My Comet runs a Amal Mk1 Premier carb, 30mm, that has seen around 40,000 miles of use. I also run the original (refurbished) Lucas magneto and the Lucas mechanical ATD. I have tuned my ATD to give 4 degrees BTDC for starting and 30 degrees BTDC at full advance.
For a long time the motor ran and idled well - nothing unusual being observed.
Around 5,000 miles ago I noticed that after covering around 10 miles that the bike would no longer idle reliably. When I came to a halt the motor would idle as expected but after just a few seconds the idle speed would increase and the only way I found to get it back 'under control' was to put the bike into gear, apply the brake then slip the clutch, thus slowing the motor. But alas, after 5 to 10 secondes I needed to repeat the same process. The situation has been slowly but progressively getting worse.
Suspecting an air leak I removed and partially disassembled the carburettor, cleaning all the jets etc and also replaced the sealing 'O' ring on the carburetor flange. Following reassembly and a test ride - nothing had changed! So again, cursing, I removed the carb and on removing the slide (again) and with the aid of a powerful LED torch and a magnifying lens I was able to see some very small burs in the carburettor body - low down in the channel in the side of the bore that the locating pin on the side of the slide travels in.
Taking care not to add to the 'damage' to the bore of the carb body, with a very small jeweler's file I carefully removed the burs, then reassembly and a test run. The idle issue was improved but not resolved. Now the motor would idle reliable for around 30 seconds before the revs started to rise but the reves did not rise to their earlier level. Just the same I needed the slip the clutch technique to get the revs to drop.
I next turned my attention to the Lucas ATD - a seemingly simple device that relies on centrifugal force to advance the ignition as the engine revs rise and then springs to retard the ignition as the revs drop.
Those of you who have ridden a bike with manual ignition advance will have noticed that when the motor is idling, if you do nothing more than apply a little advance the engine revs will rise - then if you retard the ignition the revs will fall. With this in mind I turned my attention to the ATD.
Looking at the exploded photo of a Lucas ATD I noticed 2 very small shims/washers that go between the rocking arms and the rocking arm ins in the middle of the bob weights. These are tiny; approx 8mm OD x 5 mm ID x 0.25mm thick. I was fortunate enough to locate a source of hardened steel shims/washers of this exact dimension from a local supplier,
http://www.precisionshims.com.au/ .
When I disassembled my ATD I noticed that these washers were missing. I also noticed that there are small 'dimples' on the underside of the bob weights that are apparently there to minimise friction as the bob weighs slide back and forth over the back plate.
With the aim of reducing as much friction as possible I polished the surface of the back plate that those dimples slide over and also gently polished the face of the dimples as well.
On reassembly I fitted the new 0.25mm washers, and as I had obtained extras I fitted them on all of the pivot points in the atd, 6 in total. My aim was to minimise friction wherever possible. As the ATD springs appeared sound I did not replace them.
On reassembly of the ATD I made sure that there was a generous 0.010" end float after I pressed the retaining ring back into place. A check for free movement within the ATD showed no trace of any unwanted friction with the unit returning to its retarded position with a satisfying 'clunk'
All that remained was to refit the ATD, reset the ignition timing as detailed above, refit the Timing Case cover (needed to remove it in order to remove the ATD), check oil levels then go for a test run.
Well over a 20 local mile run just completed with multiple stops to verify the idle I can report that the problem of idle speed advancing all by itself
appears to have been resolved. In a few days time I will be doing a 100 to 150 mile day run and that I suspect will be the ultimate test of my handy work.