Best after market exhaust for 3.2

[Ascari32]

What a comprehensive thread. Serious work has gone into research, development and testing here!

I've just started researching 3.2 exhaust options and it sounds as if there are no *easy* or cheap power gains to be had, or even sound improvements.

A somewhat silly question - is the stock system 2.5"?

Absolute correct about no easy or cheap power gains to be had. And Colin's observation about the standard Alfa back boxes is pretty much on the money, albeit they are quite restrictive. However, they are a substantial size so it is hardly surprising.

My problems, Post - Engine Modifications, stem from a number of factors which I simply did not appreciate at the time and only grew to understand their inter - relationship as I progressed.

1)       At start - up, the engine management system injects extra fuel to quickly get the manifold cats up to temperature. With free - flow manifolds, this rich mixture gets into the exhaust system and adds a cacophony of noise which no system can control/muffle/silence. From cold, the revs sit at about 1500 - 1600 rpm and it is only after a few minutes that is subsides, obviously the ECU predetermining that the "Cats" are now up to temperature!

         This, in my case is exacerbated by the fact that my JTS's valve timing is radically different from a standard JTS's, whereas the standard timing has 2.5deg. NVO: inlet opening 11.5deg. ATDC, whilst exhaust has closed at 9deg. ATDC. However, the crossflow created by the new cams, introduces air into the exhaust system which adds to the mischief of the rich start - up fuel.

         The effect all this noise has is to create almost panic, whilst one tries to understand what is happening. Eventually, there is an acceptance of the noise given the engine didn't explode into a million bits, but still one is at a loss to understand what is going on.

2)      Eventually, the engine settles out at a rock - steady 750 rpm. But there is power in the idle, as if the engine should be revving at 1500 - 2000 rpm. Another phenomenon which was not obvious to understand. Indeed, this was not understood properly until much later during tests on the exhaust system, Alfa's against my new one - admittedly to try and understand why the new just would not work below 3000 - ish rpm, whilst the Alfa's wouldn't work above.

         It has to be said, my thinking was being channeled in the direction of a faulty exhaust system/mismatch since the engine was modified. By which time, the Alfa Twin Cat had been replaced by the Supersprint 100cps Sports Cat and the Alfa centre section, complete with post - cat resonator replaced with Superasprints stainless steel - resonator free section.

         However, one of the differences these two elements made was to make the whole system more transparent and the racket was now firmly at the back of the car in the rear boxes.

         Not however, before realising the fuel issue extended to creating misfires, and quite severe resonating along the full length of the car, from the Supersprint cat, all the way back to the tail pipes.

3)      At this point, I began to think there may be an issue with the Standard MAF. And so, I cut the wire to pin five - output of the MAF - and placed a DVM in series, set to milliamps. It became possible to monitor the mA o/p of the MAF as the revs changed. Believing there could be some inaccuracy, I introduced various resistor values and found the tick-over subsided to more like normal and on test runs; down a particularly straight empty rural road, acceleration improved greatly - to the point where hitting the throttle a shove in the back became very pronounced.

        By this time, I had found the Bosch's data sheets and sought out the MAF for a 911 Porsche 3.8L. My logic led me to believe the 3.2 JTS MAF needed a fast response time Dr/Df to enable the ECU to set the AFR, given standard valve timing restricted the length of time the ECU had to calculate what fuel to inject.

        The new camshafts introduced 23.5deg. of crossflow (PVO) and allowed the induction stroke to begin to fill the cylinder from TDC and not 11.5deg. ATDC. Added to this, is the fact that for 2.5deg., flow through the MAF/inlet tract is "Checked" - both inlet and exhaust valves are closed. So volumetric efficiency can hardly be very good either!!!!!!!!

        However, 23.5deg. of Positive Valve Overlap (PVO) generates considerably velocity of air in the inlet tract/MAF, increased further by the downward thrust of the piston on its induction stroke. The Sports - cat and Headers actively promote this - increasing "Scavenging" enormously, whereas hitherto the Alfa valve timing had none. And this, I reasoned, was generating an error in the MAF which leads the ECU to set the wrong AFR.

        What also became patently obvious was, the Colombo Bariani camshafts greatly increased "Throttle Response" and by the same token reduced "Pumping Losses".

       My argument for choosing a Porsche MAF was that a Turbo car would have a less acute "Transfer Characteristic" than the 3.2 JTS's. By which time I decided to try my spare new 916 GTV 2.0 ltr. MAF insert. Better - much better. Idle more normal - acceleration good, but topping out at 4500rpm - just as the Bosch Data Sheet Graph indicated.

        Cue - my new 916 GTV 3.0 MAF, although it took me a while to get it off the car. At last, things were going in the right direction. Torque extending down to a little over 1000 rpm and good response. And All of a sudden - my new rear boxes are starting to sound the way they should. Lovely above 2500 - 3000 rpm and beyond - lost for words. But still, although more tolerable, not enough to retain them for anything like urban speeds. So Alfa's Boxes had to go back on.

4)     Deciding I was going in the right direction, being unable to find a Porsche insert at a sensible price, or a Bosch 0280002421 - the highest flow rate sensor on their data sheets (5), I ordered a 0280218008, which is the next one down (4) - under £32 new. Used on Volvo XC70? - perhaps this is why it was cheap - sorry Volvo owners. But being an Alfa Owner, when the boot is on the other foot, I'm gonna milk - it!

5)     The new MAF insert should be here sometime after Friday. But in the meantime, I had the Alfa boxes refitted. Gone is that glorious howl from the back boxes - dare I say "Ascari's". Colin's observation about Alfa's boxes ring true at this moment in time. True, the engine is not as free flowing as it was with the Ascaris, but they are so much more "Docile" and torque at the bottom end is noticeably more pronounced, coming in at just above tick - over. I have had a few jollies in her, given the weather is not too bad at the moment and it is a real joy to be driving her, even though it still has the GTV MAF fitted.

        The "Dead in Moscow" can return to their slumbers and with the front offside wheel bearing having been replaced a week of so ago, the quality of this car in its driving is astounding. Absolutely beautiful.

        As a footnote however, when Adam finally drove the 159 into the servicing bay to remove the Ascaris and refit Alfa's, it sounded sublime - nothing like it sounded from the drivers seat??? So could it be, one of the differences between the two is due to "Acoustic Coupling". It is very long way from the tail pipes to the input of the cat and although suspended by rubber hangers, it is a rigid piece of stainless steel. And what of the chances of coupling through to the boot floor?

       It seemed to me, apart from the obvious increased restriction the Alfa boxes create, that restriction would be "Bi - directional". So the attenuation it generates in the tailpipe direction, would also exist in the reverse direction, reducing what is reflected back towards the cat!, Coupled with the physical size of the Alfa boxes, acoustic coupling must also be less through the boot floor.

       So, is it possible, my remaining problem is to find a way to isolate the Ascaris from both the body and the rest of the exhaust system, via flexible coupling and acoustic insulation in the boot? It would be nice to think it is. Because, I still can't come to terms with not hearing the top end howl the Ascaris produce.

2.5inch! 2.0inch back boxes                       
 

[torquemeister]

I have the Wizard exo with Prodrive S tips fitted to the rear boxes. Looks and sounds stunning. Next step is to decat the precats and go either Wizard or these babies from Orque..............
http://www.orque.co.jp/tuneup/kakudai/3.2jts-mani2.html

[Ascari32]

They aren't manifolds - they're works of art.

I don't think Wizard do headers - they have just finished prototyping some some back boxes. Autodelta assured me, their headers can be fitted with the engine in situ. However, my engine was being replaced so it wasn't an issue. Not sure how well it will perform without a valve timing change though. Is there any real point in gas - flowing and exhaust system with no crossflow?

[Ascari32]

What a comprehensive thread. Serious work has gone into research, development and testing here!

I've just started researching 3.2 exhaust options and it sounds as if there are no *easy* or cheap power gains to be had, or even sound improvements.

A somewhat silly question - is the stock system 2.5"?

Bosch MAF 0280218008 arrived this morning. Quickly popped out with my wife to dispensary and shops. Car nicely warmed up and I have been getting used to the 3.0 GTV MAF fitted to the Q4. So much so, I thought it may not be a bad alternative.

However, got home and twenty minutes later, the new MAF insert is fitted. Out to my "Test Circuit" - quieter at idle, but won't know what she's like from cold till tomorrow morning.

Good at low revs - 1100 rpm in fifth really smooth and really quiet. Acceleration nice and crisp with good throttle response. Always eager for more throttle. Exhaust sound, more akin to Alfetta with no break - up from the Alfa Boxes. Miss the punchiness of the new ones, but these are really quite good compared to before. Agree with Colin's sentiments; probably best compromise.

No holes in the response with excellent progression. One can feel the urge from the engine when transiting to above 1500 rpm; lean burn technology below, so Alfa state. Generally, the impression of being a db or two quieter and a little less edgy on low throttle - would fit in with the less acute dr/df slope of the new MAF. Should make trimming the AFR easier - if I get bored.

   
I think I am done now! Just need the "Designed by Holden Badges" for the front wings and that's me finished.

Cheers, - Stay Safe!

[Stu159]

Hi Ascari,

I was reading back through this thread last night and thought I might add a little more for your research and to benefit others. I noted that you stated that you thought the guts of the centre cats crumbled owing to your engine mods and increased temperatures. The reason I had the work done on my exhaust set up that I've noted on this thread, was because I had experienced exactly this issue. Unfortunately I can't recall how many k's my car was up to when it occurred, but I estimate around the 80k mark. At the time my mechanic just removed the insides and refitted it and that's how I drove it up until recently when I had it all modified.

I'll add that I've owned my car since new (bought as dealership demo in 2008 - approx 4000k) and until the recent exhaust mods I've noted, the car - with the exception of my suspension, has remained stock. So the issue occurred in my car, with no mods. I'm not sure how common it is?

All the best
Stuart

[Ascari32]

Hi Stuart.

Thank you for that contribution - much appreciated.

Chronologically, I had concerns about the centre cats for quite a while before contemplating the exhaust system beyond the Man - cats.

The difference their removal made was substantial, not least because of the exacerbated drone at 2500 rpm. The sequence of events that followed was the removal of the Alfa centre section and its replacement with a Supersprint stainless section. This shifted exhaust system noise towards the rear boxes, but spread the range of harmonics wide-band.

I then fitted a pair of Ascari sports boxes which were intolerably noisy. Feeling I had made a bad mistake buying them, I researched the Alfa twin cats and discovered they were ceramic. They have 500 cpi and all articles I researched indicated, a: they have a lower operating temperature, and b: excess temperatures can strip the catalyst coating and cause obstructions.

By this time, I had ordered a 100cpi Supersprint Sports Cat. It seemed obvious to me I was never going to get the Ascaris working and felt the safest option was to ask for Supersprint'S advice, particularly given their boxes are dedicated, but damned expensive.

The response I got from a senior design engineer who had many years with them was, "It is unlikely there is anything left of the internals of the Alfa twin cat, due to the temperature and pressure of the gases from my modified engine. Armed with that information, I destroyed what I though was a good spare cat I had in the garage. What fell out was substantially ceramic dust and sand - like grit, the remnants of the cat canister.

I still some time to wait until the S.S. Cat arrived, but as soon as it was fitted, the engine was transformed and refitting the Ascaris, she screamed like a banshee at 3000 rpm and above.

But still she wasn't right - the noise below that was unbearable.

I was grasping at straws at this point and reasoned the MAF may not be helping because of the radically changed valve timing. At the same time, I was starting to get some coil pack codes. So a MAF with a more progressive slope was found, after much trial and error with JTS MAF, and tests with new 2.0 GTV and 3.0 GTV MAFs.

Things started to come together and exhaust noise was starting to be more sporty and less aggressive, if you get my meaning. Eventually, all coil packs were replaced: "Delphi"s", and the engine improved yet again. At this point, the new MAF with the gentle-er slope was starting to perform better than the two GTV MAFs. Low rpm Torque coming on strong and load on the engine when cruising at seventy like nothing before. There is just a little hint of growling, where hitherto, it seemed like the engine was labouring, even though it wasn't.

The MAF element is 280 - 218 - 008. However, another device is being sought, a 280 - 002 - 421. This has a range up to 1000 Kg/Hr. Importantly however, it has an even less acute slope.

On evidence, progressively reducing the slope, as was the case using the 2.0 and then the 3.0 GTV inserts showed substantial improvements in tick over and progression compared to the 3.2 JTS MAF insert proper. This improvement continued with the fitting of the 280 - 218 - 008. So I am pretty confident the 002 - 421 will prove beneficial.

I think the 3.2 JTS MAF is bespoke to Alfa, produced for them by Bosch. So they will have a tolerance spread which enabled the Alfa "Softies" to embed a "Stock" fuel map. That I believe would take into account Alfa's peculiar valve timing. There is little I can do about that and I suspect there is little can be done without expensive software rewriting. And I don't think the difference it would make could ever justify the cost.

So, given the new cams allow for much more accurate MAF flow figures, I still believe ultimate power to be in the region of 315 + bhp. So much so, that I actually think the Ascari rear boxes deserve another chance and They will be getting one in another week or two.

But first, I have to find a 280 - 002 - 421 at a sensible price.

[Ascari32]

Further to the above, it gets incredibly tiring, continually searching the internet for parts which are at sensible prices. Particularly given the degree of uncertainty as to whether they will do the job.

However, I have found a MAF insert for a Porsche 3.8L, 911 Turbo. They are no longer produced by Bosch, or if they are, they may now have a different part number. It seems however, a 280 - 218 - 055 appears in the guise of 98660612500 for Porsche.

Manufacturer Part Number:   0280218055   Brand:   BOSCH
Reference OE Number:   0280218055 HFM-5-SF 30211 EE4129 XAM4247 558322, 38.796 7516248 FDM5038 FDM897 86248 0890212, LVMA256 VMA106 93006 1882019 15395 98660612500, 98660612501 QM1030 QM738 QM900 19706-M 19708-M, 19714 EAM029-M EAM033-M MAFS284-OE MAFS445-M, MAFS450-M 61-06448-SX V45-72-0032 330870300, WG1014276 WG1495389.

"Construction year to: 07.2004. Construction year from: 08....more   Porsche   911   2011   997 [2004-2012] Coupe   4.0 GT3 RS   Coupe   3996ccm 500HP 368KW (Petrol)"

Given this is a Turbo Flat Six, logic suggests - my logic that is - the Transfer Characteristic will have a pretty linear response, whereby output voltage does not increase excessively at the lower end of it's Kg/Hr response, so turbo boost will be much more progressive across the engine range and AFR much more accurately controlled. Otherwise, there would be a very big bang!

With this in mind, making the analogy with the airflow generated by the Positive Valve Overlap and the powerful scavenging generated by the Supersprint Sports Cat and free flow exhaust system, I am hopeful the AFR will be less variable at low kg/hr. than with the standard 3.2 JTS MAF element.

Evidence from my tests; to date, indicate the dramatically changing output of the JTS MAF at low kg/hr. contributes to exhaust system noise. With such huge valve overlap - 23.5deg. it is possible, unless there is a much more progressive change in fuel variation when the throttle is exercised,  enriched exhaust gases could be ignited within the exhaust system, albeit the percentage quite small, yet sufficient to generate considerable exhaust system noise.

Previously, coil packs breaking down; electrically, probably accounted for misfiring and incomplete burn at a time the exhaust valves opened, generating exhaust system noise. So,  by the same token, if enriched exhaust gases find their way into the system, the same phenomenon is possible. To my way of thinking, it is better that the ECU adjusts the fuel in a gentler fashion as dynamically, no engine can respond instantaneously to fuel enrichment. It is better if there is a slight lag as the engine assumes its new mean power as opposed to being over - injected with fuel prior to getting there. 

Exactly what the ideal response curve of the MAF for this to happen is something I cannot calculate. So it has to be by trial and error. The range that the Porsche MAF covers is 3.6 to 4.0 litre. And given substantially it is air flow/boost that determines their increased output, I hope it won't be too dramatically different with my 3.2 JTS and its increased induction and cross - flow.

Air through the inlet tract/MAF is already moving fast by the time the engine is at TDC, courteousy of the massive valve overlap the Colombo Bariani Camshafts have created. So the ECU should have a greater length of time to determine the correct AFR.

And it didn't cost too much for the sake of finding out!

But I would dearly love to know what it's Transfer Characteristics are! Help from any quarter would be much appreciated. 

[Ascari32]

Life is full of surprises - deciding to try a Porsche MAF insert and order one yesterday and having a delivery date for 7 days +, the element arrives within 18 hours. That has to be a Christmas record!

So one hour later, insert fitted, I go for a test drive. Floods causing havoc with the roads - friends farmhouse flooded out - I spend 20mins dodging the lakes.

Initial impressions: lost a bit of drama at lower revs but exhaust system quieter. Silky progression but not as sharp as with GTV 3.0 or the 280 - 218 - 008 insert. Throttle response less dramatic, but stab it and the exhaust sounds more civilised - mundane really. Good drivability in lower gears, torque good but more comfortable than in higher gears for the same speed!

Five miles on and ecu throws a code - assuming it is the new MAF out of range. Hopefully, it will reset when ECU learns the new profile.

Conclusions? Think I am going in the right direction. Actually think I can refit my new boxes. Even the Alfa boxes are sounding more "Sporty". Not saying it is ideal, but believe I am in the right area. Needs more testing/miles driven. But need that Engine management code to clear - of its own accord - before I am totally convinced.

Bought this Porsche insert quite cheap and quality not up to Bosch's usual build quality. However, until I am convinced about the right Transfer Characteristic, I can't justify spending more on what are essentially "Test devices".

Happy Christmas everyone - Stay Safe!

[soakk]

If I've learned anything reading this thread its that aftermarket exhaust install on the 3.2 is a black hole of effort and money 

[Ascari32]

So true!

And again the car threw a code. Hasn't affected performance - at least I can't detect a change and I think, but can't say for certain it is due to MAF and being over rich at low revs.

The car is quick from the off but it has been unusually cold and she has never truly got warmed up.

I think; again don't know for certain, it has something to do with keeping the man cats at light - off temp, but they of course don't exist anymore. But on Start - up, extra fuel is injected to warm the cats, causing the engine to Rev @ about 1500 rpm for a minute or two before she settles down to 750 rpm. I think this and the cold weather is fouling the pre-cats and they are going out of limit.

Sitting in the cold in the supermarket car park while my wife was shopping, the exhaust was starting to reek of petrol fumes. I need to see if there is a temperature element somewhere which is functioning; normally for a standard engine, but just too low now with all the changes I have made.

She goes for MOT on Thursday so I will get some feedback after the test - but I'm not confident she is going to pass emissions.

To follow up on the previous post - best avoid the expense for what is really just cosmetic and little by way of increased performance. But if anyone goes the whole hog, the car may end up being used for track days.

But that really misses the point, for me. For one, it has been an exercise in trying to determine how good this engine could be. And failing emissions is not sufficient to condemn the exercise outright. It has never been an exercise in finding how good a dog of an engine can be made to sound, by piffling exhaust modifications. It's about outright performance, after the areas I believe to be poor, are addressed.

Secondly, it is also a question about what can be achieved without recourse to software rewrites. Principally, because I have seen so many differing views on so many issue, none of them supported by engineering evidence. There are experts who could resolve problems - quite quickly. But they come at a cost. And they are few and far between and certainly not local to me. So, it is a bit of a cleft stick really. But, I'll get there in the end.

[afelice001]

good luck! 

[afelice001]

If I've learned anything reading this thread its that aftermarket exhaust install on the 3.2 is a black hole of effort and money 

Seems that way. Getting more power out of the 3.2 V6 is like getting blood from a stone
Alfa should have started with the GM 3.6L block. Another 30-40kw would have been helpful in the 159.

[Ascari32]

You make have a point there. But I am still not convinced they wouldn't have buggered that up too!

Too much was going on within the FIAT Empire at the time and, in my opinion very little "Oversight" on the 159/Berra project. Hell, one minute they were climbing into bed with GM and the next GM are throwing two billion at Fiat to get out of the deal. And tossing in the 159 chassis/p platform for free.

Seems like GM came to their senses. No matter how much trouble they were in, a marriage to a dysfunctional  organisation such as Fiat was not the solution.

Since then, Chrysler were wedded and now they have taken a French Mistress! If anyone has a Busso, hang on to it, because Alfa Romeo has had so much Cosmetic Surgery, it is the only way to identify what they once were.

[Ascari32]

If I've learned anything reading this thread its that aftermarket exhaust install on the 3.2 is a black hole of effort and money 

"Round, like a circle in a spiral, like a wheel within a wheel.

Never ending, or beginning, on a ever spinning ----------- ad infinitum.  https://youtu.be/qKV9bK-CBXo

So a couple or three days later, still a standing code/alarm - presumption it is Lambda/MAF related and I abandoned the car due to bitter weather.

Spent some time trying to get my head around what is going on and in particular, what the hell the ECU has to do with all this! A couple of really good articles read, authored by company that does stand alone ECU's for racers. Not really interested in exactly what they do, more the reason why they do it!

The particular area I was interested in was Wideband Lambdas fitted into the headers - Pre - cat Lambdas. The post cat lambdas can only act upon what the exhaust gases contain, after catalytic conversion. Of particular interest is the closed loop action of the manifold lambdas. Several re-reads persuaded me the MAF element I bought was providing an output voltage which resulted in the ECU setting a Fuel figure, beyond the range at which the wideband lambdas could enable the ECU to "Trim the Fuel" to within the "Lambda Tolerance/AFR" and thus the code/alarm.

And bitter conditions meant the engine was barely getting up to temperature, thus a very heavy smell of petrol from the exhaust - extra fuel injected below operating temp to help get the cats up to temperature. This was in the supermarket car park, on tick-over, whilst my wife was shopping.

This latest code/alarm occurred within a few, miles of the garage technician interrogating the ECU for historical faults/codes and having found no recent ones, reset the ECU, although standing alarms had been present when he did so.

As I understand it, the ECU is continually monitoring the AFR and storing any changes - slowly adjusting the fuel trim over a substantial period - "Standard Deviation", or long time averaging. Clearly, the mismatch was too great for the ECU to stay within bounds.

The MAF element fitted was 280 - 218 - 008. So clearly there is a problem with it's Transfer Characteristic. However, given the alarm appeared after very few short miles, I reasoned the ECU would not have had sufficient "Drive - Time" to store/average/modify the mean fuel injection quantities, before the failure occurred.

More in desperation than faith, I decided to refit the Porsche MAF Sensor. 280 - 218 - 055: AKA, 98660612501. The difference between 008 and 055, aside from the obvious is, with my 3.2 JTS there is an output from the MAF on tick - over with 008, whereas with 055 there is not; confirming my belief that it is relatively insensitive, which ties in with the fact that the Porsche 911, 3.8L is Turbo-ed. Meaning, one can remove the MAF plug with the 055 and the engine does not respond to the disconnection - no change in basal output voltage. If however, one disconnects the 008, the engine does respond with the "Weight" dropping out of the exhaust sound - less powerful although still rock steady at 750 rpm. And it's output voltage drops to it's basal figure when the MAF element is disconnected.

Importantly, with the 055 the exhaust system is considerably more sedate and the engine revs cleanly. I am hoping my initial judgement on the 055 was clouded due to the ECU having; prior to fitting last time, had an extensive history file of lambda adjustment/averaging and thus it went out of bounds when the 055 was initially changed.

The engine still had not come up to temperature after refitting 055 and the conditions were still pretty bitter, although it was not dark. So cold, I decided I would go sit in the car while it came up to temperature. But, after a short period, and the engine still quite cold, the differences became obvious and I got out of the car to have further listen at the back of the car. She was just a little uneven at tick-over but nothing that I would not be content with.

I am talking of less than ten minutes after the MAF had been changed. Realizing the A/C was on full blast, I looked through the rear screen and the engine management failure had cleared. I had to go sit in the car, just to get my head around this and in doing so, switched the A/C off. The engine settled immediately to a gentle purr. Got out the car, walked to the rear and the exhaust note was nothing like anything I have experienced during all my travails.

Of course, all this will change again, no doubt. But, it came as a huge surprise, particularly how civilised the exhaust sounded. And it reassures me - if it is like this now, then regardless of what other issues pop up amongst the general malaise, it can be like this again. And if I can make it thus - I'll be very - very happy.

However - let's see what tomorrow brings!       

[Ascari32]

https://www.alibaba.com/product-detail/LUFTMASSENMESSER-SENSOR-MAF-FOR-PORSCHE-0_60112075050.html

I stumbled over this too! "Circles". If this is correct and I have no reason not to, seems the Porsche insert is pretty universal. Some Alfa models seem to share this element with not just Porsche, but the Ferrari 599 GT.

And yet when one thinks of Porsche and Ferrari, one would never believe something as pedestrian as a 159 JTS "Tank" could have anything in common. A 147 maybe, as relatively speaking it carries the weight of a Gazelle; the 159/Berra, a Hippopotamus.