Help! At the end of my rope
- Thread starter DarylF
- Start date
Flash the ECM/TCM back to GM files and see what it does. As others have said the ECM has nothing to do with the shifts in terms of when they occur, that is controlled 100% by the TCM and the tune in it.
How did they determine the ECM is commanding a 6000rpm set point during the shift? I can’t say I’ve ever seen that one occur but it certainly isn’t out of the realm of possibility.
How did they determine the ECM is commanding a 6000rpm set point during the shift? I can’t say I’ve ever seen that one occur but it certainly isn’t out of the realm of possibility.
This is not true...the tcm will command a target for the ecm to try and hit. The ecm calibration 100% does have an effect on shift quality(speed and rpm) and lockup.
while it's true the ECM has an effect on the quality of the shift based on defueling and information given to the TCM about fueling. the actual shift points, RPM, MPH, etc are all in the TCM. the ECM has no command over this. the ECM doesn't request anything as far as shifting.
the TCM, if certain parameters are met will send a request to the ECM to defuel so it can perform a shift and expects a certain amount of fuel (torque) during this. if the fuel amount or torque is off during the shift then the shift quality suffers
really the main information sent to the TCM is the throttle position and MPH. the TCM uses this information to look up in the tables when it should shift. All the shifting RPM, throttle and MPH tables are in the TCM. the ECM doesn't tell it when or how to shift
If your fuel quantity and torque tables are not set up properly to jive with what your tcm is commanding, then you will have horrible shifts.
I should add too, that low end torque derived from tuning and/or truck setup will change how different torque converters respond and couple, which also then may require fine tuning to get the truck to behave correctly.
The TCM can only effectively control shifts up to a certain rate of acceleration as well, which it, by itself, cannot control. Therefore, the truck setup and engine calibration play a very large part in the effectiveness of the TCM to do its job. A great shifting Allison, I'd say, is 60% physical trans setup, 30-40% engine calibration quality, and 0-10% tcm calibration quality.
There are several folks who have long touted that no Allison ever needs to have the TCM file adjusted to shift great, but that is just straight ignorance. It is usually not needed, sure, and a lot of people do more harm than good with the tcm adjustments, but changes are certainly needed in some instances.
Last edited:
That is 100% correct, the ECM does have an effect on the overall shift but the ECM does not command/control the when part of the shift.
If the truck is seeing an RPM command of 6000rpm during the shift there is certainly something funky going on. The easiest solution, in my opinion, is to flash both the ECM and TCM back to clean GM files and see if the issue persists.
Well OP if you’re convinced that this is hardware issue vs software replace the “janky” ECM and put whatever tune you have back in it. Then see what happens.
I think you are missing the point I'm trying to make. if you go back to post 21, you'll see that the OP is stating that the" ECM is commanding a 6000 rpm shift"
my reply is that, that command comes from the TCM and not the ECM. post 24
this is where I think you misunderstood what we where getting at. we know quality of the shift is dependent on the ECM. but that is not what we are trying to argue. it's the part where the OP stated that the "ECM" commanded the shift at 6000 RPM. and we know this to not me true
Didn't misunderstand it at all, I just thought that it should be made clear that when Brad said that the "ECM has nothing to do with shifts, as to when they occur", that it isn't a true statement, and the reason should be explained. Now what would be neat is if we pooled together and made a general order of operations list for each controller in each generation and make them available as free reference material on this site, so that we can direct people with these types of problems to that material, and maybe they'll actually save time and money throwing parts at it or going to the wrong techs.
Once again I think you misunderstood Brad. the 6000 rpm shift is what he is referring to when he says "as to when they occur". we are not referring to the quality or fractions of a second in changes to the shift because defueling is off or there is something going on with throttle. the TCM is just looking at the data and using it's tables to make a shift. if the data doesn't match what the TCM is expecting then it will use the wrong section of the table to calculate the on and off times, pressure, etc for the shift and that is when we get bad shifts.
as far as I know. there is no command or instruction or otherwise that comes from the ECM to manage or control when a shift will or should occur. this is all in the TCM
as far as order of operations. I think it's something like this: the ECM is continuously broadcasting over the data lines the throttle position and maybe the MPH but I think the TCM can calculate that too from it's own sensors and tables.
the TCM only needs throttle position for stand alone use so I don't think there is any other information sent from the ECM to the TCM needed for shift.
then the TCM watches the throttle and MPH. if the vehicle crosses the threshold above or below the up or downshift then the TCM checks its tables for calculated torque, RPM, clutch pressure and other metrics.
the TCM then sends a message to the ECM that a shift is about to happen (or is happening) and reduce fueling to the shift level amount.
the ECM will then look up it's defuel table and see what the mm3 should be and drops the fueling to this level. I don't know if there is any kind of message to the TCM to let it know that it has reduced the fuel as requested by the TCM.
the TCM then shifts and the ECM restores the fueling as it's non defuel tables indicate.
the TCM during a defuel shift event is expecting a certain amount of fuel or torque from the engine when it is going to perform a shift. the TCM has a set of tables that are supposed to be expected values that the ECM is going to deliver. when that table doesn't match what it's actually getting then the shifts suffer as the TCM is trying to manage applying a clutch with a certain amount of smoothness or RPM change. but when it doesn't see the expected smoothness it will increase or decrease the pressure to try and achieve that. sometimes the values fall outside it's tables or algorithm then the TCM can't adapt and the bad shifts never go away
I totally understand what y'all were trying to say. But to be honest, as soon as the OP said they had it figured out because they saw the ECM commanding a 6000rpm shift, I mentally discarded that information because it isn't plausible. It isn't real, helpful information. Whether the tcm would have to command it instead of the ecm is a moot point, It was a poor diagnosis.
So, what we really should have done is explain why that was a poor diagnosis. But instead we got into this discussion about what he said at face value...blah blah blah... I felt I needed to make the ECM calibration's role in shift quality/timing clear because I have seen several different BIG name tuners' ecm files cause un-needed clutch wear, lockup issues, and fubar the trans with a standard reputable build or stock, and stock tcm files. Some of those problems can be masked or even fixed with tcm calibration changes/tcm OS changes, just as harsh shifts from using really high coefficient of friction clutches can be. But the majority of the time, it is as simple as fixing the ecm calibration or finding the bad sensor, or boost leak, or fuel pressure issue that is really there still wreaking havoc behind the scenes.
It takes a pretty thorough diag to find and fix these issues sometimes, which is why my first post was what it was....
understood. that's the reason I told him to put the ECM back to stock before moving forward as we can't know if it is tuning or hardware related until then. then he stated that the ECM was returned to stock but the issue continued. so at that time it sounds like a hardware issue. until he mentioned the 6000 rpm command thing....
I'm glad we're all on the same page over the face value post. So technically we're both right and both wrong, it simply depends on how deep you want to get into the way the controllers work and the definition of "when".
I still stand by my statement, the truck needs to have OEM calibrations put into the controllers and see what it does from there as that's the only way to eliminate EVERYTHING except the trans build itself.
I still stand by my statement, the truck needs to have OEM calibrations put into the controllers and see what it does from there as that's the only way to eliminate EVERYTHING except the trans build itself.
That eliminates software issues, sure, but doesn't eliminate another electrical issue or weird sensor, or weird input caused by another mechanical issue. But in this guy's case, they've gone through most of everything it seems, so my bet is on a problem inside the trans.
Agreed, it eliminates the possiblity of the "6000rpm" issue being software related. It obviously doesn't eliminate any broken/chaffed wires or partially/on it's way out sensor/solenoid etc.
I too personally believe the issue is internal but the only way to know is eliminate everything else or pull the trans.
I'm going a little off topic here, but it was some of your old posts that inspired me to really dig into how all the defuel tables work in the ECM. Long story short, I think I've pretty much figured it all out (and concur with your excellent summary here). And in order for it to shift (mostly) like stock, the sum of the TBIQ, TLIQ, and Base Torque tables in the ECM have to match the Torque Limits table in the TCM. Typically, tuners make fairly drastic changes to the TBIQ table, and then either no changes to the other two, or changes that amount to throwing darts blindfolded. Outside of getting lucky, this will screw up the defuel on the shifts (which is what I suspect the OP's problem is).
I made a monster Excel sheet with all sorts of interpolations and V-Lookup craziness that will automatically calculate the Base Torque and TLIQ tables given the inputs of the TBIQ and Torque Limits tables. It matches everything up to how GM originally configured them (given the changes). In my limited and relatively mild tuning on my own truck, it works quite well. I've been meaning for months to do some more polishing, some more testing, and eventually post it up for others to see, test, and critique. Just haven't had the time.
The only thing I've never figured out is how to change the amount of defuel the TCM requests. Doing it the way my calculator does, it essentially only matches the stock requested reduction. There's no way to explicitly command more or less, and you've therefore got to manipulate the TLIQ and BT tables to do so. That could probably also be automated / calculated, but I haven't gotten around to trying that. There's surely a table in the TCM that controls this, it just isn't mapped in EFI Live.
100%!!!!
I had (emphasis 'had') a stock-trans-safe tune from a VERY well-known and 'reputable' tuner, probably one of the top 3 at that point in time. On the whole it was fine; made decent power, wasn't smokey or laggy, good fuel mileage, etc. Light throttle shifts were fine, WOT shifts were fine, but the medium-heavy throttle shifts flared something terrible. After logging it, the reason was obvious: the "requested torque reduction" would be positive, but the actual fueling would increase (in some situations, a LOT). It did this at low throttle positions as well, but the magnitude was small enough that you couldn't feel it. In the interest of keeping the trans together, I elected to discard their tune and write my own.
So in other words, this "big name reputable tuner" that claimed to have tuned many thousands of trucks, and actually was pretty decent at strictly the engine tuning....had absolutely no idea how the defueling worked on a rather simple LB7 calibration. And likely did nothing more than some 'guessing-and-checking' on the TLIQ and BT tables to get it kinda close at WOT.....
The OP claimed that he flashed stock files in and the problem still presented itself.
The OP stated the stock tune was put back in, he didn't specify if it was just the TCM or the ECM as well. The post in question is #9. It is certainly possible that the ECM was not touched.
@DarylF, was the ECM returned to stock as well or just the TCM?