Shared Port Inline engines with MPFI

[CTX-SLPR]

Howdy,

as usual a craigslist ad got me thinking about some hairbrained EFI project. Its a 71 Custom-10 Chevy with 250 I6 power and a 3spd on the column in my favorite colour, dark green. Looking at the ports on the 250 and how to EFI it made me really start to wonder how do you setup MPFI on an engine where 2 valves share the same port? Looking at the factory EFI I6s, it looks like they all have individual ports (ie the 300 Ford and 242 Jeep) so that's no help. What do ya'll think?

[Matt Cramer]

The key is preventing charge stealing. Either use some very tricky injector sequencing to prevent this, or put the injectors about 6" up the runner so the fuel and air are well mixed by the time the reach the head.

[CTX-SLPR]

That would not really be a problem the way I see it since the easy idea I had was to basically build an adapter spacer to put a Ford 300 EFI intake on the 250. The other idea I had was just to use the Ford 300 fuel rail and build an intake out of exhaust tubing and shuffle the injector order so you have the batch fire companion pairs on a common port.

thoughts?

[dieselgeek]

we're in the middle of a real life situation similar to the dilemna you propose. Inline 8 buick with a 6-71 Weiand supercharger sitting atop a custom intake manifold.

I voted for port injectors doubled up, alternating sequence, maybe even firing more than 2-squirts per engine cycle... mounted relatively upstream in the long intake tubes.

The builder got concerned, so he went the "homemade TBI" route and ended up with distribution problems. However a carb works fine on top of the 6-71. So they are switching to an EFI hat and will be back to see how it runs in 2 weeks.

It's right up your alley: this old ass engine has a 36-tooth crankshaft trigger, pickup sensor, and 8 LS1 coilpacks!

[CTX-SLPR]

M/S controlled or is it someone else's box?

[dieselgeek]

M/S controlled or is it someone else's box?


it's a megasquirt (ems-pro). So it can't do "sequential" but it can do "multiple squirts per cycle"

I think the multi-squirt is a better solution than sequential, 1-squirt-per-cycle - in the case of shared ports.

[milner351]

you guys are reading my mind.

I've been thinking about something like this for the 292 after I finish phase 1, which will be carbed.

[CTX-SLPR]

Whoo Hooo!!!! my mental builds are being useful!!!

What is the batchfire setup on an I6? Would that be the best way to cluster the injectors and get them to fire twice or what?

[dieselgeek]

Whoo Hooo!!!! my mental builds are being useful!!!

What is the batchfire setup on an I6? Would that be the best way to cluster the injectors and get them to fire twice or what?



without a cam sensor, the injector sequence starts randomly with sensor inputs. So there's no real timing you can guarantee unless you are running cam sensor or, with 2+ squirts per engine cycle, a missing tooth wheel.

[CTX-SLPR]

Whoo Hooo!!!! my mental builds are being useful!!!

What is the batchfire setup on an I6? Would that be the best way to cluster the injectors and get them to fire twice or what?



without a cam sensor, the injector sequence starts randomly with sensor inputs. So there's no real timing you can guarantee unless you are running cam sensor or, with 2+ squirts per engine cycle, a missing tooth wheel.
Exactly, correct me if I'm wrong but I would think that a well designed batchfire MPFI would be much akin to a wasted spark ignition system. System has no idea if its TDC compression or TDC exhaust, it just fires the injectors something like 15deg before local crank 0 each time on companion cylinders.

My idea is that you'd put this "companion pair" on the common runner so both are firing a with lead for one of the opening ports but that might cause some funky sequencing though on an I6 I would think it wouldn't be too hard.

[nickleone]

Go to:
http://www.starchak.ca/efi/siamese.htm

The man has been working on this for a while.
Look at the main page there is an article about his attempt at injecting a shared port engine.

Nick

[dieselgeek]

Go to:
http://www.starchak.ca/efi/siamese.htm

The man has been working on this for a while.
Look at the main page there is an article about his attempt at injecting a shared port engine.

Nick



i've been reading that link for nearly a decade and can't ever find if the guy did anything more than theorize about his setup, or actually try running an engine? there's a pretty huge flaw in his thinking, assuming that any EFI system would only fire once per 2 crankshaft revolutions. Put in 2 squirts per cycle and a lot of his concern goes right out the window...