Barney,
Thanks for your usual and thorough reply.
THE VICTIM: This was a GT6 speedometer. I was riding in the car and
noticed the speedometer's only movement were violent jerks from 0 to 40
MPH. I offered to take a look at the thing before it completely
self-destructed.
As expected, there was too much end-play in the first assembly which
allowed the spinning disk to drift outwards and contact the side (vs.
the outer flange) of the pie tin. That contact is what caused the
violent movement of the needle. What seems obvious now, is that this had
been going on for quite a while and, in fact, I had to "unwind" the
speedometer two full rotations before the needle would rest lightly on
the peg.
CORRECTING END PLAY
On dismantling the speedometer and separating the first assembly from
the odometer and speedometer sections, I could see the scarring on the
side of the pie tin where it had been brushed by the spinning magnetic
disk. I could also see a lot of end play of the disk. The construction
had the spinning magnetic disk separated from the spiral gear for the
odometer take off by a thin (stainless) metal strap. The ends of the
strap were riveted to the first assembly -- this is what locates the
disk. The play was obvious at the strap as it seemed the magnetic disk
and sprial gear had separated slightly on thier shaft.
My repair was to wedge some small screwdrivers under the spiral gear,
which pushed it up against the metal strap and, also, raised the
magnetic disk to the limit of the end play. I then used a large hammer
and lightly tapped the magnetic disk down onto the shaft. This removed
all the end play -- fixed the immediate problem.
MORE ON THAT MAGNETIC DISK
While admiring my handiwork, I noticed the spinning magnetic disk
appeared to be soldered to a brass yoke and this yoke appeared to be a
press fit to the shaft. The "magnetic disk" is not a disk at all but
actually a near-rectangular, small steel strap about 1/2" wide. The
center attaches to the brass yoke and the outer ends are rounded to
match the contours of the outer flange of the pie tin. There also is a
hexogonal-shaped, spring steel stamping, about 1/4" wide, attached to
the magnetic strap about 1/3 of the way down from the edge. I don't know
the purpose of this stamped peice.
REPOSITIONING THE MAGNETIC DISK AXIALLY
Anyway, as I was looking at how the connection between the magnet strap
and the brass yoke, they appeared to be soldered together. Referring to
other comments that speedometer accuracy could be changed by how deep
the magnet fits inside the pie-tin flange, maybe this would be a means
of adjustment. That is, a dial-indicator could measure the distance
between the side of the pie tin and the magentic strap and, by moving
the strap in or out, this would change the speedometer reading slightly.
Assuming that if you move the strap inwards, the available magnetic
torque would increase. This is something that could be done in the
garage if you had the right tools:
1) Dial indicator.
2) Appropriate solder gun (or iron?)
3) An apparatus to accurately move the magnetic disk up or down the
brass yoke a few thousands while keeping the disk perpendicular to the
input shaft.
4) A means to measure the results.
But then you stated that you thought the axial play would have a minor
impact on the speedometer accuracy. It might alleviate that ~2 mph waver
but not provide large adjustments.
ADJUSTING THE AIR GAP TO PIE TIN FLANGE
You went on to say that trimming the ends of the magnet strap to
increase the air distance between the pie tin flange and the strap would
make the greatest changes to the speedometer reading. Naturally, taking
material off the end of the magnet would cause the speedometer to slow
down. That would help my cars but it wouldn't help the GT6; I think the
speedometer is now running slow.
To "speed up" a speedometer, you would have to increse the amount of
mangetism (how?) or lengthen the ends of the magnet (add material? how?)
and either option seems dubious in the garage.
If I fired up the Dremel tool and ground off a few thousands off the end
of the magnetic disk, would the grinding operation affect the amount of
magnetisim?
ON MAGNETISM
You say you can't add magnetism to the disk. Why not? I think I was
smoking in the bathroom during that part of science class but I thought
there was a way to magentize stuff. Of course, magnetizing something a
'known' amount would be the challenge.
THAT CLOCK SPRING
I don't think it would be viable to mess with the clock spring. I can't
get to it. I believe it is attached to the needle (indcator) shaft under
the speedometer face and, as you pointed out, the speedometer needle
does not seem to be removable. It also doesn't appear viable to get to
the spring from the pie tin side. Even if the pie tin wasn't in the way
(and, seemingly) soldered together, all the little wheels and gears to
the odometer are in the way. So, my take, is that the spring can't be
messed with directly.
ON WORN PARTS
I hadn't thought about a worn bushing causing shaft wobble of the magnet
thus effectively increasing the magnetic torque on the pie tin
(seemingly increasing the magentic disk diameter). That would make the
speedometer run slower. But, as you say, to replace that bushing would
require availability of special parts.
On the other hand, I thought the bushing was the "oil impregnated" type
and I have had fairly good luck soaking the bushing in oil. I removed
the first assembly which contains the bushing, spiral odometer gear, and
magnetic disk and immerse the whole thing in Marvel Mystery Oil for a
couple of days. This once corrected the wavering in a Midget speedometer
for the 10,000 miles that I continued to own it.
I recommend this procedure. This also works well for distributors, also.
ON CALIBRATION
It sounds like you have a pretty fancy calibration procedure. I am a bit
more pragmatic. While on the bench, I monitor the relationship between
the speedometer and the odometer. I do this with a variable speed power
drill, a long nail, and a stopwatch. I use the drill and nail to run the
speedometer to an indicated 60MPH and then watch the clock to see how
fast it takes to click off a mile.
I figure success would be getting 60MPH to click off a mile in a
minute. I if I could do that, then it would be a conversion gear box to
match the unit to the car.
CONCLUSION
I'm happy with getting the GT6 speedometer to act normally even if it
ain't very accuruate. The owner is messing with other things on the car
so the unit has not yet been reinstalled and road tested. I think my
next experiement will be to find a used 'B' speedometer for my 'C'. As
the car has 'B' rear gears, it might start closer to calibrated. Then
I'll futz with this unit until I either fix it or brake it.
Once again, thanks for your contribution to the list.
Bob Allen, Kansas City, '69CGT, '75TR6
|