Alternate history: M96 Cup engine - Page 5 - 911/Carrera GT - PistonHeads UK

This is going to be long, and rambling, and potentially very boring - it's a story that currently has no end, and indeed may never have a happy one. It's the story of an engine, specifically an engine that never was, and potentially might be at some point in the (hopefully near) future. Most people have a good understanding of the history of the 996 - it was launched with a development of the water cooled flat six from the Boxster, pushed to 3.4 litres, but the decision was taken to install a water cooled evolution of the older air-cooled motor to create a high performance version that would be called the GT3. The same engine also appeared in the turbo, and the Porsche Cup cars (with changes appropriate to each, of course). Slightly less well known is the X51 Powerkit, a comprehensive upgrade package for the M96 engine that was created when Porsche was evaluating whether the M96 had what it took to be the base engine and the high performance version. For the 996.1 this kit was extremely comprehensive, encompassing an additional oil scavenge pump, heavily revised sump, X51 specific heads, pistons, cams, valves, a whole new cast aluminium intake (instead of the stock plastic unit) and much larger diameter exhaust manifolds, an additional central radiator and a revision to the DME to manage all the changes. This wasn't a red badge and a tweak to the map - the engine was gone through from top to bottom and almost every system was revised or enhanced. And, of course - it wasn't enough. The decision was made that the Mezger was the way forward and the rest was history. The X51 kit continued to be offered, but with each subsequent iteration it was less comprehensive, and whilst certainly more than a trim level it was never as comprehensive as it was in its original form. Over the years the M96 has become the butt of many jokes, the chocolate engine, the engine Porsche wanted to forget, and with the move to the 9A1 architecture it largely was. I have to admit though, I liked it - characterful, makes a great noise with a sports exhaust, plenty of shove low down but some real zing in the upper reaches of the rev range. I had bought my 996 on 60,000 miles, and decided that I was going to run the engine through a prophylactic Hartech rebuild before it hit 100,000 in order to preserve all the things I liked, but remove the weaknesses that were well known to bedevil the M96 (and some that were less well known, but still needed to be engineered out). Around this time PopPopBangBang had the engine in his very well known 996 C4 rebuilt, in his case by Autofarm, and pushed to 3.7 litres. He was good enough to publish some estimated power power and torque plots for his revised engine, and it was clear that the new unit had more horsepower and torque- both delivered lower down in the rev-range. This was clearly a strong indication that when I got my engine rebuilt I should swap from the 96mm bore to the 100mm bore in order to take my 3.4 to (the obviously much more desirable) 3.7. But, why stop there? It occurred to me that if one were to add the X51 kit to the 3.7 then it's higher lift cams, flowed heads and increased rev limit might create a unit that continued making power when standard 3.7's were already changing gear, an engine that potentially answered the question "what if Porsche *didn't* choose the Mezger?" I started to put together a wish-list for my new engine, a list that could be summarised in two key points: remove the weaknesses of the M96 and add drama in the form of revs and power. First, the weaknesses.

With the weaknesses taken care of we had to think about how we were going to get enough air through the engine. I had purchased an X51 intake, but it soon became apparent that an intake optimised for a 7,200 rpm wasn't going to cut-it at 8,200 rpm. I sold the X51 intake having never taken it out of it's box - losing a mere £500 for the privilege of having it on a shelf for six months. With the decision taken there was only really one choice - individual throttle bodies had to be the way forward. We had a look at what was available - a choice of one, essentially, with Jenvey making a kit. We decided to go our own way, and started adapting some BMW M3 ITB's: Initially our thinking was to use the variable length centre section from the later 4.0 RS as a plenum to feed the ITB's: But we may go with ordinary ram-pipes initially, more on this as we continue to develop the intake side of things.

The heads have been extensively ported, and the results on the flow bench were very promising, however Mike was convinced that the geometry of the valve seats was holding us back. New, larger valves (paired with titanium retainers and new, stronger double valve springs) allowed him to re-profile the seats for another significant increase in flow.

Fantastic read and epic project. Full marks for getting into the very upper echelons of project escalation! Some beutiful work in the pictures. But Porsche did choose the Mezger, probably with good reason. It is regarded by all as the pinacle of the water cooled engines, the definative motorsport engine for the road, with character, chatter, performance and reliability. I sincerely hope your end result gives you a lot of pleasure and that the journey does too. But unless Hartech are doing a good portion of the development at ‘mates rates’ as a research project it looks like a bit of a financial black hole.

Yes, I never claimed that this made sense - but it's fun. Bore and stroke - the standard 3.4 engine has a 96mm bore, and the Lokasil liners are an open-deck design which ovalise over time. They can be fitted with reinforcement rings, but a better solution is to machine them out and replace with Nikasil liners, and whilst you're there close the deck. And, if you are machining out the liners and replacing them - why not go for a little extra displacement? Initially I was going to use my existing 3.4 litre engine, but as development went on I was made an offer on a brand new 3.2 litre (Boxster S) block that was impossible to refuse. The 3.2 is identical to the 3.4 in every dimension that mattered for me, and the new 100mm liners combined with the 3.4 crank yielded 3.7 litres. Martin decided to build a 3.7 for his existing early 996 coupe, and a 3.9 (using a 996.2 3.6 litre bottom end) for his new project car, a 996.1 C4 converted to RWD. New 100mm pistons matched to the new bores: And three fresh engine blocks waiting to be built up: We need to finalise our con-rod dimensions and get them ordered, then we can assemble an engine and run it up on the dyno. We're so very close now - but there's still so much to do.

Only understood half of it (ok that’s a lie, hardly anything), but that sounds very cool and I enjoy reading stories like this / learning a little something along the way. I really hope it will be a success and you will be happy with the end result

I'd never thought of using anyone other than Hartech to rebuild the engine - their knowledge of the M96 is unparalleled, and Baz Hart is an extremely talented engineer and (rather helpfully) a very nice chap. I spent ages on the phone with both Baz and Grant, and their willingness to share their knowledge really made this project possible. Hartech race the M96 engines and have developed fixes which address the weaknesses exposed by both road and track use - incredibly useful empirical knowledge of the engine, won the hard way. Now, the M96 has an intermediate shaft that runs beneath the crank and transfers power to the cams via chains, and the oil pump via a 50mm hexagonal key. It's got an oil fed plain bearing at the oil pump end, and at the other a sealed bearing - the infamous IMS bearing. This is the beté noire of the M96 according to Porsche folklore, and it's said to be simply a matter of time before it destroys every engine in which it is found. This isn't quite true, and anyway for the new engine we'd use the final development of the IMSB from the M97 - a much larger bearing than used in the 996 versions, so large in fact that it can't be installed or removed without splitting the engine cases. Next on the list was crank flex- the construction of the M96 is such that the crank has a marked cantilever from the final main bearing to the flywheel, and this, when combined with vigorous use of the clutch and revs can flex the crankshaft, leading to rapid wear of the main bearings, leaking Rear Main Seals (RMS) and - in extremis - snapped crankshafts. Hartech offer an EN40B crankshaft (which a good friend who joined me in the project chose for his engine, more on this in due course) and also they can machine the engine case and install an additional bearing just behind the flywheel which constrains the degree to which the crank can flex, and essentially removes this problem. Now, I was determined to spin the engine past 8,000 rpm - I wanted fuel cut at ~8,500. This meant we had to look at the valve train, and by this point a chap from 911UK who has an incredible knowledge of racing engines had offered his assistance. We quickly identified the tappet chest as a weak link, for numerous reasons. For one, it was cheap crap - a weak casting with the bare minimum of material required, and was known to suffer from hydraulic fracturing in normal road use. As new camshaft blanks were not available for the M96 we'd also need to re-grind stock ones, hence would have a smaller base-circle, which would mean that the tappets would ride higher in their bores, and would need both more support and the oiling hole position changing. It was clear that we needed a new tappet chest, as we simply couldn't modify the existing one to meet our requirements. By this point I'd managed to infect another chap from 911UK with my obsession, and he'd decided to build not one but two engines to our new "X51 Cup" specification, and very rapidly it became apparent that without his input we'd have got nowhere fast. We ran a stock tappet chest through a scanner, creating a highly detailed (and enormous) digital image of it. My partner then created a CAD model from the digital image, and in consultation with the engine designer we created a new tappet chest that addressed the weaknesses of the OEM one. I emailed Richard Brunning (of Bad Obsession Motorsport fame) and asked if he had a recommendation for a machine shop that could do the work - step forward Specialist Engineering, who created this prototype for us: Finally, the oiling system - an integrated dry-sump in Porsche speak, disparagingly called a wet-sump by almost everyone else, but actually somewhere in-between. In it's stock form the integrated dry-sump tank is formed from a plastic baffle that sits in the centre of the sump, and whilst the engine is in operation the oil is returned to this internal tank. Sadly, the tank isn't exactly oil-tight, and under heavy G the oil can escape into the rest of the sump. This is largely resolved by fitting the X51 sump which has an aluminium baffle/tank with rubber seals that fit tightly against the internal walls of the sump and do a much better job of keeping the internal tank full under G. This just left the issue with the heads - or, rather, the right hand head (as you look at the car from behind). To save money Porsche used the same casting for both heads, so the drive end of the cams is at the back on the right, and at the front on the left. Each head has an oil scavenge pump that returns oil to the central tank/sump, and under normal conditions this works fine. However, under heavy braking, and especially if the car is turning left, oil pools in the front of the right hand head as the scavenge pump is at the back, and for that brief period is sucking air. Porsche resolved this by fitting a dual stage scavenge pump to this head that scavenges the front of the head via a port on the cam cover, to which it is connected via external hard-line (the iconic X51 towel-rail). So - buy the dual stage pump and problem solved, right? Sadly not - the currently available dual stage pump spins backwards, which we considered to be sub-optimal. So we made our own: Leading too: Edited by Dammit on Sunday 26th April 22:14

Outstanding effort. I've only skimmed so far, shall return for a proper read but wanted to thank you for taking the time to put this thread together plus of course the massive effort gone into the engine - could help many others decisions. I believe it was Jethro Bovingdon that commented along the lines of "if it wasn't for the Mezger, the M96 would be considered a great Porsche engine'.

I'd not heard that - but I immediately agree with his observation. To go with the ITB's we're going to use a lightweight flywheel - I could give you many performance reasons for those choices, but frankly I'd do it just for the sound they promise. For me the sensations of Motorsport always started with the noise - the rising intonation followed by a brief pause, then a further escalating shriek of the cars running on the straight followed by what to me was always the most intoxicating sound - the driver prepares the car for a corner, engine barking in repeated short bursts whilst they change down through the gears. That whap-whap-whap almost like a super-bike engine revving on the starting grid. It was the first thing that struck you as you walked into a racing circuit, then the smell of hot motor-oil competing with that of the doughnut stands, and finally the sight of the cars themselves, primary colours of sponsors logos, flaring brake discs and flaming exhausts. But ever present, the noise- spine tingling howls punctuated by staccato barks, the defining aural signature of the sport. My car may be a little silver bath-tub with a canvas roof, but one day soon it's going to make the noise.

Fascinating thread thank you! I have a 997 C4S X51 so watching with great interest. I have fallen completely in love with mine and have a Hartech rebuild in mind in the future as it’s definitely a keeper. The midrange howl is addictive and it pulls like a train. Yours is going to be another level by the looks of it! Good luck.

Excellent threat. Looking forward to seeing how this develops. It's going to make me sad I wasn't able to do more to my engine when it had the full works rebuild by Hartech some years ago though!

Yes yes yes. Fantastic project and thank you so much for taking the time to post this up, youve made my week and its only Monday morning! Moar pics!!!

Gearing - the standard box has pretty long ratios, adding ~1,000 rpm to each gear would make them intergalactic. Also, the engine is going to come on cam from around 4,000 rpm, then scream past 8,000 before fuel cut forces you to grab another gear. Long gears are not going to work here - they'd take forever to get into the power band, and when you change they'll drop you out of it again, a recipe for frustration. The stock box is, frankly, somewhat of a pain here- it requires a fairly hirsute press to get the gears off the shafts, and of course back on again. Albins (in Skippie-ville) make gearsets which can be swapped in, should you have an Esau edition press in your workshop, and said gears are helical so should be fairly mild mannered. There is also the option of straight cut gears from the likes of Guard, but I'm uncertain that I could cope with the noise (beyond what would undoubtedly be the hysterical first drive). We did look at using the GT3/turbo box, but the different dimensions make this very difficult (and conversations with companies such as Quaife revealed that they'd tried and failed to convert their Cup box to work with M96 variants), so modifying the stock box appears to be the way forward. Current plan is to dyno the engine, then extrapolate the required ratios from the shape of the curve that we get, in order to drop into the power band each time another gear is selected, then order a stack of suitable gears from Australia. Somewhat undecided as to whether it would be a good idea to make the whole box very short, or leave sixth as a much longer gear for cruising on the motorway.

Legend, OP! Following with interest!

Great post @dammit I will read thoroughly when I get a chance, but already know part of your project. Put my m96 project on the backburner and instead getting the suspension up to spec. I have a detailed post on pcgb forum if anyone has access to that (wibble).

Glad that some of you chaps find what we're doing interesting - hopefully at some point this year we'll have a dyno chart to show whether it worked or not. Cams. The 3.4 litre M96 engine has four camshafts, two per bank, with the exhaust cam being driven from the intermediate shaft (yes, that intermediate shaft) via a chain. The inlet cam is driven from the exhaust cam, again via a chain - the tension on the chain can be varied, which changes the inlet cam timing. This is called VarioCam, a simple system that has an annoying appetite for the (expensive) actuating solenoid. Unlike more sophisticated systems there is no variation in valve lift with this system - that came with VarioCam+ on the 3.6 litre engine of the 996.2. In the X51 package for the 3.4 Porsche changed both inlet and exhaust camshafts, interestingly the earliest engines came with lash caps whilst the later ones featured longer valve stems (valve diameters remained the same). With the 3.6 X51 package Porsche elected to only change the exhaust cam, the inlet now having variable lift courtesy of the "+". We had a good look for the lift and duration of the X51 cams, but the paucity of information (Porsche themselves were totally unwilling to provide any data) combined with the cost prompted us to look elsewhere. No one makes cam shaft blanks for these, and billet cams are a) expensive and b) accelerate wear in the valve train. That left us with Schrick cams, or re-grinds. First we looked at the Schrick offering and, whilst it was interesting, we weren't a fan of some design elements and also decided that having a grind that was 100% orientated around our requirements was the optimal solution. This does mean reducing the base circle, but we have our billet tappet chest (so we can adjust where the upper edge of the bore of the tappet is relative to the centreline of the cam, and the oiling) and our own valves (as we upped the diameter for both inlet and exhaust) so this has been accounted for. At the moment the thinking is to have three options - fast road/track/race, but I suspect we'd be unable to resist the temptation of fitting the lumpiest cams with the highest duration. We'll see - the cam designer can grind to order, so we don't need to make our minds up until quite a late stage. The cams in the 3.4 are shared with the 3.2 Boxster S, so when in a 3.7 (my car) or a 3.9 (Martins) they're a significant limiting factor, therefore this was an important thing to get right. I think three of the things that we paid the most attention to were valve and valve spring material, weight and form, every aspect of which then informed the cam design. I'm reliably informed that valve train float is A Bad Thing, so attention here is important. Anyway, suffice it to say that whichever of the three profiles we end up putting in the engines the valves will open further, for longer. Interestingly this has knock on effects in other areas - the Porsche exhaust cams are (relatively speaking) rather inoffensive little things, so choice of exhaust manifold design doesn't really make a huge difference. Moving to rather spikier cams means that we now need to optimise the exhaust for the cam design, so equal length primaries become important, although we're constrained by the packaging when it comes to primary length (there being a finite amount of room between the underside of the engine and the road), still we should be able to get a decent amount of tube under there.

If you decide not to keep 6th as a cruising gear, can you not just change the final drive ratio? Or are there also weaknesses / potential improvements in the stock box that you want to address?

Dammit said: Interestingly this has knock on effects in other areas - the Porsche exhaust cams are (relatively speaking) rather inoffensive little things, so choice of exhaust manifold design doesn't really make a huge difference. Moving to rather spikier cams means that we now need to optimise the exhaust for the cam design, so equal length primaries become important, although we're constrained by the packaging when it comes to primary length (there being a finite amount of room between the underside of the engine and the road), still we should be able to get a decent amount of tube under there. Good to know there are other lunatics hell bent on cramming as much N/A horsepower in the most unorthodox manner possible, under the engine cover of a narrowbody 996 Neil If these fit under my steed : I'm sure you'll be able to get something as, or more impressive, under yours and Martins ...

Would love to fit a GT3 RS lump in to a 996 carrera 2. Making a 996 GT3 Touring of sorts. I assume it would all fit relatively easily? Deep pockets needed just for the lump though !!

Replied to read in detail later. What great project! Bravo and thanks for the detailed post. Surely a valuable source of resource! Only one question arised already: what didn't you like about the Schricks design? Something specific about your build, or something in general. Just asking out of curiosity, since they are a highly regarded manufacturer of camshafts, at least in Germany.

The reason to have gone with Schrick would have been if we had to use a stock base-circle, in which case we could have then optimised for one of the grinds that they offer - which are off the shelf. I did speak with them directly about whether they would supply us with blanks - sadly the answer was no. We'd have gone with Schrick blanks if they'd been available. Given that, and the work we'd done with the other elements of the head and valve-train it was significantly better to use grinds created with specific reference to the weight of valve/spring/retainer, point of valve-train separation, fuel cut, desired performance characteristics and of course flow performance through the head (and many other elements that no doubt Mike will point out to me as being even more important, but that are beyond me!)

Enjoying the thread, so thank you. Sadly, I can’t add anything constructive, but I am interested!

Is this a PistonHeads thread? Somebody pinch me.. It’s these unicorn threads that give me hope.

Slippydiff said: Dammit said: Interestingly this has knock on effects in other areas - the Porsche exhaust cams are (relatively speaking) rather inoffensive little things, so choice of exhaust manifold design doesn't really make a huge difference. Moving to rather spikier cams means that we now need to optimise the exhaust for the cam design, so equal length primaries become important, although we're constrained by the packaging when it comes to primary length (there being a finite amount of room between the underside of the engine and the road), still we should be able to get a decent amount of tube under there. Good to know there are other lunatics hell bent on cramming as much N/A horsepower in the most unorthodox manner possible, under the engine cover of a narrowbody 996 Neil If these fit under my steed : I'm sure you'll be able to get something as, or more impressive, under yours and Martins ... I was wondering about this - it looks like the main part of the Mezger block (pistons and heads) sits about 2-3 inches lower than the M96 because the latter has the sump and intermediate shaft? Very interesting read and good luck!

Yes and no, which isn't that helpful of an answer I am aware. The Mezger does mount lower - if you look at a naked C4 shell and a GT3 shell then the only panel differences are in the rear of the engine bay where the engine mounts are lower down, and to the firewall behind the rear seats to accommodate the oil tank, although this latter change is apparently not strictly needed. I'm not sure whether the mounting points being lower equates to the engine being lower, though. The Mezger engine also has an intermediate shaft which drives the cams via chains, and the depth of the sump in the M96 directly under the crank is very shallow - most of the space inside it is too either side, see this excellent video for more detail: https://www.youtube.com/watch?v=4DOLuqLFOUk

Dammit, any further developments with the project?

We are looking at piston design - complex things, pistons. Ours need to hit a very precise compression ratio target and to be as light as we can make them (amongst other design goals), and only once we've resolved these issues can we finalise rod geometry and get those ordered. Hopefully once the design part is done we can get the parts made fairly quickly, but I think the goal of getting an engine in a car in 2021 is probably now unlikely, bit we may still be able to get an engine on the dyno before year end.

Piston design finalised and they're in production, new ITB design finalised and awaiting machining, and a prototype dual-stage scavenge pump is installed in my engine and has had a short test run. I'll be putting as many miles on the pump as I can before the end of the year, when we'll strip it down again to check that everything worked as expected. It was installed by Precision Porsche using the standard Porsche workshop manual for the X51 dual-stage pump, and used the OEM hardline.

Thought I'd give this a bump to ask if there was any further developments?

The pistons should be arriving in the next few days and we're ordering the rods now that we have the precise dimensions. Next stage is a dry-build of an engine to measure basically everything, then if everything is nominal and we don't need to adjust anything we put it together properly and map it on the dyno, make any adjustments that are required and then put the engine into a car and map it again.

Something of a milestone - dry build of my engine ahead of measuring absolutely everything, before stripping it back down again and then rebuilding it properly (if everything checks out).

Is this running yet? When can I buy one?

The dry build confirmed that everything was correct, so the engine was stripped back down again. We're waiting on some parts now - the billet tappet chest pictured earlier in the thread is in a stock 3.4 at the moment getting miles on it, as it was designed for the stock camshafts. We are putting a fairly significant grind on the camshafts to get the profiles that we need so in turn the base circle is reduced, which means that the tappet chest design has to be revised in order that oiling and so forth is correct. The revisions should be complete shortly, then we need to slot into the CNC places schedule, then that is ticked off. In terms of the other elements that we are waiting for the ITB's just need to be CNC'd and assembled- that's a fairly self contained process as the chap we are working with can do all that in-house. That leaves the equal length tubular manifolds and I'm uncertain where we are with those - we have found a chap whose work I've seen on a heavily customised F40 and is of excellent quality, but I don't know his timescales at the moment. However, with all that said, I believe that 2021 is the year that we get an engine on the Dyno, and all being well - in the car.

Amazing project

What's the plan with the itbs? I'd love to fit those to my boggo m96 3.4 in time, if only for the noise!

A thread of real substance.. The very best of luck to you.

shalmaneser said: What's the plan with the itbs? I'd love to fit those to my boggo m96 3.4 in time, if only for the noise! We initially looked at re-purposing some ITB's from an M3 with an adapter plate, but whilst it looked good (and would have sounded great I'm sure) the flow-bench showed that the adapter plate was killing flow. Given the not insignificant work that has gone into raising flow (both in terms of volume and velocity) of the heads themselves it would be lunacy to then write-off these gains with ITB's that were not optimised. We started discussions with a UK company that makes custom ITB's who is based close to Martin and went through a discussion that led to a CAD model, once that was signed off we went for a rough 3D print, then a higher resolution one which went to the flow-bench, and once that was signed off we ok'd the final design. I would warn - these are designed for our application, and whilst they can of course be bolted to a stock 3.4 (or indeed 3.6) the question of whether they'd give gains on a stock engine is one that we have not looked at answering. They'd sound excellent, of course.

Steve Rance said: A thread of real substance.. The very best of luck to you. Thanks Steve - I'd be keen to get your driving impressions if you would be interested in taking my car for a spin once we get the engine in?

I would be delighted

Cool project. What performance figures are you looking for to consider this engine a success?

Well, the (annoying) answer to that is that it depends. The core of these engines is (literally) a Hartech large capacity conversion, so as long as we hit our target for character (pulling hard to 8,500rpm, very low inertia, very rewarding to rev it out and so on) then I'd consider it a success if we don't *lose* any power over what the standard Hartech high-capacity engines produce. Of course I'd hope that all the work we are doing leads to more power - and I do of course have some ideas of what that *could* be, based on the flow-bench, the cam profiles and the modelling that we have done. But the situation that we are in is that as long as we don't lose any power from that which the Hartech units produce we're already at a figure that pushes a 996/997 (or a Boxster, or a Cayman) down the road very effectively. All of this is a long-winded way of saying that I'm not going to give a hostage to fortune by specifying an output - lets see what the engine dyno says. Also I'm perfectly serious when I say that character of delivery is ultimately more important than what is delivered. The other thing we're doing is engineering out the issues that Porsche left in the M96, so the goal is to end up with reliability, character and power.

Hi Dammit - super interesting threat and well beyong my technical/ mechanical understanding. I have an engine that was rebuilt with latest IMS bearing, better cams and flowed head amongts other things. It produced about 340bhp on the dyno before being run in so i think some real gain over stock. it really comes alive above 5000rpm where it feels a lot livelier than before. Although i need to get the brakes to a place i am happy with first, i was wondering if there are any bolt-on engine mods that you think could add some power? induction or exhaust (i have 200 cell cats but standard manifolds)

Dammit said: shalmaneser said: What's the plan with the itbs? I'd love to fit those to my boggo m96 3.4 in time, if only for the noise! We initially looked at re-purposing some ITB's from an M3 with an adapter plate, but whilst it looked good (and would have sounded great I'm sure) the flow-bench showed that the adapter plate was killing flow. Given the not insignificant work that has gone into raising flow (both in terms of volume and velocity) of the heads themselves it would be lunacy to then write-off these gains with ITB's that were not optimised. We started discussions with a UK company that makes custom ITB's who is based close to Martin and went through a discussion that led to a CAD model, once that was signed off we went for a rough 3D print, then a higher resolution one which went to the flow-bench, and once that was signed off we ok'd the final design. I would warn - these are designed for our application, and whilst they can of course be bolted to a stock 3.4 (or indeed 3.6) the question of whether they'd give gains on a stock engine is one that we have not looked at answering. They'd sound excellent, of course. The M3 throttle bodies are very tempting as they're so widely available. I'm not that bothered about adding any numbers to the power output if I'm honest, I would be happy with cool noises and some improvement to throttle response. I'm a design engineer so I can model the parts and print them myself at work for verification as a fun side project so the only initial cost is my time. The main issues with an ITB conversion is that the M96 has negative crankcase pressure so needs a fair amount of vacuum which can be tricky with ITB's as I understand. The mapping is clearly also an issue but there are various ways around that such as aftermarket management or a piggy back system. Excited to see this motor come together!

Agree with the others that this is a very interesting project and I believe your goals are right on target. Does Baz have any interest in offering a rebuilt based on learnings from this project ( Ie higher capacity with more character and top end. ) Or is this going to be a one(Or two) off.

Well, I wish you the best of luck with it. IMO the M96 is a fantastic engine, full of character and strong horses. We all know the X51 mods were being evaluated for competition use alongside the Mezger so there is undoubtedly great potential in these old lumps

I think it’s far too soon to even think about that - we don’t know whether these engines are going to perform as hoped yet- total* humiliation may still be a couple of Dyno pulls away. If the engines don’t turn into 250kg of high speed confetti then we still have to prove that they work in the cars, will handle a trundle to the Alps and then set your trousers on fire through the mountain passes- so much still to do before that sort of question could even be asked. /*Expensive humiliation at that Edited by Dammit on Thursday 21st January 14:41

james.a.c.911 said: Hi Dammit - super interesting threat and well beyong my technical/ mechanical understanding. I have an engine that was rebuilt with latest IMS bearing, better cams and flowed head amongts other things. It produced about 340bhp on the dyno before being run in so i think some real gain over stock. it really comes alive above 5000rpm where it feels a lot livelier than before. Although i need to get the brakes to a place i am happy with first, i was wondering if there are any bolt-on engine mods that you think could add some power? induction or exhaust (i have 200 cell cats but standard manifolds) Based on conversations with Wayne at Chip Wizards who has tuned a lot of these engines in different configurations a lot of modifications (especially to the intake side) end up costing you power - not a huge amount, but Porsche designed the whole system to work together in a highly integrated fashion, so a change that makes sense on examination of the individual part may upset the balance of the system, if that makes sense? We're checking everything on the flow-bench, so we know whether it's a step forward, a step backward or no change - but we're not doing that with a largely stock engine, so I'm afraid I can't recommend anything to you. I also have the 200 cell X-pipe on my stock engine and I will say it *sounds* great, so I don't care if it's costing me 10bhp. Hopefully it's giving me 10bhp, but we've not examined that, so we don't know.

I’m running 200 cell X-pipe cats (with switchable sports back boxes) and I’m pretty certain that it has lost something in the mid range. Maybe a tiny gain up top. Despite this I love that it sounds silly when I want it to (and quiet when I don’t) so can’t bring myself to swap back to standard cats. I have been considering taking it to Chip Wizards but wasn’t sure whether it’d be a waste and/or Wayne would just tell me I’m an idiot for fitting those cats

Thanks Dammit. looking forward to hearnig results from your testing. I used to have some regular 200 cel cats and replaced them with x-pipes and also love the sound at high RPM. Would definitely not remove!

A question for those who might be reading this with a lot more ECU knowledge than I have - the thinking was to use Syvecs, but I am unsure what model would be best, and I admit I don't understand the difference between Life Racing and Syvecs and why one would choose one over the other?

When considering an ecu, you need to think about what features you want. Drive by wire, single wideband or one for each bank, how many inputs you want to add (oil pressure, oil temperature, etc), what outputs you want to control (a/c compressor, tach signal, electric water pump, exhaust valves, etc). Then you need to think about canbus support Ideally make a list and see what suits your needs best. No point buying something expensive with features you won't use. The other consideration is who is going to be wiring it up and tuning it. Most tuners will tune anything but some stick to what they know. With regards to the M96 Cup engine build, have you thought about ditching the mechanical water pump for an electric one?

Dammit - I assume, because I know how thorough you all are, you've already seen this? https://verbotenmotorsport.myshopwired.co.uk/scave...

I'm afraid you give me too much credit - I wasn't aware of this product until I saw your post. Very interesting! It's reassuring that we're not the only people who believe that this is a key vulnerability/required fix for the M96 (as well as Porsche I suppose, but of course they don't make the dual-stage pumps anymore).

We have been running engine simulations recently - something we've done quite a lot of over the course of this project as we can't afford to make 10 intakes, 10 exhausts, then run every combination on the engine Dyno to see what works best. This one is the final sim that we've used to specify exhaust manifold primary lengths and internal diameter, these are now being made by a very pleasant chap in Silverstone. Also, because if you looked up "feature creep" in the dictionary it would reference our engine project, we're making a complete exhaust now (this was never the original plan), we always intended on making a true equal length manifold, but given the (surprisingly large) diameter that we need for the primaries and what that then meant for where the flange had to be, we've had no choice but to do as a minimum manifold and cats, so we thought we may as well go the whole hog.

Great stuff!

always excited to see developments in this project!

Dammit said: We have been running engine simulations recently - something we've done quite a lot of over the course of this project as we can't afford to make 10 intakes, 10 exhausts, then run every combination on the engine Dyno to see what works best. This one is the final sim that we've used to specify exhaust manifold primary lengths and internal diameter, these are now being made by a very pleasant chap in Silverstone. Also, because if you looked up "feature creep" in the dictionary it would reference our engine project, we're making a complete exhaust now (this was never the original plan), we always intended on making a true equal length manifold, but given the (surprisingly large) diameter that we need for the primaries and what that then meant for where the flange had to be, we've had no choice but to do as a minimum manifold and cats, so we thought we may as well go the whole hog. I mean, you should definitely try to monetise some of the development you've put into this. I would love to see some cool upgrades for the 996.

Bottom end is going together - an exciting moment (for me, anyway!): We've been doing a lot of thinking about the exhaust system, what's absolutely certain is that we need to make our own manifolds as nothing available off the peg is going to match our requirements. One possible route forward would be to use an OTP GT3 cat section and silencers with our manifolds, as the diameter of those should work. More thinking to be done.

Very cool

Those are some fancy looking rods!

Had to have rods that put the piston where we needed it, which sadly the OEM ones were not going to do. These do look nice, also, which is a bonus. We're going to run the first engines on Syvecs, specifically the S7Plus - if anyone has an unused S7Plus gathering dust in the garage please let me know and I would very happily buy it from you (chance would be a fine thing, I suspect, but no harm asking).

I'm guessing stuff is happening but I see no posts here!

Stuff is happening, yes - but achingly slowly. We're waiting for the performance tappet chest design (smaller base circle, and all the implications of that for oiling, tappet position and so forth) to reach the head of the queue at the CNC place we're using. In the meantime we have produced a run of the dual stage scavenge pumps to see if anyone else wants one.

I've just found this thread. Outstanding stuff Dammit. Viva la M96! :-D

I watched the video but am none the wiser

Well, if we're wrong then at least we are *very precisely* wrong. Next: camshafts.

It's a machine for making sure the metal is where the designer said the metal should be.

Correct - seemed sensible given that the rev limit for these is above 8,000 rpm, and therefore knowing that your valve train is where you meant it to be is comforting.

shalmaneser said: It's a machine for making sure the metal is where the designer said the metal should be. Yep, pretty standard inspection equipment but it does look impressive, whizzing around, to the untrained eye. When are you going to get one in a car though Neil?? I told you, you should have bought a GT3. You’d have been enjoying it all this time, instead of standing over CMMs.

Where's the fun in getting the benefit of other people's R&D efforts, when you can forge your own path? I'm looking forward to hearing about how this gets on as the engine comes together; I really enjoy the ambition and technical detail you're sharing. This really looks like it's being done properly ! Hopefully you're keeping notes, as once the new engine is on the road and running smoothly it feels like it should really become a video or article of some sort, whether a PH Readers Cars vid or elsewhere.

It's late on a Monday, halfway through December - must mean it's time for some charts. These are all based on flow bench data - and a caveat, before we continue, is that whilst air flow is indicative of likely performance it's absolutely *not* definitive, it's one of a series of inputs that determine overall power and torque. Therefore the following should be taken in the spirit with which it is presented - as interesting, but nothing more than that. First, the baseline - a bone stock M96 on a bone stock 96mm bore: A range of companies now offer a capacity upgrade as part of major engine repair/upgrade which increases the bore size to 100mm, so next we'll look at the effect which that has on flow (in effect, the 3.4 in the previous chart is now a 3.7, with the valves slightly unshrouded by the increase in cylinder diameter): Close, but there's a definite increase for both intake and exhaust, with the exhaust side increasingly (relatively) more. Now lets compare the stock 96mm M96 to our modified design, which is on a 100mm bore: That's showing a significant increase, but what does that really mean? Lets compare with a couple of different Mezger heads to try to put it into perspective. First, a GT3 head: And, because we had the figures for one, here's a 997 Cup head:

Very promising - great work!

Remarkable. Oh, to be a fly on the wall in Stuttgart back in the day. I wonder what decision making criteria they used to decide between the Mezger, the M96 and the X51. Versus what you’re doing, does the X51 have any common threads (from a philosophical point of view, perhaps)?

I think it's unquestionable that the Mezger is a better engine - I imagine Porsche reviewed the X51 program and decided that they wanted something with more headroom for development/more natively suited to being a racing engine. And, of course, since that point everyone who has ever campaigned a water-cooled Porsche platform has made the logical choice to use the more powerful and reliable engine, for which there's a huge performance aftermarket (try buying a sequential gearbox for an M96, for example). But I don't race, and I like the "what if" element of the X51 story - and to answer your question I think the people who worked on the X51 project would see a very great degree of similarity with our engines, as we've followed the same route as they have. Cams with higher lift and more duration, reworked heads, new intake and exhaust manifolds and so forth. We have, I admit, taken things a bit further than they did - when discussing this engine with the friends with whom I am working on the project I tend to refer to it as the X51-Evolution engine as it seems to be a good descriptive fit. And, of course, our engines may be a total, abject failure when they run up on the dyno for the first time - but the flow-bench results are encouraging.

I had missed this thread completely, and was intrigued by the references to your engine from others on 911uk and here. Fantastic attention to detail

Interesting that the GT3 exhaust shifts more air than the cup motor... Results certainly look promising but as you say this is just part of the puzzle! Exciting to see progress nevertheless.

Yes, that's a great example of raw air flow only being a part of the equation - the Cup motor has no VarioCam, so the design is optimised for race cams with fixed timing. The GT3 engine has more options at it's disposal to manage timing and so forth. Our billet tappet chests are now done, next on the list is to get the cams ground and (hard to believe I'm typing this, in some ways) then building the short block up to be a complete engine.

shalmaneser said: Interesting that the GT3 exhaust shifts more air than the cup motor... Results certainly look promising but as you say this is just part of the puzzle! Exciting to see progress nevertheless. I believe the valve sizes are the same on the Cup and GT3 heads, but the ports in the Cup head are bigger, this should theoretically mean the flow numbers would be better, though if the valves are the same size as the GT3 road car, the only way you'll get more flow (and thus power) is to open the valves for longer and open them further, which is exactly what the Cup does with high lift/long duration exhaust cams and the deletion of the Variocam to enable such cams to be utilised reliably. Edited by Slippydiff on Tuesday 14th December 08:21

Dammit said: But I don't race, and I like the "what if" element of the X51 story - and to answer your question I think the people who worked on the X51 project would see a very great degree of similarity with our engines, as we've followed the same route as they have. Cams with higher lift and more duration, reworked heads, new intake and exhaust manifolds and so forth. We have, I admit, taken things a bit further than they did - when discussing this engine with the friends with whom I am working on the project I tend to refer to it as the X51-Evolution engine as it seems to be a good descriptive fit. Good luck to you. It’s a very interesting project. To understand the value of the X51 on the options list you need to have a fairly deep appreciation of base engine design to see the “value”, as the quoted power rise was fairly low in the brochure vs the cost (almost £10k IIRC). But unquestionably I would spec it on my car and in time I might accumulate enough parts to build X51 spec motor by piece. You seem to be putting out some healthy numbers in your charts!

Dammit said: The chaps from Project Binky recommended the engineering place that machine these- for which I’m very grateful as their work is excellent. We now have a set for each of the three engines we have in build. Here with the stock version: And here with the new, larger valves, titanium retainers, and a 3D print of the ITB’s we’ve designed to check dimensions and fit. Fixed formatting. What throttle body are you planning on using?

We're making them for this project - they're not from anything else.

The chaps from Project Binky recommended the engineering place that machine these- for which I’m very grateful as their work is excellent. We now have a set for each of the three engines we have in build. Here with the stock version: And here with the new, larger valves, titanium retainers, and a 3D print of the ITB’s we’ve designed to check dimensions and fit. Edited by Dammit on Wednesday 22 December 23:15

Cool excited to see those too!

I admit, they're largely there because of the small boy inside me who wants to hear an over-square, cammed, tuned flat six howling through throttle bodies at 8,000 rpm. Of course, there are performance reasons, but the noise should be incredible - and bear in mind my car has no roof to get in the way of appreciating it.

So is this a part that's a component of the head, sitting between valves/ports and the cams? Is it a common design feature on other engines?

Yes, that's it - I think it's more commonly known as the lifter carrier. The stock version is a cast part that can fail (hydraulic fracture is the most common cause of failure IIRC) which is, as you might suspect, Bad News for the engine: Our dual oil scavenge pumps should reduce the chance of this significantly as they prevent the type of oil build up that may lead to the carrier fracturing, but there are other reasons to move to the billet ones in the picture. We looked at billet cams, we looked at sourcing cam blanks, and we came back to re-grinding original cams as the best blend of cost/performance. This means that the base circle is going to be smaller, so the lifters will ride higher in the bore, so the bore itself needs to be adjusted to prevent the lifter tipping, and the oiling needs to be modified slightly to accommodate this. We're also spinning the engine a lot faster than Porsche did for the X51 series, so valve train stability and strength is important. We can't modify the stock items to meet all these goals, so CNC'ing new ones became the logical choice.

Dammit - I'm loving this thread. Have been following for some time! I've been looking at how I can make my C4S the best it can be, and am getting there. Brakes, set-up, geo, suspension is pretty much all there now. The only thing I've got left is the engine. I was looking at the 3.9 Hartech but I can't help feeling I'd like a few more RPM. I know that torque is king etc, and why would you want more RPM, but for me its also about the experience. I do wring my C4S out to the red-line regularly and love it. How is progress since your last post? Is the engine together yet? I'll be following this closely as you near completion - and certainly giving Baz a call if/when you are hopefully successful! Good luck!

Very impressive stuff! Wrt suppliers of custom camshafts I'd like to mention the supplier to Porsche of the original M96 camshafts (and probably many other Porsche engines prior and after): https://www.schleicher-fahrzeugteile.de/en/product... They're based in Munich and have supplied BMW Motorsport too. They are equally well regarded as Schrick are and offer prototype manufacturing too. Edited by hardtailer on Saturday 22 January 10:58

Impressive - looking forward to updates

Is this finished yet?

My old x51 996 Carrera 2 loved it……

Any progress on this project? Keen to hear how you are getting on!

Are there any updates on this incredible project?

We have the cams from Piper, to end is going together and we now need to order the ITB’s and equal length (this bit is both important and non-trivial) exhaust manifolds. Progress continues to be made, but sometimes more slowly than we would ideally like.

Have you seen the posts on Facebook (yuk) by Jake Raby about cam timing retardation across the cam itself? He seems to suggest the individual cam lobes are retarded across the cam itself for some reason. I suspect it might be manufacturing or measurement issues but he seems convinced it is deliberate. I personally find him pretty irritating but that was quite interesting to me. Maybe this is a common thing on production engines for one reason or another, some cylinders getting less airflow than others for example.

shalmaneser said: Have you seen the posts on Facebook (yuk) by Jake Raby about cam timing retardation across the cam itself? He seems to suggest the individual cam lobes are retarded across the cam itself for some reason. I suspect it might be manufacturing or measurement issues but he seems convinced it is deliberate. I personally find him pretty irritating but that was quite interesting to me. Maybe this is a common thing on production engines for one reason or another, some cylinders getting less airflow than others for example. Haven't see the FB post, but a look at the exhaust manifold design may explain why there's a difference in the cam timing from one cylinder to another. For all the scorn heaped upon the 3.4 engine, it was clearly very cleverly optimised by the engine development engineers at Stuttgart. All the more so as I'm guessing there were many questions asked which were responded to with "Nope, we don't have the budget". Mr Raby "pretty irritating", ? I'd say that was being very diplomatic ... But love him or loath him, there's no doubt he knows these engines inside out, and is one the few individuals who's worked out how to extract big horsepower from them, but more importantly, to do so reliably.

Slippydiff said: shalmaneser said: Have you seen the posts on Facebook (yuk) by Jake Raby about cam timing retardation across the cam itself? He seems to suggest the individual cam lobes are retarded across the cam itself for some reason. I suspect it might be manufacturing or measurement issues but he seems convinced it is deliberate. I personally find him pretty irritating but that was quite interesting to me. Maybe this is a common thing on production engines for one reason or another, some cylinders getting less airflow than others for example. Haven't see the FB post, but a look at the exhaust manifold design may explain why there's a difference in the cam timing from one cylinder to another. For all the scorn heaped upon the 3.4 engine, it was clearly very cleverly optimised by the engine development engineers at Stuttgart. All the more so as I'm guessing there were many questions asked which were responded to with "Nope, we don't have the budget". Mr Raby "pretty irritating", ? I'd say that was being very diplomatic ... But love him or loath him, there's no doubt he knows these engines inside out, and is one the few individuals who's worked out how to extract big horsepower from them, but more importantly, to do so reliably. Hah - pretty diplomatic indeed.... As far as I know the reason for the unequal length primaries is to spread the effective amount of exhaust scavenging across the rev range, single length primaries will only tend to do that at one specific exhaust frequency/revs. I've not heard about tweaking the cylinder timing on a per-cylinder basis.

Exhaust manifolds - we have an update. Our engine designer stated that (amongst other dimensions) 27" was the ideal length for the primaries. Anyone reading that sentence who has ever been underneath a 996 will probably be wearing a slightly bemused expression - but luckily when you are designing the entire exhaust you have a degree of flexibility over where things go. Yes, we, erm, moved the silencers and used that space to fit in 27" truly equal length primaries and a nice merge collector, before turning the exhaust back toward the rear of the car. Out of the car and from the side this looks like this: Looking down through an imaginary decklid you can see the overall packaging - this system will have 200 cell cats (removable) in the straight sections that come down from the merge collector turn to the start of the X pipe: And from under the car: The silencers are repackable and the system is designed for 105dB maximum - although with the ITB's I suspect that track day noise limits are going to be interesting. As this is Pistonheads in 2023 I understand that I can't post without talking about values, so: in order to make building a car in the UK easier I moved to Singapore at the start of the year. Cars here are *expensive* - a 996 in good condition is a hundred thousand pounds, minimum. I went to see a BMW 2002 from 1976 which (as a classic) was actually fairly affordable - chap who owns it owns around 70 classic cars, and is moving some out in order to make space for some new 911's, which he is intending on importing from the UK due to the price difference (UK cars are, even with import duties, stupidly cheap in comparison). Edited by Dammit on Friday 8th December 08:51

Awesome update. Love that exhaust setup!

Lovely...

That system looks “strangely familiar” Neil ...

Typical long header design...

Wonderful, are you now close to a dyno run to establish the power curve and then look at gear ratios?

My suspicion is that there’s a logical way to package long primaries and this is to put them where the stock silencers sit- I imagine you could make it work if you put then somewhere else, but at what cost and why are two questions to answer before so doing.

Well, we are closer than we have ever been- but we now need the ITB’s to be manufactured and then we need to resolve the four thousand little tasks that will no doubt crop up. I’m spending time on the turbo thinking about vacuum, ITB’s, and things like crankcase ventilation, brake booster efficacy and so forth.

Is the engine going to have a turbo? I thought it was to be a NA screamer.

Turbo-trainer - an indoor cycling trainer. Useful for allowing the mind to wander when trying to ignore the signals your legs are trying to send you suggesting that they are literally on fire. The plan for the engine is NA, and that won’t change.

Ahh... Reminds of the pro cyclist who used to have 'Shut up, legs!' painted/taped on his bike.

TDT said: Typical long header design... For a 997/991/992 with side deletes maybe ... The usual design for a M96 or M97 is akin to this : https://www.fvd.net/adbimage/100615/shop_detail/be... Or the 996 R/RS/RSR items : Edited by Slippydiff on Friday 8th December 14:55

I’ve got some photographs of the full exhaust system, equal length manifold included, which I’ll post shortly - it’s been on a friends car and in his words makes the flat six sound rather like a (very large) motorcycle engine. In less entertaining news: I moved to Singapore for work a few years ago and I’ve gradually come to terms with the reality that I’m not going to be able to continue this project. If anyone wants an engine that (in theory!) should behave like a 997.1 GT3 engine, but will bolt straight into a Carrera chassis - please let me know. It’d be a great shame if this engine sat on a shelf for the rest of time. Head back to the first post to see what I’m talking about - but be prepared for quite a lot of reading. (Pistonheads is blocked from Singapore IP’s, so I’ve not been on here very much and therefore I’m not really up on Porsche discourse here - so if M96 with the weaknesses removed, 8,300 and a lot more power are now commonplace please forgive me).