manual trans shifting characteristics - the why

[sbrats]

Different manual gear boxes shift differently - some shift like a knife in butter, some not. In particular the notch that is felt in some boxes as the change goes thru neutral. What design characteristic creates this. Is it related to synco, gear or shifter linkage design?


tia

 

[MikeHalloran]

Yes.

Okay, not so much the gears, which are always in mesh.

There are detents that hold things in position so the transmission won't fall out of gear, or fall into gear, or rattle more than it needs to. Detents are deceptively simple mechanisms, usually comprising a recess, a ball, and a spring. The recess could be a groove, or a countersunk hole, or some similar feature. The geometry of the recess, at a fairly detailed level, determines how much force it takes to move the shifter sleeve, or the shift rod, or whatever is being detented. The size of the ball relative to the recess, and the chamfers on all the parts, and the spring rate and force, determine, whether the shifter takes a good shove to knock it between gears, or whether it's just a two- finger deal.

Wait; there's more. Gearboxes have sliding surfaces, too, where the shifter forks slide as they move the shifter sleeves. When GM converted their manual boxes to use ATF instead of real lube, they didn't adjust the length of the sliding surfaces to compensate for the terrible friction characteristics of Dexron (cheap, small, wet clutches grab like crazy; so does everything else; nothing wants to slide). It's just like the "monkey sliding on a stick" toy that you learned to analyze in kinematics class; the monkey doesn't slide real well, so the box feels balky. Real gear lube helps, but then the synchros don't grab right. Recent boxes are less awful, but nothing like a Ford top-loader with a fresh load of grease in the shifter, and nothing at all like a real (old) Mini-Cooper.



Mike Halloran
Pembroke Pines, FL, USA

 

[drwebb]

In GM's defense, MT's constitute less than 10% of the US passenger car market- and maybe evel less of US OEM sales- so they may have left their MT development up to the interns . . .

 

[MikeyP]

The force from the detent is just the small initial clunk that you feel (assuming you are starting from neutral (otherwise it is the small clunk as you pass through neutral in changing from one gear to another) The rest of the "feel" is due to the synchro. After the detent, there is a force as the cone and ring of the synchro start to bite and spin the gear up or down to speed, then there are a further two (or sometimes 3 depending on which flanks of the dog teeth touch first) clunks as the dogs initally meet and then slide into place on the gear and the synchro releases. Of course the actual force from the synchro and detents has to pass along the shift linkages before it gets to your hand so the character changes depending on the design. In AMTs (clutchless manuals) the stick is just a fancy electrical switch and the computer does the rest.

M

--
Dr Michael F Platten

 

[BiPolarMoment]

drwebb
In GM's defense, MT's constitute less than 10% of the US passenger car market- and maybe evel less of US OEM sales- so they may have left their MT development up to the interns . . .

I never would have thought this was true until I worked as a Valet. 10% is probably generous when considering the total market--at least with non-economy cars--it was probably less than 5% in my experience, which is pretty sad considering some of the automatics were porsches

.

 

[drwebb]

"considering some of the automatics were porsches"

Several advanced AT designs can compete with a MT for fuel economy (not as sure about power losses), and a double-clutch auto can shift a lot faster than a person, so expect to see fewer tradional MTs in general and on performance cars especially.

 

[NormPeterson]

I hope not.

On the street, less time required to effect a shift is almost completely without significance (with the exception of dubious merit being the rather adolescent bragging rights associated with 0 - 60 in under "x point zero" seconds vs "x" plus a tenth or two).

Quite separate is the notion that any electronic control system isn't nearly as intelligent as an alert human operator with good awareness of the mechanical bits of the machine that he is operating. It's better at following programmed directions, to be sure, but it's not very good at anticipating situations about to occur and acting accordingly.

Re: the original topic - having the linkage internal to the transmission does eliminate the damping effects of dirt and grime on the effort required and the feedback "feel". The 5-speed Tremec 3550 I now have in my V8 RWD car works far smoother and with less effort than either of the external linkage 4-speed boxes and shifter mechanisms that were in it previously. It's a bit notchy on the 1-2 shift when it's cold enough outside to make ice (using GM Syncromesh fluid, as recommended), but that goes away within a couple of miles as the fluid warms up.

Norm

 

[BiPolarMoment]

Several advanced AT designs can compete with a MT for fuel economy (not as sure about power losses), and a double-clutch auto can shift a lot faster than a person, so expect to see fewer tradional MTs in general and on performance cars especially.

I know this is true but that doesn't make me want one. I've driven a slightly older SMG BMW M3 as a Valet and the feel is terrible compared to a traditional manual. Also, the SMG Maserati was even worse, but that's probably not a good baseline since I've heard the Maserati SMGs are horrible.

As far as automatics go, I'm pretty surprised with the Cayenne's. The torque converter is set up so that the car generally doesn't "creep" at idle, which I think would make a majority of auto drivers more considerate if it were applied to other cars.

 

[izzmus]

drwebb, there will always be a mass of consumers - enthusiats involved in amateur racing - who will prefer three pedals and a wobbly stick over two pedals and a paddle. The problem is the HAL factor. The driver wants to car to do something, and the electronics balk and say "Dave, I can't do that." My favorite was the poor Audi TT driver who tried to do a handbrake turn. The computer unlocked the center differential, which was thoughtful, but it also refused to open the throttle plate, which was quite a surprise.

On the other hand, such self-preservation schemes develop a rich aftermarket for "chips" that disable them. Never ceases to amaze me how people will buy a new car, then spend hundreds if not thousands more to circumvent the built-in deficiencies, voiding the warranties in the process.

 

[Peter7307]

Don't overlook the marketing aspect of it all as well.

Some car makers wanting to have their vehicles percieved by the test driving customer (who is in the car for as little as twenty minutes in some cases) as "sporty" will ensure the design gives a "notchy" change by altering the items mentioned in the second post.

To many people a notchy change equals a sporty change.

Pete.

 

[sbrats]

Thanks ...
I wanted to know as I find that for a fast track day shift change I find that the 'notch' really set the tone for a quick change - (not that I try to shift as a drag racer - and the std cluch won't take it anyway)

I was trying to figure out why I get shifter rattle in 3rd - if internals (balance?) was causing something - It seems, but I'm not sure (yet) if it relates to how fast I shift (the faster I go into 3rd, the more likely I'm to have a rattle) As it's FWD Getrag 6spd I plan to go under and see if linkages (extenal if any) might be hitting something. Just try'n to learn...

 

[geesamand]

I rebuild manual gearboxes (pushrod shifted, FWD 5 speeds) occasionally, and I get feedback that the shifting tends to get kind of mushy and resistant after rebuilding.

I suspect the resistance is from the shift rods not sliding as smoothly in the aluminum bores of the housing, even though I deburr the rods and lube it well at assembly. Any insight as to how to improve this situation? Would some hand lapping b/t the housing and rod help?

Dave

 

[MikeHalloran]

Guesses:

- Use the trans' designated gear oil or whatever as assembly lube. The conjecture is that your assembly lube is preventing the oil from reaching the rods.

- Use a different gear oil. Redline fans report that Redline MTL works great. I have no personal knowledge of it, and they're all dealers too, so some caution is advised.

- Chuck the rods in a lathe and polish them to a mirror finish. You will love SimiChrome polish for the last step.

I don't think I'd touch the rod bores, except to make sure they're squeaky clean.



Mike Halloran
Pembroke Pines, FL, USA

 

[BowTied]

Norm:

I am curious about what car you are referring to that you have.. I have been consdering changing my Muncie for a new tremec TKO600. Any advice?

Ron

 

[NormPeterson]

1979 Chevy Malibu.

It was not a difficult swap at all, although it was necessary to modify the tunnel sheetmetal (and the already slightly custom exhaust routing) somewhat. Even the shifter location as-shipped was good enough to permit re-using the previous 4-speed transmission's shift handle, only having to drill the holes slightly oversize.

Norm

 

[BowTied]

Thanks Norm.

I have heard (in particular to the original question in this thread) that the internal rail shifting of the tremec is quite a different feel compared to external rod linkages on other transmissions like a Munice or Saginaw. What is your opinion Norm?

 

[NormPeterson]

All else equal, I wouldn't go back to an external rod linkage design. See the third paragraph in my earlier response.

Norm

 

[sbrats]

Back to my post (go up 7 posts) If I sense a oscillation in 3rd (through the shifter) by pulling the shifter off the stop, I will assume that there is a spring that holds it against the stop, into a small dead zone but not to the point that I am going into neutral, the oscillation goes away (or not felt) – but this isn’t seen in any of the other gears.
I am going to crawl under and see if the linkage is hitting something in that position but was wondering if the was a sign – or signal of some internal problem.