Mongo

First off, I'm sorry for the length of this thread and continuous resurgence of this topic. I figure I would start a new thread on this to chase down my problems.

I have changed the following on the car:

Intake Service:
-All Vacuum Lines
-TPS (Throttle/Idle Switch)
-Idle Control Valve
-Temp II Sensor
-Flappy Solenoid
-Fuel Injectors
-Knock Sensors

-Ignition Wires
-Ignition Coils
-Spark Plugs
-Distributor Caps
-Ignition Rotors
-Fuel Pressure Regulator
-Fuel Dampers
-Crankshaft Sensor
-MAF Sensor
-LH Brain
-O2 Sensor

I have verified that all of the sensors (and TPS) indicate accurate voltages according to the Factory Workshop Manual.

Now for the problem. After doing the dampers last weekend, my erratic idling issue when warmed up is not as bad as before, but very much a nuissance. I have a low temperature thermostat that stays at the 1/4 mark when driving up here in the Bay Area (it used to be 1/2 in San Diego) as well as takes a while to warm up. When the car enters stop and go or street traffic, the temperature will warm up to 1/2 (or a little above) and the fans will turn on.

Now here is where the problems begin. After driving the car and coming to a stop, if the car warms up the idle begins to drop from a steady 650 down to 400 and comes close to stalling. The ISV is attempting to catch it and regulate the idle. Revving the car shortly after brings the idle up, but once it settles it begins to get erratic and low again. Noting the temperature gauge drops from 1/2 to 1/4 after revving it.

I have tried 2 scenarios to duplicate this:

1) Cold Start the car and just let it idle to get up to 1/2 temperature. The problem DOES NOT occur when this happens. This is most likely because there is no wind directed into the radiator at road speeds and the temperature is a constant throughout the cooling system. The radiator fans will cool it only to an ambient temperature set by the fan switch.

2) Cold Start the car and drive it to operating temperature. When the temperature reaches 1/2 it idles erratic and low only when driving at low speeds (25 mph) and coming to a stop shortly after it hits 1/2. Obviously, since the car is in motion, there is air directed at a higher speed to the radiator. The water in the radiator is cooler than in the engine significantly. Forcing the cold coolant from the radiator into the the water pump and engine may be a significant variance in temperature being read by the Temp II sensor and the LH is assuming the engine suddenly has become 'cold.'

I have been reading on some sites that a colder thermostat can cause engine temperatures sensors to give inaccurate or inconsistent readings to the ECU and richen the fuel mixture assuming the engine is still cold. Clearly, if the idle drops, it's because of the richness in the mixture. Could a 75C thermostat, which is the coldest available 'splash' cold coolant causing the computer to overfuel my car?

Has anyone ever heard of this? I am looking forward to putting the original 83C thermostat back in the car soon. i don't know if this will cure my idle issues at operating temperature, but it really spoils the fun of driving it.

Again, everything else checks out and I hear no vacuum leaks, nor are there any old vacuum hoses/lines.

Thanks for everyone's patience on seeing these threads. I know I keep beating this like a dead horse, but I"m determined to solve this before I go to an independent mechanic.

danglerb

The job of the thermostat is to keep the temp stable so the brain can fine tune the fuel and spark. The lower thermostat is exposing a weakness in the cooling system which is allowing the temperature to rise above the control setting of the thermostat. Going back to stock is a smart step in sorting out your problem.

dr bob

In theory, yes a too-cold thermostat might cause it to run a little rich. But the amount of enrichment difference between 175º and 185º is not significant enough to cause the symptoms you are chasing. Plus, the reason that the mixture is a little richer when cold is because the engine needs to be a little richer when the engine is cold.

To make the point, consider that you can check your idea in the driveway with the car idling. The fans don't come on until radiator cold-side temps go up to what the fan controller considers normal plus a little, regardless of what the thermostat is doing so long as it's open by about 190º, which your 'cold' thermostat is according to your dash gauge. If the car is idling poorly at the time the fans start in the driveway, your problem is not going to get cured by a warmer thermostat.

Another broader consideration is that a slightly rich mixture, even a significantly rich mixure, is not going to cause the poor idle you describe. It takes a LOT of extra fuel to make the idle rough.

Remember that you can get an idea of mixture by looking at the oxy sensor with a decent DMM. In a perfect world, the terminal voltage from the sensor will be around 0.7 or so. If the sensor thinks the mixture is lean, the voltage there will drop, the LH controller will try to richen the mixture, voltage comes back up if it is enough correction, and the sensor will be seeing the correct mixture again. If everything is balanced the voltage wil be stable at about 0.7. If it's running lean, as it might from a failing MAF, an intake air leak (vacuum side), an un-metered air source like a crankcase or oil vent system leak, the voltage at the sensor will be low. If you have leaking injectors, bad/leaking FPR or dampers, vacuum hose off of the regulators/dampers, etc, causing a too-rich mixture, the sensor voltage will be high. The sensors are extremely narrow-banded and have very non-linear response, so a small change in measured oxygen makes a big change in sensor voltage. Bottom line: You can use a DMM to at least help narrow your search to things that make it rich vs things that make it lean.

Mongo

I understand what you mean Dr Bob. Now this just has me thinking if this is a relay now or not??? I have not experienced a no-start condition with this issue and I sure do hope it doesn't happen either. This counters the theory of a failing relay. It will also start right back up if it ever stalls due to low idle.

At least I know there were absolutely no metal shavings in the oil I changed and no TBF. Flexplate was also released about 5 months ago too.

dprantl

The lowering idle speed makes me think that somehow the engine is being deprived of air, fuel or is being physically bogged down? Idle speed should never drop below 675rpm. A vacuum leak will raise the RPM. The O2 sensor has a pretty large range of adjustment so it should not have problems keeping the mixture fluctuating around 14.7:1. O2 sensor problem? Idle control valve? Pinched air hose from ICV to throttle? TBF? Weird transmission problem causing drag on the crank? Some kind of wiring cross-talk fooling the LH? When the car runs like this, it would be good if you could peel the LH connector boot back and measure some inputs while it's running.

Dan
'91 928GT S/C 475hp/460lb.ft

John Speake

Even with a somewhat "cooler" thermostat, as long as water temp gets to 70C or so, the car will be running closed loop, and A/F ratio will be well regulated at 14.7:1.

I think your O2 loop is not working, for some reason. Also, as Dan says, your idle speed should be a steady 675rpm under almost conditions (slightly higher when water temp is below 25C or so).

Everything points towards LH not able to adapt O2 loop and hence idle loop.

What happens if you disconnect the O2 sensor ?

Mongo

Dan - if it was TBF, there would be metal shavings right? The car cranks slow when cold only because the oil in the pan is cold too. When warmed up, cranking is no problem! Flexplate release only had a difference of about 4mm movement on that pinch collar also. That was back in 2007. Nowadays the movement is not even apparent or is negligible when I release it.

Mongo

I haven't tried to disconnect it recently after changing the dampers. When I did do it 2-3 months ago, the car ran great! I even put a video up of it in the forums here assuming it was a bad oxygen sensor. Now with a new one... same sh*t different season. LOL

I am 100% sure it will run the same in Open Loop with the O2 sensor disconnected too. What would cause the LH not to read and adapt the oxygen sensor though??? I don't see any breaks in the wiring either. I have had this idle problem since September 2009.

dprantl

Leave the fuse cover off. Next time it starts to run bad, reach into the passenger footwell and disconnect the O2 sensor as fast as you can. If it immediately runs better, your problem is O2 sensor, LH or their respective wiring.

Dan
'91 928GT S/C 475hp/460lb.ft

dr bob

Andy--

Go back to simple stuff. The ISV is relatively new, right? So we don't really suspect it. It is a chore to change so let's put that out of the option list for now.

The throttle switch tells the LH to use the ISV to stabilize idle speed. IF the connection at the idle side of the switch is funky, the brain will not know it needs to use the ISV. Test the switch connections again. Looks like LH connector, terminal 2, should have continuity to ground when the pedal is up, goes to open circuit when the pedal is moved off of idle position. Simple ohm meter test.

The running mixture rich or lean is easy to test in a broad-brush way with the oxy sensor voltage as I mentioned in my last post. What's it read at full temp, at idle?

Mongo

I guess it wouldn't hurt to test the throttle and idle switches for a third time... I got the same readings the past 2 times that indicate it's functioning normally.


How do you test the O2 sensor?

dprantl

Disconnecting it will tell you a lot about how the loop is functioning. I was just thinking that with an ST2 it would be pretty easy to figure out your problem. Anyone with an ST2 close to Andy?

Dan
'91 928GT S/C 475hp/460lb.ft

Mongo

Believe me, nobody responded when I asked for an '88 LH locally. I don't think Bill Ball is on here anymore. He is the only one I know with an '88 car and a Sharktuner.

Mongo

Smog is right around the corner, like 2 weeks around the corner. Can I get this car to pass without the O2 sensor plugged in an running open loop???

dprantl

No way.

Dan
'91 928GT S/C 475hp/460lb.ft