87 TR BAD IDLE WHEN COLD

(All -- Ken and I have already exchanged some PMs, but I need to post some jpegs so I’m moving the conversation here.)

Ken -- I’ve never heard of a F113B040 TR engine before (I only know of the F113B and F113B046 types). For your F113B040:

Does it have cats?

Does it have O2 sensors?

(And anyone with info on the F113B040 engine configuration please speak up).

Can you please verify the print number of the TR OM that you have access to -- it will be in the form of XXX/YY (usually on the inside of the front or back cover).

Almost unfortunately, when the "E" part (the Electromagnetic pressure actuator = the EM valve) of a KE-Jet isn't working, it's still possible to tweak the fuel mixture with the mechanical adjustment screw so that the warm-running is quasi-OK (i.e., it runs like a regular K-Jet when warm, and it runs OK warm because it was tweaked when warm), but if the EM valve isn't working, there is no enrichment added when cold. Consequently, I’d start with confirming/denying the resistance measurements described in these two jpegs (the jpegs show the 25-position injection ECU connector used on a US version – your injection ECU connectors might have fewer pins installed, but you should still be able to figure out who’s who).

Let me know if you pass or fail these resistance tests.
Image Unavailable, Please Login
Image Unavailable, Please Login

 

This is good -- F113B046 is a Swiss (CH) version KE-Jetronic with Lambda

This is actually a US version TR OM -- but it will work fine for helping us to communicate. If you go to the Owners Site you can download the "partial" Swiss OMs 467/87 and 519/88. If your engine has the "pumpless" air injection set-up, 467/87 would apply, and, if you have the "with air pump" air injection set-up, 519/88 corresponds (there are good figures in each for you to compare with). They don't have English text, but the good thing is that both these manuals include an injection system electrical schematic so I'd encourage you to go thru the hassle of getting a copy of the one that actually matches your TR (and with a similar US version TR OM in hand to look at, you can generally figure things out OK).

Understood (and fits with being F113B046). As I suggested earlier, you might try unplugging the single wire coming from each O2 sensor during warm idle and see what happens. Most likely nothing (i.e., it will go "open-loop" and still run fine), but, if it runs poorly, this would be another clue in a different direction.

Sorry, I don't understand this. For each test, you should have 2 separate resistance values -- one for each loose injection ecu wire harness connector when unplugged. (These are the injection ECU connectors over the RH rear wheel -- not the ignition ecu connector over the LH rear wheel.) Please explain further or remeasure.

 

Sorry, but I have to press you on those resistance measurements one more time (but my suspicion is everything is OK):

The thermister measurements seem fine if your meter was actually measuring KOhms rather than Ohms -- is this possible?

The EPA coil measurements would also be OK if it was really .02 KOhms -- i.e., 20 Ohms. If it really was 0.2 Ohms or 0.2 KOhms (200 Ohms), either would be bad, but I think it highly unlikely that both would be so equally bad.

With regard to what happens/happened when you unplugged the red wire at the water thermoswitch during cold-running, I would need to know which Swiss type you have -- "pumpless" air injection or "with air pump" air injection.

 

Ken -- I'm not asking you to unplug the O2 sensors during cold-running -- do it during WARM-RUNNING! In fact, during cold-running the signals from the O2 sensors are completely ignored by the injection ECUs -- so your result of "no effect" is exactly what should happen during cold-running if you unplug them (on a good cold-running TR or a bad cold-running TR).

If during warm-running, you unplug the O2 sensors, and it runs poorly -- this would indicate that your mixture(s) is(are) improperly adjusted -- and that would have to be corrected before anything else was done or evaluated. Proper cold-running is absolutely dependent on the manual mixture adjustments being correct (which are done warm with the O2 sensors unplugged) since, as I stated, the system has to run open-loop when cold because it can't use the O2 sensor signal outputs and the EPAs to compensate for a bad mixture adjustment (like it can when it's warm).

TRs have a small half-cylinder sheetmetal shield that snaps onto the O2 sensor body for protection -- so that might explain the visual difference you've noted.

If you measure the voltage on the single wire from each O2 sensor (relative to a good ground like the engine block or cylinder head), this can also give a clue about things:

1. When cold and plugged in at idle, voltage on single O2 wire = meaningless

2. When cold and unplugged at idle, voltage on single O2 wire = meaningless

3. When warm and plugged in at idle, voltage on single wire = should be banging around between ~0.1VDC and ~1.0 VDC -- this behavior shows the EPA actively adding and subtracting fuel, trying to hold the O2 sensor output at 0.5VDC (which it really can't do without overshooting in each direction).

4. When warm and unplugged at idle, voltage on single wire = should be a constant 0.7~0.9VDC. If it went to something like a constant 0.1V or a constant 1.0V this would also be a clue that the mixture is not properly adjusted.

 

Well the good news is that:

1. You’ve identified a clear problem that will definitely cause bad cold-running -- i.e., if warm-running idle with the O2 sensor(s) unplugged is bad, cold-running idle can’t possibly work well, and

2. You’ve learned how to measure the O2 sensor output (and read the tea leaves) during both closed-loop warm-running idle (plugged in) and open-loop warm-running idle (unplugged) -- IMO every FChatter with a K-Jet with Lambda or KE-Jet with Lambda F model should learn how to do this -- even if they don’t DIY.

The mixture is solely controlled by adjusting the mixture screws. The air bypass and throttle stop screws are used to control how much air is entering the engine, and then the mixture screw is adjusted at idle to add the amount of fuel that gives a 1.0~1.2% CO concentration in the exhaust entering the precats. The desired target is:

A. The engine runs at warm idle at 1000 RPM with the O2 sensors unplugged, with the mixture screws adjusted to create the aforementioned CO level, and there is equal pressures in the intake manifolds, then

B. when the air bypass screws are fully closed, the idle drops to ~700 RPM with still equal pressures in the intake manifolds (the mixture is not measured nor adjusted during this test), then

C. the air bypass screw are reopened (usually about ½ a turn) to get back to condition A, then,

D. the O2 sensors are plugged back in, confirm the O2 output voltages are swinging as expected, and that the %CO is still OK (or lower).

If you are confirming that the “bad” O2 sensor measures ~0.1V during warm idle when unplugged and the engine runs poorly, but when plugged in, the engine now runs well and the O2 sensor output is still 0.1V -- that would be a sign that the O2 sensor itself is bad. If you search the Internet on “testing O2 sensor” you should get information on other methods how to test just the O2 sensor itself. If you’ve got 30K+ miles on your TR, I think you could get a new set anyway with no regrets (even if they/it aren’t the problem).

However, even if your O2 sensor is “bad”, and you replace it, you will still need to retweak your mixture (and probably check the airflows) after it is replaced (i.e., warm-idle with the O2 sensor unplugged must run well).

Give a shout if I can help, but you’re getting to the point where I would recommend that you get and read a copy of Section D of the TR WSM.

PS You should restore your air injection air pump operation. On TR (with air pumps) there is an EM clutch in the air pump pulley assembly that disconnects the air pump drive during warm-running so you get virtually no performance benefit during warm-running by removing the air pump belt (and, unlike 308, even if the air pump freezes up, there is no risk to the engine). However, during cold-running, your cats will suffer more HC fouling if the air injection isn’t working properly -- JMO.
Image Unavailable, Please Login
Image Unavailable, Please Login

 

Glad to hear that you're making some "progress", but 1/2 turn of the mixture screw is a huge amount -- "way too lean" to "way too rich" is something like 1/8 of a turn CW IME.

Are you now confirming different O2 sensor voltage results for the test conditions of #3 and #4 as described in post #12?

 

Yes...clockwise richens....and make sure you plug the access holes to the mixture adjustment screws.

 

Actually, if your old O2 sensor was always (wrongly) reporting ~0.1V (lean) to the ECU during warm-running, this would cause the EHA to add a lot of extra fuel so the "compensation" would have been to have the manual adjustment set to a very lean condition -- i.e., if warm-running was OK-ish with the old O2 sensor plugged in, it does make sense that adding richness was the right direction to go when the O2 sensor is unplugged or during cold-running. However, although an IC engine runs OKish when set very rich, you will be putting a lot of unburnt fuel into the cats and they will run at a higher temperature (which is not good).