Hi, I have a fresh ej22  with about 1000 miles on it now. It is outfitted with 1994 Legacy externals. (all reused Legacy parts: wiring harness, ECU, throttle body, MAF are from a single complete donor car)

 

It is installed in a 1990 vanagon and it runs great. No complaints.

 

Much is new: CA-certified Cat, Denso O2, OEM fuel injectors, idle control valve, plugs NGK irridium, plug wires, etc. all sensors new OEM,  all hoses belts, and pulleys new OEM.

 

The only hitch: I took it in for CA smog today and the  HC and CO were super clean, but the NOx was barely OK on the 15 MPH test, but was about 30% above limit on the 25MPH test... FAIL

 

The van is automatic and the engine speed is close to 2000 RPM on both tests, but at 25 MPH it is in top gear.

 

The smog guy suspects the O2 sensor. My mechanic suspects that the throttle position is set improperly.

 

Any insights would be most appreciated.  

 

Thanks,

 

Eric

 

 

isn't high nox a sign of lean mixture?

Thanks, Tex!

 

That makes sense, but what factors will play into the lean condition? (is there anything besides the obvious intake leaks?) bad O2 sensor? 

 

Although every hose was clamped, checked and double checked, (after the engine compartment photo was taken)  there is the plastic intake part that clamps to the remains of the subie intake duct...  Only God knows why the MAF is located so far from the throttle body (increasing the potential for leakage) I'm going to have to find something that sticks to polypropylene and seal that joint better...  Has anybody had any luck gluing/sealing polypropylene? Product suggestions?

 

Back in the day, working with weber carbs, we used to spray some ether around the intake manifolds while idling and see if it idles up a bit. does this work with the subie, or does the computer compensate?

 

 

Eric

I'd think starter fluid, propane or even carb cleaner should get a detectable increse/change in the idle - even if just for a few moments.

 

be careful

A MAF leak will have its most noticable affect at idle. Once the engine speed runs up a bit a small leak will have less effect on the AF mixture.

 

The automatics need EGR basically for this reason. The engine load conditions at a steady cruise speed cause a lean condition which increases combustion temperature and thus increase NOx production.

 

This is part of why EGR was implemented on 95 and later years.

OK, rather than leave another dead-end thread, here is what solved the problem...

 

 

Having been through the intake leak maze when the engine was first installed, all the intake stuff was double checked and found to still be tight and free of leaks.

 

Based upon the ease of changing the part, and direct relationship to A/F ratio, I swapped in a new Denso O2 sensor. (BTW, this what the smog guy suspected was the problem)

The HC and CO rose a bit as anticipated, and the NOx dropped to almost zero. The engine passed rigorous CA smog with flying colors.    

Apparently the first Denso O2 sensor was defective, as it only had 1000 miles on it, and was not worn or abused.

 

The biggest hurdle was that everyone assumed the problem couldn't be with the O2 sensor because it was new. 

 

My buddy ( a 40 year veteran mechanic) says that: "the quality of auto parts in general has declined in the last several years, and failures like this are becoming more common." 

(even from previously well-respected brands) This is not good news for those of us who maintain our own vehicles, as it adds time and uncertainty to our efforts.

 

On the bright side...The new EJ22 engine is now running even better than ever! 

 

 

 

A hearty Thanks! to those members who generously shared their knowledge and analysis! 

Good to hear you got it figured out!

Sad to hear that a Denso sensor was a bum out of the box, haven't ever had any problems with them myself.