wstrohm

In this old thread, I described our '94's high/normal idle problem and found that others had experienced the same issue. The problem has now existed for a couple of years, and I would like to chase it down if possible.

My next shot at it will be to remove the AWS valve and plug the AWS manifold ports, on the possibility that the valve is randomly leaking and sealing when the engine is running. But there is a problem... simply removing the valve and disconnecting the connector will cause a fault code #38.

Looking at the wiring diagrams, it appears that the ECU "looks at" the presence of the +12 volts at the "cold" side of the solenoid coil (terminal 4P of the ECU) when the coil is not activated. If the +12 volts is present when ignition is ON, the ECU decides the coil is intact. If the voltage is missing, the ECU decides the coil is open, and sets a fault code.

One way to prevent the fault code setting would be to connect a resistor across the wiring harness terminals of the AWS connector, letting the connector float. Unfortunately, to simulate the coil resistance exactly requires a resistance value of 11 ohms or so. Then when the ECU grounds its terminal 4P to activate the AWS, over 1 amp will flow in the resistor, and it would have to dissipate 12-14 watts. The physical size of a 15-watt resistor is too large for a neat installation into the wiring harness, so I would like to use a much higher value, like maybe 1 Kohm. That would allow only 12 - 14 milliamps to flow when ECU terminal 4P is sinking current, or about 200 mW maximum resistor dissipation. Thus, a 1/4 watt, 1Kohm resistor would fit on the connector's harness end easily.

But what if the monitor point for the +12 volts, internal to the ECU, is a voltage divider off the +12 volts, i.e. like a TTL "1" value of +4 volts? Then the new 1Kohm resistor might drop the monitored voltage so low that the ECU would still store a Code #38 fault. (I guess I could just try various values and see what happens to the 4P voltage, keeping both the resistor value and the voltage as high as possible.)

Does anyone know what the circuits that monitor external 12 volts look like inside the stock ECU, i.e. what is the highest resistor value I could use (to minimize resistor power dissipation) and still not trigger a fault?

wstrohm

OK, I have now removed and blocked off the AWS system, and put a new OEM gasket under the blockoff plate (which I made from 1/8" thick aluminum plate). I removed the AWS air intake tube and blind-capped its 5/8" supply nipple on the big air intake. I also bridged the two terminals in the AWS wiring harness connector with a 2.2 kohm, 1/8 watt carbon resistor. I wrapped the connector with vinyl electrical tape, capturing the resistor and holding it in place. There is no stored AWS fault in the PCME after a test drive, which was the point of adding that resistor.

I removed, cleaned, and tested the IAC valve with my variable voltage power supply. It was quite dirty, and took a number of inundations with WD-40 before coming clean. Results of power supply testing were that the IAC valve began to move at 3.0 volts, at which voltage the coil used 240 milliamps. It continued to open until 9.8 volts (770 mA). Going up to 10 volts, I heard a "click" from the valve and current was 800 mA at 10.1 volts. Movement was smooth (this was after cleaning) and repeatable... no sticky spots. I used a new OEM gasket when re-installing the IAC valve.

I had previously implemented this circuit at the PCME to allow disconnection of the EGR control, and had capped off the vacuum hose connection to the EGR valve. This time I disconnected the hose from the rat's nest to the UIM, and capped the UIM nipple. There is no EGR fault reported when codes are pulled from the PCME.

Using my MityVac, I tested the various hoses I had disconnected when removing the UIM. All held vacuum. Could not find any leaks I could pin on the rat's nest or the secondary air control.

I also replaced the UIM to LIM manifold gasket, the throttle body to UIM gasket, and the intake air tube to throttle body O-ring, all with OEM parts.

After all the above, the engine acts exactly the same as before the changes. Still idles at around 1500 rpm when warm... sometimes. And sometimes it idles at 750 rpm or so, which is where I set it with "TEN" jumpered to "GND" in the Diagnostic connector last time I worked on the car.

Power is very good at 98000 miles (I don't have a boost gauge), and engine is very smooth off-idle all the way up the rpm range, although I have the infamous 3000 rpm stumble on gentle acceleration (but not under full acceleration). The engine runs very rich when cold (i.e. smelly), but that goes away when warm. Gas mileage is about 13-14 in city, 21-23 on highway.

So once again I am stuck trying to find the source of the high idle problem. Any suggestions?

7dust

iTrader: (14)

^^^ Is your fast idle cam fully disengaging?

Pust it open with your finger after the car is warmed up and see if it drops the idle any - it should have no effect.

mad_7tist

/\ wax rod issue, fast idle cam, tps settings, vac leak

wstrohm

7dust and mad 7tist,

Thanks for your replies. I'm sure the tps settings are correct, since I measured all the voltages and they have stayed in spec and have been steady. I have found the operation of the wax-driven rod, which bears on that spring-loaded screw and moves the throttle linkage off the fast idle cam, confusing. For one thing, I cannot figure out what it is that should be pushed manually to check that the linkage has separated from the cam. I did adjust the linkage when the throttle body was off the car, per the manual. But I cannot see the operation when it is mounted on the UIM. So maybe there is a problem there. The difficulty with assuming that, is that it doesn't explain how or why the idle would vary from stoplight to stoplight. As I wrote in my previous thread (linked above), sometimes the idle is at 750 rpm or so, and sometimes it is at 1500 rpm, with the engine fully warm. It's the cycling for which I can find no reasonable cause.

Six Rotors

iTrader: (17)

Sometimes the coolant passage around the thermowax gets crapped up and the coolant flow is variable,hence the variability in idle.This may also occur if there is air in the cooling system.
It is worth removing the thermowax and cleaning out the coolant passages.If you do it,try not to destroy the rubber O ring seal.

wstrohm

Six Rotors,

How does the thermowax rod unit come off the throttle body, and does the TB have to be removed first? Can't find anything in the '94 factory manual describing testing or replacing it.

[Edit: Never mind, the parts catalog shows that the thermowax assembly just bolts to the throttle body. Rats, that means I have to remove the TB to remove the thermowax and clean it. Just had TB off the manifold yesterday!]

7dust

iTrader: (14)

^^^ You can see where the thermowax button pushes the cam down...

Push the cam all the way down with your finger when the car is warm and there should be no change in the idle.

My car does the same stupid ****. I put a piece of rubber under the screw and effectively disabled mine until I can replace the thermowax. The car needs a little coaxing when cold, but the idle is more predictable until I can make the repair.

wstrohm

7dust,

OK, I saw that... will try it, thanks.

wstrohm

Well, pushing down the throttle flange while engine is warm does not change the idle speed, so I guess the thermowax unit is not the problem. Stuck again.