Ignition Missfire and knock count question

[lagos]

I have a question for all of the guys that tune cars for a living, or have the ability to data log.

Say you have a car with a known Ignition problem that is causing missfires under boost/high load. Would this ignition missfire be picked up by the knock sensor as knock? Or would there just be a hesitation, but the knock count would read zero?

[SpendOne]

Originally Posted by lagos

I have a question for all of the guys that tune cars for a living, or have the ability to data log.

Say you have a car with a known Ignition problem that is causing missfires under boost/high load. Would this ignition missfire be picked up by the knock sensor as knock? Or would there just be a hesitation, but the knock count would read zero?

On my car, DSM, if you ignition breaks up on high RPM's it will not read it as knock.

[lagos]

Interesting.
I cant monitor knock count directly, but I can monitor my car falling back on its safe fuel maps which normally happens from knock sensor input. Lately this started happening more and more, and now its at the point that it happens with the boost controller off and the car at 7psi (10psi is stock boost). I did some trouble shooting and found out that my distributor is leaking oil into my dist cap. So that would very likely be the cause of my issue, but if thats the case, why would the car fall back on the safe boost map if its only triggered by knock input.

hmm... Either way, I'll have to rebuild my distributor and see what happens.

[MKIIISUPRA]

The primary purpose of knock sensors ARE to pick up preignition symptoms. Typical readings are 0-10hz sometimes up to 25hz. With enough knock, the computer will go into limp mode. This could possibly throw a code if its bad enough. Usually the higher readings = bad gas. What octane do you DD on? Also, what kind of plugs are you running? List some mods.

[lagos]

The car runs on 93 octane daily, along with alcohol injection, so octane should not be an issue here, especially since the problem happens with the boost turned all the way down to just 7psi. AFR doing a 7psi pull is in the 10s.

Plugs are the standard one step colder ngk bkr7e's. I normally run .028 gap, but i tried a fresh set @ .026 with no change.
I ended up rebuilding my distributor so that it no longer leaks oil (it was a pain in the ass to do) and installed a new rotor. Tested the distributor cap and wires with a multimeter and they both passed. Ignition timing is correctly set at 10tdc.

The ECU IS going into limp mode. It normally happens at around 5k on a wot pull.

I have to admit, I know my engine like the back of my hand, and I'm a little stumped by this issue. Im going to replace the Dist cap, and wires, even though they tested in spec and see what happens. Im also going to try to check my mechanical timing to see if the belt didnt skip a tooth.

[MKIIISUPRA]

Originally Posted by lagos

The car runs on 93 octane daily, along with alcohol injection, so octane should not be an issue here, especially since the problem happens with the boost turned all the way down to just 7psi. AFR doing a 7psi pull is in the 10s.

Plugs are the standard one step colder ngk bkr7e's. I normally run .028 gap, but i tried a fresh set @ .026 with no change.
I ended up rebuilding my distributor so that it no longer leaks oil (it was a pain in the ass to do) and installed a new rotor. Tested the distributor cap and wires with a multimeter and they both passed. Ignition timing is correctly set at 10tdc.

The ECU IS going into limp mode. It normally happens at around 5k on a wot pull.

I have to admit, I know my engine like the back of my hand, and I'm a little stumped by this issue. Im going to replace the Dist cap, and wires, even though they tested in spec and see what happens. Im also going to try to check my mechanical timing to see if the belt didnt skip a tooth.

well first off, on a side note, that AFR is a bit rich. low to mid elevens would be much better, especially at the price of gas.

Next, what codes are you throwing.

Skipping a tooth wouldn't cause this but still check it just in case.

What kind of knock readings are you getting WOT throttle?

Describe the misfire when it happens. Does the power just cut off throwing you forward some? is there afterfire? maybe backfire?

[lagos]

well first off, on a side note, that AFR is a bit rich. low to mid elevens would be much better, especially at the price of gas.

Thats stock ecu tuning for 7psi. I normally see a/f ratios in the mid 11s at 15psi of boost.

Next, what codes are you throwing.

None. Ecu is code free.

What kind of knock readings are you getting WOT throttle?

I cant monitor an actual knock count. I have an led wired up to my stock tvsv system so that I can monitor the ecu going into limp mode.

Describe the misfire when it happens. Does the power just cut off throwing you forward some? is there afterfire? maybe backfire?

Its just a big loss of power. I see the ecu switch over to limp mode, and it stays there for a few seconds/minutes after the even has happened. No backfires, or anything like that.

Its pretty much what you would expect a factory ecu to do if you were running too much boost.... problem is, this is happening even at stock boost levels.

[MKIIISUPRA]

Originally Posted by lagos

Thats stock ecu tuning for 7psi. I normally see a/f ratios in the mid 11s at 15psi of boost.



None. Ecu is code free.


I cant monitor an actual knock count. I have an led wired up to my stock tvsv system so that I can monitor the ecu going into limp mode.



Its just a big loss of power. I see the ecu switch over to limp mode, and it stays there for a few seconds/minutes after the even has happened. No backfires, or anything like that.

Its pretty much what you would expect a factory ecu to do if you were running too much boost.... problem is, this is happening even at stock boost levels.

Do yourself a favor and do a boost leak test (a real one, pressurize the system to whatever you normally have it at with a compressor). This sounds like a classic case. I would bet you have a decent size leak somewhere even if your boost doesn't bleed much.

Personally i have the same issue, except my system only detects the leak at low throttle and higher RPM, right around 5K RPM. I assume it doesn't detect as WOT since the turbo can keep up with the size of the leaks. I happen to be leaking out of three injectors and BOV. ARGH. My injector insulators didn't seat right during the last rebuild.

[lagos]

Originally Posted by MKIIISUPRA

Do yourself a favor and do a boost leak test (a real one, pressurize the system to whatever you normally have it at with a compressor). This sounds like a classic case. I would bet you have a decent size leak somewhere even if your boost doesn't bleed much.

Personally i have the same issue, except my system only detects the leak at low throttle and higher RPM, right around 5K RPM. I assume it doesn't detect as WOT since the turbo can keep up with the size of the leaks. I happen to be leaking out of three injectors and BOV. ARGH. My injector insulators didn't seat right during the last rebuild.

Its not a boost leak. I've already boost leak tested the car more then once.
The car is going into limp mode. Im not saying that based on feel, but based on the fact that I can monitor when it DOES go into limp mode.

It has to be either and ignition related issue, or something is getting picked up by the knock sensor. I do have a very noisy clutch, and im starting to wonder if it could be a phantom knock issue.

[MKIIISUPRA]

Originally Posted by lagos

Its not a boost leak. I've already boost leak tested the car more then once.
The car is going into limp mode. Im not saying that based on feel, but based on the fact that I can monitor when it DOES go into limp mode.

It has to be either and ignition related issue, or something is getting picked up by the knock sensor. I do have a very noisy clutch, and im starting to wonder if it could be a phantom knock issue.

You're not getting this. It can only be caused by so many things. It doesn't matter how many times you've boost leak tested, a leak can spring ANYTIME. Take my car for example 400miles ago no leaks what so ever. Now im leaking from 3 injectors and a legit HKS SSQV all of which only have 6K miles on them....

A knock sensor being triggered would not throw enough timing to cuase you to be moved much, fuel cut however, would. A past car i've run without knock sensors all together for a few thousand miles, after i connected them, the power difference wasn't very significant. Timing advance only goes so far on a stock ECU.

[SovXietday]

Is there a way to get rid of your knock sensor with that car? If so, that's what I would do. Seriously.

Knock sensors are barbaric at best when it comes to collecting knock data.

[MKIIISUPRA]

Originally Posted by SovXietday

Is there a way to get rid of your knock sensor with that car? If so, that's what I would do. Seriously.

Knock sensors are barbaric at best when it comes to collecting knock data.

knock sensors will save your engines life. And no you can't get rid of the knock sensors unless you create a device to mimic them to trick the ECU into thinking they are present, this is HIGHLY illadvised.

[lagos]

Originally Posted by MKIIISUPRA

You're not getting this. It can only be caused by so many things. It doesn't matter how many times you've boost leak tested, a leak can spring ANYTIME. Take my car for example 400miles ago no leaks what so ever. Now im leaking from 3 injectors and a legit HKS SSQV all of which only have 6K miles on them....

A knock sensor being triggered would not throw enough timing to cuase you to be moved much, fuel cut however, would. A past car i've run without knock sensors all together for a few thousand miles, after i connected them, the power difference wasn't very significant. Timing advance only goes so far on a stock ECU.

Ive owned my car for 7yrs and I do all my own work on it. I would say I know it VERY well, so trust me when I say that the car is pulling timing and going into limp mode, and that its not a boost leak or fuel cut.

I pondered the idea of giving the knock sensor a nice quiet new home... haha. But that wouldnt work because the ecu expects to see a certain noise level to determine that your knock sensor is working correctly.... not to mention that its a very bad idea.

I have a new set of oem plug wires on order, so that will be my next step.

[MKIIISUPRA]

[QUOTE=lagos;863612]Ive owned my car for 7yrs and I do all my own work on it. I would say I know it VERY well, so trust me when I say that the car is pulling timing and going into limp mode, and that its not a boost leak or fuel cut.

I pondered the idea of giving the knock sensor a nice quiet new home... haha. But that wouldnt work because the ecu expects to see a certain noise level to determine that your knock sensor is working correctly.... not to mention that its a very bad idea.

I have a new set of oem plug wires on order, so that will be my next step.[/QUOTE]

let me know what happens, though i highly doubt this is your problem.

[240sxDann]

possibly....a bad knock sensor? lol just an idea! I gotta replace mine.

[vividk20]

With my turbo'd civic I know that I have to replace my knock sensor...the reason is, is that my check engine light is on about knock count...with honda's, knock sensors will pick up knock and send a message to the ecu to adjust timing to eliminate knock, if and when you get a CEL there is an issue with the actual sensor...they are extremely sensitive so i would recommend checking the sensor.

[igo4bmx]

Originally Posted by SovXietday

Is there a way to get rid of your knock sensor with that car? If so, that's what I would do. Seriously.

Knock sensors are barbaric at best when it comes to collecting knock data.

thats a pretty bad statement since now owning an evo there are less blown motors in the evo world than in the obd1 knock sensorless honda world

[SovXietday]

Originally Posted by MKIIISUPRA

knock sensors will save your engines life. And no you can't get rid of the knock sensors unless you create a device to mimic them to trick the ECU into thinking they are present, this is HIGHLY illadvised.

No, I just turn mine off with my S300.

I do not like knock sensors. Pulling timing for knock is a bandaid, the engine shouldn't be knocking in the first place.

[lagos]

Originally Posted by SovXietday

I do not like knock sensors. Pulling timing for knock is a bandaid, the engine shouldn't be knocking in the first place.

We also shouldn't be running more then stock boost, and should be using race gas all the time.
In the real world, knock sensors save your ass.

[SovXietday]

Originally Posted by lagos

We also shouldn't be running more then stock boost, and should be using race gas all the time.
In the real world, knock sensors save your ass.

Ok, no offense man but have you ever tuned a car? Do you know what goes into tuning a car?

There is ZERO reason with a good EMS that you can't tune for a good powerful car that runs with some cusion beneath the detonation threshhold. There is just no need for the innaccuracy of a knock sensor. Being as they work by sound waves the sensor can just as easily disregard knock as it can mistake another sound for it. I don't know about you, but I don't trust a sensor that reacts the same to ping as it would to hitting my block with a hammer. The idea is great, the method is flawed.

If you have to rely on the knock sensor for your engine's safety than you need to rethink what you're doing. Simply put, how does the knock sensor work? It detects knock, and THEN pulls timing. That's like putting on a kevlar vest after you get shot, lol.