Home Gallery Fire Specs Vac diag Hardware Seals Faults

OZ951 Modification Notes - Last updated October 2004

Pre fire                                                                   Post Fire


 My 951 has gone through a few different stages of work and modifications. Initially I was rebuilding the car because there were lots of broken parts.  In doing so I upgraded various parts to aftermarket performance parts. That got me some decent power improvements. Then I realized I needed a better ‘system’ of parts so I had to change various things to make sure that all the parts were working together effectively. That worked well and I ended up with plenty of power. Then I took the car to the track and realized that I would like it to stop and go around curves as well so I went into a number of suspension upgrades. Then I started to wonder about reliability and drivability so I went into a bunch of reliability and drivability improvements that were not aimed at more power. And finally after having the engine fire and some threatening looks from my girlfriend I decided to improve the safety aspects of the car as well. Not all of the work is complete as I type this but I was already very pleased with the car before the fire and I know that it is going to be simply awesome to take it beyond that level. I think that I have ended up with a well rounded 951 that is an awful lot of fun to drive hard and yet still be highly street-able and reliable, and I still have one arm and one leg left!

Garrett TO4E Turbo

to4e.jpg (121698 bytes)

Garrett TO4E




101-0108_IMG.JPG (81462 bytes)

Oil feed line to turbo (top) and coolant outlet (bottom) Oil

The turbo in my car is a Turbo Engineering Co (TEC) Garrett hybrid TO4E. This turbo combines a Garrett cold side with a KKK hot side around a center housing constructed from Ni-resist (22% Nickel). T4s were originally designed for diesel engine applications which operate with different characteristics ie temperatures around 1400F, than gasoline engines with turbo temps around 1700F. The Ni-resist  center housing allows the TO4E to withstand the hotter temperatures experienced in the gasoline application. The turbo is both water cooled and oil lubricated.

The turbo is working well, its slightly laggier than stock but delivers power at the top end and shows no sign of abating at redline. The Garretts are smokier than the KKKs. I have gone to a crank breather to eliminate drawing oil laden crank gases into the turbo and that reduced smokiness but didn't eliminate it. 

Several 944 owners have gone ahead with an oil restrictor plug in the oil feed to the turbo and that has eliminated the smoke issue. The restrictor cuts down the feed outlet internal diameter to around 1/16th".  The manufacturer doesn't recommend it (no real surprise there) but those who have done it have not had any longevity problems that I have heard about. The problem is supposed to arise from the fact that the 944 produces such high oil pressure that some of it is forced past the seals in the turbo center section and is subsequently burnt with resultant smoke. The KKK turbos seem to be designed to cope with it whilst the Garretts seem to prefer much lower oil pressure (as is consistent with Corky Bell's Maximum Boost). I know of one or two club racers that have eliminated the water cooling for their K26 turbos and rely on the large volume of oil flow to the turbo in the  944 to provide both the cooling and lubrication for the turbo.

My Garrett turbo has a larger compressor housing than the KKK K26 and so requires a few modifications to fit. I performed  the following  modifications to fit this turbo 

a) raising the coolant pipe to clear the compressor housing by using common brass fittings. 

b) Grinding some material away from one of the lugs on the compressor housing (to allow more space between manifold and compressor)

c) Replacing the screws on the throttle return spring bracket with countersunk machine screws. (I tried rivets but they exhibited a tendency to loosen up and that could have bad consequences (potentially throttle stuck open).

d) Grinding some material away from the alternator casing to allow greater clearance  for the intake plumbing to the turbo (others have gotten away without having to do this)

e) Use of larger diameter intake plumbing to suit the larger turbo inlet.


turbomods2.jpg (50817 bytes) turbomods1.jpg (72765 bytes) alternatorfiled.jpg (51178 bytes) turbopath.jpg (73640 bytes)


SEP 04

turbo_coolant_fitting.jpg (61245 bytes)

oil-restrictor.jpg (59855 bytes)

At my third track day at Mallala with about 8,000 miles since the rebuild I had a failure of the turbo coolant line brass adapter. Basically it sheared off leaving part of the brass nipple still threaded inside the turbo center section. In the pic which is above and to the left, the brass joiner can be seen and the brass nipple (with the thread tape on it) can bee seen connecting the brass joiner to the turbo. It basically fractured in the middle of the thread tape. I subsequently learned that this has happened to others who have done this type of modification, particularly if  the support bracket was not utilised after performing the mod. That's likely to be common since the  mod means that the factory bracket which normally supports the coolant pipe, does not line up after the mod is performed. My solution is to replace the brass nipple with a thicker and stronger steel item and to fabricate a bracket to provide additional support to the coolant pipe in its modified location. There was no steel nipple available to I had to fabricate that too in the end but at least that meant it had the correct standoff and the correct M16 x 1.5 thread to suit the turbo.

Whilst I was in there I also popped off the oil inlet line and installed an oil restrictor to the turbo oil inlet. The restrictor was a brass allen socket head plug that I drilled a 2mm hole through and then used a countersink bit to provide a conical profile to the restrictor hole on the side that faces the turbo bearing. This is to improve the oil spray pattern as it exits the restrictor under pressure. Since the allen plug has a tapered thread it will not screw all the way into the threaded hole that already exists in the turbo center section hence it is necessary to grind away several mm of the plug before installing it. That still leaves sufficient material for the allen key to fit into and drive the plug into place. As expected the oil pressure at idle is higher than it has been in the past, and my car has always had quite good oil pressure anyway.


Boost control. 

Like many folks I turned to manual boost control initially in preference to the stock setup. The manual boost controller (MBC) is a simple valve that restricts how much boost pressure gets to the wastegate. Hence if you set it so that almost no boost pressure gets to the wastegate the wastegate will open late and thus boost will build high. If you set the MBC so that all of the boost pressure gets to the wastegate then it will open at whatever pressure the wastegate spring is set at. I ran mine with the stock wastegate initially. That setup is simple and works pretty well, but I did find there were a few things that I didn’t like about it. It had a tendency to overshoot the set boost level by about 1PSI when you stomp on the gas. This tendency may well have been a combination of the MBC and wastegate characteristics. A better wastegate may not have had the same overshoot issue. I also found that setting the things to achieve the correct boost level involved a lot of trial and error because the boost controller was sensitive to small changes in the control knob.. It was a PITA for me because I mounted the MBC in the engine bay and had to stop and open the hood to make the changes every time. I mounted it there for two reasons, 1) to keep the tubing runs as short as possible so that wastegate response could be improved (this could be an urban legend, I never experimented with it) and 2) to keep inquisitive hands from playing with the controller and setting some ridiculous boost level that would grenade my engine.

 After I did my first dyno runs in the car I saw a couple of things that needed attention, 1) the fuel mixtures were quite rich (good for safety of my engine but not for performance),  and 2) boost could clearly be seen bleeding off once revs started climbing. Another thing I noticed at the dyno is just how much smoke the Garrett turbo produces, but that is really only a cosmetic thing. I sent the dyno charts to Danno and Travis and they sent me back some new numbers to plug into the Link MAP via a hand controller. Then I shimmed the stock wastegate. These two things improved the way the car pulled but I could still see that boost was bleeding off. Next I pulled out the stock wastage and had the performance shop weld in a deltagate wastegate. That has fixed the boost drop condition at revs. A few people have turned their noses up a bit at the deltagate but I really cant say I have had any real problems with it at the 300 ish RWHP level. My current round of mods is aimed not at more power but at getting to the power quickly, that may be a reason to swap out the deltagate for a bigger wastegate but I wont do that until I see how things are going with the Apexi AVCR which I have just installed. My plan is to provide the means for my engine to rev up and develop boost quickly and then be able to dump that boost quickly when it gets to the desired level. That will require free flowing intake and exhaust, and a large wastegate and blow off valve. (More Information) For the BOV I went with a Forge 006. It is not a rebuildable unit but has a sturdy mechanism that should withstand high boost levels

wastegate.jpg (47028 bytes)  

Deltagate now replaced with TIAL 38mm.

mbc.jpg (68420 bytes)  

Manual boost controller 

apexivalve.jpg (48099 bytes)

Electronically controlled valve for Apexi AVCR

bov.jpg (41430 bytes)

Forge 006 BOV

Apexi AVCR (Electronic Boost Controller)  













I installed my Apexi AVCR with the control solenoid in the old cruise control area and the pressure sensor in the DME area. The control unit I mounted on the dash where I can see it clearly. I referred to the Apexi setup guides that were posted by some Rennlist folks. They covered most of the pertinent information but it is also handy to know that during the initial setup of the AVCR you must tell the AVCR how many pulses are generated per revolution (I think) anyway the answer is that the 944 does 6 pulses but the AVCR does not have an option for 6 so you pick another number and accept that the speed displayed by the AVCR  will be incorrect. The final setup task is to nominate the TPS characteristics in the setup (etc) menu of the AVCR. The TPS voltage should be set to the UP ARROW. That particular statement makes more sense when you can see the AVCR setup menu in front of you. For more information on the APEXI AVCR install click here 

AVCR Setup here are the learn gear values that the Apexi measured for each gear ( I noted the reading at about 3000rpm in each gear)

1st = 129

2nd = 78

3rd = 52

4th = 38

5th = 30

 HEADS UP  - Exercise caution when adjusting the duty cycle as small increments can make a very big difference to the boost /over boost. I found that 50% duty cycle worked fine to get 17 PSI but when I changed the duty cycle to 55% I got substantial overboost (22PSI). Also note that the duty cycle that works to achieve say 12 PSI of boost will not be the optimum duty cycle that works to achieve say 17 PSI of boost. Hence if you change the target boost level make sure you verify (on a conservative basis) that the duty cycle works for your new target boost level. I would recommend 1-2% changes at a time in the duty cycle value until you get to the desired boost level without spiking.

I also found that if I set a flat duty cycle of say 50% to get 17 PSI then the turbo lag could be fairly significant. I found that two things remedied the situation. I adjusted the feedback speed from 6 to 4 and I adjusted the duty cycle bar chart so that in the lower rpms from say 2000-3500 I had a slightly higher duty cycle. I made the duty cycle progressively lower as RPM increased. The AVCR lets you specify the RPM points at which to change the duty cycle. These RPM points are called the NE points and there are 8 points that you select. I ended up with a table like this for my 17PSI configuration

                            17PSI            15PSI

NE 1    2000 RPM    54%                51%

NE 2    2500 RPM    53%                49%

NE 3    3000 RPM    51%                47%

NE 4    3500 RPM    50%                46%

NE 5    4000 RPM    50%                45%

NE 6    4500 RPM    49%                45%

NE 7    5000 RPM    49%                45%

NE 8    5500 RPM    49%                45% (scramble boost for the 15 PSI setting is +4% duty +0.15 kg/cm2 (to give 17PSI) for 15 Secs)

(Feedback speed 4 for all gears) With NE 1 at 2000 all RPMs below 2000 will use the same duty cycle that is set for 2000 RPM and with NE 8 at 5500 RPM all RPMs above 5500 will use the duty cycle that is set for 5500 RPM. With the table set this way the AVCR will display *** for duty cycle underneath the target boost figure.

Going from a flat duty cycle curve to this simple graduated duty cycle curve made such a big difference in the way my 951 got on boost that I had to adjust the fuel map because the engine was transitioning though load/rpm zones of my MAP system differently. Once your comfortable with the basics of the AVCR its well worth while learning how to set up the duty cycle table/chart properly.


avcrsetup.jpg (99468 bytes)

AVCR Connections for Porsche 944 Turbo

Apexi Manual

TIAL Wastegate Manual

http://members.rennlist.com/tazman/AVCR%20Setup.txt , http://members.rennlist.com/tazman/Apexi%20Setup.txt 


AVCR on the dyno, one thing I have noticed is that when I put my car on the dyno I tend to get a higher level of boost that it would develop on the street. With the manual boost controller that might equate to a couple of extra PSI. With an electronic boost controller that can equate to a big time boost spike. In my experience when the AVCR controls boost it does so very well and consistently. But when something changes like when you use a new duty cycle or put it on the dyno the AVCR can actually fail to control the boost. I tend to use (or start ) with a slightly conservative duty cycle when the car is on the dyno. After confirming that the boost doesn't get beyond the desired maximum I wind in a more aggressive duty cycle if needed. Normally I will have the dyno guy monitor the AFR and call out if there is a problem, whilst I monitor boost and RPM during the run.


icduct.jpg (61694 bytes)

Ye old home made aluminum intercooler ducting.

The stock intercooler is known to have some restrictions that can be improved upon but I haven't gone down that road just yet.

Intercooler ducting  

After seeing my intercooler ducting a friend with a supercharged M Coupe experimented with a very basic duct for his intercooler. Even though the duct was not that good a fit he observed an 8 degree drop in intake temperatures from 118F to 110F. With a better fitting intercooler duct the temperature drop may be a few degrees greater. 8-10F might not sound like a lot but for track work the better cooling  provides a greater margin of safety against detonation and when combined with some other cooling and oiling improvements the degree of reliability should be improved.

Intercooler spraying:

I was contemplating using an Aquamist water/methanol injection system but decided against doing so as I did not want to end up relying on the water injection to run high boost only to have a malfunction in the water injection system put me instantly into severe detonation. I will probably go for a less efficient but also less critical system like an intercooler sprayer to reduce heat soak during track days. Whilst a track intercooler sprayer can afford to be pretty basic I found an Australian company that has developed an intelligent intercooler sprayer that looks affordable and well designed. Here is a link to the intercooler sprayer kit from autospeed.com.au   


injectors.jpg (46521 bytes)

The car has been through a number of fuel management systems. Of course it started life with the Bosch Air Flow Meter system. From there it went briefly to a MAF/hotwire system for a short period of time. The MAF only remained on my car briefly due to two factors. One is that some MAF systems were experiencing drivability problems - this later seemed to be identified as inappropriate positioning of large MAF sensors which resulted in poor sensing across the hotwire. More significantly for me was that a system became available which used a lookup table approach to fuel programming rather than the analogue controls that were common at the time. The lookup table system was provided with the Link Electrosystems MAP (manifold Air Pressure) system, marketed primarily by Guru racing .

So I ended up exploring the issues of integrating a MAP based system into the 944. The install was straight forward, the tuning however was not. I did manage to achieve pretty decent results from my MAP based system but there were always parts of the system that I felt were not quite right. As a driver you end up compensating for the behaviour of the car in certain respects. That gets you by day-to-day but ideally your chosen system should provide you with the means to tune out any specific problem areas.

The problems that I experienced included unstable idle, poor idle mixtures and hesitation on throttle application. I worked on all of these issues and solved them to varying degrees but was left thinking that something was still missing. I believe that the piggy back MAP systems like the Link unit suffer from being forced to work between a chip based fuel map and a Motronic DME that was not designed for the very quick response of the MAP sensor (the DME programming was designed to accommodate the characteristics of the AFM sensor which has significant differences to the MAP sensor). There is nothing inherently wrong with MAP sensors, its just that forcing the MAP sensor to work with a dated ECU is the problem.

Chips. I have used Guru Racing chips in my car and they have worked satisfactorily . The programming of the chips has not been ideal for my car - but I never expected them to be since my car is modified. I do believe they were probably about the best available (non-custom) chips for my car at the time. I used the chips in conjunction with the MAP systems programmable fuel controller. The controller was absolutely essential for me to be able to plug in my own fuel corrections and achieve satisfactory fuel mixtures. 

At this point I decided that the real solution to the problem is to go straight to a stand-alone ECU. The standalone ECU is designed to use a MAP sensor and so will not suffer the same sensor latency issues that piggyback MAP systems experience. There are no "chips" in the stand-alone ECU - the ECU also has its own tables which specify what fuel and timing to use as well as how to compensate those figures for temperature, warmup and a good range of other conditions such as aircon loading etc. This provides me with a transparent set of features to manipulate, a well matched sensor/ECU and no unexpected fudging via the programming in the DME or chips. These are the reasons that people who go to standalone ECUs can get such consistent and predictable performance - good or bad, it just depends on how well you have tuned it. In the 944 community I have heard of several who have gone the stand-alone route and had great success. None of the competent folks that I know of have had any significant problems and most have claimed idle etc, to be as-good-as or better than stock.  

The system that I have chosen to install is the Haltech E6X. I had read about some issues with the early Haltech product line dating back a number of years but could find only positive comments about the newer products. The details on the Haltech installation and tuning are being detailed on a new web page that will be added to my website. In hindsight I would have been far better off to go straight from the stock AFM to a standalone system. The expense seems greater, but in the long run, the time, money and effort involved in playing around with piggy back systems has been far greater and caused a lot of drivability issues along the way. 

FUEL INJECTORS Having seen some of the evidence regarding stock injectors maxing out (duty cycle) at high boost levels I had decided to replace the stock injectors with 75lbs/hr injectors. My injectors are Delphi units that run with a 1 Ohm ballast resistor per injector pair (when used with the Link AFM only).

ECU Update Feb 2006 - Haltech E6X Installed.

I have now replaced the Link AFM / DME / KLR with the Haltech E6X. The setup is nice with laptop control over fuel, timing, various enrichments and configuration data and of course data logging. The install was not particularly problematic and the removal of the DME harness and replacement with the Haltech harness was started one weekend and finished the next with the car being started on the second Sunday. I had installed the trigger and mounted the coils/sensors prior to that. Idle tuning was very straight forwards, now I am awaiting some track time for the next stage of the tuning.

On the first drive I found that the car had a hesitation at about 4000RPM and did not want to rev higher than that. On reviewing the data logs it was apparent that the logs showed that the ECU thought the RPM was spiking to 10000RPM so the ECU was cutting fuel as a consequence. The underlying problem was that the trigger signal to the ECU was getting noisy at 4000RPM. Adjusting the software gain/filter settings did not immediately rectify the noise on the trigger.  The cause of the trigger noise turned out to be RFI from the XXXXX and the solution to that problem was to isolate the xxxxxx power supply by powering it from the battery via a relay rather than directly from the ECU. That eliminated the trigger noise and 4000RPM rev hesitation.


Initial harness layout

Crank Trigger setup

Completed Install (other than some powder coating)

Data log during warm up (The spare AD channel is AFR millivolts - divide  by 100 for actual AFR ratio)


(Background information – Fuel Management)  

My process for tuning the Link AFM/ MAP  

Excel worksheet with data logs and comments

dynoexcel.jpg (80294 bytes) dyno321.jpg (101524 bytes) zones-nov.jpg (194883 bytes) 

!!!! NOTE - The zone worksheet here is with 75 lb Injectors but FQS set for 65lb injectors to offset the fact that I run 2.5 bar of fuel pressure rather than 3 bar as the fuel map in the guru chips was originally designed for.

Dyno results prior to and after sorting out fuel curve. Some Excel data has been overlaid on the initial dyno chart but it is only roughly correlated with the dyno data - ie the boost curve may be off by a 100 rpm or so. I have included my zone worksheet / fuel map, this is the map that produced the 321hp run when under boost but it has been tweaked to get a better (smoother) transition to boost. - 11 Nov 03 These numbers are not likely to yield the same results for anyone else but what they do show is is which part of the chip map is rich and which part is lean when I run the car at 15PSI. I have seen the same trend in another modified 951 with Guru chips & MAP, 3" exhaust,  TO4B and 55lb injectors - also at 15PSI. When I get around to sorting out a fuel map for 18 PSI, the fuel I am removing in cells 940 to 965 will move closer to zero or extra fuel will need to be added.

After doing this fuel tuning the underlying fuel map is close to where it needs to be, I then put the car back into closed loop to maintain a stoich mixture for day to day driving whilst off boost.

Fuel System

I had heard several tales of  woe regarding engine fires in 944’s - unfortunately they are quite common. The two causes are old, dry & cracked fuel lines or fuel rails that leak over the exhaust manifold and old battery wiring with cracked insulation that shorts out against the engine or body. Either of these events usually reduce a 944 to a complete smoldering wreck. I decided to address both problems. I replaced my fuel lines with braided lines initially but I was still not happy with the clamps that still had some potential to leak. The fuel rail itself is also subject to cracking so I decided to replace the complete system with a billet aluminum fuel rail from Lindsey Racing . That system incorporates a series of AN fittings between the rail, dampener and adjustable fuel pressure regulator and connects to the factory steel hard lines via durable compression fittings. End result.....peace of mind on the track...Amen to that.

As the power level of the car has increased I became increasingly concerned about being able to deliver the required quantity of fuel to keep the air-fuel mixtures healthy when operating at high boost and high RPM. The limiting factors become fuel pump delivery volume & reliability and fuel injector duty cycle. To address these  issues I  upgraded the stock fuel pump with a higher capacity Bosch unit and I upgraded the fuel injectors to Delphi 75 lbs/hr units. I also reverted the static fuel pressure back to 2.5Bar which is what the factory pressure is set at (As boost rises the fuel pressure regulator allows the fuel pressure to rise by an amount equal to the boost pressure  so that the pressure differential across the injector remains relatively constant ie its a rising rate pressure regulator. Although it seems a bit backwards, the reason for dropping the fuel pressure is to ensure that delivery volume is not compromised as the fuel pressure rises. At the lower pressure the large fuel injectors are still providing a decent idle - I was concerned that idle quality might be poor but that turned out not to be a problem.

eng-nov.jpg (115228 bytes) 

This picture was taken after the Lindsey Racing billet fuel rail system was installed.

UPDATE - This is one piece of equipment that I definitely have mixed feelings about. I really do like the fact that the original hoses have been replaced with new hoses and the fact that clamped hose connections have been replaced with AN fittings. The AN fittings that come with the fuel rail package are made from aluminium, some are of the opinion that the aluminium fittings may crack where those fittings are directly exposed to engine vibration. Steel fittings are available for those that are concerned about this possibility.

The issues that I have had with the Lindsey Racing (LR) kit include, a drop of 2 PSI in fuel pressure from the regulator that led to a slight lean fuel condition. A failure of the dampener that took in the order of two months for Lindsey Racing to replace because they did not hold any spares and Marren told me I'd have to wait until the next batch were produced. Users of this system have also experienced a very annoying knocking/reverberation in the fuel lines that is transmitted loudly into the cabin. Lindsey Racing produced a replacement line to solve the problem but in my case the knocking has not been eliminated.

I got to a stage of my car being unable to run with the Lindsey fuel rail system installed due to the problems associated with the FPR/dampener. Theoretically it should be possible to swap the LR fuel rail/FPR/dampener for the stock fuel rail components and have the engine run in the same state of tune. In my experience this has not been the case. After removing the LR system from my car it ran very rich with the stock fuel system back in place even though I repeatedly checked that the same fuel pressure was set the same in both systems. At that time a significant amount of dyno time had been invested in tuning the car with the LR fuel rail installed. All that was wasted and I was forced to revert to earlier tuning maps that I had developed with the fuel rail setup I had been using prior to the LR system. 

All-in-all I spent a great deal of time trouble shooting the LR system and checking associated fuel system elements including the DME electronics and wiring and sensors to the fuel pump, its wiring, check valves and hard lines, the FPR, dampener, injectors and their wiring. The whole process was perplexing and I have still not yet found a satisfactory explanation of why the fuel dampener failed or how its failure could lead to the situation where the two different fuel rails require two markedly different sets of tuning at the same pressure. Observations I did make with respect to the FPR are that the tiny threaded vacuum port on the FPR should be sealed carefully or else it can allow pressure to bleed. Be careful not to allow any thread sealant to ingress the FPR though as that could restrict some of the FPR internal orifices and effect its operation.

At this point I have a new dampener and FPR to go along with the LR Kit however the whole kit will remain off the car until I have tuned in the Haltech ECU with a stock fuel rail.


When I first bought the 944 it was not boosting properly and I spent quite an amount of time trying to work out why that was. In the end it turned out that my catalytic converter was largely blocked. The catalytic converter material is only about 1” thick, the rest of the space in the catalytic converter serves as a resonator. Anyway, there was no way to see inside the cat, it was only  a suspicion of mine that it was blocked because I had eliminated so many other potential causes of no boost. So I took to it with a crow bar and was rewarded with clear evidence that it had in fact been blocked. The chunks of honeycomb like material that came out  were largely melted over. There is also some wire mesh that needs to be hooked and pulled out along with the catalytic converter material. For me gutting the cat was the first step to getting any boost so it did improve my boost but I wasn’t in a position to be able compare normal boost with a working cat to boost with a gutted cat. The gutted cat and most test pipes do tend to create a popping sound in the exhaust when backing off the accelerator from WOT. This is the result of a momentary rich fuel mixture igniting in the exhaust. It bothered me at first but has become less of an issue. It may also be occurring less due to the fact that I have improved my fuel mixtures.

I rebuilt the engine, turbo and fuel management system before I made the next change to the exhaust. I had the car running about 16 PSI of boost with a manual boost controller and that was very stable and predictable. Boost would come on and overshoot perhaps 1 PSI while the wastegate responded. With the engine in that configuration I replaced the cat back exhaust section. The MPO (moron previous owner) had actually had a smaller than stock muffler welded into the car. So I replaced the cat back section with a full 3 inch system using a straight-through Borla XR-1 muffler. I had this system custom made at a local performance shop that builds dragsters and was reputed to be good. I took the car out for some WOT runs to see what difference that made. On the way out I was pleased to hear that the exhaust not was not droning or overly loud like I feared it might be, it did in fact sound throaty when I got on the gas a bit though, so that was good. It seems like the turbo also thought the exhaust was good because now when I planted my foot, boost shot straight past 20PSI and into overboost protection! I was simply amazed at how much difference that exhaust made. Going from a 2 ¼” to 3” rear section had opened things up and the turbo was now breathing an awful lot easier. I could not be happier with the 3” cat back & XR-1. A friend of mine forked out over a grand for a fabspeed 3” testpipe and catback and he is happy with its performance but it seems to hang pretty low under his car. When we went for a run in the Texas Hill country his car would bottom out and scrape the exhaust which prevented him from having as much fun. I now have a SFR 3” test pipe in place of the original cat section. The car is a few weeks away from being back on the road but that bigger and smoother section of exhaust should yield a slightly better flow compared to the smaller diameter original piece  and gutted cat. The final thing I did with the exhaust was to have the headers and crossover pipe coated with a aluminum ceramic coating to retain heat energy in the exhaust gases and reduce under-hood temperatures. For the turbo hot side I used regular exhaust wrap, primarily with the intention of keeping under-hood temperatures lower. 

Tips – I found that the clearance for a 3” exhaust was very tight where it passes by the mounting point for the caster block. It was possible to get a tiny amount of clearance by twisting the exhaust a little but that put the flange (the one that mates to the cat back) slightly off center. To avoid potential vibration noises and contact between the body and the exhaust I used a carbide bit to grind back  about 2mm of the lip of the caster block mounting area. Other than that it was very easy to fit the test pipe, the fact that the front section of the test pipe splits from the back half makes life a lot easier to maneuver the flange up to the downpipe.

Ceramic coating benefits

When I originally built the car I wrapped the headers with exhaust wrap and was very pleased with the results. The under-hood temperatures dropped measurably. This time I decided to take it a step further and ceramic coat the headers and crossover pipe which should provide an even greater reduction in temperatures under the hood. I like that because it means less stress on the rubber hoses and insulation on wiring etc. The coating also means that more exhaust energy is maintained in the exhaust gasses and transferred to the turbo. I had Airborne Coatings http://www.airborncoatings.com/  do my coatings, the communication was light but they did a great job, turned the parts around in about 2 weeks (inc shipping both ways) and packed the coated parts exceptionally well. Their coatings are actually a mixture of ceramic and aluminum and the finished product looks like polished aluminum. Prior to having the coatings applied I removed the heat shields from my headers by grinding away every spot weld in the shields . That turned out to be worthwhile though because I discovered some cracking in the union between the No 2 and No 3 pipes, so I had that welded up prior to having the coatings applied. It turns out that my headers were a bad fit to my head and was pre-stressing the headers – that explains why it was always a bit difficult to slide the headers onto the exhaust studs. My local Porsche guy actually completely cut the pipe and mounted the headers onto a spare head and then welded the header pipes back together so that I ended up with an easy fitting stress free header. That became obvious when I assembled the headers and crossover pipe to the head and turbo and everything lined up nicely.

UPDATE - Well now that my 951 has some mileage on it I can say that I am not particularly happy with the coatings that were applied to my headers. They went from the polished look to a frosted look very quickly - no big deal there unless you particularly want shiny headers. However the underhood temperatures are definitely hotter than they were with the factory heat shielding plus DEI wrapping. At this stage the next time the engine comes out the headers will be getting re-wrapped.

Cylinder Head / Block

I made a couple of improvements to the cylinder head. First I polished the intakes and chambers, the exhaust ports were left untouched due to their ceramic coating. Then the  intake manifold runners were also ported and head ports were matched. I planned on running higher than stock boost so I also had the head o-ringed when it  was in for its valve grind and decking. The head sits on a Wide Fire Ring head gasket. Valve train is all new OEM.

I am running first oversize (100.5mm) OEM/Klobenschmidt (sp?) iron coated forged pistons. Apart from the overbore the basic bore/stroke are standard hence my engine is at factory compression ratio of 8.5:1

head1.jpg (30512 bytes) head2.jpg (57948 bytes)

 Engine mods / Lightened parts.

 By cutting down reciprocating mass with the lightened crank shaft and flywheel I am hoping that the engine will rev more quickly as well as loose less power in driving those parts around. If the engine can rev faster then boost should come on sooner. I was a little concerned about going with the lightened crankshaft and flywheel but some research showed that the 944 crank shaft is about 15lbs heaver (at 55Lbs) than typical 4 cylinder crankshafts so a lightened crank is more like a typical crank for any other car. Mine weighed in at 42 lbs after lightening. I also spoke to some folks who were running lightened cranks and lightened flywheels and the verdict from those people was that there was no loss of street drivability. Hence I decided to proceed with a stage III crank and a Fidanza Aluminum flywheel http://www.aluminumflywheels.com/ . The fidanza flywheel only has one grub screw for the timing / reference system.... that is fine for the 951 as the other grub screws that you see in the stock flywheel are not necessary. The DME gets its info from the teeth of the ring gear and from the grub screw in the Fidanza flywheel.  I also saved a bit of weight (3.75lbs) by replacing the original pressure plate with a heavy duty unit  from KEP http://www.kennedyeng.com/. Their aluminum pressure plate has 60% more clamping force than stock and is more of a necessity for my 951 given the extent of its modifications. I am using the KEP pressure plate with the standard friction disk (no small springs).

The stage III crank should also improve reliability of the engine for track work. As many 944 owners know, the 944 engine has a tendency to spin #2 rod bearing, particularly on high mileage engines. The stage III crank has an improved oiling system that should significantly reduce the likelihood of the #2 bearing problem. The crank is also knife edged so that it tends to slice through the oil in the sump rather than bashing the oil which can increase foaming of the oil  and splash it around.

crank.jpg (71121 bytes)

Stage III crank, lightened, knife edged, crossdrilled.

flywheel-pp.jpg (46334 bytes)

Aluminium Fidanza flywheel & KEP pressure plate.

Engine reliability:Accusump:

 Other steps I took to improve engine reliability included installing an accusump oil accumulator system and adding an additional  Mocal external oil cooler. The accusump is a cylinder full of oil with a piston in it and a charge of pressurized air on the other side of the piston. That cylinder of oil is connected to the engine oiling system by a hose and a valve. Once the valve is open the pressurized air will force the oil out of the cylinder through the hose and into the oil galleries of the engine if the pressure in the oil galleries is less than charge air pressure in the accusump. So when the  engine is operating with good oil pressure that pressure will force the oil to stay in the accusump. When the engine suffers a drop in oil pressure the accusump will feed oil into the engine to make up the lost oil pressure.  The accusump also has the benefit of being able to feed oil into the oil galleries prior to startup, ie it pre-oils your engine to provide good lubrication before starting. The benefit there is less wear on engine parts during engine start. 

I ended up mounting the accusump on the cover that goes over the rear seat area of my car. The plumbing is then connected via section of AN10 hose to a bulkhead fitting in the floor of the seat area. I considered running aluminum pipe from the rear to the front of the car but I decided against it as I figured that the aluminum line would be less tolerant of knocks from debris under the car. So I ran AN10 stainless steel braided hose from the rear to the engine bay where the solenoid is mounted. I wanted the solenoid to be mounted close to the engine rather than close to the accusump, that way if the line between them (under the car) was damaged I can shut off the solenoid which would be located before the damage, that would enable me to prevent oil from bleeding out of the engine if the oil line was damaged somehow.

The hose from the solenoid then runs to a 10AN to ½”NPT 90 degree adapter to connect to the oil input adapter plate that sits between the oil filter and its original mounting point. That area is a little crowded with the headers on one side and the power steering lines and water pump on the other side but the aforementioned fittings work nicely and there are no additional problems installing or removing the oil filter. There was still sufficient space to mount the NPT fittings for the oil pressure gauge line. Given the proximity to the headers I made that line a copper line rather than the Nylon line that originally came with the gauge. Its worth noting that the cost of the braided hoses and AN fittings can add up very quickly to $100-$200 depending on where you fit the accusump. I recommend spending some time to think through where you will mount the system and which fittings you will need to route the hoses at the correct angles to clear all obstructions, each bad guess will cost you $$ in buying fittings that do suit the application… ask me how I know.

Oil Cooling:

Given the hot track temperatures in Texas and Australia I thought it would be prudent for a track car to have some additional oil cooling capacity. So I added in a 13 row Mocal oil cooler to perform that cooling.

Tony Morris has looked into 944 oil cooling and did some excellent work to come up with an oil cooling solution for 944’s. His page at  http://944oilcooling.com/ documents some of the significant reductions in oil temperatures that can be achieved with a good auxiliary external oil cooler. There are also some pics on Rennlist that show an installation of the kit that Tony Morris has put together. http://members.rennlist.com/drive944/cooler.html. I already had a bunch of AN fittings so I made my own plumbing to the new oil cooler but the Tony Morris kit looks like a bargain and he has done all the work of figuring out the right lengths of hose etc. My factory hoses were in poor shape so I had my local hydraulic supply co manufacture some new hoses that had metric fittings on one end (Porsche used metric fittings) and AN fittings on the other ends of the hoses. That way I was able to add in the extra lines to the oil cooler using AN fittings that I already had.

RESULTS - I am very happy with the results of the oil cooling setup, the combined oil coolers are dropping the temperature more than 60 degrees. These results I can see every day as I switch between the temp sensor in the sump and the one at the exit of the factory cooler (the last cooler in the loop). In highway driving in Texas that means that the oil heading to the bearings is at 135 degrees F compared to the 195 degree F oil in the sump.

oilcooler.jpg (67104 bytes)

Mocal oil cooler & Accusump

oilcoolermounted.jpg (62309 bytes)


oilinputfinal.jpg (26504 bytes)

Oil Input adapter for Accusump, also shows tap for sensing oil pressure 

oilinputinstalled.jpg (51167 bytes)

vacoz951.jpg (100774 bytes)

This is the vacuum system as it currently exists on my car with various elements removed for the sake of reliability and simplicity.


Given the importance of oil pressure for track use, the car has a mechanical oil pressure gauge to give reliable oil pressure indications independent of electrical variances. The car has an oil adapter plate at the oil filter which allows the Accusump to provide residual oil pressure in the event of an oiling problem, so I tapped that adapter plate so that I could install the gauge oil pressure feed at that point. It is pretty hot in that area of the engine bay so I ran copper tube from the oil pressure port past the headers and into the cabin. Whilst on the subject of gauges I tapped the oil level sender block off plate to install an oil temperature sensor in the sump area. The oil temp and pressure lines feed into the cabin through the DME harness.

oilinput2.jpg (53708 bytes)

Accusump  oil input  adapter with 1/8" tap for oil pressure gauge line. See previous picture for what I ended up with as the final configuration for oil pressure sensing and oil input by the accusump.

oiltemp.jpg (31798 bytes)                    oiltempaux.jpg (73154 bytes)

Oil level sender plate tapped for oil temp sensor #1 and oil temp sensor #2 is inserted into a fitting on the exit from the oil cooler (to measure the temp of the oil that is on its way to the bearings)

Crank vent/oil filler

One of the things that I have not been real happy with is the amount of oil that gets spread through the intake system. That is a consequence of the crank breather system. Manifold vacuum is used to draw air from the crankcase and that air has a quantity of oil mist in it as well as the possibility of some blow by combustion gases. All of that heated air/oil gets mixed in with the intake air with an effect on the mixture/combustibility. I believe it also contributes to level of  smoke in the exhaust. To address those issues I installed a Lindsey Racing crank breather system which employs a catch can. I enlarged the port on top of the air/oil separator and crank fumes vent through that port towards the catch can. The catch can has a drain that allows accumulated oil to be drained away. Since this system does away with any connection between the crank and the intake system my intakes should remain a bit cleaner and hopefully I will end up with better combustion. The catch can has a drain with tap that is located near the passenger side caster block.

RESULTS - Lindsey crank breather – I found that the top of the filter just managed to get close enough to my hood to give an occasional vibration under load. This may not have happened if I had foam under the hood. To fix that vibration I trimmed away 1/8th of the rubber that slips over the neck for the catch can. That let the filter sit 1/8th lower and there was no interference with the hood.

firenozzle.jpg (84968 bytes)  

Crank case ventilation system / catch can



This is a problem that doesn't get thought about too much until you experience a few perforated hoses or other wear issues. If you give it due consideration ahead of time you can certainly save yourself a lot of grief. 

Braided hoses- These hoses are tough, no doubt about it. Unfortunately they can be so tough that they can lead to other problems. I had one case of a braided hose that was rubbing against the aluminum bar at the front of the engine bay that also serves as a coolant passage for part of its length. The braided hose from the accusump filed away a section of the aluminium tube as the engine rotated slightly under load.

Other problems I noted due to engine rotation include:

A groove (would have been a hole if not detected) being worn into the turbo compressor housing by the housing coming into contact with the turbo coolant pump bracket.

Exhaust vibrations at the wastegate, downpipe adjacent the driver side caster block and where the rear sway bar crosses over the exhaust. These problems arose when the exhaust was increased to 3" diameter.


The 944 Turbo  has one of the hotter engine bays that I have have seen. Consequently rubber and plastic parts tend to deteriorate or harden relatively quickly. Its a definite problem for electrical wiring and is one of the main causes of electrical shorts in the unfused battery/alternator wiring. I found that the addition of fiberglass sheathing prevented the complete electrical shortage that would have been terminal for my 951 when it had its engine fire. Since the fire I have applied the same sheathing to key battery wiring, fuel lines and hydraulic lines that are subject to adverse heat or abrasion.


After putting my car on the track for my first DE I realised the value in having good stopping power. I was fortunate enough to ride with Jason Burkett in his 944Turbo for 10 laps in a DE and experience just what good braking is all about. Up until that point n time my efforts had been directed at HP improvements but thanks to that experience with Jason I started to look at addressing more of the handling aspects of my car. At that point my car was fitted with 300lbs springs & Paragon adjustable height perches, 29mm torsion bars and Koni shocks. So the suspension was reasonably firm and balanced but the braking power was not so well matched for the Hp and speed that my 951 could develop. My track experience had left me with a spongy pedal due to the heating that the standard 87 Turbo brakes underwent. I investigated various options and decided to replace the worn front rotors with  993TT brakes (Big Reds) from Paragon Products as they offered the best value improvement. Jason also recommended changing the bias valve to 5/33 from 5/18 to increase the rear brake bias and overall stopping power of the car. The rear rotors were worn as well and so were replaced with slotted version of the stock rotor. I left the rear calipers stock since the rear brakes seem to be under-utilised as it is.

The big reds give me front rotors that are 322mm in diameter and 32mm thick, these rotors actually have cast in ventilation holes rather than drilled holes. The Brembo calipers use 4 pistons of 36 & 44 mm diameter. The rear brakes are standard turbo dimensions with 299mm rotors, 24mm thick and 4 piston Brembo calipers at 20mm & 30mm.

The 993 brakes required the rotor dust shield to be modified by trimming away the outer 10mm or so edge (the raised portion on the dust shield). That was accomplished with Dremel cutting disks. The brake calipers themselves required an adapter block because the 993 brakes have mounting holes which are at 90 degrees to the 944 mounting holes. The brakes came with the proper mounting block and hardware to ensure a good solid mount. One point to be aware of is that the adapter blocks are drilled in such a way that it is possible to mount the caliper incorrectly - I did that and it was not obvious until I put the pads in and saw that they protruded about 3mm past the outer edge of the rotor. The problem was solved by flipping the mounting block 180 degrees. Jason also recommended using the short hard lines to go between the caliper and the flex hose. I ended up swapping my cheap braided lines for the ones that came from Paragon because the Paragon product had end connectors that rotate thus enabling the brake lines to be tightened up without having to rotate the entire caliper (impossible if caliper is bolted in place).

Tom Pultz has put together an excellent web page on 944 brake upgrades, check it out here 

7inch clearance.jpg (51030 bytes)

clearance with 7ET55 wheels

9inch clearance.jpg (49383 bytes)

clearance with 9ET55 wheels

front caliper.jpg (98475 bytes)

Factory 993 rotor & Caliper

rear caliper.jpg (56004 bytes)

Slotted 944T OE caliper

For street use I use Pagid street pads on the front (they squeal) and Metal Masters on the rear. For track use I use Pagid RS14 (Pagid Blacks) front and rear. To bed in the brakes I used the procedure from StopTech's web site, I read the procedure from the installation manual for the BMW E46 M3 but the procedure is generic.

Brake Cooling:

I also switched from the stock cooling ducts to 968 cooling ducts which scoop air from around the control arm and direct it to the rotor. This modification is hopefully an improvement in cooling as well as freeing up some space in the bumper area so that I can direct some of the airflow at the front towards the auxiliary air cooler mounted behind the fog light.  

The Part numbers for the 968 Ducts are 944 341 555 01, 556 01, 051 01 and 052 01.  

968 Brake cooling ducts – . The ducts cost around $70 from the dealership. Unless you have 968 M030 control arms you will need to purchase some brackets separately from Frozen rotors ($50) or make some brackets yourself. If you make brackets yourself I would recommend not drilling into the control arms due to the potential for introducing weak spots that could be subject to failure. The upper parts of the ducts bolted straight up to the same point that mounts the deflector plate that comes with the original 951 brake cooling setup.

I use Pagid black pads (RS-14) for track work and they are very effective but they also squeal a lot even when warm. Anti squeal compound  doesn't help either.  I also use the Pagid S (SRS-14) street & light track pads, even they squeal at times despite having anti squeal shims already attached. "Light track" is key here, they don't hold up to any decent track work.

Porsche 968 Brake cooling scoops

I have used Valvoline high temperature brake fluid in the past for a weekend of track work but have found that the pedal was spongy by the end of a DE weekend. On the other hand after a full weekend of DE work with ATE fluid the pedal was still very firm. I am quite happy with the ATE and will stick with it. I also like the fact that the ATE comes in two colors so it is a lot easier to complete a brake bleed without wasting a lot of fluid. In fact it was not until I used ATE blue for he first time after clear Valvoline that I realized just how little fluid there actually is in the system and how quickly the new color fluid gets from the reservoir to the wheel cylinders during a bleed. Previously when I was bleeding the brakes and using the same color fluid as was already in the system I had to let a lot of fluid come through the caliper to be sure that it was fresh fluid coming out, I wasted a lot, now I alternate colors and have very little wastage. I use a hand pump power bleeder to bleed the brakes, I cant recommend those things highly enough. They are simple, reliable and don’t end up causing fluid leaks at the cap like some home made or lesser quality products. They really make bleeding the brakes painless. On the other hand I wouldn’t touch the mighty vac brake bleeder with a 10 ft pole, I would be more inclined to beat a mighty vac with a 10ft pole though. I have one and it gave me no end of grief and highly variable results and was just a complete PITA. I do use the vac pump for pressure testing other things though.


 Porsche Battery wiring is dangerous as it is so there was no way I was going to buy a used one of those harnesses. I purchased one of Iceshark’s battery and alternator harnesses with supplementary earth and positive leads. One look at those things and you know you are looking at good stuff. Iceshark makes it as good as it gets. His kit also made it a lot easier and tidier to connect in a battery kill switch. After the fire I realized just how important a kill switch is so I decided to install the type that will isolate the alternator as well as the battery because if you just kill the battery the alternator will keep the car running . Its also important that the switch disconnects the alternator before the battery if you want to avoid frying the alternator. Anyway the Icesharks kit allowed a tidy connection on the main battery wires to the kill switch and the supplemental positive provides a very good location to connect the ancillary positive connections which would have been a bit of a mess if they had all been connected to the post of the kill switch. The kill switch is installed adjacent to the battery in the area where the vacuum canister used to sit and it is operated via a pull cable that goes back to a  bracket which I manufactured and mounted on the side of the console near the drivers right knee. The pull cable for the kill switch runs along the back of the firewall on the fresh air unit side and feeds in through the same hole that I drilled adjacent to the fuse box for the supplementary positive cable. Routing of that pull cable takes some thought to ensure that it does not interfere with any of the pedals etc. Speaking of pull cables, that one was a custom order from a supplier of aircraft type pull cables (ie Iceshark grade) in Houston but still only cost $28 delivered. It has a similar nice T handle to the one for the Halon extinguisher.

UPDATE - WIRING OCT 04   well if you look at the adjacent picture you can see why its so common to have electrical problems in 944's. This is the important pair of connectors for the timing and reference sensors. My sensors are new but the wiring loom is as old as the car and the rubber boot is like hard plastic. I discovered this problem when the car decided not to start after I had the manifold off. The wiring had been disturbed and the insulation around the wires basically had flaked away and a single strand of wire was shorting two of the connectors after I had moved the wiring out of the way - End result No Start..... well not for long anyway.

timingandreference.jpg (134126 bytes)

cracked insulation exposing all 3 wires to the timing / reference sensor



 When my car was last running at the track I found that I had a bit of a stumble at around 5500 RPM which I thought might be ignition related. I replaced the wires, cap and rotor but that did not solve the problem. Since then I have replaced the Magnecor wires again because one of the wires pulled apart after removing the wire from the plug only one time, that and I was not real happy with the design of the boot that fits over the plug. I also discovered that the factory wires are zero resistance and that the Beru connectors provide the necessary resistance. Non Beru type aftermarket wires are normally resistive and so the different length wires will have different resistances and that may cause variances in spark seen at each cylinder. So this time round I bought 8mm Beru based wires from Lindsey Racing. I also swapped the old coil for a higher output MSD blaster coil (no ballast resistor). For plugs I am using NGK type R gapped to 0.7mm, these are a traditional looking plug.  

UPDATE - When the Haltech E6X goes in the ignition system will be replaced by a pair of direct fire coils operating in wasted spark mode. The coil pack I have is the coil pack from the TEC 3, the Ignition module is the Bosch unit provided by Haltech. Plug wires to suit the DFU coils haven't been purchased as yet.


Full Article on replacing Torsion Bars (RIGHT CLICK and SAVE AS)

I replaced my stock Torsion bars with 29 mm sway away torsion bars from Paragon Products, this gave the car a much more balanced feel under acceleration, previously the nose always felt like it was lifting significantly when I got on the gas. the lifting phenomenon has been eliminated and the ride has stiffened up all round. The 29 mm torsion bars are a good match for the 300lb front springs. I was expecting the ride to be fairly harsh as a result but it is not as bad as I expected. On a 4 hour drive along the highway the ride is fine, but there is one very potholed/patched, rutted section of road near my place where the ride does feel bad. It is the only piece of road that I have had the car on where the ride is bad but if all your roads were like that then you probably wouldn't like 300lbs springs and 29mm torsion bars. For the track the spring rates seem to be pretty reasonable. I am sure that a little stiffer would make things even better but I am very happy with my slight compromise from a full track setup. My front springs are made by Hypercoil and purchased from Paragon along with adjustable perches. At present I still have the stock spring bases attached to the Koni struts. This was fine until I put camber plates on the car but now that the camber plates are on, the ride height is higher than I would like even though the adjustable perches are on the lowest setting. What this means is that I will have to pull the struts out again and cut off the stock spring perches, that will allow the ride height collars to sit about an inch lower and subsequently the front of the car will sit an inch lower as well. Trimming off the stock spring bases will also allow the fitment of 9" rims up front (55mm offset) so that I can put 255s up there.

 The tracking of my 944 was not particularly good from the outset. So one by one I replaced many of the wear items including control arms/ball joints, tie rods (both ends), caster blocks (with 968 versions) and I resealed the power steering rack. The sway bar bushings were replaced early on, all these steps have eliminated any excess play in the suspension. The stock upper strut mounts were replaced with Paragon Camber plates, this became somewhat necessary when I found that lowering the car resulted in a situation where I could not return the camber to factory spec. The camber plates allow me to set the camber to close to factory spec and still quickly set a more aggressive camber for track days.  Front and rear shocks are single adjustable Koni yellows, these are good for the street but may turn out to lack some adjustability for the track.

I eventually replaced the stock turbo front sway bar with a Weltmeister unit. The Weltmeister bar is substantially heavier than stock and has adjustable drop links on the ends. When the car is off the ground there is some interference between the drop links and the flexible brake hose when the wheel is turned fully outwards. On the ground the geometry changes and the issue is less of a problem. The Weltmeister bar also does away with the rubber bushings where the drop links connect to the control arm. Instead of rubber the drop links have aluminum bushings to connect the drop link to the control arm.  

M030 Rear 19mm Sway Bar – When I went to install the bar I thought I had the wrong bushings because when the bushings are fitted to the bar there is a good gap between the end of the bushing and the metal plate that clamps over the bushing also looked like it would be a struggle to fit. Well it turns out they are the correct bushings (964 part #) and the gap is normal. I used a 5mm longer bolt on each side so that I could get the bolts started – that made life easier.


torsionhousingdropped.jpg (68222 bytes)

Torsion bar replacement


newtorsionbars.jpg (53759 bytes)

29mm Sway Away Torsion bars


newsuspensionsteering.jpg (65354 bytes)

Replacement control arms, tie rods and the steering rack undergoing reseal.

camberplates.jpg (75136 bytes)

Adjustable Camber plates

coilovers.jpg (52056 bytes)

Adjustable height coilovers and 300lbs 2.5"  hypercoils


castor block.jpg (57934 bytes)

Upgraded castor blocks & Weltmeister control arm bushings

120-2061_IMG.JPG (74306 bytes)

 Oct 03

swaybar3.jpg (157403 bytes)

Factory Turbo Sway bar

 SEP 2002


weltmeister.jpg (54342 bytes)

Weltmeister adjustable sway bar - Oct 03

Brey Krause strut brace 

strutbrace.jpg (100641 bytes)

Corner balance

The car has been corner balanced in its present street/track configuration and weighed in as follows (Full tank of gas, 6x9 speakers &  trim still fitted) figures in yellow and with me in the car at 162 lbs figures in blue. Believe it or not we actually didn't make any adjustments - these figures were what we obtained as soon as we put the car on the scales. I was pretty happy considering I set the suspension up by eye and measuring tape (got to get lucky sooner or later).

                            FL 744      (1534lbs 51.2%)      765 FR                         FL 814     1588lbs 50.4%         774 FR

                                                   3007lbs                                                                   3148lbs

                            RL 740     (1468lbs 48.8%)       728 RR                        RL 808     1560lbs 49.6%         752 RR

Suspension comments / update

When I first put the car back on the track I took the first couple of events pretty easy. I wanted to see how the engine and drivetrain were holding up. I also had to get myself into the swing of things and bring my driving up to speed. As time went by the car held together and I started to drive a bit harder. As those things happened I was able to experience various handling issues with the car. It gave me an opportunity to adjust front and rear sway bars, tire pressures, and camber. I have limited adjustability on my shocks. At the same time the seasons were changing and the track conditions along with them.

I would say that I have only just begun to push the car hard enough for some of the suspension changes to make a significant difference to my lap times. At my last event in Texas I had an interesting and expensive lesson. I had stiffened up my front sway bar to about the half way point. The Weltmeister bar that I use is a big one so that equated to quite stiff. I was running a brand new set of front victoracers for this event. The previous event I had been to had been cold and wet and I had been running a higher tire pressure. The current event was dry and mild conditions though and the tires warmed up a lot more quickly. I was also put into the fastest run group and the instructors & club racers got combined into the same group for various reasons - so everything was moving along a bit faster than usual - me included. I noticed that the track felt a bit greasy is spots and other drivers had made similar comments. Unfortunately for me I was missing the signs that my car was starting to understeer due to the way I had my suspension setup rather than the track condition. As the weekend progressed the understeer became more and more apparent as I pushed the car harder. I realised that my tire pressures were too high to start with and they were overheating in the warmer conditions and harder driving. I started backing the pressures down quite a bit but was still getting a significant understeer. Towards the end of the weekend I found that I had taken both the front tires down to the cords on the inside. I was only running -2.2 and -2.5 camber on the front tires - that's hardly a lot. But it seems I had my sway bar so stiff at 50% that it was not allowing the tires to roll onto the full contact patch when cornering. That meant the inside of the tire was trying to do all the work (and failing - hence the understeer) and saw accelerated wear as a consequence. My fronts are only 225s which is somewhat narrow for my car but the poor contact patch resulted in those tires being trashed in a single weekend. Annoyingly - I had a tyre pyrometer with me that weekend but was not using it to check my cross tire temperatures.

Tire Pressures: My experience is that the victoracers start going off around 38PSI hot. To that end I set the cold pressures around 32-33 PSI. I will experiment with this a bit more once I have tuned the suspension a bit more and I am sure that I have a good contact patch.

Straight entry - this is where I think I can make the single biggest improvement to my times. Presently my 951 scrapes the edge of the rear tire as it hits or leaves the banking on the main straight at TWS. The harder I hit the bank the worse the rub is. Unfortunately by not entering the straight as fast as possible I am cutting down my average speed the whole way down that very long straight. 275s fit under the rear guard just fine on a wheel with the correct offset so I need to go down that route and upgrade the fronts to fit 255's to the front at the same time. Alternatively I will have to roll the rear guards to fit the 275's in. I also still have to manage the throttle coming out of the esses in 3rd or the 275's will still let go.

UPDATE - WELMEISTER POLY BUSHINGS - Oct 04  I had heard various opinions on whether Weltmeister RED poly bushings are suitable for track use but I installed a set anyway since one of the rubber bushings that came in my refurbished control arms was looking pretty sorry after a short period of use. Anyway after 11 track days on the Welts they still look fine but I noticed a detectable play in the suspension - in so far as I can grip the top and bottom of the rotor and rock the suspension slightly. It took a while to isolate what was moving but eventually it could be seen that the control arm itself was moving at the front bushing. I removed the bushing and discovered that the wear was actually on the inside near where the two halves of the poly bushing face each other. It doesn't look like much but it can certainly be felt. I suspect that my time at Mallala with its many tight corners has been the main reason that the bushing has worn so quickly. Incidentally its all on the drivers side (drivers side bears most of the load in the mainly right hand turns at Mallala), no movement can be felt in the passenger side suspension. I only found this today so now I guess I am going to have to find some Delrin bushings or fork out for some Racers Edge spherical bearings for the fronts of my control arms.

welt_bushing.jpg (35631 bytes)

Play between PU bushing and steel insert


I have two sets of factory Turbo twists ET55 17x7 and ET55 17x9. I run an old set of Bridgestone RE-71 on the street set, these are 215 up front and 245 at the rear. These are not bad tires but the 245's are traction limited in a straight line when the boost hits in second gear. I would like to replace them with Bridgestone Potenza SO-3's 225 and 265 for increased braking traction up front and increased rear traction under boost. On the track wheels I have been running  Khumo Victoracers with 225's up front and 275's in the rear. 

I have to run 17 inch wheels on my 951 to clear the big reds. I also trimmed back the stock strut spring perch so that I could fit a 9" wide wheel up front. See the pic below, the wheel does fit inside the fender, I have not yet road tested that configuration though. When I can pick up an extra set of 9" wide rims I will put wider track tires up front.

UPDATE - as per the suspension comments its now time to go with wider tires up front, 255s will fit without any steering obstruction and a better rear offset to pull the 275's in a few mm. Most probable  option here will be to go with 17" CCWs (Complete Custom Wheels). Wheel dimensions: 

Front ET 65 17 x 9.5 to mount 255/40/17 and  Rear ET 60 17 x 10.5 to mount 275/40/17

turbotwists.jpg (59411 bytes)

17x7 ET55 Factory turbo twists


 wheels.jpg (109427 bytes)


wheels9inchfront.jpg (38065 bytes)

Front mounted 17x9 ET55 Factory turbo twist

Audio system

Note - most of the audio gear described here gear is no longer fitted due to weight and fitment considerations such as the door speakers interfering with the roll cage once the roll cage was fitted.  I have left the section in the web page for reference for anyone that might want to use the seat area as a sub enclosure.

At first I wanted to put a big stereo in the car because the car was to be primarily a street car. As soon as I started to track the car though it became obvious that there was a definite weight penalty to be had and it was also difficult to fit a lot of gear in the car when that space was taken up with audio gear. The gear that I did fit is listed in the table below.

Head Unit

Kenwood KDC MPV7019 MP3 
Sub Amp Kenwood KAC 749S 250W per channel
Main Amp Kenwood KAC 749s 250W per channel


1 Farad Digital
Rear Infinity Kappa series 692.5i 6x9 2 way
Front Infinity Reference series 5.25 with remote tweeter
Subs Infinity Reference series 1220W 12 inch subs

I built a custom sealed enclosure in the rear seat space for the subs. That worked out at an excellent volume for the Infinity subs that I was using. Quite a bit of shaping had to be done to create panels that would seal of two chambers for the subs. Expanding foam and silicone sealer were used to create air tight working surfaces that the other panels could attach to. The top panel of the enclosure was constructed from 16mm thick MDF and sat on a weather strip soft rubber seal to provide an airtight seal between it and the rest of the enclosure. The amps sat on a separate raised board so that wiring could be routed underneath and to provide some isolation from the vibration of the subs. The rears were mounted in standard 6x9 sealed MDF enclosures. Those enclosures are small but they made a huge difference to the sound from those rear 6x9's. I constructed a cover board to fix the rear 6x9 enclosures in place and to provide a barrier between the cargo area and the speakers. For the front I had to manufacture custom enclosures as I was not happy with the dismal space available in the doors. The tricky thing was to have an enclosure that would still clear the lower door rail as well as the dash when the door is closed. After much trial and error this was successful and aesthetically pleasing.

subside.jpg (44989 bytes)

Expanding foam to seal edge and provide solid mounting point for edge.

subinsides.jpg (53858 bytes)

Foam packing was used to establish the volume of the enclosure.

subspadding.jpg (33212 bytes)

Fibre fill in Sub box

speakers.jpg (48226 bytes)

Amps can be seen mounted on a raised panel which provides isolation from speakers

rearspeakercover.jpg (50981 bytes)

Hatch area showing custom cover for rear speaker boxes

front speakers.jpg (49308 bytes)

Custom sealed MDF speaker enclosure and map pocket

Body / Aerodynamics - Splitter

When I bought this car it was missing the cover that goes underneath the front bumper cover, I looked around for a replacement for ages but could not find one so I installed a GT racing splitter instead. These units are delivered unpainted and come with no mounting holes. It is necessary to mark the position of the nuts along the bottom of the bumper cover using some tape. Then jack the splitter up into place and mark the where each hole needs to be (next to the pieces of tape) The splitter can then be lowered. Two sets of holes have to be drilled. A large one along the bottom edge of the splitter so that a flat washer can fit through and a smaller hole in the other edge of the splitter so that the screw/bolt can poke through and connect to the nut which is attached to the bumper cover. Since the splitter is made of fiberglass the flat washers are needed to spread the load of the fixing screws.  After installation I decided to trim away some excess splitter that protruded into the wheel arch area. A dremel tool works well for this and other parts of the splitter installation

splitter-under.jpg (41828 bytes)

The large holes allow the flat washers to pass through the bottom edge of the splitter.

splitterend.jpg (32427 bytes)

Excess splitter needs to be trimmed away.


Initially I had not planned on taking the car to the track but lifes events changed all that and I subsequently decided to forego having a show car in preference to having a car that could be driven, fully. As I take the car to more DE's I am trying to gain more driving experience and improve the 944's setup to suit the needs of track driving. 


seatbrackets.jpg (90892 bytes)

I fitted a Sparco Evo to the drivers side and a Sparco Evo 2 to the passenger side, the Evo 2 is 4cm wider than the original Evo. I had originally purchased a pair of EVO2’s but my butt was definitely too narrow for the EVO2 and defeated the purpose of having a race seat. The EVO 2 brackets mated right up to the seat. The EVO 1 seat was 4cm too narrow though. After confirming that I had the correct brackets I had to fabricate spacers between the seat mounting brackets and the sides of the seats. I used 19mm MDF 2 inches high by the length of the bracket. That provides the required contact area between the seat and the mounting points. The seat brackets were mounted on Sparco sliders as well. I had to drill one hole in 1 slider on the drivers side and 1 hole in 1 slider on the passenger side as well. There is some movement in the shoulder area of the seat so I may still attach a seat back brace from the seat to the roll cage.

Roll cage:

I managed to pick up a used autopower roll cage, it is a bolt in unit that anchors on the rear wheel arches in the rear seat footwell and up by the kick panels. The cage only came with one harness bar so I had another one welded in for the passenger side since some DE events require passenger harnesses if the driver has a harness. The roll cage fit without too much drama but does necessitate removal of  the sun visors. It also prevented me from being able to use a gauge pod so I had to install my gauges in individual gauge cups that bolt to the roll cage. I had previously built some custom speaker enclosures in the doors and they too had to go when the roll cage was fitted due to interference. Depending on your application the biggest hassle might be the fact that the roll cage also prevents the emergency brake from being applied. The side intrusion bars go straight over the top of the E-brake handle so that you cannot raise it.


I installed RCI 5 point harnesses. The shoulder straps fix to the harness bar of my roll-cage. I had to drill 12mm holes in the transmission tunnel and through the floor for the sub belt. The other side strap I connected to the original seat belt mounting point. Note that seat belt mounting bolt is a 7/16 fine thread, not a metric bolt. The sub belt bolts were 55mm on drivers side and 70mm M12  on the passenger side. The bolts that go through the transmission tunnel need to be about  45mm M12 depending on how many washers you use. Points to consider with the harnesses are that the buckles need to be able to clear the appropriate holes in the seat so that the buckles dont catch against the seat and interfere with the proper load bearing of the harness. Next time around I would use pull down up lap harnesses in preference to pull down versions because there is not much space to pull down.

Halon Extinguisher: 

Having been through a fire in a 944 I now appreciate just how quickly these things can burn. Even at big track like TWS it can take time for a fire marshal to get to your car and every second makes difference. I decided to accept the weight penalty and install a 10lb halon in-car fire extinguisher system. The unit I purchased is a Safecraft LT10 system. I mounted this system in the rear seat area and it is activated by a pull cable near the drivers knee. The unit discharges in up to 3 locations ie engine bay, cabin and fuel tank area

After mounting the 10lb extinguisher bottle in the rear seat space I had to plumb the nozzles to their respective locations in the car. I ran one line through the floor in the seat area and it runs along the top of the rear sway bar where it goes to a T. One line runs from the T straight up to discharge in the fuel tank area. The other line from the t goes across to the fuel line area and follows them to the engine bay. That line then runs along the frame rail and comes up the strut tower and the discharge nozzle itself faces the fuel rail. The other outlet from the extinguisher will be plumbed to discharge in the cabin area.

firenozzle.jpg (84968 bytes)

Fire nozzle - it has three ports and is clamped to the strut tower.


rollcage.jpg (74329 bytes)

Autopower roll cage

disconnect.jpg (60620 bytes) batterykill.jpg (87913 bytes)

Electrical disconnect switch in place of vacuum canister

fireext.jpg (75531 bytes)

Safecraft 10lb Halon fire extinguisher


Gear Knob / gauges / interior

 I replaced my stock gear knob with a Momo unit. That knob slides straight onto the stock shaft but it requires some shims to produce a tight fit. I also put one of the provided rubber caps and a little silicone on the top of the shift lever and trimmed the cap slightly so that the knob could still slide on there. I made the shims from some masonite about 30mm long, 12mm wide and 3-5mm thick. The masonite is easily sanded by hand until they are a good fit between the shaft and the inside of the knob cavity. Prior to sliding the knob on I also put a little silicone on the backs of the shims where they contact the shaft. The base of the knob has 3 tapped holes where grub screws fit. One of those lined up with the front of the shaft so I filed a small notch in the shaft and the grub screw mates up with that notch thereby preventing the handle lifting off the shaft. The other grub screws bite into the masonite shims and prevent the knob from being able to twist. I purchased a Momo shift boot with the knob and the shift boot has a hard piece at the top which screws onto the base of the knob.

I wanted my oil pressure and temperature gauges close to my line of sight whilst driving on the track, my roll cage prevents the firment of a gauge pod to the A-pillar so I had to fashion a bracket to attach to the roll cage. It interrupts the flow of air from the vent but hey..... its just another one of those compromises.

interiorbefore1.jpg (70282 bytes)    BEFORE     interiorbefore2.jpg (75584 bytes)

interior.jpg (102018 bytes)              interiorafter2.jpg (69132 bytes)   AFTER   interiorafter1.jpg (68445 bytes)                interiorafter3.jpg (88796 bytes)


Copyright © 2003 , — All Rights Reserved
Porsche® and the Porsche Crest® are Registered Trademarks of Dr. Ing. h.c. F. Porsche AG
All other trademarks used within this site are property of their respective owners