Monthly Archives: May 2014

Gear Linkage and Rear Light

OK, so the two are in no way related, but that’s sort of how it goes, Once it’s down to the little bits, one works on what comes to hand or needs doing next.

The Gear Linkage
Once the engine was reunited with the frame there was the small matter of hooking up the gear linkage. It didn’t fit. Not only did it not fit, but it showed no signs of ever intending to fit. Just about everything was in the way. Firstly the gearbox selector shaft was about an inch inboard of the frame member between it and the pedal and the Aprilia  lever was straight. That one was solved by turning the original lever through 90 degrees (so that it pointed downover) sawing the end off and drilling a hole to take the ball end. Then I had to grind away about half the actuator plate for the side stand switch, to get a straight(ish)  run from pedal to lever. That looked OK and, as luck would have it, the original linkage was the right length. Once that was in place it was obvious that the up and down movement of it when changing gear hit everything remaining.  In the end the only answer was to remove the side stand switch and its mounting bracket  and to slightly crank the linkage.

GN250 rear set gear linkage

Rear Light

By this time it was pretty much down to cleaning things up and sticking it all back together. The original rear light and bracket are admirably visible, but about as huge and ugly as only a 1980s designer could produce, so an alternative had to be found. A little LED light from China cost about a fiver and a convenient lump of ally was fashioned into a bracket to hold it and the number plate. The contrast between the two is quite dramatic.

rear light comparison

The Noisy Bit – Engine Re-assembly

No, those are not pans and vegetables in the background at all. I would never dream of rebuilding an engine in the kitchen while my girlfriend was visiting her father.

The first engine task (whilst waiting for the shiny new barrel and piston to arrive from China) was procuring a cylinder head. One would think it would be a simple matter, but they’re clearly kept on the same shelf as the hens teeth and the rocking horse poo.  I’m guessing, by the lack of heads and the number of short engines available, that the head must be the weak point. I do know that they tend to break camshafts if you run them without oil. Eventually I found one from an LT 250 quad bike, which everything on the web says is exactly the same.  Of course it isn’t. The head casting is the same, the LT however has no tacho. This means that the tacho drive gear on the camshaft is missing, as is the hole in the rocker box for the flexi drive. What’s more, there is an extra hole in the rocker box for the decompressor, which the GN doesn’t have. The solution looked simple enough. I could use the GN rocker box,  and get hold of a cheap electronic tacho. The principle was fine and would have worked a treat if not for the BIG mistake. More on that later.  In the meantime a quick decoke using cheap paint stripper (plug up the holes, fill the ports with it and let it soak for an hour then flush it out. What it doesn’t remove it softens up well) lap the valves in and it was ready to receive a quick coat of Ger Ger Granvilles cylinder black and be bolted on.

It turned out to be a significant stroke of luck that I decided to paint the engine casings too. When I removed the primary drive casing I found this rather substantial bit of steel (which should have been under the camshaft) floating about in it. That would have made a really interesting noise.

I did see one warning on the web to carefully check the size of the barrel locating pins when fitting the Chinese 300cc barrel. Apparently they need to be at least 12.1mm or the barrel misaligns itself and gets sick. Mine were fine and assembly was very straightforward, except for the extreme shock at the price of Moly paste. The only other engine(ish) mod was the fitting of an engine sprocket that was one tooth larger.  I really hated running the little thing so close to the red line and I’d read that they could cope with higher gearing, which seemed entirely reasonable since first gear was only good for about the first five feet anyway.


The leathercloth on the was well and truly past it’s best, so I thought it would be nice to continue the colour scheme through the seat. I’ve been through three iterations of designing a custom seat for my ST1100 (I’ve got short legs and a high wide bike becomes an issue) so working in Vinyl is no problem. The first thing you need is the right sewing machine. You can just about get away with a large domestic one, but the little 12V one I had definitely couldn’t cope. A bit of research revealed that there are two types of machine that will do a good job on it. You can get a second hand heavy duty commercial machine for about £1500 to £2000 or any hand cranked Singer built between about 1890 and 1940. I got mine ( a 1930 Model 28 Singer) from the inevitable source for tenner. It ran straight from the box, though it took me a bit of time to figure out the right tensioning for everything. New needles for £5 are a wise investment. Using a heavy duty thread is a good idea if you can find it, but it doesn’t need to be the fancy twine that you’d use for leather. My current Pan seat was stitched with Tesco’s heavy duty thread, has done about thirty thousand miles and none of the stitching has given way.

Whilst the thread isn’t super critical, the structure of the seam most certainly is. It must be overstitched, like the outside leg seam on your jeans. That means that you stitch a normal seam, turn one side over and then stitch through the whole lot again (the picture makes it clear) that way the pull on the thread (and the holes) is very much in shear and doesn’t pull the seam apart.

overstitched seams

The exception in this case is where I put the black stripes on, but they just sit on top of a single piece of white vinyl, so there’s no structural element to the seams. Incidentally, I can’t feel the stripes, even on the odd occasion when I’ve nipped out in a pair of thin work trousers. The other trick for the seat was to make the pad on the bum stop bit hinged, which gives me access to the hump, where I keep a disc lock and a bit of towel. The flap is just held up with Velcro. The original foam was rather nasty and fell apart when I took the vinyl off, so I replaced that with a Yoga mat from a charity shop (£3). It’s quite firm but, because of the way the riding position works out the riders weight is spread along a fair bit of thigh and I’ve never had any numb bum problems on runs of an hour or less.

The off white vinyl had faired quite well through the winter and still scrubs up clean with a bit if effort, which it very seldom gets.


cafe racer build

the towell flap



Paint is going to come up a good few times, but for the first attempt…

Having got all those bits to the right sort of shape, I was faced with the problem of colour. No spray booth, no compressor and limited funds. I’d read a few different stories on the net of people successfully painting cars with Rust-Oleum. For the uninitiated, Rust-Oleum is pretty much an old fashioned, oil based, domestic gloss paint. Thinking back to days of carriage painting, it was all done with a very soft brush, lots and lots of coats and lots of rubbing down, starting with fairly fine abrasives and working down to pads lovingly made from the fur of baby squirrels dampened with virgins tears. SPOILER ALERT !!! I’m tempted not to point this out, both for the sake of the story and to catch out those who don’t read to the end of a set of instructions before following them, but that would be cruel. The paint job was not a success, but I now know why!

Step one was to rub the tank down. Always use a sanding block, even on curved surfaces, and always use wet and dry wet, or it will clog instantly. I think I rubbed the original paint down with 400 grit to provide a good key for the new paint. I then applied a couple of coats of Rust-Oleum primer, thinned with white spirit until it took four seconds to drip off a teaspoon. It takes remarkably little paint to cover a small fuel tank, so don’t mix a huge batch up. After a couple of coats of primer it was sanded down with 600 grit and then 800 grit to give a good, flat finish. I sanded through the primer several times and had to re-coat it, so unless the finish is really bad I’d recommend no more than a quick scuff over. The next coats settled into a routine for a while. Home from work, rub it down with 600 grit, dry it off, apply a coat of paint and then dinner. At first it seemed to go swimmingly, until I managed to sand through the finish. No problem, just a quick coat of primer again, sand it off then back to the top coat. Well, the quick coat of primer was fine, a quick coat of colour was fine, then the next coat might as well have been paint stripper. The whole area just bubbled up in a complete mess. Removing it left what can only be described as a hole, about an inch long and three quarters wide.

With hindsight, I reckon that I overcoated it too soon. There is a point where it is just dry enough to cover, but not so dry as to require sanding. Too early and the solvent dissolves the layers underneath.

cafe racer build

GN250 tank painted and flat

This was definitely too big a hole to fill with paint and body filler would be the wrong colour and take a million coats to hide. What I needed was a filler the same colour as the paint. This is where my memory went back to toy aeroplanes. There was a product available called micro-balloons and that’s just what it was. It looked like a fine powder, but was hollow, so you could mix it with glue or paint to make a really light filler. Of course a few grams wasn’t going to matter on the tank, but a fine finish in the right colour did. Micro balloons were somewhat expensive and I had no idea where to find them so an alternative had to be found. The answer came from the bathroom and was Johnsons baby powder. Mix it with paint and one instantly has a very fine filler in the exact colour that you want it. That idea was definitely a success.

So, from there I continued along the same lines, made the same mistake twice more and finally had a very flat, very smooth paint job. I took that down with 1200 grit and was really quite chuffed.

Three vinyl stripes gave me the colour scheme I was after and then it was time for the clear coat. I decided that clear coat would probably work best sprayed, so I elected to use a rattle can of the same brand.

Once more It went on fine, once more I sanded through it, once more the repair bubbled up like something out of a low budget sci-fi movie and left me with a huge repair job. Once more, I repaired it. Finally I got it down to 1200 grit, rubbing compound and a good coat of wax. I reckoned it looked pretty good.

The seat proved a bit simpler, except that there was a big hole in the hump to fill up and it was full of pinholes because it was originally vinyl covered. The former was a simple matter of ally mesh and filler and the latter, back to the baby powder. The side panels were just as simple, but I didn’t seem to be able to get quite the same standard of finish. Maybe I was just a bit lazy by then.

I also decided to paint the engine. I was originally going to polish the engine casings, but my early attempts were not encouraging. Essentially, the alloy used in Japanese bikes is a whole lot harder than that used on the old British ones (presumably to cater for thin walled vacuum cast casings) and develops over the years that strange mottled grey look. That surface oxide layer is very thick, so we’re talking about using something like 220 grit to move it. It simply wasn’t practical in the timescale, so a coat of black paint from a rattle can was the way to go.

Cafe racer build

GN250 Engine casings

What went wrong!

All that went fairly well until I refitted the tank, after some minor tweak, forgot to reconnect the fuel pipe and turned the tap. By the time I noticed, it had removed a large quantity of paint from both the engine and one side panel. It’s pretty well known that very little short of two pack paint will be fuel proof and, being pretty new, it didn’t stand a chance.

I made up some vinyl decals for the side panels. I got the inkjet printable vinyl from Ebay and can’t say that I was impressed. The black came out quite light and a bit blue tinged and then dissolved instantly when I blew clearcoat over it.

Once I had it all sorted and reasonable again it seemed to be fine for a while. After a couple of thousand winter miles, however, the paint on the tank began to crack and now looks like a dried out dessert landscape. The most likely reason is that I got carried away and put too much paint on. That causes a lack of flexibility in the paint and cracks soon appear. Another possibility could be that I didn’t leave enough flash time between coats, but I did follow the manufacturer’s instructions, so that seems less likely. Yet another possibility is that I went straight from painting to finishing. You’re supposed to wait about a month after painting before cutting and polishing the paint. I didn’t.

I guess the conclusion is that the Rust-Oleum route can work, but I didn’t get it right. I’m certainly convinced that a good finish can be produced with a brush and some patience. I’m now the proud owner of a compressor (Aldi – £90) and a “new” tank, but before that job goes forward I’ve got to sort out the dent in the tank and make an air cooler/drier for the compressor. More on that considerably later.

Changing the Look

Having got the heavy stuff out of the way it was time to start looking at the cosmetics and that leads us to the vexed question of the “clean triangle”.

I read on a web site somewhere the words “ an authentic café racer must have a clean triangle”. What????
Well, my first thought was that I can certainly imagine Mitzi doing her bit with a lady shave, but Ogri?? No way, though I can imagine the facial expression.
Then I realised that the chap was talking about the whole business of removing the airbox, buying a very expensive gel battery and hiding it under the seat, to leave the centre section of the frame empty.
So having cleared up what a clean triangle is, let’s look at the rest of the statement. What the Flamin’Ada is an “Authentic” café racer?
Well, to me it means one of two things, either we’re talking about a really nicely prepared Triton, Norvin, Bonneville etc. with a big Ally tank, high level pipes, clip ons, rearsets and a big chrome headlight or, and far more commonly (which to me means more authentic) a 1960’s British bike with clip-ons and rearsets, possibly a bum stop seat, if they were a bit posh and a few extra shiny bits if the owner could afford it. Pride and Clarke ally mudguards (at least at the front) would be de rigueur, as would a reverse cone mega from the same source. In truth, there never was an “authentic” café racer, it was what we could afford.
Now here’s the important point. They were generally Triumphs, BSAs, Nortons or Velocettes, I actually had a 500cc Ariel, but it was a bit of a dog. The crux of the argument being that they were all dry sump engines. That means that when they came out of the factory one side of the “triangle” was the oil tank, the battery usually lived in the middle and the other side was the tool box. Even when converted, at least the oil tank had to go somewhere and in the posher ones it was a rather nice custom aluminium job bang in the middle of said triangle. Most of them just retained the original oil tank and tool/battery box. So even if there was an “authentic” look, it didn’t ever have a clean triangle. The only bikes that came close to a truly clean triangle were genuine circuit racers and I suspect only then the early Japanese ones with wet sumps, though there were a few people putting the oil in the frame.
Having said all that, I have to confess here, that I really like that clean look myself and would have no compunction about building a bike that looked that way, but one simply cannot be prescriptive about these things and it can certainly lay no claim to authenticity. Personally I don’t feel the need to aim for an “authentic” look anyway. I’m building a 1995 Suzuki. Just do what works and looks nice to you.
So where was I? Oh yes, studying Suzi’s triangle. Now that has a better ring to it. What we have there is a huge airbox with a fairly small, snail shaped inlet, quite a large battery, the regulator/rectifier and a few sundry bits of electrical stuff.
To be honest I’m loath to significantly change the inlet arrangement until I start on engine mods. The snail shaped inlet must just be for noise reduction (with a large plenum downstream it can’t possibly be part of a tuned setup) and looks a bit restrictive, so that could come off without more than perhaps a change of main jet, leaving the original filter (well, a new one) in place. The reg/rec is actually in a pretty sensible place and I can’t afford a fancy battery just now, so the original side panels will be retained.
The tank is the wrong shape completely, but I’ve seen a couple of bikes where people have lifted the back end of it up a bit and it makes a huge difference. It does mean that some fuel will become inaccessible, which jars painfully against my general rule of function (preferably enhanced) always taking precedence over aesthetics, but I’ll have to swallow it and promise myself that the tank will be replaced later.
So getting the tank and seat to fit looks like the next move.
The tank is rubber mounted, so the simplest way to raise that is to use old technology, by which I mean a lump of wood under the rubber mount and longer bolts. There may be many people who regard using a bit of wood as a frightful bodge, but it has its place. As an engineering material, wood has two big drawbacks. One is that it only works in certain planes (i.e. it has precious little shear strength across the grain) and the other is that it isn’t dimensionally stable, meaning that it will expand when damp and contract when dry.
The flip side of this is that it has tremendous resilience in compression (I also build guitars and the tension on a set of steel strings is huge) it’s very easy to work, it’s light and it’s cheap. In short, if all you’re doing is jacking up the back of something non-critical, it works.
The seat presented three problems. It was too short to fit nicely at the back, it lacked the right mounting points and it left an awkward shaped gap above the side panels.

Re-Mangling the Seat

gn250 cafe racer build

GN250 seat and tank before rebuild

The rear mounts were a simple matter of riveting and glassing in a couple of brackets, which I left over length to provide mounts for the rear indicators as well. The front and sides were done by fabricating an aluminium “nose” which pop riveted into place, providing a front mounting plate (which would share the tank mounting bolts) and skirts to fill in the gap over the side panels. I’d describe the result as less than perfect, but once more it would suffice until I found/made a new fuel tank. Time was very much of the essence.

GN250 cafe racer build

GN250 café racer seat modded

Sorting the Brake Linkage

I reckoned that the gear linkage would be pretty straightforward (yes, you’d think I’d know better by now) so I started with the brake.

The problem with the brake linkage is that it’s a good old fashioned drum brake with a rod linkage. Incidentally, it works extremely well, even the MOT tester was surprised. Anyway, the issue with a rod gear linkage is one of matching the geometry of the linkage to that of the swinging arm, so that,
a) it doesn’t bang the brake on at full bump or droop  and,  
b) the brake still works at any suspension position.
I must confess to messing around with bits of welding rod, string and whatever else came to hand for some hours before finally coming to my senses and seeing the blindingly bleedin’ obvious. The original linkage worked perfectly and had all the right bits to connect to the brake light switch.  Once that penny had dropped it was the work of a few minutes to do a couple of quick sums to get the ratio of levers right. I wanted a rather less movement than on the original lever, but not so little as to reduce my ability to brake hard. On measuring the various bits and pieces at my disposal I opted for a pedal that would require one and a half times the pressure of the original (I’d be in a better position to push on it and I reckoned that the bike was designed for small eastern youths, not European adults) but with only a centimetre or less movement. 
Once the sums were out of the way I simply chopped off most of the original brake lever and drilled a hole in the remaining stub to accommodate a straight linkage from the pedal.  No problemo!

The gear linkage would have to wait until I had an engine in the frame, but that would be no problem at all, as the linkage and lever  from the Aprillia would fit, honest.

Starting the real work

The forks will need to be shorter. I see a good few bikes with the forks shortened by slackening off the pinch bolts and sliding the fork legs up in the yolks. Whilst that instinctively feels (and to me, looks) like a bit of a bodge, it has a lot to recommend it.
Firstly, it’s quick and easy. More to the point, it’s also quickly and easily reversible.
I decided therefore, to do just that as a first pass. That would give me time to assess the handling change, save me having to rebuild the front end in a hurry and I could always come back and shorten them properly afterwards. So it’s time to start on the rest.

This is the bit that I usually dislike intensely. Years of grime, rounded off nuts, bits that are seized up… and that’s just me! To be fair Little Suzi wasn’t too bad at all. There was very little in a particularly poor state.
Now, of all the new and clever tools that have been invented over the years there is one which stands out head and shoulders as the most useful aid to rebuilding and restoration.
The digital camera.
Gone are the days of a million tie on labels, bits of masking tape that either fall off the wiring or become illegible under grubby finger marks, copious notes on lost scraps of paper. Just take lots and lots of pictures of everything. Wiring is an obvious one, but the positions of washers and spacers, even introduce a measuring device for scale. I love the thing. So…


With the engine out and the front footpegs removed, the dog gets to see the proverbial rabbit.


I took the centre stand off early in the proceedings and gave it a clean and some new paint. Since the bike was going to be sat on it for most of the rest of the job it saved hassle.
This picture gives a good view of the original brake linkage… and reveals a fair amount of Suzuki standard ferrous oxide finish.


But after being sent off for bead blasting and powder coating she was looking much better. NOT A CHANCE!!!! This is just not that sort of restoration. What she got was a good rub down with wet and dry and coat of warm Hammerite smooth, cost roughly £10.

Footpegs and linkages
On the procurement front, that famous auction site yielded a pair of ex-Bantam clipons for £6 and a set of Aprilia 125 rear sets for £16. The rearsets were a bit bent and much repaired, having also lived on a Moto-team racer for some time, but they’d do the job.
Well, actually they wouldn’t.
The mounting points for the standard footpegs are set back a little from the level of the frame, see piccy. Fine if you’re going forwards with them, but it doesn’t work if you’re trying to go backwards. Needless to say, the mounting holes don’t line up anyway and whatever goes in there has to leave clearance for the end of the swinging arm spindle.

footpeg mounts

I did come up with a shape for a mounting plate that would do it, but to be really right it would have to be machined and whichever way it’s made there’s still too much cantilever (to clear the frame) for me to feel happy about it.
The answer to that problem, and another, came from the “Normous Newark Autojumble”. It was a fine event. We spent pretty much the whole day there and even my girlfriend enjoyed it, which is just as well, since she has entirely adapted her natural shoe and handbag spotting ability to be able to locate motorcycle parts at great distances. I cannot overstate what an invaluable and fulfilling feature in a lady this is.
Anyway, two bargains came out of it. One was a rather sad looking bum stop seat for £20 and the other was a battered old Clarke welding set for £25. I learned stick and gas welding as part of my apprenticeship (in electronics!) and then added MIG many years later when I was racing cars, so this looked like a good solution.
The right place for the footpegs was going to be somewhere within the triangle of frame that supports the rear pegs, so it should be a simple matter to weld a hefty fillet into the gap on which to mount the new ones.
Apart from one tube of one the triangles being slightly cranked, grinding off the rear footpeg mounts and sticking in a couple of bits of plate was as straightforward as it sounds.

P1020575 P1020576

P1020580 P1020581

Positioning the footpegs so that they were in the same place on each side of the bike took a couple of minutes thought, eventually achieved with two bits of string. One from the petrol tank mount and one from the rear seat mount. That gave me a triangle, the bottom of which was where the footpeg lives, simples!
The Aprilia pegs themselves were a bit knackered, so I chose to fit a £5 pair of new rear pegs instead and welded a 10mm nut to the back of the mounting hole, so that I could screw them in with lots of locktight and then fit a locknut behind them.
The levers were slightly more complicated, as they needed to stand out from the plate (to clear the linkages) and had plain shoulders on the mounting bolts which used to fit into a “bearing” area on the aluminium mounting.
In the end I had to drill out a couple of nuts to provide the “bearing” bit and then weld those and threaded nuts to the mounting plate. It was a little bit of a fart on grinding everything down to just the right height, but once done it all seemed appropriately secure and happy. The pictures tell the story better than I can, I used an old footpeg that I found in a box for fitting up.


Then of course, there are thelinkages…