Turbo Engine Fuel Injection Maps

[cybernck 400 1 Cars]

ok, i'll start a bit theoretical but then we'll get on to the practical part.

 

if you take a look at the normal (n/a) engine's fuel injection map (for instance, Mi16 Motronic 1.3),

basically there's a map based on RPM, airflow (load) and corrected by lambda sensor (in closed loop).

on full throttle the AFM/load is at it's maximum, pulse width as well and the system doesn't run in closed

loop lambda correction mode. oh, and it's important to note that on Mi16 injectors fire in batches of two.

 

 

now imagine you add a turbo to this engine, with the AFM on the airfilter side, pumping out 0.4 bars.

turbo sucking the air through AFM would make the AFM max out as soon as there's any boost present,

so the engine would have max load & max injection pulse width nearly all the time when throttle is applied,

but this ofcourse wouldn't be enough, so an additional injector would need to be added. also, the system

should be in an open mode so the lambda correction shouldn't be messing it up.

 

the way i see it, the additional injector would have to be triggered each time one batch is fired (to have it

"mapped" against revs). but what bugs me is the pulse width. ofcourse, the most important thing is that

the 5th injector would have to be controlled by boost (a map sensor) so it doesn't fire when there's less

than 0 bars, at idle, on over run etc, and so that it gradually increases pulse width from 0 to 0.4 bars.

but what i'm not 100% sure on is would, from then on (when it has 0.4 bars), pulse width stay constant

or would it have to rise even more? my logic says constant.

 

 

now on to the plumbing - regardless of is it a single additional injector or four of them, should they

be plumbed in serie (output from additional fuel rail to the input of the main fuel rail) or in parallel

(a T-split between the input of the main rail and joined back together after it) and how many FPR's

would i have to use in either case (thinking just one in series and a couple of them in parallel)?

 

--=i=i=i=i=---=i=i=i=i=[FPR]---

 

or

 

.--=i=i=i=i=-[FPR]-.---

`--=i=i=i=i=-[FPR]-'

 

or maybe even just

 

.--=i=i=i=i=-[FPR]---

`--=i=

 

 

it's late so i may have overlooked some basic things but you'll get the point of my questions.

 

 

p.s. "mapping" of the additional injector would be based on using a VEMS WBO2+EGT sensor.

 

 

thanks in advance (to anyone who know what they're talking about ).

[jackherer 543]

oh, and it's important to note that on Mi16 injectors fire in batches of two.

 

I thought they were simultaneous?

[petert 554]

The 5th injector is all a bit hit and miss. Regardless of how it's fired, and for how long, how can you be sure you've got even fuel distributon to each cylinder?

 

Phase 1 Mi16 injectors all fire simultaneously. Whilst they are wired in pairs, to share the load over two drivers, they all fire together. How would the know ECU know which pair to fire next? There is no reference signal.

 

Phase 2 Mi16 injectors fire sequentially.

[TEKNOPUG 3]

Surely if you had anyway of properly mapping the engine, you wouldnt bother with an additional injector just fit 4 bigger injectors that can cope with the fuel demands on boost?

[jackherer 543]

Phase 1 Mi16 injectors all fire simultaneously. Whilst they are wired in pairs, to share the load over two drivers, they all fire together. How would the know ECU know which pair to fire next? There is no reference signal.

 

Phase 2 Mi16 injectors fire sequentially.

 

i.e. all 1.9 Mi16s are simultaneous and all 2.0s are sequential?

[cybernck 400 1 Cars]

jackherer> hmm that's a good point, i got so much into studying wiring and impulses than i'd forgotten it

doesn't have a cam phase sensor (mixed it up a bit with the T16 ).

 

petert> that's why i want to use 4 additional injectors, one in each branch of the manifold, so i could know

fuel is getting evenly distributed across cylinders (not taking into account injectors having different flow etc).

 

teknopug> ofcourse, but as a temporary solution, i have to go this route.

 

 

still, this doesn't answer my question on pulse width nor how to plumb the additional fuel rail in .

[dee205 3]

Would 4 bigger injectors work on standard managment? Can the standard ecu be remapped to suit? Surely you would get over fueling at certain rpm's.

 

 

Damien

[cybernck 400 1 Cars]

i can't remap the standard ECU, but apparently Chipwizards in the UK can, which is of no help to me .

 

if i got the injectors right for on-boost, they woud be overfueling off-boost & on idle.

[petert 554]

I can only see two choices to do it properly:

 

1. Eight injectors Mi16 injectors, two lots of four staged with an aftermarket ECU. The two fuel rails are in series, one fuel regulator.

 

2. Four large injectors, with a quality aftermarket ECU that will make them work properly at idle.

 

There's other work arounds but there's not much time between a lean mixture and a melted piston if you get it wrong.

[kyepan 291]

i second petert.

 

In the pulsar world the MAF (mass air flow meter) would max out above certain boost pressures (read air flow rates). So they switched over to another one that was capable of detecting the flow of more air z32MAF.

 

To cope with fuelling issues larger injectors were the order of the day, they would just flow more fuel for the same duration of opening (duty cycle) If your injector is at 100% duty its working pretty much like a carb needle with constant flow and you loose all the advantages of timed fuel injection.

 

Then either piggy back afr controllers lie to the ecu and get it to richen the mixture at certain rev ranges, or a propper after market ecu was used to map the injection, ignition etc.

Edited May 7, 2007 by kyepan

[cybernck 400 1 Cars]

i'm not talking about a permanent and perfect solution, just something to get it working.

 

if TT manages ok with their device and 5th injector, my device and 5th or another 4 injectors will do it too.

 

 

would two fuel rails in series with one FPR work better than a rails in parrallel with a FPR each and why?

 

 

thanks.

[petert 554]

As the two rails are very close to each other, there is negliable pressure drop in the line. If the fluid was stationary, there would be equal pressure everywhere, but as it is moving, there is a pressure drop from the friction of the walls of the pipe, bends, etc. The two fuel rails would have to be 10m apart however, to have a measurable pressure drop with that pressure and that size fuel hose.

 

Why don't you buy the F3 Haltech I have for sale? http://forum.205gtidrivers.com/index.php?showtopic=74894

 

It's the perfect way to control a 2nd set of injectors. Just trigger it off the neg. side of a coil. You won't need any sensors. In fact, you can even have it cheaper without sensors. The std. ECU would control the original injectors as normal. Then as the boost starts you start to turn on the 2nd set, with mapping to suit any situation up to 1.5 Bar of boost. Complete control!

Edited May 10, 2007 by petert

[cybernck 400 1 Cars]

well i've changed my mind a bit - i'm going straight for 4 larger injectors route,

controlled by Megasquirt. well, that's when i get it running fine with standard

injectors and no boost first.

 

i'll keep Motronic for ignition side of things, for the time being.