To my other TR6 pages

December 18th, 2014

Tachometer and Speedometer

[Click the pics for a better view]

My speedo and tach were both intact, but I remember having some trouble with both of them before I put the car away.




I started with the tach because it is simpler.  When I inserted a small flat screwdriver into the drive spindle of the tach and apun it between my fingers, there was sometimes a chatter or stutter to the rotation, suggesting to me a worn bush.  

The front bezel twists, and pops off, while the internal mechanism is released by two screws in the back.




Two more small screws release the front face after the pointer is carefully pried off.




The innards reveal how the tachometer works:  When the input spindle is spun, it turns a bar magnet inside an aluminum cup.  The moving magnetic field of the magnet induces eddy currents in the cup, which in turn create their own magnetic field.  The two fields interact such that the cup tries to rotate with the magnet with a force that is proportional to the speed of the magnet.  The cup can rotate, but is restrained with a small spiral spring so that the deflection of the cup is determined by the speed of the magnet.  The pointer for the tach is attached to the small spindle that holds the cup.

The second picture below shows the magnet and cup separated.




Since I needed to replace the bushing that the input spindle rides in, I had to get the spindle out of the bushing.  Unfortunately, there was a keeper on the spindle that was riveted to the main casting.  The two rivet heads are visible in the second picture below.




I drilled out the two copper rivets and removed the magnet and spindle.




I wasn't able to locate any replacement bushings, either as replacement parts or from general industrial suppliers.  It is a bit of an odd size--3/8" OD and 6 mm ID, and fairly long.  My fallback was to make a bushing.  I started with a hunk of raw Oilite rod, which is what the original bushing appeared to be made from.  Oilite is an oil impregnated pourous bronze, which is ideal for use in places that probably won't get external lubrication.  I brought the bush to final ID with a 6 mm reamer after it was pressed into the casting.  The spindle fit was good, and showed no more tendency to chatter.




The spindle assembly then had to be remounted on the casting.  I didn't have the right kind of rivets, so I driled and tapped the rivet holes for some 6-32 screws.  I really should have used smaller screws since the holes in the spindle keeper had to be enlarged.  Also, I had to reduce the height of the button head screws slightly so the magnet wouldn't drag on them.  It seemed like a good idea to use non-magnetic screws to replace the copper rivets, so the screws were non-magnetic stainless.




With that repair done, I turned to the case.  The case was apparently bare zinc plated steel on the outside, but had some rust. There was yellowed white paint on the inside.  There are also a couple of translucent green plastic domes riveted in the case for the instrument lights.  I wanted to bead blast the inside of the case, so the plastic domes had to go.  I expect that the green paint patch was to subdue the light intensity close to the bulb.




After bead blasting and painting the inside and outside.  The little domes were remounted with 2-56 screws and nuts.




Before I put everything back together, I wanted to check the operation and calibration of the instrument.  I remounted the dial face, and when I retrieved the pointer, I was horrified to find it snapped in two.  I'm not really sure how it happened.  The pointers are not available new, and used ones are getting rare.  A generous member on one of the popular TR6 forums kindly donated a junk tachometer for the pointer.  I noticed that the donor pointer was much whiter than mine, but still yellowish compared the the white lettering on the dial face.  Before I installed it, I masked and painted the pointer a whiter white.




This is a calibration rig I put toghether when I rebuilt the instruments on my Triumph motorcycle.  It consists of a univeral motor (a Sears Craftsman chain saw sharpener) modified to run in reverse, a speed controller, an optical tachometer to read out the speed of the motor, and a stand to mount everything on.  I can dial in a speed for the motor, and note the reading on the instrument.  This tach is marked "3.55:1" on the dial, which means that the indication on the dial should be 3.55 times the RPM of the tach inner cable.  (The tach drive gears on the distributor have a ratio of 16:9, and since the distributor runs and half the speed of the engine, the RPM ratio between the engine and the tach cable is 16/9 x 2 = 3.5555..., which rounds to 3.56.  Smiths made a small math error.)




The tach operation was smooth and steady.  However, graphing the curve of dial reading vs input speed, the tach was indicating consistently about 10% low.  I have not run across a source that reveals how to adjust calibration on the instruments, but I did try a few things.  There are some arms that connect the two ends of the magnet, and they appear to have been bent slightly up toward the disk at some point.  I surmised that this was a calibration adjustment.  I played with this, and it did seem to help a little, but I still couldn't get the reading to be on the money.  I suspect the magnet has weakened a little in 40 years.  Some have gone so far as to piggyback an additional magnet, but this seemed a little anal, even to me.  In the end, I decided to be content with a good-looking, smoothly operating instrument that reads a little low.  In practice, I rarely use the tach anyway.

Final assembly continued.  The original rubber mounting pad was still in excellent shape.




The bezel components were still usable except for whatever was in the bezel next to the front side of the glass.  It came out in small shards, and I'm not sure if it was some kind of gasket or a caulk.  I found some rubber cord used to repair window screens that was the right size to provide a seal against the glass.




The speedo is next.  It's internal workings are very similar, but it has all the odomoter gear, too.  One good thing is that the input shaft bushing on the speedo seems to be OK.



I'll update for the speedo if there is anything relavent.



Update--January 7, 2015

The speedometer, though of a design similar to the tach, presented its own special problems.

The first issue explained the rattling sound I heard from inside the case.  The pointer stop wire is normally fixed to the blue lens for the hi-beam lamp, but the barrel behind the lens had broken, and the wire and back part of the lens were loose in the case, along with the little brass washer that holds the wire down.

 


Next issue came in removing the pointer.  That little brass ferrule is supposed to come off with the pointer, not stay on the spindle.  In my life, I've worked on four tach/speedos--two on my Triumph motorcycle restoration, and these two.  On three out of the four, I had pointer problems.




Both of these were fixed with a little of the appropriate adhesive--a solvent based plastic adhesive for the lens, and a metal filled epoxy for the pointer (OK, it's JB Weld).




I did also decide to replace the input bushing like for the tach, but the speedo bush has a different OD.

 I really wanted to clean the odometer wheels.  This is simple for the trip odometer because they can be rotated with the reset shaft.  However, there is no way to turn the main odometer wheels on a later instrument other than through the input shaft (which would take a while), or by removing the wheels.The stack of wheels has a lot of tiny internal parts that caused some hesitation, but I forged on anyway.  There was a while there where I was really sorry I did that.  After a tense hour or so, everything was cleaned and back together.  I decided to set the wheels back to 00000, since the number that was on there had no significance anyway--the speedo cable was broken for much of the time I drove the car.




Then on to calibration.  I took a deeper dive into calibration than with the tach, and it payed off.  




I graphed the response curve for the instrument so I could compare it to what it should be.  Most speedos have a number on the face that indicates how many revolutions of the cable should be needed to register one mile on the odometer.  A little thought will suggest that this number is also the RPM of the cable that should register 60 MPH on the dial.  In my speedo, this number was 1120.  This allowed me to generate the response line that the instrument should have.

My speedo read nearly 20% low throughout its range.  




Since the operation of the pointer was smooth and free, without catches or oscillations, I thought I could rule out a mechanical problem, at least to start.  A weak 40-year old spring would probably make the readings high rather than low.  This left the magnet.  Older magnet materials exhibit a decline in magnetic strength with heat, shock, and just time.  On the theory that this was the causeof the error, I researched ways to rejuvenate the magnet.  This is mostly done with strong electromagnets, sometimes using very high current pulses.  All that is really neded, though, is a relatively strong magnetic field to boost the strength of the magnet.  

I have some pretty strong rare earth magnets, and in playing with them, I learned something new about magnets: If you have two magnets, one a lot stronger than the other, their like poles will not repel, they will attract.  This really confused me since I couldn't  determine which was the north and south poles of the speedo magnet by using a strong magnet.  It turns out that this is a known effect.  The strong magnet overwhelms the field of the weak magnet, temporarily reversing its poles.  Further, there can even be some residual effect on the weak magnet.  

After this learning experience, I checked the speedo magnet with a compass and found that the compass would react to the speedo magnet's north pole, but not its south pole.  It was as if the magnet didn't have a south pole.  Now, the physicists among us will tell us that this is really unlikely, so probably I'd just displaced the south pole to somewhere else on the magnet.

Fearing that I'd ruined the magnet, I put the speedo back together and re-ran the calibration check.  I'd succeeded in making its performance a lot worse.  It was then reading almost 40% low.




Since I at least learned that I could manipulate the strength of the magnet, I thought if I designed the operation a little more intellegently, I might be able to restore the magnet to its full strength.  Knowing from the compass which was the north pole of the magnet, I made a strong magnet with known poles that would bridge the speedo magnet.




I applied it to the speedo magnet, and stroked it back and forth for a while.  I really had no way to control or modulate the process, but I surmised that when the magnets were originally made, they were probably magnetized to saturation--that is, they were as strong as the material would allow.  If this was true, there wasn't much danger of making the magnet much too strong.




Back into the calibration fixture, and I was delighted to see that the response line was now a little above the theoretical.  The speedo now read a little high, which I believe was the way most speedometers were in new cars.




This is the way I'll leave it, and I'll re-open the tach to take care of its low readings.

I did all the calibration without the odometer pawl in place, so none of it registered on the odometers.  I installed the pawl and keepers, and then ran it in the cal setup to make sure the odometers worked OK.




Ready to go back in the case.  I use a rubber cord to seal the bezel to the glass.




Ready to go into the dash.




Comments to:  elhollin1@yahoo.com

To my other TR6 pages