High force motor % and late 1-2 shift

[Just Chuck]

P59, SD 2 bar OS, 4L80 segment swap, about a 2800 circle D stall, built 4L80. Look at my log around 8:11:59

Been using bluecat and overall, the 1-2 part throttle is pretty good, In my log it shifts into 2nd at 24MPH. The 1-2 shift in the tune is 19 @31 TPS and 22@ 37 TPS. Usually accelerate normally about 33%. I can take 2-3 MPH out of 1-2 in that area and probably be happy, but want to understand the delay.

The second issue is the force motor current is at 70% in 1st and at the 1-2 shift. Delivered engine torque is in the 180-200 ft/lb depending on the RPM, my base shift pressure is vs tq is around 12-17 in that range, I guess I can zero it back to stock to see if that softens up the part throttle 1-2

Any observations, suggestions?



ls2drveNov8_Id850_4.hpl

LTD_50lb_boostref__LS2_ID850_45PSI_3.hpt

[kingtal0n]

When trans is command to shift, the valve needs to move, and fluid needs to fill the drum associated with the clutches that cause the shift. The higher rate of chance of engine RPM (first AND second derivatives of engine RPM) the more difference between command and resulting shift (longer delay caused by higher rate of change of engine RPM)


As for force motor situation,
here is OEM Pressure vs force motor
Attachment 115094


And OEM pressure ranges for WOT at OEM torque.
Compare these with your unit by installing pressure gauge to outlet port (usually blocked by square head plug on shifter side of transmission)
Attachment 115093


At part throttle and WOT the pressure requirements may vary with engine torque, drivetrain rotating mass, momentum and kinetic energy.
For example very heavy towing requires more pressure even if engine torque output is the same, the trans is more likely to slip during a gear change.
Gear ratios multiply torque so the final gear, tire size, and difference in engine rpm vs output shaft rpm is a consideration in the pressure requirement to hold clutches from slipping during a shift.
Some minimum slipping is expected and it is impossible to eliminate all slipping completely since the moment the gear change occurs there is always a difference to overcome via the friction action of clutch materials.
Therefore it is up to the owner/operating/tuner to decide how much pressure to use and what sort of slipping is tolerable.
The fluid volume flow is increased at higher pressure and clutches are held tighter by higher pressure, however some of the flow is also in-part controlled by orifice sizes such as shift plate hole which is why they are drilled for performance apps. Thus higher volume of flow for drums is possible without increasing pressure but this will not increase holding power or decrease slippage it just brings on the gear faster by filling the drum faster to engage the clutches.

Pressure is control by EPC which is fed by filtered actuator valve pressure thus in order to fully control pressure the actuator feed must be healthy
reverseiswat.png

Notice that in reverse pressure is 'boost' much higher than forward gears.
Note that design of boost and PR valve as well as internal mods to transmission all may affect final pressure outcome. EPC is not a reliable source of pressure data until it is compared with real-world pressure values from the side of transmission using proper gauge (0-300psi or 0-400psi)


How I recommend tuning:
Increase pressure for part throttle shifts until the feeling is positive, no slippage, as quick as possible, but not jarring or slamming/harsh into the next gear. I directly mod force motor table to achieve these results because the algorithms which control pressure are not super reliable or predictable, and torque management will inaccurately skew the resulting Commanded EPC to the point where it may be detrimental to the transmission.
Therefore I Recommend using a gauge on the transmission in combination with Force motor current changes to achieve necessary shift pressure, and using torque management to make a very fast shift feel more comfortable and gentle to preserve the life of drivetrain components.

[eXo3901]

How I recommend tuning:
Increase pressure for part throttle shifts until the feeling is positive, no slippage, as quick as possible, but not jarring or slamming/harsh into the next gear. I directly mod force motor table to achieve these results because the algorithms which control pressure are not super reliable or predictable, and torque management will inaccurately skew the resulting Commanded EPC to the point where it may be detrimental to the transmission.
Therefore I Recommend using a gauge on the transmission in combination with Force motor current changes to achieve necessary shift pressure, and using torque management to make a very fast shift feel more comfortable and gentle to preserve the life of drivetrain components.

So all pressure tables remain stock using your method?

[kingtal0n]

So all pressure tables remain stock using your method?

Ah, sorry no. I've significantly raised the pressure in areas of my individual shift tables.

My file is here
https://forum.hptuners.com/showthrea...l=1#post648378


The thing is- you cannot depend on those values all the time. In Gen3 applications the computer sometimes ignores torque estimate and uses TPS% instead for determining shift pressure from the table (apparently). Other times it seems to blend them together. And when torque management is involved it can dramatically skew the result as well, depending how much timing your engine needs at WOT. For example my engine needs like 12* of timing at WOT and pulling maybe 1 or 2* results with a massive drop in torque. Whereas the torque management tables... well just look at them. They are like crazy talk, nonsense, pulling 10* or 14* of timing sometimes when I really only want 1 or 2*. And if you tell it to pull only 1 or 2* it estimates incorrect torque reduction.... The whole thing is a mess which required trial and error to nail down for my specific setup.
Another issue is the sort of 'predictive algorithm' for transmission pressure. For example as you step down TPS% rising the pressure will seem to shoot much higher than the torque and even TPS% from a table would have you expecting it to rise. The computer has some kind of mathematics which are 'helpful' to get the pressure up before it is actually needed. However these algorithms sometimes interfere with, or seem to ignore, the numbers from the tables. Often times it would wind up completely 'wrong' compared to the situation I was in, either too much or too little pressure.

The final fix was simply adjusting the force motor current table to reasonable values. Ideally, do NOT slip the transmission or it will burn clutches, and do NOT use so much pressure that it shocks the drivetrain which could damage something. In between those two you need to make sure the pressure is rising and STAYING where it needs to be, by monitoring the real-time Force Motor commanded % AND the current (mA) applied to the pressure control solenoid. Its kind of simple but may require trial and error.

[eXo3901]

Ah, sorry no. I've significantly raised the pressure in areas of my individual shift tables.

My file is here
https://forum.hptuners.com/showthrea...l=1#post648378


The thing is- you cannot depend on those values all the time. In Gen3 applications the computer sometimes ignores torque estimate and uses TPS% instead for determining shift pressure from the table (apparently). Other times it seems to blend them together. And when torque management is involved it can dramatically skew the result as well, depending how much timing your engine needs at WOT. For example my engine needs like 12* of timing at WOT and pulling maybe 1 or 2* results with a massive drop in torque. Whereas the torque management tables... well just look at them. They are like crazy talk, nonsense, pulling 10* or 14* of timing sometimes when I really only want 1 or 2*. And if you tell it to pull only 1 or 2* it estimates incorrect torque reduction.... The whole thing is a mess which required trial and error to nail down for my specific setup.
Another issue is the sort of 'predictive algorithm' for transmission pressure. For example as you step down TPS% rising the pressure will seem to shoot much higher than the torque and even TPS% from a table would have you expecting it to rise. The computer has some kind of mathematics which are 'helpful' to get the pressure up before it is actually needed. However these algorithms sometimes interfere with, or seem to ignore, the numbers from the tables. Often times it would wind up completely 'wrong' compared to the situation I was in, either too much or too little pressure.

The final fix was simply adjusting the force motor current table to reasonable values. Ideally, do NOT slip the transmission or it will burn clutches, and do NOT use so much pressure that it shocks the drivetrain which could damage something. In between those two you need to make sure the pressure is rising and STAYING where it needs to be, by monitoring the real-time Force Motor commanded % AND the current (mA) applied to the pressure control solenoid. Its kind of simple but may require trial and error.

Thanks for that. It seems like something I could try on my setup and see how it plays out. I have an E40/T42 so may not relate exactly to gen3 but I have noticed the pressure/tq management tables never match the delivered tq PID (I am not assuming that they are supposed to). I also noticed the related trans pressure PIDs seemed to do whatever they wanted sometimes, close to what you described. Now that I have it working, torque management pulls a relatively close, but to me random amount of timing during part throttle shifts. I have been adjusting the tables directly and fine tuning was mostly me moving my overall pressure or tq management slopes left or right on the table. Lower pressures and tq management seems to have a "cliff" of sorts, under which nothing seems to happen, more-so with tq management.

[kingtal0n]

.... Now that I have it working, torque management pulls a relatively close, but to me random amount of timing during part throttle shifts. I have been adjusting the tables directly and fine tuning was mostly me moving my overall pressure or tq management slopes left or right on the table. Lower pressures and tq management seems to have a "cliff" of sorts, under which nothing seems to happen, more-so with tq management.

When my car was slower 350hp~ I would see torque management properly in the graphs. Now that I've increased to 650hp~ I no longer, or rarely see the torque management in my graphs. At first I thought it was not working- because the tire would still spin on a shift. I sent my tune and logs to the HPtuners software development (whatever they are called) because I thought it was bugging out. They couldn't find anything wrong with my tune file or application integration. So I went back to my car and set up some experiments to determine what was going on. The gen3 computer has two different transmission maps, so you can swap between with the press of a momentary ground. I used that feature and programmed 2 different maps, same shift pressures but one with no torque management and one with a lot of management.

Turns out it works just fine. There is definetely a difference with it on or off. But it refuses to show in my logs, and the tire still spins on a shift.
The tire spin I can attribute to the high holding power of the 4l80e (200psi of WOT pressure it grabs pretty good) plus I just rebuilt it with all the tricks (build thread in sig has list of mods somewhere) which probably helps. With so much power its enough kinetic energy (or momentum) in the parts that even with no engine torque the tire still spins pretty good. But the car stays straight which is nice.

The missing timing dips in my log- now that is a mystery, maybe. It could just be that the instant it happens isn't being caught because of the logging rate, you know you can set the rate of the log up to be every 1 second or every 10th of a second, or every 1000th of a second, etc... But I am pretty sure I tried a very fast rate of logging and still only catch one timing dip out of say 10 runs. I guess it has something to do with how fast the car is accelerating (the rpm rise is very fast?) that the log doesn't catch it. If I turn the power down (boost controller down to 3psi) the torque management always shows up (or almost always).

Bottom line: Seems fickle and I don't depend on the log to show me everything, or even anything. Always back up the scanning data with real world data, like from transmission gauge pressure for example. There are probably times when the graph will show something that isn't true in reality, it either lags behind or holds onto something for too long, I've noticed injector pulse lines rarely match with what the car is doing on other graphs for example. It may have something to do with the speed of the computer we are using, and of course if you have too many PID logging data points at once that will tie it up and keep it from showing things. So try to limit the number of active things you are logging and pay close attention to the rate of logging, slow the ones down that you can spare to free up process for other more important ones.

[eXo3901]

Been using bluecat and overall, the 1-2 part throttle is pretty good, In my log it shifts into 2nd at 24MPH. The 1-2 shift in the tune is 19 @31 TPS and 22@ 37 TPS. Usually accelerate normally about 33%. I can take 2-3 MPH out of 1-2 in that area and probably be happy, but want to understand the delay.

You can try manually calculating and inputting your wheel size information. I have a E40/T42 and 4l65e, relevant or not. I did a rollout test and calculated the values to input. My shift speed matched much closer to what I have in my tables after that. Sometimes the tire size calculator in HP Tuners wont help as the original values in the computer from GM are wrong. There should be a few threads on here about this topic.

Bottom line: Seems fickle and I don't depend on the log to show me everything, or even anything. Always back up the scanning data with real world data, like from transmission gauge pressure for example. There are probably times when the graph will show something that isn't true in reality, it either lags behind or holds onto something for too long, I've noticed injector pulse lines rarely match with what the car is doing on other graphs for example. It may have something to do with the speed of the computer we are using, and of course if you have too many PID logging data points at once that will tie it up and keep it from showing things. So try to limit the number of active things you are logging and pay close attention to the rate of logging, slow the ones down that you can spare to free up process for other more important ones.

I will have to get a gauge hooked up to the trans and compare. My pressure and FM readings rarely make sense in the scanner. Though my experiences are limited compared to yours.

Some information in the scanner is useful. It seems like there might be parameters that control when tq management is allowed to act during/before the shift, the duration of said tq management, and ramp in/out rates. Those three items seem very consistent through all scenarios I have encountered. I have notice tq management will not act early in the shift and as a result will not occur before/during a rev limiter situation but will after the rev limit process is complete. Based on feel, it is always late. Late enough that it feels like the trans is in the next gear before the timing pull has an effect. Probably more of an inertial thing. Having more ability to control tq management would be beneficial if it exists.