edmcguirk

I tried to pass NJ inspection in December and it failed emissions. I wasn't really suprised because the AFR gauge was acting funny. I thought the wire to the gauge was bad but the failure was closer to the sensor.

No big deal. I bought a new oxygen sensor because they are cheap but the problem was a broken wire. I replaced the sensor but I couldn't fix the break in the wire because it was under the manifold. Instead I cut the oxygen sensor wire at the ECU and connected the wire there. The wire's color code in my 87 didn't match the color code from my 86 workshop manual but I think I got the right wire - the neighboring wires all matched.

Anyway, the NOx measurment was low (80 vs 1378 pass) and the HC and CO readings were high (HC 372 vs 180 fail and CO 2.14 vs 1.01 fail). Those measurements imply the engine is running rich but the AFR is reading lean. That usually means that I have some kind of misfire but I can't find any problem. The car idles smoothly and in all respects feels normal. It gets normal MPG.

I replaced the spark plugs and also reinstalled the old oxygen sensor and it didn't make any difference.

I understand that narrow band AFR gauges are not accurate but under normal operating conditions I expect the AFR gauge to look like this:
Off throttle should be full rich until fuel cut kicks in and then it should be full lean.
Full throttle should be medium rich.
Cruise should enter closed loop mode and pulse rythmically rich to lean.

Instead of that my AFR looks like this:
Off throttle is full rich until fuel cut kicks in and then it is full lean.
Full throttle is on the lean side of the stoic range.
Cruise is on the lean side of the stoic range and never goes into closed loop.

Are there any sensors that might be preventing the ECU from entering closed loop mode?
What elso should I look for?

Monday evening I will take it back to NJ inspection in case the gauge is lying to me but I think the car will fail again if it will not go into closed loop mode. This is a race car and I want to fix this problem regardless of whether I pass NJ emissions.

ed

spot_skater

I'm not sure if any sensors could be screwing around with your AFR's, but a quick check to see if your car is running in closed loop is to let the car warm up to operating temp, put a DVOM/Multimeter on the O2 sensor wire, and if it is constantly cycling from .1v-.9v, then you're in closed loop mode.

Closed loop is basically when the O2 sensor reads whether the motor is running rich/lean, then compensates for it using other sensors on the car. If it's cycling in that voltage range, you should be in closed loop.

If it's constantly low (between .1-.3) then you're lean, and the computer will compensate by injecting more fuel.

If the voltage at the O2 sensor is constantly high (between .6-.9v) then you're constantly rich, which will send a signal to your injector to put in less fuel.

If it's a problem you can't diagnose, try to get it on a scanner and check fuel trims, if possible, and go from there. I don't remember if OBD-I has this capability, I know OBD-II does, though.

Good Luck,
James

edmcguirk

I don't think any other sensors are messing with my AFRs. I think maybe some other sensor is telling the ECU that it's not time to go in closed loop mode yet.

I didn't see that information clearly spelled out in the manual.

I guess the temperature sender could be broken. I think I remember that the ECU will assume normal operating temperature if the sensor is unplugged. But if the temp sensor was reporting a cold temperature, I don't think the mixture would be enriched to the point of misfire.

ed

edmcguirk

Boy, don't I feel foolish. I spliced the wrong wire at the ECU.

The wrong wire was loading down the oxygen sensor and the ECU wasn't getting the signal so it couldn't go into closed loop.

The signal still looks a little funny but it does go into closed loop. I'll run a few more tests and then try to get inspected again next Monday.

ed

RotaryEvolution

iTrader: (30)

the O2 sensors on these cars don't really affect emissions much except under certain circumstances which even a loaded dyno test usually isn't one of them. it is an extremely primitive control system, basically tuned from the factory so any sensors out of whack or faulty will put emissions through the roof.

try checking the timing, adjust the TPS to spec, check your TID(if you have one, didn't see a car specified) for cracks, check for vacuum leaks, check the wiring harness for damage and check to see if the sensors are within spec by ohming them out at the ECU connector, if everything checks out ok try leaning out the variable resistor by turning it counterclockwise until the engine just starts to misfire and have it retested, if it still fails i would consider replacing the catalytic converter.

also be sure the ignition system has had a complete tune up.

i passed emissions without an O2 sensor hooked up, so don't beat your head against the wall looking in the wrong area.

edmcguirk

In NJ the emissions test is on the dyno and it will not pass unless it goes into closed loop. I have watched the AFR gauge when the car is on the dyno and saw the rythmic pulse of closed loop in operation.

It is highly unlikely you could tune the car to such a fine edge that it would hit the right numbers without closed loop operation.

ed

HAILERS

There's a fun little test tool you can make out of a single LED and two wires. You connect the wires to the six socket check connector in the engine bay and you can run the LED and wires into the cabin.

What happens is that when the ECU goes into the closed loop, the LED will flash on/off/on/off several times per second. It usually happens when in gear and over 1500 rpm.

But a common digital volt meter backprobed into the 02 wire on the ECU will do the same. Not as flashy as the light though.

The sockets on the green check connector where you'd insert the two wires would be the two center sockets.

edmcguirk

I thought there were two leads in that connector. A rich and a lean signal.

I've already got the AFR gauge so there is no need for a voltmeter on the sensor. The AFR gauge is really just a voltmeter.

HAILERS

********AFR gauge is really just a voltmeter*********.........not in my predjudiced opinion.

That green diagnostic connector............There's the socket below the black/white wire in the middle. That socket will pulse a ground signal several times a second when the engine is in the closed loop mode. The black/white wire just above just supplies the voltage to the LED.

That light pulsing would mean more to me than the afr gauge as far as whether or not the 02 was in the closed loop mode or not.

There's no Rich/Lean signal in that diagnostic plug (as far as I'm concerned. Depends how you look at things).
'
That LED will work just like the GREEN LIGHT in the diagnostic tester works as described in the FSM. The chapter called DIGITAL CODE CHECKER.

You don't even have to drive the car to get in closed loop mode if you follow the instructions in that chapter.

edmcguirk

I don't have the manual with me. I thought I remembered it differently. I thought there were 2 terminals and the description led me to believe it simply followed whether the oxygen sensor was rich or lean. I will read it more carefully tonight.

Inside the AFR gauge I have is just a 10 segment voltage divider chip. For all intents it is a 1 volt voltmeter. It's not as accurate but the function is the same.

The characteristics of the oxygen sensor signal when the ECU goes into closed loop are pretty unique and obvious. Even though you can't really trust a narrow band oxygen sensor for mearuring how rich or lean you are operating at, I find the AFR display fairly expressive of what's going on.

I even considered converting to a heated sensor just to help it along a little. Since the header mounts the sensor farther away, my sensor cools off at idle. Also, the dedicated ground in a heated sensor is not a bad thing.

ed

RotaryEvolution

iTrader: (30)

california emissions are among the toughest to pass. *shrug*

edmcguirk

I didn't notice you were in California. Up to now I would have bet money that you couldn't pass without an oxygen sensor but I stand corrected. (but maybe you just have a particularly clean car)

However, if all the sensors are working, there aren't really too many adjustments you can make to the mixture (I agree, the ECU is not very sophisticated). I got the impression that the mixture adjust resistor was mainly for idle mixture. I don't notice a big change when driving.

I was more concerned when I had the wires mixed up because the readings were not making any sense. Now that I have the wires right, it looks like I should be able to track down any remaining problems easily.

ed

HAILERS

I'd check the air control valve to see if it's working right. It oculd be the air that should be going to the exaust ports is getting dumped overboard out the silencer in the right fender.

Also if it's a old acv, the diaphrams tend to crack due to heat . I don't know what you have, turbo or non turbo but you might check the diaphrams out.

On a non turbo the diaphrams are fed by two metal nipples just above the acv. PUt a vacuum hose on each and suck on them to tell if the diaphrams are good/bad.

On a turbo one of the nipples is ON the ACv itself on top and sticks straight up. It is for vacuum. The other nipple is at the very bottom of the acv on a mushroom device and it gets PRESSURE to move it.

The two solenoids on the acv itself have little to do with passing emissions. The front one, the Port Air Solenoid is disabled after twenty thousand miles and the other solenoid is activated when in fifth gear only.

I'd say if the two diaphrams are good and the signals from the Relief and Switching solenoid valves are good to the acv, then ..............buy another cat and your guaranteed to pass.

And the tps should be set right. You probably already know, but if using the two LED light method of setting the tps, that the ground signal to the tps connector to light up each of the two lights is ALSO the ground that turns the Relief and Switching solenoids on/off.

HAILERS

And if it helps any.....if the acv is working right, then a 02 sensor should read LEAN at idle because, the acv is feeding raw air to the exaust ports at idle. If you pull the Blue plug off the Relief solenoid, the 02 reading should read richer.

If you were looking at a wideband you might see something like 14.3 afr at idle and when the blue plug is pulled it will drop down to approx 13.0 afr (no raw air being mixed in, just exaust fumes). That's at idle.

GTR

iTrader: (6)

Sorry to bring the thread back but

My suggestion would be:

If you see no apparent misfire, you secondary air injection is probably not working correctly. I would check that. If the air injection working correctly, the HC and CO readings will drop easily. I would just check if there is any air going to the catalyst

edmcguirk

Actually, I haven't fixed the problem yet.

I have not yet figured out what is preventing the ECU from going into closed loop. The car idles normally and WOT is normal but cruise is a little lean and a little weak with no hunting characteristics of closed loop operation.

I replacerd the plugs but I have not yet tried replaceing the primary injectors.

I have run with disconnected sensors to see if any of the sensors are bad. The car actually ran slightly better when I accidentally knocked the MAP hose off (but still no closed loop). I have gone as far as replacing the ECU and removing the ACV completely. No change.

It's not that the car is running too rich, it won't go into closed loop. It won't hunt around Stoic on the AFR gauge.

I suspect a vacuum leak because I can't think of anything else. The cold weather seems to make things worse and I am waiting for some evening warmer than 40 degrees (as of January 31st the inspection has expired so I do my testing after dark when the cops can't see the sticker)

This is annoying as hell.
ed

GTR

iTrader: (6)

If you still haven't found out why it doesnt enter closed loop, you might want to check

Engine Coolant Temperature - If coolant temp never reachs operating temperture, you will never reach closed loop.
TPS signal - If TPS signal is WOT, it will never enter closed loop. Make sure TPS voltage is in range.

Those are the two common problems.

I would check the signals from the PCM by backprobing the PCM connector.

edmcguirk

I took the easy road of disconnecting the temp sensor and later the TPS and then both.

The ECU is supposed to default to medium safe values when the sensor is missing.

I also adjusted the TPS until only one light lit.

So far nothing has helped.

ed

edmcguirk

This is driving me nuts. I have been working on this car weekends when the weather is good but I have done just about everything and I can't find what's wrong.

I'm not just trying to pass emissions, since this is a race car I need the car to run right. It will pass emissions easily when it is running right.

Let's recap:

The car first failed state inspection with a broken O2 sensor wire and I screwed up the wiring at first but that has been corrected. At that time the problem seemed intermittent but now it is constant.

I have installed a new O2 sensor and spark plugs then I swapped out the air flow meter, the ECU, and the fuel injectors.

I have not swapped the ignition coils because I don't have any spares on hand and spark problems should be worse at WOT than at cruise. Same with the fuel filter.

I have run with the MAP, water temp, and throttle position sensors connected and disconnected.

I have removed the Air Control valve and blocked it off just so any stray air injection into the exhaust won't confuse the AFR.

None of the changes have had any effect.

The car idles perfectly. The narrow band AFR shows correctly rich.
The car runs full throttle just fine. The AFR shows correctly rich.
Fuel cut works just fine. Dead lean on the AFR.

At cruise and mild throttle the AFR gauge shows the mixture just falls off the scale lean. The car runs a little rough and the power is a little off.

If I slowly increase the throttle, it stays lean until it suddenly snaps rich like normal WOT with normal power.

Today I noticed that if I run the car over 4000RPM at any throttle opening it snaps rich and smooths out with normal power. I think this means something important but I don't know what.

I can't think of anything to swap or test at this point. If anybody has any ideas, I'm open to suggestion.

If anybody in the north NJ area can help, I would greatly appreciate it.

ed

driftin8ez

well 4k is about when your secondary injectors kick in. So maybe your primaries are dirty or somehow causing the car to run lean. Maybe a broken/unplugged connector on one of them. just a thought. Im in central nj but i would be off little help to you.

GTR

iTrader: (6)

I would go back to my same suggestion, check the tps. get a graphing multimeter and scope it.
That would be my #1 concern if you already checked the ECT. If all is good, continue on to the MAP sensor. Make sure 5 volt ref is there and do a voltage drop on the ground circuit. Get a vaccum pump and using a graphing multimeter, scope the MAP.

As far as diconnecting sensor, you can not always depend on a PCM to default values when sensor is missing. A lot of PCM will not do that. Have you backprobed the sensors at the PCM? Since our cars have no scan data, you will have to do it the old fashion way.

HAILERS

*****If I slowly increase the throttle, it stays lean until it suddenly snaps rich like normal WOT with normal power. ********

That is not the secondary injectors coming online if you SLOWLY increasing the throttle. The secondary only come online under LOAD as sensed by the boost/pressure sensor. In other words you have to be ON IT for the secondaries to come online at approx 3500 rpm.

IF your secondaries come online under low/slow throttle, then you boost/pressre sensor is not working or you have a n/a one on a turbo car.

edmcguirk

I calibrated the TPS using the 2 lamp method but I did not measure it's full range.

Tha manual says that the TPS, water temp, and MAP will default to neutral values and the problem occured exactly the same with any and all of those sensors disconnected.

The manual even says that the TPS should be disconnected in order to test closed loop operation while standing still so closed loop should work without TPS.

ed

edmcguirk

I agree, the secondaries should not come online if I slowly increase the throttle but 4000RPM is highly suspicious because that is the point that a lot of things change.

Closed loop should stop at 4000RPM.
The secondaries should kick in at 4000RPM WOT.
There is usually a 4000RPM (approximately) dead spot.
I think a lot of the emissions functions turn off at 4000RPM.

However, like I said, I swapped the fuel injectors from another car and I have to think that the car would not run well at WOT in all gears and all RPM if all 4 injectors were not able to deliver fuel. (I'll look at the full circuit resistance to check wiring)

This car is unchanged since the last state inspection so I do not have the wrong parts on the car.

I will look into testing the MAP sensor because I believe the MAP default is WOT when disconnected. I have never heard of one failing though.

I would imagine that the results of a failed MAP could be very similar to what I am seeing. (You would think that the air flow sensor would have a little more authority)

ed

edmcguirk

I'm still not getting anywhere. I tested the MAP sensor. Various levels of vacuum change the voltage to what the manual expects.

Right now it's a little difficult to troubleshoot because when I split the intake manifold to replace the primary injectors the gasket tore and I have a small vacuum leak now. The idle is up around 1500 and it's a little rough.

I also just received a set of block off plates so that I can remove the Bypass Air Control valve. That should have no effect on my problem but I feel better to strip the BAC and the Air Bypass Valve from the intake air path.

Can anybody suggest someone in the north New Jersey area preferably near Wayne that I can take a street legal race car to? (I'm not looking for any help passing emissions, I just want to fix the problem)

ed