K87ZZ4

Thanks, Cliff! I need to get some closed loop data, but it won't happen until later today.

K87ZZ4

I finally got some closed-loop data recorded today. What I found is that it's not staying in closed loop very long. It went into closed loop about 2 minutes after cold start (maybe too early?), but then went back into open loop 20 seconds later. For the first 10 mins, it was in open loop most of the time, switching to closed loop about 5-6 times for only about 10 seconds each time. Finally, about 20 minutes into the trip, it stayed in closed loop for about 2 minutes (while I was cruising on the freeway).

The O2 was typically between 400 and 600 mV. It got as high as 850 when I was the freeway. Cross-counts reset about 20 times during the trip, but were stable when it was running rich for a period of time, (seems to make sense, to me).

BLMs were typically between 108 and 124, but hit 128 and 130 at times, while I was on the freeway (about 15 mins).

From what little I know, it's supposed to stay in closed loop once it warms up. Is that right? Does it seem like the O2 sensor is working correctly? What else should I be looking at?

Thanks, so much, for the help!

K87ZZ4

BTW, I also recorded on the drive home. It started out much more normal, it was in open loop for about 4 minutes, running lean with BLM = 118, O2 between 350-450, no cross-counts. Then O2 started climbing to 700 and it went into closed loop. O2 was in the 750-850 range while I was cruising on the freeway. When I got stuck in stop-n-go traffic, the O2 came back down to 4-500 range and it went back into open loop with BLM 108 most of the time. It toggled in-n-out of closed loop a few times. It looks like once I took my foot off the throttle for 10 seconds, it goes back to open loop.

pcolt94

I am not the expert on the 87, but in general the O2 voltage should be constantly moving and swinging quickly. 100 to 800 is about the norm and I don’t think the 87 has heated O2 sensors. Are there any mods done to the exhaust system?

If the voltage swing of the O2s are not correct, the ECM will not be happy and take it out of closed loop.

K87ZZ4

Yes, most of the exhaust system was replaced a few years ago. It now has mini-cats and 90's mufflers behind them.

There is a single O2 sensor and it is not heated. IIRC, I believe that's the intent of the A.I.R. system on the 87, to help the O2 get up to temp quicker.

Cliff Harris

These are the conditions to go into closed loop mode:

Coolant temperature > 40.7°C = 104.4°F
Engine run time > 300 seconds (cold) if coolant temperature < 58.5°F
Engine run time > 206 seconds (warm) if coolant temperature > 58.5°F and < 158.5°F
Engine run time > 50 seconds (hot) if coolant temperature > 158.5°F
Oxygen sensor reading between 0.195 and 0.686 volts for > 10 seconds

My experience is that the engine run time and coolant temperature will meet those criteria before the O2 sensor gets hot enough to start working. The O2 sensor voltage will be low when it's cold and will rise as it warms up. When the ECM goes into closed loop mode it will vary the AFR continuously and check that the O2 sensor is following the AFR changes. It's like what the automation folks call a "bang bang" controller. It richens the mixture and when the O2 sensor shows a rich mixture it leans out the mixture. When the O2 sensor shows a lean mixture the ECM richens the mixture again. This happens continuously while the ECM is in closed loop mode. It's pretty fast -- it takes about 1 to 2 seconds to go through a rich-lean cycle.

If you look at the location of the stock O2 sensor, it is as close as possible to the exhaust manifold. My FSM mentions that the O2 sensor can get too cold at low RPM and cause the ECM to go into open loop mode. That's why it's important to use a heated O2 sensor if it is located farther away from the exhaust ports in the heads (like the usual place in a header collector).

K87ZZ4

I put a new O2 sensor in Friday, at the same time I did an oil & filter change, and the initial data didn't look good. I collected more data today on the drive to, and from, work (30 mins). The O2 is spending ALL it's time between 950 and 1,000 mV!! So, it's running rich ALL the time and I have virtually NO cross-counts.

The old O2 sensor was pretty coked up with black soot. My O2 sensor is installed on the drivers side, right after the manifold connects to the exhaust pipes.

I finally checked the resistance on the injectors... after figuring out how to release the plugs (push inward, toward the plenum, on the u-clip and pull up on the wires at the same time) a pretty consistent 17.6-17.7 ohms each.

I checked the fuel pressure... 42 PSI on prime, running at 40. Holds pressure for a few minutes, but was at 20 PSI after 15 minutes and 10-15 PSI after 30 mins.

I pulled a couple of plugs (#5 and #7) and found both to be pretty fouled with black soot. I cleaned 'em up because I didn't have any replacements on hand.

Wondering what this points to next...

VikingTrad3r

subd

my 87 is doing exact same thing.

K87ZZ4

Tonight, I checked for vacuum leaks by spraying carb cleaner around the intake and every vacuum hose I could find. No surges in engine speed.

I suspect, since both 5 and 7 were equally fouled, that the rest are fouled, too.

I'm sort of at a loss at this point...

This is still nagging me, in the back of my mind... when the car failed emissions (high Hydrocarbons (of 450; limit 220)) at idle, a few years ago, a shop said a component of the A.I.R. system wasn't working right (the diverter or switching valve, I think), but also said a replacement part was not available. Even though it failed emissions, at idle, it idled and ran fine. So, if the valve isn't working right and there IS air to the ports in closed loop, that means it's potentially injecting air into the exhaust, right? How would air injected into the exhaust, downstream of the O2 sensor, cause a high hydrocarbon reading at idle? If this is the true source of the problem, it could have been just getting worse over time.