There are some reasons to do this, and a lot of reasons not to.
Who is writing this and why? (Updated February 2022)
Steve Timmins, Ph.D. Mechanical Engineer. Started as just a regular PCA member and track junkie in 1988 finishing up my undergraduate engineering degree and moving on to my doctorate, I was on the other end of the wrench, as you might say. Don’t worry, I’m not some trust-fund jackhole…I was working full time at UD in IT as a systems programmer, and taking a full course load and I did 18 track weekends my first full year. 42 days…which is how I ended up as a PCA instructor in 1991. In 1992 I transitioned from 944turbo to 3.0 liter 911, then a 3.4, then a 3.9, and finally to a 3.8 Varioram. I currently own 30+ Porsches including three GT3 cups, a 2007 GT3, four 996tts, three 993s, a couple of 964s including a turbo, a pair of 944 turbos, and the rest 3.6/3.8 conversion 911s, all of which have VariorRams.
Over these thirty plus years, including the twenty five since the shop opened (November 1997) I have owned pretty much every engine combination you can imagine, and I absolutely put my money where my mouth is.
As you may or may not know, I teach Automotive Powertrain Theory (engine design and tuning) at the University of Delaware, in addition to Vehicle Dynamics of Race Cars. and Engineering in Motorsports and FSAE Senior Design. These are senior-level mechanical engineering design course centering on internal combustion engine design and involves thermodynamics, kinematics, statics and dynamics, heat transfer, fluid mechanics, combustion theory…blah blah blah. In the Fall semester, in addition to my full-time gig as a UD IT server administrator, I teach the double-credit senior capstone course FSAE Senior Design, in which a dozen very dedicated students design and build a race car from scratch in a single semester. I am also the FSAE Faculty Advisor and am off to the international competition with our team in a few weeks. It’s a bit complicated, but let’s just say I know a bit about cars. What I say below is what I have learned by spending the money myself, not by marketing products. We don’t advertise, never have. Our customers come to us…just as you did…by seeking us out. (Update May 2019) In a few weeks, I retire after 32 years from UD IT, retaining my UD Mechanical Engineering faculty position, and my primary focus will be the shop.
The Basics – it’s all about the air
There are only so many ways to make power but they all boil down to the same thing. Burning fuel. Bottom line is that at some basic level you can burn 1 part fuel for every 14.7 parts of air by mass (weight, less precisely), which is known as the stoichiometric ratio. If you try to burn more fuel you simply cannot, because it’s a chemical reaction. Performance engines can run a higher fuel ratio, say 13:1 (air is always stated first) but that just means that some unburnt fuel goes out the exhaust pipe. Throwing more fuel at a motor means each molecule of oxygen stands a better chance of hooking up with a fuel molecule…at some point it’s too much.
(NOTE: The following is no longer politically correct, too bad) Think of it as a high school dance, and some boys will go home alone, if you have the same number of boys as girls, some of each will go home disappointed, so if you want to make all the girls happy (O2) you want some extra boys around. Not the most efficient but oxygen is the limiting factor. At the end of the day (or combustion cycle) what really matters is how much air you can get into the motor AND burn with fuel, which explains turbocharging…another matter for another day.
Volumetric efficiency (Vf) is the ratio of air inducted (sucked in) in any one combustion cycle versus the displacement of the engine. If you do real well you can get a Vf of 1.0 and it’s possible to get more. This depends on induction system (intake manifold, air filter throttle bodies, efficient scavenging (exhaust sizing)), etc. and cam and valve selection plus combustion chamber and piston design. Lots going on. Maximizing Vf is the primary way to get more power out of an engine of a given size and design.
Back in the day, like the 1980s, the only way to get more power was to modify the 3.0 or 3.2. 2.7s and smaller motors have some design flaws and, other than fitting into a certain race class, are poor things to start with. So now for the engines. If you have a race class that demands a 3.4 you can mod a 3.2, but many have found it better to de-stroke a 3.6 and take advantage of the built-in superiority.
3.0 Carrera/SC and Variants.
The basic 3.0 was the expansion of the 2.7 but was a return to the aluminum case, which Porsche had used in the 356 engines and in sand-cast form on the 2.0S. The 2.2-2.7 engines had the lighter but much weaker magnesium case (engine block, sorta). The new aluminum case was introduced for production in 1975 with the Carrera 3.0 and turbo and had a 70.4mm stroke and a 95mm bore resulting in 2994 ccs. They came with CIS (read vacuum driven fuel injection) and a rudimentary and completely separate ignition system. They are reliable until something in the CIS system goes bad (thanks ethanol) and expensive to fix. A fuel distributor, warm up regulator and injectors put you over $2K for rebuilt units plus labor. So you can upgrade these by using 98mm cylinders (a.k.a Max Moritz) to get 3.188ccs. The increase of displacement of 6% or so will increase power a similar amount. 3.0s made between 180 and 208 HP based on US or Euro compression ratios and intake runner size. Anyway…small ports and low compression or larger, these were great when a Corvette or Mustang made 200 hp, but…well…it was the still emissions era when Jimmy Carter also forced those 85mph speedometers on everything sold in the US.
Throttle response with the CIS is terrible (you have to open the throttle, lift a metering plate then it gets more fuel and the intake is restrictive, so the HP numbers don’t tell the story). A common misconception is that CIS is somehow electronic fuel injection because in 1980 Porsche introduced “Jetronic”, which is a hack job that modifies fuel pressure based on oxygen sensor reading and really only does anything at light load to help with emissions. In general it does more harm than good. To get any real performance you need to get rid of the CIS in favor of carburetors (terrible gas mileage, finicky in cold weather, there’s a reason Porsche ditched them) or “real” fuel injection, which is expensive. Carbs cost about $3,000 (well they used to!) and will increase power and throttle response, kill fuel economy (12 mpg ouch) and even on race cars are nowhere near as good as “real fuel injection”. More about that later as it applies to 3.2s. In reality getting more than 245HP out of these engines on a reliable basis gets very expensive.
Beyond port sizes and intakes the other things that affect volumetric efficiency (power) are cams and valve size, exhaust etc. Because of the shape of CIS pistons, which have an odd lump on them to get the fuel and air to mix properly, aggressive cams (again with the poor gas mileage and difficult idle/low RPM performance etc.) can’t be fitted without replacing pistons and aggressive cams mess up the CIS system so, in for a penny and all that making real power means replacing pistons, cams, valves, enlarging ports, fitting carbs or fuel injection. Compression ratios on these motors are generally low (8.7:1 up to 9.2:1 Euro) which also affects power, bumping compression ratio which means twin plug (look that up, two spark plugs per cylinder, which means a stand-alone or integrated ignition system, so to make, say 300 HP you will be socking away $15-20K in mod, injection stacks, management, tuning, etc. to get a 3.2 that makes all its power above 500o RPMs, idles at 1500, does not like to run cold and gets terrible gas mileage. The exhaust on a 3.0 is terrible so that gets tossed first…let’s call it $20K plus the value of the engine itself and lots of downsides. You can also bump the size to a 3.4 by using a Carrera crank and the larger P&Cs, which cost $4K). It’s not hard to blow $30K all in..again for a motor that isn’t all that wonderful and can be downright painful to drive at reasonable street RPMs and gets crap gas mileage.
The 3.2 Carrera engine is a 3.o with a longer stroke (74.4mm yielding 3,164 cc) and real fuel injection/ignition management. The HP is rated at 217 or 235 depending on what you look at and US vs Euro, but the 235 is pretty optimistic. They have larger intake ports, the same valves, similar cams, and the same design lump pistons which make fitting aggressive cams impossible. A stock 3.2 with good exhaust and a set of 964 cams and tune will make 245HP. Beyond that the same route as the 3.o applies, with carbs or fuel injection stacks, management, etc. etc. and while a 3.2 is a better starting point than a 3.0, the twin-plugging needed for higher compression ratios and cams and…..whatever. The largest displacement reasonable (same as above) is a 3.4 and you have now spent $15K+ to get anywhere near 300 HP and that is all above 5000 RPMs. Here’s some discussion of my very expensive 3.4, which I ditched in 2002. It broke at a track event at VIR on a Friday and I drove back to the shop (360 miles), pulled a 3.6 out of a crate, did the conversion, and was back at the track (360 miles) by noon the next day. Never looked back. That motor still sits on a stand.
Turbocharging and Supercharging
Forget it. Modifying a normally aspirated engine for forced induction is complicated and expensive and can also have reliability issues. If you want a turbo motor buy one or plan on $35K to build one right. Tons of power. Not relevant to mortals. If you want that route that’s fine. Take it off the table for this discussion. Stock Porsche air cooled turbo motors are really laggy and have terrible throttle response at low RPMs, yeah I own a couple of 930s, more CIS and up until the 993 twin turbo they are archaic. The 993 tt motors are bringing $50K now (2016) and building one can approach that with the proper management. 996 twin turbos are wonderful but the plumbing or both air and water are not for the timid. Supercharging, been there, done that and its sitting on the shelf. No thanks, this isn’t a Mustang. If you have the courage and a stout budget let me know, we have 7 996tt engines in stock as of May 2019.
Enter the 3.6
In 1989, midyear, Porsche introduced the 964 (Carrera 4) with a brand-new 3.6. The pretty mich the only parts they kept were the connecting rods and rockers, both from the 930 3.3 turbo model introduced initially in 1978. Almost everything else, from the case to the oil pump to the magnesium chain housings is different. Many parts used metal and/or rubber gaskets so leaking is reduced. The engine uses twin-plug ignition with 11.5:1 compression ratio, features lots of plastic parts for light weight and features 76.6 mm stroke and 100mm bore for a true 3600 cc displacement. That’s 14% more than a 3.2 and 20% more than a 3.0 so just based on displacement those power increases are to be expected. The increase in compression and efficient induction, twin plug ignition and more modern management mean that while the 964 motor is only rated at 245 HP, that is mainly because they kept the same junk exhaust system that limited the 3.2 to 217 HP. On displacement along the increase from 3.16 to 3.6 puts the motor at 268 HP if you believe 235 and mods scale appropriately. The reality is that a 964 motor with good exhaust and a set of appropriate chips will make 285HP completely stock. With that it will get 25 mpg highway, start when it’s freezing cold out and the drivability is lightyears ahead of the 3.2 because torque at 2500-5000 RPMs is 50% higher. No more downshifting to pass, just step on it and go! These are great motors, just watch out for 1989-1991 models since they have serious head gasket issues .
The 993 engine was introduced for the 1995 model year and has a few changes from the 964. Most notable (and difficult to upgrade) is the use of a hot-film mass flow meter, as opposed to the old “Barn Door” meter introduced with the 3.2. In reality it’s more accurate but has almost nothing to do with power. The increase in stock rating from 247 to 272 (10%) is almost entirely due to a real dual exhaust system with better though still unequal length headers and a split catalyst system. The engine also got hydraulic lifters (no more valve adjusts!) and a lighter rod/crankshaft package but retained the 964 later model plastic intake. The part number changed because of a mount but that plastic intake lived through the 996 twin turbo models.
In 1996 we got the Varioram intake. This and some larger valves upped the HP rating to 282 HP. Again, in reality the 95 motors and 96-98 Varioram motors are indistinguishable above 5,000 RPMs on the dyno, but the 20% increase in torque is amazing. These engines are the best, but you can get nearly the same performance with a 95 motor and with a Varioram intake strapped on and the appropriate chips. See this article for a detailed explanation.
Net Cost and Reversability
When considering a 3.6 upgrade, think about bottom line cost. Realistically the conversion, all in, costs somewhere in the $17-22K range. Rebuilding a 3.0 or 3.2 stock will set you back 10-12K. Additionally if you upgrade to a 3.6, remember that you have that as an asset which you could sell, keeping your original motor, or, if you don’t care about going back to original, there is a substantial core value to the 3.0 or 3.2 and nice examples are starting to nudge the $10K mark. Our 3.6 conversions are completely reversible and use NO parts from your existing engine.
Still interested in more power?
There are tons of upgrades to the 3.6. Cams make a lot of difference and will get you to 320 HP or so but need solid (not hydraulic) lifters and may need to have the pistons modified. We have lots of pre-configured solutions up to 3.8 liter. We have built 3.9s but have had reliability issues. If you need more power hotrod the 3.8, the extra displacement beyond that gets expensive and thin cylinder walls are unstable. Here is some information about the 3.8.