BSA B21 BUILD
John Lay started with a stack of spares from several motorcycles. Over several years he painstakingly created a single 1938-era 250 Beesa which looks like it’s never been apart in its life. Last month’s first test ride revealed that while it pulled well e
John Lay started with a stack of spares from several motorcycles. Over several years he painstakingly created a single 1938-era 250 Beesa which looks like it’s never been apart in its life. Last month’s first test ride revealed that while it pulled well enough the brakes still needed further attention…
Fitting a Bantam front wheel had certainly moved the project along and allowed the machine to be wheeled about. But clearly if the BSA was to be used on public roads it really needs something better. The correct wheel has the same diameter brake drum as the Bantam but its shoes are twice as wide, and this greater braking area should improve things. However, the project came without a front wheel and two years of searching failed to produce the correct one. I did discover a
hub and brake plate, less brake shoes. Before this wheel could be rebuilt the hub itself required some attention.
First: clean and inspect for damage, likewise the bearings, then repack with fresh grease. Second: find or modify a set of brake shoes to fit. An original set is probably out there somewhere, but those of a C11 are of the same width and diameter and much easier to find. The only problem was that they didn’t fit… not because they were the wrong size but because of the shape of the shoe casting. The original shoes were made of fabricated steel and the part of the shoe that bears on the brake plate pivot pin is quite narrow. The new shoes are an alloy casting and were too wide for the pin. This was easily overcome by grinding off the head of the pin – it’s riveted onto the brake plate – and turning a new, longer replacement, fixed in a similar manner to that used on the rear brake.
Next came the problems associated with rebuilding the wheel. Even a professional builder needs some basic information; the rim offset, spoke lacing pattern, length and gauge of the spokes. All of this is available if you have an old complete wheel to start with, but rather more difficult when all you have is a loose hub and rim. Not that I had the right rim. There were a number of old wheels that came with the bike when first acquired, none of which had anything to do with it.
One of these had a suitable rusty-looking rim fitted. It only ‘looked’ rusty – if rust had advanced to the point of affecting the rim’s integrity then it would only be fit for the bin. The inside of the rim (the part you don’t normally see) was carefully cleaned beforehand and checked prior to being dismantled. With the clean rim and hub to hand, I constructed a ‘taverner’s jig’. This device is described in Radco’s The Vintage Motorcyclist Workshop (well worth searching out a copy).
With the hub and rim set up in the jig and with the help of some rather fuzzy pictures found on the net, I was able to work out the pattern: cross two on the brake side, cross three opposite side. My first plan for the spokes had been to buy a set of blanks and cut and thread them as needed. I have a spoke rolling machine but I don’t have the correct 10-gauge rolling head (if you’ve got one please let me know). All attempts to find one came to nought. I was even told by one supplier that they are no longer made because ‘motorcycle wheels don’t have spokes any more’! In the end I had to make do with such spokes that I had to hand, even resorting to buying an old wheel at a jumble just for its spokes. The same caveat applies to wheel spokes as the rim; any that show the least sign of damage, stress, cracking, etc, should be thrown away.
The next task was to get the wheel running true, both radial and lateral run-out. This could be done in the bike’s forks, but is easier done away from the bike in a jig. I made a simple affair from wooden battening, held in a vice. If I planned to do many more wheels it might be worthwhile making something more elaborate. In the photos you’ll see the set square clamped to one leg of the jig which is to check for concentric alignment (up and down), while the dial gauge for runout (side to side). Once I was happy with the results, any spokes protruding through the nipples were ground off flush. Lastly, the inside of the rim was given a coat of anti-rust paint. Water can and does get between the tyre and rim, which then merrily rusts away, sight unseen.
A visit to the National Motorcycle Museum showed that I had indeed got the spoke lacing correct, but I wasn’t entirely happy with the result. Riding a machine with an incorrectly built front wheel isn’t something I (or anyone else) should be doing. The problem seemed to be with the first spoke crossing on the brake side.
In the normal way of things, spokes are lacedup alternatively inside/outside of their flange. This means that where the spokes cross they are separated by the flange thickness so the problem doesn’t arise. On my wheel the spokes on the brake side are all laced from the same side. This arranged the spokes at the first crossing very tight against each other. Consequently either one spoke is pushed out of its flange seating, or the spokes have to bend to accommodate each other. Neither was ideal, but what to do about it?
At a visit to the Classic Bike Show at the NEC I inspected a restored 1937 B20 (they use the same wheel as my B21). It showed the lacing pattern as my wheel and had the same problem, bent spokes at the crossing points! If that’s the way they were then maybe BSA weren’t worried about it. Alternatively, as this was a restored bike, perhaps the last wheel builder wasn’t worried either. However Central Wheels who were also at the show were, like me, not keen on the idea. Their solution? To make a set of 10 new spokes of the same length but with a slightly longer offset bend to give the required clearance.
When the new spokes arrived I replaced the inner ‘crossing’ ones. Note for any wouldbe wheelwrights: unless the holes in your hub flange are of the keyhole shape (in which case you can simply lift them in or out one at a time) you will find those laced-through plain holes are all but impossible to fit, as the other spokes get in the way. Don’t be tempted to bend them into some sort of halfmoon shape to get clearance. You need to remove one or two from one side or the other to make room. Then don’t forget to put them back and re-tension.
It’s a bit annoying to have to redo something all over again. Removing and replacing a number of spokes will almost certainly put the rim out of true, so that had to be done again as well. Maybe I am worrying over nothing, but I felt it was worth the effort just for peace of mind. That meant another session in the jig,
and the final truing would be done in the bike’s own front forks. It could be done in the jig if you know the correct offset… which I of course didn’t, not having an original wheel to measure. Just because a wheel runs true on its hub does not mean it is in the right place. Many wheels can and do have a fair amount of offset to allow room for brakes, drive sprocket or what have you. It’s important that the wheel runs centrally in this case within the forks.
The bike was put back upon the workbench with a temporary wooden stand under the footrests to raise the front wheel. Despite my best guess the new wheel was a little out of line. Happily I was able to move the rim over to run centrally, which meant a little more rim readjustment. Then I checked all the spokes for tension and looked for any spokes that had protruded through their nipples. Any of these were ground off flush and the anti-rust paint reapplied.
With a replacement rim tape, tube and tyre fitted it was time to fix the wheel into the forks. This generated a few more items which had to be fabricated, including a new torque arm and a ‘front hub distance piece’. This in effect is nothing more than a short spacer, counter-bored on one end to allow clearance for the bearing cone that protrudes through the brake plate. I also made a new set of wheel nuts from hexagon bar. As seems to be the way of old bikes, someone in the past had used an incorrect size spanner and all corners of the nuts were rounded off. Then I could fit a replacement brake cable and adjust it. At last, the rivet counters could stop telling me that I’ve got the wrong wheel fitted!
The weather wasn’t ideal for further test rides, so I tackled other tasks like the brake plate blanking plug. I made up one complete with a milled hexagon head before I saw an original, which took the form of a slightly domed head with a simple slot across its centre. This wouldn’t make much difference in use but I decided to make another one in the correct style.
Another task on the to-do list was to improve the headlamp: a 6V 30W bulb coupled with an aged reflector was just a bit on the dim side. Most cars have brighter sidelights than my headlight. Converting to 12V and fitting a more modern headlamp would be one way of improving things, but the solution proved to be much simpler, by fitting a LED lamp from Dynamo Conversions. The difference was amazing, the beam went from a dim yellow puddle to a brilliant white light. As an added bonus the current load is only about 1.5A as opposed to 5 to 6A for a conventional bulb, which should give the dynamo an easier time.
Unlike more modern bulbs which are pre-focused, the BSA’s needs to be done manually. It was common practice pre-war for the bulb holder to fixed by a small clip. This, once slackened off, allows the bulb holder to be moved backwards or forwards to get the best result.
While the BSA wasn’t being ridden and was left standing in one place, it became obvious that it was still losing oil (they all do that, sir). A sheet of newspaper underneath revealed that it was coming not from the engine but the gearbox, around the final drive sprocket in fact. Modern boxes have proper oil seals of course, but not it seems in 1937. This may be something I have to live with: keep a drip tray underneath and check the levels prior to each ride. Oil is probably finding its way past the main shaft bearing. BSA clearly thought this might happen so they fitted an ‘oil thrower’. These were once commonly used wherever a shaft passed through a casing. The thrower takes the form of a reverse scroll machined onto the shaft, or in this case onto the extension of the final drive sprocket. As the shaft rotates, the reverse scroll tends to throw the oil back whence it came. Great when the machine is running, but not much good when stationary. When these machines were used every day the small bit of leakage overnight may not have been much of a problem. There may well be a long term solution possibly involving machining the box and fitting a modern oil seal, but for the time being I plan just to ride the bike and top up the levels as required.
Now I just needed an opportunity to ride the bike again to see whether my efforts had improved things. Boxing Day dawned bright and mild and my local VMCC section was holding a road run; a good chance to give the bike a more complete test.