A friend from the local pub has recently bought himself a new toy - a 1972 black Cadillac. 8.2 litre V8 :-)) with cream leather interior. It is apparantly the one used in the last of the 2001 series of 'Monarch of the Glen' (BBC TV UK), when Archie is chasing around the highlands of Scotland looking for Lexie in this black Cadillac.
While sitting in the pub one evening, he happened to mention that the dashboard clock was not working. He described the clock to me and from his description I thought I knew roughly how it worked and what might fix it. He dropped it in for me to have a look.
The clock mechanism is a mechanical movement, which is wound electrically using a solenoid arrangement. From an initial look at the coils of this solenoid - they did not look at all healthy, and they measured 0 ohms - burnt coils!
Further investigation also showed that there should have been a small relay/switch type contact on a small sprung brass/copper arm between the solenoid arm and a contact on the clock mechanism which is hit and winds the clock spring. This contact had burnt out.
The clock mechanism itself was luckily working perfectly - this was tested by winding the clock manually.
The solenoid part of the clock was easily removed using three small screws. As the solenoid part is removed, care must be taken as the main spring unwinds itself. With the solenoid part on the bench the full extent of the missing contact could be seen - a suitable donor contact from an old relay was soldered in place on the existing sprung brass/copper arm. The coils looked more complicated. Firstly the solenoid had to be dismantled - easier said than done as the unit had been pressed together and the ends of the coil center bars had been crossed to make them increase in diameter so they would not pull back through the mica and washers they were fixed with. This can be seen on the above photo as little crosses on the end of the bars with the coils on.
With some gentle persuasion from a pair of sharp wire cutters the ends of the bars were worked so the assembly could be dismantled. With the unit appart the coils were unwound. One coil had 168 turns and the other 170 turns on it. The coils were connected in series - I assume this is for 12 volt operation, with the coils being connected in parallel for 6 volt operation. Both coils were wound the same way, clockwise (when looking from the rear of the clock) starting from the center connection. Both the center connections were soldered together and the two coil outer connections were soldered on to the small terminals near each of the respective coils. The wire as it was removed was measured with a micrometer for its diameter - 0.31mm.
A suitable replacement wire was purchased from MAPLIN Electronics - 32 SWG enameled copper wire, 0.28mm (MAPLIN part YN88V). The 0.28mm figure does not include the enamel - with the enamel the wire is ~0.32mm in diameter. The replacement wire was the easy bit - how to wind the coils?
With the micrometer I measured the coil core bar - it was 5.5mm diameter, and the coils need to be no more than 7.5mm in length. I found in my bits box a small length of steel bar 5.52mm in diameter. I then got 2 small pieces of wood and drilled a 5.5mm hole in the sides of both and forced each end of the bar in to the wood. I set the pieces of wood at 7.5mm apart. I put a small piece of insulating tape on each piece of wood - the reason for this will become apparant. I then wound the coils as tightly as I could and tried to make the wire sit as neatly as possible on the former. I knew I would not get this is tight or as neat as the original coils as they had been machine wound.
With one coil wound I painted the outside of it with enamel paint - I only had a car touch up stick in yellow to hand - this was to hold the coil together and stop it unwinding. Once this paint had dried I then took one piece of wood off the bar - the insulating tape stopped the paint from sticking to the wood. Gently then the coil can be slid off the bar and put on to the solenoid core bar - then a little more enamel paint is applied to help the coil stay in place on the bar. This is repeated twice, once for each coil.
I'll be honest and say that the first few attempts I had at making the coils were unsuccessful!
With both coils on the solenoid it was reassembled - the coil bars were spread again by using the end of a sharp possidrive screw hit in to the center of the end of the bars. This increased the diameter of the bars enough to hold them firm in the washers and the mica plate. The coil ends were then stripped of enamel and soldered. The two coils together now measured 2.6 ohms.
The solenoid assembly was then put back on the main clock assembly - which was just the reversal of dismantling it, with the main spring being partially wound during the process.
Now came testing time - what would happen when volts were applied?
Being a chicken - knowing how long the coils had taken to wind - I started off using only 9 volts. 'Snapp!' - the clock wound itself and ran for ~30 seconds - as the contacts came together at the end of this time, another 'snapp' - it wound again. YIPPEE! I then gradually increased the voltage.
12 volts was a slightly louder 'snapp' and the clock winds for ~55-70 seconds.
The clock kept time for the 1 hour bench test - I event tested it on 18 volts with success - the clock winding itself for ~90 seconds.
The clock was put bar in the car - the housing in which it sits was stamped with the date 20 Aug 1971. The clock ran perfectly while I was taken for a test drive - I very nice smooth ride, and the noise of the V8 :-)
No modifications were applied to this clock - but in hindsight I could have added a diode accross the contacts to reduce the back EMF when the clock snapps and the contacts are opened.