The pump is a 35mm wide cog with a sloping top. Looking at it from the side and rotating it to see maximum slope, the high side is ~6.8mm and the low side is ~4.4mm. This cog sits in an oil sump, which has an input oil feed and an output oil path.
There is a pin which changes position with throttle, travelling along the radius of the cog (from between 6mm to 14mm from the centre – judging by the marks made on a used one). The cog is sprung to push upwards against the pin so that, as it rotates, the “up down” travel changes with the slope. When the pin is at 14mm from the centre (at full throttle) the travel is ~2mm and when it is 6mm from the centre (idle) its travel is 0.4mm.
The oil pump, “pumps” once every 83 revs of the crank (i.e. the cog rotates once)
Let’s look at an engine that idles at 1,000 and red lines at 7,000. Note that without actual measurement of fuel flow and actual pump flow (as opposed to calculated flow), this is all theory.
Regardless of throttle position changing pump travel – the oil delivered at full revs is 7 times that delivered at idle (assuming constant pump efficiency). If fuel delivery was directly proportional to revs, then this may indicate a fairly constant ratio of fuel to oil (i.e. more fuel = more oil).
However, if we now take into account the pump stoke ratio – at full throttle the pump is pushing more than 5 times the oil through (2mm / 0.4mm). So if we assume that 7,000 revs as 2% ration, then idle is 0.4%!
This sort of agrees with the ratios stated on Dan's Motorcycle Two Stroke "AutoLube" Oil Pumps page, where “At an idle the pump mixes the oil with the gas from a 120 to 1 ratio on up to 20 to 1 ratio at 8000 or more RPM.”, which is roughly difference of 6 times.
Here is a excel spreadsheet used for some calculations: oil_pump_calculations.xls
More Information?
To make make this information more accurate it would be good to:
- measure actual oil throughput with the pin at idle and full throttle (on a lathe maybe)
- measure actual fuel flow under different throttle positions/revs and conditions using a Fuel Flow Meter
- with oil/fuel known for different rpm's, we can calculate ratio for the whole range
- Is the pump cog case hardened? Can the slope be adjusted to pump more oil for tuned engines?