One of the first jobs to do before wheeling the bike out of the workshop for the first start was to make a detachable prop stand. I did not want a stand as a permanent fixture on the bike as it would be somewhat out of place, even if it is a replica, of a bike intended for an attempt on the World Speed Record. There is also not much space to fit a stand.
Although I had a paddock stand, seen in pictures in my last blog posting, this needs 2 people to be used safely and is not ideal if I am alone. It’s also not feasible to carry a paddock stand around if out for a ride! I therefore constructed a prop stand to go onto the nearside footrest that can be easily detached and carried in a rucksack. The bike stands well on this.
The stand itself consists of 2 pieces of aluminium, shown in the pictures below with the bike on the ramp and before chamfering the top and bottom surfaces.
Time to start the bike!
My good buddy John Taylor came round to help (John has been a constant source of encouragement throughout the project – Thanks Mate) and we put the bike on the starting rollers …and, with my wife on the phone camera, the bike fired up for the first time.
Well, it was a great relief that it actually runs! …and with no discernible vibration (the effort on the crankshaft design and manufacture has paid off). However, that was the one and only run; the bike would not start for a second time. It’s not difficult to see that the engine was somewhat reluctant to start at all and would only run on 2 cylinders at higher engine speeds and with full ignition advance.
Now, it is well known (at least, if you have a V-Twin) that V-Twin magnetos can be problematical in the respect that you can get one good spark and one considerably weaker spark (see eg https://www.themagnetoguys.co.uk/v-twin-magnetos for an explanation in the context of K2F magnetos on Vincent twins). There was certainly evidence of this here in the respect of only wanting to run on one cylinder at lower speeds and with the ignition in the fully advanced position.
As I am now 70 years old, my days of push starting a substantial motorcycle by running alongside, dropping the clutch (or valve lifter) and jumping on are over! The bike has to be able to start using the kick-start. I owned a Vincent Black Shadow for many years and had the same issue with starting in the respect that the bike would only fire on one cylinder at low speeds. I replaced the original Lucas K2F magneto with a new BTH self-generating magneto and this absolutely transformed starting – one lazy kick on the kick-start and the engine would fire instantly. I therefore decided to fit one of these on the AJS. BTH make magnetos in a variety of shapes and sizes and I ordered a 50 degree, base-mounted, anti-clockwise rotation, 45mm spindle height self-generating magneto to replace the 50 degree BTH magneto that I had originally fitted
The new magneto, on the left in the above picture, is slightly larger than the original but the important dimensions – distance of spindle taper to flange, securing bolt threads and location on the base and spindle height are identical and the new magneto bolted straight on.
with perfect chain alignment. It is extremely easy to set up: the magneto is locked with the 6mm spindle (seen inserted in the hole – the brass blanking screw has been temporarily removed) and the crankshaft position set to give the desired fully-advanced timing (36 BTDC) on the rear cylinder. As I had made Vernier adjusters to set both the valve and ignition timings all that was now needed was to make sure the upper run of the chain was in tension and to insert the pinned washer through both sets of aligned holes. Job done. The electronics takes care of the ignition advance vs rpm with a pre-programmed curve .
The magneto uses external coils and a small junction box which need to be packaged neatly – yet to be done.
After reassembling the timing cases, oil pumps etc.. it was time for another start. The 2nd start is not captured on film (the Chief Photographer had gone shopping) but I can report that the bike fired instantly on the rollers and we were able to check the oil return to the tank and to play around with the tap that controls the oil pressure in the line that supplies oil to the camboxes and the crankshaft via the crankcase drilling.
With the tap fully open, there is free flow to the pipe that
delivers oil directly into the crankcase between the cylinders and the pressure
gauge registered at the bottom of the scale at about 5 psi.
With the tap fully closed, all oil from the feed pump goes into both the vertical drilling that connects with the crankshaft and the supply to the camboxes (the pipe on the right side in the above picture delivers oil to the flow restrictors at the camboxes) and there is substantial oil pressure at around 60 psi.
It didn’t take long to realize that the tap-closed position delivered far too much oil to the camboxes and it came cascading out of the gap between the rocker and the cambox on the front exhaust and the rear inlet. This was not really surprising for 2 reasons: firstly, I had only inserted a single 0.6mm diameter orifice into each flow restrictor (the AJcette has 2 orifices in series) and, secondly, a consequence of the orientation of a 50 degree V-Twin means that the oil can more easily exit the cambox because of the orientation at +/- 25 degrees of the rear cylinder inlet and the front cylinder exhaust .
The single flow restrictor was therefore replaced with 2x 0.6mm orifices in series, shown below at the far left of the picture. These have been drilled at different radii to ensure that the upstream hole cannot feed directly into the downstream hole. The fibre washers provide both sealing and, in the case of the restrictor plates, allow a small, narrow plenum to exist between the holes.
With the new BTH magneto fitted and the revised flow restrictors in place it was time for the next test.
The bike would now start on the kickstart, I have balanced the carbs, the engine will idle and oil no longer pours out of the camboxes, although this will require some further calibration and checks.
Overall, a successful set of test runs and it shouldn’t be too long before I can start road testing.
