Question About Logging Misfire Data

[swingtan]

I've been logging the misfire PID's for the past few days and was wondering what we can use the information for and how do we interpret the data? There are 2 main PIDs that I've found...


GM.MFCSTAT: Misfire Current Status

This seems to be quite active on my ECM (E38). Normally this PID shows a value of "0", but at times the value jumps up in isolated spikes or "clustered" groups.
This PID always goes to peak values when the APP% is at 0% and the engine is not in "Idle Mode"
Quite often, at light throttle cruise, I get a lot of activity on this PID with values in the mid range ( min=0, Max=255 )


GM.MFCYLIDX: Cylinder Mode Misfire Index

This PID never seems to have a value of "0", usually is has low values of between 35-1500. Peak values are 65535 (IT Tech guys, take a moment to notice the significance there...)
Quite often the PID reaches MAX value and stays there, but so far I haven't worked out what is the correlation for this, but there are times when MFCSTAT is active, so is MFCYLIDX. At other times they seem to be quite exclusive.



While these PID's have been quite active, I've never throws a DTC for a misfire. I have had some light throttle surge though and when logging these PID's, it does show that something is going on. I have "tweaked" the DTC settings for misfires so it throws a code and freeze-frame, but the data is not telling me what cylinder may be at fault.

So, here is my question...

Is there any way of mapping these events back to a specific cylinder?

Simon.

[joecar]

Subscribing.

[swingtan]

Here's an image from one of my logs....

Note that that both MFCSTAT and MFCYLIDX can be isolated as well as both active at the same time. The indications though are very strong in indicating "near misfire" as a cause of light throttle surge. Perhaps this shows an incomplete burn of the mixture, causing a lows of performance but not a full misfire.

8723

Sorry about the size, but if I made it too small, it's loose too much detail. Note the oscilation in the RPM / Spark and even the commanded AFR ( as I'm playing with OL Fueling and load modifiers). This surge is present no matter what AFR I command though, I've been down as low as 13:1 and it's still there.

Simon

[5.7ute]

Simon, have you logged maf airflow during these conditions? (Since airflow not airmass triggers the misfire AFAIK)
You can get caught in a chicken/egg debate here. Is the misfire the cause of the surge? or is the surge triggering a misfire?

[swingtan]

Yes, MAF airflow was logged. For the first section, the readings were 15.05-17.07gm/Sec during the surge. The second section was 15.83-18.46gm/Sec, so in both case there was only a minor change in MAF airflow. If I expand the scale of the MAF Airflow chart so it has min/max values of 10/18, it doesn't look like there is any correlation in this case. While I'm pretty certain that the misfire data is an indicator for the surge, I'd really like to know if there was a way to pinpoint a specific cylinder. Otherwise it's a case of swapping coils, leads and plugs first, then Injectors to see if there is anything there. I'd really love to pinpoint it if possible...... I get a bit pedantic like that at times.

[5.7ute]

I haven't had the pleasure of playing with one of these E38 cars so I can't be of much use.(Apart from asking silly questions lol)
I gather then that the individual cylinder misfire pids aren't supported on these OS's.

[FeetDry]

On my 411 ECM, the GM.MFCSTAT is an high/low indicator. I don't see any mid range data, it is either low (0-6) or high (255). When it is low, the current MF count appears to be valid and counts toward the total count. When it is high, the current MF counts appear to be ignored.

I found this by using the DVT->Engine control->fuel inj control->cyl off
When I turn off a cyl, the MCSTAT goes high, when all injectors are ON, the MFCSTAT stays low.
It is an easy way to track/log when an inj has been turned off.

How the misfire detection system works:
If the MFCSTAT is high the misfire detection is turned off.
As long as MFCSTAT is low the individual "current cyl misfire count" (MFCYL#C) can increase and the MFCYCLE continues to count-up. When MFCYCLE hits its limit (100 for me), the individual MFCYL#C are summed up and this becomes the new MFTOTAL. Then the individual MFCYL#C and MFCYCLE are reset back to zero and the process repeats itself.