Chasing knock problems

Razor

Forum tech Advisor
Staff member
Joined
Jul 31, 2001
After a bazzillion threads on this board of knock happening here or there, I want to make a thread to help those looking for solutions.

First, read this paragraph twice... Your knock sensor is basically a microphone that is located on the rear of the engine block. It is connected to an ESC module that "filters" out specific frequencies that are metallic sounding. Depending on the severity/intensity of this sound.. will vary the output of the ESC module. Your vehicle's computer picks up this information and displays the knock readings. This is why knock detectors will show knock before it is displayed on your scan tool. This is becuase knock detectors are connected directly to the output of the ESC module. Some chips have the ability to "mask" out the knock readings under certain conditions.. this is also known as "knock ignore". Example the Extender when you downshift.. it ignores knock for a predetermined amount of time.

Now here comes the kicker. Real knock escalates in intensity. It will increase in severity. Call it avalanche.. the little pebble that turns into the huge snowball. And monitoring knock on scantools your to look at numbers and severity that increase. Example 1, 3, 7, 11, 20... kaboom. Numbers that flash all of a sudden "20".. is false. Just as numbers that come then disappear. The only thing is if your on the edge of knocking.. numbers will hover in the very low digits coming and going. Like 0, 0, .5, 0, .7, 0.. etc. This means your on the edge and need to make a tuning change(timing/octane/fueling). I'll get into this latter.

So is it real or false? Hopefully re-read the above until it sinks in.

Now.. if your still in doubt ask your self the following questions.

1) Is the boost set at a level that the motor should not knock based on fuel octane? Pump gas 91-93 octane should get you typically 12-14 PSI boost with a normally functioning engine. Note I use the word "typically". Some may get 24 PSI.. Just like some win the lottery ;)

2) Is the chip working properly and matched to the injectors used? If your not 100% sure on this question.. stop and buy a chip from someone reputable. Period. Stop wasting your time and the time of others trying to get an unknown chip with unknown injectors to properly sync together. If your chip is matched to the injectors, make sure fuel pressure is set correctly and BLM numbers are within normal ranges. I'll get into BLM latter in tuning.

3) If your boost is low, your octane is premium, you have a matched chip for your injectors, all scantool numbers are withing normal limits, engine idles quiet, stays quiet when you rev it, you should not be experiencing any "REAL" knock. Assuming your engine doesnt have a blown headgasket, huge leakdown, wiped out cam, etc. Easily confirmed with leakdown and compression tests.

4) At low boost your motor should not knock. Unless something internally is coming apart. I can pretty much make this a blanket statement. Just like it shouldnt knock when rev'ing in park/neutral.

So far so good.. so you followed the above and still see the lights and numbers climbing on your scantool...

Lets take it one step further.. There are way too many conditions that can create knock readings. There are certain conditions that aggravate trying to ascertain whether its real or false. I typically when working on this issue try and limit certain scenario's. The number 1 scenario is trying to do a boosted launch on a street surface then seeing the numbers climb. What this means is slamming the pedal down in first gear and spinning tires in hopes to get real data. My suggestion is to perform this test in manual 2nd gear or manual 3rd. Whereby the car cannot spin tires or "shift" gears. And then ease into the throttle watching the boost guage climb while paying attention to knock readings. If the boost guage climbs slow to the preset boost level and the knock readings dont appear.. you have your answer. SO if you set to 14 PSI and it runs with zero on the knock readings your chasing false knock.

There are some scenario's whereby all the idle and drivability would appear to be normal.. and the car will start showing knock as it goes into boost. Example is bad mass air flow sensor(MAF).. whereby it doesnt read the airflow correctly and an improper amount of fuel is sent into the engine. This is typical with remanufactured MAF's sensors. And only way to ascertain is by either swapping for a known good piece, or watching air fuel readings on a wide band sensor. If the car leans out as it goes into boost.. it can/will knock.

Another is a bad O2 sensor whereby your fuel settings are "learned" improperly. That is why I always preach the use of GM pieces in your Buick. There is nothing wrong with an AFS20 single wire O2 sensor.


Now.. lets see if we can figure out where the metallic noise is coming from. This would be to address false knock issues.

So.. you've lowered boost.. assured proper octane.. car is running slightly rich.. chip and injectors are matched.. all plugs look even and light brown/tan.. gets great gas mileage.. no drivability issues coming from stops.. etc..

Lastly.. 3 degree's of sustained knock is not severe. You will not blow your engine up at low boost with 3 degree's showing. Anything under 15 PSI..probably can sustain 8-10 degree's for a pretty long time.

If you see .1 knock.. the sky is not falling. I have seen multiple times whereby someone see's .5 degree's and lifts screaming... ohh noo :eek: .

And if your hearing the "marbles in a can" type of sounding when putting a load on your engine.. aka.. pinging.. YOU BETTER LIFT OR ELSE>>> :eek:


Lets work on deciphering the flase knock..

The best thing to do is make a list of what can create this and work down the list. Crossing off what is and isnt.

And you need to be able and replicate the condition without undo stress on the motor. Putting the engine at 25 PSI is foolish looking for a problem.

So we get back to the microphone listening to the engine.

The following is a list of possibilities.. and there can be even more.. but this should cover the majority.


Engine mounts, downpipe clearance, TQ converter bolts loose, pulley/tensioner problems, wrong knock sensor(microphone), wrong/bad ESC module, wire harness banging on knock sensor, bad timing chain, bad rocker shafts, noisy lifters, bad thrust on crank, transmission pump, bent pushrod, bad bearing, bad piston/wrist pin, MAF pipe hitting alternator, alternator, power steering pump, etc.. hope you see the complexity and time involved in diagnosing this type of issue.
 
Ahh tuning and BLM's

Rule of thumb, you never while driving the vehicle want to see BLM(BL) numbers that go to either extreme of your chip. Most chips will either be 90-160 or 105-150 as min/max numbers. 128 is the center point. And if all is working according to plan.. your numbers should hover +-10 from 128. Some can argue if the BL is lets say 110 or 145 it is still adjusting and not to worry. This statement is fine unless your chasing a problem then a ? mark comes up. Typically on a car if the BL's are 118-138 under all driving conditions.. I look elsewhere for problems. BLM=long term fuel adjustment. Setting your fuel pressure too low creates a higher BLM meaning the ecm is having to add more injector duty cycle to get the air fuel where it needs to be.

Now understand your start and end points. It all starts at the MAF, and ends at the O2 sensor. Your ecm, adjusts fueling based on what the O2 tells it. Ever heard garbaje in=garbaje out. Wrong or bad O2 will make the ecm make improper fuel adjustments. Perfect example is a , bad injector, cracked header, hole in IC, wrong PCV, vacuum leak, etc. Air leaks dont allow for proper adjustments to be made. And the chip is what reads this information through the ecm and makes commands to correct the issue.

So before you start beating on a car.. make sure the foundation has been set.. this foundation is critical if you plan on beating on the motor and it is to live. Any issues must be dealt with now..

If you search High Blm's on this board.. it goes for pages.. always the same thing.. Rule #1.. understand whats happening, then systematically address what it can or cannot be. Sometimes its a simple hose that blew off.. other times it can be an electrical error in the vehicles wiring. Both are extremes.. and while I have chased this issue on many Buicks.. its no walk in the park with a preset amount of time to get sorted out.

Now you know why I mentioned chip and injectiors must be matched ;)
 
Last tidbit.. some will hate me for saying this, but it is what it is.

When trying to establish a baseline with your vehicle.. get rid of all the gizmo's that are attached to the electrical part of your engine management. This means volt boosters, TPS enhancers, heated O2 sensor, aftermarket sensors non-GM/AC Delco, specialty spark plugs, aftermarket ignition module, remanufactured MAF, etc..

If you have access to a known good MAF.. swap it in.

Get your car as close to 1987 Electronics as possible. It didnt have this issue in 1987.. lets get a baseline without issues.. then complicate ourselves accordingly.
 
hey julio, great info I have quite a list of things to check on. question about ecs module can this be tested to see if it's bad. I am getting low boost knock and have been bangin my head trying to find it. I am goin to try to borrow a good maf and try that first just curious. thanx
 
Thanks Joe and Chris... This subject is so beaten to death.. and everytime what I just wrote I felt was needed..

ESC module amplifies what the sensor see's. If you disconnect the knock sensor and still see knock numbers the ESC module is bad. It could also be bad and picking up the wrong frequencies.. other test is replacing it with a known good one.

HTH

PS.. a little birdie told me your TTA less :eek:
 
Nice writeup Julio. Clear and to the point. I hope everyone appreciates the time and effort you put in to it. You may not be as dumb as you look. :wink: Actually, I don't even know what the he!! you look like. :smile:
 
Thanks Bob.. its tough to address this particular issue. Hopefully this becomes a tool that can help someone out.
 
Hey Julio I didn't see any mention of OIL ( PCV related ?? )in the intake in your write up. From your experiences , do you have anything to add in that regard?? I know I have seen threads in the past where you mention it. I was wondering how often you've run into the oil in intake problem with customer cars and is it the carbon itself from the oil burnning that we need to be more worried about??

BTW: Great article ;) :biggrin:
 
bsdlinux said:
Hey Julio I didn't see any mention of OIL ( PCV related ?? )in the intake in your write up. From your experiences , do you have anything to add in that regard?? I know I have seen threads in the past where you mention it. I was wondering how often you've run into the oil in intake problem with customer cars and is it the carbon itself from the oil burnning that we need to be more worried about??

BTW: Great article ;) :biggrin:

Thx..

Absolutely.. but at low boost levels its not an issue unless major major oil is sucked into the motor. Like someone driving a crop duster for weeks cuase their turbo seal let go and they insisted on continuing to drive it.

At high boost it plays a major role. This is why most newly built engines dont detonate as easy as 150K mile motors.

And its the simple stuff everyone overlooks.. I remember this old man worked at a garage told me.. " I have fixed more carburator problems replacing plugs than replacing carburators" :wink:

So making sure the engines tune-up is up to date.. is critical :redface:
 
Educate me further.
Is there any way that running too cold of a plug - or a loose plug wire - could cause knock - false or real?
Thanks
 
Loose/bad plug wire wont cuase knock.. but your engine will noticeably wont run properly. Boost or not.

Cold plugs will foul.. but that would cuase some sort of drivability issue as well.

Both conditions should not cuase knock sensor to register.

Maybe stated like this, unlug a cylinder and drive around.. you wont want to put the car into boost.

If your having mis-firing.. you can tell. It is very obvious.. and the higher the load on the engine, the worse it gets... and the knock sensor shouldnt go off unless your really trying to hurt the engine while its mis-firing. Its like breaking an anvil.. some can try really hard and eventually succeed.
 
Julio YOU ARE THE MAN!!!!!!!!!!!!!!!!

I love the KISS method ( keep it super simple )

thanks for the GREAT info!!!!!!!!!!!!.

JD :D :D :D
 
Great job man! But if you get everybody trained, I'll never work again.LOL Just wanted to add one more thing. You did mention the importance of having proper fuel pressure, and there are a million and one threads out there to address what proper base pressure should be. My comment is this. I see alot of stock, or close to stock cars with knock issues. When I take it out for a spin I make sure I have a fuel pressure guage where I can see it before putting a load on the engine. It's all part of making sure all the basics are covered before chasing demons. The last one had a new filter, and a new Walbro pump with wiring upgrade. The customer(A shop owner)had been beating on it looking for the problem for some time. It took 1 minute to hook up the guage(with him swearing it wasn't fuel related the whole time)and then drove it one block to watch the fuel pressure drop as soon as it started to boost. He needed a new fuel tank. The car had sat for a few years, and the tank had started falling apart from corrosion. Later
Danny
 
Top