Carb to efi fuel conversion help needed

[Testament]

On 2/13/2018 at 13:06, Roman said:

Yes but its a lot nicer to tune when your fuel pressure differential is stable (As in, your FPR is manifold pressure referenced) 

As it keeps your injector characterization more stable.

In which case you obviously need your regulator near your manifold.

Admittedly I'm fussy but I'd only run returnless if I had 3d deadtime data from the injector manufacturer. (And the ECU capability to use this)

you are being fussy

the flow of the injectors is still going to be repeatable, its just as boost rises/vacuum decays the amount of injector opening time for the same additional volume of fuel wont be perfectly linear but neither is is engines need for fuel necessarily - hence the need to tune rather than just going about this whole business 100% mathematically. plenty of modern cars run like this too.

but the main point of it is that it is alot tidier in some respects, a little cheaper (less fitttings and line needed) and if the car never had a return line you can use the original hardline which makes things super easy.

so nah not "ideal" but definitely works just fine

[Yowzer]

Returnless systems tend to run much higher pressure than your standard 90s spec 3bar. Some little turbo thing I looked at awhile back was running a constant 6bar fuel rail. I can't remember what the Mazda 2.3T runs but I think it was 4 or 5.

[Roman]

On 2/17/2018 at 15:44, Testament said:

you are being fussy

the flow of the injectors is still going to be repeatable, its just as boost rises/vacuum decays the amount of injector opening time for the same additional volume of fuel wont be perfectly linear but neither is is engines need for fuel necessarily - hence the need to tune rather than just going about this whole business 100% mathematically. plenty of modern cars run like this too.

but the main point of it is that it is alot tidier in some respects, a little cheaper (less fitttings and line needed) and if the car never had a return line you can use the original hardline which makes things super easy.

so nah not "ideal" but definitely works just fine

Yes being fussy.
But more specifically what I am meaning is that when differential pressure changes, so does your injector deadtime by a much bigger swing than if you keep differential fuel pressure stable. 
So if you tune your fuel map at a particular IAT and ECT, when you come back later and these have changed your compensations wont work as expected.
Especially if you had a VE based fuel model where IAT and ECT are part of the fuel calculation at the start, rather than a fudge factor afterwards.

Now of course pays to keep in mind that ECUs like early Links didnt even have deadtime information at all, and a car will still run perfectly fine for the better part either way. 

But if you try to get your cold starts and your hot idle and so on setup nicely, having this work as good as you can to begin with saves chasing your tail around...
Again though each person has a different idea/expectation of what they want from an aftermarket ECU. 

[ajg193]

On 2/13/2018 at 12:19, BlownCorona said:

I also dont recommend putting a t piece in the supply to the pump, you'll recycle the fuel heating it more and more each time until one day it'll vapor lock and be a cunt to figure out

Had vapour lock for the first time today. Was just sitting in the garage after driving and spent like 15 minutes fiddling with AE tuning and it suddenly lost pressure and went lean / pump free spinning. Fuel lines were hot. Two minutes later it started fine though once it primed itself.

Gonna lower the tank in a few days when it is a bit more empty and pull out the pickup and modify it to have two pipes.

I can't imagine this happening under normal driving/traffic conditions though as it would only really happen when you aren't moving and are using very little fuel.

[BlownCorona]

called it. 

but yeah, cheers for adding weight to my statement, it does happen! ive spent longer than 15 mins in a burger king drive thru. i remember because i spent the whole time watching my temp gauge rise. pretty sure i nearly cooked the old motor that day

[Testament]

On 2/18/2018 at 18:07, Roman said:

Yes being fussy.
But more specifically what I am meaning is that when differential pressure changes, so does your injector deadtime by a much bigger swing than if you keep differential fuel pressure stable. 
So if you tune your fuel map at a particular IAT and ECT, when you come back later and these have changed your compensations wont work as expected.
Especially if you had a VE based fuel model where IAT and ECT are part of the fuel calculation at the start, rather than a fudge factor afterwards.

Now of course pays to keep in mind that ECUs like early Links didnt even have deadtime information at all, and a car will still run perfectly fine for the better part either way. 

But if you try to get your cold starts and your hot idle and so on setup nicely, having this work as good as you can to begin with saves chasing your tail around...
Again though each person has a different idea/expectation of what they want from an aftermarket ECU. 

so you are saying there is a material difference in the injector opening time with differing fuel pressure DP in the range of 4 to 6 bar?

while we are talking opening times in the order of 1ms i suppose, but I'm going to need some convincing on that.

it doesn’t seem to me that it should effect the operation of the solenoid that much.  if that made that much difference then so would engine bay/injector body temperature

[Roman]

With the data from an ID1000 and a 10psi pressure drop, the variation from just the deadtime changing on account of not referencing vaccum means:

At 1ms, you have 5.8% error in fuel delivery estimation 
at 2ms, 2.8% error and so on (Error halves as MS doubles) 

With a 440cc injector my engine idles at around 1.2ms effective pulsewidth so I would be idling at just over 0.5ms effective PW with an ID1000

So nearly 12% error in the calculation for fuel delivered at idle.

Not to mention that at at 50psi pressure its a 1085cc injector and at 40psi it's a 980cc injector. (I would assume these measurements are absolute fuel pressure, squirting into atmospheric pressure)

So that's a 10% difference in CC rating coming into the mix as well, which ECU may or may not be smart enough to compensate for. As opposed to CC rating staying (more) stable.

Again though if you dont care about chasing your tail around to get your idle really nice, or whatever else... Its splitting hairs for sure.
Even more so if your injectors are a lot smaller than 1000cc.

The data sheet for ID1000s is here if you want to take a look / check my dubious maths skills

http://injectordynamics.com/injectors/id1000/

Also, as an interesting aside. 

The factory LS1 ECUs that are reflashable, for a long time people struggled with reflashing the ECUs because when you reflash it.
It would go momentarily super rich, and then over the period of about 10-15 minutes all of the values would return to normal.
Someone found out that the ECU models injector tip temperature, and when you reflash it sets injector tip temperature back to 0 Kelvin or something like that hahaha. 
And then comes back to modelling a temperature that's a bit more normal over the next little while. 
So most people just turn off the injector tip temperature compensation. 
But apparently GM knows something about injector temperature that the rest of us dont... Perhaps for the factory injectors the orifice diameter increases or decreases in relation to temperature or something like that.