K Jetronic questions

[Bernd Felsche]

Brian Michalk tapped away at the keyboard with:

> I thought my first problem was coming up with a way to manufacture
> some sort of mechanical actuator either via a lead screw, or cam.
> The problem is that I don't know what the cam profile needs to be.

> I think what I'll do now is instrument the distributor with a
> linear transducer for testing.  I'll run the engine up and down,
> and get the plunger displacements for RPM, MAF, and TPI.  I can
> take that data and have a cam ground just for the task.  That way
> I won't be wasting my time in front of the mill shooting in the
> dark with different profiles.

May I suggest?

You need an accurate fuel flow meter, a fine-pitch screw and a
dial gauge. Set up a small jig above the flow meter so that the
metering plate (and attached lever with plunger) will move as you
turn the screw. Switch on the fuel pump and measure the fuel flow
rate over the range of movement. No need to run the engine; just run
all the outlets through the fuel flow meter and back to the tank.

Keep in mind that you're dealing with volatile fuel!

If you don't have a fuel flow meter, then you can run the pump at a
constant setting for a fixed time and measure the amount of fuel
delivered in a glass measuring beaker. Half a dozen points should be
enough; you can interpolate quite safely.

Once you know the fuel flow rate at a given plunger displacement,
you can calculate the corresponding air mass.

You can set the fuel distributor's plunger entirely mechanically;
but the mechanisms involved aren't exactly simple to build, let
alone fit into an engine bay; an old distributor with centrifugal
and vaccuum advance can be "re-curved" to the corresponding airflow
and throttle profiles; you then have to change the upper part of the
distributor shaft (where the rotor sits) so that shaft will extend a
corresponding amount as the flyweights move with engine speed.

Top off the shaft with sliding bearing and a thrust bearing to take
the rotation out. You then have a plunger that moves an amount
according to engine speed. Rotate the base-plate of the distributor
with the plunger and use the "floating" vaccuum advance to rotate
the action plate according to load. The movement of the action plate
now corresponds to engine speed and load.

Run a bowden cable from the action plate to the lever of the fuel
distributor and you're done. Not pretty.

OR

If you don't mind putting a fairly-powerful solenoid actuator off to
the side, powered by some hefty switching transistors, then you can
set the correct plunger position according to your whim by
electronics. (I hope the road doesn't get too bumpy! A fast stepper-
or servo-motor would be more stable but slower to act.) A simple
analogue computer that takes into account engine speed and throttle
position would be the minimum. 

Anything more complex and you're better off going for a digital
computer. You can then use whatever air-metering takes your fancy.
If you use MAP, then fuelling will have to be "conservative" (on the
rich side) unless you're also checking EGO to close the loop.
If you use n-alpha, then EGO measurement is practically indispensible.

The more you measure, the less you have to guess.

> Once I have that profile, I should be able to fine tune with the
> pressure actuator.
> 
> What I'm still not clear on is whether I need a MAF, or just take
> everything from tables.

-- 
Bernd Felsche - Innovative Reckoning, Perth, Western Australia
----------------------------------------------------------------------------
To unsubscribe from diy_efi, send "unsubscribe diy_efi" (without the quotes)
in the body of a message (not the subject) to majordomo at lists.diy-efi.org