From: John De Armond
Subject: Re: The Mileage Miracle
Date: Thu, 10 Jul 2008 03:24:00 -0400
On Wed, 09 Jul 2008 15:22:06 -0500, Elliot Richmond
>On Wed, 9 Jul 2008 15:08:42 -0500, "S. Barker"
>>The engines need for fuel hasn't changed. The EFI systems do the same thing
>>as the acc pump during pedal movement. Enrichment HAS to take place during
>>a throttle opening to prevent the lean mix backfire.
No it doesn't. Not with a proper dry manifold design.
>>They also enrich
>>during warmup just as a choke would do.
Technically there is some but it's orders-of-magnitude less than carbureted
systems. Starting enrichment is usually over within 5 seconds of ignition,
the time constant being temperature dependent. Speaking now from working on
and benefiting from the works of others in reverse-engineering GM PCMs. And
by having the source code for the firmware in the GM vehicles that I drive.
>Yes, but it is a carefully metered amount of fuel. The accelerator
>pump just squirted raw gasoline into the throttle bore. This fuel was
>incompletely vaporized so most of which went right out of the exhaust
>pipe. The computer on a modern EFI system enriches the mixture by
>just the amount needed. It is much more precise.
>This constant monitoring of the mixture is one of the things that
>makes the modern EFI system so much more fuel efficient than the old
Yep. Further, most EFI systems don't need any acceleration enrichment at all.
Even back as far as 1975, Datsun designed the engine in the 280Z not to need
any. Nada. None. The key was an extra hole drilled through the head to each
intake port to allow the injector to squirt directly on the intake valve. No
wetted intake surface at all.
The accelerator pump squirt (and the TBI's electronic replica) was necessary
only to wet the manifold walls. An engine doesn't inherently need any
With wetted intake walls, when the absolute pressure suddenly rises in the
manifold during throttle opening, what gas vapor there is condenses out on the
previously dry walls. Absent any additional fuel, the mixture is way too lean
by the time it reaches the cylinder and lean backfire or hesitation results.
On the opposite side, when the absolute pressure suddenly drops when the
throttle is closed, gasoline on the intake walls evaporates in the reduced
pressure environment. That causes rich backfire in worst cases and high CO/HC
emissions during over-run otherwise.
That most annoying of early emission controls, the throttle dashpot that
slowed the throttle closing and caused momentary engine racing when the
throttle was blipped, was the band-aid solution. Spreading the evaporation
event out over a longer period of time lessened the over-fueling and gave the
cat or AIR a fighting chance at controlling the emissions.
An extreme example of this wall-wetting phenomena is the
almost-never-successful suck-through carbureted turbocharger system. With
this architecture, not only did the intake manifold have to be wetted but also
all of the blower side of the turbo and the manifolding leading up to it. That
required a massive acceleration enrichment, more than any carburetor could
In my early hotrodding days before I understood fully what was going on, I
tried several suck-through conversions. All were more or less unsuccessful,
despite some rather clever approaches to get more fuel into the system than
even the largest accelerator pump, nozzles and cams could provide.
Blow-through solved that problem but that design had a bundle of its own
problems. Well-designed port EFI put an end to all those problems, even back
in the simple analog days.
Back in the Z's heyday, there was an erstwhile competitor (he considered me
competition but I didn't reciprocate) who doggedly stuck to wetted manifolding
in his turbo kits even after EFI became available. He removed his awful
suck-through Holly carburetor and replaced it with an equally awful
throttle-body injection system. The same problem persisted.
Even with the injectors turned full-on (not pulsed at all), the >200 lbs per
hour injectors could not supply enough fuel to wet all that surface area.
Horrible hesitation and occasional backfires on open throttle and billows of
black soot on closed throttle were trademarks of his system.
I once did a turbocharger conversion to a small japanese car, a Subaru I
think, though memory's fuzzy. It had the oddest EFI that I've ever seen.
There was a huge balloon-like plenum after the throttle valve but before the
individual runners. Into this plenum fired two huge injectors. Even more
interesting, they fired asynchronously to the engine speed. When the
accelerator was pressed, the injectors suddenly sounded like buzzing bees,
they fired so often. This was the early 90s so I can just imagine what they
had to do to make that setup meet emission standards.
I yanked all the OEM EFI out, had the runners machined for individual cylinder
port injection, cut away much of that plenum and added one of my homemade EFI
computers plus the turbo. The transformation was remarkable. The car was a
dodgy transportation appliance before with, shall we say, tepid throttle
response. After the conversion, throttle response was instantaneous even
before the turbo spun up. No acceleration enrichment at all was necessary.
Sometimes you stand there looking at the stock engine and you just have to
scratch your head and say "What tha hell were they THINKING?"