BMG DESMO KIT

BMG Motor-Cycles Ltd of 352 High Rd, Ilford, Essex, developed a desmodromic valve kit for Velocette 'M' series engines (Viper, Venom, MSS), and applied for a patent (#939,895) on May 17, 1962. The drawing reproduced here is taken from their patent application form, and shows the parts which they manufactured as a direct bolt-on accessory, with no major machining required.
The kit cost £38, or £46.10 if they fitted it to your machine.

The BMG setup is ingenious in its simplicity and clarity of purpose, and is fairly well made, although some have called it a bit agricultural. Regardless, the kit worked as advertised; this was borne out in a road test by Bruce Main-Smith in 1963 (The Motor Cycle), who said 'My summary is that this is the tool for the rev-happy rider'. His road test isn't exactly a ringing endorsement, although he is able to rev the test bike (a 350cc Viper with Butler dolphin fairing - see period photo) to 7500 rpm in the intermediate gears, although he can 'only' pull 6200rpm in top, as the bike is slightly over-geared - this equated to 93.9mph on the Viper, which is going very well. BMG had tested their own Viper and found 6800rpm/102mph was possible.

Having said all that, Main-Smith seems strained to find a reason to spend money on the kit, as the power gains were difficult to gauge (the fairing would have added to the top speed considerably without the desmo kit). The road test was taken in poor, windy weather, and as far as I know, no other road test was published in the day, or since. Supposedly, a Thruxton equipped with the BMG kit took a speed run at Bonneville, and managed a record top speed, but I haven't found firm documentation on the results. All other indications point to an overall moot gain in power with the kit, although I'd love to hear otherwise, or from someone who actually fitted the kit to their Velocette.

The lower photographs are courtesy Dennis Quinlan, showing a display cutaway engine built by Australian Velo technical guru, Norm Trigg. The BMG kit has been fitted and the installation can be clearly seen, including the cam shaped, paired cam followers, and the positive valve stem connection. You can see, slightly in shadow, the lower rocker arm which lifts the valve off the seat. The pushrod connection at the other end of the rocker arm is fairly substantial, and is where the valve clearance is adjusted.
As there is no valve spring, the cam isn't fighting 100lbs or so of spring pressure to open the valve, which theoretically would give the cams a much easier life, and have the effect of vastly lightening the whole valve train, allowing the engine to rev more freely.

These BMG kits come available now and then, and I've always thought, 'what if?'

DESMO VELO #1

While Ducati has a lock on the Desmodromic valve system, they weren't the only motorcycles to use double-acting valve management. Many Velocettes have been subjected to springless valve actuation - first in the mid-1950's, when Sid Willis in Australia made an experimental sohc desmo KTT, then Harry Hinton (also of Australia) in the late 1950's, finally in the mid-1960's, when BMG made a commercially available push/pullrod desmo kit for the M-series Velos (Venom and Thruxton - the BMG kit deserves a post).
Interestingly, in both cases, no power advantage was found, and the conversions were short-lived.

Sid Willis was a champion 250cc rider in Australia, using several ex-works Velo engines and parts. In 1953, he joined the Continental Circus (as the European racing season was called in the 1950's) with friend Tony MacAlpine (pictured with his Vincent racer - second photo), taking his 250cc dohc Velos, built around pre-war Velo ex-works cylinder heads. Even with this advanced engine spec, his bikes still used rigid frames - very light, but difficult to handle over the cobblestones of the Mettet circuit in Belgium. To overcome handling issues while retaining the light weight of his racer, he asked frame builder Doug St.Julian Beasley to bring his chassis up to date. Beasley built the now famous all-welded lightweight double cradle frame which housed several 250cc Velos after Willis' (see top photo), which is similar the Norton Featherbed, being an all-welded double cradle, but isn't a merely a copy, being much smaller in scale, lighter, and with slightly different frame geometry.
With this new frame and his dohc engine, Willis came 5th in the 1953 Isle of Man Lightweight TT, the second privateer home behind Arthur Wheeler's Moto Guzzi, and a field of very rapid Italian lightweights including works Guzzis, dohc NSUs, and DKW two strokes.

At some point in the mid-50's, Willis considered the theoretical advantages of desmodromic valve operation for racing (mainly, unlimited revs without fear of valves dropping), and built up his own desmo head. The lower photo shows his thought process made metal; it shows his cobbled-up model, built to see how/if the project would work. Satisfied that it COULD be done, he set about fabricating a new cambox, which he fit atop a KSS mk2 alloy cylinder head, which he had machined flat. Needing a very tough metal for the rocker arms, he cut up steel wrenches and silver soldered them into shape! Bush engineering at its very finest.
The finished engine would rev to 10,000rpm, and started easily enough, but it had no gas sealing when push starting and thus no compression, as a small amount of valve clearance needed to be maintained when cold, so that the valves would have room when the cylinder head warmed up. When the motor was turned over quickly, gas pressure sealed the valves. Willis tried the completed motorcycle at his local racing circuit (Mt. Druitt), but found that it was no faster than his ex-works dohc engine. Plus, the cobbled-up rockers proved troublesome (he used silver soldered wrench handles for the finished product as well), with breakages at the joints, and as he was only interested in quicker lap times, he decided to give up on the project. The bike languished in his shed for a year, with one further attempt to cure its ills. When this failed, Sid melted down the cambox for another project! No photos exist of the completed desmo machine, although Dennis Quinlan, when interviewing Willis for an article, was able to rescue the model from his scrap metal bin!

This post is based on a story written by Dennis Quinlan which can be found here.