When the time came to run my version of the Seagull things did not go as I had hoped.  After many adjustments as per the Westbury instructions I did eventually manage to get the engine to perform with a fairly docile performance, which I felt that could be improved upon.

During a brief interlude after I had managed to get the Seagull running successfully I decided to revisit a side valve engine that I had designed; having jettisoned the old carburettor I decided to make a mixing valve type carburettor as Westbury used on his Atom 1 engine. This certainly gave a big improvement over my original attempt. A new cylinder head for this engine was under construction along the Ricardo guidelines for side valve engines, this being the same as the cylinder heads used on the Seagull. in addition a new style of carburettor was being designed for this engine that more closely resembles the B&S version and using wisdom gleaned form the following account.

In the design of the Westbury carburettor for the Seal or Seagull there is a small hole

across the spray bar that becomes more effective as the throttle is closed, this hole is fixed and thereby I felt caused one major problem as all the subsequent recommended  alterations then have to revolve around the throttle barrel, either opening up the closing edge of the throttle nearest the engine to increase airflow, or if this is overdone then opening up the opposite side of the throttle which is open to the atmosphere to compensate for being over zealous.

While this procedure undoubtedly works it was I found a long and protracted method. I did wonder to myself was this perhaps why so many constructors decide to fit a commercial carburettor from a glow plug engine?

This tuning of the throttle barrel needs to be carried out methodically and a great deal of patience needs to be exercised. I felt it was time to learn about how these commercial glow plug carburettors did in fact work; more importantly the various design characteristics of each design.

As if by fate I was handed one such variety of carburettor to design a manifold to suit a Chenery V-twin. While it was a very clever design the overall length of the throttle and jet adjustment made it almost as long as the pair of cylinders on the Seagull and did little to encourage me to fit such a carburettor to my engine. Given that Westbury’s original design was so compact, very much in proportion with the rest of the engine, and it had been shown to work, I therefore thought there ought to be a design that could incorporate the features of the commercial carburettor with the compactness of Westbury’s design.

Once again as if by fate I was handed a design of a carburettor by Shelley Curtis, this follows what I would call the conventional Glow plug carburettor that I have seen fitted to the Jones 605, a 10cc 2-stroke engine designed by Colin Jones and serialised in Model Engineer. I believe the original design for Shelley’s carburettor is based upon a scaled up version of those used on the Vega V-Twin that was designed by Dave Parker.

These two designs gave me a starting point and a carburettor was constructed along these general lines, a photograph of this carburettor is shown. While the engine ran quite successfully compared to the initial Westbury design I still did not think that the tick over was as slow as it could be and there were still issues with regard to the flexibility, as this appeared to deteriorate as the level in the fuel tank dropped.

Perhaps to some I am being a little too picky, but after all if there is room for improvement then why not strive for that improvement, after all in achieving a better performance there would be more knowledge to be gleaned. I felt that the reason for the high tick over was down to the airspeed through the Venturi, if this dropped below a certain flow rate then the fuel drained back towards the tank, this was supported by the drop in performance with the dropping level fuel level in the tank. Fitting a non-return valve would stop the reverse flow and this was indeed the case but there was still the need to provide sufficient suction at lower revs.

With the Westbury design the small hole across the spray bar was to a certain extent providing the suction needed at lower revs that would not be available from the Venturi alone arrangement, if this refinement could be provided in the Curtis/Jones style of carburettor then I would have the best of both worlds. I did however think rightly or wrongly that in the original Westbury design this hole across the spray bar was too large, as no matter how well I adjusted the needle valve the engine still seemed to be running too rich at Idle. It therefore follows that if this orifice could be made adjustable then this would overcome my concerns about the hole being too large, as this would allow some fine adjustment to the air supply at low revs.

While I was mulling this problem over as to how to incorporate this feature into the design, I was re-reading Len Mason’s Model Four Cycle Gasoline Engines and the chapter on Carburation, in this chapter he talks about the “Submerged Jet” principle used in full size carburettors. This started me thinking about keeping the jet of my carburettor “submerged”. It also made me wonder about how the fuel was affected in a normal horizontal spray bar. I reasoned that the suction from the engine should completely fill the supply pipe to the spray bar right up to the end of the needle valve. However, if that was the case then why was the fuel so eager to run back to the tank at the slightest drop in engine revs?

I then thought: just suppose that the fuel was running along the bottom of the fuel line in the area of the needle jet at low engine revs, after all the fuel is affected by gravity just like everything else, this would mean that there would be some air admitted to a small extent at the top of the needle valve and hence allow the fuel to drain back to the tank.

If it were possible to engineer a small well to trap a small quantity of fuel such that it always kept the needle valve orifice completely filled then this would go some way to stop the fuel draining back at low engine revs. The small reservoir of fuel would also be beneficial when opening the throttle, as fuel would be readily available to meet the increased airflow through the carburettor.

Soon I realised that if I had my main needle valve vertical then the fuel reservoir could sit at the bottom of the needle valve, also by putting the fuel supply pipe into the jet body at an angle rather than horizontal it would also help to provide an extra reservoir. This would be acting not unlike the bend in a sink waste pipe and ensuring that any air in the system would be higher than the jet. Furthermore if provision was made for the adjustable air supply across the needle jet orifice to be arranged directly inline with the jet outlet then this would be a very short distance from the Venturi.

By manufacturing a new spray bar, jet assembly, along these lines I was able to marry this new design to the existing Curtis/Jones design which would allowed me to trial the new design without having to manufacture a whole new carburettor and it would also allow a direct back to back comparison with the previous setup. There was a marked improvement “straight out of the box”.

From a cold engine I had a nicely tuned engine within about 15 minutes of running, now I could open the throttle with not the slightest sign of hesitation. Armed with this newfound enthusiasm I made a completely new carburettor with a revised jet securing method, rather than having a thread on the jet assembly to screw into the body of the carburettor the jet assembly is now held with a single capscrew.

A silicone O-ring on the jet assembly where it fits into the main body looks after any sealing problems; a further silicone O-ring is fitted to the needle valve adjustment to ensure that no air enters by this route. By altering the position of the tapped hole holding the jet assembly into the carburettor main body, say through 90º, then it will soon be realised that this carburettor can now be used in a horizontal position.

There is no reason why this tapped hole that retains the jet assembly cannot be at any position between these two positions thereby giving unlimited positioning possibilities for the carburettor, the only position which would prove a problem as the carburettor is currently designed would be in Westbury’s original updraught position, as the manifold is in the way of the jet adjustment needle. As this position went out of favour many, many years ago in full size practice then I did not consider this to be too much of a draw back, but that is not to say it cannot be done.

For the final carburettor I decided to make the body out of aluminium and apart from the flange shape presents no real problems other than the tapering of the throttle bores which requires a small boring tool. It will be noticed that the manifold side of the throttle bore is smaller than the atmosphere side and care should be taken to get  these diameters as close to size as possible. I think with a precision instrument such as a carburettor these bores need to be machined rather than just drilled, if reamers are available then these could be used or a D-bit could be quickly made to suit.

Once these bores and the outer external turned portions were complete the rectangular section of the body can be milled to size, finally the flange can be milled or filed to shape and the two mounting holes drilled. The pocket for the throttle barrel was also bored using a small boring tool, the depth of the pocket wants to be no more than 0.05mm deeper than the length of the throttle body, ideally the two parts want to be the same size. That way the 4mm diameter Nylon stop that restricts the throttle barrel from coming adrift can also act as a friction device as well and will hold the throttle at a preset position, which is handy when starting.

The bore for the jet assembly is also machined at this setting, thus ensuring absolute concentricity and can be reamed to size if a reamer is available. This work can be carried out in the 4-Jawed chuck or, as I did, in the milling machine with the aid of my smaller boring head. The reason why I chose this option was that I wanted to rough out the pocket and machine the three connecting holes that break into the pocket first and thereby relieving a tedious deburring operation as the final boring operation to size the throttle barrel bore would save me the bother.

The throttle barrel was turned from EN1A Pb, it will be seen that the throttle barrel bore has a reduced section in the centre and is flared out to the outside diameter. A small D-bit was made on my cutter grinder to produce this feature and to ensure both sides of the throttle were identical the work being held in the dividing head and the D-bit fed in from both sides.

When the Venturi bore was completed attention was turned to the two steps that form the throttle stops. The one on what will become the inner face of the throttle barrel was done using a larger slot drill than the actual width of the step machining partly in the throttle barrel and partly in the parent bar stock, this step butts up against an M1.6 screw in the carburettor body to locate the fully open position. The outer step nearest the throttle lever is for the idle screw to abut and also to allow clearance for the barrel to rotate.

Having the luxury of a toolpost grinder the outside diameter of my throttle barrel was then ground to size, but even if I did not have this luxury I would still finish turn the outside diameter, as taper D-bits and reamers do have a tendency to distort the material around the top of the hole, more so when the hole is across a diameter as in this case, after which the barrel was parted off and faced off to overall length, in my case by using the edge of the grinding wheel.

To complete the barrel the hole for the spray bar to protrude through into the Venturi needs to be drilled, which will throw up a small burr in the Venturi, these can easily be removed by lightly twisting the D-bit around in the Venturi from both sides, but do not apply too much pressure otherwise the care taken to get the Venturi truly symmetrical will have been wasted.

Continued in part two

MEWS IS SPONSORED BY - CHRONOS -   PAULTHECAD.CO.UK - ECCENTRIC ENGINEERING - WARCO - BRISTOL EXHIBITION - HARROGATE EXHIBITION - MERIDIENNE EXHIBITIONS - CAMDEN MINIATURE STEAM - TRANSWAVE CONVERTERS - CHERRY’S MODEL ENGINES -

Model Engineering Website is edited by David Carpenter