Four stroke
carburetors have accelerator pumps that produce a squirt only while the throttle
is being added, not at a constant throttle or trailing throttle. The purpose is
to make up the fuel lost to low vacuum until the revs build and recreate the
vacuum. Like any other carburetor circuit (e.g., pilot circuit, needle circuit,
or main jet circuit), there may be more fuel added or less fuel added than is
needed. In addition, the AP squirt may be the right amount but not last long
enough (to build revs) or may be too long (being rich after revs build).
The amount and
duration of the AP squirt is dependent on a few things:
When you twist the throttle, the AP linkage pushes a rod down against the AP
diaphragm which pushes gas through a passage going to the carburetor venture.
The squirt of gas compensates for the big gulp of air the carburetor sucked in.
The sudden drop in vacuum causes less fuel to be sucked through the normal jets.
To help in fine tuning this squirt, there are adjustments. The leak jet is like
a bleed hole in the AP squirt passage. A slow twist of the throttle will push a
small amount of gas through the leak jet (which is the path of least resistance)
and almost none will make it through the whole passage into the venturi. If you
twist the throttle quicker, it tries to force more fuel through the passage
which can't all go through the leak jet, so the rest flows through the whole
passage and squirts into the venturi. The timing screw (on the external linkage)
lets you time when the squirt starts and to some extent the duration, i.e.
earlier squirt equals slightly longer duration. The length of the rivet on the
diaphragm will also control the duration. A longer rivet will cause the
diaphragm to bottom out sooner and limit the travel of the diaphragm, therefore
reducing the duration of the squirt. A shorter rivet will allow longer travel
and therefore a longer squirt duration.

The figure is a schematic sketch of the accelerator pump (AP) circuit. When the
bike is running with the AP fuel reservoir full, and you whack the throttle, the
actuator rod (green) gets depressed. This pushes the AP diaphragm (blue) down
forcing fuel out the bottom passage. As you can see, the passage allows the gas
to go in one of two directions. One passage leads directly to the AP nozzle in
the carburetor venturi. The other passage leads to the Leak Jet (red) and back
into the carburetor bowl.
A couple of things become immediately apparent. One is that by varying the size
of the leak jet, we can vary the amount of fuel coming back to the bowl and,
therefore the amount that goes to the AP nozzle. A larger leak jet allows more
fuel back to the bowl, and less fuel to the AP nozzle. A smaller leak jet allows
less fuel back to the bowl, and more to the AP nozzle. A such, a larger leak jet
leads to shorter squirt duration/volume and a smaller leak jet leads longer
squirt duration/volume.
Also, the rivet on the bottom of the AP diaphragm limits its travel. A longer
rivet would equate to shorter squirt duration. Length of the AP diaphragm is
measured, measured from the top of the AP diaphragm to the bottom of the rivet.
Stock diaphragm length is 7.5 mm.
It is also important to set AP timing screw (on the left side of the carburetor,
under the cover). If you set it too tightly, it partially depresses the actuator
rod. This holds the AP diaphragm down, and reduces the volume of fuel in the AP
fuel reservoir leading to shorter squirt duration/volume. If the AP timing screw
is set too loose, the AP squirt lags the movement of the carburetor slide, also
potentially producing a bog.
When the AP squirt is not correct in volume and duration, the bike will stumble
and potentially stall with a quick blip of the throttle off of idle. Searching
the list, and this site, I found that several other people had the same problem.
Also, there appears to be no one fix that works on all bikes.
Especially on the 03 and later models (on which the BK mod is difficult or
impossible), it is likely that the AP can be tuned pretty close using different
leak jets, AP diaphragms, and adjusting the timing linkage (see below).
Approach to
Curing the Bog
The big issue is that the same fix does not work across the board. There is no
"quick" fix. This is because like all jetting, the AP function is dependant on
multiple factors such as (but to a lesser degree than other jetting circuits):
So the only option you have is to meticulously work through the process until you get satisfactory results. Here are several things you can check and/or adjust that will affect AP function and the degree of off-idle bog you experience:
Now, I clearly understand that all of this feels daunting to those who are uncomfortable with jetting and for those who have never delved into their accelerator pump. But it is really not rocket science once you start digging into it. The goal is to achieve an AP squirt of 0.5-1.0 seconds that just misses the slide. Also keep in mind, that for most it is quite easy to cure 90% of the bog with:
Everything else listed above is for those who need that extra 10%.