After having suffered a number of unexplainable problems with erratic running on my T160 Trident, I decided that I really had to get to the bottom of it all.
First, I gave the T160 a full service with new oil and filter, set tappets at 0.008 and 0.010, stripped the three 626 concentrics, checked all the orifices, set slide height on the manifold and replaced the manifold rubbers.
I splashed some fuel in the tank and after some difficulty getting the carburettors to flood managed to get it started. Disappointingly, nothing had changed. Tick-over was non-existent and adjustment to the throttle stop screw resulted in fast running at 2000 rpm at which subsequent point it would inexplicably die. The carburettors then had to be flooded again for a restart on the kick-starter. (There wasn’t a hint of life on the electric start, even though it was cranking over OK.) Despite the problems I did manage to adjust the mixture screws to obtain relatively even firing.
Back to the drawing board. Re-strip the carburettors and double check that the float height is set in accordance with Triumph Service Bulletin 2/73 – Checking and Adjusting Amal Concentric Float Level. All OK at 0.080″. Retest. No change.
Believing that the Lucas Rita ignition system fitted years ago had now reached the end of its working life, I replaced all the ignition components with a TriSpark system from LP Williams. Some improvement. I was now able to start the engine on the starter, as long as the carburettors were flooded prior to starting. No improvement with regard to tick-over, etc.
As it was clearly a fuel delivery issue, I needed to see what was going on in the float chamber. I therefore set up a test rig off the bike, which allowed me to feed fuel to the float bowl and measure the float level.
One issue became obvious immediately. The head of fuel necessary to lift the Amal brass viton tipped 622/197 float needle off the seat without assistance from the float was a full 15 inches. Why is that a problem? The float rises until the tab drops the needle on to the seat and fuel flow ceases. However, the float and float levels are adjusted on the assumption that the float tab presses the needle to the seat to stop fuel flow. The actual level under these conditions is therefore much lower than expected. Thus my splash of fuel in the tank was quite insufficient to obtain a good result.
Problem solved? Not exactly, but at least it was now understood why the bike would run well at the beginning of a run and then become problematic at the end. Much of it was down to the head of fuel in the tank. Though I don’t understand why this has manifested itself in what seems to be the last 10 months or so when it appeared fine before that.
During testing I also experienced a problem where previous heads of fuel were sufficient, but were now ceasing to lift the needle from the seat. Checking through the fuel feed I found that the fuel filter at the bottom of the float chamber had rotated such that the plastic upright holding the gauze was obscuring the hole in the banjo. This was sufficient to reduce the pressure at the needle. Thus requiring an even greater head of fuel to lift it from the seat. Another note to self.
I then decided to revisit the issue noted in the Triumph Service Bulletin. I don’t know the circumstances for the bulletin or the Amal components being used at the time, but it’s not obvious to me why a carburettor would need such a radical adjustment from the designer’s optimum setting (fuel level 0.17″ to 0.24″ from the lip of the chamber) in order for it to work in what is a standard application. In any event, setting the float height to 0.080 below the lip exacerbated the problem with the float needle, as the actual level in practice (with the brass needle) was then lower still. It also explained why the carburettors were so difficult to tickle, since the ticklers couldn’t reach the floats to depress them.
The float levels were therefore set back to Amal’s recommendation. i.e. The top of the float is slightly higher and parallel to the lip of the chamber when the tab is depressed until the needle is held on the seat by the tab.
As they say, the proof is in the pudding. Filling the tank to the top ensured that I had a sufficient head of fuel to flow past the needle until the float forced the needle to the seat. With choke and no tickling the T160 burst into life from cold on the starter and almost went onto to immediate tick-over at around 1000 rpm. Definitely a result.
I have since tweaked the mixture setting and have run a wide open throttle test. The plugs show no sign of the over-rich condition alluded to in the service bulletin and the engine ticks over smoothly at about 600 rpm. In addition, I have replaced the brass needles with the lighter 622/197A aluminium version offered by Amal. These only require a head of fuel of around 6 inches to clear the seat, so should work for all normal fuel levels in the tank.
In summary, my observations are as follows:
1. Set the float level to that recommended by Amal.
2. Replace the brass 622/197 needle with the aluminium 622/197A needle (earlier nylon needles appear fine).
3. Ensure that the holes in the banjo are adjacent to the filter gauze and not the plastic uprights.
4. Fill the tank with fuel to at least the halfway mark before doing any tuning.
Naturally, this problem doesn’t only apply to the T160, but would likely apply to all bikes with MK1 concentrics, as it sorted a similar sporadic fuel starvation problem with my trials Cub.