[Diy_efi] injecting 5 port head

[Daniel Nicoson]

Here's a stupid question, Does it actually have to be sequential?  We tend
to focus on sequential since most autos use that but the OEM went to
sequential primarily for emissions reasons at low RPM.

What if you thought more along batch-fire lines.  Put your injectors back up
in the intake runner a few inches (no machining on the head itself this way)
and then fire all the injectors every time an intake valve opens for 1/4 the
PW (on 4-cylinder).  Not claiming to have it all worked out but trying to
keep minds open.  Batch fire systems have worked great for a long time,
especially at other than idle RPM.

These engines probably won't know the difference between fuel injection or
carburetion as long as they get the fuel they need.

KISS!

Good luck, interesting projects!  LOVE to hear more about the Spitfire
engine project!

Dan Nicoson

-----Original Message-----
From: diy_efi-bounces at diy-efi.org [mailto:diy_efi-bounces at diy-efi.org]On
Behalf Of Carter Shore
Sent: Tuesday, December 26, 2006 8:54 PM
To: diy_efi at diy-efi.org
Subject: RE: [Diy_efi] injecting 5 port head

On 12/26/2005 Jean Belanger wrote:

> The problem with the way you want to do this is that
> you need injectors that
> flow enough fuel at < 25% duty cycle and not 50%. So
> your idle problem is
> even worse that you thought.

Yah, that's true. One way to extend the dynamic range
of the injectors at low RPM is to put fuel pressure
under ECU control; the Ford return-less setup uses a
fuel pressure sensor and PWM to the fuel pump, so this
is do-able.

> You don't really need to have 2 injectors per port.
> If you only inject when
> the intake valves are open then the fuel will be
> carried to the right
> cylinder. This will simplify packaging and injector
> selection, and help
> idle. The scheduling algorithm will then be required
> to inject for each
> injector in a sequence such as pulse, pulse, wait,
> wait.

I'm still up for 4 injectors, if only to preserve the
option of tweaking 'per cylinder' as needed. I'm not
subject to 'bean counter' constraints like the factory
engineers. And I also don't need to worry about CAFE
and other regulatory issues.

> I agree about the variable timing aspect. I'm
> planning to time the injection
> pulse with respect to the middle of the pulse and to
> have a timing table
> that varies timing with respect to RPM and load
> (either MAP or TPS).
>
> By the way, are you designing your own ECU or are
> you using an available
> one?

The control hardware will be custom, and I have not
settled on a platform yet. I'm considering a
distributed architecture, where semi-independent nodes
manage fuel and ignition on a 'per-cylinder' basis. A
central node sends out high level status/commands, and
serves as a collaboration node on common issues
through a shared communication network.

This gives me a lot of freedom in platform selection,
as the timing and control issues relate to just one
task, and autonomous operation is possible for
'limp-home' mode.

By defining the ECU as an architectural interface and
functional responsibilities, I can implement any node
with the appropriate hardware/software to do the job,
and change those out with newer/better ones as needed
without throwing away the whole thing and starting
over.

Carter Shore



__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around
http://mail.yahoo.com
_______________________________________________
Diy_efi mailing list
Diy_efi at diy-efi.org
Subscribe: http://lists.diy-efi.org/mailman/listinfo/diy_efi
Main WWW page:  http://www.diy-efi.org/diy_efi