Fuel in Oil; MFI... continued

[Tspringer]

For those new to the saga... my '69 911S with 2.7 RS MFI engine was puking fuel into the oil in quanties large enough to raise the oil level and blow it out the breather. A bad problem.

We know the problem is not any sort of internal leaking within the MFI pump. The pump is setup to NOT share oil with the engine so there is not way for a pump related leak to foul the engine oil.

That leaves only 2 possibilities: A bad injector is allowing fuel to leak past under low pressure and then flow past the rings into the oil OR the car is running so rich that it is washing fuel past all the cylinders and fouling the oil.

The car runs great. Lots of power. Does not foul plugs. Does not throw black smoke.

Now for the latest news. I got my new toy, a Gunson gas tester. This little green box allows me to test the CO content of the exhaust and thus tell whether the car is running rich or not. Here are the result of my using it this afternoon:

First Test:

The car was driven for 20 minutes to bring it to full operating temperature. It was then shut down for 40 minutes while the tester was setup. The tester was mounted higher than the exhaust on the car, not strictly as instructions call for. I tested by letting the car run at the specified rmp until the tester had settled down to a steady reading and then letting it stay there for at least 3 minutes. Here are the CO % test results:

Idle: 9.5
2k rpm: 12.9
idle: 7.9
2K rpm: 12.0
idle: 3.4 to 2.6 drifted down slowly
2k prm: 12.0 drifted down slowly to 9.9
idle: 2.2

I found these results frustrating and strange. I was doubting the gunson at this point and went back to the instructions. I decided the car may have cooled down too much before starting and thus was rich due at first plus the gas tester needed to be on a level or lower than the car. I positioned the tester again and set it up to recalibrate while I took the car for a 25 minute drive. I tested immediatly upon return:

Test 2:

idle: 2.3
2k rpm: 8.0 hunting somewhat
idle: 3.0 to 3.2
2200 rpm on hand throttle: 13.3
idle: 3.2 to 2.7
2200 rpm on hand throttle: 13.2
idle: 2.8 to 3.2

These results were more stable and using the hand throttle was more steady than my attempting to hold the throttle steady at the linkage while watching the tach. I also decided to recalibrate the test again after moving it to a new position lower than the cars exhaust and with a straight hose. I was using a 15 foot piece of hose to attach the exhaust probe to the tester and I insulated the tester in a box protecting it from any wind or movement. I let the tester calibrate for 30 minutes. The car had been stationary not running for about 10 minutes when I began the test:

idle: 8.1 after seeing this I gave the throttle a few quick blips and let it idle again for 4 minutes, it drifted down to 7.0
2200 rpm on hand throttle: 13.2 - 13.5
idle: 2.4 - 2.6
2200 rpm on hand throttle: 12.5 - 13.0
idle: 2.2 - 2.1
2200 rpm on hand throttle: 12.5 - 12.7
idle: 2.5 - 3.1. most stable in 2.8 region

Based on these tests, it would seem the idle mixture is pretty dang good. The off idle mixture seems way too high?

Im not at all sure what all these test results mean? Any ideas?

Terry

[Tyson Schmidt]

You should lean out the main rack until you get about 8% consistently at part throttle. This will effectively lean out the idle mixture too much, so compensate by richening the idle mixture. Shoot for around 4% at idle.

[brokke]

Hi,

I am using the gunson tester to confirm the road tests. Just lean out the main mixture to the point the car stumbles on the road at the transition from 1500 rpm to 3000 rpm (mainly at part throttle, less at full throttle). Then rich up the main mixture untill the car drives fine. Then confirm the road test with the Gunson tester and dial in the idle mixture. I found:

7% - 8% CO at 2000 - 2500 rpm
3% - 5% CO at idle

Mostly it is 1 or 2 clicks form slight stumbling to running fine. The fuel consumption figure is also a good measure. When the car is adjusted spot on I am doing on a long trip on the open road about 8 km per litre gas (speed 80 - 90 mph), short trips 6-7 km per litre gas.

Plugs are dry but sooty around the threads but the insulators look fine ( sometimes on the lean side, rather white). I am using Bosch wr4dpo's.

I thinking she is dumping some fuel in the oil, but not much. I use the oil preasure as a measure for the amount of fuel in the oil. With 15w40 I am getting at 180 degrees F 55 - 60 psi at 3000 rpm. It takes at lot of km to lower that figure to fore example 50 psi at 3000 rpm.

Peter
911 2.4S

[Tspringer]

Im going to mess with the mixture adjustments when I next get time to mess around in the garage. I'll combine the road testing with the gunson and shoot for 4% at idle and 8% at 2200rpm on hand throttle.

Still.... even with the numbers I am seeing now I dont see any way this slightly rich off idle condition could be responsible for putting so much fuel in the oil. I do think it could put some fuel in the oil..... but probably not enough that it would not evaporate off.

At this point, Im thinking the problem is injector related.

Now I will throw in another little kink. I have driven the car about 250 miles since the last oil change and I ran it for at least 1.5 hours yesterday during driving and tuning. When I was done, I inspected the oil by repeatedly inspecting the dipstick. The oil had no gasoline smell and its consistency did not indicate any dilution. Hmmmmmm

I have run a bottle of techron through and am doing another bottle now. Is it possible there was a carbon deposit partially blocking an injector and allowing fuel to leak past under low pressure? I can imagine a scenario where under the 220 psi the injectors fire at it could work fine but then at low/static pressure it could allow fuel to drip past. If this happened after shutdown until fuel pressure was gone, it would certainly put fuel in the oil. Perhaps I am lucky and the carbon deposit has been cleaned off?

Im going to adjust the mixture and then watch the oil very closely. If I get any indication of fuel in the oil, I will order a new set of injectors. If after replacing all injectors its still dumping fuel into the oil.... well, I'll really be stumped!

After tuning the mixture, and taking more gunson readings, im going to take it down the street to the emissions testing place and get some accurate readings from then as see how the gunson compares.

Terry