I know that the knock sensor could trigger the ECU to retard ignition timing. What else could cause the ECU to retard ignition timing, specifically under high load?

Basically I have noticed that the ECU would retard ignition timing when the engine is under high load. For example when I am accelerating quickly, going up a hill, or accelerating in a higher gear. Also the ignition would only retard timing up to 3000rpm. After 3000rpm the timing would advance as usual and the car will keep pulling with full power.

I have also noticed that when I am accelerating lightly or cruising around, the ECU would advance ignition as normal and everything drives fine.

There are no check engine light, and I am running 94octane fuel. The engine is not running lean and the long term and short term fuel trim is all normal close to zero.
AF is normal around 14.7 in closed loop.

There are also no knocking noise, at least not that I could hear it.

Also at IDLE and neutral the engine would rev fine and the ignition timing would advance as normal, usually around 30~40 degree. This lead me to believe that the ignition system itself is working fine.

Any one here have any suggestions on what could be the problem?

 

Also does anyone know if the knock sensor only work in closed-loop?

 

A modified intake could cause the ECU to retard the timing under load. If you made any changes to the intake try putting it back stock or restricting it and see if it makes a difference.

 

How are you getting all this data from the car?

I don't know if you drive a TL or TL-S, but a lot of TL-S owners, including me, are complaing about pinging especially above 3k rpms.

Check out this thread, there is a lot of good info there about timing:

https://acurazine.com/forums/showthr...ighlight=knock

 

I have a TL and all my info are from OBDII scan too and on board AEM wideband gauge.

Can you explain why a modified intake cause this?
I have a supercharger but is currently not installed. So we just attached a tube from the throttle body with an air filter on the other side. The MAP sensor is still correctly attached via the supercharger elbow.

 

Thanks for the heads up. I am actually having problem under 3k rpm. I tried to get my friend to listen and we both couldn't hear any knock.

Mine is not a TL-S, but I do have TL-S head and cam.

 

until the SC is installed, just run 91 octane. You have 11-1 compression now with the s heads?
In street gas, there are additives to retard combustion- higher compression engines require those.
I am going to ~guess~ that 94 has more of them
If you can get 91 locally- try that
Have you done a cleaning of the intake system since the heads changed
Have you tried resetting the ECU? pull the clock fuse in passenegr footwell, or disconnect the battery NEG cable for a few minutes

 

When the SC is installed- you will NEED that 94!!!

 

I have lower compression piston in so it should be around 10.5cr. I will try the 91 gas. However I don't believe its the gas because the ECU is only pulling timing based on load.

 

I dont know how many sensors are involved
but it has to be taking info from both sides of the cat- the throttle position- the inlet air temp- and a few others at minimum

 

Today I did some testing. All testing were done by Accelerating from a dead stop in 2nd gear to produce a high load.

First test, with knock sensor connected. Acceleration was weak and timing advance was stuck at 2degree. There are no CEL for the whole time.

Second test, disconnected the knock sensor. Acceleration was much stronger (but obviously not as strong as starting in first gear). Timing advance was in the 10 to 20 range and then slowly go up to the 30s as load decreases, which looks normal. Problem is, after 5 second of accelerating the CEL would come on indicating a knock sensor malfunction. (obviously because I disconnected the knock sensor on purpose.)

What do you guys think? A faulty knock sensor?

I checked all the other sensors, MAP, IAT, Coolant, TPS, and they all seem to be working fine.

 

Did a few more test and datalogged the ignition timings.



The first two row is with the knock sensor connected. The run is done right after a ECU reset, VSA OFF, and a 2nd gear launch.

The next two row is with the knock sensor disconnected. Again the run is done right after a ECU reset, VSA OFF, and 2nd gear launch.

Last two row is with the knock sensor connected, but this time the runs are done without reseting the ECU.

Seems like with the knock sensor connected, the ECU is pulling a lot of timing. With the knock sensor disconnected, the ECU advance timing as normal but eventually the check engine light will come on and retard the timing again.

 

lookinco, do you still have the ACM hooked up or anything else besides the wideband?

A few members that live in the higher mountain ranges complain of a loss of power, but when the enter VTEC, the power is all there. I would think that the ACM or vtec point is down to 3000 RPMS just as the comptech ACM makes it. That should be why the timing returns to normal. But in a period when you are not in VTEC, I would think that the same sensor that causes the loss of power in mountain conditions would cause the same problem you are having.

 

I don't have the ACM. As a matter of fact, the supercharger is not even installed right now.

So right now everything is pretty much all stock except for the swapped motor. Stock ECU, fuel pump, injectors, and all stock sensors.

 

Maybe it might be some sort of incompatibility with the stock J32 motor ECU and that swapped motor, if I remember correctly, isn't it the one with the MDX short-block, cls heads and a few other different parts? Or did you change it and to what?

Also aren't there multiple stages of VTEC? This could be the relation with the 3K RPMS.

 

Right now i am using a TL block with CL-S piston, RL crank and rod, and TL-S head and cam.

There shouldn't be any compatibility issues as others have done it without problem. My last engine was the same setup and didn't have this problem as well.

Also I have noticed that the retarded timing is based on load, not rpm.

I am suspecting that the knock sensor was damaged when we worked on the motor which caused the ECU to pull timing.

 

Damaging a sensor with all of that work would be easy to do. Why did you swap out your motors?