I don't know how much PI gasoline will be vapour before it has passed the intake valves. Molar mass of air is 29g and gasoline about 105g. So if the gasoline was vapour at the PI tip (worst case and unlikely?) then at 12:1 AFR by mass you would expect 2.3% FAR by volume.
Direct Injection Delete
- Thread starter 135iTX
- Start date
Can anyone suggesting that DI be completely removed for ultimate power kindly explain how removing a fuel source directly from the combustion chamber is better than not having it there at all? I don't really understand the logic, since there are plenty of people running mixed fuel delivery systems (tech seems well understood and effective). Feel free to pick one specific use case and make the explanation of why it's better. I'm not arguing that it is not; I don't understand why the case for "delete" is specifically made. Thanks!
Filippo
Filippo
You got this notion from a different conversation, which was when someone stated that PI held back the platform. It did not, DI has.
This hasn't been true all these years up until just now, in fact, still not yet AFAIK. It may have happened, but I have not seen one syvecs N54 make a 850+ pull. They are still trying to get them installed and tuned. Trying to get the DI and PI to play nicely together at the highest power levels is not a beautiful Symphony at the moment, it would be better to remove one of the injection methods from that equation to produce maximum power, particularly before Syvecs. From the time we needed more fuel up until today, running PI only would have simply had more advantages than disadvantages in the quest to make maximum power. Due to the complexities with blending them, if you had to choose ONE injection method to make the most power, which would you choose, given the current state of our platform. I rest my case.
More boost.
It would be great if we had some DI break through on our platform, but DI had its chance. Nobody ever came out with anything, nobody invested the money in creating a custom pump or special injectors. So really, it doesn't matter how many generations of diesels have been using them if there is no innovation, save the shotgun. The time for the expensive investments like a redesigned HPFP has passed and I don't see anything occurring in the near future, particularly with the direction this market's audiences has headed. The time to make a custom HPFP was back when they were all failing, that would have been big time money to come up with a fix and an upgrade simultaneously and yet nobody did. So if it didn't happen at a time when EVERYONE needed a new pump, that investment sure isn't going to come now when a few purists want to keep their DI for their big power projects, all seven of you. Everyone else will just go Syvecs, etc. There were big names on the platform, AMS is probably the best example here and with companies like that you never know what type of products might come out. I remember speaking to Eric on e90post about various topics and they left because of our fuel problem after they talked up their Forced Performance turbo upgrade. So for now we have companies like VTT/ADE/Motiv etc who will be carrying us the rest of the way, and all I see is port injection being used on the big end. Our injectors may be capable of flowing a lot on a spec sheet, I was there for those 'white paper' posts on the other forum so no need to remind me. I was a top ten reputation there at one point, but I'm not sure they were actually tested that way or capable of withstanding the pressure or flow rates required to see that volume of fuel delivery in a production environment.
So ideally, yes DI would be better to retain. Its better for those who aren't race cars, its better when you have proper control of the fuel systems with something like a Syvecs figured out. But for making big power up until the Syvecs/Shotgun era, DI has been just a pain in the ass. It flows all of what, 450whp on E85? The other 500whp has been coming from port fuel!!!!
I'm not sure I buy this line of thinking. How does one commercially compete with BMW installing new HPFP's for free? There is no market there.
I bought a single barrel. Those buying single and double barrels are probably more likely indicator of some of the HPFP market. I would replace my 7-year-old HPFP with a better one.
I completely agree that an HPFP improvement does not exist because until now, the cost/risk to bring to market is seen as greater than the reward (market size / price point). I don't know that it is a foregone conclusion that the time was at some point in the past when BMW handed out free candy. More so I wonder if the sweet spot would have been in the last few years and currently, when all these cars became cheap and the mod market went beyond the seven people with $50k new cars under warranty willing to pour in a bunch more money into drivetrain mods.
On your other comments (about PI/DI mixed at high HP) thanks. That clarified my prior question. I assumed that all the cars running around with PI/DI were proof that it was well sorted out. Did not realize the higher HP complexity of that marriage.
Filippo
Last edited:
Cost is always a factor, but the difference was motivation. Why would you want another free HPFP that is going to fail? That is like getting a free re-tune by a shop that blew up your last motor, no thanks.
Until Syvecs, that combination has been run by two systems that are unaware of each other, can not speak to each other and only function because of the stock DME's capability to adjust fuel trims very well on the fly. Ask someone what happens when you enter open loop at 7000 RPMs while spraying over 50% of your fuel through an AIC controller, you'll get various answers that don't reflect an ideal, controlled or favorable scenario. I'm no expert in this area, I simply know that when you have two separate fuel systems spraying fuel that can't talk to one another, that is bad and it doesn't take much to realize that.
There is debate over whether or not the DME cuts spark or if it continues to fire on those lean cylinders, and that is why there are about three companies right now working on solving these problems for those who don't go standalone. Some of what I say is still open for debate or may be incorrect but my concerns are in the ball park.
For the guys running PI and DI I wouldn’t chance it with AIC. There has to be a way to run the PI from the DME by hijacking the signal. From there its just like rescaling for different injectors. Syvecs for the win or replace the DI with PI or replace the whole lot from a different make/model
Problem is enough outputs and injection timing for transients. The PI needs information in advance of DI and a fuel film model.
I've looked around but haven't found anything yet, though with more turbo N52 setups coming someone is bound to cross that bridge.
Another fun plus to the N52 head is it has a stock metal valve cover...
Now a turbo N52 is just dumb. A low boost centric supercharger makes much more sense for that engine.
That metal cover won't work for our engine by the way.
Another reason I said it's pointless because you have to remove the head plug the DI holes since you can't leave the DI injectors in since they become glow plugs if you do. If you factor in the cost of doing that plus a standalone ecu that can run the canbus system on a bmw you're almost at the cost of a syvecs. No need to reinvent the wheel just buy the syvecs and called it a day. We are not the only platform pushing high hp with both di and pi. Lambos and R8s have been doing it for years.
Removing it won't add power, but replacing the DI N54 head with a port injection on N52 head 'could' theorhetically be better for crazy high hp applications. As stated before, the head flow on the N54 is very bad, and but no one knows the N52 head flow to know if it really is better.
Yes I know but the N52 head will fit on the N54 engine. So it would be a bonus if the swap was ever made. Again, if the N52 head flow sucks then it's absolutely not worth it. But at first glance it would look like it may significantly outflow the N54 since there is no DI injector taking away valuable real estate from the valves.
But the coolant passage and bore size is different. Not really a viable solution. That horse has been beaten to death already.
Because BMW improved the HPFPs. And lots of folks went and got them done. most every BMW platform has suffered some kind of issue that was significant, M42 profile, 3 liter v8, HPFPs on N54/55/M57/N57, S54 bottom and top end, and on and on. It doesn't seem to have stopped anyone from continuing to buy the cars. In fact sales continue to increase at a rapid rate. I'm pretty certain every member of this forum still owns one of these cars with a BMW HPFP. That's how it has and will continue to be.
Filippo
I'm not saying people would stop buying the cars, I'm just saying that its too late to make an upgraded pump from scratch and I think its reasonable to assume that some of those people who had failed pumps would opt to install an upgraded unit instead and sell their new warranty pump they get from BMW. So going full circle, it would have been a better time to invest the big bucks to create an improved HPFP while the pumps were failing, because there would have been more buyers and more motivation. The N54 isn't even sold anymore, that is clearly not a better time to launch an expensive R&D project than when the cars were flowing out of the factory.
My points were largely theoretical, however race teams make billet engines, so removing the DI in CAD shouldn't be a problem, though entirely impractical for anyone else. The concept of running PI only is not one for someone who intends to drive their car normally. I made this assumption anyway.
Last edited:
I don't think it's been beaten to death...? I think there's only 3 posts ever on this, or at least that I've came accross.
Go to 12:50 on the video below, it looks like they are able to align. Some minor modifications to the gasket and/or ports, but nothing too radical (at least not in comparison to the vanos/injector controls haha). Best solution would be if we could retrofit the DI injector into the N52 head (again, assuming that it out flows the N54 head by a lot).
i don't see why it would be difficult to tap the injector ports, thread an aluminum insert in the hole, bench it, weld it, bench it. It would be very minor work to remove the DI ports from the head.
3 posts you are able to find. there were plenty of discussion on this in the late 2000s on this...
If you think retrofitting the DI on a head that was never designed for it, let alone getting the valvetronic to work... you are seriously underestimating things...
There are only 3 viable solutions that have happened so far :
1. half-assed piggyback (AIC, JB4, etc) with PI and DI
2. use synvecs or motec to control everything instead of DME plus piggyback
3. get rid of DI all together
everything else is speculating on parts and setup that don't exist yet... and IMHO, unless you are the one putting money and effort developing such things, discussing these sort of things purely base on some speculation is rather pointless. people have been doing that on this platform for more than a decade now.
Why the attitude? The thread is titled Direct Injection Delete, I'm pointing out that the only gain from this would be if you could retrofit a non direct injection head.
I am sorry I came across as condescending in my previous post. This sort of discussion just kind of irrates me since I have been seeing people talking like it can be done for years. Yet no one has the balls to get the actual project rolling and prove the naysayers wrong.
Instead of debating it on the internet, why not actually go ahead and try it out on your with real parts if you think you have the right idea.
