Return of a Tuning Tale. Chasing classic knock

[Neat]

The lean condition you are seeing may be cylinder misfire. I'm not sure about the narrowband O2 sensors, but I have seen the wideband on my car show lean when a cylinder drops out.

It was explained to me that the O2 sensor is just that, an oxygen sensor. It is not an air/fuel meter, even the widebands that are sold as air/fuel ratio meters use an oxygen sensor - which, not surprisingly, only senses oxygen. When the spark drops out for some reason, there is no burn to consume the oxygen or the fuel. The unburned oxygen and unburned fuel make their way into the exhaust, but the oxygen sensor only senses the oxygen and reports a lean condition; even though raw fuel may be coming out the tail pipes.

The momentary lean condition you describe sounds like it might be just that. There may be an intermittent loss of spark (which jives with your carbon traced coil) which causes a non-fire on one or more cylinders, the O2 reports lean because of the excess oxygen in the exhaust. I don't know how fast the ECM reacts, or if it even does closed loop correction during your knock / misfire event, but my LS powered car with a Holley stand alone will make a closed loop correction for the false lean condition in about 1/25th of a second. The sudden addition of fuel (again, on my LS car, not sure on LT5 stuff) causes a dopey rich effect, and one or two combustion cycles later I'll get multiple flame fronts from the excessively rich mixture hiding out in the ring pack / ring land area. The ECM then reports knock. The problem is exacerbated if the dodgey coil comes back online after only one or two misfires and then drops out again.

The scenario above only happens at light throttle. At WOT the extra fuel from the closed loop correction of the false lean is consumed enough to prevent extra fuel from hiding from the combustion event.

That's all real world stuff I've experienced with my LS car and the Holley EFI. I don't know how much, if any, of that is applicable or helpful to the LT5 / stock ECM; but maybe something in there will be helpful. Please keep us posted!

[XfireZ51]

Originally Posted by Jagdpanzer

Dom,
MSD 8224 DIS coils?
How old?

Phil,

I got the MSDs when I first ported the top end on my motor which was back around 2007-2008.
Got the motor put back together today, started logging again, and saw the same lean condition again, and yep knock retard.
I?ll also check the spark plugs later today after cool down and then probably taking plenum off again and run injector check.
Sometimes this gets pretty GD frustrating.
I?m also going to run Marc?s knock sensor check just for the hell of it altho that would not explain the right bank O2 sensor reading lean. Knock retard comes on when motor gets to about 3k rpm. Happens under no load as well.

[XfireZ51]

Just a quick update. I did swap out the coils for new AC Delcos. I also swapped
back in my MSD 8.5 wires in place of Magnacores based on some comments from Hib on the subject.
I had also run a test on the injectors and found that two secondaries (2&4) tested at 48ohms when hot while the rest were in the 16.9-17 ohm range after a 30 minute heat soak. Thank you FIC for warrantying the injectors and shipping replacements quickly plus they sent prepaid package for returning defective injectors. They want to take a look at them.

I use the ALDLdroid app to datalog these days. Along w a .csv file the app also produces an .xdl file which can be read into TPRT and then used to produce a History Table. You can build a table of your own which I did to "map" where the KNOCK RETARD was occurring rather than hunting thru a spreadsheet. I've attached a sample of the output from that. Its really helpful altho I don't see a way of breaking it down by Port Throttle status (Open/Closed). It seems to only handle 2D tables.

[XfireZ51]

A followup on this. I have been corresponding w Marc and mapped out a plan to help narrow things down, mainly to help determine whether this is true knock or something mechanical. So the plan was to:

1. Make sure my prom had the correct KS filter circuit. Marc provided part# of the acceptable filters that would be found onboard the prom. Mine has part#
2061252 stamped on it which came specifically w the 92 proms.

2. Filled tank w 93 octane and poured a bottle of Lucas octane boost. see attached TPRT KR mapping from datalog 12.09.2020. As you look at this, the SA table being used is at or below SA Marc uses on his performance setup for modded motors. Most of the "knock" recorded here is occurring in steady state cruise with little to no change in parameters like TPS%, MAP, or vehicle speed. Marc tells me
LT-5s will exhibit tip-in knock.

3. Ordered KS from Jerry just to make sure it is the correct sensor. Inexpensive enough (thx Jerry for quick shipment). Qjacks make it a pretty easy swap. In the process of doing this, I did check resistance values of BOTH KS. They were both at 3.9k ohms. In the process of installing the Jerry's KS, I found that the in place sensor was hardly torqued. It appeared that only 1-2 threads were engaged w the block. I installed the new KS(an AC Delco) and torqued it down paying attention not to over torque it. Using the same calibration and existing fuel you can see the resulting knock "map" w the new KS and properly installed. See attached from 12.10.2020. Altho still showing some "knock", the pattern is significantly minimized and less random. But I believe it still points to having some kind of mechanical noise setting off the KS rather than detonation occurring. This is also confirmed by inspecting the plugs. They show no indication of detonation occurring. No metal flecks embedded in the ceramic portion of the plug tip. Going WOT produces no knock.
Motor sounds and runs great with no apparent pinging noted.

N.B These knock maps are derived from logs which comprised about a 1hr drive of 40-50mi incorporating stop and go, hills and highway driving at
70-80+mph. In both logs, I had a total of 7-8 KCs during the logging session, most of which were .35-.7* KR. In some cases the ECM will increment the KC without any actual KR being invoked which did occur during these logs.

[wfot]

Have the Knock tables been altered on the chip? I from what I understand, the factory knock tables are fairly aggressive and can been tamed down some, but there needs to be some safety margin in them or if you eliminate all knock ability from the ECU, the engine will knock... If you have a stock bin file, you can compare the tune you have and see any differences.

is it possible that the air temp sensor could be bad/wonky? if the ecu thinks it is getting a one temp "x" and it is really getting temp "y" that could cause a problem. again I recommend GM parts for sensors, the aftermarket stuff can have much wider limits than OEM I have found

TPS adjusted to .53 volts?

and I think you did say you modified your accelerator map - maybe compare it to a stock bin and see if there is an unusual difference in the lower KPA settings ??

first time reading this thread and it does sound frustrating

there is also a table to adjust the fuel flow rate of the injectors, I think FICs are really close to stock flow and usually that value does not need to be adjusted if you are running their stock replacements but if they have a different flow rate then stock that could be a place to check as well...

my 4 cents

John

[XfireZ51]

Jon,

Thx for ur input. I?ve been doing my own tuning for the last 20 years and the LT-5 in particular for the last 13. I am aware of the areas u point out. The motor is fairly well modded w porting, cams and exhaust. The driveline is modified w SM flywheel, sprung hub disc and 4.10 Dana 44 gears.
My latest effort has been to confirm what I am seeing, or not seeing, on the spark plugs, which is a lack of evidence that would indicate detonation. In conjunction w Marc?s suggestion I mixed Lucas Octane Boost w my 93 octane and am running a less aggressive SA table than Marc does for his upgrades. Also installed an AC Delco KS from Jerry?s Gaskets just to make sure. As u can see, even after that, some knock still is registered in low load, steady cruise conditions. That strongly suggests that the KS is picking up a false knock from some mechanical source. There may be some kind of harmonic developing which the KS is sensitive to. At this point I just may ignore it because it doesn?t appear to impede performance and does not have any consistency other than occurring at 77-80+ mph range when it does occur. Other than that car runs and idles great.
One thing I did find before this last datalog, is that the previously installed KS was hardly torqued into the block. It was hanging by just a couple of threads. This may have contributed to the knock pattern u see in the first scan and why its so different from the second scan.

[XfireZ51]

Hoping that the saga here ends. The corollary in tuning is that you first make certain all the mechanical bits are operating on nominal levels, ie they're working the way they're supposed to work. Otherwise you'll be chasing your tail tuning around the deficiency. As part of a recent rehab of my motor, which included a freshly rebuilt short block, it was necessary to replace my Accel injectors. After 10 years, they'd pretty much given up the ghost with the #7 secondary having a deteriorated aerator which caused fuel to simply stream into the cylinders creating a lean condition. That in turn brought about knock ultimately cracking the top compression ring in 4 places. So enter a new set of injectors which are advertised as flow matched to within 1% of each other.
In the course of tuning, the plugs did not show an even burn among the cylinders however, so I had my suspicions. I waited for an reason to pull the plenum and a blown alternator gave me the opportunity. I had the injectors flow tested and I have attached pics of the actual flow for both left an right banks. IOW, each primary was flowed along w corresponding secondary. The diagnostic report w the values for each injector is also attached.
As you can see, these injectors were NOT within 1% of each other, essentially confirming what I was seeing on the plugs. Steve Schroeder recently went thru this same process w his injectors, altho I believe they are Bosch.
With the information of the flow test I was able to match up pairs of the injectors so that I was able to tighten the range of fuel flow(on a cylinder to cylinder basis) from 10ml down to a difference of only 2ml. when operating on all injectors. To me this also strengthens the argument for running all injectors during engine operation.
For idling operation, I was able to lower overall fuel and reduce the disparity between banks by 50%, albeit 1 ml.
As Steve will also likely attest, the SOTP is readily discernable. Engine smoothness and responsiveness, is pretty obvious. It feels like it has more power altho I don't have any measurements yet. Along w a change in calibration and spark plugs (NGK BKR6E @ .035"), knock is pretty much eliminated. The LT5 is a noisy motor, and when modified becomes even noisier so my objective was to eliminate any knock that was consequential, which appears to have been achieved.
I would make injector flow testing a basic part overall "tuneup" regimen. The $112 the flow test cost me, it was more than well worth it and would recommend it to anyone.

N. B. My injectors have 5k+ miles on them, so I am not being accusatory of any vendor's claims. Just saying that "trust but verify".

[jss06c6]

Good for you Dominic! Very happy for you and admire your persistence! I have made this part of my rebuild process on all ZR-1's as well as LT-1's. I too find the LT-5 to be much smoother. Our engines present a unique challenge by using 16 injectors intermittently (8 then 16 under load)..

Sent from my SM-G950U using ZR-1 Net Registry mobile app

[XfireZ51]

Another update on this. The re-balancing of the injectors has had clear impact on how the motor is running. Much smoother and quieter. However, I have found that in doing the necessary re-mapping of the VE tables, I have taken out about the 3-5% fuel, resulting in better mpg(>1.5mpg) and cleaner exhaust as evidenced by my exhaust tips. But since all fueling is dependent on the VE tables, I am also needing to re-visit my Acceleration Enrichment. Removing fuel from the VE tables is requiring me to put it back in for AE. I have been getting audible knock in certain circumstances resulting from an insufficient ?pumpshot? due to reduced fuel. So AE now needs to be not only increased, but also lengthened in terms of duration.
This also would affect the PE parameters for WOT, but I?d rather do that on a dyno.

[Erik]

Great info. Thanks!

I don't think it can be overstated the importance of properly operating injectors to keep an LT5 singing happily. Injectors are usually not something I worried about much in other cars, but the sure seem to matter a lot here.