Setting the Stage.
4r70e/4r75w tranny play - page I
Findsomethingandburnit.com 2004-2020. Doc J.Vicious & Murph-O-Minion enterprises. Godfrey, ON.
Content may be used on the condition that the original author be clearly credited.
Last updated July 10, 2020.
Who? What? Why?
In days of ye olde when knights were bold, a
shade tree auto tranny swap was for all intents
little more than matching the bell-housing bolt
pattern and occasionally addressing driveshaft
length.

And of course humping the bastard things up into
the chassis; the 250 pound bench press - trying
to start a bolt with tail shaft to mouth, bell-housing
resting on the boys, fluid pan balanced on your
chest.

These days? Who the hell knows what will fit what
anymore, and if it does bolt up - what are the
chances that the computer will play nice?

Recently finding myself arsed when my "new" '07
Crown Vic Interceptor shit her slushbox less than
a hundred miles from the curbside dealer.

Caveat emptor.

But no worries, There's was the carcass of an '04
Vic in the garden on blocks, although shy a
number five connecting rod, with a perfectly tight
juice box.

At a glance the (transmission) units appeared
identical  - at least from beneath. Let's just swap
'em.

Only
after pulling both transmissions in the August
heat did I realize that there were in fact subtle
differences - the '07 (4R75W) box had an
additional two pin connector above the selector
lever missing on the '04's 4R70E. Will it work
anyway?

Consulting with a higher authority returned
information to the effect that the two transmissions
were in fact totally different and
not
interchangeable.

So I was told. Sure - But I was also once told that
I'd go to hell for playing with my willy.

And so the defunct unit was sent six hours West
for a $2300 master kit rebuild. Unfortunately
however, upon reinstallation, It still didn't function
correctly.

And so this saga begins.

Less than looking forward to pulling the tranny
again for a warranty tear down, and suddenly
realizing that I wasn't even certain if the problem
itself actually lay within the rebuild or at the other
end of one of the various mystery connections
between the transmission and the chassis,
(Indeed realizing that I didn't even have a clue as
to what signals the transmission was expecting to
receive), it became immediately clear that to be
able to properly diagnose and repair the
problems, I would have to better understand the  
transmission.

This is not a story about the internals.

The '04 (4R70W) box is a known to be good unit
that physically fits in the chassis.

I just need to know how to speak to it.

There are black box aftermarket solutions that will
do exactly that, but I want to understand the
mother tongue, not insert an alien fish into my ear.

This knowledge as well will serve me in the future:
I like to build cars. It would be nice to utilize a
relatively modern EOD offering instead of being
shackled to 50 year old auto trans technology.

At a glance, the method in which I've decided to
tackle this project may arguably appear somewhat
long winded, ham fisted, and with a series of Rube
Goldberg solutions.

Indeed it is. I could simply write a bit of open
source for an Arduino, publish a schematic with
five or six components, and problem solved.

And indeed this may very well be the resulting
final chapter - if only for one of the hundreds of
available popular transmission flavours.

I assure you however, that it will be the knowledge
gained during these practical hands on
experiments - for better or worse - that will be of
eventual importance to the savvy wrench; either
making one a more effective diagnostician, or in
removing the fear factor of selecting a modern
"electronic" transmission for  a 'rod project.

Of course it's also entirely possible that this
particular project will conclude with a thoroughly
destroyed transmission - perhaps two.

For the record, a disclaimer: I don't really know
what I'm doing.

But Mark Twain was quoted as commenting that a
man who carries a cat by the tail will learn
something which he can learn in no other way.

The ultimate goal of this project is to reverse
engineer the signals required to operate the
transmission as a "stand alone" black box entity,
and then construct a simple circuit capable of
controlling the trans without the ECU. In effect, to
spoof the transmission.

Here kitty kitty...
How difficult could this possibly be?

At the end of the day, this is still a hydraulically operated
automatic transmission not at all that far removed from
the Hydra-matic offerings of the late nineteen thirties.

The major difference being that this particular unit (A
Ford 4R75W) is at least partially controlled by a
computer - when and to which gear to shift, line
pressure, and converter lock - via a convoluted bushel
of injected ones and zeros based upon a metric assload
of processed chassis and engine data.

As far as I've been able to determine however, there are
only s
even possible global parameters able to actually
affect what this transmission does.

T
wo, can for all intents, be disregarded as inherent to
operation. They are that the trans is connected to an
input source (a steam engine or electric motor), and
that
the orientation is
reasonably pan side down and level.
The third is the mechanical gear selector input shaft,
rather straightforwardly cable operated by the gear
selector lever. It has an internal six position positive
detent and occults absolutely no nefarious voodoo.

For all intents, suffice if you fastened a broom stick or
Vise Grips protruding through a hole in the transmission
tunnel to the selector shaft. The positions are
(clockwise, from the forward most position closest to the
engine): PRND2L.

The tandem switch module sends signals to the chassis
and computer to, among other things, disable the starter
function as appropriate and operate the reverse
indicators - but it does not directly send any signals to
the transmission internals.
This connector on the transmission passenger side
adjacent to the gear selector is where the remaining four
command signals of interest to this project are sent.

It is not an overstatement to conclude the following:
Irrelevent to ponder the hydraulic magick within. That all
universally possible internal clutch, servo, band,
converter, shift, brake, and coast configurations rely
entirely on the t
hree "inputs" aforementioned, and these
four wires.

Indeed there are electrical output signals egressing the
trans case, and we may discover them helpful in
developing a practical useful daily driver friendly circuit.

But theoretically .. We only have four wires to figure out.
And here they are. A pair of shift solenoids - two
solenoids each with two possible states (on or off) giving
four possible gear selections, a third for converter lock,
and yet another to modulate line pressure.

All four solenoids are electrically isolated from the case
and referenced to a common pin.

For the purposes of our experiments, given that the
Crown Victoria mule vehicle is still my daily driver - a
method of switching the transmission between the
factory ECU signals and whatever test circuit I imagine
will be required, as well as bringing the four signals into
the passenger compartment so that they may be easily
"sniffed" in various normal driving modes.
Four generic SPDT automotive type relays with their
coils wired in parallel function as a four pole double
throw relay.

The relay poles are connected to the transmission, and
the normally closed contacts to the factory ECU. In this
manner the circuit when unpowered has no effect and
the vehicle's transmission is commanded normally by the
ECU.

Energizing the relay bank will switch the transmission
control from the ECU to whatever we create.
The harness between the ECU and transmission has
been cut, and both ends brought through a hole cut into
the transmission tunnel to the relay bank.

A bracket was quickly fashioned to mount a clam shell
hobby box which will both protect our breadboard circuit
from coffee spills and cigar ashes, as well as keeping
the experiment easily accessible for modifications and  
relatively tidy.

A lighted pull switch will serve as a "master" switch
ensuring non ambiguous engagement and status.

Circuit power will be drawn from the cigarette lighter
harness, and a small 120 Volt inverter will be added to
operate the various bits of test gear that will be required
to look at data being sent to the transmission while
driving normally under various conditions so that I'll be
able to eventually emulate the signals.