Carrera 3.2 Performance mystery and Chips (long)

[Steve W]

With all this talk about 3.2 performance and chips, I often stay away from these types of posts, but I’d thought I chime in here before too much misinformation gets posted.

On the 84-89 programs, there are a large multitude of maps, with over 400 data points. Talk about ignition timing at 1000, 2000, 3000 rpm whatever is so meaningless; it’s like staring at the branch but missing the forest. Porsche provided a small multitude of chips for the U.S. 930/25 Carreras. Off the top of my head, there are the 1267355102, 1267355028, 1267355180 for the 84-86 Carreras, which are absolute dogs in throttle response and overall performance because of the excessive lean mixtures and conservative ignition maps for emissions compliance and expected worst of conditions for use with 91 RON, which is equivalent to U.S. 86 CLC octane. All the above chips are basically identical, except that later on Porsche started using 4k vs 2k eproms as memory prices dropped. The 84-86 chips are programmed to idle at 800 rpm with the idle ign timing set at 2.5 degrees ATDC. For the 87s, with the 4k eproms, the two most common are the 1267355236 and 1236355302 chips, with the 236 chip also being an absolute dog, but the 302 chip providing good throttle response. These newer programs increased idle from 800 to 880, with the 302 chip having it’s idle ignition timing set at 3.5 degress BTDC, and the 236 chip at the old 2.5 degrees ATDC. For the 88-89 Carreras, there also were several chips, the two most common being the 1267355357 and 1237355358, with the 357 chip being the absolute dog, and the 358 chip being the best and identical in every respect to the 302 chip with the exception that the 8K 28 pin chips had double memory of the 24 pin 4K chips to contain the 8051 processing code onboard, instead of a separate onboard chip. The 358 and 302 chips were factory spec for use with 95 RON octane, equal to 90 U.S. CLC octane. The 358 and 302 chips also had the greatest ignition advance at full throttle, with 23 degrees advance at 6000 rpm, vs about 17 degrees for the 357 chip, and about 20 degress for the 84-86 chips. There are also the 28 pin 1267355330, and 1267355910(32K), but these are less common so we won’t discuss them. So if you are a club racer and need to stay stock, go with the 302, 358, 330, or 910. If you have a 2K 84-86 box, it is a simple modification to move a jumper on the circuit board to read the full 4k chip of the 87 cars. However, many 84-86 DMEs are already 4K capable.

With regards to pockets of performance left on the table, Porsche did not leave them on the table because they were unaware, but because of having to perform a balancing act with respect to meeting emissions, low fuel octane requirements, potential poor maintenance by owners, intake air leaks, clogged fuel injectors, carboned intake valves and piston tops, clogged cats and contaminated and non functioning O2 sensors, etc., and still function to meet emissions requirements and still operate without major issue. But for a truly optimally performing engine, this leaves many compromises that I am sure Porsche was very well aware of.

All one has to do is log the air/fuel ratios using a digital wideband data logger while you are driving and see the differences with the different chips and why they behave the way that they do. In addition, the part throttle ignition maps are much more optimized in the 302 and 358 chips over the others, which also contribute to their much improved low part throttle response. Almost all the stock chips will try to maintain the stoichiometric ratio of 14.7:1 under part throttle conditions, especially with the O2 sensor providing closed loop feedback. This makes for good emissions, but especially on the older chips, with their very retarded part throttle ignition maps and super lean fuel curves, creates a lean bog on throttle acceleration.

[Steve W]

continued:

By now almost everyone knows that what is good for emissions is not good for maximum power. A stoichiometric 14.7:1 afr provides complete combustion for best emissions, but on the aircooled 911 motors, maximum power occurs at a richer afr of 12.6 to 13.0. And as every turbo tuner know, a richer mixture provides a much cooler combustion chamber which allow increased boost without predetonation, often running as low at 12.0:1, however in our case, with the normally aspirated motors, the richer mixture allows a more optimized ignition timing without(edited) predetonation. It is a misconception that the O2 sensor always keeps the part throttle mixture at stoichiometric, no matter what you do to the part throttle fuel mappings. The O2 sensor only has a certain margin that is modulates under, and also has a certain amount of latency, so when fuel enrichment is provided for that momentary burst on throttle acceleration, the proper richer mixture can be delivered for that moment to eliminate a bog, and provide full power, no matter where in the power curve that you are in. One can also remap the part throttle curves which forces the afr to go to an optimum power mix at high part throttle load, no matter what the O2 sensor says.

The ideal situation would be maintain a stoichiometric mixture under cruise and light loads for best fuel efficiency and emissions, yet the moment you jab your throttle, either by just cracking it a half inch, or down to the floor, the air fuel mixture instantly goes to 12.6-13, and stays there until you are done and have reached your attained speed or torque level, at which point, your afrs drop right back to 14.7 at steady state. At full throttle, you want to maintain approximately a 13.0:1 afr for optimum power for the 911 motor, but above 4000 rpm, it is probably desirable to go a little richer down to 12.6:1 to prevent predetonation due to uneven air distribution, especially cylinder #5 which gets the most air from the Carrera intake manifolds. We may loose a little power on the other 5 cylinders, but better safe than sorry.

Doing all this may seem like a daunting task, but is really not that difficult and quickly tuned with the right equipment. Using a digital wideband afr analyzer, one can instantly monitor the afrs from cruise, to part throttle acceleration, to full throttle, and with data logging, I can bring the data back to my computer and analyze the results at every rpm range, and retune. In my case, using a laptop interfaced into the Motronic box emulating the chip, I can perform realtme adjustments to the fuel and ignition curves on the fly, while monitoring the afrs on the display in realtime. Too much fuel at 4000 rpm full throttle … just dial back the full throttle curve at 4000. See a momentary lean afr at 2000 rpm when you crack the throttle, leading to an acceleration bog? … Just recurve the part throttle map at 2000, and get rid of it. Cruising along the highway and seeing a perfect 14.7:1 afr – beautiful. When all is said an done, and the ignition curves are reoptimized to work with the remapped fuel curves with the octane fuel you intend to use, the cars performs amazingly, very smooth, and so much more drivable over the stock programming, and with reliability.

There has been talk of playing with the sensors, i.e. disconnecting the O2 sensor, readjusting the spring tension in the AFM, etc. – been there, done that – which to me is Mickey Mouse and a compromise at best with the ability today to do everything in the digital domain. The basic principle of loosening the AFM spring is to increase fuel delivery on acceleration, by making the door swing open faster to compensate for the lean mapped factory chips in part throttle. Disconnecting the O2 sensor also makes the car run open loop and keeps it from leaning out in part throttle situations. But bear in mind, especially on the older chips, this provides only a fractional improvement over reprogramming, and only works at part throttle. Also, changing the relationship of the AFM door swing to what has been programmed will shift off the part throttle ignition maps to a different load range that it was never intended to work in. The DME will actually think with the faster swinging door, that it is seeing greater load, and start retarding the part throttle ignition more than it is programmed to do. With the O2 sensor disconnected, you will also probably not meet emissions and fuel economy in steady state situations. Motronic also ignores all sensor inputs at full throttle, including the O2 sensor and AFM anyways, so no matter what you do by tinkering with those sensors, your car will deliver no more maximum power than was delivered from the factory by the stock chip. The fuel is fixed, and so is the ignition at full throttle. The only way to optimize at this point is to program within the chip. Those who already have performance chips should not alter the sensors, as it will throw off the delicate balance of fuel enrichment and ignition that has already been remapped in the chip.

In the past there has been much discussion about chips and the tradeoffs associated with them. Having dealt with may owners, I can say they work fine for the majority of Porsche owners out there, but there are some situations where it may not work well. By optimizing the fuel and ignition mixtures to work with either 91 octane, 93 octane, or 110 octane race fuel, one must assume their car is fully tuned without all the mechanical problems as outline in the previous paragraph. Carboned pistons and clogged injectors can lead to lean uneven fuel mixtures and light predetonation. Occasionally, after a chip install, the initial runs knocks out some of the carbon that has accumulated on top of the pistons over time, and the problems clear up never to resurface again. However mechanical problems can exceed this, and a chip won’t run well with this until they are fixed. In this case the stock chip is very tolerant of mistuned motors. To give you an idea, back in the day, a certain very famous Porsche tuner here in California when performing their 10.3:1 CR 3.4L, twin plug, 964 cammed, polished intake, sport exhaust conversions, did nothing more than use the original 1267355102 chip and bump the full throttle ign timing by 3 degrees, and set the rev limit to 6720. Nothing else was changed, not even any of the fuel curves or any of the part throttle ign curves. This give you an idea of how conservative the ignition and fuel margins are available on the stock chip, and still not predetonate, even when running so lean with this much compression. It is the chip user’s responsibility to ensure that he is using the proper fuel, and maintaining his vehicle properly, or if not, decide the aftermarket chip is not for him.

The key to optimizing the ignition curves is to keep the max ignition timing below the threshold of predetonation, yet still maintain an adequate safety margin. The ignition timing that produces optimum power in a 3.2 is not optimum for a street car running on pump gas. The successful club racers are now running anywhere from 35-36 degrees max ignition timing at 6-7000 rpm with 100-110 octane race fuel. But this will cause nasty predetonation with 91 or 93 octane fuel. Some of the early performance chips run their advance anywhere from 30 to 36 degrees. In programming chip, I generally like to keep max advance to about 27.5-28 degrees for 91 octane street gas, and about 29.5 for 93 octane fuel at the full throttle 4000-6500 rpm zone. This may be a compromise, and not deliver the maximum power on a dyno over what another chip with 33 degrees advance uses, but keeps the margin of safety I prefer. In addition, to control predetonation in hot weather and high temp condition, I implement modified knock protection which retards overall ignition timing based on temperature, so you can have both optimized timing for cool conditions, yet dynamically provide more conservatism when it gets hot. This is something I think few chip tuners utilize.

I know this has been long, but hopefully it is able to answer most of the questions most may have with regards to stock chips and aftermarket chips. My intent is not to start a flame war, but to provide the information as clear and as concise a possible yet not give the farm away.

Steve

[Jim Smolka]

Steve,

Thanks for the info.

[mjshira]

notice that there is no rubber left on the rear tires of Steve's car! (check his link)

[goat]

Steve.
Very informative
Didn't know about different chips used in the same years.
Bravo!

[alf]

I did not understand 95% of what Steve said

But i like the way his chip makes my 911 feel and drive.



alf

[mwbaum]

Great post Steve....thanks.

[88-diamondblue]

I have had the pleasure of speaking with Steve several weeks ago. This information above is what we talked about then and with more facts about how he programs his chips (detailed above) is why i will use the 911Chips chip in my car (replaces Autothority Chip). There is a balance between safety and power programed into the chip and most important, overall drivablity. The 911Chips chip has provided that for me. And as a plus, has customer service second to none. I will recommend Steve's chip to anyone without hesitation. The above post is what has been claimed all along, no huge HP and torque gains, but gains non the less which has greatly improved overall drivablity no more, no less. This upgrade to my car has only made that ***** eatin grin bigger than ever when i am driving

[tomt]

Uh.... what he said. I'm glad he's the guy that programmed my chip. My explanation goes like this:

"Drives real nice."

[Steve W]

My pleasure everyone. It's nice to share useful information without getting into personal attacks.

[NY65912]

DITTO

Thanks Steve!

Mike

[Wil Ferch]

I was one of the earlier Steve W chip pushers on this board.... I guess.

It was because of his *combined* knowledge of how an internal combustion engine works...and for his chip programming savvy. Can't just be good in one area.

"Conservative aggressiveness"....good label for his chips....

---Wil Ferch

[JeremyD]

Amen brother! and can I get a big amen from the gallery?

[Mark sP]

Still lovin it Steve! You Da Man Buddy!!!

[KFC911]

Thanks Steve! As I've said before, both you and Loren (and MANY others on this board) know more about this stuff than I ever will, and I read this board as an attempt to 'hopefully' make informed decisions. Thank you, Loren, and everyone else for sharing your knowledge with the rest of us. "This" is why I keep coming back to this place... !

[toolman]

Thanks for the info Steve, very informative in a way I can understand. I love the way the chip works with my modifications. You are truly a wizard !

[anh911]

Excellent post Steve. Thanks for all the information. And congrats on Pelican carrying your chips.

saved and printed.

[real550A]

Thanks Steve!
That's the most concise, informative school on remapping I've seen.
I knew I wanted one of your chips before....now, I have to have one!
I'll be calling you soon.

[Adman]

Originally posted by Jeremy964
Amen brother! and can I get a big amen from the gallery?

Amen....Steve!

[danmatar]

wow! great info. With regards to the different stock chips, by opening the ECU, can I tell what chip number I have. I'd like to see if I have one of those "dogs".