Results 1 to 9 of 9

Thread: 8L90 Harsh 1-2 Shift Fix?

  1. #1
    Potential Tuner
    Join Date
    Apr 2020
    Posts
    4

    8L90 Harsh 1-2 Shift Fix?

    GMC Service Manager here. We have tons of complaints of a harsh shift from 1st to 2nd, as well as a delayed engagement when coming to stop, or partial stop (seems like it doesn't know which gear to select?)
    GM is aware of it, with no fix. Is this something we could tune out? Maybe its a delay in shifting that we could make quicker? Line pressure?

    Thanks

  2. #2
    I do notice that the 1-2 shift quality is worse than all the other shifts. I just contribute most of that to it not being a freewheel clutch engaging anymore, I know the 4L60 was that way and I think the 6L80 is a freewheel going into 2nd from first? Most of the time I notice its either a unnecessarily hard shift or its either a flare.

    A lot of people won't agree with this, but if the issues are in fact in the software, we don't have the necessary parameters in HP Tuners to properly fix these issues. The extremely inconsistent and overly harsh/hard 3-4 shift versus all the others in the 8L90 at part to medium throttle drives me insane. I suspect this is due to lazy or overly conservative calibration tuning, since it will go away with clearing the TAPs or when you force the desired shift times higher in those regions on the 3-4 shift. However, the TCM always finds away around what you have done for some reason and goes back to its normal behavior of unnecessarily hard and very inconsistent shifting.

    As a service manager, I think you would be better off not resetting relearning the transmission or resetting taps for customers that complain, since I think it will only make them more mad when the issues come back after a few weeks. Unfortunately I think you might have to agree with them and tell them you hate it to but there's nothing you can do. Even if we get proper access to the necessary parameters they may be so complex and so many of them we may never figure it out anyway without some engineer on the insides help.

  3. #3
    Another issue is we don't have correct PIDs defined to log what the transmission is doing. Not sure if these are available to look at with a GDS tool or not, I have a Chinese copy of one but it never connects or works right and is basically useless.

  4. #4
    Tuner in Training
    Join Date
    May 2020
    Location
    Mn
    Posts
    11
    I'm a GM world class tech at Chevy dealer and I also own one of these trucks. I have almost completely eliminated mine it seems. The issue is from a normal bleed down of the clutch overnight, but after I dialed in a particular sensor for air flow (not giving too much away but you can figure it out) my 1-2 morning slam/trans jumping out of the bed feeling is almost gone. It only happens now 10% of the time and it's much softer.

  5. #5
    Quote Originally Posted by HardH20 View Post
    I'm a GM world class tech at Chevy dealer and I also own one of these trucks. I have almost completely eliminated mine it seems. The issue is from a normal bleed down of the clutch overnight, but after I dialed in a particular sensor for air flow (not giving too much away but you can figure it out) my 1-2 morning slam/trans jumping out of the bed feeling is almost gone. It only happens now 10% of the time and it's much softer.
    I'm not sure with the 8L90 if the normal fill volume strategy applies (or maybe too much cold pressure modifiers), but have you noticed the extremely inconsistent and overly fast/harsh 3-4 shift in the 8L90 vs all the other shifts at part to medium throttle? I can reset the taps and it will go away for a few days until it learns back in the habit of a .250 second shift at part throttle (30-40% accelerator/ around 2000rpms). This is way to fast of a shift specially when the 2-3 and 4-5 shift at this throttle level is in the .400 second range. I can watch the latest shift time PID and see if learn back always to this value despite what you command in the shift times and try to override in the pressure offset tables.

    I really want to try full writing my TCM with a Corvette or maybe CTSV operating system and see if there is some hidden table. I have noticed these inconsistent shifts in a lot of GM transmissions including the Allisons too. And the parameters we have available in HP Tuners and EFI Live have absolutely zero effect, the controller will always override it. Now shift times at WOT may actually stick when you changed the commanded target, but maybe there is more tables at play when at part throttle.

  6. #6
    Advanced Tuner
    Join Date
    Aug 2010
    Location
    Detroit, MI
    Posts
    408
    Maybe I can weigh in. In this case unfortunately, there is no 100% solution to the 8 speed or 10 speed shift quality issues. The 8 speed has obviously earned itself a worse reputation than the 10 speed, but they share a common fault. That is that the controls themselves are flawed. GM dropped the ball on the software that controls those transmissions and no tweaking of the calibratables, available in the editor or not, will sufficiently fix the shift quality. I worked with a few people involved in the 8/10 speed shift quality investigations at GM that were pretty vocal in their disapproval of the trans algorithm team on these transmissions. The fast trans learn and on-road service procedures to learn clutch pressure/volumes is realistically the best bet to get rid of the shift issues. But it's a bandaid at best with the underlying algorithms not controlling the transmission as well as they should.

    Probably not what people want to hear, but it is what it is - so just manage your expectations.

  7. #7
    Advanced Tuner
    Join Date
    Apr 2017
    Location
    Meridian MS
    Posts
    891
    So increase the amount of torque reduction during shifts in the region you want softened. Works a treat even when shift times are very short - take away all the engine torque for ~.300s and the trans can shift as fast and hard as it likes, and you won't feel it in the seat.

  8. #8
    Tuner in Training
    Join Date
    May 2020
    Location
    Mn
    Posts
    11
    We get a A LOT of complaints and no problem found issues at the dealership for 8 speeds. People do not like it when we tell them what they experience is normal operation. There have been so many tcm updates, the updated fluids, and faulty trans temp harnesses, but at the end of the day they just have a lot of quirks. Most of the issues we see are abrupt throttle changes at or near a shift event. Try it out....you can induce some crazy stuff just by flopping the pedal around a bunch.

  9. #9
    Thanks smokeshow, I figured bad programming was the case. Do you think it was 100% just sloppy programming maybe from beauracracy? budget limitations? or do you think compromises were made for durability or whatever reason?

    I find the absolute worst part about it is how inconsistent it like to be on how good or bad it shifts, there is just no rhyme or reason. I guess its the same reasons why we see unnatural curves with random discontinuities in the stock maps as well. I know there aren't too many examples of random discontinuities in natural physical processes, but apparently whoever programmed the SOI modifiers thought they existed haha

    It seems like there is something I am missing as it doesn't seem that difficult of a control engineering problem to just put in some kind of gain (pressure increase or decrease) in a feedback loop (transducer in feed lines) and then just tune it by trial and error and experimentation. I guess an engineer was probably trying to fix it when some project manager probably put a stop to it after they found they were over budget on engineering hours from their unrealistic planned engineering hours.

    Like the OP said originally the 1-2 shift has some issues, a lot of random flares with some occasional overly harsh shifts with rarely a good smooth shift. However, it just makes no sense why the 3-4 has to be that fast and hard of a shift at part throttle. It is pretty consistent in that you have to push torque up past say about 150 ft-lbs (shifting at a little over 2000rpms on stock shift points) to get it to happen, but its so wildly inconsistent that it only happens about 75% of the time and the other 25% will randomly be very smooth and match the same feel and ramp down time as all the other shifts.

    Smokeshow, I have heard that apparently when you reset taps on a transmission if you clear all the shifts at once it will eventually learn back the same way or diverege the shifts. So you have to let it learn enough to where it stops learning then individually only reset the taps for the problem shift. In fact, I believe this is what the TSB that they released about this problem says exactly that you only should reset the taps for individual problem shifts and not try to correct a bad individual shift by resetting the whole thing. This leads me to believe that the control algorithms for each shift are not independent and other shifts can maybe preemptively "learn" from other gears that have already been shifting through, is this correct thinking?

    Another problem we have is our ability to reset the taps in the HPT scan tool is flawed in that it dosen't work correctly and it appears they just took the same stuff they developed for the 6L80 and copied it over to the 8L90. The Chineese GDS tool I have also only has the ability to do the learn procedure while driving and dosen't give you the actual ability to just reset each shift taps or even all of them at the same time. Is there another way we have to individually reset only a certain shift tap value?

    Thanks for the info, I know there is probably a lot of people who could care less, but I love inside info like this and im always searching for knowledge like this. Thanks