|
|
||||||
|
Introduction
My speedometer has never quite worked correctly. I attributed this at first to tire selection. Then eight years ago I installed a five speed (JT5), and it became even worse. Paul was very helpful with selecting new speedometer gears for the transmission, but I quickly ran out of adjustments at the transmission end. That meant that I had to attack the speedometer itself. I had been happy enough making the adjustment in my head, and had never bothered sending it out. Recently, I read Anthony Rhodes terrific write up on calibrating Smiths speedometers, and became inspired to fix the problem once and for all. It took a while to drum up a sacrificial speedometer with the correct calibration for my car, and attempt the job. The first attempt didn't work, but then I had a D'Oh! moment, and found a solution. Here's the story.
The Origin of the Problem
It's the same story as many of the peculiarities of these cars: the
British handled speedometer calibration in the most convoluted possible
way. To understand the point, consider any American car. Every US speedometer
that I've come across is calibrated for 1000 turns per mile. (That's turns
of the speedometer cable). So adjusting the speedometer is a simple matter
of changing the driven speedometer gear in the transmission. US speedometer
gears are mounted to the tranny in removable housings, and the big three
tended to stick with the same gear patterns across transmission designs
and through the decades. As a result, if you want to calibrate an American
speedometer, you have a wide selection of gears available from your friendly
dealer. Problem solved for under $5. This is true of the JT5, which takes
GM speedo gears. The only problem is that the nasty British speedometer
at the other end isn't calibrated for 1000 TPM.
|
|
||||||
|
How did Smith's solve this problem? In the British world, the ratio
of the speedometer gears on the transmission end is whatever the transmission
manufacturer likes. Angle drives may be reduction boxes or 1:1. In this
unfriendly world, the calibration has to be done at the speedometer head.
Each speedometer is calibrated for the car, taking into account tires,
differential, and transmission. The number of calibrated turns per mile
is written somewhere on the face of the speedometer, and it's different
for every car: on my '69 with 3.54 rear, it's 1312. Smiths intention seems
to have been to sell you a new speedometer if you ever needed a recalibration.
What with 40 years, different gearing, tire sizes, and transmissions, this
isn't really sensible today.
| Each Smiths Speedometer is Calibrated Differently | ||||||
|
Calibration Basics
To understand how calibration is done, it's important to understand that a Smith's speedometer is actually two or three instruments in one case. Speedometer calibration is completely independent from odometer calibration. And depending on vintage, the trip odometer can be a separate instrument from the main odometer. Each calibration is independent of the other. In my case, the instrument was badly out of adjustment: the odometer was 10% slow, as was the trip odometer. The speedometer was 20% slow.
To adjust the calibration, I began with the assumption that the calibration noted on the dial must be the actual odometer calibration, because the odometers are calibrated with hard gearing. So I went for a ten mile drive on the interstate, comparing the odometer reading to the measured mile markers. Since I had a 10% error, I knew that the rate at which the cable was turning had to be 10% less than 1312, which is the noted calibration. This works out to around 1180 TPM. Looking at the chart in Rhodes book, it turns out that there's a Smiths odometer gear that would calibrate for 1184 TPM, close enough if I could find the correct gear. Eventually, I turned up a cheap junk speedometer with the right calibration.
The next step is simple right? Swap gears. Not so simple. As Rhodes notes, there are three styles of Smiths speedometer. The one in my car is the newest style, the junker was the oldest. The old style has separate main and trip odometers, so two gears. The new style as in my car has a combined trip and main odometer, so only one gear. No problem, I'll use one to fix my odometer, one for spare, right? Wrong! The problem is that between the two designs, the dimensions of the gear subtly changed. So after hours of patient disassembly, back together it went with nothing accomplished.
That not done, it's on to the speedometer. The speedometer consists of a magnet bar that's rotated by the cable. The rotating magnet creates drag on a steel cup, which is attached to the indicator and a spring. The faster the magnet rotates, the more force drag is applied to the cup. The cup will pull the needle around the dial until the force of the spring equals the force created by the rotating magnet. Calibration can be accomplished by various means: changing the strength of the magnets, changing the distance between the magnet and the drag cup, and changing the strength of the spring. It can also be fudged by simply repositioning the needle. It should be noted that if the rest of the instrument isn't in calibration, repositioning the needle will mean that the speedometer only reads correctly at one speed! But then, it's most important to know your speed around 60 MPH: below, it's not a police matter, above you're already in trouble.
Out of curiousity, I did a standard calibration on both instruments (the one from my car and the donor). This is done by removing the needle. With the spring completely at rest, the needle is reinstalled so that it points to the white dot just below the zero. Then the spring stop is depressed, and the needle turned to up until it rests on the stop. Then I spun them up on my drill, which turns at 1200 RPM. The result was a surprise: both speedometers were off by about 10% of the expected reading. To check, I used an optical tachometer to measure the speed of my drill. To double check I spun up a spare Stewart Warner speedometer from my Plymouth. No doubt, both Smiths speedometers were reading low. My guess is that the magnets loose strength in time, and I bet it's happening to all of them. Keep this in mind as you scoot down the road.
The early speedometer, as it turns out, appears to have a simple method of calibration. The upper end of the shaft (the one directly behind the dial) passes through two plates. There is a screw that tensions the two plates, which has the effect of adjusting the end shake. This changes the distance between cup and magnet, and allows for fine adjustment. The newer speedometer, the one on my car, has no easy adjustment. The correct way of changing calibration is by unsoldering the hairspring, and feeding more or less length into the coil. A delicate and problematic process if you aren't practiced. It would have been so much easier if they'd simply pinned the spring, but no, it's the hard way or no way for Smiths. Another way to achieve calibration is to add magnets to the rotor. But as a stop gap, I chose the simple method of repositioning the needle.
The D'Oh! Moment
Since my first attempt at calibration failed, I began to search for a correctly calibrated speedometer with new-pattern gears. It occurred to me that XJ's and XJS's used American automatic transmissions through the 1970's. That meant that Jaguar had to have solved the calibration problem. Sure enough, it turned out that the speedometers from pre-1980 XJ's and XJS's were the same old Smiths units, calibrated for 1000, 1020, or 1040 TPM. This made things easy, since it meant that the speedo head could be calibrated near the "sweet spot" of the transmission. With a 1000 turn head, you're in the middle of the adjustment spectrum of available transmission speedometer gears. A trip to the GM dealer, and it's just about calibrated.
I obtained a very well preserved speedometer from a Series II XJ, for all of ten bucks. When I took it apart, I had the D'Oh! moment. All along, I had been looking for a donor to exchange gears in my stock speedometer's odometer. The XJ speedometer turned out to have exactly the same internals as the E-Type speedometer (which isn't much of a surprise, on reflection). The only difference was the odometer set shaft, which could be swapped without disassembly. So rather than take the time and risk of disassembling odometers and swapping gears, I swapped the whole insides! That was easy!
The last remaining problem was calibrating the new speedometer. The reason that the natural calibration doesn't work is that the XJ speedo was a 140mph unit, the E-Type was a 160mph unit (admittedly, off-calibrated). Once again, I chose to calibrate by "fudging the needle"... simply remounting the needle until the reading was correct at a test cable speed. Using my electric drill, I was calibrating at 1200 TPM. I needed to work out the setting required for the speedometer to read 60 when the car was moving at exactly 60. This is a simple matter of working out all the ratios: the distance traveled for each turn of the tire, the rear axle ratio, the ratio of the transmission speedometer gear, and the ratio of the speedometer head. You can get fancy by adjusting for tire wear. To see how I worked out the calibration point, you can go through my calculations by downloading the spreadsheet (click here). The idea is that you want to pick the transmission speedometer gear that comes as close to the 'ideal' solution as possible, then you need to work out the needle displacement. It's all in the spreadsheet. To use it, insert your data in the green input box. This includes the tire width and aspect, tire wear, differential ratio, transmission drive gear (you may have to speak to your transmission builder for this), the RPM of the tool you are using for calibration, and the nominal calibration of the meter. Once all that is entered, the answers pop out in the blue table, which will show you the calibration for every possible transmission driven gear. Simply choose the one that comes closest to 'ideal'. It works.
Solutions for Stock Transmissions
Speedograph Richfield advertises ratio adapters that can take the place of angle drives in Moss boxes and Jaguar four speeds. I'd be interested in hearing from anyone who's tried this solution.
Final Points
1) If your E-Type is getting 25 mpg, don't feel smug: the reason may
be odometer error.
2) If your car feels solid at high speed, you may not be driving as
fast as you think. 100 could be the new 60.
3) If it feels like everyone's in your way, your speedometer may be
reading low. 60 could be the new 100.
4) It's easy to do a crude speedometer calibration, hard to do it the
correct way. And you can't do an odometer calibration at all without the
right gears. There are no replacement gears available for Smiths, unless
you cannibalize an old speedometer.
Acknowledgements
Hats off to Anthony Rhodes, who's done an incredible service by documenting the innards of Smiths Speedometers. And many thanks to Paul Cangliosi and the gang at Medatronics, who invented the BW swap and helped me with many things transmission related.
Copyright©2007 Michael Frank
