SDFever

Knock retard parameters are adjusted in this area. They can be adjusted so ignition retard actions are more or less aggressive when knock is detected. Adjust, as desired, based on the applications.

Knock Sensor Tuning

Holley EFI systems can receive input directly from either Resonant (1 wire) or Non-Resonant (2 wire) knock sensors. Make sure you properly configure them in the Ignition Parameters section. The following must be configured:

• Sensor Type – 1 wire (Resonant) or 2 wire (Non-Resonant)
• Number – 0 if none are used. 1 or 2 if knock sensors are being used.
• Bore – Engine Cylinder Bore Diameter (inches). Used only if a 2 wire sensor is used.
• Sensitivity – Used to tune/adjust the sensor signal so the ECU can properly detect when real knock is present.

Knock sensor signatures vary and depend on the sensor itself as well all facets of the specific engine and chassis component design. Just changing the engine mounts (or many other components) can change the knock signature of an engine. This is one reason why it is always recommended to program in the “proper” ignition timing curve. You should not fully depend on knock sensors to pick up all knock events as well as always trust that you are not getting false knock triggers.

Due to the fact that this EFI system is designed to use any type of Resonant and Non-Resonant knock sensor on any engine and chassis variation, a tunable parameter must exist such that the ECU can recognize what is “real knock” and what is “false knock”. This parameter ultimately must be “tuned” by the end user. This parameter is the “Sensitivity” parameter which is found in the System Parameters ICF under “Ignition Parameters” and “Knock Sensors”. Increasing the value of this parameter DECREASES the point at which the ECU will recognize a certain signal level as knock.

• If you are seeing “false knock”, DECREASE this parameter.
• If you are not seeing “real knock”, INCREASE this parameter.

There are parameters that can be logged in order to assist with the proper tuning of the knock sensor. The only caveat is the user must know when “real knock” is occurring. If you have open headers, etc, this may be impossible to audibly hear knock. Knock occurs before it can be audibly heard, but if you can audibly here it, you know it is occurring.

There is a parameter that can be viewed in the data monitor and logged to assist in this tuning called “Knock Level”. This knock level is a scaled knock sensor reading level that is dependent on the Sensitivity parameter entered in the knock sensor setup. When the “Knock Level” value is over a value of 80, the ECU sees this as a knock event and will retard timing (whether a real knock event or false).

• If you are getting “false knock”, DECREASE the sensitivity value in the Knock Sensor Setup.
• If you are NOT sensing “real knock”, INCREASE the sensitivity value in the Knock Sensor Setup.

NOTE: You want to see the “Knock Level” value go higher than 80 when you have real knock occurring, and stay below 80 when there is no actual knock present.

Perform data logging to view these parameters as it is almost impossible to view them real time.

A sensitivity value of 35 is a good starting point for some applications.

SDFever

They have. The Holley System does this nicely but even that should not be left in full learn all the time cause it too can create problems. But using it as an inactive monitor works nicely. Many guys think they have a leak-free exhaust cause it hasn't cause any visible problems. But i can tell you that i bought a brand new set of headers that leaked first day. Went to our welder and had them repaired. I took my sensor out of my tail pipe yesterday after many hours of running in the salt; it was bone dry. There was a certain amount of water that just became "acceptable" to me until i realized that you can have a dry system if it's in good order. This also does not take into account condensation which is heavy in marine engines. More than you might ever think.. The sensors in road cars is far upstream though. If you were to place those sensors closer to the back (which would be similar to a boat) you would be replacing them far more often.

Rage

Would you please explain how and with what specific equipment you are able to observe live and historical knock codes on a PCM555?

SDFever

Simple -

Merc CDS and/or Diacom

Historical is saved perm in freeze frame and any live faults show as current trouble codes. When you pull the data in freeze frame it shows all vitals at time of inception.

I promise if you have a lean condition using that box the knock system will not save your azz -

Have you still not purchased a scan tool after all this time?

SDFever

The name of this thread should be changed to something a little more positive imo. It's not a bad platform for many owners. It's certainly less-than-desirable perhaps if you started with no engine or accessory drive parts at all.

Not to mention the fact that there are many owners out there with Raylar parts that now have a way to get a better block once theirs finally lets go. You now have 2 really good stroker options and a bore size that will rival anything out there.

Take your 496 to 588 or even 632 if you feel like it and you do save repurchasing a ton of parts and all the problems that come with it.

Rage

I got the Diacom in 2008 and borrowed one before that. I have used it extensively over the years for both data logging and reading trouble fault codes. Early on I looked for a knock/detonation fault code generation capability from the PCM555 and found none. Attached is a Mercuiser 496 Fault Code listing. The only codes I see on knock is the Codes #10 & #11 which I suspect is to report a sensor integrity defect.

As an alternative I watched the data logged PCM555 spark timing for any ECU initiated retard that would indicate that a knock event had occurred. I believed then that if no retard had occurred there had been no knock sufficient to exceed the ECU's trigger threshold. Then Dustin Whipple told be that the PCM555 data logged spark timing was only the base timing from the spark table and the ECU's retard of individual cylinders is not recorded.

So I am very interested in how you read a PCM555 fault code reporting a knock/detonation event.

bobl

Let me clarify. My comment about 13.5 was referencing light loads. I never tune a naturally aspirated engine leaner than 12.5 at WOT.

Bob

SDFever

That's what I'm yacking about.. There is none. The sensors are there and they are hooked up but it will let your engine burn to the ground. The knock codes I have seen (as stated earlier) were only at failed cranking speeds along with cam, crank etc.

Rage

The 13.5 AFR was in answer to my initial question and the reason for my starting this thread. ie Question. What is the leanest naturally aspirated engine AFR that is safe from detonation damage from idle to 2200 rpm? Plane out is ~2200 RPM. I also noted in the initial post that my significant MAP/engine load increase begins at ~4000 rpm.

Rage

OK. That clears that up. Thanks.