L5P - ECM Swapping Options

[kidturbo]

I'm gonna toss this here since it's currently a technical study. But hopefully this helps with stand alone and engine swaps going forward. So please add anything of value along those lines.

Last year I had the change to check out some "stand alone" L5P's being controlled by E35 ECM's. When GM released the L5P, they attempted to lock tuners out of the ECM. While it's now possible to have the E41 ECM modified, cost remains steep and tuning options are limited. And now GM is moving to encryption of canbus data in an effort to stop hacking of any onboard modules.

However when GM ditched the LML's piezoelectric injectors for the Denso solenoid injectors, they actually made controlling the L5P hardware with an older ECM much easier. No need for swapping or modifying the heads to use old style injectors. So when I saw the L5P engines running with a modified E35 ECM, had a major Duh Huh moment...

Now that's all great dandy if like me you spend your days sticking Dmax's in things GM never intended. Hacking up a engine harness is no big deal. But what if you just wanted to swap the E41 ECM to a E35B in a newer truck lets say. Would it work?? Besides the harness connectors needing modified, would it even talk to all the other modules on the bus. Well today I finally had an opportunity to answer that question..

Buddy of mine stopped by with his well used 2018 L5P, and said "what can we do with it?" I said you leave it parked here, I'll drop that powerplant in a boat... But he was thinking keeping the wheels attached might be easier. So I grabbed the laptop and couple hacking tools and started examining some raw high speed canbus data. At first glance, as expected, GM kept all the protocol and ArbID format from previous years. In short, the E41 ECM speaks same language as older E35B, LML, or even the E38/E67/E92 used on gassers.

Great news for all the stock chassis swap gurus. See no reason why you couldn't drop a stock L5P engine and harness into any modern GM chassis and have it run and drive just like GM built it. If ECM and TCM are happy with the BCM commands, then it will start, run, and dash should work as normal. However that's what I plan to focus on next.

With the data captured today, I was able to verify all the gauge and chassis data looks identical to all other model years back to 2007. Now I need to verify the Start/Run commands between ECM and BCM haven't changed. Next is layout harness pinouts and see what matched up to the older LMM harness.

Below is a screenshot of the L5P HSGMLAN data capture. I highlighted some common data like RPM, TPS and such that other nodes on the chassis use. I'm gonna do a side by side to an LMM truck data and see if anything clashes. Now as for change in actual sensors or engine hardware, we know they changed the VNT controler, but what else has anyone noticed that wasn't on a LMM?

 

[2004LB7]

That's fantastic. I was speculating back when the L5Ps where new that you should be able to drive those injectors with the LMM or LBZ ECM. If everything else was compatible then maybe making an adapter or swapping the connectors out would get you a tunable "L5P" using efi live. I think the important sensors and regulators such as the HPFP regulator could probably work with just a rescale on the appropriate tables. I'm glad someone is finally looking into it.

 

[buick455]

Been thinking about this. Great work so far. I personally do think being able to swap an L5P into an "older" chassis might be pretty slick if they do end up being as stout long block wise as some people are testing now.

 

[Cougar281]

Would probably have to use a LMM ECM. A LBZ ECM would probably run the engine fine, and might even communicate with the TCM, but anything that the newer trucks use LSGMLAN to communicate wouldn't work with the LBZ ECM as that doesn't speak LSGMLAN, just HSGMLAN and Class2.

Sent from my 2PZC5 using Tapatalk

 

[2004LB7]

^^ true, forgot about that

 

[Duramax racer 8]

Fantastic!!

 

[kidturbo]

Would probably have to use a LMM ECM. A LBZ ECM would probably run the engine fine, and might even communicate with the TCM, but anything that the newer trucks use LSGMLAN to communicate wouldn't work with the LBZ ECM as that doesn't speak LSGMLAN, just HSGMLAN and Class2.

Sent from my 2PZC5 using Tapatalk

Actually after LBZ, and not counting Serial Data Wake Up, the ECM & TCM only have HSGMLAN ports. LSGMLAN is mostly found in BCM links to other network nodes like doors and head unit. All the main stuff such as stability control communicates directly with ECM/TCM over HS. Which is good, because that limits the possible difference between chassis components that GM uses.

The only case where I've noted a completely different HS ArbID structure is in the Corvette. Up until 2014 I believe, they used a German made BCM that was part of the fuse/relay block. MB called em a SAM module. But for whatever reason, GM had to rewrite all their HS canbus traffic to match that BCM.. So depending on OS #, an E38 ECM might run one of two different protocols. Which is contained somewhere in the tune file, because I've swapped the OS on purpose just for shitz and giggles.

Along with the common GM diesel and gas ECM's, I have a full 2016 LS Camaro bench setup with ECM, BCM, TCM, IC, and few other nodes wired in. I haven't had much time to spend on it, but all the ID's match what I saw from the L5P truck.

I've also noted some of the LBZ HSGMLAN data is slightly different from the LMM. Though probably not an issue, the main problem I find with LBZ is some missing sensors like EGT probes that we know the LMM and L5P both uses. I'm pretty certain an LBZ would run a L5P in stand alone just fine, but as for a direct ECM swap, likely be much harder than the LMM.

 

[slowlml]

Thats great to hear. Definitely a cheaper option than the current tuning options out there. Unfortunately ive heard of some tune only l5p having issues cracking pistons. So i dont think they are as strong as people think. Guess we will wait and see if they start snapping cranks lol

 

[kidturbo]

Below is a quick side by side HSGMLAN ArbID comparison between a 2017 1500 LS powered truck

, and the 2018 2500 L5P truck.

As you can see, pretty much identical ID's. Besides adding some nodes/ID's, this same Hex ID and byte structure traces back to 2004 on the gasser side. The Chevy Volt differs a bit due to all the battery stuff, but it's pretty unlikely anyone is looking to drop a L5P into Volt/Bolt...​

 

[Yngdmax92]

So am I understanding this right?, That maybe one day a l5p could drop into my lbz truck and have a custom harness installed to run l5p injectors everything off from my lbz ecm?

 

[NC-smokinlmm]

You found the back door!

 

[PureHybrid]

More like: Well the door is locked, guess I'll sawzall a hole in the wall :thumb:

 

[2004LB7]

More like the door is locked so we will put a new door on it without a lock

 

[1FastBrick]

Thats Cool!0

I would like to know what you use to capture that kind of data.

 

[countrycorey]

I would love to see an L5P swapped for an LBZ and run off the LBZ electronics!

Sent from my SM-G973U using Tapatalk

 

[clrussell]

This just got more interesting

 

[JoshH]

That's cool and all, but with the recent ECM unlock price drop, I'm not sure it would be that much more economical once you factor in the work it takes to make it functional.

 

[kidturbo]

Ok we know for certain now that dropping an L5P into a LBZ/LMM chassis with said electronics can and has been done successfully. Unfortunately it requires good bit of hardware modification just to make it fire.. Or does it??

Issue is, besides moving the crank tone ring to rear of the crankshaft, GM also saw fit to change the tooth count on both crank and cam rings. Hope to learn the reasoning behind this modification some day. But they did and ya can't just poke a screwdriver through the pickup hole and knock a tooth off the wheel like old days. Seems they also phased em differently, so TDC is off now too. Well screw the pooch, end of story...

Not so fast. Since I've found a niche stuffing Duramax's into things GM never intended, you meet people along the way who also enjoy a good challenge. Just so happens one of those people who helped me with some custom OS cax file stuff in the past, rang me up today. He decided to buy a junk yard L5P to see if he could power it off an LBZ ECM a while back. Rather than tear it down and swap out the cam reluctor, and figure out how / where to put put an LBZ crank ring, he started tackling this issue from the ECM software side. With pretty decent success to date, like it fires the injectors.

So in couple hours today, I learned more about the L5P challenges than ever expected. BUT at the same time, came up with some ways to address this in a more scientific manner. Right now it's not really the tooth count difference holding up the wheels of progress, but more the overlaps and offsets from TDC that are different between LBZ and L5P. If you can read this, 647572616d617820726f636b73 then ya get the idea. We get use to the comfort of EFI or HPT doing all the heaving lifting. Cause if you only modified tunes in hex/dec, there would be far far fewer modified tunes in the world.

So basically the approach we are considering is two fold. I'm gonna put the ocilly scope on the crank / cam signals lines of my buddies L5P and record some data so we have the exact pulse and TDC of each wheel sets down to the us. Then compare that to saved pattern data I have from LBZ/LMM. Hopefully that helps in the quest to modify the OS with correct phasing. Second, gonna fwd both those logs to an electronics guru and explore the option of a plug and pray external pulse converter. Basically a modified harness and pcb that takes in L5P pulse signals, outputs correct LBZ signals. Been done successfully on gassers before, so no reason couldn't work on a diesel. That's the down low on what I know today..

 

[2004LB7]

duramax rocks

what do I win?

 

[kidturbo]

duramax rocks

what do I win?

Good work, sending you 5000 lines of code to find 4 bits of data we need..

Sent from a reclined position using Tapatalk