E38 - 24x Reluctor - 1x Cam - $$ Reward

[Tre-Cool]

Throwing this out there for the clever ducks with PowerPC knowledge.

Basically I'd like to get 2 x E38 Oses capable of running the ls1 24x crank reluctor. so 12624402 (2009) & 12639270 (2011) to support the early/late TCM's.

I've found that there are switches in the factory code that allow for multiple crank / camshaft options & while a mate has tested that the 1x cam sensor seems to work. we get nothing when trying alternative crank solutions.

Internet posts alluded to a guy called Jesse from wait4me performance getting it to work back in 2013 or so on e38/e67 ecu's, so i reached out to him but unfortunately he lost a lot of his stuff from a hard disk crash years back & no one was really interested in it at the time, so he never got back to it, but the theory is that while the switching is there, the pattern info is not actually in the underlying OS binary. (not confirmed)

Now to offer some incentive, I'm willing to offer some of my disposable income up to 2k AUD to help get it done. (Money talks, bullshit walks) I wont be keeping this info secret, it will be free for dissemination & potentially patching into other oses via something like Universal Patcher. My main purpose for wanting this done is to give another alternative option for conversions & being able to use the 6l50-90 gearbox's since i absolutely despise the 4l60.

I will obviously share the relevant factory calibration locations here to get you started (or can just give you ghidra repository access) as an example here are some of the options i have mapped into an efilive cax file for the 270 os since i primarily use that as my main tuning tool. (Units may be incorrect for some of the below until i correct)

e38-crank.PNG (94.17 KiB) Viewed 1897 times

[Cincinnatus]

4l60 doesn't deserve any hate, but i'm all for getting an e38 to run 24x.

[BillNobody]

I have an Arduino sketch that converts a 24x/1x to 58x/4x and vice versa if you are interested.

[Highlander]

The effort to do this is a lot more than 2k. Way more.

[antus]

I didn't want to jump in this thread too early, because I don't know what I don't know, but since nobody has jumped in with a solution yet here are my 2 cents.

the theory is that while the switching is there, the pattern info is not actually in the underlying OS binary. (not confirmed)

Generally speaking, because these are not aftermarket PCMs, I don't think the code is written in a way that that has a hardware abstraction layer for the trigger pattern where there is a set of timing data for the trigger that is looked up and used and translated to something common for the rest of the OS. I think it's more written in to the code to expect a certain pattern with certain timings. The switch might be a jump between two code paths in the relevant place or places. If that is the case (I am assuming the E38 functions similar to the previous ones which I am more familiar with) then it's a case of finding any of the code that has the trigger pattern baked in to it, checking the type switch (if that test is in the right places, adding it if its not) and then jumping to some unused code space and processing the other type. However because its timing related it's possible you find the code isn't executing at the right time for what's needed and you might have to make some heavy changes to the OS internal timings. If that's the case, then it would be very difficult without source code.

Anyway I don't know for sure this is the case in the E38, and it wouldn't be the first time someone with fresh eyes finds a way to do something amazing, so if anyone still wants to have a go.. don't stop because of my currently unfounded thoughts.