CASE problems P0315

Hi Guys, Ive been having problems with this lately which is weird..
First of all I mostly do LS conversions and make from scratch brand new 6L, LS3, and LSA harnesses right here in Perth, West Australia, to suit vehicles such as Hotrods, any sedan, you can squeeze a LS into, Patrols, Land Cruisers etc and have supplied several hundred of these including, either customer supplied e38's that I have reflashed with decent start tunes in, and others where I supply the e38 ecm.... meaning I am very experienced with these ecu's.
Occasionally I get to see the above CASE relearn required, which is impossible (that I know of) in a conversion to run the procedure and clear it, but I have performed the procedure with success in a few in original Holden Commodores, and these were mainly approx 2007 models with YPRZ broadcast code... the 12597121 service number. I have also managed to clear it by doing a Full Flash again on the bench.....
However, lately, especially with e38's becoming rare here in Australia, we have been importing from USA used and serviced e38's... and have been regularly getting this code in later model 12633238 2010 onwards... and the thing I have noticed is, this will set on the bench, and/or on my test engine bed without even running the engine or trying to start it.
So the question is what sets this condition???? as its clearly not the Crank Angle Sensor.... and is there a way to clear it??? I have tried Full Flashing again but no good.
I realise not many of you would have seen this but hoping a couple of you more experienced or even Ross may know.
I have 6 e38's in front of me that have done this in the last few days...
Regards Al

The CASE (Crank Angle Sensor Error) is data that is stored in the NVRAM area of the ECM (an area of memory that remains intact even when battery power is disconnected). The data contains information required by the ECM to help it decide if a spark event misfired.
When the CASE learn procedure is performed, the ECM records the slight variation in the crank acceleration for a non-misfire v's a real misfire. It uses that data as a reference when attempting to detect actual misfires.

If that data is determined by the ECM to be invalid then it will set that P0315 code.

I don't know what criteria the ECM uses to determine if the data is invalid. Just taking a wild guess, it could be something along the lines of: When checking for misfires using that data, if every spark event appears to be a misfire but the engine is still running (i.e. no actual misfires, or only a very few) then the data is probably invalid.
If/when the ECM is swapped from one engine to another, the CASE data stored in that ECM (which matched the previous engine) may not (probably won't) match exactly with the new engine and the ECM may determine that the data is invalid and set that code.

The only way to clear that code it is to successfully perform a CASE relearn. Re-flashing should not make any difference to that code being set or not.

What is stopping you from performing the CASE relearn?
Are you using V8 bi-di controls? They have better error reporting (then V7) and may display the actual problem that is preventing the CASE learn from happening.

I.e. I just tried it on a bench unit and I see this error:
Attachment 24039

In this instance, the ECM is configured for an auto transmission but no TCM is connected so the CASE learn fails since the ECM does not have any input from a TCM telling it that the trans is in Park/Neutral.
There are other criteria such as the ECT must be at operating temperature, the hood (bonnet) must be closed - assuming the ECM requires a sensor to detect that, the handbrake must be applied (or possibly the brake pedal must be depressed, can't remember).

Just FYI...
I'm pretty sure you have about 5 seconds after clicking on the CASE relearn button to start the engine speed ramp-up to 5000rpm. Once you feel the engine misfire (at around 4500-4800rpm) let off the accelerator and switch off the ignition for 30 seconds to allow the ECM time to store the results in NVRAM.
I think also that if the CASE fails, you must cycle the ignition before attempting again, as you can only make one attempt per ign cycle.
Sorry if I sound vague but it's been years since I worked with the CASE learn stuff.

Regards
Paul