This isn’t a problem, just something I’m curious about. On all the car engines I’ve tuned prior to my 4AFE Corolla, the timing mark (when using a timing light) stays at constant position (say 10 deg BTC) at idle. It varies maybe +/- 0.5 degree, no more than that. But on my Corolla it jumps around from 5 to 15 BTDC at idle. On the Corolla there’s a special test mode to set the ignition timing. In that test mode it doesn’t jump around. It acts like my other cars in that test mode. But after returning the computer to normal mode, then the timing jumps around ± 5 degrees at least. Just wondering what motivated Toyota’s engineers to add that complication to the computer’s timing algorithm. Is this a common thing observed on other computerized ignition system cars too?
What year is your Corolla?
How many miles are on it?
How’s the maintenance been… done, or put off to some future time?
Timing is usually pretty stable once the engine’s warmed up… unless there’s lots of mileage or an anomaly. I know from your posts that you’re conversant with the basics, so I’ll skip the dissertation. But if it has a variable valve system, poor/erratic oil pressure can be a variable that can cause poor idle. The way to verify it is to unplug the solenoid that operates the VVTI system and see if the idle smooths out. If it does, you’re well on your way to solving the problem. The next step would be to test the oil pressure at idle with a test gage.
92 Corolla/4afe engine. No VVT. It’s not a malfunction, it’s supposed to do that, according to the factory service manual. It’s done it since day 1. I’m just curious why the engineers designed it that way. Improved MPG? Emissions? Power? Or maybe it isn’t really doing that, some kind of timing light testing optical illusion?
I’m stumped. That’s truly a new one on me.
You and I aren’t the only ones stumped about this.
I’ve read that the same thing happens on a GM TBI 350. If you don’t unhook a brown wire under the glovebox, disabling “electronic spark control”, you cannot set the timing correctly. Sounds similar to what you’re describing from what I’ve read, although I’ve never had to set the timing on either (or any) engine.
All I can think of is the computer is adjusting the timing similar to what the centrifugal advance used to do unless you disconnected it when timing. But my timing light broke years ago.
The timing is used to maintain a steady idle. The engine reacts to timing faster than it does to idle air changes and it allows Toyota, and others, to set a lower stable idle speed for emissions and fuel economy.
That makes sense, but what change in the engine at idle is the engine computer sensing to make it change the ignition timing from 5 btdc on one firing the 15 btdc the next? I can see how it might change the timing slowly over time, for example to advance the timing if the rpms start to slow. But maintain idle rpm by changing the timing by 10 degrees from one firing to the next? I just don’t see what the engine computer could be sensing to make it do that. In the test mode where the timing is rock-steady, the idle rpm is still good. I don’t see the idle quality being any better in the normal mode than the test mode. Maybe it is done for emissions reason only, when idle emissions was the key thing back then. When the car was new, all Calif required was an idle emissions tailpipe test. Now-a-days they measure the tailpipe emissions with the car moving on a treadmill at 15 and 25 mph, so maybe these days that timing trick doesn’t work.
Since the crank and cam position sensors read every tooth plus the spot where there is no tooth, the ECU knows exactly where the engine is at all times. The ECU loop time is likely in the 1 msec range (or better) for read and respond. Now consider a Chevy crank sensor with 53 teeth - so every 6.8 degrees plus TDC. Add in the cam sensor and you know where in the stroke each piston is to within 6.8 degrees. Actually better, since you know the rpm as well, and can estimate within that 6.8 degrees. At 5000 rpm, engine fires a cylinder every 3 msecs. So you have a very accurate picture of where every piston is in the 4 strokes and plenty of time to control it.
Now consider sequential fuel injection. The precise amount of fuel is injected into each intake port a the precise time for each cylinder. The timing of that injection is very importance to horsepower and emissions. Sequential injection made more HP than batch fire injection. Now we have direct injection. The timing is even MORE important and the results are again more HP and fewer emissions. I used injection to show how fast the ECU needs to be (and IS) to make it clear that the ignition timing is pretty easily controlled to that level of precision as well.
Each spark advance can be adjusted in real time, every stroke, to maintain a low, smooth idle. Since emissions are regulated not as a percentage of the exhaust but the total amount emitted, This favors small engines for obvious reasons as well as explaining Stop/Start systems. A low idle makes meeting the standards easier even with a large engine. Also stopping the engine at the “best” point to re-start using Stop/Start systems uses that info, too.
As for state emissions checks, they don’t have the necessary equipment to measure all the things required by the regs. They are just taking a snapshot to prove the car is still in compliance. Some states don’t even use tailpipe sniffers or a roll test, they just look for the CEL.
hmmm … The 1992 4afe engine uses batch firing for the injectors, all 4 are fired by one signal. And the crank/cam sensors are inside the distributor, so the resolution steps probably aren’t quite as precise. Are you saying MM that most cars w/electronic ignition systems these days do this same thing? i.e. if you examine the idle ignition timing w/a timing light on the harmonic balancer, it won’t be rock steady at idle, instead it jumps around +/- 5 degrees? Thanks for the description of how it works, but I still don’t quite understand all that’s involved I guess.
State emissions inspection programs for those states (and in some areas counties) required to have them must be and are approved by the EPA on a case-by-case (state-by-state or, in some cases county-by-county) basis. When NH went to OBDII verification rather than combustion analyzers (sniffers) some years ago, we were required to create and get approved by the EPA a replacement plan.
For the record, I learned about the county-by-county detail when I moved to my current town and no longer had to pass emissions testing while the county I moved from still did. I was happy. 
It isn’t a matter of insufficient equipment. The inspection stations don’t test based upon the technical requirements of the Clean Air Act as modified. They test to the EPA approved program mandated by their respective state.
I don’t know if all managed engines do this. I learned of this technique from a Delphi friend that did engine management software. It was just a normal tool in their toolbox to control the engine. If similar tools were not in every tuner’s toolbox, I’d be very surprised. Your description of the timing bouncing around on your older engine matched that “tool” to a T. I can’t think of any other reason to flail the timing around at idle except to steady the engine. Heck we USED to see that as a sign the centrifugal advance springs were broken!
Consider if the engine slows just a bit, more timing can be added to boost the idle speed - we did that manually back in the day, right? Backing off the timing slows the idle speed. That’ why you adjusted dwell first, timing second and idle last on a tune-up. Even if Toyota did that for the bulk timing rather than individual cylinders, it would still be an effective tool for balancing idle speed.
I get the sense this isn’t a common thing, at least not in most newer cars. B/c it would be obvious to anyone pointing a timing light at the harmonic balancer; folks here would report they’d noticed it happening on their own cars. I’m guessing the main reason isn’t to steady the idle, but for idle emissions reasons. Like I say, back when the Corolla was new, the only tailpipe test required was at idle. If the manufacturer could somehow squeak by at idle by hook or by crook, they’d be home free. Later on the Calif Air Resources Board must have figured out the trick, and as a countermeasure required cars like mine using this trick to undergo treadmill tailpipe testing at 15 and 25 mph instead. Speculating is all of course.