Erratic Knock Retard Index Readings

[Shabadoo]

This is going to be a long story.

I recently did some data logging on my expedition using an SCT X4 and was getting the following readings for the Knock Retard Index



As you can see, it is quite erratic. Being relatively new to analyzing data from the vehicle I wasn't sure what to think of this. I did some research and from what I've found I know that the higher the number the worse the situation basically. However, from what I read on other forums it seemed that readings should not be this erratic when experiencing knock. In addition, true knock seemed to correlate with engine load, rpms, etc... The readings I was getting seemed to correlate with nothing. As you can see though, there is a pattern and the pattern does change somewhat with RPM, load, etc....but it is still very erratic. The particular sample you see in the screenshot is when I was cruising at about 45 mph (I was using cruise control).

A little background now. I did an engine rebuild over the summer for fun and to learn a little more about the car. In the process I decided to swap out the PI pistons with NPI to increase the compression. You may be thinking "What an idiot" but hey, I like to experiment. I have not yet gotten a custom tune though. I know it is highly recommended with modifications of this type but I just haven't been able to dedicate the time I would like to to it. So, during this particular data logging session I'm running the canned SCT 91 octane tune.

More background info. Since the rebuild, I have put nearly 3000 miles on the truck. In that time, two spark plugs have gone bad on me (Cyl. 2 then Cyl. 1 about 700 mi later) and caused misfires. Here are two pictures of the most recent one I had to replace:





From what I have seen, that is normal wear for a spark plug. The first one I had to replace looked the same. Please let me know if there is anything abnormal about it.

I idled the car for a bit recently and tried listening for any indication of knock. I heard nothing on the driver side and a very slight tap on the passenger. Nothing that would indicate the engine is about to blow up like the data would have me thinking. Plus, I know I have an exhaust leak somewhere near the headers so possibly it is linked to that. I haven't had the time to fix it lately.

Anyways, if you have read this far, thank you. I appreciate it. I tried to include as much info. as I could but I know I'm probably leaving something out. If there is any more information you would like I would be glad to get it for you as soon as I possibly can. Hopefully someone is able to provide a little insight.

 

[docraymund]

What you could do:

Examine the spark plugs for signs of pepper specks -this confirms engine knock -use a magnifying glass.

Test the knock sensor.

Examine short and long term fuel trims for signs of a lean condition.

Repeat the test again, this time under heavy acceleration.

 

[docraymund]

You could also make what tuners call a det can. It's basically a short piece of brass or copper tubing flattened at one end and attached to the engine block. A length of hose is connected at the other end. You could make it long enough to go inside the cabin. There you could listen for engine knock while the truck is accelerating.

 

[Shabadoo]

Thanks for the reply, I hope to test the knock sensor this weekend and maybe try out the det can. I re-examined the spark plug without I magnifying class (I need to go get one still) and there seemed to be a few dark specs here and there but nothing substantial. If you happen to have a picture or a link to one that I can compare to I would appreciate it. I wasn't able to find anything very good online.

I was able to log some data from the Short Term Fuel Trim and found some interesting stuff. Below is a screenshot of a WOT run. I was cruising at about 45 and gunned it to 60.



So you don't have to totally strain your eyes:

Dark Orange = Load
Light Orange = Spark Source
Dark Blue = Calculated average A/F Ratio ((SFTF Bank 1 * 14.36) + (SFTF Bank 2 * 14.36))/2
Light Blue = Spark Advance
Purple = Throttle Position

I'd be glad to attached the full data log if you would like to view it in excel or Livelink.

First off, I'm unsure if 14.36 is the correct number I should be using for the A/F Ratio. I believe that's supposed to be the target stoichiometric ratio. 14.36 was from a Honda forum I think...I've seen others say use 14.64 or something similar. From what I saw it seemed to be about where it should be, showing 12.34 to 11.86:1 during WOT and averaging 14.34:1 over an 11 minute drive.

So, moral of the story, A/F Ratios seem normal but I have no idea if I even calculated those correctly or what "normal" is for a 2000 Expedition. I haven't checked the knock sensor yet but I'm starting to suspect something is up with it. I may just replace it to be safe. Once that is all said and done, I'll do some more data logging and see what I get. If the results are the same I may contact 5 Star Tuning and look to get a custom tune to see if it irons some of the kinks out. Like I said, I increased cylinder compression but haven't gotten it tuned so maybe the car computer just can't quite handle it.

 

[rtbrjason]

If you raised the engine compression without re-tuning you would likely see some knock. Hopefully you're running 93 octane at least. If this is the culprit, you'll most likely need to reduce ignition timing to account for the higher compression.

It could also be something rattling around in the engine bay that the knock sensor is picking up. Knock sensors are just microphones tuned to pick up specific frequencies.

Hard to understand fuel trims with the data you provided. The AF ratio (while in closed loop) will always target 14.6 or so. It's a matter of how much fuel the ECU has to add or subtract to get to that ratio. Fuel trims should be within 5% or so of 0. Fuel with ethanol will raise fuel trims slightly as you'll need more fuel to reach 14.6. Environmental factors like humidity and temperature can change fuel trims sightly as well which is why I say within 5% or so of 0 should be fine.

If your fuel trims are really negative, that means the engine is pulling fuel due to a rich condition which could also cause engine knocking.

 

[Shabadoo]

I went back and looked at the short term fuel trims I logged. They look like they meet the specifications you laid out. Here is a little blurb I found on another forum about reading STFT:

An STFT value of greater than 1.0 or a "-" percentage means the ECU is removing fuel [1.1 (-10%)]
An STFT value of .90 (+10%) means the ECU is adding fuel.

(http://www.hptuners.com/forum/showthread.php?17611-Resetting-Fuel-Trims)

Let me know if what I found is incorrect for this application but if it is true then it looks like I should be good. Over an 11 minute drive, I averaged an STFT of 1.00 on both bank 1 and 2. The lowest it dipped being 0.82 (at WOT) and highest being 1.20.

Regarding engine compression, I'm thinking I need to get a good custom tune. With the mods I put on it couldn't hurt. I plan to contact 5 Star Tuning before the years end. That way I can hopefully eliminate that as a suspect.

As I do more research, I think the fluctuation I'm getting on the knock sensor is just a lot of noise. Like I said, I did an engine rebuild so it's possible something got moved and is now rattling, as you stated. I did read on a few forums that over tightening or under tightening the knock sensor can cause false knock. Is this true? I couldn't find a torque value for the knock sensor anywhere so I simply tightened it as much as I could without breaking it. If anyone has this information off-hand or has a shop manual, I would appreciate it.

 

[devincox]

subscribing to this thread. I just had an engine rebuilt for my 97. I went from NPI to PI heads, but Im having the same problem with the knock sensor readings. Adding fuel makes it better on mine but still not perfect.

If you make you "WOT run" on the highway where you have some type of retaining wall on one side you can here the knock more easily. You can also put the truck in 2nd gear, from a stop, and floor it.
I got a tune from 5 star but it didn't help much, however, I changed my injectors to 24lb injectors. (I have hot cams and ported heads) I still need to work on my maf transfer function.

I highly recommend a book by Dan LaSota; The Ultimate Ford EFI Tuning Guide. If you are going to get into tuning, this is a must read. I'm pretty sure he helped SCT's software and their training program.

I just bought SCT's pro racer software and a wideband, so I can tune mine myself. I would rather have someone else tune it but its been hard finding someone locally that wants the business.

 

[Shabadoo]

I'll head out to the highway and try that. Right now, when driving, all I can hear is what I'm pretty sure is an exhaust leak at the headers. Need to get it looked at.

Which particular tune did you get from 5 Star? I know they have 91 Perf., 91 Economy, etc... Did you need to send them any data logging files or did you just tell them you increased compression, etc, and they were able to build the tune? I plan on getting into custom tuning as well so I'll definitely get that book but for now I would still like to get a decent custom tune.

This upcoming weekend I'm going to attempt swapping the knock sensor without removing the intake manifold. I know it can be done but we'll see if I have the will. Not expecting any miracles but it will at least eliminate it as a suspect. If the erratic readings continue I may just chalk it up to false knock. Despite the readings, the car isn't retarding the timing so I'm assuming the computer is making some kind of decision as to what is real and what is not.

 

[devincox]

I'll head out to the highway and try that. Right now, when driving, all I can hear is what I'm pretty sure is an exhaust leak at the headers. Need to get it looked at.

Which particular tune did you get from 5 Star? I know they have 91 Perf., 91 Economy, etc... Did you need to send them any data logging files or did you just tell them you increased compression, etc, and they were able to build the tune? I plan on getting into custom tuning as well so I'll definitely get that book but for now I would still like to get a decent custom tune.

This upcoming weekend I'm going to attempt swapping the knock sensor without removing the intake manifold. I know it can be done but we'll see if I have the will. Not expecting any miracles but it will at least eliminate it as a suspect. If the erratic readings continue I may just chalk it up to false knock. Despite the readings, the car isn't retarding the timing so I'm assuming the computer is making some kind of decision as to what is real and what is not.

I got the 91 perf and 87 perf tunes. I had to work with them a bit since my combo is far from stock, though not radical. I didn't expect a finished tune but wanted a nice idle and wanted it to run ok. I did send them many data logs. You log exactly what they want to see, they can give you the pids/dmrs and then you go through some procedures, like stepping through the rpm levels or starting in 2nd or 3rd and going WOT.

You should be able to swap out the knock sensor on a 2000 without taking off the intake. I have a 1999 intake and have removed my knock sensor twice. I just had to remove the alternator so my hand would fit under there.

 

[Shabadoo]

I got the 91 perf and 87 perf tunes

How did the 87 performance tune run? Were you able to run on 87 gas with the increased compression?

I just had to remove the alternator so my hand would fit under there.

Good to know you were able to do it. I was a little nervous about trying. Did you disconnect any of the heater core hoses to get to the electrical connector? They seem in the way but not to the point that it would prevent me from getting a hand to the rear of the engine.