ZEUS+

It does not offset in the same equal voltage. You did not specify that situation is even present. I only replied to what it could do if it was. It would not have complete combustion. To many variables and assumptions.

stealth

My 86.5 is failing NOx repeatedly, as well. First test right after I bought it (May) showed very low CO and HC, and NOx of 3.05 grams/mile (limit 3.0)...I can't find those test sheets right now so I can't give exact CO HC...this was with factory cats and NO O2 sensor at all (it ran kinda crappy) and Autothority chips.

So, any excuse to get new exhaust...replaced factory cats with Motorsport X with 100cfm cats and installed new O2 sensor. Retest showed almost identical results (3.06 g/m NOx). Mechanic adjusted CO per shop manual after installing crossover pipe. These cats also do not require the line from the air pump.

Just last weekend, I finally tried again, having replaced the Autothority computers with my spare factory brains and driven with them for a couple of months (.03 rev of LH, i.e. '85, not the .04 rev that came in '86). Still low HC (.43 g/m, I think this was lower, .20 or less, for the first two tests, limit 2.0) and CO (.82 g/m limit 30.0) and NOx at 3.53 g/m (and horsepower quite noticeable reduced). The car was at least as warmed up, if not more so, for the last test versus the first two (first two were at same shop, last week's at different shop).

This car idles high but not ridiculously so, 850 or so, and idles stably. Can't hear vacuum leaks anywhere though I knew they're there. Replaced intake boots prior to first emissions test, those were in bad shape and leaking unmetered air into the intake for sure. Fixed vacuum leaks at charcoal canister diaphragm and NLA hose from there to the front of the strut brace from parts car. No hard starts since those were replaced, slight sputter at startup (probably leaky injectors?). Temperature gauge (no direct measurement taken) shows where a good-running early 32V should show...slightly less than halfway up the gauge, even in traffic, a/c off.

All the mechanics have helpfully advised me that I should check for vacuum leaks in the EGR system...

I have another engine with marrenized injectors and refurbed top end destined for this car so I don't care too much (if I were experiencing cognitive dissonance about emissions, I wouldn't drive the 928 in the first place) as long as I don't get a ticket...but it's quite a match for this thread. Note that I've replaced LH and EZF computers and still have the problem. Haven't replaced MAF or hooked my ARM-1 up.

928saregreat

Dave, Shane may have inadvertently hit on something, this is assuming that the wiring harness for the O2 sensor is functioning properly of course but a chipped 928 with a modified exaust and no other mods to the fuel injection will run lean. Adjustable fuel regulators are very handy with these sorts of changes. Could this apply to the 87 your working on?

Mark

Bill Ball

Just a comment about the "air" in the rails. I never paid any attention to what comes out when loosening the endcap on the fuel rail other than to capture that gas in a cup. Today, working on a car with a hard start warm issue but otherwsie runs well and passes emissions, we decided to put a pressure gauge on the rial. The car was still warm when we loosened the cap. There was a good amount of hissing and bubbling when the cap was slightly loose. So, I think this is normal behavior.

worf928

I think you're right Bill. I think it's 'normal' behavior for warm fuel system that experiences a pressure drop after the engine stops. The pressure drop - due to leaky injectors or a sticky check valve allows the warm fuel to vaporize. The check valve's job is to keep that from happening barring leaky injectors.

worf928

To recap: '87 AT. new high-flow cats, no air-pump, new 02 sensor. A known not-too-bad Mass-air Sensor. Good idle fuel pressure. Failed emissions with very low HC, very low CO, and very, very high NOx.

New info as of a few minutes ago:

The owner used my ARM-1 A/F gizmo and observed cycling behavior. But, the cycling was not centered on stoichiometric. The cycling was centered on the lean side - just slightly. With increasing RPM (static tests - car in the garage) the centering moved slightly more lean. So, the A/F measurement would indicate lean running while the car's supposed to be in lambda-loop control mode. It's lean at idle and leaner at a steady 1200 and leaner still at 1800 rpms.

The owner's going to change the fuel filter. It's at least three years old. Can't hurt. If the filter's so clogged that at idle the motor's lean, wouldn't you expect to see something reflected in the fuel pressure? Or wouldn't you expect the car to have NO power at WOT?

If the HC and/or CO was higher I'd suspect injectors in an instant. But, with these low HC/CO numbers 3 injectors on each side would have to be exhibiting the same behavior. Possible. The owner religiously uses Techron.

Too much air or too little fuel. Intake leaks - false air - seems like the other thing to contemplate.

John Speake? Do you know if the LH has a limit on idle or part-throttle fuel injector pulse width? Or, in other words, why cannot the LH compensate (provided that it's running in lambda loop mode) for even a huge amount of false air by increasing pulse width to the maximum possible duty cycle? And if it was so able to compensate wouldn't the expectation then be a too-high idle?

Mass-Air Sensor. Certainly the MAS could be so far gone as to cause this problem. Yet, wouldn't we then expect the car to exhibit other symptoms consistent with a necrotic MAS? Such as stalling during overrun? And starting problems? The MAS that's in the car now is the 'old' one from my '89. It's been sitting on a shelf for a year, but prior to that my '89 passed emissions with it.

I think I also want to see what the 02 voltage is at the LH harness. The bit of exposed copper on the LH-side of the harness got a bit of electrical tape prior to the test. Since only the signal wire was exposed and the heater wires were still nicely-insulated, the 'copper trail' was a false lead.

I must be missing something. Ideas?

Bill Ball

My narrow band A/F meter cycles slightly to the lean side too. Cycles from the top of the stoich area to full lean. Only goes into rich zones on WOT and then just barely. I think that's also normal.

I have a buddy with an 87 failing with high NOX and low everything else too. So far, we have not figured it out yet either despite new O2 sensor, one non-firing injector found and fixed and a cat swap. New cats on order.

the flyin' scotsman

Dave............going back to basics what does a running combustion engine do to produce NOx? Given that its a product of the fuel being burned is it then the expectation that the cats would remove the high levels down to what would pass the test?

You mention techron; if used in sufficient quantities could that alter the readings?

worf928

And when did you last pass emissions? (IIRC, wasn't your NOx creaping up over the past few years??)

Actually, it's weird though. Your description is pretty close to that given to me by the owner if this '87. However, the behavior of my '89 GT was very different. It would cycle past the top of stoich and never all the way back to full lean - it was pretty centered on stoich. Just breathing on the throttle would give me an instant of 'blue' in the rich zone. That was my baseline for 'normal' ARM-1 readings.

I'm hoping your description is NOT normal. Else I have no clues left.

Man I hope we're not going down exactly the same road together. (Or, on the other hand, maybe I do )

You found the one bad injector with a noid light? How was the car running otherwise? With the one non-firing injector?

rjtw

Bad O2 sensor? Sorry if it's been covered; I just jumped to the last page.
If everything in the system is checking out then maybe it's time to go back to stock cats.

worf928

My understanding is as follows:

NOx output levels are low if the A/F mixture is rich. NOx levels rise as the A/F mixture approaches stoichiometric. In this context, the primary purpose of the cats is to reduce the 'unacceptable' levels of NOx produced by the engine when it runs stoichiometric. NOx levels rise exponentially as the A/F mixture leans past stoichiometric. (There's a nifty graph of this in the Bosch book. And I googled one up the other day, but cannot find it now.) The cat has a limited ability to convert NOx compounds to 02 and N2. The exponentially rise of NOx with increasing A:F ratio means that a mixture that is only a little lean (say 10 or 20%) will produce NOx above that which the cat can clean.

Note that while stoichiometric refers to a 14.7:1 A/F ratio and theoretically would result in perfect combustion with only N2, CO2 and H20 as the result - the reality is that nothing's perfect.


I dunno. 'spose it could. But, I doubt techron has been run continually. I will add it as a variable to track though.

Here's a nifty page on Catalytic converters. I just found it and haven't read it all yet.

jon928se

Dave

attached diagram from Bosch indicates what should happen emissions wise, Although I don't think it helps with your issue.

The right hand image has O2 sensor volts in red - top of the straight line is about 0.8v and the bottom of the straight line is about 0.1v

I still am inclined to think that the issue is the wiring between the O2 sensor and the LH - to the best of my limited knwoledge the O2 sensor can't produce anything other than it's normal 0.8v at the rich side of stoich and 0.1v at the lean side (before the outputs curve off the straight line)

I would say a duff sensor had you not just changed it for a new one with no improvement.

Something is causing the LH to see an O2 output as if say a constant 0.1v had been added to the cycling 0.1-0.8v thus the LH sees 0.2 - 0.9 and tries to adjust the AF lean to make it back to 0.1-0.8v

Bill Ball

Funny, the O2 sensor in my car is new and I do now recall the old one cycling more like your GT, using the full meter. Car still runs like a champ, although the recent smog test did show NOX mid-range rather than low. Plugs look perfect.

Do the plugs on the car in question show evidence of leanness?

Regarding the non-firing injector - I could tell my buddy's 87 was only running on 7 cylinders. Narrowed it down to one cylinder by pulling spark plugs wires. Plug was dry, so we pulled the fuel rail cover and found the clip had fallen off the injector lead and it was compltely unplugged. I was feeling pretty certain we had found a decent possible cause for high NOX, but although the car ran MUCH better, NOX did not improve.

worf928

That's the one for which I was looking.

But, it gives my descriptions a visual aid.

Yup. (I didn't claimed otherwise. All indications are that the condition is 'slightly' lean. If it was VERY lean then we would expect high engine temperatures and perhaps even knocking. Neither of these occur.)

Yes. The observed behavior is completely consistent with an offset being added to the measured signal used as the basis for control.

Failing any other ideas, my plan as of this instant, would be to get the car back so that I can get my hands on it, read the 02 voltage directly at the ECU harness connector and perhaps swap the LH for a 'new' one I have on a shelf.

I am also curious about Bill's observed ARM-1 behavior as contrasted with my observations of my '89.

Yet, if Bill's observations are truly normal, and my '89 is abnormal, then the question arises: If the lump is not lean then why the @#$% is the N0x so high? (Given the other variables eliminated so far.)

I still have not ruled out:

- numerious clogged injectors
- air getting sucked into the pressure-side of the fuel system
- MAS (but seems unlikely to me)
- demons and spirits

worf928

More ARM-1 observations are called for me thinks...
Reminder to Dave: add pulling plug(s) to plan of attack when next my paws are on the car.
I would have been suprised if that were not the case. In my case this '87 seems to run nice and smooth.