Index Home About Blog
From: John De Armond
X-Source: The Hotrod Mailing list
Date: Mar 1992
Subject: Re: Transmissions Questions

>The other issue is smoothness of downshifting, and that's a fairly
>big one in road racing, which is why the NASCAR cars still use all
>three pedals when they run at Sears Point in June.  

Probably as important is NASCAR's (oh hell, let's call names, Bill 
France Jr's) abject fear of the big E. (electronics, of course.)
I sometimes marvel that MSD boxes are allowed.

>The F1 cars
>use electronics to match revs so that there's no driveline shock
>when the driver picks a lower gear; a conventional automatic isn't
>going to be that gentle, and if the rev mismatch between gears is
>high enough, you *will* lock the rear wheels.  I guarantee you it's
>a very disorienting experience to be going into a turn and suddenly
>to find yourself face-to-face with the guy who was just trying to
>outbrake you on the corner entry; you wouldn't believe how big a 
>pair of eyes can get.  And letting a conventional automatic
>shift for you through a set of turns seems like a really good way to
>get to know Mister Guardrail.  


Understatements of the year :-)  The ultimate auto is of course a 
CVT.  Though not available in production for cars, CVTs have been
available for motorcycles at various points over the last 20 years.
I've raced two example.  

First was the Rokon dirt bike.  This motorcycle had a 380 cc snomobile
engine (complete with pull cord!) and a snow-mobile-style torque converter
(CVT by another name.)  For those unfamiliar, this CVT involves a 
belt drive between two variable diameter pulleys.  The engine driven
pully is actuated by centrifugal force (RPM) and tries to increase
its diameter as the engine speed advances.  The other pulley is actuated
by a cam that responds to torque and tries to increase its diameter 
according to the torque requirement.  The net effect is a balance between
RPM and torque.  In practice, the engine revs to near its peak torque
RPM immediately and stays there while the vehicle accelerates (fast!).

The second bike was the Husky automatic.  This bike is actually a 5 speed
transmission with a series of centrifugal clutches that sequentially
shifts up according to ground speed.

The Rokon, despite its small engine, was the hardest accelerating dirt
bike I've ever been on.  It would loft the front wheel the entire
length of most any straightaway one encountered on a motocross track.
The Husky is fast but not quite as fast.

Riding either bike required a significant change in style.  The Husky
freewheels on closed throttle as a byproduct of the clutch design.
The Rokon effectively does the same because the torque pulley will
shrink when the torque demand goes away.  Several techniques were
developed by riders but mine was to stay on the gas a bit and use the
rear brake to control torque application.  I fitted a hand brake lever
to my Rokon to allow me to do that in either type turn.  A little power
would keep the gearing correct while the brake is large enough relative
to the bike's weight to make it practical.

From this experience, I can extrapolate to 4 wheels.  A car is too heavy,
of course, to use brakes for power control.  I believe electronics can
do the job here.  A solution might be to use accelerometer 
and accelerator pedal rate input to control gear selection.  When
lateral acceleration is present or when the pedal is being lifted, the
ratio is frozen.  When lateral acceleration is present but the petal
is being advanced, the engine can be allowed to slowly speed up.
When there is only forward motion (straightaway) and the petal is 
pushed or is wide open, then the engine speed is optimized for maximum 
acceleration.

Clicking down a notch in complexity, the transmission control computer
could simply delay downshifting until throttle is again being 
applied and RPM is matched, such as during a turn exit.  This is how I 
manually shift an automatic on a road course.   I'll blip the throttle
while downshifting and then save the last downshift until I apply 
exit throttle.  

Clicking down another notch in complexity, the shift is manually triggered
like the F1 situation but the shift is delayed until the engine speed
matches. The driver must push the button and blip the throttle in order to 
cause the downshift.

Either of the last two could be implemented with existing computer controlled
production transmissions.

I'm assuming here that throttle control is strictly manual.  If 
drive by wire with an integrated power management system is implemented,
then the driver's gas pedal input could simply be a demand for 
acceleration and the computer controls the engine and transmission 
in order to deliver. One might have to mask the engine sound to the
driver in order to not confuse him :-)  I know one of the hardest
parts of adapting to the Rokon was learning to ignore the exhaust 
note.

>My experience with PGs, though, refers to the car that I had in mind when
>I joined this list.  I'll be describing that in the coming few days
>as I get to know the list and its constituents; I just had to chime in
>on MGBs in road racing, since I actually know something about that. 
>Meanwhile, is there an FAQ file for this list, or an introductory 
>package?  I'd like to get started on the right foot...

Pretty much anything having to do with high performance goes here.
You got a hot rod 'tronmobile that you plug in at night?  Talk about it.
Me, I hotrod Datsun Zcars and am interested in automotive electronics and 
engine engineering.  We have several drag racers, a few road
course drivers and a few classic hotrodders on the list.  Noise factor is
practically zero.  Best list I've ever been on.

FAQ?  No but if you'd like to volunteer :-)  while on the subject,
what does everyone think of developing a resource list?  We've been
building one over on the Z-car list for quite some time and it is an
invaluable resource.  Someone want to volunteer to collect the list 
and either post it monthly or send it to me for automatic posting?
And while we're at it, how about someone volunteering to set up an
archive site?  We've had about 3/4 megabyte flow through the list so 
far (yeah!) so there's a lot of good poop available.  I would do this
but I'm not on the Internet proper.

Hey, let's have fun... Rapidly..

John

Index Home About Blog