[Diy_efi] New project

[Dennis Noordsij]

Greetings!

I am new to this list, having browsed the archives many times but now doi=
ng=20
the formal introduction thing :)

Let's start! The engine is a Mercruiser 470, a 170HP 4 cylinder, 224 cid =
(3.7=20
litres), aluminium block (Merc design) with a Ford 460 head as well as so=
me=20
internal parts from Ford, obviously used in in a marine setting. I have o=
ne=20
myself on a 17' boat, a friend has one on his 20' version.

Let's say my friend is the mechanically inclined one, and I'm more useful=
 on=20
the electronics side. He's scrounged up from somewhere a TBI body, 2=20
injectors, built-in pump and regulator, built-in MAP and TPS sensor. I ha=
ve=20
no doubt it'll take him all but an hour to have the whole thing hooked up=
=20
replacing the carb, throttle cable adjusted, fuel return line installed, =
etc.=20

The engines run perfectly fine carbed, so it's completely a for-fun proje=
ct,=20
where a working EFI setup at carb quality would be a complete success. Th=
e=20
TBI is a bolt-on, so if everything goes wrong the carb can be reinstalled=
 on=20
the water if necessary.

His side of the deal ends with a connector with sensor wires and injectio=
n=20
trigger wires, which is where my part starts.

The sensors on the TBI body are set for a speed-density setup, which shou=
ld be=20
pretty straightforward.

Because of the wet exhaust I am not sure if we can use any kind of lamda=20
sensor, don't really know how far they should be from the exhaust valves =
(wet=20
part starts about say 30-40cm away from the valves) and if they like seaw=
ater=20
:-) Knock sensors are still not common in marine applications so I am sur=
e=20
there are reasons, and probably shouldn't be tried at this point.

An advantage for a marine application is that throttle response is nowher=
e=20
near as important as in a car, as you mostly stick to a certain throttle =
at=20
cruise.=20

I am wondering if there is an advantage in going with a synchronized setu=
p, or=20
just go asynchronous based on a small timing loop. Since it's a TBI setup=
 it=20
is only replacing the carb so there is no real direct relation with=20
individual cylinders.=20

I have no problem with the theory of speed-density sounding fine and dand=
y,=20
but of course I don't know how it really works in practice, and about any=
=20
hidden gotchas to be aware of. Obviously the volumetric efficiency is a=20
question mark, say it is 90%, but then is it a constant (easily tuned) or=
=20
depending on TPS, RPM, MAP, phase of the moon, etc? How much benefit woul=
d=20
there be in adding a MAF sensor? (expensive). Marine engines (carbed)=20
traditionally like to run a bit rich, typically 12:1, and because of the=20
constant load and lack of cat. convertor is it necessary to persue a=20
perfectly stoch. mixture?

Ideally, the VE factor would be constant through TPS/RPM ranges and tunab=
le=20
with a dial, say between 80 and 100%, and the ECU will persue the set mix=
ture=20
(probably also a dial) basing it on ambient pressure, VE setting and sens=
or=20
inputs, with of course special compensation for a cold engine and for sud=
den=20
throttle increases.

I would like the ECU to be based on a PIC, since I have most experience w=
ith=20
that as well as burners, compilers, etc. Even the 4 MHz part should be fa=
st=20
enough to calculate injector pulse times in real-time, A/D inputs are mor=
e=20
than fast enough, program and data memory should be plenty, and the circu=
it=20
simple enough (only few parts), can be flashed in-circuit. So spare ECU b=
oxes=20
can be carried as well, just in case :)

What kind of results would that give?
Anything obvious I am missing so far?

Cheers,
Dennis

_______________________________________________
Diy_efi mailing list
Diy_efi at diy-efi.org
http://www.diy-efi.org/mailman/listinfo/diy_efi