Has anyone any clues as to why the ASR/VDC/Hill Start and Petrol Gauge should fail in unison on my 3.2 Q4 JTS. I am inclined to think it is related to some bad connection or moisture - it has been getting colder and damp of late. It would be nice to think it is something obvious, but history tells a different story. I do hope it's not some terminal software failure. Aside from these failure indications, the car is running nicely.
And advice would be most welcome.
Cheers,
The USUAL cause is you have lost the connection to your MAF.
Either the MAF has died completely (rare) or the connection to it is bad.
Try disconnecting and cleaning the plug and reconnecting. If the fuel gauge suddenly works you have found the culprit.
OK, here's the story. My New camshafts have such a massive valve overlap that the air through the MAF is such that it creates an error which the ECU is unaware of. New MAF flow based on cylinder fill + Crossflow.
So, after much testing I buy a new MAF sensor Bosch 0-280-218-008; fits Volvo XC70. But, it still needed a little "Tailoring" to work - get tick-over volume down.
Anyway, after much investigation trying various test components, I settle for 10k ohms in series with pin (5) the MAF output signal. Lovely, just the job. However my tinkering around the MAF has been going on for about a week and although the alarm indications came up yesterday, I made no connection with them and the MAF as the car/engine continued to respond nicely to my tests, getting better, then worse, then finally better as I swapped components in and out.
So earlier today, having settled on the final value, I phone my local garage to book the car in to have this problem investigated. At the same time, I post for any help in trying to understand why I now have these failures.
To which, a prompt response points me in the direction of the MAF, with the prospect of it having failed at worst. However, was I being sent - up? What could possibly be the connection between the Hill Start handbrake function, the ASR, the VDC and the Fuel Gauge?
Feeling a bit of an idiot, I go out to the car - having just got back from the garden centre - and do as was suggested by bazzbazz, disconnect the MAF plug, give it the once over and firmly plug it back into the element.
Key - in, depress the clutch and press the starter button. Bingo!!!!!!!!!!!!!! The logic is lost on me, yet another mystery surrounding this enigmatic car.
But, none the less grateful to bazzbazz for his insight - top man! One more issue ticked - off the list. Car is now running very much quieter than when I started all this malarkey. Just marginally noisier than a standard 3.2JTS. No resonance, sporty character to the exhaust system - right across the revs - Alfetta 3.0 like. Bags of pep on the throttle. In fact it is so different - I am persuading myself, rather than cancel the garage, to give my new boxes another chance. If they work, I am going to be one very, very happy bunny.
Thanks again Bazz!
When I had ASR/VDC/Hill Start issues the root cause was failure of the brake switch, or more to the point, poor adjustment.
A simple resetting of the switch position cured the issue.
Quote from: Colin Edwards on November 25, 2020, 07:49:37 AM
When I had ASR/VDC/Hill Start issues the root cause was failure of the brake switch, or more to the point, poor adjustment.
A simple resetting of the switch position cured the issue.
Yes, but the key indicator here was in addition the fuel gauge crashing to zero.
Well, true to form, things aren't as simple as it seems. Bazz pointed me in the right direction and that was that - or so I thought.
It was pure hypothesis about increased air flow due the changed valve timing and it seems a reasonable one at that. However, the ASR/VDC, Hill start and fuel gauge failed again. So further investigation was needed. Bazz is absolutely spot on wrt the MAF association. However, in reducing the current in the output to the ECU, I prompted a loop - sense fail. But the fact that I could reduce the "Power" in the tick - over, by trimming the current indicated air flow was affecting the ECU's ability to set the correct air fuel ratio. So I am on the right track. Just I can't trim the output of the MAF to adjust it, and expect the ECU not to display these failures.
Although the engine runs well, one cannot drive around with these failures permanently displayed on the dash. Although, it has to be said, with the 10k ohm in the output, the tick - over became the same volume as when I disconnected the MAF connector.
So, it seems I am being drawn towards remapping, which is not something I would willingly choose to do. The new Volvo XC70 senor is working brilliantly but the idle airflow is still sufficient to cause an output from the sensor, which it should not. This not only affects the tick - over, but the overall economy, given on lift of, fuel injected should fall to that which it is at idle.
So as an alternative, I may try to modify the flow characteristics through the MAF Venturi tube, or even change the whole of the inlet ducting. I shall have to do some more research/testing in the hope there may be some clues on how to achieve it, by way of a different MAF Venturi or reducing the airflow through the internal Venturi, in which the element sits. Essentially, the MAF body is a Double Venturi, the smaller enclosing the sensor, whilst it is located in the main body.
There must be a solution to this problem, even if it means a complete re-work of the inlet ducting. Just not yet thought too deeply about how to achieve it. The problem seems simple, yet the solution much less so.
Hey! Ho! The project goes on! Both fascinating and frustrating. Would I want it any other way? Probably not. Simple things aren't projects, they are just routine - and pretty boring to boot!
Hi!
Well given I started this thread, it seems only right that I should close it out, unless of course anyone else has something to add. Quite pleased that I gave sufficient detail in the title for BazzBazz to point me in the right direction. Given I was experimenting the with the MAF of my 3.2 JTS, it just tied in nicely and from then on things fell into place. Meaning, there was sufficient evidence to persuade me my outstanding problems revolved around the MAF output signal.
My current thinking is, whilst the Butterfly (Throttle) was closed, the MAF output voltage did not correlate with the conditions for proper idle and so the ECU was still reading the manifold lambda probes and attempting to set the AFR, despite both the inlet and exhaust VVT Solenoid Valves being in their "Default" positions, ie., inlet camshaft, fully retarded; exhaust camshaft, fully advanced. Hence the "Audibly Powerful tick - over at 750 rpm.
It is clear from the Bosch Data Sheets, the devices which have the greatest Kg./Hr. flow rates, do not specify and output voltage at the lower end, whereas, MAF elements, clearly intended for lower capacity/less powerful cars do! And because my original element was producing an output voltage, but behaving normally in every other respect, I am convinced it was the wrong element all along.
With the increased flow rate due to cross - flow/PVO, which my Colombo Bariani Camshafts introduce, an output voltage was being created which fooled the ECU into continuing to read the manifold lambdas, and attempting to set an Air Flow Ratio appropriate for minimizing emissions. This; I believe, caused the Lambda output signal (voltage) to induce the ECU to modify the fuel metering to the extent that it went beyond the range which could control emissions. Hence the Engine Management Code failures.
The element now fitted is a 98660612501, AKA 280 - 218 - 055, which; as it transpires is fitted to 911 Porsche 3.8/4.0 Litre Turbo's and Ferrari 599 GT. This according to published details is also fitted as standard to the 3.2 JTS, although Alfa have gone to great lengths to disguise the fact.
None the less, the engine is now performing beautifully, particularly at Idle where one would be hard pressed to believe it is the same car as before the element was changed. Tick - over is particularly quiet now and the engine has remained free of any management failings/codes. I have concluded, the MAF element which was originally fitted was not the correct device, raising suspicions about some aspects of engine maintenance, prior to my purchase. However, the engine in the car now is a fully balanced and modified Brera engine, with cam, exhaust and header modifications. The Alfa rear boxes remain fitted but I shall soon be refitting the Ascaris. given I now have confidence that they will perform better than hitherto.
Thanks again to Bazz for a very swift response to this issue which saved an awful lot of time and grief. It would have taken me considerably longer to resolve this issue, had he not picked up on it very quickly. Particularly when he intuitively linked all the elements together.
Cheers.
One is glad to be of service.
I'm not sure where you have gathered the published details for MAF inserts( i suspect chinese site) as 0280218055 is not the correct MAF insert fitted to 3.2 JTS. The 3.2 maf insert can be found on engines with less than 75Kw output. You could always contact Bosch for accurate information, in the past i've found them very obliging.
Well, it really is a question of what to believe, particularly when it comes to anything on the internet.
The response curves for the 2.0, 3.0 GTV and the 008 are taken from Bosch data sheets. So too the "421" which is the highest rated I could find. I could not find anyone who could supply it. Also, the Porsche is no longer available from them. But I was given the 98660612501/055 as direct replacements.
I have contacted Bosch and they were far from helpful, suggesting I speak with the manufacturer - Alfa, suggesting to me, the precise details were exclusive to Alfa. And of course Alfa won't disclose any information beyond selling me a replacement, at their ridiculously over inflated price.
No matter, it was the 98660612501/055 I had been looking for which I am informed by several sources is the correct one for the 911 up to 4.0 litre. Also quoted for the 599 GT. And the 3.2 JTS.
It would not surprise me that this element has been "Tweeked" for Alfa as their valve timing creates the need for a very fast response at the bottom end. That would tie in with what you say of engines up to 75 kW., which is evidenced by the much more acute slope of the lower Kg./Hr. elements for smaller capacity engines - as per Bosch data sheets. Indeed I have written, "The Alfa Insert response is probably more acute than that of (1) of the Bosch data sheets".
Essentially there is little difference between all of the devices I have mentioned. It is just the way they are aligned that differentiates them. With a calibrated source, I could align them my self to match the engines characteristics.
None the less, I am happy with the 055 performance, both because of the super smooth idle, which I confess seems unusually subdued for a sports system, and the really crisp way the engine responds to the foot. It, undoubtedly will not be bang - on, but until I get it on the Dynamometer, I won't know precisely how far it may be out.
The important point being, with such dramatic changes to valve timing, it required, in my opinion, a gentler Transfer Charcteristic. With the Turbo of the Porsche doing all the work in shifting the intake air, I reasoned the response of the MAF would be much less acute.
This appears to be borne out as my SS cat, Autodelta headers and in particular the CB cams will create much higher flow velocities, sort of mimicking what a Turbo achieves. Cruise control is very smooth in the way it maintains the speed setting, with very fine interventions an economy is better than I hoped for, although unable to do further testing due to lockdown restrictions.
Yes, I do believe she is still a little rich, but she was on my last Dynamometer run, with the same ECU. And I am impatient to try my Ascari back boxes again. However, "Slowly, slowly, catchee monkey". There is plenty of mileage left in this project to keep me entertained for some time to come. And Although I set out principally to get engine temperatures down, I think I can say, I have ended up with a pretty quick motor. And yes, I still want to fit a Holden badge!
Quote from: Ascari32 on January 14, 2021, 07:30:20 PM
And yes, I still want to fit a Holden badge!
THE HORROR! :o
Well the simple question is .. What is the part number of the MAF insert you had orginally prior to all the experiments?
Quote from: specimenX on January 16, 2021, 05:35:25 PM
Well the simple question is .. What is the part number of the MAF insert you had orginally prior to all the experiments?
I am sure that Ascari32 is aware of the MAF that was originally installed in his vehicle. If you read any of his posts, it will become apparent to you that he has made a number of modifications and is looking to improve upon what was spec'd by Alfa.
Might I suggest you start another thread for arguments, requesting if you want the five minute or full half an hour.
I said, nothing was ever going to be easy - but Hey Ho, yet another post cat alarm.
The performance of the Porsche MAF insert is very good across the range and the exhaust sound at tick - over is very subdued, by comparison to what it was just after the exhaust system changes. But failures do keep popping up.
Ignoring this, for the time being, an error does exist; even were the MAF a perfect match, insofar as with the Colombo Bariani camshafts the induction stroke is now a full 180deg., where hitherto as the inlet did not open until 11.5deg. ATDC it only amounted to 178.5deg.
The difference being an increase of 6.44% in potential air flow into the cylinders. Given air is swiftly moving through the inlet tract, courteousy of a 23.5deg. PVO, the relative effectiveness of this is substantial.
So I always believed, there would be a need to add a correction factor to the MAF output signal (voltage).
By simple trial an error, the output of the MAF: all the elements I have tested, has been reduced by way of a potential divider network between the output signal - pin 5, and the input to the ECU. I chose a total resistive value of 5,000 ohms to ensure this network would not load the MAF and drag the voltage down. This was convenient as at a peak output voltage of 5 volts, 5k ohm would result in 1 milliamp output current. The output of the MAF is low impedance and the input to the ECU is high; I believe, so the system should tolerate what amounts to very little extra loading.
By calculation, a 330 ohm resistor in series with 4k67 ohms, across the MAF pin 5 and pin 3 reduced the output signal (voltage), to a pretty close approximation of 6.44%:actual error: 6.6%, so .016% discrepancy is not bad.
But the engine is not as happy with that value as it is with 266 ohms in series with 4670 ohms - the output to the ECU being taken from the junction of the 330/266 ohm resistor and the 4670 resistor, the bottom end of which goes to pin 3: ground/0V. This amounts to a correction factor of 5.39% It may well be worth testing a lower value than 266 ohms.
I do not understand why this is the case but there is so much that I still do not know about this engine and in particular the software/ECU, I can only go with what evidence comes out of my tests. It may have something to do with improved Volumetric Efficiency, which a change in valve timing has generated.
It is true, millivolts output signal change can make a tremendous difference to the AFR and it will need to be trimmed eventually with a software recalibration. But given, I seem to be in the right region, such that I can provoke code failures and clear them, with small changes in value, I am not unduly worried, as long as I do not run the engine lean: which was always a worry despite my believing it was on the rich side.
Economy appears to be better than anticipated and a step - change occurs much more obviously at ~1500 rpm with the 266 ohm resistor, than with 330 ohms, whilst the exhaust system - Alfa Boxes - seem less laboured. Ideally, a diode in series with the output of the MAF causes the ECU to run the engine as if the MAF plug was disconnected. But the forward voltage required to overcome the internal junction voltage of the diode creates a flat spot which cannot be overcome. But the difference with the 266 ohm is not so different and persuades me I should try the Ascari back boxes again.
The car is quick - very quick! So, Air Fuel Ratios across the rev range; given I can detect no flat spots whilst accelerating, cannot be too far adrift from where it ought to be. But with this Pandemic and the need to do more tests, it will be some time before I can get her in for dynamometer tests.
Footnote:-
As this engine is direct injection, at tick - over and certain periods, fuel is injected at, or near TDC, it seems logical to assume there is ample time for the ECU to accurately meter the intake air sensed by the MAF. So despite my initial concerns, an accurate AFR should be possible.
This being the case, the air flow through the inlet manifold at the beginning of the induction stroke will suffer much less inertia and greatly improve VE. So, it may well be that any fuel metering needs a compensation factor applied to take this into account.
Perhaps this why I am seeing an improvement in engine behaviour with the lower value resistance, ie., the difference between cylinder fill and overfill, due to the velocity of intake air causing compression of the charge air, before the inlet valve finally closes on the compression stroke.
Just a Hunch!
Hi Kierenc,
Nice to know you are pursuing this. Although I have always felt the Spider was a "Bit Large", my affection for it has continued to grow. So much so that I genuinely believe it to be a car that will still turns head 30 years from now. And the Q4, from my experience with my 159 is second to none - forget your Audi's, this system is the Bees Knees.
The engine however, has never been right, IMHO, but has the potential to be great. The Block is a masterpiece and the fact that it will continue, albeit with some changes, to be produced until 2026 speaks volumes for it. I am convinced that will not be the last we will see of it either. With the swing towards "Environmentally" friendly transport, there will not be the impetus to come up with new "drawing - board designs", so this engine could be pressed into further service beyond 2026.
However, back to your Spider!
I shall try to deal with your questions, not necessarily in the order you asked and I shall add a little by way of cautioning you about the overall cost.
The MAF issue has and continues to occupy quite some time. However, these issues revolve around code alarms and not engine performance. As you would expect, making modifications to the MAF, as I have done, is bound to affect performance. But they have been done; and continue to be done, to see how much more can be gained from my engine modifications. And also, I confess, to defer a visit to a good software company for remapping. But regardless of the issues surrounding the MAF, there is no question about the tremendous difference between my engine now and as it was prior to modification.
In every respect, I consider my 159 to be "Sports Saloon" and the engine, given I have exclusively owned Alfas for over forty years, performs better than any I have owned. It does not have the charisma of my 2.5 Alfetta GTV, but it has a balance which is extraordinary for such a big car. It can never be as great as the Busso V6, but it is better. The flexibility the VVT System brings to the engine can never be overstated. And while it does not sing the same song as the Busso, it's power is expressed in the exhaust note much more forcefully. It never runs out of breath as some variants of the Busso do, although the Busso can be made to perform extraordinarily well for such an old design. But, one mustn't detract from Busso's achievement and I still have my 3.0 GTV after all these years, unwilling to come to terms with parting with her.
It has been quite an exercise in resolving the MAF issues. But, I believe the Porsche MAF insert, 98660612501, AKA 0280218055, AKA 0280218192 suits the engine and as such it will remain my final choice. My original element, whatever it may be, is history and will remain so. It is also reassuring that the clip of the SW with C.B. cams and Brembo's has been well prepared after considerable money was spent on it and no doubt performs really well, particularly given the expense on the exhaust system, let alone all other components.
I recently posted on the Aussie site my latest thinking wrt the MAF and I now believe I have got my head around the issue of flow rates/AFR better than hitherto. The potential divider network using 2670 and 266 ohms gave noticeable better performance than the 330 and 2670 ohm resistors. Adjustments, I now realise, need to be incrementally small as one can go through the optimum point without realising it. And when all is said and done, I can always fit the Supersprint tail boxes to quieten the exhaust system a tad more.
It seems strange to say, but in truth the MAF issues have been at least the cheapest aspect of all the work which has gone into the engine. I firmly believe the Manifold Cats are real killers of this engine. I am informed by Autodelta that their headers can be fitted without the removal of the engine. However, as my spare Brera engine was being rebuilt it was not an issue for me. There will be a marked reduction in engine temperatures alone if only their removal is undertaken. But, as Supersprint pointed out the secondary cat under the car, being Ceramic, have a lower operating temperature and the catalytic wash could be stripped and cause blockages due to exhaust gas temperature after the Man Cats are removed. Particularly if valve timing is modified in the way that the C.B.'s do. One commentator has stated the angles are very much in the same territory as his Ferrari! Certainly, the sound they generate is glorious.
However, the work I have been doing surrounding the MAF has now led me to believe there is no issue with the Colombo Bariani Exhaust Valve Timing. My initial concerns were wrt the way the air is metered through the MAF with such a large PVO and the fact that the C.B. inlet is now open 0.5deg. BTDC, thus giving a complete induction stroke from TDC to BDC - volumetric efficiency being improved further given the inlet also closes later: compressive action of the rising piston upon the fast moving inrush of air through the inlet manifold, ably assisted by the massive valve overlap/scavenging and the Supersprint exhaust system, with sports cat.
But given at tick - over and under certain conditions the fuel is injected at, or around TDC on the compression stroke, the ECU has ample time to read the MAF flow rate and set the correct AFR, in conjunction with the wideband lambdas and certain other parameters. And I believe the C.B.'s will also afford a greater degree of accuracy of AFR - when the fuel maps are corrected, which will extrapolate as increased power output.
However Colombo Bariani confirmed to me, changing just the inlet camshaft would net an improvement only fifteen percent less than if all four camshafts were changed. With this in mind, whilst not maximising gains, considerable money could be saved with only a two cam (inlet) set up.
It could actually be cheaper as peculiarly Alfa have used hydraulic lifters of different heights on the inlet and exhaust camshafts, inlets being 34mm. and exhausts 37mm., if my memory serves me correctly. So fitting exhaust lifters to the inlet side, a regrind of the Alfa Inlet camshaft may well be possible. to achieve either a higher lift or a longer duration, whilst the Alfa Exhaust camshaft remains unchanged. And the exhaust lifters are more readily available.
Yes, the Colombo Bariani Camshafts are expensive and their performance is very much dictated by accuracy of valve timing. With this in mind, it is really essential to renew the timing chain system. I think you can see the direction this is heading. Combined with the very first issue I addressed wrt this engine, the tensioner system is not sufficient to maintain accuracy due to poor oil flow/pressure distribution across the block. Were I to do nothing else to this engine, I would definitely modify the oilways with supplementary lines - again. This, aside from improving chain tension, greatly improves the oil feed to the VVT System and makes a marked difference to its speed of operation.
Another problem with is engine is its relative low sump capacity. It is designed around four litres capacity but the return of spent oil from the system is very slow and the oil pick - up snout in the sump is quite high above the sump floor. At high revs, the level in the sump can dramatically drop and, particularly when the car is moving laterally, the pick - up can become exposed and the bearings suffer oil starvation. The latest version of this engine has a massively redesigned sump of much greater capacity. They have also included baffle/swage plates to ensure oil is always moving across the face of the pick - up, whatever direction the car is traveling in. In addition, they have extended the windage plate across the full length and I believe they have redesigned the return ducting for oil from the cylinder heads, away from the rotating crankshaft journals.
Simple modifications to the VVT Solenoids and the camshaft position sensor on the rear bank exhaust camshaft improve the accuracy of timing further. But, looking back, it seems such a daunting prospect. In truth, the mechanical side was very straight forward, especially given I had some fantastically good people involved - Scholar Engines, Fast Test, Brian Randal, Kevin Johnson, Supersprint, Autodelta, Mauro Bosio and Colombo Bariani, and numerous others who helped source parts.
Kierenc, you are about to undertake a project which will cost a lot of money. And I fear, you will not be satisfied with "Half Measures". That being the case, a spare engine may be the way to go, so you can progress it in an orderly pace and spread the cost over a longer period. But if you choose to go down the modification route, I would start with the Manifold cats. According to Autodelta, that will save about 8 bhp in losses, but your engine temperature will drop dramatically and it will stop destroying its oil. Intake air density will increase as temperature drops so actually it may well be worth 10 - 12 bhp.
I don't think accuracy of valve timing can be achieved and sustained long term unless the timing chain system and vvt's are addressed, which means "Oil Way Mods". Beyond that, C.B. cams are still high up the list, particularly as they can be reprofiled for even more performance. But, oil flow/pressure needs to be improved to get the best from the C.B.'s. Unless Mace have moved the Lobe centres I cannot see them being even better than Alfa's stock cams. Their blurb implies the cams give a pleasant exhaust sound but say nothing of increasing power. You pay for what you get and Mace are cheap, I suspect because they are more cosmetic than performance related. If the lobe centres stay the same - which I believe they do, and duration is less than Alfa's, then that won't generate PVO.
I hpe I have covered your questions. However, you can always PM me if there is anything you feel I may have omitted.
Kind regards,
Further to my above, I started this project with little knowledge of what I was letting myself in for. The improvement to the lubrication system I felt worthwhile but at that time had not anticipated the failure of my engine, which led on to the Brera rebuild.
However, knowing what I know now, would I undertake to do it all over again?
The answer to that question is, most definitely, YES. Having been back on the road again for a year tends to diminish the difference between my car now as opposed to how it was. So I must get to drive a standard 159 or Brera, which would bring home just how different the two are. None the less, the car feels right both for it's size and performance. It sounds so much sweeter, melodic even. But I understand the real cost of achieving this and couldn't recommend it to anyone that did not view the a car with the intention keeping it for a very long time.
Logically speaking, removing the man - cats and fitting Autodelta's would be a start. Removing the secondary cats and fitting a sports cat - Supersprint. - and their centre section could be a good second move. Then replacing the rear boxes, so that the whole system was much more free flowing, not having Alfa's central resonator.
A marked improvement would be achieved with these changes alone. which may then whet the appetite for replacing the camshafts. At which point, I believe it essential, to maintain valve timing accuracy, the timing chain system should be replaced with the uprated Mace system, whilst oilway and vvt modifications can be incorporated - without engine removal, to ensure one gets the best from the improved breathing and increased power the camshafts will bring.
It may not to be everyone's liking, preferring not to believe there are issues surrounding the oil flow/pressure, but throughout the time it has taken to rebuild this engine, so much more has come out of it to reinforce my belief that this issue alone is an overriding factor which limits how well this engine can perform.
Still struggling for a badge and a name. So I suppose I can just fit whatever suits me at the time. So I picked up a Maserati Ghibli boot badge - as cheap as anything else around.
I spent a while in North Africa - Libya being one of my haunts - Benghazi being one place where I did an HF communications project on the old war time airfield. I sent my old man some pictures as he spent a large chunk of his war time service in Libya and Tobruk. Actually he was with the Northumberland Fusiliers and got caught in the Tobruk siege along with a bunch of Aussies. He asked me to visit the Memorial to them at Mersa Matruh.
Although he rarely spoke of his war time experiences, on occasions when he did, the affection he held for his Aussie comrades was patently obvious. Although the British Government never awarded a medal for either El Alamein or Tobruk, Australia did issue the "T" medal and they awarded it my father who was with them during the siege - one of the few, true Desert Rats.
Throughout the campaign, he must have endured many a Ghibli. I certainly recall how it would whip up the sand and dust and blow for hours on end. It got everywhere. I drove the coastal route from Tripoli to Benghazi twice and recall how many of the minefields still remain a serious hazard and WWII machinery littered the sands, only parts exposed as the Ghibli would blow the sands against it, and blow it off again.
Robert Menzies wrote to Churchill saying the Aussie guys had "done their bit" and deserved some R&R. Strangely, my father in law, serving on the Devonshire I believe, escorted the Aussie troop ships back to Aus. He was a handy boxer and there was a boxing tournament organized on route. He recalled making it to the final but his opponent was an Aussie Officer. He spoke of his concern at the time, about "Belting" an officer. I suspect a lot of guys would have relished the chance. But my father in law, although tall and physically very strong, was a gentle soul - a polite Welshman from the Rhondda Valley.
My father however, remained, moving on to Sicily and then Rome. Jesus, the things that generation had to endure!
So Ghibli it is!
After my "experiments" with the MAF insert, which led me down a few blind alleys, I think I have gone as far as I can. In the process of trying to understand the changes to the engine with the CB's, the Autodeltas and the Supersprint exhaust system, eventually I got the MAF insert I had been searching for - 98660612501, alias - 280 - 218 - 055, alias - 280 - 218 182. 9866062501 is quoted as being the element fitted to a Porsche 911 Turbo up to 4 litre. Elsewhere it is quoted for a Ferrari 599 GT. But it is not produced anymore and one eventually ends up with an equivalent being quoted as 280 - 218 - 182.
However, I have been running with the 2501 for a little while, with the standard Alfa back boxes. Generally speaking they perform well but still break - up on moderate to hard acceleration. They seem to introduce a little hesitancy which I can only assume is due to their restrictive nature. But once the car is fully warm this reduces, yet still seem to be choking the engine a little. Light load cruising is pretty subdued and tick - over is actually quite pedestrian. Although quite responsive, it is really not till above 2500 rpm that they seem to flow better, although a bit raspy. In honesty, I don't think they will ever be quite right despite nice flexibility at low revs in urban areas.
However, I felt to conclude my experiments I needed to retry the Ascari's, if for no other reason than to give them another chance. Particularly since all the coil packs have been changed and the new MAF element fitted. The coil packs have led to a cleaner sounding exhaust with the Alfa boxes fitted, so would the Ascari's equally be cleaner sounding, and perhaps a bit more civilised at low revs?
With this in mind, I had the Ascari's refitted this morning. I checked the exhaust before I left the garage and it sounded pretty subdued at 750 rpm. Returning an hour later and it sounds menacing, quite clean, but non the less menacing and two - three times louder.
Driving away, it is obvious how less restrictive they are but one pays for that with the level of noise. The flexibility is surprising and the willingness to pick - up the revs; very, very cleanly and quickly, is immediately apparent. They suit the engine that much better than the Alfa boxes. 3000 rpm cleaner and less characterful than before but still quite startlingly different from the Alfa's. Obviously less restrictive as they create seamless progression as the rpm climb, with absolutely no hesitancy at all. Torque -er at low revs than the Alfa's - but god the noise!
I shall have to live with them for a while now, for despite thinking they are probably much better suited to the engine now than the Alfa's, I think they are best suited to the track. And the Alfa's, in all honesty, unsuitable for either. The exhaust system is very clean insofar as all the sound emanates from the rear of the car now, with either the Alfa's or the Ascari's. Gone is the 2500 rpm drone/resonance and the engine seems content as it is beautifully smooth and sounds silky.
My concerns about the exhaust system were allayed after communicating with Supersprint and the engineer who dealt with my concerns was damned accurate in his predictions of what to expect from my changes, particularly when the Alfa Twin Secondary Cats came off. So much so, I feel compelled to trust them further and order a pair of their rear sports boxes.
In the mean time, I have a few weeks to live with the Ascari's and although I think I will miss their performance above 3000 rpm, they make the my 3.2 JTS sound incredibly crude below!
Having got the car through MOT and dash no longer lit up like a Christmas tree, generally it feels much happier. But I cannot get over the niggling doubt about the air flow at tick - over, not matching the figures published for the MAF output. This is, I believe a function of the revised valve timing. Needless to say the output volts; a function of Kg/Hr air flow in the inlet manifold is above the figure quoted on the data sheets.
I have been down this route before, with varying degrees of success and failure - notably when I introduced a resistive network between the output of the MAF and the input of the ECU. This was during a period when the Alfa boxes were back on the car. Worst case was provoking Hill Start/ASR and fuel gauge failures, which Baz quickly identified as a function of my tinkering's.
However, during that period I was having some success with exhaust noise levels and but for these failure, i would have been happy to leave the resistor network in place.
But the Ascaris are back on and although after MOT the car performs a lot better, it is the bottom end which spoils the whole experience. So I decided to try yet again to achieve a compromise with noise levels. I cannot help but worry, the boxes may not be entirely to blame and genuinely believe, because of the higher kg/hr output at tick over, this results in a noticeably more powerful idle. However, by definition, if this raised MAF output level at tick over is creating a rich mixture at 750 rpm, it must follow that at all other points, where the MAF is involved, the mixture must be rich away from tick over also.
So taking what Gaz suggested was the initial problem, I returned to putting a resistor in series with the output between the MAF and the ECU: my very first attempts were this method, before a chose to use a Potential Divider Network.
I thought what GAZ intimated actually was a very good tool in determining the "Threshold" for MAF output voltage at idle. I also knew, from early experiments, 270 - 330 ohms were good values for performance, although this was during the period of PDN testing. Never the less, those values should be a good starting point for further tinkering.
270 ohms - good, exhaust level came down and no ASR/hill start/fuel gauge failure. Ditto, 330 ohms with exhaust level dropping by probably 3 - ish db. further.
Next, 360 ohms. Fuel Gauge to zero, although it did recover, (just back from filling the car up), ASR and Hill Start fails. So, I thought 360 ohms must be close to the threshold. 3 X 1k ohm in parallel + 30 ohms in series. Fitted a further 30 ohm resistor in parallel with the existing one, value now 345 ohms.
No failures or alarms and on taking her for a drive - sublime acceleration with still elevated noise levels, but no where near as bad as without the resistors. And no "Hiccups", which I attributed to slight misfiring due the over richness.
Came home, bonnet up, A/C and radio off, 750 rpm and a warm engine. DVM reading direct from MAF, 1.345 volts +/- 0.015 volts. DVM reading beyond the 345 ohm resistor value - at the input to the ECU, 1.2 volts, +/- 0.015 volts.
Bosch Sensor Data Sheet - 1.2315 volts.
OK, I will settle for that. Not turning heads as much and it all sounds a bit more civilized inside the car. Well almost, but still not good enough. However, this I believe I can fairly blame on the Ascari Boxes. I shall monitor the performance until a get my new boxes, in the hope that now, the MAF characteristic is more "In - Tune" with the true flow rate through the inlet manifold.
I now have four MAF inserts to test after further experiments with the 055 element, which has the 345 ohm SOT in series with it. As I said, it has made the Ascari's at best - liveable at tick - over, if not entirely so when driving. So I continue to try and find a more definitive answer to my exhaust problems.
Somewhere on the Forum I read someone testing their MAF using the OBD II and it was suggested the kg/hr for the Brera should read seven hundred Kg. That is enormous, particularly as I have a chart which states ~ 300 @ 6500 rpm.
However, I wanted to explore the issue of "Over - rich running" during normal driving. So I revisited the Transfer characteristics published by Bosch. The improvement in noise level with the 055 element and the 345 ohms resistor in series corrected the voltage to the ECU to what it would be, but for the increased inlet manifold flow at tick over. It also reduced enrichment beyond tick - over although I think it is still too much. I not using much oil at all after the engine was rebuild, so I can only conclude it is HC's.
However, only 280 - 218 - 019 comes close wrt the shape of the slope, albeit their starting points differ: 019, 1.239 volt @ 8kg/hr and 055, 1.2315 volt @ 15kg/hr.
Needless to say, 019 runs out of puff at 1000 kg/hr, whereas 055 is rated for that flow. But according to my chart, that flow rate is way in excess of what it should be. I know response times are important but from the data I have, there is little to differentiate between any of them. So I reasoned, with such close threshold voltages 1.239volt and 1.2315volt, what significant difference would there be if I fitted 280 - 218 - 019, inclusive of the 345 ohm series resistor, in place of the 055?
Please see attached plot of their comparative responses. It is clear to me, 019, having a lower Kg/hr characteristic beyond tick - over, should result in a less rich burn and if the richness of the burn is in part responsible for rear exhaust box noise levels, the exhaust system should become quieter.
In short, it is! The Ascari's are now; in terms of sports boxes, tolerable. The noise they generate beyond tick - over is not as objectionable as hitherto and the way they scream above 3000 rpm put a smile on my face. All I need to do now, is leave it alone for a few hundred miles to access just how acceptable they are, long term. I will however, still be seeking to fit some good straight through silencers. But for the moment there is no need to refit the Alfa Boxes.
I should make clear, the Porsche 055 Insert response is the Pink Trace. The Blue - the 2.0 GTV. + 345 ohms in series. I feel it is a much better response and that is reflected in the performance of my car.
Spoken - All too soon! Christmas Tree ligh(T) appeared again on the dash. Only went for a gentle jaunt to a local farm shop - 36 mile, round trip. Came up on the way there and I wondered if there was some obscure reason it happened when doing only 30 - 40 mph. None the less the car seemed to be running ok. But as we got nearer home, the exhaust became a little more coarse - I think. It is true, there is a huge difference between the kg/hr of the 2.0 GTV insert and the 055.
However, I felt the compensating resistor 345 ohms would have negated this at low revs. I really don't know, other than I would not want to risk running lean at any time. So when we got home 055 was refitted but I left the 345 ohm resistor in circuit. I checked the engine ran ok but was in no mood to do a run in it - I'd had enough for the day.
Today however, after a short run to Tesco's and the bottle bank, I waited by the car for my wife to return from the supermarket, with a triple espresso!!!
After a little while, a youngster drove off in his little Honda, with the "Ubiquitous" large diameter tail pipe. To him it must have sounded great and I could hear him 1/4 mile away as he joined the bypass. Which got me thinking - the Ascaris don't sound that bad - noisy yes, but not that noisy. And inside the cabin, it is a lot different from when I first started this exercise. Is it me? Do new boxes get quieter after a few miles? I have been in this loop for so long, it's almost as if I am losing track of what I am chasing.
So, at the risk of it all changing again, this is where I am at! 345 ohms restores the output voltage to the ECU to nominally 1.2 volt, dropping about 110 millivolt across it, with 280 - 218 - 055 fitted. That 110 millivolt accounts for considerable volume in exhaust sound and the engine seems to lose a substantial degree of silkiness/smoothness when accelerating - indeed it sounds coarse. without it. Also, I think I can detect an element similar to a misfire - can't really describe it any other way. Not really misfiring but perhaps hesitant is the best way to describe it. A "Timing Error" - slow burn - due to over - richness?
Also, a greater resistive value reduces tick - over volume, but provokes an ASR/VDC/Hill Start/Fuel Gauge Failure. However, it seems there is little point in presenting a voltage to the ECU that is lower than that specified by Bosch in the data sheet for the 055 device.
In this regard I feel I can do no more and it seems clear to me now, I shall have to have a remap, combined with a pair of decent Back - boxes. The rear/boot area no longer resonates in sympathy with the Ascari boxes and it is possible to drive at urban speeds without too much intrusion from the exhaust. Where there is noise, it is appears to be emanating from the Ascari's and not in the rest of the exhaust system.
For the moment however, I am going to have to live with the situation as it is, until I can get a good pair of rear straight through boxes that are relatively quiet. 74 db seems to be a bench mark for SPL, so I will have to explore my options. But I could try strengthening the Ascari's with some "Ribbing" as Colin suggested.
Rear boxes and MAF - issues and conclusions.
With BazzBazz's input, came the realisation that there exists a voltage from the MAF; which feeds the ECU, below which a failure occurs, evidenced by ASR/VDC/Hand brake/hill start/fuel gauge fails. What I failed to add was; under certain conditions, the engine starts to "Hunt" - fails to tick-over smoothly.
With the valve timing changes, Autodelta manifolds and Supersprint exhaust system - not forgetting the Ascari's, the output from the MAF is; in the case of every MAF sensor I have tested, well above the ECU threshold.
However, the Bosch MAF data sheets indicate, only one, 280 - 218 - 008, has an intrinsically higher voltage at the bottom of its Transfer Characteristic. From this, I have concluded, the output voltage can be reduced from the MAF when 008 is fitted, by a much greater margin before the onset of "failure" such as that covered above.
It's starting point is 1.3395 volts @ 15kg/hr flow rate. For 055, which also starts at 15kg/hr, the voltage is 1.2315 volts. It is true, 055 goes to over 1000kg/hr, whereas 008 runs to 850kg/hr, but the characteristic curves are very similar despite this. And 850kg/hr is much more than a 3.2 litre N/A engine would digest. The other devices, with the exception of 280 - 217 - 123 which is wholly unsuitable, have threshold voltages similar to 055, which also limit the amount of correction any attenuation may afford.
So, I have chosen to use 008, to try and optimise the AFR before any software remap.
Making adjustments to the voltage the ECU sees @ 750 rpm, left considerable margin/headroom before the onset of failure, such as above. The Ascaris also became much less strident below 3000 rpm but it still required attention as to how the throttle was exercised, to prevent a horrendous cacophony. In short, I decided, the Alfa boxes were to go back on and the Ascaris - modified.
Today, the Alfa's were refitted at the same time as the new front discs and pads.
I alluded to the fact that it was very easy to adjust the MAF output voltage and completely miss the optimum point at which exhaust system noise was acceptable. In fitting the 008, it has been possible to attenuate the output voltage from 1.45 volts to 1.3 volts, whilst the response stays pretty sharp. But not sufficiently to make the Ascaris acceptable.
However, when the Alfa boxes went back on, I have to ask myself - what has all the fuss been about, and why didn't I manage to find the sweet spot before?
Needless to say, the Ascaris will not be modified - eBay or the bin. A second new Alfa box is being sought and some fine tuning around the 530 ohms fitted to the 008 MAF over the next few weeks will start in earnest, before the car goes for Dynamometer testing.
Conclusion - considerable exhaust system noise is generated by incorrect/over rich AFR. But beyond this, it really does depend upon the quality of the rear silencer. In this, the Alfa boxes may well have stripped a little off the out and out performance of the engine - because they are more restrictive. But not to the extent that they are unacceptable - particularly given they are now handling the upper revs in a much more refined manner.
On a final note, After picking up the car, Adam left a message saying he would ring me later - had to go out on a call. However, between getting that message and him ringing, I had another Code failure.
On speaking to him, he said he had not extended the leads for the Lambdas as there were no recorded failures and suggested it may be related to the MAF, but would not speculate as to what that might be. However, the engine continues to perform flawlessly. So, for the time being, I shall ignore it and hopefully diagnostics will pick it up when it goes for remapping.
I continue to be occupied with this damned exhaust system. And the dawning realisation of the differing performance the Ascaris boxes and the Alfa boxes create.
It is fundamentally; I believe, due the differing restrictive properties between them.
I am becoming more and more convinced about the "de Laval Nozzle" phenomenon. The shape of the sports cat seems deliberately intended to assist the already low restrictive properties of the 100 cpi cat element. And the post cat stainless steel section appears to perform as a condenser. Lowering the pressure in this section can only increase the velocity of gas appearing across the canister of the cat. It is analogous to the steam condensers, post steam turbine as used in power generation, where it is desirable to achieve almost vacuum conditions, thus maintaining steam driving force up - stream of the turbine.
Whether or not this is absolutely accurate - wrt the Supersprint cat and centre section - it is true that the free - flow this system creates; in conjunction with the Autodelta manifolds and the PVO of the camshafts - it is sufficiently accurate as to positively state, changing the restrictive element of the rear boxes makes a direct change to the flow rate through the inlet tract and thus the output voltage of the MAF.
Not having access to ECU software has meant, using Bosch Data Sheets for various new MAF inserts, which detail the kg/hr against MAF output voltage (to the ECU).
I have cross - referenced various voltages in attempt to find a Transfer Characteristic that best suits the new engine characteristics.
The rationale for this is based upon the premise that my original Dynamometer plot is indicative of a good ECU program.
This being the case, the improved engine breathing characteristics will simply change, substantially reflecting itself in increased performance - across the rpm range. It won't be absolute, but should be pretty damned close.
What I have slowly began to realise; given I have been at pains to give the Ascaris as much opportunity to show their worth, is the difference between the various MAF elements, and hence their changing kg/hr, to the actual air flow through the inlet tract, depending upon whether the Ascaris or the Alfa's are fitted.
So I never got the same result twice - until this final phase!!!
The Alfa's being more restrictive, reduce the flow rate substantially, when compared to the Ascaris. Thus, although I kept a record of the idle flow - rates; as I did various tests with all five elements, it only began to dawn on me when I at last decided to abandon the Ascaris.
Aside from the dreadful sub - 3000 rpm cacophony the Ascari's produce, I needed a valid technical reason for their unsuitability. This, I now believe is due to the inability of the MAF elements to reciprocate the flow characteristics the Ascaris generate in conjunction with the camshafts, cat - free manifolds, sports cat and stainless steel centre section. Suggesting the whole system requires an element of restriction and the Alfa boxes, whilst not ideal, are better suited. This tends to be borne out by how much quieter the whole car now is.
However, it has raised another spectre wrt Dynamometer Tests and the probability of doing some being premature, until rear boxes with characteristics which compliment the system are found.
It is frankly astonishing just how transparent the Autodelta and Supersprint Sports Cat and Stainless centre section; in conjunction with the Colombo Bariani's massive P.V.O., has made the whole system - engine included.
The term "Transmission Line" springs to mind, whereby Maximum Power Transfer occurs when all components are matched.
From this statement, it follows that if any one component in the transmission line is "Off - Tune/Mismatched", it is immediately reflected in increased power losses within the transmission system, and thus power to the load falls off. The load being in this case, the road.
So optimising a matching resistive network between the various MAF Elements and the Ascaris, did not yield the same results as with the Alfa boxes.
Consequently, the output voltages differed and it only became obvious, despite the boxes being changed over five times, when I finally cam to justify my decision to abandon the Ascaris to the scrap heap.
However, with the Alfa boxes now fitted, in conjunction with the Porsche 055 element - 98660612501, and a series resistor value of 733 ohms between the MAF output and the ECU input, the performance is pretty close to spectacular. Relative to the cars weight and it's performance hitherto.
Torque is enormous - courtesy of the VVT system. True, when warm, the engine is ticking over at 950 rpm., but I am unconcerned as I think I need to do some fine trimming of the resistor value +/- a few ohms.
This does mean however, buying a pair of Supersprint rear boxes would create yet another shift in the "Tuned - point", due to the differing back-pressure/ restrictive properties, between them and the Alfa's.
So it is my intention now, to find a second new Alfa box and enjoy the car whilst doing a little more to trim the response. And maybe have the car Dynamometer tested with Alfa boxes fitted.
Delays created by Brexit means it is unclear how long delivery of goods from the E.U. will be. So, until timescales improve, I will make do without.
Sent from my iPhone
Pointing me in the right direction as Bazzbazz did, when I provoked the above failures with my MAF tinkering has turned out to be a really useful tip/tool.
Now the Alfa boxes are re installed on the car, a similar exercise has been undertaken, to try to determine the right "threshold" for the 055 MAF element.
The combination of the Ascaris and an unknown/unsuitable Transfer Characteristic made it incredibly difficult to pin down exactly what the problem was. Although the true AFR remains unknown, and will until the car is re - dynamometer tested, it is reassuring to know where exactly the threshold of the MAF Transfer Characteristic begins. Clearly there is loop sensing of the MAF circuit which provides a tell-back to the ECU, in the event of it failing.
But the question remained - "what is the preprogrammed threshold of the ECU for fuel metering?" I cannot know that without having access to the ECU Program proper. So, until such time, the best I can do is guess-timate.
But a healthy low end - torque is a good indicator. So that is the first clue.
Secondly, trying to marry - up the actual voltage, as seen by the ECU, cannot be determined precisely, whilst the increased air - flow through the MAF at tick-over, is generating an output voltage some way above the ECU map threshold, resulting in a over rich AFR.
Unless the MAF signal/output voltage is reduced. in an attempt to make them match, this too will result in a fuel metering error which will have consequences right across the Rev range of the engine.
So, by attenuating (reducing) the output voltage, at tick - over, with a fixed value resistor in series between the MAF output and the ECU input, the knee/threshold of the MAF Transfer Characteristic can be brought into alignment with the MAF input.
The beauty is, this fixed attenuation/reduction will have a pro-rata effect upon the AFR right across the Rev range.
This can only be an approximation at best. But reducing the rich fuel/air ratio, will manifest itself in, a) improved economy; within reason, and b) improved power and throttle response.
But it is subject to finding the right degree of attenuation/right value of resistor.
In introducing attenuation between the MAF and the ECU, progressively, power and volume of exhaust noise is reduced. An initial resistor of 733 ohms resulted in substantial refinement in both acoustics and engine performance.
Changing the value to 900 ohms provoked ASR/Hill Start/Fuel Gauge Failure. However, they did recover and the performance of the engine seemed to step up a notch or two. But it did result in tick - over climbing to 1000 rpm. I wouldn't be unhappy with this and it probably deserves more investigation. However, on a positive note, it seems at least the threshold for the loop sense was discovered.
Until I can research this further, I made a general assumption that the best that can be hoped for is a value between 733 ohms and 900 ohms. Thus, I increased the value to 793 ohms. This value brought the tick - over back to 750 rpm.
A half hour trip around my test track gave really good acceleration, with superb progression- no holes/flat spots up to 5000+ rpm and wonderful low end torque, with super -quiet noise levels.
It would be nice to think I am getting close to optimum, albeit without a re - map, but non the less it is very satisfying. I shall leave any further testing for a couple of weeks, just to allow me time to digest exactly where I am in the whole process.
Hope you folks don't mind, but all this is pretty much a diary, for a project which seems to be taking forever. I find myself wading through previous posts, just to re - check if I have "been there before", with symptoms.
Such as, for instance - when trying various resistive values and deciding on a series resistor or a Potential Divider Network". Above a certain value; excessive attenuation, on pulling away the engine became very easy to stall.
This problem had been a characteristic of my 159 before the engine was replaced, but at no time after.
So, why has the car started to do this now?
Well I can only hazard a guess and say, by reducing the MAF output voltage with the series resistor, it appears the voltage falls below the MAF default level. Meaning, the base - line figure at which point the ECU alone is deciding upon what fuel to inject, to keep the engine running when the car is stationary and the throttle is closed.
Thus, when the throttle moves from the closed position and the ECU is now reading MAF Air Flow, it is a lower figure than when the car is "at - rest"; thus, I stall the engine. Which makes it necessary to apply throttle to quickly get past that initial stall point.
This became a issue with 900 ohms fitted although I did not make the connection at the time - particularly as exhaust noise became much sweeter, if that is the right way to describe it. Also, it seemed the car was quicker - a gut feeling!
And so, I returned to 793 ohms: 733+30+30 in series. Again, it seemed there wasn't the same "sparkle" and the exhaust noise increased a db or two. And, although less obvious, stalling the engine was still a possibility, if one wasn't getting pulling away quite right!
This issue does not exist with 733 ohms, but the exhaust is louder, but still pretty damned good when compared with the Ascaris. Bottom-end torque is very "beefy" as well.
So, it seems; by default, I am pretty close to where the inherent ECU tick-over fuel setting, transits over to where the MAF derived setting should be!
From this, I concede it now looks increasingly obvious I can no longer avoid the issue of "Re - mapping". The two conflicting issues being, sufficient richness at low revs to give good "Urban - drivability"; and it is great with 733 ohms. True also with no resistor but massively over - rich and louder, and a leaner AFR at higher revs for apparently more torque and quieter exhaust.
I love problems like these. The easy solution being to acknowledge one is at the limit of what can be done without software changes. However, this apparent "Nonlinearity" I detect - rightly, or wrongly, is somewhat different from what the Ascaris introduce.
This nonlinearity is solely down to the inherent mis-match introduced by the C.B. Camshafts, the Exhaust System, the less than ideal MAF Transfer Characteristic and the ECU Maps.
However, these factors are "Defined". But, one element can, if modified, have a considerable positive effect upon all others - "The MAF".
How? By attempting to create a nonlinear characteristic in the MAF output to counter the ECU - essentially to "Tilt" the MAF Transfer Characteristic.
Maintaining the degree of richness at tick -over, which 733 ohms; or there about's, provides, whilst modifying this value at higher rpm - lowering the dy/dt slope of the MAF.
On a further "Positive" note, I came across a company by the name of "Cybox", who do flexi-sections to create equal length headers. And they are software specialists. They are based close to where my son lives. So, he will go check them out for me. He himself being a software specialist - in a different area, should be able to discover how good they are. They also do similar sections for the Busso V6's, so they are starting with a few positives in their favour.
Now, I have to figure out the right degree of nonlinearity for my attenuation network. Or, at the very least, remove any notion from the equation and just get the ECU re-mapped.