auto trans question

[mrjones]

With the car in Drive, stopped pointed uphill, take the foot off the brake, and the car will roll backwards. Pull the trans into 2, and it will not roll back. Why is that?

[SergntMac]

Jyst thinking out loud...If this was my MM, I would check the fluid level first, then my idle speed. That sensation would make me wonder if the front pump inside the tranny was reacting to the low load or not. Does the car creep forward at idle, no brake, on a level surface?

[Racerx88]

Our cars have a slightly higher stall speed in the torque converter than your average car. In first gear, it'll take more than idle speed to get the car moving, especially on an uphill start.
In other words, do what the Sarge said.

[TripleTransAm]

I'll take a stab at it... sitting in a hotel room in Austin, no access to the manuals, I think I recall a few details.

From what I remember, there is an extra clutch band or perhaps some other friction element that is applied when manual-2 is selected. Perhaps to increase torque capacity or perhaps just to deliver augmented engine braking, but this extra element seems to override some sort of sprag-like operation within the tranny.

ie. there's most likely some sort of mechanical sprag-clutch that allows application of power in one direction, but freewheeling in the opposite, in the normal 2nd gear application found in 'D'. In this case, when the engine is pushing against the tranny, this sprag locks up and allows torque transmission through the gear set(s), but if the engine is decelerating, the sprag freewheels which means the decelerating engine places little or no reverse torque / drag on the driveline, and hence little engine braking. However, in manual-2, this sprag action is overcome because of that extra friction element locking things together in both directions, not just one as in the case of the sprag. It's my understanding that this only happens in manual-2...

So in 'D', at a stop the tranny is in 1st, and there is no reverse torque available so there is little to hold the car in place (and the 4.6's idle speed is so low and the torque output so low that the loose converter cannot provide enough 'push' against the tranny to lock the sprag in the driven direction). Put it in 2nd, all of a sudden the whole tranny is locked together and the torque converter is now acting directly on the tranny... no freewheeling so whatever torque can be multiplied at that RPM is delivered solidly to the transmission.

I think this whole sprag issue is also what results in that secondary thunk that occurs when one quickly gooses and releases the throttle from a complete stop while in 'D'... a combination of this sprag behaviour and the 4.6s idle management.

My MM displays this sliding backwards behaviour at idle on a ramp, and the occasional movement of the steering wheel will counteract this due to the artificial raising of the idle RPM in response to power steering loads.

Well that's my theory anyway... hopefully someone with the service manual(s) or practical 4R70W experience can confirm or shoot this down.


Edit: in rereading my post (can't always trust my grammar skillz), I thought of a good example of the sprag-action... a normal 10-speed (okay, 12 speed, I guess I'm showing my age) bike. When you're not forcing against the pedals, the pedals freewheel. When you force against the pedal, the sprag locks up and you can apply force in that direction of rotation. Stand on a bike on an uphill and you can hold the bike in place because the sprag will be locked up (you're applying force on the pedals to keep the bike stationary and not rolling downhill, right?). But imagine no solid link between the pedals and the rear wheel, instead there's a fluid coupling called the torque converter. Might not be so easy to keep the car on the hill unless you're pedalling fast enough.

Edit2: Still not entirely convinced my logic is good, will need someone with internal 4R70W experience to confirm or disprove that latency of the sprag action. Please?

[MENINBLK]

The 4R75W in my 2K4 acts the same way as TTA describes...

[stevengerard]

sounds right to me triple, boy you are really old, road racing bikes are now 20 speeds! 2 Chainrings, 10 in the rear gear cluster

[JACook]

Well that's my theory anyway... hopefully someone with the service manual(s) or practical 4R70W experience can confirm or shoot this down. Well, I'm not gonna take any shots at you TTA, 'cause you actually got pretty close.
Not bad for someone that hasn't been inside one of these boxes.

Actually, though, the answer lies in how first gear works.

But first, it's helpful to understand the differences between the Simpson gear train in a typical 3 speed box,
vs the Ravigneaux gear train in an overdrive unit like the 4R70W. In the Simpson design, there are two
separate planetary gearsets, usually designated front and rear. In the Ravigneaux design, the two
planetary gearsets share a common planet carrier. Even so, you can still think in terms of front planet/
rear planet as far as function is concerned.

In both gear trains, first gear is actually double-reverse, and it's what's happening with the rear planetary
gear that prevents the car from rolling backward when in second gear.

When the transmission is in D, and in first gear, the rear planet carrier is held stationary by a one-way roller
clutch that's attached to the transmission case. (The 4R70W uses a roller clutch, not a sprag.) The sun gear
is the input, while the annulus gear is splined to the output shaft. The roller clutch allows the transmission to
overrun when you're decelerating, but that's only one reason why it's there.

The other reason for the roller clutch is because of what happens when the transmission shifts into second.
In second gear, all the same conditions exist in the rear planet, but now the sun gear shell is held stationary.
If the rear planet didn't have the one-way clutch, you'd bind the gear train. In a Simpson gear train, you
could use the low/reverse clutch or band in place of a one-way, but this kind of synchronous shifting
is harder to coordinate than the non-synchronous 1-2 shift. Because the Ravigneaux gear train has the
common planet carrier, you have to use the one-way.

Because we do have the one-way clutch, when the transmission shifts into second gear, the rear planet
carrier simply overruns the roller clutch.

As long as you're moving forward, that is.

In second gear, we have the rear planet sun gear held stationary, the annulus gear connected to the
output shaft, and the planet carrier overrunning a one-way clutch that's connected to the transmission
case.

What I've just described is a reverse-lock, no?
And this is why, when the transmission is in second, the output shaft physically cannot turn in reverse.

NB this condition exists whenever the transmission is in second gear, regardless of whether you're in drive
or manual 2nd. Of course you're only likely to notice the behavior in manual 2nd...

[TripleTransAm]

Awesome, Jeff. Thanks so much for the reply, and your explanation was crystal clear.

Do you (or anyone else, like David M etc.) know if the Hydramatics behave this way as well, in second gear only? Reason I ask is that I don't recall my GTA doing this in 1st gear nor the old '85 Parisienne (which probably weighed as much as the Marauder). But I think I did experience something like this every time I'd go 'un-store' my '78 T/A from its sub-basement dungeon by driving it up that (what seemed like) 80 degree incline! (I swear sometimes I thought that dungeon was heated by the Earth's core, it was buried so deep). When I get home, I'm whipping out my GM books to compare the torque flow in this reverse situation and see if it matches in any way.

Thanks again.

[JACook]

Do you (or anyone else, like David M etc.) know if the Hydramatics behave this way as well, in second gear only? Reason I ask is that I don't recall my GTA doing this in 1st gear nor the old '85 Parisienne (which probably weighed as much as the Marauder The THM200, 250/350 and 400 are all Simpson geartrain boxes, so yes, they would behave the same way.

But...

I don't ever recall driving a Hydramatic-equipped car that would actually let you start out in second gear.
That is to say, if you put them in manual 2nd, you'll still start off in first, then shift to second like normal,
but won't shift into third.

AFAIK, All the Ford automatics, from the old MX Cruise-O-Matic on up, have allowed you to start out in
second gear, supposedly to make for easier going in slippery conditions.

But yes, if the valvebody were calibrated to allow you to start out in second gear when you're in manual
2nd, the Hydramatics would exhibit the same reverse-lock behavior. Likewise for the newer 4Lxxx/4Txxx
Ravigneaux geartrain 4-speeders.

If it'll let you start out in second, you can use it to hill-hold.


But I think I did experience something like this every time I'd go 'un-store' my '78 T/A from its sub-basement dungeon by driving it up that (what seemed like) 80 degree incline! (I swear sometimes I thought that dungeon was heated by the Earth's core, it was buried so deep). When I get home, I'm whipping out my GM books to compare the torque flow in this reverse situation and see if it matches in any way. Hmm... Not sure what you're referring to here.
Even though the reverse-lock behavior is caused by how first gear works, it only actually reverse-locks
when you're in second. In first, whether in D or manual 1, the transmission will still allow the car to roll
backward, if there's enough pull to overcome the torque converter.

[TripleTransAm]

Hmm... Not sure what you're referring to here.
Even though the reverse-lock behavior is caused by how first gear works, it only actually reverse-locks
when you're in second. In first, whether in D or manual 1, the transmission will still allow the car to roll
backward, if there's enough pull to overcome the torque converter.


Sorry, unclear english...
What I meant to say was that in D, in 1st at a complete stop, I was pretty sure I'd witnessed my '78 roll down a really steep incline, albeit very slowly, and what perplexed me was that I can't say I remember the same occurring on my GTA on steep inclines, and likewise for my old '85 Parisienne (although memory is beginning to fade on this last one: it'll soon be close to 7 years since I last drove that car, and only in short bursts to move it around). Most likely torque converter tightness?

Same for why my MM seems more likely to roll back down a hill than my GTA in 1st, I'm thinking it's a matter of torque converter tightness... the looser converter in the MM allowing less 'grip' on the motor than in the tighter GTA, and with the GTA's healthier low end, it can hold the car up better.

In second, this would be replaced by a solid sprag / roller clutch action, as you described, negating any torque converter effect on the scenario. Correct?

And yes, in general the Hydramatics are not given the capability of starting in 2nd, with the exception of the computer controlled versions (the "E" at the end of 4L60-E, for instance) which were allowed to start in 2nd via a console-mounted button in 6-cylinder applications in the f-bodies, if not other applications.

Again, thanks for the detailed explanations.