project

Peter Wales pjwales at magicnet.magicnet.net
Thu Feb 23 21:12:39 GMT 1995


The first project which needs doing concerns the needs of a well known tuner
of vehicles. He has fitted a turbo to a Miata and has sold a large number
over the years. Now a competitor has advertised higher power outputs and he
needs to compete, but being a man of integrity, he won't tell lies. He must
have the higher power. The limiting factor is the fuel available through the
stock injectors. Changing them is too expensive. The car already has a
rising rate regulator which increases fuel according to manifold pressure
times a factor, but it isn't enough.

So, the solution I am arriving at is that we need a pressure sensor to
detect an on boost situation. Nothing, but nothing must change off boost.
Once we have detected boost, we need to increase fuelling according to boost
pressure. This can be a simple multiplier of boost pressure volts and a
"mixture" pot. If the trailing edge of the injector pulse is used to trigger
the start of the enrichment pulse, the system will never run leaner than
stock. The pulse needs to start within a few hundred microseconds to stop
the injectors closing and reopening. The stock system runs a maximum duty
cycle of 65% at the moment so there is a potential 30%  increase in fuelling
available. The injectors are fired 2 at a time so we need 2 drivers.

The existing system needs to be interfered with as little as possible and
putting snubber diodes across the injectors to kill the back emf is not
allowed. This alters the way the injectors fire and upsets the fuelling.
Good input filtering and high voltage drivers are needed for the output, but
the final component specification is some time away yet.

We have a 12v source, terrible environment and it must have a watchdog
system which works.

There is a second aspect to the problem, and that is of boost retard. This
would be a very desireable feature if attainable, but as the system is not
reprogrammable, it will not be easy.

So, I need ideas, outlines and a price for doing the development. If you can
make these items, then I need a unit price in 10 off lots. There is however,
the potential of 3000 per year and that starts to get more interesting.

Talk to me guys

Peter




More information about the Diy_efi mailing list