Bizarre ignition timing problem

[JPerez]

I'd originally started into this in the MeMyMod section, but perhaps I can spark some interest (pun intended) in here.

I have recently acquired an '87 Celica GT. ST-162, 3S-FE, manual tranny.

The engine knocks noticeably under normal driving.

I've got the timing set dead-on at 10 BTDC warm, at idle, with the diag jumper in.

Tonight, I actually went in and watched the ignition advance with my timing light as I ran the RPM up and down. At idle, with the diag jumper out, the ignition is around 15 BTDC. Then, I reach over and slowly open up the throttle (reeeeal slowly):

The timing stays where it is until about 1,400 RPM (or about 5 degrees of throttle opening), and then it suddenly jumps to damn near full advance (it has a few degrees left for once it gets up around 3,000 RPM, but I mean it when I say it jumps straight from 15 BTDC to about 30 BTDC.) At that point of course I start to hear the knocking.


Some notes:

1- The jump in advance is exactly co-incident with the throttle position switch moving out of the IDL position. (this engine doesn't have a true throttle position sensor, just two switches. It's either fully closed, wide open, or somewhere in-between.) If I manually open the IDL contact, it advances immediately. If I hold the IDL contact closed, it never advances.

2- There is no MAP sensor on the '87 3S-FE. All it knows is RPM, air volume, and temperature.

3- I have replaced the following:
- Coil
- Rotor
- Cap & wires
- Spark plugs
- Air flow meter

4- The engine is hard to start when cold. (cold = it hasn't been running within the past couple of hours, regardless of outside temperature)

5- The vehicle passed an Ohio emissions test last month.

At this point I'm really stumped, and I'm getting close to buying a new ECU.

[88 LE]

Originally posted by JPerez
4- The engine is hard to start when cold. (cold = it hasn't been running within the past couple of hours, regardless of outside temperature)

Check the cold start timing switch and cold start injector.

[JPerez]

pgmomni raises a few good observations, but unfortunately been there, done that-

I've tried running 93 R+M octane gasoline with no change. I've also run through a couple tankfulls from different suppliers.

The engine does have a timing belt. It was changed in 1999 at 75,000 miles (about 15,000 miles ago).

I removed the valve cover and timing cover to verify that the cams are correctly aligned and belt correctly tensioned, and I also verified that the crank pulley is giving a true reading by sticking a dowel into the #1 spark plug hole at TDC and rocking the crankshaft back and forth. Mechanically, everything seems to be fine.

I'm convinced that whatever is happening is purely electronic in nature. The real clue for me was when I found that I could make the ignition jump from 15 BTDC straight to 30 by simply opening the IDL contact at the throttle position switch (the equivalent of opening the throttle just a few degrees). That just can't be right, can it?


Oh, and 88LE, I actually have not checked the cold start injector yet. It may be the cause of my starting problem, but if so, then I think it's probably unrelated to this ignition thing which is my primary concern. From what I can tell by the schematics, the cold start injector isn't connected to the ECU or the ignition system in any way. I'll check it out tonight.

[automatic]

So, did you ever figure it out?
Are you still lurkin' around here?
I have had a similar problem lately w/my '92 Corolla; will advance a large amount only when @ idle, creates a miss-like stumble just momentarily.
Next is to replace entire distributor w/one from another vehicle; have replaced MAP sensor plugs,wires cleaned contacts on TPS, cleaned out PCV valve, drained some gas from the bottom of the tank,and have adjusted everything up and down to try and find anything that will trigger the symptom.

[JPerez]

I have made some progess, but I'm still not quite satisfied.

The cold start problem was easy. The thermo-time switch in the coolant manifold wasn't working (always open). Rather than spend $110 for a new one, I simply bypassed it by splicing into the harness and grounding the green wire which is the low side of the injector. Since the high side of the injector is powered by the starting circuit, the injector now squirts whenever the starter is cranking. Doesn't seem to have caused any problems. This trick probably won't work on 3S-GE and 3S-GTE engines whereby the ECU senses the green wire (it'll think the engine is always cold), but on the 3S-FE this terminal is used soley to fire the injector, and the ECU only senses cranking voltage.

As to the ignition, I ended up replacing the ECU with one I got off ebay. The new ECU did change the character of the ignition advance somewhat, although at 10 BTDC is still pinged a bit. I found that with the timing set to 7 BTDC the engine is driveable, and pings only very occasionally.

Have I mentioned how much I miss carbeurators and mechanical advance distributors?


One thing I discovered along the way, which might be helpful to others, is that it is impossible to accurately test an AFM with just an ohmmeter. The resistance jumps all over the place.

What you have to do is test the AFM with a voltmeter while it's hooked up to the car. Turn the key on (engine not running), and measure the voltage between terminal VS and ground. A working AFM will produce a smooth, linear transition between 0.5VDC and 4.5VDC (or thereabouts) as you open and close the flapper by hand. It helps to have the AFM uncoupled from the air box and hose, with just the wiring harness plugged in. It's also difficult to get to terminal VS without cutting the caulk and opening the AFM, since the body of the connector is sealed by a rubber boot. In my case, I already had it open.


Now, we move on to the transmission, which grinds into second when warm. Anybody ever try using GL-5 instead of ATF on an S53 5 speed?

[automatic]

So the ECU solved the rapid ignition advance problem then?

I have had very good luck with GM Synchromesh fluid, available at Chevy dealers, doesn't take much(half a bottle) to really help out with the shifts.Costs around $14 a quart.