Engine knock

[mikec03]

I can hear my '95 subaru [2.2] knock, especially when the auto trans shifts from first to 2nd. So far, I put in premium gas and, not surprisingly, it stopped. I have two questions for the experts on this forum:

 

1. Once the engine knocks, how long does the knock sensor override the ecu settings--5 sec or 5 min or what?

 

2. Why would it start knocking? It doesn't burn oil so there shouldn't be deposits in the cylinder heads.

[ShawnW]

If it goes away with load or fuel its not considered KNOCK is more likely PING.

Could indicate you need to do a tuneup, your knock sensor may be bad, or your gas is junk. Might even have excessive water in the fuel tank.

[mikec03]

Of course, I meant a ping. The gas is good, the car is tuned up, and the knock sensor is apparently working by reducing the engine power [retarding spark].

 

Does anyone know how long the knock sensor is in force once it activates? It has to have some sort of delay or it would be continuously clicking on/off.

 

And why am I getting pinging?

[Rooster2]

The knock sensor is electronic, so it's input to the ECU to retard or advance engine timing is super quick within milliseconds. That is way faster then the seconds or minutes that you are thinking about.

 

It has always been my understanding that ping is caused by carbon engine deposits on the face of the piston or on the head. Short of pulling the heads to look for the deposits, you may want to try a gas additive. A trip to a car parts store should yield you a product to add to your gas tank. I would ask the counter people for advise on which product to buy. It might help.

[mikec03]

It's just hard to believe that in this day of unleaded gasoline that there would be significant carbon deposits in the head! It's even harder [for me] to believe that gasoline additives would remove carbon deposits anyway.

 

Any other thoughts? Surely I can't be the only owner with a 15 yr old Subaru that is getting pinging?

[OB99W]

[...]It doesn't burn oil so there shouldn't be deposits in the cylinder heads.

 

It's just hard to believe that in this day of unleaded gasoline that there would be significant carbon deposits in the head![...]

Carbon deposits don't require oil burning to form, and removal of tetraethyl lead from gas didn't change the situation, either. Gasoline, being a mixture of various hydrocarbons (and additives), by itself is a source of lots of carbon.

 

Do a Google search on "engine carbon removal" if you'd like some info. One of the products often suggested at USMB is Sea Foam, but it's not the only one.

[abog]

If you were to pull the heads and remove the valves, you would be amazed at the amount of carbon buildup that is on the valves.

[mikec03]

While it could be carbon deposits, I don't see anything when I change the spark plugs, admittedly the view is somewhat restricted.

 

Here's a post I found from 8 years ago by GeneralDisorder:

 

..the fuel delivery system is a feed back loop controlled primarily by the O2 sensor and the coolant temp sensor. When the engine is cold, [before the O2 sensor begins reporting], the fuel delivery is controlled based on the engine temp, and tends to be rich-thus no pinging. I would put my money on the O2 being bad-once it switches over to closed-loop operation, the O2's data is telling engine that it's running too rich.

 

Well, this makes some sense to me since I have good pick up and don't hear any pinging on acceleration when I first start out. I read somewhere else that the O2 sensor doesn't start to report until it reaches 600 F..

[OB99W]

I'm going to avoid the detonation/pre-ignition determination, since one can often lead to the other. I'll just discuss the causes of either -- they share in common that the air/fuel mixture is not ignited and burned at the right time or smoothly.

 

Carbon deposits may or may not be the cause of your engine's problem. A layer of carbon in a combustion chamber can raise the compression ratio. It can hold heat from the combustion process into the following cycle, when the cylinder should be (relatively) "cold". A piece might even act like a glowing ember. Just inspecting spark plugs often won't reveal carbon, since the plug tends to run hotter than the cylinder walls and head. Speaking of plugs, if their heat range is wrong (high), that alone can be troublesome.

 

Assuming that carbon isn't the issue, there are plenty of other things to consider. A too-lean mixture can definitely cause a problem. However, just because it doesn't occur when the engine is cold and the system is running open-loop doesn't necessarily mean the O2 sensor is at fault -- it might be, but open-loop richness can mask other conditions. For example, if there's a sufficient vacuum leak after the air metering, the ECU doesn't know about that air and the mixture can be too lean.

 

A cooling system that's not up to par can give you grief. Even if the gauge doesn't indicate above midpoint, it's possible to have localized hot spots. The coolant concentration being significantly wrong will sometimes impact cooling enough to lead to knock.

 

Have the heads ever been milled? (Sometimes in order to get it flat again, a warped head will have too much material removed.)

 

One cause that's often overlooked is a malfunctioning EGR system. If there isn't enough exhaust gas recirculation when needed, that alone might be your engine's problem.

[FleaDog]

Sorry to bump an old thread but instead of starting a new one with a similar issue I thought i'd post here.

 

I am getting what i presume is pinging (knocking?) at around 2-3000 rpm range under moderate-heavy load. This is right around the shifting point(s) during normal driving. Seems to be most noticeable just after shifting into 2nd gear, but can be heard in other gears. Doesnt seem to occur until engine is warm. Most pronouced under moderate throttle say going up a hill. Under minimal/light throttle/load it does not seem to occur. It was less noticeable when i filled the tank up with 93 octane, but is again more noticeable after a tank of 87. Mileage seems stable. No CEL's, but i havent hooked up the scanner to see if there's anyhting that could be there (if thats possible).

 

95 auto wagon 170,000 miles.

Bought the car 2 months ago dont know service history.

changed plugs, didnt seem to make difference.

 

I need to pull the motor to do the flex plate b/c i have leaking oil in that area, and will do the t belt/tensioner/pulleys b/c someone mentioned earlier in this thread that could be a source. A poster also mentioned the oil pump could be it, so I will change that also.

 

Another mentioned check the EGR system, so will look into that.

 

Could this be indicative of the knock sensor goign bad?

 

thanks!

[FleaDog]

Filled up with tank of 93 and it again pretty much went away totally. (going from 87 oct).