Considerable knock retard when revving the engine with no load, what gives?

[BoxheadTim]

Quick background summary: I'm trying to trace another issue, namely the engine running hot on prolonged light-medium load (uphill at 40-50mph with up to 5psi boost). I data logged a slow drive and noticed that under mild load I was getting loads of knock retardation. So I figured I'll start with the basics and check from the idle log up.

Anyway, I noticed that revving the engine in neutral at idle, standing still (so obviously no load on the engine) would result in the timing getting considerably retarded due to knock. Keep in mind that I'm using 91 octane 'cos that's all I can get out here unless I want to resort to either octane boosters or E85.

After reading a couple of similar posts on here I was wondering if those symptoms look like a bad knock sensor or if the log suggests that I'm looking at a different issue (like the timing being too aggressive, for example). The PO mapped the car himself on the road (with a pyrometer, not a WB, ) and my plan is to make the car safe enough to drive over the Sierra to FFtec to get it mapped on the dyno there - I plan to do the occasional track day and don't really need the map to be off in areas you don't reach in normal road use.

Anyway, I would really appreciate if someone more knowledgeable could cast their eye over the attached ECMlink v3 datalog and check if there's something amiss.

As an aside, I haven't managed to fit a wideband yet; I'm planning to do so in the next few weeks but at the moment the logs are NB only. Also, I noticed that the IAT values don't make sense (77F on a 80F day with a hot engine?) so I need to look into the IAT sensor or lack thereof as well.

 

[BigRSnail95]

check your knock sensor and make sure there is not a lot of gooey looking junk coming out of it? the reason you may see 77 degree iat temp is the person tuning may have locked it at that temp to tune it. it is an option on the v3 as thats what my buddy tunes mine with. thats all i can tell you.

 

[way2fastlaser]

reach behind the block and check to see if its tight in the block i had mine rattle loose and it caused spikes like i see in your log it has to be TIGHT and no goo

took me weeks to figure that one out just getting random knock

 

[GsxMan98]

Are you running the stock maf. If you have the gm maf it doesnt use iat so it will just say 77. Its probably just your knock sensor i would just check that first.

 

[BoxheadTim]

Thanks for the advice everybody - I'll try to check the knock sensor this weekend.

@GSXMan98, didn't even think about that but the MAF on the car awfully looks like a GM MAF, which would make sense with the rest of the mods. I guess I'll better add an IAT to my shopping list.

 

[BoxheadTim]

Finally managed to spend a little time with the Talon last night and noticed that the 'sealant' that's filling the knock sensor is definitely pliable and gives when you apply finger pressure. I guess that's a sign that I want to replace the knock sensor? Keep in mind we had temperatures in the 95F range yesterday.

At least it hasn't gone all gooey and leaked out but it didn't feel too confidence inspiring, either.

 

[gofer]

What color is the sealant coming out of the back? Take a picture if you can.

Odds are you need to replace it, with the new one be sure to torque it to FSM specs (17 ft/lbs). If you still get PK get a plastic washer from Home Depot and put it between the block and the knock sensor, then torque it again.

Should be problem solved after that.

 

[BoxheadTim]

I can't currently tell if there's any sealant leaking out as I had to do most of the investigation by feel and not by sight. It didn't feel like there was much if any sealant missing, it was rather soft at the current ambient temperatures (around 90F).

Do I need a special socket for the knock sensor? I can't believe that trying to squeeze the cable into a socket with the sensor would do the cable much good, but maybe I'm overly cautious.

 

[gofer]

You can use a crow foot tool to torque it without doing any damage to the wire going to the back of the sensor. I believe the knock sensor is 24mm but I'm not 100% on that, maybe someone can second that or knows the actual size.

Get a 24mm crows foot (assuming thats the size you need) which looks like this.

You must be logged in to view this image or video.

Just remember, that when you torque it to 17 ft/lbs with that crows foot the settings will be different. This is the calculator I've used in the past to determine torque settings using a crows foot socket.

Torque Wrench Crow Foot Calculator​

Don't shrug off the torque specs on this either. Too loose and the knock sensor will register PK and too tight it will do exactly the same, you need to make sure its torqued to 17 ft/lbs as per the FSM.

 

[chanley talon]

The harmonic balancer will cause knock if the rubber starts to seperate inside the balancer. Happened to me once.

 

[BoxheadTim]

Well, looks like the new sensor that I finally fitted is duff (I'm getting an error code the the knock sensor circuit now) - back to square one, pull it and try again .

 

[DeadlyScone]

Wanna try to bring this thread back alive so I can get this issue solved.
as previously mentioned the CEL comes on at about 2000rpm stays on until the knock will stop which is after you let off the throttle. I figured to compensate with the knock just retard the timing(Noticed BoxheadTim adjusted -1 on the timing tables) but this is where another issue could occur. When the car is at normal operating temp the knock will not happen anymore at 2000rpm or higher. So basically it just quits after 195 - 200F. Any thoughts would be great.

I don't know how well you fitted the knock sensor BoxheadTim and i have not got around to actually looking into feeling around for the knock sensor but maybe you could clue me in on your process since i own this car now

EDIT: i'll post a log on a 1st to 5th pull.

LINK HERE

 

[BoxheadTim]

I actually had the knock sensor fitted by a local mechanic who I trusted at that point. They found a wiring issue which caused the fault mentioned which were fixed by then and made the sensor functional again. The sensor was a NAPA OEM-spec one...

I would have another good look at the sensor and its installation if I were you, I've since discovered some rather "suboptimal" work these guys did on another vehicle of mine so I can't guarantee that it had been installed correctly in the first place. However given that the symptoms are the same between the old and the new sensor, I would expect that the sensor is working.

It's also been mentioned (not sure if it was on here though) that some of these sensors can be overly sensitive and might need a fibre washer when installed to bring them into the sensitivity range that the ECU expects.

One of the things I would suggest to look at (which was next on my list if I had kept the car) is the fuel compensation on warmup. Given that it was only indicating knock at lower revs during warmup, it might be that it needs a little more fuel at certain points there. But that's a semi-educated guess on my part, not real knowledge .

 

[DeadlyScone]

I actually had the knock sensor fitted by a local mechanic who I trusted at that point. They found a wiring issue which caused the fault mentioned which were fixed by then and made the sensor functional again. The sensor was a NAPA OEM-spec one...

I would have another good look at the sensor and its installation if I were you, I've since discovered some rather "suboptimal" work these guys did on another vehicle of mine so I can't guarantee that it had been installed correctly in the first place. However given that the symptoms are the same between the old and the new sensor, I would expect that the sensor is working.

It's also been mentioned (not sure if it was on here though) that some of these sensors can be overly sensitive and might need a fibre washer when installed to bring them into the sensitivity range that the ECU expects.

One of the things I would suggest to look at (which was next on my list if I had kept the car) is the fuel compensation on warmup. Given that it was only indicating knock at lower revs during warmup, it might be that it needs a little more fuel at certain points there. But that's a semi-educated guess on my part, not real knowledge .

ya i was reading some threads about knock sensors and say that you can use a plastic washer as you explained, so i might give that a try before doing anything else. hopefully it's not becuase the car's base time is not set correctly(cross my fingers) if so all that pre-ignition is going to kill the engine

 

[BoxheadTim]

I have this vague recollection that I did check the base timing and it was OK, but it's always worth checking. That said, I'd expect more knock all the time if the base timing was off.