Unique mixture unit behavior. Any ideas?

[vwdmc16]

Working on another car that runs okay but cannot pass smog due to high CO.

New ignition system: plugs, wires, cap rotor, coil, timing set to 13.

Great fuel pressure. Correct pressures at WUR Hot and cold

Cleaned injectors. Double checked while installed in the car. Flowing evenly and with good spray pattern.

Smoke tested for vacuum leaks. None

Now it gets odd. Installed a wideband O2 right before the stock Cat. Set the idle fuel mixture to 14.7:1 runs great.

Go for a test drive with the wideband recording info. As soon as you tip into the throttle the fuel ratio goes lean, very lean: 17-18:1

Return to idle it holds 14.7:1 all day all temp.

Try to richen up the mixture. Now its 12:1 AFR at idle. Start driving and again any time off idle it's super lean but it did improve.

Really richen up the mixture. 10:1 AFR at idle, the engine hates idling. Go drive the car and finally have good fuel mixture under load. High 13:1 AFR. But that's not sustainable as it will foul plugs at idle.

I've never seen any fuel injection system get lean as throttle is applied. It's as soon as the idle microswitch is turned off it go into a totally lean setting everywhere.

Fuel pressure and voltage only go up when above idle.

Ideas? I'm thinking a bad fuel distributor.

[DMCMW Dave]

Verify that the Lambda system is really working. You can set AFR with a CO meter and get it to read correctly, but if you are overcompensating for a dead Lambda system (O2 sensor/computer/frequency valve) it will be very lean the rest of the time.

Verify that the vacuum advance solenoid is working correctly. If that fails and you get vacuum all the time, you'll be setting the timing extremely retarded. Normally unplugging the vacuum line from the distributor won't make any difference at idle. But if the control solenoid is defective or unplugged, you'll have vacuum at the distributor all the time. If you set the initial timing like that, timing will be very retarded and cause some really screwy CO readings.

[Bitsyncmaster]

You will see a lean spike when you punch the throttle when recording. Most every engine will do that because the system can not respond fast enough. The wide band has fast response. The throttle plates open so fast the lambda system has no time to react.

[Ron]

Verify that the vacuum advance solenoid is working correctly. If that fails and you get vacuum all the time, you'll be setting the timing extremely retarded. Normally unplugging the vacuum line from the distributor won't make any difference at idle. But if the control solenoid is defective or unplugged, you'll have vacuum at the distributor all the time. If you set the initial timing like that, timing will be very retarded and cause some really screwy CO readings.

+1

If it does make a difference when you unplug the distributor at idle, check if the vacuum solenoid lines are reversed. If so, the solenoid will not vent properly which can hold the timing high until you shut the engine down. How this affects the timing/AFR depends on whether the timing is set before or after the first time idle speed is broken....

[David T]

You will get a lean spike if the CPR is not hooked up right or is not operating right. I agree, start with checking the timing, vacuum and mechanical advance. Verify the Frequency valve is buzzing. In fact, there are tests in the Workshop Manual D:04:15 to make sure the Lambda system is functioning correctly. The mixture screw sets the BASE (lean) mixture and the frequency valve adds just enough to get the A/F ratio closer to stociometric. If the Lambda system is not working you will always be lean unless you over-richen the mixture screw to compensate. That will only work at idle and you will lean out as you increase the RPM's.

[vwdmc16]

You will see a lean spike when you punch the throttle when recording...

This engine isnt producing a lean spike but full lean condition. It plateaus and stays consistently much leaner than it is supposed to, if I barely tip in the throttle or punch it it goes lean and stays lean, never returning to stoichiometric until it returns to idle.


The vacuum solenoid was clicking with the idle switch, timing was set with vacuum disconnected.
I remember feeling the suction at the distributor hose when the throttle was opened.

I'll double check that it is venting the vacuum.

[Ron]

To be sure:

Installed a wideband O2 right before the stock Cat. Set the idle fuel mixture to 14.7:1 runs great.

It was going lean before this, correct?

10:1 AFR at idle, the engine hates idling.

It liked idling before this setting, correct?

======

Fuel pressure and voltage only go up when above idle.

Normal voltage at idle?

Which fuel pressure(s), system, primary, and/or control?
The system pressure rising with RPM can be normal.
Once the engine is warm, the control pressure should not change.
The primary pressure should never change!!

[vwdmc16]

To be sure:

It was going lean before this, correct?

It liked idling before this setting, correct?

Normal voltage at idle?

Which fuel pressure....

Unknown how lean or rich it was. The car ran decently. Connected the wideband to get more info.


Yes when set to 14.7 AFR at idle it does idle well. No hunting, very smooth.

13.4 v at idle and up to 14v when revved up.

Fuel pressure measured at fuel distributor inlet from the pump. This confirms the pump isn't just falling off when fuel load increases.

[Ron]

Unknown how lean or rich it was. The car ran decently. Connected the wideband to get more info.

So, fwiw, it is possible that it could have passed before you put the wideband in?

Fuel pressure measured at fuel distributor inlet from the pump. This confirms the pump isn't just falling off when fuel load increases.

If you mean you used a tee joining the line from the filter, the FD main inlet, and the gauge, we might be on to something.
If I'm following you, that would be primary pressure which shouldn't change. If it rises when you raise the RPM, the control pressure would rise and press the plunger down which would make it lean.
What readings are you getting?

This also fits the frequency valve, as Dave's suggested. (Might not be able/told to close enough...)