What I'm getting at is that I think QSilver had improperly adjusted valves with old (weakened) valve springs, and a "bug" in the ECU's code allowed an overrev in neutral. I'm sure the ECU can handle WOT at the red line in gear (I've spent quite a bit of time bouncing on it), and I'm sure it can handle WOT in nuetral starting from a low RPM. However, I wouldn't be surprised at all if the ECU doesn't have time to project future RPMs in nuetral at WOT when it is already near redline. The crank accellerates much quicker at 8K RPMs at WOT than it does at 2K RPMs.

 

ASS or not, thems is the realities ....... what do you think a realistic sample rate is for the ECU on the crank sensor?

 

I'm sampling the crank sensor 68 times per rotation at 9K RPMs (770 times per rotation at idle), AND averaging the result over several rotations to get a "decent" RPM measurement. Of course, the results from my fuel controller are vastly superior to the results of the OEM ECU...

 

Yet another thought ...

An '02 ECU is less than $500 retail. My microcontroller, which spends MOST of its time calculating RPM and does nothing with injectors or spark plugs, IAT, OBDII, etc. was $200 from the manufacturer. I seriously doubt the OEM ECU is better at calculating RPMs than mine is.

 

Yet another another thought going back to your baseball example ...

If the ECU calculates the speed of the ball .04 seconds before the ball is released from the pitcher's hand and "smooths" the sample out with the samples taken previously ... does that blow your example?

 

 

As for the subject of does the PCM recall Max RPM in case of a overrev, yes and no.

The Early S2k PCM's (00-03) don't. The K motor'd Civic Si and RSX do.
I've got a pretty good source that tells me the 04 S2k PCM does. Anyone want to downshift a 04 at 75 to 2nd gear? That will anwer it for us.

All OBDII PCM's will store freeze data at the time a DTC occurs. IE, you have a misfre at 8900rpm, the PCM stores what DTC and what RPM as well as all of the other current PCM data.


I'm more than a little skeptical that the OEM PCM was unable to stop an overev under acceleration because the RPM's where allready at 8k. If that was the case, almost every s2k built would have bent valves.

The engine in question had a pre-existing problem. Spinning it to 9k found it.

 

I'm skeptical as well, but I still think it is possible that the ECU's rev limit is triggered by some low priority, infrequently checked calculation. I could be wrong, but I recall UL mentioning something about our ECU actually having fuel and timing maps extending beyond the rev limit.

 

The ecu does NOT record data unless a dtc is set.I know this because i had an s2 at the dealership with bent/broken valves and i wanted to see any stored data.I asked techline about this and they told me that the ecu is an older design and it does not have that (black box)feature.In this case i think there must have been previous damage from a misshift.

 

Of course a ball doesn't increase in speed after force ceases to be applied to it. However, look at a pitch this way.

OK, the ball is the car, the mass. The pitcher's arm muscle is the engine, the energy source. Arm speed is measured at the hand. The pitcher's brain is the ECU, controlling things.

During a pitch, the brain says, "Put max energy into the ball up to 100 mph." The arm is busy doing this, the pitcher's hand accelerates to 99 mph, and then the ball slips out of his hand - a missed shift occurs. The brain didn't expect this and keeps the arm working for a short time, continuing to push, but there is no resistance. The hand exceeds 100 mph.

You might say it doesn't exceed 100 mph by much, but imagine the same scenario with a 50 lb ball, yielding more like the same mass/energy ratio as in the car. I can see a very considerable increase in hand speed.