Friday, May 24, 2013

New PVL twin outlet coils

I dropped by Sprint yesterday for some new ignition coils. When hot, the bike had been stalling at traffic lights, idling very unevenly after a run. The bike ran perfectly well for the first five miles or so. But then things would deteriorate so the stalling was definitely related to how warm the engine was, or else how long it had been working. The problem had been coming on earlier and earlier over the past week so I decided I wouldn't risk it becoming a permanent feature.

I was convinced it wasn't a fueling problem, after all the work I'd done on the tank, and the carbs, and the fuel pipes ... ! Starting from cold was OK - firing after maybe four or five spins of the engine on the starter.That's why I thought it was time to revisit the ignition system to look for an explanation. I'd previously had a problem with my Daytona cutting out when warm and that was due to electrical breakdown of the ignition pickup coil when the engine was hot. It had always started perfectly from cold and, after cutting out, would re-start from cold. In that case though, it was total failure with the engine stopping completely while riding at normal speeds.

With my Trophy, it was different because it would run reasonably well at normal speeds though would cough occasionally accelerating from low revs. I'd read about similar problems on some internet forums. They pointed to failing ignition coils as the potential cause.

On my Trophy, these are a pair of twin-outlet resin-encapsulated units manufactured by PVL in Germany. The PVL coils on my bike were the originals. I know because PVL stamp the production date just above the low-tension connections on their coils. Mine read 0491 - April 1991.

So that means they are 22 years and one month old. That's old enough to justify retiring them. The coils I bought from Sprint were marked 1112 and 0313 - November 2012 and March 2013. I was somehow amazed to see that their construction appeared to be identical in every way to the originals - same cases, same potting, same terminals.The only difference is the white print on the body, with the code 356 100. The PVL website lists that code as "High performance double Ignition Coil 2/4-cylinder Triumph, BMW", 1.1Ohm primary winding, secondary 7,66kOhm. My old analogue multimeter sort of agreed with this, given that the needle waves about quite a lot without ever seeming to settle. I noticed that the secondary windings on the old coils were reading about 10kOhm - higher than the new ones anyway. This measurement doesn't mean much though because the bike was cold when I measured the resistances and the problem only surfaces when hot.
I smiled to see 'Made in W Germany' still moulded into the plastic cases. It was hard to believe so much time could have passed with absolutely no evidence of design change. Of course, changes could be hidden inside.

I've been for one ten-mile run with the new coils so far and already the difference is noticeable. No more coughing on the throttle at low revs. Whilst idling on its centre stand after the ride, I got the bike hot enough or the fan to cut in. Whereas it would have been stumbling badly yesterday, there was just a hit of lumpiness. I was able to turn the idle speed right down to an indicated 500 rpm without the bike stalling. I must have left it running like that for a minute just to see before turning it back up to its regular 1000 rpm idle speed.

Sunday, May 19, 2013

Adjusting preload on front suspension

I had guessed at the best length of the spring spacer on my front forks, after deciding the standard spacer was making the front end too harsh. The standard spacer is 110mm long but, as my replacement springs are 20mm longer than the springs I took out, this was preloading the replacement springs quite significantly. So I made up a pair of 90mm spacers. These turned out to allow more sag than I wanted so I decided to add an additional packing for the replacement springs. Also, the 90mm spacers were from a narrower diameter tube than the standard spacer (which has a diameter of 38mm). 
90mm spacer shown loose in stanchion
 I decided to use o-rings to keep them centralized in the fork tube. I bought a Hilka O-ring kit some time ago and it included some that were just right.They made the spacers a snug fit in the stanchions so I have better confidence the forces will be centralized and distributedevenly.

Top 3mm disk being installed over 9mm spacer and lower 3mm disk
I added the additional length to the spacer I made up before with two 3mm thick disks and an extra section of tubing 9mm long. I cut the disks out of sheet aluminium with a drill attachment and a jigsaw: a very noisy business. It took a fair amount of filing to sort them out after the power tools but I was pleased with the result. This means the total spacer length is 86 + 3 + 3 + 9 = 101 mm.

I rode south to the Sammy Miller Museum today and was pleased with the result. The ride is firm but not overly firm. This is a matter of personal preference of course. It feels more controlled than with the 90mm spacer but just that bit more compliant than 110mm.
TriMoto 1200 outside the Sammy Miller Museum

Sunday, May 12, 2013

Posh badge

 The top yoke (or triple tree, if you prefer) always looked a bit Spartan to me. The pic below left was taken the day I brought the bike home, the pic below right was taken this weekend, after all the work I've done. Much cleaner and shinier. But I like the Triumph plaque on my Daytona and thought my Trophy deserved something similar and felt I could do something similar.
I found several Ebay vendors of Meriden Triumph patent plaques and couldn't believe my luck when I found one specific for the original Trophy. 

Patent plate and ally cover ready to fit
Unfortunately, it wasn't quite the right shape to fit exactly over the truncated triangular recess between ignition key and steering head nut. So I made up a cover from 1mm aluminium sheet to go between patent plaque and top yoke. If found the shape amazingly tricky to get right. I can't exactly explain why, other than that the rounded corners of the triangle made measurement very difficult.
Patent plate with top two rivets just locating into ally cover
The patent plate came with three brass hammer-drive rivets, each with an OD of 1.6mm and 4.5mm long. The patent plate and ally cover were both 1mm thick so I could expect a max of 2.5mm to secure into the top yoke. The top two rivets could thus locate into the top yoke, but the bottom rivet only to my bespoke ally cover because it was over the yoke's central recess. I used a 1.5mm drill bit to make sure the rivets would be a tight fit in the top yoke, drilling to a depth of 4mm. They took a fair old whack with a hammer to get moving.
Top two rivets driven though patent plate and ally sheet into top yoke
The result pleases me no end. So I consider my posh badge mission to be accomplished.

Reducing front fork preload

When I fitted stiffer fork springs, I noticed that the unloaded length of the new springs was 490mm. That was 20mm longer than the ones then installed (470mm). This meant that the new springs would be preloaded by the existing spacers quite a lot more than the springs I took out. At the time, I decided to leave it that way to see how things went. After about 900 miles since the rebuild, I've come to the conclusion that there is too little initial movement in the forks and the front ride height could be lower. It feels as though I'm not getting the most out of what the forks can offer to ride quality. I measured the static sag (how much the forks compress when off the stand) and found it hardly moved. The full 145mm stroke dipped hardly at all - too little to measure reliably though I think about 20mm. I think the ideal figure is said to be around 30% with rider on board (hence difficulty in measuring!), though opinions vary.
With the front wheel jacked off the ground, I removed the fork top caps and measured how far the original spacer tubes came up above the stanchion for reference. It was about 31mm. Add that to the threaded length of the top nuts and it means the springs were compressed about 50mm at zero loading, which is about 10%, compared to about 5% for the old springs.  
Original steel spacer tubes are 110mm long
The original spacers are 110mm long. I decided to reduce the length of the spacers in the fork legs by 20mm to compensate for the longer length of my replacement springs but I didn't want to cut the originals up. I had some spacers left over from an Ohlins kit I'd used on my Daytona and decided to cut them down for this application instead.

The Ohlins spacer tubes are made of an aluminium alloy, rather than the steel ones fitted to the Trophy, and they were of a smaller external diameter. That meant there was a risk that they would not seat neatly against the fork top caps or on the spring seats. As it happens, the forks have a spring seats that fits between the top of the springs and the bottom of the spacer tube. It has a lip that by coincidence fitted neatly inside the Ohlins tubes. So the lower end of the spacer wasn't a problem. It wasn't as easy to be sure about this for the upper end of the tube, where it buts up against the fork top cap.

Measuring up some aluminium disks as seats of the spacer tubes
I made a couple of disks from 1mm aluminium sheet to serve as seats for the top of the new spacer tubes and then cut the Ohlins spacers down to just under 86mm, allowing for the thickness of these disks and also with the idea that I could add extra disks in progressively to vary the preload more easily in the future.

Shortened alloy spacer tubes and seats
With the shortened spacers fitted, I checked how much they protruded from the unloaded fork tubes: the answer was 8mm, so I had reduced the preload by 23mm. The disks and caps went on a treat afterwards.The suspension immediately felt more compliant, pumping the forks up and down with the front brake held on. The ride tomorrow will give me a better idea still.

Note: I later decided that this was too soft and added an extra 10mm of preload spacers to the forks. 

Tuesday, May 7, 2013

Out and about in the South West

I took a trip out over the Mendips on Saturday. It was a beautiful Spring evening - ideal for riding in the hills. I stopped in Wells just as the last of the light was fading. This is outside the Wells Town Hall.

Friday, May 3, 2013

Fuelling fixed - "This time I know its for real"

So I took the float bowls off to check the float heights. The first thing I noticed was the amount of sludge in there. This was despite having completely cleaned them out 18 months ago and running the bike with two additional filters!

A good dose of carb cleaner soon shifted the goo but still ... I'm hoping that it was just a very fine rust from the tank and that my efforts to seal the tank will mean the end of it.

The floats were about 1mm too low, at 13.5mm compared with the spec in the Haynes manual. That might explain why the bike was hesitating just off idle and the amount of fuel evident when I pulled the carbs out. The floats are measured with the carbs inverted so they are bearing down on the valve that admits fuel to the float bowls. Low floats mean high fuel - can be confusing. So I set them all to 14.5mm.

I remember having trouble setting them when I rebuilt the carbs last. I had removed the plastic cages that hold the floats and remember that it was difficult to re-seat them afterwards. They push into an O-ring, which was springing back. No such trouble this time.

I did two other things to lean off the mixture as well. This is silly from a scientific perspective: any change could be due to any one or a combination of any two changes. But I thought, "while I'm in there, I might as well", as you do ;-). These were:

  • to reset the pilot jet screws which control slow running. Haynes says they should be two turns out from fully seated. As it was, they were between 2.5 and 2.75 turns out. So I set them all to 2 turns out.
  • to lower the jet needles. They were on the standard setting, which is position three of five. I set them all to position two. This really wasn't necessary but I wanted to see what effect it would have on the middle rev range and progression off idle.

I was pleased to see that the carb diaphragms were all in great condition - clean and supple.

I put new fuel pipes in (again!) with new t-pieces but no filters this time. I also fitted a new thick-walled vacuum pipe between carb 3 and the fuel tap. The routing of these pipes to the carbs is now clean, with no horrible bends - and a shorter journey!

She started up when the float bowls were full - and idled really rough. Not what I wanted to hear. The carb balance was way out after these changes. I suspect resetting the pilot screws was the main reason. After balancing them up with my old Morgan vacuum gauges, nice and even. I managed to get a tick over at an indicated 500rpm. Sweet.

I've done about 100 miles since and there is no sign of fuel starvation. I'm guessing the big change was the new vacuum pipe for the fuel tap. Better still, that hesitation off idle is no longer there. Result. I think the mixture is probably a bit lean in the midrange but nothing drastic. I'll run the bike like this for a few hundred miles than may put the jet needles back to position 3 again. That's easy to do with the carbs in place.

The sun is out. It's time to ride.