Optimal AFR for 1980 911SC

[mradovan]

Steve and others,

I am looking for some advice on how to obtain optimal performance from a 1980 911SC. This car is NOT part of the bi-annual Ca smog program because it is in a rural county. First some background:

Car has now has approx 100 miles since I rebuilt it. What started as a two broken head studs ended up being a complete (everything !) top to bottom rebuilt. Partly due to my curiosity. Current spec are:

964 Cam profile put onto sc cams
SSI heat exchangers
recurved dist by Jerry woods.
MSD 6200 ignition with MSD coil

I have added an innovate wide band AFR meter which has a narrow band output hook to the CIS analog computer. Additionally I am using the innovate data logger to record freq valve duty cycle and RPM. All vacuum lines, hoses, boots etc on the CIS are new. I have been experimenting with control pressure and primary pressure and I have stacks of interesting data logs that I can make available to any interested parties.

First question- Idle richment adjustment. O2 sensor unplugged on a 1980 puts the duty cycle at 85%. If I set the idle richness with it unplugged the car leans considerably when the O2 is reconnected because it defaults to a 65% duty cycle. See attached jpeg for an example. If the 15C switch is open this car starts at with 85 percent duty cycle but not for long. Once it goes to the 65% setting it is pretty lean. FYI the primary pressure for this graph was 5.4 bar and the cold/warn control pressure were 1.6 and 3.5 bar

What is the best way to set this? O2 connected or disconnected? What CO or AFR?

I have since modified the warm up regulator to make it adjustable. Based on a WOT test and the duty cycle of the car at constant rpm under light load I concluded the car wants more fuel. See attached jpeg " 5.4 primary and 3.5 control". In this log at time=17.0 to 17:10 I am cruising at light throttle at 3000 RPM (black line). The AFR is is between 14 and 15 (appox) (magenta line) and the duty cycle is between 65 and 75 % which I think is high. Just after 17:15 I accelerate at WOT for a few seconds and you see the duty cycle go to 65% because the 35 degree (throttle plate) switch has put the system in open loop. The AFR during this period is lean as shown. 5 or so seconds later i do the same test up to about 4400 rpm with the same result.

Question for Steve- I have read that during break in I should stay under 5000 rpm. Is this test safe or should I wait until I have more miles on the car to be doing this experiment?

Based on these logs I have dropped my control pressure to 2.9 bar and things have improved. I also have experimented with bumping the primary pressure a little. It is currently at 5.8. Interestingly the AFR variation (high frequency changes) at idle and during operation is increased with higher control pressure. My first logs were with primary pressure at 4.6 Bar. I upped the primary because it was quick and easy to do. This was before the WUR was modified to be adjustable. I plan to repeat the tests with lower primary and control of 3.0 if it is safe to continue this type of testing.

The last log for now is for a primary pressure of 5.8 bar and control at 3.0. In this the WOT response is much better all the way to 4500 RPM. Duty cycle at cruising constant RPM has dropped to 55 to 65%.

My concern is that although this might be acceptable what will happen when I operate at WOT to red line? I suspect the 65 % duty cycle will not be able to provide enough fuel. Comments? I could lower the control pressure to compensate but I what point will that cause other problems. I also had an idea that i could add a limit swith disconnect the o2 sensor signal thinking that would trick the CIS controller to going into 85 percent duty cycle for extra enrichment. I tested this theory with a toggle switch on that signal and discovered that once the 35 deg throttle switch puts the unit into open loop mode (65% duty cycle) it doesn't see the O2 switch being disconnected. So I say what about using the 15deg c switch? Will that for 85 percent duty cycle? Wiring would be simple. Add a switch and use a relay to switch the 15C circuit. The last option is to disconnect the 35 deg throttle switch and use the O2 or 15 C switch circuit to get 85 percent duty cycle at WOT. Has anyone ever tried or done this?

Thanks for you help and wisdom.

Matt

[Steve]

Hi Matt,

Thanks for your very detailed post.

I've not had great success tuning these late SC's due to their use of closed-loop fuel control. They simply run too lean for good driveability & best power, especially when modifications are made. CIS systems all share one major issue; they are either too lean or too rich at one end of the RPM/load range and its indigenous to these FI systems. I've had better success with the early, non-oxysensor equipped versions as they deliver more fuel, however they still have the same limitations.

Best (safe) power for these engines is found between 12.7-12.9 AFR and this is difficult to achieve, even with the digital, programmable WUR's.

Keep the max engine RPM below 5K until the rings have seated and you'll know that from your leakdown tests.

Given all your efforts, you maybe be much better off using Engine Management to achieve what you want.

[mradovan]

Steve,

Thanks for your recommendations. My plan is to explore the possibility of using the 85 percent duty cycle mode of the 1980 to do the extra enrichment. I might do this with a double pole micro switch on the throttle plate. If I set it to activate at some angle greater than the 35 degree switch I can use it to trick the computer to deactivate the open loop mode (65 percent duty cycle) and to close or open (whichever is required) the 15C switch to make it think it is in cold start up mode. This will provide more fuel.
I also plan to explore what happens if i lower the control pressure a little more. When I do this I will disable the 15c feedback so the car never starts at 85 percent duty cycle. The 81-83 start at 65 percent and they have the extra enrichment feature. I will compensate for that by adjusting the system to run in the mid 13's on AFR

Because I have no info on the logic of the O2 sensor controller i am doing some reverse engineering. I suspect that the best i can hope for is to get closer to an ideal situation but never there. You and I talked once on the phone about the problem with the 911SC and doing any performance upgrades. The next step is a big one in terms of cost and might not be all that logical.

If I decide to go down the slippery slope what would you recommend as the next step form engine management and what should I expect for material costs?

I will post AFR curves for further experiment in case other readers are interested. One interesting point I touched on in my earlier email is the closed loop response of the system as a function of primary pressure. There is a small gain by raising this value based on the average AFR but there is more high frequency variation. Since the O2 controller is an analog computer with a fixed control loop I can't tune the PID terms. I suspect I am under damped and that is why the high frequency variation is increased. I don't notice it when driving but I suspect there is an upper limit to the primary pressure at which point it would become a problem.

Once I get the CIS to the best possible configuration I will dyno the car and post the numbers. I really want to know what the best you can do on these cars using the stock configuration plus the few small changes I did during the rebuild.

Thanks again for your help and for hosting this forum.

[Steve]

[mradovan]

Update on my quest for optimal AFR:

After doing some additional investigating which was prompted by a nagging feeling that the recorded AFR numbers vs duty cycle were a little too far out of range I made an interesting discovery. The #2 fuel injector on my car was stuck closed. This is the side that the wide band O2 sensor is on so the reported AFR numbers were incorrect. All the data presented in my earlier post is wrong and should be ignored. I will re post new curves in the near future for anyone who is interested. The bottom line is now that for a warm control pressure of 3.0 bar I have plenty (probably more than I need) enrichment for full throttle operation. Because the engine is not broken in I have not ventured past the 5000 rpm mark.

I am now encouraged that with this tool I will be able to optimize the systems. Eventually I will dyno the car and report he numbers at which point I will know that this is the best that can be done with CIS.

Matt

[Steve]

[mradovan]

Rennsport Community,

For those who are interested in the subject I am providing this new set of AFR curves for my car with 6 brand new fuel injectors which I know are working correctly. Total fuel delivery was checked for each injector to confirm nothing funny was happening. All the curves bellow are for a system pressure of 5.4 bar. When I stated this process the car had a system pressure of 4.6 which I boosted to 5.8 for the above curves in an attempt to get the car to run richer however all that data was flawed due to the #2 fuel injector being stuck closed. With the new injectors in place I decided to lower the pressure to put it just a little past what Porsche considers in spec. (Bentley 201-5 shows 4.5-5.2 bar as within spec).

The first curve is for a warm control pressure of 48 psi which is slightly lower than the spec according to the Bentley manual which lists the value as 50-55 psi (page 240-14). Black is RPM, Red is the Lamda duty cycle and magenta is AFR. The time scale is extended to and there is no data smoothing on any of the graphs. Between T=9.7 and 9.9 the car is closed loop just briefly and is maintaining around14.7. This first full throttle acceleration is in third gear to about 5900 rpm. As soon as the system goes open loop (duty cycle=65%) the car quickly goes much richer and stays flat around 12.0

In the second graph I increased the control pressure to 51.5 but I did not adjust the IDLE afr which shouldn't effect the full throttle operation however I did not confirm this with a test. The important thing to note is that the full throttle AFR was 12.44 and holding steady.

The last curve the warm control pressure was increased to 54 and I adjusted the idle back to 13.7 because the car was running bad at idle (too lean). For this test I did full throttle accelerations from 2-4 gears. Now the AFR is around 12.7 which is more like it.

The last graph is the before and after idle AFR adjustment. The lean setting was correct when the control pressure was 48 psi but when I went to 54 the car went from an idle AFR of 13.7 to almost 15.4.

All these adjustments were made on the warm-up regulator which I added m5 tapped hole to so I can pull up the warm of cold plunger. The warm and cold setting are coupled some so I will have to adjust the cold control pressure to get the car to behave during cold start up. That remains to be done because the car needs to sit for many hours to be sure the warm up regulator is at ambient temperature. Because I got sick and tired of connecting my gauges each time I made a small block to carry a fixed fuel pressure gauge as shown. It is also worth noting that my CIS system DOES NOT run at 85% duty cycle on start up or when the oxygen sensor is disconnected as reported in some of the literature. I added a toggle switch to the circuit so I could disconnect the O2 sensor and I always saw 65% not 85%. As it turns out my idea to trick the system was not required because my original conclusions were flawed due to the bad data from the stuck fuel injector.

Conclusions: CIS is capable of delivering as much fuel as these cars need which is no surprise considering the fact that Porsche stuck with it on the Turbos for some time however it is pretty sensitive. I can't saw enough about the convenience of the wide band AFR sensor for making these measurements. For those who care this is a very cheap must have.

Now for the expert opinion-
Steve- What do you think about the last curve? Is this right or should I lean or richen the car at WOT.

Thanks again for hosting this forum.

48 psi with 13.7 AFR idle.png
51.5 PSI with idle afr approx 14.7.png
54 PSI with 13.7 afr idle.png
IDLE AFR Adjustment.png
IMG_0720.JPG

[Steve]