The ball and disk integrator principle was used in the famed Norden
Bombsight in WWII. There was something referred to as a Tommy Box (SP?)
that I think used surplus Norden Bombsight parts to make a mechanical rally
computer.

Furthermore, these variable ratio friction drives go back to the early
1900s. There was a "car" that I read about once that a transmission based
on two wheels at right angles. One was like a flat surface flywheel on the
end of the engine crankshaft. The other was like a small tire that could be
slid along a shaft across the surface of the spinning flywheel. If the tire
was near the rim of the flywheel, it was driven fast. If the tire was slid
toward the center of the flywheel it was driven slower and slower with more
reduction ratio. If you went past the center to the other half of the
flywheel, you had the same ratio possibilities in reverse! The tire was on
a splined cross shaft that drove the wheels through a chain. I think the
engine ran at constant speed. Who says continuously variable transmissions
are a "new" thing?

SD







>Message: 3
> Date: Thu, 28 Apr 2005 15:22:55 EDT
> From: WaltKammer@...
>Subject: Re: Weird machinery
>
>In a message dated 4/28/2005 1:28:12 PM Eastern Standard Time, satch@...
>writes:
>Bill Jonesi has directed us to the "Navichron" clock (cheaper than a Heuer
>Sebring, anyway!). I wonder if any Halda geeks on this list know anything
>about a strange black box that Rusty Link installed in my Saab 96 over
>thirty years ago (along with his Halda Twinmaster): this box had two cable
>connections, one coming in from the dead-wheel drive and the other going out
>to the Twinmaster.
>
>Inside the box was something called a ball-disc integrator, and the front of
>the box had a vernier dial. Its purpose was to adjust the relationship of
>the incoming and outgoing cable speeds, giving you infinite adjustment of
>Halda. No swapping of X and Y gears! Pinpoint accuracy!
>
>Link, of course, does not remember exactly where it came from, and I've
>never encountered another like it. Anybody ever see one of these or know
>anything about it?
>
>Satch Carlson
>Team AFRICA
>Satch,
>
>How you doing up there? Been a while!
>
>Anyway, relative to your topic, yes these mechanical variable ratio devices
>used to be fairly common in the process control industry as well as some old
>machine tools.
>
>There were two different designs, one which had a "ball" which could be moved
>down the length of a cone. One side, usually the cone, was the input shaft
>and the ball end was the output shaft. Depending upon where you positioned
the
>ball via a mechanical screw adjustment, it would ride around the cone at what
>would be a certain circumference.
>
>To change the ratio up or down the adjustment shaft would in essence move the
>ball to the "fatter" end of the cone in order to make the ratio go in one
>direction and to the "pointy" end of the cone to make the ratio go the
other way.
>
>
>The design you have (disk/ball) is the one which was mechanically more
>complicated, although not much more. The disk could be positioned at the
"equator"
>of the ball for a ratio of unity, and could be moved to the north pole of the
>ball for making the output shaft turn fewer turns for a given input shaft. In
>reality the "one to one" point was, by using different diameter disks, made
>at the 45 degree north latitude position for models which needed a +/- ratio
>available. Usually the ball was the input end of the shaft and the disk the
>output.
>
>Also, most of the designs did not actually have a full "sphere" as the ball
>inside but rather a hemisphere since mathematically there was nothing to be
>gained by having the ability to adjust in the "southern" hemisphere of the
>arrangement.
>
>Similar principal for the ball/cone models that for rally use usually had the
>unity ratio a little less than halfway up the cone, since the "tip" of the
>cone was really not a workable position to use.
>
>Anyway, these were also used in some older mechanical aircraft instruments,
>although even when I had my single engine pilot license I never personally saw
>the "guts" of an instrument which actually had one. I was told by an old
>timer that they were used in there somewhere...
>
>In fact, there was a chap in Toronto who used to sell these things specially
>packaged for rally use and they worked great. When I ran with Iain he always
>had one installed in the car usually either to his Halda/Speedpilot setup or
>internally via the Langwell odometer he used, which accomplished the same task
>with a slightly more compact design of infinite ratio input to output shaft
>adjustments.
>
>Some of the later Halda Speedpilots (maybe the MkX or something) had an
>option which likewise did offer a limited range of infinite adjustment but
I don't
>recall the details of that design but I seem to recall it was a ball/cone type
>deal, albiet in a modified execution. I had Martin rebuild mine so never
>took them apart myself.
>
>Do you remember exactly what the little blob you used in ths Saab looked
>like? I presume like everything else of the period it was crinkle black in
color,
>but I do have some recollections of the setup Iain had.
>
>I had a ball/cone one installed in one of my really early rally cars (...get
>ready...) a 1965 Mustang Coupe, but to make it work in a worthwhile manner I
>still had to use one of the little inline adjustable plastic gearboxes that
>Gene sold in order to get the total number of rev's/mile down to a number that
>made sense for the ball and cone. The big problem was that if you got too
crazy
>with the RPM of the cables, the disk would slip on the cone (or ball) and
>unfortunately damage the rubber coating that was there, thus making the setup
>suitable for the trash bin. Fortunately when I used to buy these things there
>was an Army/Navy surplus joint in Syracuse where I could get plenty of
spares...
> only fabrication was to cobble something up for the end to convert the fat
>keyed shaft into the more typical speedo cable of the day. I must admit
that I
>don't know what happened to them all since between busted ones and "surplus"
>spares I probably had a dozen or so of them... the corrected cable originally
>went to a microswitch creating a contact closure for a 12VDC Veeder-Root
>resettable electromechanical counter that we used for odometer readouts of the
>corrected official miles. I still have those, and even the microswitch
setup, but
>the ball and cone things are gone.
>
>Take care and hope our paths cross soon...
>
>Regards,
>Walt Kammer
>
>
>[Non-text portions of this message have been removed]