Do speedometers get out of calibration with age and or mileage?

[TJ Hopland]

Do speedometers get out of calibration with age and or mileage?  Its reading high so maybe some sort of lubrication getting hard so what normally slips to make it work now sticks?  I have no idea did speedometer technology change much over the years?
 

I was recently working on a car that has more or less been sitting for the last 7 years.  I had worked on it around 7 years ago when at that time it had also been sitting for many years.  Working the bugs out back then I drove it a couple hundred miles and just don't have any memory of the speedometer being off but this time its significantly off, its reading high by like 15-20 mph on the highway. 

Back then I did have the transmission rebuilt and put slightly smaller radials on than the originals.  Most of my driving was after those changes and like I said I just don't remember thinking it was that far off.  My memory ain't what it used to be but after driving it again that would usually be enough to jog my memory and it just hasn't.  About the only hardware change since back then has been a fuel pump. I bouncy now which I also don't remember but I have and have had other cars in the last 7 years with bouncy needles so that is less noteworthy just thought I would mention it incase its a clue.   

[The Tassie Devil(le)]

G'day TJ,

The changing to smaller diameter tyres will definitely have an effect on the speedo reading, as the tailshaft will be rotating faster than with the originals.

As there is no physical contact between the Speedo Cable and the Needle because the Magnetic force of the spinning magnet on the Needle rotator controls the rotation.   The Hair Spring adjusts the tension of the magnetic force.   But, a weakening Magnet will only create a slower reading Needle.

The cause of the Bouncy Needle can be of two actions.   One being a sticking Speedo Cable (Lack of lubrication, or kinked) causing the inner cable to load and unload as it turns within the outer tubing.   The Second cause is by the lack of lubrication inside the Speedo Head, with the Odometer Drive Gears wearing from lubrication that is too old and sticky, and causing the small nylon gears to occasionally slip a tooth.   This mainly occurs through lack of use, by that I mean sitting and not being used at all and the grease just hardens over time.

I purchased a Speedo for my '60 CDV out of a car that had been sitting in a wrecking yard for a couple of decades, and boy, did that needle swing from zero to 120 (Needle Stop to Needle Stop) and made a hell of a noise that I had to pull over to the side of the road and disconnect the cable.   This flapping started about 10 miles from home on a test drive after installing the "new" speedo in my car.   When I drove out of the drive, the needle was working beautifully, but as I drove further, after about 5 miles, it started to "wobble".   The further I drove, the worse it got.

Turned out that the internal grease had set like glue, creating great resistance on the Odometer gears backwards to the Speedo Cable.

Bruce.

PS.   In the pictures, the needle axle is broken at the tip, but just long enough to fit the needle onto. 

[Jeff Rosansky CLC #28373]

I don't know if they get out of calibration or not. But, our 55 reads about 5 high. Our 70 reads about 5 slow. Go figure.
I think the grease in the 70's is bad because it doesn't read when cold, but the odometer works.
The 55's needle has a slight bounce. It was really bad and noisy but we replaced the cable and that seemed to fix it.

[Eric DeVirgilis CLC# 8621]

An out-of-calibration speedometer doesn't necessarily mean the accuracy of the odometer is any less. As Bruce points out, the mechanics of the speedometer are completely different than those for the odometer. Typically, a working odometer has less chance of being off than the speedo. 

[TJ Hopland]

Most of my driving back then was with the new smaller tires.  Its seeming like it must be my memory.  Maybe after I changed the tires I just figured that was the cause and didn't really think about it again.

I will have to see if I can find a tire calculator that lets me input old sizes and see if that is how far off its reading. 

So speedometers have always been the magnetic deal?  And when did they start?  20's?

[bcroe]

Bad grease will cause problems, maybe even break
parts.  The back bearing for the indicator needle
rests in a bearing in the center of the rotating
magnet, so when that sticks the needle gets
thrown against the pin. 

I took apart the speedo on a car (I bought new)
after 300,000 miles, and absolutely EVERYTHING
from the head down into the trans was worn out.

All speedos here needed to be calibrated for
my needs.  They can be made 10 times as good
as the factory delivered.  I start by getting
the odometer exactly right, then adjust the
needle to agree.  Note the rolling circumference
of a tire is tied to the belt inside, and tread
wear has very little impact.  Tire size changes
are not your biggest problem, axle ratio matters.
good luck, Bruce Roe

[James Landi]

Way back decades ago when "Consumers' Report" was called "Consumers' Union,"  I recently read one for their testing reports on a mid 50's Cadillac that INCLUDED a report on the accuracy of the speedometer.  On what was then a new mid century Cadillac, the speedometer was rather inaccurate... I don't recall how much, because they included several referenced speed points of the subject car, but I do recall that the faster one drove the more inaccurate the speedometer read.   (One personal data point:  Having owned a bunch of well used Cadillacs from the mid 50's through the late 70's, of the dozen or so, I don't recall any of the  speedometers or odometers lasting much past aproximately 130k miles. They were rather poorly designed and tended to lock up solidly at the odometer's tiny gears, thus rendering the entire cluster useless). 

[Lexi]

Way back decades ago when "Consumers' Report" was called "Consumers' Union,"  I recently read one for their testing reports on a mid 50's Cadillac that INCLUDED a report on the accuracy of the speedometer.  On what was then a new mid century Cadillac, the speedometer was rather inaccurate... I don't recall how much, because they included several referenced speed points of the subject car, but I do recall that the faster one drove the more inaccurate the speedometer read.   (One personal data point:  Having owned a bunch of well used Cadillacs from the mid 50's through the late 70's, of the dozen or so, I don't recall any of the  speedometers or odometers lasting much past aproximately 130k miles. They were rather poorly designed and tended to lock up solidly at the odometer's tiny gears, thus rendering the entire cluster useless). 

That is what I have noticed on mine. The faster you go the more innacurate they are. Fortunately I lubed my cable last year and at least it no longer jumps or makes noise. Clay/Lexi

[tluke]

but I do recall that the faster one drove the more inaccurate the speedometer read

So that time in the '70s when I was in my 20s and I got my '55 75 Series up to 100 mph just to see if it would, did I not really break 100? I've been telling that story for 50 years!
Terry

[The Tassie Devil(le)]

So that time in the '70s when I was in my 20s and I got my '55 75 Series up to 100 mph just to see if it would, did I not really break 100? I've been telling that story for 50 years!   Terry   

G'day Terry,

No.   Speedos were designed to read faster than real time, and up to 10% faster.   This was done to ensure that there was no accidental chance of anyone being caught over the posted speed limits.   This way, the Makers could not be sued for any speeding tickets that one received.

In Australia, the Law enforcement vehicles were a different matter, as in the past, it was a requirement that these Speedos had to be accurate to within + or - 2 KPH at 200 KPH and was part of the Tender Specifications.

These days, they are more accurate as the computers require accurate readings for all the parameters required for performance, fuel consumption and other readouts.

It was a real letdown back in the day when one thought their speedo was accurate.

Bruce.    

[axleech24]

My speedometer was reading about 5 miles slow while the odometer was reading correct.  Just last week, I took apart the speedo assembly and cleaned up the tar like grease and re-greased all the gears.  I aligned the needle perfectly with the 0mph line.  I took it for a drive and now the needle is 10 mph off.  How do I fix this?

[Jon S]

Graphite and oil is still the best lubricant for the cable

[bcroe]

My speedometer was reading about 5 miles slow while the odometer was reading correct.  Just last week, I took apart the speedo assembly and cleaned up the tar like grease and re-greased all the gears.  I aligned the needle perfectly with the 0mph line.  I took it for a drive and now the needle is 10 mph off.  How do I fix this?

If the odometer is accurate, your cable gearing is correct. 
Reading low, your rotating magnet may be getting weak, they
can be remagnetized.  Also be sure no layers of the
hairspring are touching.  Beyond this, some adjustment of
the spacing of needle cup from the rotating magnet will
change the sensitivity.   good luck, Bruce Roe

[The Tassie Devil(le)]

What you need is a speedo which is known to be correct, a Battery Drill with a fully charged battery, a square piece of shaft the same square as the Speedo Cable. (I use a square copper boat nail), and a small screw driver/tool, to adjust the slider at the end of the hair spring.

How I adjust the Speedo Needle when recalibrating is to use my Battery Drill to operate the corrected speedo and read the speed registered.

Then go to the one needing adjusting, and do the same to see how far it is actually out.   To adjust the hair spring, the speedo has to be partially disassembled so you have access to the hair spring adjuster, which is the piece of steel with the dab of red colour sealing the spring end.

Bruce.    

[bcroe]

I have done some past work to improve speedo accuracy.  On
the bench I needed a steady motor to drive the cable.  I used a
drill press with an AC induction motor for constant RPM.  These
are closely locked to the fixed line frequency, tiny bit of slip is
pretty constant with no load.  The multi level belt pulleys give a
few speed choices.   A problem is normal reverse rotation.  But
I fixed that with a reversing switch on the starting winding.  Also
use reverse for other things.  Note a brush (universal) motor has
very poor RPM stability.   

My approach is to get the speedo and odometer to agree.  Later
when installed in the car, no need to hold constant speed, just find
measured distances and compare to your odometer.  Agreement
will also make the speedo right. 

With the drill press spinning the odometer input shaft, take a
stopwatch and find time for a 1 mile odometer increase. For
example at 60 mph, it will take 60 seconds, at 30 mph, it will take
120 seconds.  Since there are 3600 seconds in an hour, divide
3600 by your stopwatch time to get the exact shaft speed.  NOW
look at the speedometer needle (unit oriented the same as in use)
and see if it is the same number.  Try this for different shaft speeds. 

If the speedo is off by a certain number for all speeds, it can be
brought back by turning the hairspringanchor a tiny bit.  If the error
varies with speed, the problem is the relation between the magnet
and the hairspring interaction.  I have already mentioned some
problems with these.  It might be possible to move out the needle
shaft bearing with a spacer (something like a sheet of paper) to get
lower readings.  I should take pictures. 

Once the speedo and odometer agree, it is a matter of getting the
right ratio speedo cable drive gears at the trans.  There are a lot of
options with a TH400, changing both the drive and driven gear. 
For some other transmissions it might be necessary to use an
external cable gearbox to get it just right. 

It has been suggested that cable gearing can just be calcuated by
using axle ratio and tire diameter (circumfrence).  That can get you
into the ballpark.  But there will be extra error, because the rolling
circumfrence is tied to the tire internal belt, sometimes that number
is available.  So a road test is needed.  Interestingly, that accuracy
is hardly affected by the external tread wear.  I found that out after
considerable driving.  But it can be affected by a different type tire,
with a different internal belt design.  Note, a tire in use is FLAT on
the road, not that perfect circle.  Think about a caterpillar track on
the road, forward distance for once around is the same as if the
track were layed out flat, from pin to connecting pin.  The height of
the track has no effect. 

good luck, Bruce

[axleech24]

Here is what my speedo assembly looks like.  Looks like there is a brass pin capturing the spring. So do I release the pin and pull the spring leg thru the hole a little bit and re-secure? I dont have a good way to test other than driving the car with a GPS.  So this may take a few tries.

[Lexi]

If this is analogous to that of a balance driven clock mechanism, shortening the hair spring will speed up the mechanism, while the reverse will slow it down. This is dependant on the hair spring being original to the clock. Clay/Lexi

[The Tassie Devil(le)]

Here is what my speedo assembly looks like.  Looks like there is a brass pin capturing the spring. So do I release the pin and pull the spring leg thru the hole a little bit and re-secure? I dont have a good way to test other than driving the car with a GPS.  So this may take a few tries.

NO NO NO.

You only need to push on the metal piece that the spring is held, and that has sufficient friction to be held in place when moved.   The piece is held by the central part that is attached to the central frame crossover.   It does move - trust me.

As I said above, "To adjust the hair spring, the speedo has to be partially disassembled so you have access to the hair spring adjuster, which is the piece of steel with the dab of red colour sealing the spring end."

Tightening the spring increases tension, allowing the needle to show slower speed.   Backing it off, shows faster speed.

Bruce.