Well I go an auto tap, just one more way for me to learn more about forced induction.

I logged 2 runs from a dig 0 - 100 mph and noticed this, which I think is a little strange.

1st run with traction control on: no knock retard at all

2nd run with traction control off: 7deg of knock retard taking off from a dig, I was spinning the tires pretty good, 1-2 shift 7.2deg of knock retard, again spinning the tires pretty good. Also once I hooked up the knock retard would instantly drop to .02deg or 0deg again


My question. Is the knock retard I logged when driving with the traction control off a biproduct of losing traction and taking the load off the engine, it appears this is where it starts pulling timing? Secondly, is pulling 7deg of timing in this situation normal and safe for the engine?

 

It is my understanding that traction control cuts fuel and retards timing..maybe with it happening at the same time...fuel cut off and timing retarded, maybe knock does not regester because timing is already being retarded when traction control is activated. Now with traction control off all systems are a go...and timing is being retarded. What would be nice to know is how much knock retard there was on the dyno with the correct AFR. If its the same than I would not worry about it...if it is not the same something else is going on...the sensors I believe can be de sensitized..I am not sure though somebody with LS1 edit or proprietary software needs to chime in on that one. We always look for zero degrees of knock at WOT....I don't know if that is realistic with a FI setup.


[Modified by Shinobi'sZ, 10:25 PM 7/12/2003]

 

[

Detonation detected as KR will kill a FI setup fast. If you are consistently getting KR with your FI setup at WOT, then you need to back off your timing and proabably riched up the AF mixture. :yesnod:

 

:iagree: My magnuson was pulling 12 deg on mine and falling on its face after the holeshot. :crazy:

 

:iagree: I just hear a lot of people saying they get a few degrees of kr with FI. The goal is Zero at WOT..That is why I was wondering if he had kr on the dyno, and what his AFR is at WOT. Lets assume that the AFR is 12.0-12.5 (12.5 being on the edge for FI) and at WOT above 5500 rpms (peak power) is getting kr...what would pulling timing do to the AFR...does it do anything or does the AFR stay the same..just with less timing?

 

Generally pulling timing will richen up the mixture (and adding timing will lean it out.)

Also regarding the KR sensors, you can desensitize or even completley turn them off with LS1 Edit. For an FI car I would not recommend turning them off!

 

Was this after tuning? I have been able to tune them for 0 KR. It sounds like you maybe running WAY too much timing. Keep in mind the Mag delivers non-intercooled boost. With all that hot air being forced in, you can't run much timing at all. In fact 12 degress total ignition advance maybe too much for some cars.

 

Yes this was after LPE install and tuning. I took it to the track and tested it out and found it running slower than normal after the initial boost came on. Back to LPE it went for 5 weeks. They tested it at Muncie Dragway themselves, dynoed, etc. and could not remove the kr under load. :crazy:

 

This is what I thought...so what do you do if your AFR is 12.0-12.5 at WOT 5800-6200 rpm and you are getting 7 degrees of kr...do you desensitize the sensors....or back off the timing..which in turn will richin up the mix?

 

KR is bad, you don’t want it and it is typically a precursor to detonation. I think a few degrees here and there are inevitable, but 7 degrees is pretty major.

The knock detection system on our cars is pretty sophisticated and the knowledge out there is very limited.

Here is what I know about it: There are two microphones positioned in the valley cover. The PCM listens to these microphones and has an algorithm for determining what frequency patterns constitute a knock event. Each knock event is calibrated in the PCM with some amount of timing retard (depending on RPM, temp, MAP…). I think it is generally about .5 degrees of timing in the high load range. So, if you are seeing 7 degrees of KR, it would mean that the PCM saw about 14 knock events. There is also a decay algorithm that attempts to restore the normal timing every 2 engine revolutions. It is typically set at 50%, so the amount of retard will get divided by 2 every 2 revolutions if no more knock events occur. The last feature is a knock learn algorithm. The PCM has a way of determining repeated patterns of KR, so it actually applies KR before it detects the knock events (very little seems to be known about exactly how this feature works).

Knock is more likely to happen under heavy load and increased cylinder pressure, so I doubt that spinning the tires and decreasing load would be the cause of your KR. It is more likely that your traction control is retarding timing and therefore limiting cylinder pressure.

Here are a few things that I would log to help understand your situation.

1. RPM
2. KR
3. Timing
4. MAF frequency or gm/s
5. Demanded AFR
6. IAT
7. Vehicle Speed

This will allow you to see what type of timing is being demanded just before the KR takes place. It will also allow you to see if the timing curve ends up being different with traction control on. MAF will be useful for seeing if/when your MAF maxes out and at what load the KR starts. Demanded AFR will give you some indication of what the current calibration is demanding for fuel.

BTW, I never had any luck with autotap giving me consistent and timely data. It is better than nothing, but I saw a lot of inconsistencies. EFI live on the other hand is awesome and well worth the extra money.

 

I would never desensitize the sensor. The only desensitizing I know of is reducing the amount of timing it pulls for each knock event. I don't think there is a way of changing what it does and does not consider being a knock event.

Backing off the timing if more for reducing cylinder pressure than it is for increasing AFR. The remaining fuel is a side effect of decreasing cylinder pressure.



[Modified by QuickSilver2002, 11:37 AM 7/13/2003]

 

I respectfully disagree with never...sometimes other engine parts have caused a false knock ie..noisy valve train from big lift cams...I have heard of many tuners changing the values in the range for the knock sensors so that they do not pick up false knock. Just like changing the value for the cam position sensor when you change a cam..if you don't desensitzed the sensor you can get a cam misfire code. I have had it happen a couple of times and the tuner adjusted the values for the cam position sensor..I believe it is the same for the Knock Sensors...I am not saying turn them off..I am saying they can be adjusted so that they only will pick up true detonation...being right in the center of the block is ideal location as compared to the earlier L98 motors they were on the passenger side near the oil pan..(1) sensor.

That is the only part I disagree with or would like to know more about. The rest :iagree:


[Modified by Shinobi'sZ, 10:24 AM 7/13/2003]

 

I totally see where you are coming from and agree that it would be nice to desensitize the sensors under certain situations, but I don't think it can be done the way you described it (this is why I would never do it). Maybe someone who has GM software can do it, but I don't think it is possible with ls1edit. Like I said, I know you can reduce the amount of timing for each knock event, but the parameters for controlling the knock detection algorithm are yet to be discovered as far as I know. If someone knows a way of doing this please educate me.

 

If your air fuel is that rich then you probably shouldn't be having detonation unless you are running way too much timing or you have false knock. The first thing to always do is add some race gas and to check the spark plugs (proper heat range and gap.) If the KR goes away, then it was real. You will either need to lower the CR of the engine, or retard the timing more to run on pump gas. If it is still present with the race gas then is probably false do to exhaust banging around.

With LS1 Edit you can desentize the knock sensors in many ways to correct for false knock. I would not recommend you do this until you have determined for sure the knock is false. You can adjust the attack and decay rates at which the KR pull timing and then put it back in. Additionally you can set the max amount of KR allowed to be pulled out.

Regarding the misfires I do this to with large cams. Basically you just inrease the tolderances in all the tables for the amount of misfires allowed per cylinder. This can prevent the infamouns p300 codes, fix tq converted lock and unlock problems, and solve idle issues with flashing check engine lights.

 

:lurk: this is a great thread, I recently found that I have a problem with my car not performing as well as it should, I hooked up my scanner and found I have alot of timing being pulled under load. I'm trying to learn as much as I can about this and so far this thread ranks right up there as #1 thanks :cheers:

 

Yes this is a great thread.

Godspeed thanks for the breakdown on how to perform what I was describing to Quicksilver. I am not got into the computer stuff just the bolt on and motor stuff with our cars. I rely on what programmers tell me, what you described to me is what MMS did with my P300 codes. I am thankful that you verified what you posted because I couldn't verify what Quick was saying..what he was saying sounded Logical so I could respond with the specifics. Looks like we all gained some valuable info.
:cheers:

 

O.K. heres one for you all. I took the car out this morning with a friend and did some more logging. KR stayed at zero for the most part and sometimes at the very high engine speed just before up shifting it would jump to anywhere between .04 and 1.4, I belive this would be acceptable but opinions would be appreciated if someone disagrees.

Here is the interesting part of our R&D from this morning, After the run say up to 100-120mph / 3rd gear shift I would let off the throttle completely and KR would jump to 7 or 8, but not every time. My passenger noticed that the KR appeared to jump up only when the exhaust would pop on decel, I thought he was nuts so we did it a few more times and guess what every time the exhaust popped on decel the KR would spike. The sensors are picking up the exhaust detonation and treating it like engine detonation.

Which brings me to the next question, since the engine is not under boost and decelerating this shouldn't be a concern, right?

 

Based on the information, I will assume it is picking up drivetrain noise.
You stated it picks it up (KR) when its spinning the tires etc, which has a lot less load on it then when hooked up and pulling the mass of the vehicle. The give away is it drops to zero when its hooked up under full load and nothing on a traction control run.

Phil

 

Wait, I think you misunderstood.

What I said was right, but obviously not very clear. There is no way to desensitize the knock sensors with ls1edit. Read what godspeed wrote carefully. He is talking about reducing the amount of attack and increasing the amount of decay against the knock counts. This is not at all like what you can do to desensitize the misfire detection. In misfire detection, you can adjust the tolerance in microseconds between two normal firing points to say what is and what is not a misfire.

So to be clear, you can adjust the action taken when knock is detected (at a given load/rpm), but you cannot adjust what is considered a knock event. This may seem like a trivial distinction, but I feel that it is not. If you reduce the action taken when a knock is detected the pcm won’t do anything different if a stronger more pronounced knock event occurs at that same point. Does this make sense? Again, if someone has specifics that prove this wrong, please post.

BTW, there is also a type of knock that has nothing to do with hearing knock, There is also a type called burst knock based on the detection of a delta in airflow (but I have never seen it happen).

There is no doubt that the KR detection system is a double edged sword.



[Modified by QuickSilver2002, 4:37 PM 7/13/2003]

 

Right, no worries on KR at decel. I would check it a few more times under different weather conditions to see if you have a problem or not. I've personally seen mine go from 0 to audibile knock with just a slight change in weather. It sounds to me like you might be right on the edge.

Something to remember, most of us are maxing out the maf and have our fuel calculation basically hard coded to RPM. Maxing out the MAF means that the computer can no longer adjust for changes in air density or flow, so you can easily go from a safe AF to and unsafe AF. The trick is to be on the way safe side. Most people don't do this because they want the most HP possible during the dyno tune.

Maybe I’m just ignorant at this point, but I think most of the tuners have people fooled into thinking that this tuning thing is pretty consistent on an FI motor. I have personally found it to be very inconsistent and non repeatable from day to day. Again, this could be something specific to my setup or my lack of knowledge, but only time will tell. I have several more things to try (but my laptop and clutch are down right now). I would just suggest not being greedy and keep it on the safe side.