[Nniftyfour]

So I wanted to start a thread to discuss what advantages we could have running just DI. Only because I do believe if we all want it it could possibly happen.

I'm also going to have to supplement my fuel system this spring and at the very least want to learn a bit about the safety, so dual purpose.

Safety/cut off- So as I understand it there's two control schemes for PI right now. AIC6 by itself and AIC6 controlled by JB4. Using AIC6 if the DME cuts fuel I'd assume the only reference's AIC6 has is boost reading and spark timing. So it would have to wait until it recognizes boost falling then react to that and cut PI. I'm not sure how quickly it'll pick up on the loss of spark, maybe thats faster and it uses that? Wouldn't we end up with several combustion events where the motor is running on just PI.. So super lean? JB4 is great and maybe Terry can answer but I doubt throwing that in the mix is gonna make it react that much faster.

Power- Would we make significantly more power using just DI? Obviously the atomization would be a lot better. I know that's good for efficiency and a better burn but at high load levels would there be a difference in power output? It would be more resilient to knock at least right? Letting us feel comfortable with tossing some more timing perhaps? Or even just helping to eliminate timing drops.

Head Flow- Our intake tract was obviously designed for just air, not a fuel air mix. Therefore our runners are quite small. Not an issue I'll run into, more for you big single guys but pretty soon won't you guys run out of head flow. N54 is at what ~860whp max currently. At what point do these runners just run out of flow? I'm sure there's math that can be done but I'd imagine looking at them we have to be close at this point. I'm sure you could port the head to expand the working range some but at the loss of low end drivability and possibly throwing the intake cam optimization out of wack.. Which can't be fixed because aftermarket cams that play nice with VANOS don't seem to exist. Just seems like the most complicated path to be going down.

What we know about our fuel system.

Injectors- Our injectors are huge to allow the motor to run on just low pressure fuel (72 psi) in case of a HPFP failure. Theoretically our injectors should be good for 1000whp easily. We also know that raising injection pressure is NOT recommended, a Continental engineer made it clear he wouldn't raise the pressure due to leaking concerns. Injection window is the likely cure here and injectors shouldn't be an issue.

LPFP- there's obviously plenty of solutions for this. Including Vargas's custom 1200hp system. Non issue.

HPFP- This is where our problems lye. Our HPFP are mechanically driven and run out of breath fairly quickly. They also historically aren't that reliable. It's important to realize how they're not reliable. Our pumps consist of a mechanical bellow pump directly driven by the Engine. This is at set ratio compared to the motor and output rises with engine RPM. Fuel flow is therefore controlled by an electrical valve. The fuel control valve (FCV), THIS is what fails on our HPFP NOT the mechanical element.

Vargas turbo has come up with a possible solution. The solution is to drive the pump at a faster ratio compared to the Engine, this is done by moving the driving force from an internal chain in the engine to on the accessory belt. This shouldn't play into the HPFP failures we've seen as once again the problem has been with the electrical control valve which is uneffected by this change. Pump weaf has not been an issue stock and Tony says hes tested it for over a year on the same pump. I know not everyone loves Vargas but this approach IMO might be the best one.

This is good for 550whp on E85 or less. (Ethanol seems a requirement at this point) The other solution would be leaving the OEM pump alone and adding a second pump in Vargas shotgun fashion. I believe this would be our ultimate solution. Theoretically should be good for ~900whp. Enough that anyone wanting more then that could use port injection. (Remember at 900whp our intake would only be carrying the air, not the air plus about 350whp of fuel, allowing margin for PI) these real OG guys looking for that power could deal with the associated sketchyness that we deal with now, thats life at the top.

Now the FCV system is complicated so running anything other then OEM pumps doesn't look likely for us. Even running two OEM pumps seems to trip up the DME. Jyamona thinks he can work on that though.. And that's something I'd love to see, so what do you guys think?

Personally the possible lag in control of the PI erks me and I dont really like it. https://youtu.be/cDGlN6mluGA

[ThaCuda]

Has anyone measured flow or spray angle of our injector tips?

[jyamona]

Would someone like the calculate the theoretical max injection time at various RPM for me? Not too sure on the math here. In milliseconds for say 2k, 3k, 4k, 5k, 6k 7k, 8k rpm.

Then I'll compare to the tables, and see how much more room we have in them.

As for PI lag, I recommend not running your boost reference off the chargepipe, as this is not a measure of true boost, sees spikes (throttle closure, etc). Boost reference to an AIC controller should 100% be off of the intake manifold. If you have the AIC that takes in a voltage, buy yourself an AEM Map sensor, plump it to the intake manifold, and then voltage output to the AIC. Then the AIC will see boost cut at least as soon as it happens. There is still the question of when the DME cuts fuel to a cyl or two due to misfire, etc, that the PI will still be injecting it. I have some ideas for this.

As for dual HPFP control, I feel this can be sorted eventually. Time and $$$ though.

[Stucks]

This is a very interesting thread. I would eventually like to see more going on with the hpfp side of things, while finding the limit of the stock injectors. Never really liked the idea of port fuel injection and all the scenarios that could possibly happen (failed injector, lean, ect ect). thankfully there havent been many of those yet. Many other platforms, just overdrive the hpfp (think vw gti's and stuff). Others just swap in bigger injectors (for example, id850s on the LSA cars) and keep the factory fuel system components in the original place. as jake said, time and money will eventually develop this area of the platform more.

[AbacusRacingN54]

Originally Posted by ThaCuda

Has anyone measured flow or spray angle of our injector tips?

Big gains in diesel world from just tips...always wondered why noone has any upgrades for n54...

Anyone ever source a 3rd party hpfp? I kno cps (again disel world but just reference) are belt driven but make over 30k psi...I would assume we need in the 3 to 5k range..

[Nniftyfour]

Originally Posted by AbacusRacingN54

Big gains in diesel world from just tips...always wondered why noone has any upgrades for n54...

Anyone ever source a 3rd party hpfp? I kno cps (again disel world but just reference) are belt driven but make over 30k psi...I would assume we need in the 3 to 5k range..

As soon as we ditch the OEM pump with the fancy FCV were probably also talking standalone ECU. The DME would bug out.

[Terry @ BMS]

Made this post in your BB thread where there seems to be a lot of "action".

Originally Posted by Terry@BMS

Looks like a long thread. Too long for me to read all the posts. But if you have any questions about the JB4 PI integration I'm happy to discuss that in more detail.

The JB4 considers a wide range of criteria in determining the injector MS including engine speed, gas pedal input, boost, low fuel pressure (directly effects injection pressure), intake temperature, throttle position (don't want to inject when the throttle is mostly closed), etc.

Right now we don't read the factory injector PW but I think we probably should add that as a criteria.

[Unklejoe]

Perhaps my standalone meth controller project could be leveraged to improve how port injection is implemented on these cars.

My controller reads the current fuel mass delivery rate through the OBD port rather than trying to derive it based on a bunch of other parameters. I was able to sample the value at over 500Hz with my microcontroller, but I could probably get that even higher, I just never tried. That should be quick enough to capture any transients. What I'm not sure about is whether the "current fuel mass delivery rate" value that I'm reading is requested or actual. This didn't matter for my application, since they're usually equal.

If it's requested, then the value would be based off of the ECU's load/RPM calculations.

If it's actual, then the value would be based off of the current rail pressure, injector pulse width, and RPM.

What we really need to know is how the ECU behaves when the fuel system is maxed out. Do the fuel trims go positive? Or to they remain normal since the ECU knows that it's not actually delivering the amount of fuel that it should be, which therefore means that for the current air flow/fuel flow, the AFR is actually valid.

I have a couple ideas about how this could work.

I have the ability to read many many more parameters, so anything is possible.

For reference, the source code is at:

https://github.com/jakemoroni/BMW-CA...eth-Controller

Feel free to copy it and roll your own implementation! That's what open source is all about.

[prelude2perfect]

So the FCV is not the restriction on the HPFP? Only the fact that it's chain driven by rpm? Sending higher voltage to the FCV doesn't do anything right?

I ask because I've been having HPFP issues and replacing it didn't cure them. I've traced my problem to the power wire going to the FCV. Sometimes the power cuts out completely. Replacing the fuse fixed it once. Symptoms come and go such as long cranks. Even with a brand new HPFP I struggle with even an E40 blend at only 19 psi.

[Steve @ BMS/Fuel-It!]

No, playing with the PWM to the FCV doesn't do anything.

The HPFP is limited by the volume of fuel that the individual chambers (3 of them) of the HPFP can accommodate. Those chambers have compressible bellows that have hard stops, consequently more pressure, flow, etc. has no affect on the capacity of the HPFP. As long as your LPFP pressures remain above 50psi or so and can completely compress those bellows, that's it, you are tapped out.

Bigger turbos = more power = more fuel used = HPFP crashes sooner. It's not a problem with your HPFP, it's the volume of fuel you require. 19psi on RB's is not that same as 19psi on stockers.

[Unklejoe]

I'm guessing that's a consequence of it being a positive displacement pump? Makes sense, which is why the only solution is to spin it faster or run dual pumps.

[Patrick4588]

Are u saying u can ADD the shotgun hpfp instead of moving the factory one?

[Tony@VargasTurboTech]

Originally Posted by Patrick4588

Are u saying u can ADD the shotgun hpfp instead of moving the factory one?

The shotgun does move the factory pump to the belt. Right now we are only shipping single barrel units, as the limitations of the DI system are not solved enough to be able to make use of the double barrel set up. The Single barrel set up is a perfect combo with Port Injection, as it gets rid of the HPFP crash down low, and allows you to inject less fuel with the PI.

[Motoxracer1551]

Originally Posted by VargasTurboTech

The shotgun does move the factory pump to the belt. Right now we are only shipping single barrel units, as the limitations of the DI system are not solved enough to be able to make use of the double barrel set up. The Single barrel set up is a perfect combo with Port Injection, as it gets rid of the HPFP crash down low, and allows you to inject less fuel with the PI.

Thats a shame....

I really want to see upgrading the DI System the solution to extra fueling on High-HP cars, (would love to see up to 1000whp on upgrading the DI system alone).

[Cortez08]

A reliable twin setup making 600whp on DI only would open it up for many of us who DD these cars. I was hoping the anticipated injector table break through would make this a reality but apparently not. Time will tell I guess.

In the meantime the recent 9 second pass by a Pure powered F10 M5 is making me move ever so closer to that platform.