Engine/Transmission (1998.5 - 2002) Inside the ECU

[Jdonoghue]

Jim,



How did you figure out this neat little trick? Was it by examining the boot loader software?



Great work!



John L.

Yep. I have a complete disassembly listing of the boot loader and the main software, with nine months worth of comments.



I also had to do a bit of tracing on the ECU hardware itself, namely the QADC inputs and multiplexer lines. Eventually I figured out that channels 2 and 6 were being looked at by the boot loader when it was powered on, right before it checked the '60 0d'.



Jim

 

[NoSeeUm]

I agree, but it would make even more sense if it was OAT instead of IAT.



Paul

I had thought that as well, but OAT does directly effect IAT and OAT does has no control in the ECU for timing.



Under most driving circumstances, I typically saw IAT as OAT plus about 40F. This delta T would be even less with a cooler OAT. So IAT under those conditions is most always less than 100F.



If you look at Jim's charts, when IAT is less than 100F and boost is low it seems to fall right in the red area constantly. Because the performance box never lets the ECU see higher boost as what it would expect to see at higher engine RPM / load. This prevents the ECU from retarding some timing based upon the boost vs RPM table until RPM gets pretty high, as well as basically a static timing advance from the boost vs IAT table.



The use of IAT for timing makes sense to me, because you could derive a table using IAT that has its foundation on the calculated combustion air temperature using the compressor ratio. Additionally, the IAT is compared to boost which also seems like it would be a more accurate / self correctly calculation in some fashion.



Fascinating stuff to say the least.



Jim

 

[JLandry]

Hi Jim,



I have a few more questions for you if you don't mind...



I'm under the impression (possibly mistaken) that you can't download the parameter software from Cummins ECM without "hacking" the ECM code to get it... I thought this is why you had to open up your ECM.



If my impression is correct, how is it possible that performance programmers (like Hypertech, Smarty, etc. ) can claim to be able to return the ECM back to the "stock" programming software after uploading performance software?



Is it possible they don't actually download and store the stock ECM software, but instead upload flash files they obtained from OEM's? If this is the case, then is it possible you could find a different (possibly updated) OEM flash file in your ECM after returning it to stock programming with a performance tuner?



Just thinking...



Thanks,



John L.

 

[Jdonoghue]

Hi Jim,



I have a few more questions for you if you don't mind...



I'm under the impression (possibly mistaken) that you can't download the parameter software from Cummins ECM without "hacking" the ECM code to get it... I thought this is why you had to open up your ECM.



If my impression is correct, how is it possible that performance programmers (like Hypertech, Smarty, etc. ) can claim to be able to return the ECM back to the "stock" programming software after uploading performance software?



Is it possible they don't actually download and store the stock ECM software, but instead upload flash files they obtained from OEM's? If this is the case, then is it possible you could find a different (possibly updated) OEM flash file in your ECM after returning it to stock programming with a performance tuner?



Just thinking...



Thanks,



John L.

Hypertech, Smarty, etc. all have flash files they obtained from OEMs. If you're like me, and your truck has it's original flash from when it was manufactured (never been reflashed by a dealer), using any one of those 'programmers' will cause the latest,greatest stuff from DC to get loaded once you return to stock.



FYI, you too can get DC's latest-and-greatest flash files with a subscription to http://www.techauthority.com - you just need a DRBIII, a StarScan, or some kind of home-brewed software and cable :-laf to download it into your truck .



Jim

 

[JLandry]

FYI, you too can get DC's latest-and-greatest flash files with a subscription to http://www.techauthority.com

Or you can purchase a Chrysler "ITIL/ISIS" DVD on eBay, which contains *every* Chrysler computer module flash file available... and much more! :-laf

... to download it into your truck.

Technically speaking, are we "downloading" or "uploading" flash files from a programmer to the ECM or PCM?



John L.

 

[Jdonoghue]

Or you can purchase a Chrysler "ITIL/ISIS" DVD on eBay, which contains *every* Chrysler computer module flash file available... and much more! :-laf



Technically speaking, are we "downloading" or "uploading" flash files from a programmer to the ECM or PCM?



John L.

I guess I need to take a peek at eBay more often.



Downloading/uploading? Programming? I would vote for 'uploading', because you're sending a file from your programmer to the ECM/PCM.



Jim

 

[Jdonoghue]

More timing information

I've been driving around with the laptop quite a bit, and have figured out some more stuff related to timing.



First, a correction: the timing adjustment looked like it set a base value of 512, which equals a 0* advance. This is wrong - it only does that when in a certain mode (possibly the 3-cylinder idle mode). Normal operation I never see that flag set - this means it allows adjustment values starting at zero. Remember, anything less than 512 is considered a negative value (retards timing).



This solves the mystery of the low timing numbers I was seeing on the highway under light to moderate acceleration - what good is less than 12 degrees of timing at 2200 RPM? And more importantly, where was this 'retarded' value coming from?



There were two tables I had labeled 'UNKNOWN 2' and 'UNKNOWN 3' - at first glance, they appeared to be used in the code under cold engine operation. This was not true. They are used all the time, and are what the initial timing adjustment value comes from. It is within these tables I found the magic '-4. 99' value I was seeking.



Which table is used depends on some unknown condition. If the engine has been in a steady state for a certain amount of time (idling in the driveway for 10 minutes, cruising for 10 minutes on the highway), table 2 is used. Otherwise, table 3 is used. The tables are slightly different. I'm sure I'll figure out exactly how the code decides which table to use.



More importantly though, is how to take care of this retarded timing issue. I decided to alter both tables, eliminating almost all of the timing retardation. Then I went on a long highway drive for an unscientific test.



Performance: It seemed to require less throttle to get up the small hills (overpasses, there aren't really any hills around Houston).



Economy: Tracking actual fuel rate using the laptop, less fuel was used this trip compared to last time (same traffic and weather conditions). Overhead 'lie-o-meter' went from 20. 6 mpg to 21. 2 mpg (really doesn't mean much, but shows an improvement). What I really need to do is go on vacation, then I'll get some real mpg numbers.

 

[Cummins724]

What I really need to do is go on vacation, then I'll get some real mpg numbers.

You are doing great work, vacation is in order after everything is figured out. :-laf

 

[Jdonoghue]

More timing, again

You're all probably sick of my posts about timing, but right now, with no EGT gauge and stock injectors, timing is really all I can play with right now. If/when I figure out how to reprogram the vp44 or make my own 'wire-tap', I'll get more into 'rolling the coal', although I can do that pretty good, if I lug it down low, with the basic fueling increases I've already done.



I've added some timing information on my project page Dodge/Cummins ECU



With what I've learned about the ECU timing generation, and what I've read about various add-on boxes (and some of the troubles with them), I think I've got the timing stuff figured out. Eventually I'll get some kind of flow-chart or diagram that describes how it all happens.

 

[Spunkys Ram]

This is great stuff. I love reading about all the timing information. Keep up the great work.

 

[Gary - K7GLD]

John, it took a couple of years in our APPS testing and playing around, to eventually arrive at a rock-solid fix that totally eliminates the expensive and failure prone DC APPS. And THAT fix was greatly motivated by info that another TDR member provided late in the game!



Keep at it, every tidbit of info added by you and perhaps other contributors is a plus that may well open up FAR better ubderstanding and options regarding the ECM than what we mow have.



And no, we're NOT bored with your efforts or info!

 

[Jdonoghue]

I finally decided to replace the original ECU on my other truck ('01 auto 4x4) with a spare and confirm my identification of the torque management stuff.



The good news is, I was right on with the identification of the code sections and tables responsible for the tq. management. There is a single configuration bit you can set to '0' to turn it off completely (and in fact, the manual transmission files have this set to zero).



So, one can either modify the tables to their liking if they want to retain some of the tq. management functionality, or just set that bit to zero to bypass it altogether.



I left it bypassed for now, I just hope I can control my lead foot enough to last until I get some transmission upgrades.