First, I don't know have any experience working for Smith or doing this commercially. I have done several of my own instruments, with success. They are much like lobsters, hard to open and full of mysterious stuff.
I started this adventure before reading the excellent post from Mike Taglieri from the Brit-Iron. mailing list. Read it first. There are also some very good comments from Ron Thompson from the same list, and some correspondence with Ian Bardsley of the BMOC in Vancouver.
This site is very much a work in progress. Please contribute - tell me what worked and didn't work and we'll add it in...
If you have any more knowledge or experience to add, please email me at randell@shaw.ca
updated: June 21, 2000
and again on March 10, 2002
and August 22, 2002
and March 15, 2004 and Nov 22/2009
There is a good site on Smith car instruments, they are more lightly built, but have the same principles http://ourworld.compuserve.com/homepages/arhodes/Speedo.html
There are many excellent rebuilders of Smith
instruments out there.
I would strongly suggest that you send your instrument to one of them.
In Atlanta Georgia - http://www.joellevinecompany.com/index.html
| A Gagg & Sons - Specialists in repair and renovation of Smiths speedometers | |
| Foreign Speedo in Palo Alto | |
| Nisonger's | |
| APT instruments http://www.gaugeguys.com/ | |
| Gower Oaks (who will also sell you a new one) |
In addition, most Brit part suppliers will sell you a rebuilt speedo, with or without a core charge.
End of Disclaimers- You're on your own now.
This page is intended only for members of the Pete the Cheap society. With the Atlas speedo, I got lucky and found a loose screw that was easily removed. In the process, I cleaned and lubed the instrument and cleaned the glass (badly needed). If you need to replace the glass on an instrument, it will probably be dirty inside. Go ahead and take the rest of it apart. These are not complicated. Odds are a careful cleaning and lube (use sewing machine or other light oil sparingly) will give you a working instrument, and you can spend your $200 on more fun stuff, like a Drouin blower..
Incidentally, the original glass is an odd thickness, window glass will either be too thick or too thin. Take an old piece of the glass with you. A new glass will cost you about $5 at any glass shop.
Smith magnetic speedos and tachs are essentially the same instrument. A reducing gearbox is driven by the engine or rear wheel. It spins a cable composed of a spring wound wire turning inside a lubricated sheath. At the instrument end the cable spins a rotor with embedded permanent magnets. The rotor supports a needle like axle on which is mounted the indicator needle. On the axle is mounted an aluminum disc and a very light spiral wound spring. Although aluminum is not a magnetic metal, it is conductive. The rotating magnetic field induces an eddy current in the disc, and this current interacts with the rotating magnetic field to exert a small torque to the axle. This force is resisted by the spring et voila - a measurement of the rotation speed of the magnet. A steel adjustment plate above the disc can be moved up or down, varying the strength of the magnetic field and thus the measured force.
The clearance between the magnet and disk is critical, and not very much, just a few thousandths of an inch. For more detail, see Ian Bardsley and Ron Thompson's emails here. If something (like a bug, they just love to crawl in there and die) gets in there, it can cause the disc and thus the needle to attempt rotating at magnet speed, becoming very broken in the process. If the inner cable is too long, it will push on the magnet, causing it to wear away the pot metal retaining surface and make contact with the disc, also not a good thing. The force generated to turn the needle is very small, so a bug, cocoon or just plain dirt can easily hold the needle from rotating. If the lubricant dries out after only 30 years or so, the needle axle to disc bearing surface can get twitchy- although twitchiness it probably a cable problem. This quote is from a former Smith's technician, courtesy of Chris Ghent on INOAList; (the original post is here)
"Cables," he said, "should never be coiled tightly, but left hanging. The inners take a set and this causes a flick in the cable operation leading to wobbly needles."
" When people say, it can't be the cable, I've just replaced it, " he says, "Yes it can," and shows them the flick test. This is a Smiths test and is laid out in a pamphlet he gave me a copy of. What he referred to as "flick", they called "snatch". It goes like this... pull the inner out of the cable and hold it in a loop, one end in each hand, pointing up and held like you would hold a straw. Gently roll one end between thumb and forefinger and feel with the other hand whether it starts to turn immediately. If it does it is OK. When they are not OK they don't turn for a split second and then flick around, as if something is bent.
On the subject of cable length- "The inner should protrude no more than 7/16ths out of the top of the outer when the crimp is pushed back inside the outer as far as it can go. At this point the inner should stick out the bottom of the outer by 5/8. "
He made the point that the 5/8 sticking out all needs to be square, if some of it is still round then it will push the inner up into the speedo/tach.The consequences of the inner protruding too far into the speedo/tach is that it destroys the mechanism.
The toughest part is getting the bezel off and
back on. The truly cheap
can re-use the bezel two or three times. I find universal linear
gasket material (electrical tape) will cover these sins up,
particularly for Commando instruments, since the bottom of the bezel is
covered by the cup. This
is the simple and basic method of getting the bezel off with an old
screwdriver
- go around 4 or 5 times with small bites. If you
are
slow and careful, you will be able to get it back it on fairly
neatly.
Be aware that the 'new' crimp on bezels are straight up and down on the
bottom lip - without a special tool it is hard to get a clean
result. A lathe is recommended, yr friendly local machinist may
do it for a sack of beer. If you are careful getting your virgin
bezel off, it may go back on more neatly than a new one.
![]()

This is a 'grey face' speedo from a 67 Atlas. The tach is essentially the same instrument, without the odometer spur gear, axle and odometer. Note the provision for a trip odometer below the magnet rotor.
These are all the parts. As you can see, there's not much inside the case. Generally, I was impressed with the design and robustness of these much maligned instruments. There really isn't very much to break down, wear out or gum up. I can see that the most common problem would be the pressure put on the magnet rotor retainer by an over long cable. The gap between this rotor and the aluminum disc is critical, and if it fails, much can go wrong quickly.
This was improved for the next generation, the 'green dot' instruments found on commandos. See the next page for a few photos from a tach. Green_dot
![]()

![]()
This is what's under the face.
The drive cable turns the magnet rotor, which creates a rotating field between the magnet and the adjustment plate. This causes an eddy current and electric field in the aluminum disc. This field interacts with the magnet, and tries to rotate with the magnet rotor. This is resisted by the hairspring. The needle measures the force created. The steel adjustment plate is pivoted in slots on the right side, and fixed on the left with the adjustment screw and spring. Turning the adjustment screw varies the strength of magnetic coupling between the spinning rotor and the plate, which changes the strength of the eddy current acting on the aluminum disk, which changes the force thus adjusting the measurement.
The brass plate carrying the spindle and
hairspring mechanism is held
in by two more plastic rivets. Push the pins through and remove
the
rivets. Note the height of the adjustment screw for
reassembly.
You will have to adjust the reading if you disassemble this
plate.
The spindle plate and adjustment plate will lift off together.
![]()



![]()

This is the odometer drive. The axles for this lie in a slot
and are retained with punch marks. See Mike
Taglieri's Brit-Iron post for details on disassembly.
![]()
The adjustment screw has a serrated edge, and should be held in place with a punch mark just to the right of the white line in this photo. You will need to do the same, after adjustment.
![]()
This is the highly technical test bench and stand
at my modern facility.
I used a good speedo hooked up to a spare drive cable to calibrate my
electric
drill (conveniently, flat out it does 45 mph). I then hooked up
the
rebuilt and adjusted it to read 45. One turn of the adjustment
screw
is about 4 or 5 mph, so this is not rocket science.

The only other possible adjustment is needle position. If your speedo is out by the same amount (say 5 mph) everywhere on the dial, you need to reposition the needle appropriately.
![]()
Reassemble the bumpers on to the guts - I used loctite, since this was the cause of my grief.

Insert into case - it should be tight, so make
sure that the internal
screw rubbers are in place, and the foam gasket for the drive, and that
the alignment is correct, it only goes in one way. Replace
external
screw rubbers and install pillar screws. Replace reflector ring,
gasket, glass, gasket and bezel. Whether you crimp the bezel back
on is a measure of your confidence. Tape works good.