Gosh, it has been a year and I've progressed a bit. Thought I should come back and share my progress as it might help a few fellow tinkerers out.
Firstly, I have been experimenting more with fuel levels and feel that Keith Franck's advice (25mm below the milled top of the jet block) is the right fuel level. I've had good progression and pretty well eliminated the flat spot I had.
I've also acquired a second pair of carbs (40DCOE18s) that appear to have OE Lotus jet specs. They are a bit tatty on the outside but very clean inside, where it matters.
- I bought a spare pair of carbs
My attention at the moment is on balance and I'll explain why.
Even with the fuel level correct, at very small throttle openings, I have been getting spitting back (lean misfire) on one barrel. This corresponds with the cylinder with the highest airflow when setting the balance via the throttle lever. On the original carbs, the imbalance is between barrels 3 and 4, on the 'new' set, it is between barrels 1 and 2 (2 being lean).
But how to correct it?
With the carbs off the car, emptied of fuel and thoroughly dried, I removed the auxiliary venturis and main venturis, unscrewed the idle adjust screw until the throttle plate was 'as closed as it can be' and looked down the barrel with a strong light behind the throttle butterfly.
- Looking down the carbs, this is where I could see light around the butterfly unevenly against the bore
I could see that there was more light showing on one barrel than the other and unsurprisingly, it was barrel no2 serving cylinder no 2. My next move was to loosen the two staked brass screws that secure the butterfly to the throttle spindle and adjust both of them (mainly rotating them slightly (they have oversize holes hidden by the throttle spindle that can accomodate quite a lot of movement). After some trial and error, I came to the conclusion that they would never be a perfect seal but there was the same amount of light getting through with the butterflies closed.
How can I prove that I've made an improvement without fitting it and trying? This is where I think I've done something new. Remember I said I had thoroughly dried the carbs of fuel, inside and out?
I've drawn up and 3D printed an adapter that bolts to each carb in turn and connects the pair to a garage vacuum cleaner.
- OpenSCAD design for the magic adapter. I have some learning to do as I couldn't see how to make the flange a perfect match for the carb, but the fit was perfect first time.
- Magic adapter printed... got to love this machine!
The logic is that both barrels on the carb under test are now seeing the same vacuum input and I can now take meaningful measurements of airflow (using the sychrometer) with them off the car, saving about two hours of disassembly and reassembly each time.
Photos can do the talking...
- Magic adapter in use
- Here's the synchrometer I use - I just alternate it between the barrels
These are the 40DCOE18s, remember that the imbalanced pair was 1 and 2, which are pretty well perfectly overlaid in the attached plot, barrels 3 and 4 didn't show any perceptible imbalance on the car, so I have probably got them acceptably close. I'll still adjust 3 and 4 now to try to get them identical, in the pursuit of excellence!
- Pretty darn close now, with barrels 1 and 2 pretty much perfectly overlaid. 3 and 4 seemed not to have an imbalance on the car, so perhaps I am close enough. The measurement range is 0 to 3 turns on the throttle idle adjustment screw, with zero being the point at which the lever starts to move