Sanity Check on 4L60e settings

[kingtal0n]

Im not sure what your trying to say. I have tried to look up exactly what desired torque really does and have not found anything. So your saying this is why its staying retarded for so long with tm on?

King pretty sure the ecm knows that the shift has happened. i have a pressure guage that i watch. I can always tell from it when its about to shift. I will see it gain about 70psi right before a shift then see the shift happen and the psi go back to normal. then there is a couple secs of retard still happening after that.

Wow, cool, you have a trans pressure gauge. That is a huge bonus. Hmm let me think...

When the pressure is rising prior to commanding a shift (simply monitor PWM duty cycle % at EPC solenoid using scanner) This I can understand.


However, knowing when the actual shift will occur is a mysterious consideration. There is no way the ECU can know when it will actually happen. Lets say theres a problem with a drum or the shift orifices are enlarged or something and it takes longer or shorter than usual. The ECU doesn't know there is some issue or a shift kit installed, so it can't predict *when* the shift will happen.


Let me ask this then. In the scanner there is a 'time of latest shift' reported. Do you see the correct time reported for the shift? Does the time reported match actual shift time from the log?

And finally,

Does the timing reduction stick around even after the 'time of latest shift' updates with the correct time for the previous shift?

This is where I would start looking to diagnose this issue.
The next place would be something related to the timing retard recovery rate. I am not sure if the timing recovery from 'knock' settings will reflect into Torque Management but its worth considering you have some extra delay there because of some retard recovery value somewhere.
Also, I've seen in my ECU sometimes the timing reduction is sooo deep it just maxes out and you get a flat line of timing reduction instead of a dip or valley. Is that what you are seeing? and if so, the issue could simply be too much timing reduction from torque management tables.

Sorry I am not HPtuner expert, just a tuning expert. But our friend here BlindSquirrel IS a HPtuners expert and if you follow him closely I bet he can help you sort it out.
Im just trying to learn what I can

[49 hoopty]

I dunno. How much does it cost to change it back and see what happens?

I'd put all the Abuse Mode stuff back to stock too, but then again I'm a boring square who doesn't do sweet WOT neutral drops.

I'm not trying to fight here. I was just asking a simple question. If that makes a difference tell me. I know you guys don't post tunes here and i get that. At least if you know can explain what a certain thing does if you move it one way or another. I googled, watched, and read all kinds of stuff on desired torque. no one ever said exactly what it does or how to set it. thats how the tune was when i showed it to you. I showed you the log of what was going on with all of the tm with the tune you told me to try. I am not sure why the retard was lasting 3-4 seconds. A lot of people say if you have a built transmission to disable all tm and run it. I believe it is there for a reason if set properly. I really don't want to blow up my transmission. Hence why i asked the question. The motor stuff is easier to tune hard to blow it up, but i am afraid set up something wrong in the transmission and its done.

[49 hoopty]

Wow, cool, you have a trans pressure gauge. That is a huge bonus. Hmm let me think...

When the pressure is rising prior to commanding a shift (simply monitor PWM duty cycle % at EPC solenoid using scanner) This I can understand.


However, knowing when the actual shift will occur is a mysterious consideration. There is no way the ECU can know when it will actually happen. Lets say theres a problem with a drum or the shift orifices are enlarged or something and it takes longer or shorter than usual. The ECU doesn't know there is some issue or a shift kit installed, so it can't predict *when* the shift will happen.


Let me ask this then. In the scanner there is a 'time of latest shift' reported. Do you see the correct time reported for the shift? Does the time reported match actual shift time from the log?

And finally,

Does the timing reduction stick around even after the 'time of latest shift' updates with the correct time for the previous shift?

This is where I would start looking to diagnose this issue.
The next place would be something related to the timing retard recovery rate. I am not sure if the timing recovery from 'knock' settings will reflect into Torque Management but its worth considering you have some extra delay there because of some retard recovery value somewhere.
Also, I've seen in my ECU sometimes the timing reduction is sooo deep it just maxes out and you get a flat line of timing reduction instead of a dip or valley. Is that what you are seeing? and if so, the issue could simply be too much timing reduction from torque management tables.

Sorry I am not HPtuner expert, just a tuning expert. But our friend here BlindSquirrel IS a HPtuners expert and if you follow him closely I bet he can help you sort it out.
Im just trying to learn what I can

King i am not sure what all the values were on that when i had the tm all applied. Didn't have the correct pids running then to see all the other stuff that was going on. Just the log that i posted with it all enabled with those pids.

[49 hoopty]

and sci85 i am sorry i hijacked your post.

[kingtal0n]

You def want torque management. Its by far my favorite feature of the OEM ecu and will absolutely keep your trans alive much longer than it would have lived without it.

People say all kinds of stuff, but I like to think and consider what is happening and decide for myself. When those paper clutch materials come together there is always a bit of slip for an instant. The less torque at that instant the less they will slip in theory, because the trans can more easily pull the engine RPM down to keep from slipping the clutches or friction. If torque output is staying high then the only other option is to speed up the tire suddenly to match the new ratio.

Thus we must always consider drivetrain intertia, the rotating mass and acceleration of that mass, all the way to tires, which may need to rapidly change rotating speed for a new gear ratio. When the trans goes to shift and applies some kind of friction materials, even if the torque of the engine drops out there is still the rotating mass that needs to accelerate or somehow deal with the difference in speeds, depending on the RPM drop the engine can experience (different converters = different drop in RPM rates, some can slip more easily and deal with a big difference in rotating speeds) so either the tire needs to "catch up suddenly" Which results with a tire slip or 'chirp' usually. Or the engine needs to slow down or the converter needs to slip to compensate for the new rotating speed differences. Or something between the three. In between all the parts of the trans and those paper frictions, making up the difference for an instant, trying to pull the tires up or the engine down. So, just thinking about it, the weaker the tire, the easier they break traction, the longer the trans frictions will last. Similarly, a looser converter with more ability to accept the difference in speed of flexplate to input shaft, the longer the trans frictions could last. And again similarly, the lighter the rotating mass of the rotating parts between the trans and tires, the longer the frictions should last. So torque management is only one of these 4 major components to preserving frictions, I guess. We want to 1. reduce torque which makes the engine easier to pull down, and also 2. make the tire easier to slip which makes it easier to pull up, 3. make all the parts that rotate as light as possible which reduces their momentum/kinetic energy, and 4. use a converter that can slip during a shift easily. With everything setup nice I bet we can easily 700 800 900hp through a 4l60e because I believe what really kills them isn't the power so much as the abuse during a shift they encounter when using powerfully sticky tires or massive input torque during the shift, the energy has to go somewhere and they don't call those things 'frictions' for nothing, the energy applied suddenly or without care can tear metal like paper if it isn't handled and accounted for properly. On the other hand without a tire you aren't going anywhere either... so there has to be some give and take between power and traction when you discuss a trans like the 4l60e, which isn't really 'beefy' no matter what parts are inside, its got some weak points and flaws that are just inherent due to it's economical design/intended use. Its a great transmission when you are interested in some semblance of economy, that is the main reason to use a 4l60 and not a 4l80. The 4l80e although it may be stronger will also take a good 3, 4 maybe 5mpg off a setup due to it's massive rotating mass and internal friction coefficient being so much more than a 4l60. The 4l80e will absorb about 5hp more than 4l60e which doesn't look like anything on a dyno graph (it won't hurt your power output significantly) but 5hp during a cruise is like running 2 or 3 extra A/C compressors all the damn time!

[eXo3901]

I cannot find it right now but there are few patents that explain some of the shift torque function.

This is not in perfect order or exactly how things operate and I do not even know if what I read made it into any production GM vehicle. Just what I can remember reading in the parents and seeing it operate similar to the description in the real world, most of the time. Also probably not fully applicable to anything but 4l6xE platforms.

Coming from the transmission side, the trans knows when it commands the shift, the engine side has no idea what is happening. Trans calculates torque or gets a torque signal from the PCM. Uses a lookup table to determine how much torque reduction is needed based on torque and current shift(1-2,2-3,3-4, ect..). Uses gear ratios to predict something. (Seems like this might be a multiplier or some sort.) Activates/deactivates solenoids. Measures RPM at point A to B and B to C to determine when and how much RPM rises/drops. Then if the calcs are good, the trans sends a signal to the PCM to reduce torque by xx%. You see this as spark cut as the RPMs are dropping and ending right around when RPM stops falling and after the new gear is fully engaged.

During all that, if torque is below a value, no torque management is requested. (Like a 1-2 shift in a parking lot when torque is very low, if you drive like a normal person) There are time delays, from my understand, between shift start and RPM peak, RPM peak and RPM fall, which if exceeded the trans will start or end its torque reduction request. It seems like there are also RPM rise/fall thresholds. Also described was timeouts for how long the torque reduction signal will be sent.

The above parameters I have never seen defined in HPT. But there are similar parameters in some very, very well defined (and recently updated) Gen3 xdfs.

It is possible the converter keeps the RPM fall small enough that the trans wigs out and just keeps requesting torque reduction, but then it should hit a timeout and stop. Sucks to assume so many things though when it could be easily explained or fixed if all these settings were defined, assuming they exist.

Links below to GM patents. Again, no idea if these were ever applied to real life applications.
Control method for managing engine torque
Shift torque management

[kingtal0n]

Awesome thanks! I'm glad to not be the only one in the dark here with the way the ECU is thinking.

Uses gear ratios to predict something. (Seems like this might be a multiplier or some sort.)

Hey! This reminds of something. I wonder if thats what the transmission selector box does, you know the one that you change from 4l60 to 4l80 in the drop down menu? That does nothing? Maybe it tells the ecu what the gear ratios are for the trans so it can predict rpm drop or whatever

good info keep it coming

[blindsquirrel]

I don't understand the resistance to just set the table to '0' instead of '640' and see what the hell happens.

I don't know if it's related to your specific TM weirdness or not, but then again I still wouldn't know even if it were in front of me and I could put my actual hands on it - what I would do to find out if it was related is change the table and see what happens.

[blindsquirrel]

I know you guys don't post tunes here and i get that.

...what? I don't know who 'you guys' are or how I got put into some kind of group, but I download, make tweaks, and re-upload other peoples files all the time to show what was changed where. Like this:

after iac changes but back to original braf minus 2 points - stuff changed.hpt

[49 hoopty]

...what? I don't know who 'you guys' are or how I got put into some kind of group, but I download, make tweaks, and re-upload other peoples files all the time to show what was changed where. Like this:

after iac changes but back to original braf minus 2 points - stuff changed.hpt

I will try your changes and post a log of it.

[49 hoopty]

here is a log i did today with all the tm setting on. Ignore the lean afr didn't calibrate after i changed some other stuff. Im not sure if the stall in the converter is messing up the computer or what. on the 2-3 you can feel it engage immiediatly but on the log it sometimes says it takes 5-6 seconds.

in 3rd gear selector shifting.hpl
shifting to coffee and cars.hpl

[blindsquirrel]

Why are the shifts taking so long to complete?? Post the tune file used for those logs. That's way too long, it's burning itself up applying that slow.

screenshot.19-03-2022 21.45.34.png

What it should look like:
screenshot.19-03-2022 21.46.12.png

(edit: yes, both screenshots were taken at the same zoom)

[kingtal0n]

I thought the shift solenoids are either *On* or *Off* so how can there possibly be any delay between turning something on and off? In other words, it doesn't apply a PWM to a shift solenoid, right? Its just an instant kick on or off. So you can't have a delayed or long drawn out apply. Its not possible. So wth am I looking at with the long slopey curve in that log file?? Some adaptive bs maybe? I hope he turned of adaptive shifting or whatever because from what I think I see, the computer is confused, not the trans

[49 hoopty]

Why are the shifts taking so long to complete?? Post the tune file used for those logs. That's way too long, it's burning itself up applying that slow.

screenshot.19-03-2022 21.45.34.png

What it should look like:
screenshot.19-03-2022 21.46.12.png

(edit: yes, both screenshots were taken at the same zoom)

I don't know why the computer is showing it lasting that long. You feel it shift quickly. It just hangs with that retard lasting forever, and you can see on the pressure gauge ever after the shift it holds onto the pressure till the retard stops and it thinks it has finally shifted. Its like it thinks it didn't shift for some reason. I need to to a log when i get back home and see how it looks with no tm settings. If its the same then its gotta be some weird adaptive or computer glitch going on, And if it shows quick shifts again something in my computer or tranny does not like tm. I will post the tune that this log was on. I know its just stock 4l60 tm settings. Something is not jiving. I know you change one thing and it effects like 4 others.

iac correction, over under spark moved 2, startup air flow decresed, full tm added.hpt

[kingtal0n]

Don't give up. I'm sure blind can figure it out. I'm also going to try to change some stuff in my trans tune today to see if i can replicate the issue because this is very very interesting to me. I think it will answer some questions about how the computer keeps track of shifts and it is one of the big mysteries to me for half a decade still.

Its worth the effort when you get it fixed the trans will shift *very* nice and last a long time.

[49 hoopty]

well i finally got a chance to do a shifting log with all the tm off. It looks like it is not the tm that is delaying the shifts. 1-2 averages around .8-.9 seconds, 2-3 is 1.6, and 3-4 is usually .2-.6. I have no idea if it is something internal doing this or one of the random settings i have not figured out yet.

shifting back from the bank.hpl

[kingtal0n]

Its a good mystery. My shifts look like what blind posted, there is a sudden notch and .1 seconds later the rpm drops and it reports me a shift time .08second or something.

The only one that doesn't work right is 4th, I think it always says some crazy number for shift time. I always wondered if it was because of my aftermarket 2800stall 9.5" converter. But even though the shift time is "wrong" the notch for gear change is still the same, there is no delay. And the trans doesn't ever slip.

I'm wondering at this point is maybe a restart is in order, go back to an old file or original file and start over. See if you can narrow down a change which caused the behavior, if any.

I hate to ask this also but, are there any codes? Perhaps a code for tcc slip or something. There are setting to throw a Code if the tcc slip is 'too much' I just wonder.

Or, I wonder... Maybe the issue is with the scanner. Are you logging too many things at once for the scanner to keep up with? I Noticed it isn't exactly super reliable. Maybe remove some stuff, slow some things down. Right click stuff in your scanner and select "polling interval -> " And choose a slower logging rate for stuff you don't really need. Like keep the ignition timing polling interval high but slow stuff down that isn't important like coolant temp logging rate can be 2 seconds to update, its fine. If you get the scanner 'more alert' it might give you some new information.

edit
actually now that I think about it a little more, I bet your polling interval for "trans current gear" is too high and that is what causes the long drawn out line to appear for the gear.

Unfortunately this would not have any affect on your shift time or torque management I'm afraid.
If I have to guess I would say its probably the converter 'hiding' the rpm drop which is what the ECU uses to tell each gear ratio, again a guess.
That means at higher rpm shifts (a shift at higher rpm) with a significant RPM drop should report a more accurate shift time and use torque management correctly. Still weird that it works for mine even with poor shift time estimates in 4th, so I'm still not sure wtf is going on.

If this is the case it seems you would be restricted to using torque management only in places where there is a significant RPM drop for the ECU to tell where the shift occurs. Like during a full power or high rpm pull. In for results.

[Klay]

I am not an expert, but a stiff 1>2 shift is a classic mechanical sign of the valve body wearing out on a 4l60e, and you may want to look into a new valve body. You dont have to drop the transmission to swap it, but it does make it easier. If you let it go and don?t fix the valve body then a full blown rebuild is the next step when 3>4 goes out! The valve body still controls a lot of the functionality of the transmission even though when the e was added to 4l60 it was the electronic revolution of the transmission.

I have thrown out 2 4l60e transmissions. First was the 3>4 clutch plate wear out of a 99 lm7 in a 4x4 GMC Sierra and then had the 1>2 go stiff before the 3>4 03 l59 4x4 Chevy Silverado bought used from up north with fiberglass bedsides and fenders stock. Currently building a twin turbo 00 lm7 4x4 Chevy Silverado that I have been doing research on to prevent having to rebuild the transmission since I actually rebuilt the top end and put this much money forced induction into. Again not an expert but I have read that some where.

[SilverBullet6.0]

My 2006 p59 controller is doing this very same issue of pulling torque management for multiply seconds even though the trans has shifted quickly and successfully. I also have a built trans, with as high stall converter.

[SilverBullet6.0]

https://forum.hptuners.com/showthrea...agement-Issues