Originally Posted by athet
Would there be any way of more directly controlling shift pressure than the 'desired shift time v. torque' setting? Aside, of course from raising the line pressure wholesale via the force motor current settings... ::)
From my friend at GM Powertrain:
You are mentioning the shift times as if
it is an input. The FWD transmissions adapt on the shift time by modifying
the shift line pressure from one shift to the next. However, the 4L60 has
never used the adapts. That is because there is no input speed sensor, so
the 'shift time' can only be calculated by looking at the engine speed
pulldown, which is not good enough. The 4L60 shift line pressure and
torque management are whatever they are calibrated to be, and the shift
time that results is the shift time - an output, not an input, is what I am
saying here. So I don't fully get it.
As for line pressure, there are 7 tables that are in the calibration.
The first column of each is 'gearbox torque' from zero to some large
number. Gearbox torque is continuously calculated using engine airflow,
engine speed, and trans output speed. The second column in tables 1-4 is
the line pressure command in gears 1-4 respectively if no shift is taking
place, called 'steady state line pressure.' The second column in tables
5-7 is the line pressure command in shifts 1-2, 2-3, and 3-4, respectively,
called 'shift line pressure.' The shift line pressure is always much lower
than the steady-state line pressure surrounding it.
So one of these tables is being used at all times. Which entry in
the table is being used to create the line pressure command at a given
moment depends on the gearbox torque calculation at that moment.
So if you followed all that, here is my advice. Don't increase the
steady state line pressure. It won't get you anything but extra heat in
the fluid, and maybe wear out the pump. If you reduce the steady-state
line pressure, clutches might slip, and that can be very hard to perceive
until it is too late and you smell it. Bump up the shift line pressure as
far as you want. If you only bump up the entries corresponding to higher
levels of gearbox torque, the light throttle shifts can be smooth, but the
heavy throttle shifts can make that pleasing tire squawk noise. The amount
of tie-up in the 2-3 is controlled by hardware, so it is ok to add huge
amounts to the line pressure for the 2-3 if you want to.
The last stage of the calibration
is the table that represents the response of the line pressure to the
solenoid current. I did not even mention that, because that is the least
logical to change, although maybe the easiest to backwards engineer for
these aftermarket folks. The problem is that different gears and different
shifts require different amounts of line pressure for a given amount of
engine torque, such as full throttle. So a full throttle 1-2 might be only
using 45% of the max current, but a full throttle 3-4 would use 90% of the
max current. But then a light throttle 3-4 would use 45% of the max
current. So you would have no way to get a firm full throttle 1-2 without
also getting a firm light throttle 3-4, by modifying that table. That is
what I think these guys are referring to. The 18x18 table that you have
there must be the lookup for that based on temperature or something. A lot
more job security goes with a 2-d lookup than just a 1-d lookup.
I'm guessing that the vertical axis of that table is temperature and the
horizontal axis corresponds to line pressure, and the values in the table
are milliamps. Clear enough? Ok, just jack up the pressure all the time
and have at it.
Also, beware of aftermarket shift kits for the 4L60. There is an orifice
they have you open up to make a faster 2-3. The problem with that, is that
the fluid going through this orifice to move the 34 clutch piston is also
the same fluid that is holding on the 24 band, but before the orifice. The
downstream of orifice fluid acts against the other side of the 24 band
piston to push it off, once the 34 clutch piston actually contacts the
clutch and some pressure builds up. Unfortunately, the fluid before the
orifice , which is keeping the band on, has some pressure drop already from
going through the manual valve. While the fluid is moving the 34 piston
against only the return spring, the orifice is like an open drain in the
system. So when you open up that orifice, the pressure drop through the
manual valve becomes enough that the band starts slipping, before you have
torque capacity on the 34 clutch. So aftermarket folks then give you a new
band piston with more area. That fixes the above problem, but then when
the 34 clutch starts to get capacity, the band has way too much force on
it, and you get terrible tie-up. But it's a firm shift, darn it.
You can successfully use the orifice changes in a shift kit to get a really
firm 1-2 and 3-4 however. But it is much more logical to do it with line
pressure.
On the 4L60, and all the other current transmissions, one solenoid event
causes a shift. In the transmission hardware, only the 2-3 shift requires
something to come off for the shift to occur, in D4 anyway. The off-and-on
for the 2-3 and 3-2 are coordinated in hydraulic hardware, using the method
patented by Simpson, around 1950, which is really quite interesting, but
too hard to put into words, although I tried anyway in the previous email.
So anyway, that can't be the shift time.
We could drive around your jeep using alligator clip wires to the solenoids
and a battery sliding around on the floor to make the shifts, and it would
work just fine, as long as we didn't, for example, put it into 2nd at 120
mph.
Lots to digest, but I thought it might be useful, especially in dispelling the 'setting the desired shift time to 0 sec = sorta two gears at once' myth Last I talked to him, he was working on the 5 speed Caddy automatics, but I'm not sure what he's doing now.
-Anton
