Hi Mike,
We'lI, I opened up my tachometer to see what was causing it to read high and I found that the disk was rubbing against the top of the magnet. The
questions of the moments are why and what can be done? Since you're one of the few that have ventured inside the magic boxes, I thought I'd share what
I've found and pick your brains.
I have removed the disk/spring carrier assembly and find evidence of the contact with the magnet on the lower outer side of one part of the disk. I
assumed this was due to the magnet riding too high and I went on to remove the magnet/worm/cable spindle assembly. It is a pressed-up assembly and the
worm is held against the retaining bar by a very thin wave-spring washer. There's no real evidence of wear.
In your web page, you say that the adjuster moves the disk higher, but I can't see how. The adjuster plate moves up/down with the screw, but it's not
fastened to the disk at all. As far as I can see, it varies the magnetic field return path which flows through the disk and returns via the adjuster
bar. Shortening the path would increase the overall field strength, increasing the coupling etc. The disk spindle seems to sit in a bearing at
the centre of the magnet and thus the distance between the disk and magnet should be controlled by the bearing & spindle - being brass, perhaps its
worn down.
Assuming I am correct - and I hope I'm not, it seems that the magnet could be riding too high because of wear in the worm-retaining bar interface &
disk spindle/bearing - probably due to lack of lubricant (it was quite dry in there), or upwards pressure from the cable. Being unable to separate the
worm/magnet & retaining bar, the only remedy I can think of is to shave a few thou of the top of the magnet - it has what I looks like an aluminium
slotted face of similar thickness to the disk. This would restore some magnet to disk clearance, but I am less than enthusiastic.
The metal on top of the disk could also be mumetal - a magnetic shield. This would channel the magnetic field through the slots making a more defined
field pattern and aiding with the eddy current flow.
Anyway, enough rambling. Any thoughts you have will be appreciated.
Regards,
Ian Bardsley
Don't have a whle lot of experience, but will share what I have.
Is this an early i.e. grey face pre 69, or the later Commando type? There were quite a few improvements made some time between 68 and 73, notably the way the rotor is retained.
Ian Bardsley wrote:
Hi Mike,
We'lI, I opened up my tachometer to see what was causing it to read high and
I found that the disk was rubbing against the top of the magnet. The
questions of the moments are why and what can be done? Since you're one of
the few that have ventured inside the magic boxes, I thought I'd share what
I've found and pick your brains.
I have removed the disk/spring carrier assembly and find evidence of the
contact with the magnet on the lower outer side of one part of the disk. I
This may indicate that the ball/socket the spindle passes on the adjuster plate is/was stuck and causing the disk to tip. It should stay square to the rotor, and the eddy current would tend to keep it that way as the alum disk doesn't spin.
assumed this was due to the magnet riding too high and I went on to remove
the magnet/worm/cable spindle assembly. It is a pressed-up assembly and the
worm is held against the retaining bar by a very thin wave-spring washer.
There's no real evidence of wear.
In your web page, you say that the adjuster moves the disk higher, but I
can't see how. The adjuster plate moves up/down with the screw, but it's not
fastened to the disk at all.
Should be..
As far as I can see, it varies the magnetic
field return path which flows through the disk and returns via the adjuster
bar. Shortening the path would increase the overall field strength,
increasing the coupling etc. The disk spindle seems to sit in a bearing at
the centre of the magnet and thus the distance between the disk and magnet
should be controlled by the bearing & spindle - being brass, perhaps its
worn down.
Assuming I am correct - and I hope I'm not, it seems that the magnet could
be riding too high because of wear in the worm-retaining bar interface &
disk spindle/bearing - probably due to lack of lubricant (it was quite dry
in there), or upwards pressure from the cable. Being unable to separate the
worm/magnet & retaining bar, the only remedy I can think of is to shave a
few thou of the top of the magnet - it has what I looks like an aluminium
slotted face of similar thickness to the disk. This would restore some
magnet to disk clearance, but I am less than enthusiastic.
Now that I think about it, it may only be urban legend that too long cables cause premature failure by wear. There would have to be a _lot_ of wear on the retaining bar to rotor interface to use up the rotor disk clearance. That much metal would be very evident inside. More likely is a rivet failure- which should also be evident.
The metal on top of the disk could also be mumetal - a magnetic shield. This
would channel the magnetic field through the slots making a more defined
field pattern and aiding with the eddy current flow.
Anyway, enough rambling. Any thoughts you have will be appreciated.
Regards,
Ian Bardsley
Sounds to me, though, that the aluminum rotor/spindle assembly has lost whatever retains it to the moving adjuster plate. It should have very little float on the adjuster plate- IIRC, there's a kind of ball/socket affair where the spindle passes through the adjuster plate, but I don't recall what holds it in place. It moves quit directly with the adjuster.
I'm attaching a few full size closeups of the grey face atlas speedo- you can zoom in pretty close, and maybe notice any differences. This was/is a working instrument, other than stray parts bouncing around. What I do notice is how very little clearance there is between the alum and rotor- and this is adjusted correctly.
Tsawassen is coming up- there will probably be a few dead Smith clocks on offer,
good parts donors. The good lord and SWMBO willing, I might be there.
--
Miker
Ian Bardsley wrote:
I made a couple of obervations today that I thought you might find
interesting. Drilling out the rivets naturally generated some swarf that attached itself
to the magnet. Couldn't get it off with my fingers or the vacuum cleaner, so
I resorted to using my magnetic pick. The bits congregated in the magnet
slots and they have an affinity for the pointed tip of my pick - probably
due to focussing of the pick field at that point. Worked quite well.
Because the other end of the pick has an embedded magnet, I used it to sense
the polarity of the speedo magnet. The slots in the magnet top cover are
poles and they are arranged N - S - N - S. So I think my theory on the top
cover being to focus the field may be correct.
The contact area is between two of the slots and I noticed that the
interface between the slotted top cover and the magnet itself was just a
shade wider where the contact had been. So I measured the combined thickness
of the magnet & cover between each slot. This showed that the cover was
slightly bowed to about 0.005", with the widest part where the contact has
been made (and opposite, where no contact has been made). Putting it face
down on the cover allows me to rock it slightly.
I put the bits back into the casing without fasteners and I find that the
magnet pulls itself upwards to the adjusting bar. If I hold the magnet
retaining bar down against the casing, the magnet "levitates" towards the
retaining bar and there is clearance between the magnet & disk. Hopefully
this continues to be true if I ever find the right sized rivets to put it
back together again. I may flatten the magnet top cover just to be sure.
Ian
Interesting stuff.
It looks like I won't make it to Tsawassen, as the weather forecast has improved greatly for the weekend, and I have farm stuff that I have to get at. I asked Ken Jacobsen to pick up any cheap non functioning Smiths for me, and intend to do more surgery on them- partly to accumulate some spares for the internals (I think most parts are common. The rotor for the tach has a worm machined on- and it doesn't drive anything). I am making up a jig to roll bezels back on with a lathe, which I now have access to. Will take some photos and post them.
--
Miker
Hi Mike,
Too bad you can't make Tsawwassen. Here's the latest - sorry about the length, but I need to share it with someone who is interested:
I trial assembled the magnet & disks assemblies using screws & nuts (#2 & #3
respectively), less than ideal, but no rivets to be found. The disk still bottomed on the top of the magnet. So, I got aggressive and carefully took
2 -3 thou off the top of the magnet cover using wet & dry on my glass plate.
Getting the swarf off the magnet was quite the chore (I carefully plugged the bores). This restored the air gap and it then seemed to run cleanly. I
have since lubed & assembled the body and replaced the dial & needle.
This am, I trial calibrated using my drill and a green spot speedo I borrowed from a club member. My theory is that the speedo and tach use the
same drive mechanism, so the calibration marks - those little white dots at 30, 72 and 123 (or so) mph should line up with the corresponding marks on
the tach at 2, 4.8 and 8 x 1,000 RPM. Using my drill, the speedo gives about 123 MPH, so I adjusted the tach to correspond. BTW: my tach is marked 4:1
and my drill say's 2500 rpm, so the 8,000 RPM this indicates is in the ball park (the drill's getting a bit tired).
The good news is that the tach now runs smoothly throughout the range. When I first revved it up, it read way up at the top of the scale, but I had the
adjuster plate cranked down. It took almost all the adjustment (i.e unscrewing the adjuster) to get it calibrated (should've took more off the
magnet). Hopefully, it will stay set. I plan to check my calibration against a tach - if I can find someone locally with one hanging loose.
FYI: after running the disk/spring/adjuster assembly though my fingers quite a bit, I am now convinced that there is no connection between the adjuster
plate and the disk - it works through varying the magnetic coupling. The disk - magnet clearance is set by the disk spindle - magnet bearing. On
close examination, the bottom of the spindle has a conical hub (several times larger in diameter than the shaft), and the magnet bore has a
corresponding brass bush - probably in the top of the worm gear. So the spindle rides on top of the magnet. As the magnet rises, it lifts the disk.
If this bearing wears, the air gap disappears - and the instrument reads high, firstly due to increased magnetic coupling, then due to drag (recall
that my disk & magnet had wear marks). Also, I'm convinced that the wormwheel is intended to ride against the magnet retaining bar - since it
had a spring to keep it there. This isolates the magnet/disk bearing from cable end thrust.
So what do I take from all this: my problem was due to loosing the air gap, probably due to wear in the spindle/magnet bearing, again probably due to
lack of lubrication. Unfortunately, this looks likely to happen again - you can only put a little lube in this bearing. I should probably have taken
more off the magnet (at least .005"). I measured the penetration of the cable stub into the magnet spindle, and there seems to be 75 - 100 though
end clearance - so I doubt cable thrust is the problem (no brass from the worm gear in evidence).
Anyway, now for the bit I dreaded from the beginning, replacing the bezel. FYI, my old one was quite rusted and bits fell off as I un-crimped it. Need
to figure a technique to roll down the edge on the new one. I'm not declaring victory yet, but I'm reasonably sure it'll work for a
while. I found your web page and Mike Taglieri's epistle invaluable. If you're interested and this works, I'd be pleased to write it up so's you can
add it to your web page.
See you at Positive Earth. Cheers,
Ian
Ron Thompson wrote:
>I trial assembled the magnet & disks assemblies using screws & nuts (#2 &
#3
> respectively), less than ideal, but no rivets to be found. The disk still
> bottomed on the top of the magnet. So, I got aggressive and carefully took
> 2 -3 thou off the top of the magnet cover using wet & dry on my glass plate.
> Getting the swarf off the magnet was quite the chore (I carefully plugged
> the bores). This restored the air gap and it then seemed to run cleanly. I
> have since lubed & assembled the body and replaced the dial & needle.
I'd be surprised if you actually took .001 off with sand paper. It takes
plenty to do that. It almost sounds like you have a lot of slop ply in the
bearing on the worm gear side. Or you have a lot of vibration wear on the
alloy case where the adjusting bar sits as a fulcrum.
> This am, I trial calibrated using my drill and a green spot speedo I
> borrowed from a club member. My theory is that the speedo and tach use the
> same drive mechanism, so the calibration marks - those little white dotsat
> 30, 72 and 123 (or so) mph should line up with the corresponding marks on
> the tach at 2, 4.8 and 8 x 1,000 RPM. Using my drill, the speedo gives about
> 123 MPH, so I adjusted the tach to correspond. BTW: my tach is marked 4:1
> and my drill say's 2500 rpm, so the 8,000 RPM this indicates is in the ball
> park (the drill's getting a bit tired).
They are similar but not the same. Of course I use a slower drive sourse,
1200 RPM. My Triumph black face OIF tach, 4:1, registers 5500 indicated at
1200 RPM cable speed. Also you are assuming that your tach drive from the
engine is a 1:1 output.
> FYI: after running the disk/spring/adjuster assembly though my fingers quite
> a bit, I am now convinced that there is no connection between the adjuster
> plate and the disk - it works through varying the magnetic coupling. The
> disk - magnet clearance is set by the disk spindle - magnet bearing. On
> close examination, the bottom of the spindle has a conical hub (several
> times larger in diameter than the shaft), and the magnet bore has a
> corresponding brass bush - probably in the top of the worm gear. So the
> spindle rides on top of the magnet. As the magnet rises, it lifts the disk.
> If this bearing wears, the air gap disappears - and the instrument reads
> high, firstly due to increased magnetic coupling, then due to drag (recall
> that my disk & magnet had wear marks).
Yes this is the premise that the disc and magnet operate by.
>Also, I'm convinced that the
> wormwheel is intended to ride against the magnet retaining bar - since it
> had a spring to keep it there. This isolates the magnet/disk bearing from
> cable end thrust.
You lost me here. I have a RSM 3003/C0A in my hand and have no "magnet
retaining bar." The worm gear / magnet is held by the alloy case in a
bearing.
>
> So what do I take from all this: my problem was due to loosing the air
gap,
> probably due to wear in the spindle/magnet bearing, again probably due to
> lack of lubrication. Unfortunately, this looks likely to happen again -you
> can only put a little lube in this bearing. I should probably have taken
> more off the magnet (at least .005"). I measured the penetration of the
> cable stub into the magnet spindle, and there seems to be 75 - 100 though
> end clearance - so I doubt cable thrust is the problem (no brass from the
> worm gear in evidence).
This is quite likely. The spindle on the needle is steel and the bearing
"cup" is brass. Either wear or pounding. The abration on the magnet wheel
and aluminum disk are your worse enemy. They increase the turbulence effect
and make the needle read high and erratic. You can lube this bearing area
but take care that you don't fill it and that what you use is not so thick
as to induce more movement on the spindle.
> Anyway, now for the bit I dreaded from the beginning, replacing the bezel.
> FYI, my old one was quite rusted and bits fell off as I un-crimped it. Need
> to figure a technique to roll down the edge on the new one.
I would check your tach drive ratio first to make sure you have the
calibration correct. unless you want to drill an extra hole in the back to
make adjustments without tear down.
The best way to re-crimp is to "roll" crimp the lip back down. I do this on
a lathe. Chuck the body of the case near the base where the cable comes out.
I made a wood disc that fits over the bezel and I run the tailstock into it
to clamp the bezel to the case. I run the lathe slowly and with an old
router bit with a bearing on the end clamped in the tool post, I run the
bearing in and slowly roll the edge down. You may be able to figure another
way to do this without a lathe. Banging it down with wood block gives you a
lumpy crimp job, and you take the chance of cracking the glass.