower, so the spark ignition is inhibited by the Anti Auto - ignite additives. The flame - front spreads too slowly to consume the fuel in a uniform burn. So there is a quantity of fuel and additives which end up being exhausted - hence my sooty tail pipes and the lambda errors. 

But this is not because the fuel metered is too rich - a lower and AFR ratio. It is correctly metered, just the flame front is too slow. I think the flame front travels fastest at around 10.5/11:1 if I am not mistaken. But that is dictated by optimum cylinder temperatures and fuel compression. Compression hasn't changed, so it must be fuel temperature.

Yours however, appear not to be related to fuel temperature, but perhaps directly to the anti - knock action of your engine management system. It is quite a step from 91 to 98 so it isn't like it's minor adjustment.

Have you tried a fifty fifty mix 91/98 = 94.5. Then trying both ways - with plug in and without. With the plug in, the system will tolerate a smaller increase in octane rating - I would have thought. And with it out, your anti - knock will not be invoked so frequently, or savagely - I think.

I'd be interested to know if that plug is a resistor and what value it is.

Cheers!

PS. My understanding of modern Anti - Knock devices is, they are tuned elements which resonate at a frequency which is excited by the characteristic of anti - Knock "Ringing" - for want of a better expression. Any engine will tolerate a small amount of anti - knock. But there is a threshold where it becomes dangerous. I expect there will be graphs and data sheets which plot an engines characteristics. And the Knock threshold will be related to; amongst other things - Ron Rating.

So I am hazarding a guess and speculating, that plug is a resistive value, which modifies the sensitivity of the anti - knock devices, which ultimately the engine management system will refer to when it auto - adjusts the I.T.       

[philpot]

My understanding of the Bosch Motronic ML 4.1, and the fuel quality plug (S30) in the Alfa 75 is that the four 'programs' are fixed and can't 'learn' ?

ie. they are 91RON with CAT;  95RON with CAT;  95 RON NO CAT; and 97RON NO CAT.

Also there is no knock sensor to inform the ECU of any pre-detonation?

http://www.users.on.net/~craigf/fuelquality.htm

The above also applies to my 33 16V's.

I run a 'Yellow' plug which is the standard pale blue plug with the wire snipped to make a (95RON with CAT) equivalent.  This is the same as the simple wire connecting pins 85 and 87 as you snip the wire between pins 87 and 86.

Tried 97RON, open loop, no plug setting whilst running 98 RON and under certain loads do get pinging, like revving out in lower gears while climbing a steep grade.

nb:  'Open Loop', NO CAT, have learnt means input from lambda probe in exhaust is not utilised.

[Ascari32]

Open loop, doesn't utilise the current sense which varies, based on 350mV representing Stoichiometric. Current in loop at Stoic is 1mA. Below 1mA, mixture lean - too much oxygen. Above 1mA, mixture rich - too much fuel.

Utilising lambda signal, more likely to inhibit pinking - I would have though?

Putting engine under load, mixture is playing catch up, so better, if lambdas are helping. With no lambda signal, system must be slower to respond to changing engine load.

Aside from additives playing a part in the equation, I don't see how the AFR could ever be right, without utilising all sensors. The system is supposed to be Auto-dynamic. Without the lambda - it's not. Well not to my way of thinking! Closed loop - continually adjusting. Open loop - intolerance to changing  conditions.

MAF- reference air flow. Lambda - sensing quality of burn, enabling ECU to auto - adjust. No lambda signal, no auto - adjust.

If anyone remembers the old SU carburettors, hit the throttle when on load - engine pinked. No different with open loop. 

Sorry if it's sounds too basic.

[philpot]

Makes sense. Use 98 ron all the time. Not tried running 95 only over a couple of tanks to see the result. Yes, both my 33 16v's have sooty tailpipes!  Choke warm up program is pretty crude too !

A more detailed home page of Craigs Place re. Alfa 75 stuff.

http://www.users.on.net/~craigf/

[Ascari32]

Fitting links for RON95 should optimize the ECUs working range around the Spark Ignition - C.A. ? So load variations on the engine, would theoretically allow the Lambda signal to auto-dynamically adjust; within the range of variations of exhaust gas emissions, the links have predetermined the ECU should control.

The links, I would have thought, do not just control whether or not the lambda signal is used. It also presets the ECU spark advance - C.A.

And RON 95 should be used. Then theoretically, engine load variations would be reflected in the changing signal from the Lambda Probe - upon which, the ECU can act.

Without setting this parameter, other issues/problems could be missed/misinterpreted as something else. As with any lab trial, one has to set the fundamental parameters of the test - setup,  before what is being tested is applied.

Small changes in RON 95's absolute figure should then be within the scope of the ECU Closed Loop Parameters. Whatever, 95+3/-4 has to be better than 91 +7 or 98-7.  But with RON98 and open loop conditions, there is little room for error. Flame front is determined by Temperature and Pressure, amongst other parameters. 10:1, I would have thought too low for RON 98. So the burn could well be incomplete, given the engine doesn't run hot either. I am not suggesting you change your thermostat for a higher operating temperature.

Plus, Spark Advance for RON 98 would suit Ron 91 with the lower grade having a greater propensity to Auto - ignite, if the spark plugs have sooted up. The deposits are likely to modify the spark plugs operating conditions - almost "Glow Plug - like".

But as the plugs progressively clean up, the spark gap restored, the burn may well be starting too late to generate optimum pressure for sustaining the initially perceived increased performance.

Setting links for RON 91 and using higher grades, the spark is retarded - C.A., so one is delaying the start of the burn of a fuel, which has additives to reduce its propensity to auto-ignite, which is counter productive.

[Ascari32]

Open loop, doesn't utilise the current sense which varies, based on 350mV representing Stoichiometric. Current in loop at Stoic is 1mA. Below 1mA, mixture lean - too much oxygen. Above 1mA, mixture rich - too much fuel.

Utilising lambda signal, more likely to inhibit pinking - I would have though?

Putting engine under load, mixture is playing catch up, so better, if lambdas are helping. With no lambda signal, system must be slower to respond to changing engine load.

Aside from additives playing a part in the equation, I don't see how the AFR could ever be right, without utilising all sensors. The system is supposed to be Auto-dynamic. Without the lambda - it's not. Well not to my way of thinking! Closed loop - continually adjusting. Open loop - intolerance to changing  conditions.

MAF- reference air flow. Lambda - sensing quality of burn, enabling ECU to auto - adjust. No lambda signal, no auto - adjust.

If anyone remembers the old SU carburettors, hit the throttle when on load - engine pinked. No different with open loop. 

Sorry if it's sounds too basic.

Correction to the above:-

Stoichiometric voltage is 450mV, not 350mV.

[kaleuclint]

Had a new sensor fitted yesterday to rid myself of the warning message / check engine light.  Possibly forgot that I've tried this before without success.  Within a day they are back on!

A long journey burning 95RON may be called for.  Alternatively can the message trigger be bypassed?