And first fire video. Still has that darn fuel knock, so ya know it's alive.
Source: https://youtube.com/shorts/EC7M6VOtsJM?si=lUx_pQa6aYmhhJ-k
Sent from my SM-S901U using Tapatalk
Sent from my SM-S901U using Tapatalk
LLY Ficm
- Thread starter mplecha
- Start date
For first run, captured some decent data. Didn't sort out the firing order yet, as it didn't seem to complain, and I managed to capture some good J1939 canbus data. Have a pretty good understanding of what the FICM is saying to the ECM while it running now. Next we'll start fiddling with some control wires and see how the FICM reacts.
Before calling it a night, couldn't help giving it one good romp to the redline. Which with APPS stuck setting at zero %, took a bit of trickery. But if you turn the key on, while feeding it a crank / cam signal, which says it's spinning 8000Rpm, it will start firing the injectors instantly. And continue firing until you drop below set idle RPM again. So I tried to spin 9000R's and it seems to have a hard shut off of around 8500R. Much higher than the LBZ/LMM units. In the example below, ECM stopped processing and regenerating the outgoing crank speed signal at same moment it stopped generating injector control signals. So either it couldn't read my emulator crank signal being received, or the ECM shut it down on purpose. Will investigate more later. But either way, will certify stock LB7 hadware is good for 8500RPM if anyone wants to give one a spin...
More Skittles Porn
Before calling it a night, couldn't help giving it one good romp to the redline. Which with APPS stuck setting at zero %, took a bit of trickery. But if you turn the key on, while feeding it a crank / cam signal, which says it's spinning 8000Rpm, it will start firing the injectors instantly. And continue firing until you drop below set idle RPM again. So I tried to spin 9000R's and it seems to have a hard shut off of around 8500R. Much higher than the LBZ/LMM units. In the example below, ECM stopped processing and regenerating the outgoing crank speed signal at same moment it stopped generating injector control signals. So either it couldn't read my emulator crank signal being received, or the ECM shut it down on purpose. Will investigate more later. But either way, will certify stock LB7 hadware is good for 8500RPM if anyone wants to give one a spin...
More Skittles Porn
Yeah that looks alot more like what I have for wire colors. And the #'s match up.
Pinout I have show a LT BU / WT and two different pins as Injector Control #6. The other diagram shows like 3 to 5, 1 to 3, yet same wrong wire colors as in the pinout.. I was confused just trying to figure out how they got so confused..
Pinout I have show a LT BU / WT and two different pins as Injector Control #6. The other diagram shows like 3 to 5, 1 to 3, yet same wrong wire colors as in the pinout.. I was confused just trying to figure out how they got so confused..
I'm a certified idiot with the worst education that my poor state could provide me with so none of this makes a lick of sense to me but I just wanted to say keep up the good work, boys! Really looking forward to seeing this all come to fruition.
Thanks for the vote of confidence. I, for one, enjoy hacking around on old OEM hardware from about any manufacturer. Especially if I don't have to depend on it for a ride to work tomorrow..
While laying all this out takes a lot of time, it's fun just to share in these types of science experiments. A decade ago, I was scared to death of breaking these expensive freaking things. Yet glad to say I've overcome that paranoia, with many years of common rail therapy.
While digging through a box for wire connectors, I found this 10 pin to 6 wire piece with Pedal tag stuck on it. Shares same style connector to our LB7 APPS. Like it's was an mid model year Adapter Cable from something GM. ??
Sent from my SM-S901U using Tapatalk
While laying all this out takes a lot of time, it's fun just to share in these types of science experiments. A decade ago, I was scared to death of breaking these expensive freaking things. Yet glad to say I've overcome that paranoia, with many years of common rail therapy.
While digging through a box for wire connectors, I found this 10 pin to 6 wire piece with Pedal tag stuck on it. Shares same style connector to our LB7 APPS. Like it's was an mid model year Adapter Cable from something GM. ??
Sent from my SM-S901U using Tapatalk
Made some more progress today on the Kodiak bench setup. Switched the Arduino Emulator to run off 12v power from the chassis harness fuse block. Was having issues with ECM not powering down by switch, when the emulator was being powered from a separate USB supply. Ground loop something was enough to mess with it. Now it powers up and starts sending crank/cam signal with ECM switched on.
Verified all of the control wires between the ECM, FICM, and bench scope match up to our latest schematics. Captured some more data using the Kodiak ECM, but was still unable to get it firing above a 640 idle. And when I added some throttle, boost, and MAF range to the equation, it still stops firing at about 650R's. Now know it's not just trying to maintain idle and killing the fuel. So I read DTC's while holding above those RPM settings, and noticed it poping a new VATS related code. P1626
Did a little reading, but nothing about this Kodiak wired version directly. Looks to me like it's either Circuit 429 Off/On/Start needs to remain powered on while my switch "key" is in the Off position. Or it's unhappy that it never gets a start signal from the starter relay. I don't have crank / start relay connected. Figured we would just push start it.. lol. Tomorrow I'll try the other ECM with VATS disabled and see how it goes. But for now, luckily this one seems to operate correctly, so long as you don't exceed 640Rpms.
Which brings me to our relevant data for the day. As mentioned, I verified the control wire #s to each physical cylinder location in our firing order. Looks a bit strange on a scope, but our factory firing order of 1,2,7,8,4,5,6,3 is what I see below. The biggest catch is the timing offset. My math says we need to change our #1 TDC tooth back one notch to start at #25 rather than #26. But when we do that, the ECM supplied #1 timing PID, and our captured injector pulse timing on control lines starts to line up... Somewhat.
With some fuzzy math, and knowing 1 tooth on the crank wheel equals 6° of crank rotation, one can start to see where the ECM consider our injector Pulse Width starts. In captures below, we are running 615Rpm, with 10° of timing, and a 1700us main pulse, according to EFI. When the timing marks didn't line up for Control #1, I used EFIlive to kill injector #1. Which it did by stopping that control #1 signal from ECM, and verified what I was seeing on the scope. Control 1 is Cyl 1..
Above I use Control 2 to measure distance in teeth to known TDC. Firing 90° after #1, it's just easier to measure the gap to TDC.. IMO, there is a 90° offset or delay built into our ECM side programing, which is then corrected by the processor in our FICM. And can't forget the roughly 50ms delay between actual crank signal from front cover pickup, and the replicated Crank Speed output signal from our ECM to FICM. It's seems our little box isn't plumb dumb after all. At least it can keep time, and do basic math.. That's everything interesting I know today.
Verified all of the control wires between the ECM, FICM, and bench scope match up to our latest schematics. Captured some more data using the Kodiak ECM, but was still unable to get it firing above a 640 idle. And when I added some throttle, boost, and MAF range to the equation, it still stops firing at about 650R's. Now know it's not just trying to maintain idle and killing the fuel. So I read DTC's while holding above those RPM settings, and noticed it poping a new VATS related code. P1626
Did a little reading, but nothing about this Kodiak wired version directly. Looks to me like it's either Circuit 429 Off/On/Start needs to remain powered on while my switch "key" is in the Off position. Or it's unhappy that it never gets a start signal from the starter relay. I don't have crank / start relay connected. Figured we would just push start it.. lol. Tomorrow I'll try the other ECM with VATS disabled and see how it goes. But for now, luckily this one seems to operate correctly, so long as you don't exceed 640Rpms.
Which brings me to our relevant data for the day. As mentioned, I verified the control wire #s to each physical cylinder location in our firing order. Looks a bit strange on a scope, but our factory firing order of 1,2,7,8,4,5,6,3 is what I see below. The biggest catch is the timing offset. My math says we need to change our #1 TDC tooth back one notch to start at #25 rather than #26. But when we do that, the ECM supplied #1 timing PID, and our captured injector pulse timing on control lines starts to line up... Somewhat.
With some fuzzy math, and knowing 1 tooth on the crank wheel equals 6° of crank rotation, one can start to see where the ECM consider our injector Pulse Width starts. In captures below, we are running 615Rpm, with 10° of timing, and a 1700us main pulse, according to EFI. When the timing marks didn't line up for Control #1, I used EFIlive to kill injector #1. Which it did by stopping that control #1 signal from ECM, and verified what I was seeing on the scope. Control 1 is Cyl 1..
Above I use Control 2 to measure distance in teeth to known TDC. Firing 90° after #1, it's just easier to measure the gap to TDC.. IMO, there is a 90° offset or delay built into our ECM side programing, which is then corrected by the processor in our FICM. And can't forget the roughly 50ms delay between actual crank signal from front cover pickup, and the replicated Crank Speed output signal from our ECM to FICM. It's seems our little box isn't plumb dumb after all. At least it can keep time, and do basic math.. That's everything interesting I know today.
Last edited:
640rpm sounds low, I thought idle was 680? And 1700us sounds high. Is it in a cranking quantity mode?
Been mostly focused on the class 2 data, and why it wouldn't exceed 600. But I did record some data on the J1939 side. As noted several page back, still just the 2 lines of interesting data. And from what I can tell, first ID is trouble codes related to crank signal status, while second ID is injector related data, and some couple bit clock source.. From the DTC list above, the ECM does use that data to identify possible issues downstream of the FICM. IE: P2146
As for the chip itself, I outsourced that job to someone with a more appropriate skillset..
Last edited:
Best I can tell, it's in cranking mode and VATS doesn't kick in until it reaches set idle speed. Reading up on the code, typically a fire and die issue in the vehicle. However in this case, I have control of the Fire side, so can keep it alive under set RPM. This thing will actually start firing at like 20RPM. So from the ECM standpoint, it has no idea there is no physical engine attached. Soon as it sees a good crank / cam signal, it starts spraying. Once it hits that VATS trigger RPM, it stops.
Last edited:
Changed to the ECM @Cougar281 supplied, and it fires right up now. Was VATS related. .
Personally think I'm getting it tuned in, pretty close to stock??
Source: https://youtube.com/shorts/VnKRzZ18JSc?si=io5wx7v9hWhFhE_J
Sent from my SM-S901U using Tapatalk
Personally think I'm getting it tuned in, pretty close to stock??
Sent from my SM-S901U using Tapatalk
Ofcourse. Was just waiting on an, accelerator....
Sent from my SM-S901U using Tapatalk
Sent from my SM-S901U using Tapatalk
It's funny... it seems half of the Duramax combustion noise is actually the injectors lol. What's it like cranked up to 8k?
I just spent the last hour or so reading through this entire thread. I have to hand it to you guys @kidturbo, @2004LB7 , and @1FastBrick. The combined knowledge y'all have in vehicle electronics is simply mind blowing! Just wanted to say keep up the good work and I hope you can get this project working. Would be great for the Duramax community to have other options for Ficems(as I guess we're calling em now) than overly priced rebuilt units. I wish I had some knowledge to share on the topic, but it's over my head. Best of luck.
Some little steel caps in place of the plungers, plus a little Gorilla tape to hold them in place, it replicates the sound, with much less load on the system than in real life. I actually added the Gorrila tape to quiet them down.. With just masking tape, they sounded like an old 7.3L
I haven't been able to get this ECM spinning above 2000R's yet. Having the throttle pedal assembly from @1FastBrick all calibrated in, I would have thought we could at least hit the factory 3600R rev limiter, but so far no luck. It keeps shutting the fuel off, even at WOT..
With this new ECM came new problems. While VATS issues are resolved, seeing some missing data now on the J1939 side, and worse, on the Class 2 side. As noted on first power up, have some normal LB7 PIDs that are not showing up in EFIlive stream. Have switched it to PWM signal, slow speed, but no luck. The ECM seems to respond to changes on these sensor, however I have no way to read the values for Fuel Rail Pressures, BARO, Fuel Rate, and the MAP is always stuck at 46KPA..
So could use a little help from the masses on this one. What am I missing here? Works fine on the Kodiak ECM, so I am going to hook up some other tools and see if is possible EFI bug.
Don't see any DTC's set now that would cause this issue. The MAF DTC is from it spinning 2000R while being set at 50gs output. It works, and I can read the PID. However this P0093 does have me a bit worried, being it''s running in my basement...
But overall I can get usable timing an PW on the injectors now. Using the throttle now allows us to vary those parameters in the usable RPM range. So I will pack up that Kodiak ECM, and send it over for further inspection. Hopefully we can get this one sorted, and then load a good Hot tune in it...
Soon as get the PIDs sorted, plan to hook the APPS to replace the POTS on that Arduino emulator. Then it shall rev by throttle input..