123quattro

Yes, but you still can't built boost with no load on the engine. Anti-lag gets you full boost in neutral. Really it's only useful for doing stuff like this...

http://youtu.be/tdXhvTwUa8U?t=27s

ptuomov

A small fuel system update. Before I went heads down in front of a computer for two years at work, I was corresponding with a firm called Fuelab about my pump control. They make the pump I use, Fuelab Prodigy. Turns out that during that time they introduced some interesting solutions to their product line which we're trying to test right now.

The background for my problems is that for high powered turbo cars the fuel system can be a real headache (and that I am a noob). On the one hand, when cruising or idling, the engine consumes very little fuel and the fuel rail pressure should be something reasonably low such as 45 psi minus the manifold vacuum. At peak fuel demand you have to produce a lot of fuel and pump has to pump it against a pressure that is the base pressure 45 psi plus the boost pressure. Batch injection pulsing isn't making the problem any easier. How both these requirements are usually met in cars of this vintage is sizing the pump by peak demand and then returning a large amount of fuel via the return line most of the time.

The main problem with this arrangement comes from the fuel heating as it's churned from the tank to the pump back to the tank; thru filters, dampers, rails, lines, etc. As the fuel heats, it's density goes down and the car runs leaner. So now you have to tune for too rich when fuel is cold. The A/C based fuel cooler int he 928 helps a bit, but it's still the case that churning fuel at the rate of 100+ gallons per hour heats it up. On the suction side at the tank pickup, hotter fuel cavitates easier, leading to all sorts of additional problems. The pickup diameter and other restriction requirements upstream of the pump become more stringent as the fuel temperature goes up.

I've tried to solve this problem with many different ways. I've got -10AN fuel pressure line, and there aren't a lot of awkward turns in it before the fuel rails. I bought a electronic fuel pump controller, but never got it to work because the PWM frequency of the controller was not compatible with my pump. I put together an Arduino controller with code referenced to the MAF signal but that required all sorts of drivers and stuff to get the voltages to match and that also ended up not being of the quality that I'd want to have on my car. In other words, I tried all sorts of things that a noob like myself would try and (un)success was at typical neophyte level.

Fast forward to the turn of the year, I emailed the Fuelab main guy about where they were and turned out that they've introduced the product that I was going to pre-production test for them before me falling off the face of the earth. It's an electronic fuel pressure regulator that combines some of the good features of both the traditional mechanical regulator system and an electronic flat head system. At a very short latency, the EFPR is just a regular mechanical regulator. On top of that, though, it measures the amount of fuel that is returned to the tank thru the return line. This measurement is used to produce a control PWM signal for the Fuelab Prodigy pump that I have to control the pump speed. So the EFPR automatically turns up and down the pump as the fuel demand fluctuates. In addition, there's another 0-5V control signal that can be logged that indicates how much capacity there is left in the fuel system -- are you running out of pump?

So _in theory_ I have a solution. In a week or two well know how the solution works in practice.

Here are some low quality cell phone photos.

The regulator. It's a single line in, single line out, and the bottom "line" is really the cable for power and signal wires.



Here's the bracket that youknowwho made for me. The assembly now has a mechanical fuel pressure gauge as well. The logic is that since the EFPR is adjustable as well, we need a gauge.



Here's the EFPR installed. The gauge is visible and the attachment is very sturdy. Because the gauge is mechanical and easily visible, it should be a reliable and convenient debugging tool whent he need arises.





Here are the lines that hook up to the rear of the fuel rails. The rail still speaks German metric but then it gets translated to 'Merican AN:



In terms of the plumbing, the batch fire system with 60 lbs injectors and occasionally very high fuel pressure requires a lot of damping in the fuel system to keep the resonances from screwing up the mixture ratios at certain rpms. The car now has four dampers and a regulator which also has some damping qualities. There are two 944 dampers in the front, upstream of the rail. There are now two stock 928 dampers in the back, downstream of the rail. The output from the two dampers in the back goes thru a y-fitting to the EFPR. The EFPR return port is plumbed to the stock A/C based fuel cooler, and the return line is stock.

Here's one quick shot of the dampers and lines at the return side:



_If_ this system ends up working, it's one template for high-powered forced induction cars, especially those running E85.

UPDATE: The car is running with the mechanical regulator working but the pump controller not yet hooked up. The good news is that at least at lower frequencies the fuel pressure is very steady.

John Speake

George Suennen and Bill Ball also battled with these problems with George's ORR supercharged 5.0 litre 928.... he could be interested in this development.

ptuomov

I believe they have solved their problems. I believe (without any knowledge, having never seen George's car, so just going based on hearsay and his posts on this forum) that George had a number of problems in the system initially but that he has since then knocked them out one by one. Which is a good thing, as the fuel system is also a safety issue. Among other things, George has an Aeromotive pump with Aeromotive controller, which seems to work. (Aeromotive pump is not an option for me in the long run as it can't be guaranteed to pump enough fuel against high pressures.)

Here are the easiest ways to screw up your fuel system. I have personally made almost all of these mistakes:
- Not thinking about the fuel system as a _system_ that is part of a bigger _system_.
- Puting in a bigger pump with the stock in-tank pickup. The bigger pump cavitates the fuel and it boils / vaporizes whatnow. One must modify the pickup, youknowwho has an excellent solution for this in his twin turbo kit.
- Puting in a high resistance filter upstream of the pump. This is another reliable way to boil fuel. At most, one should have a strainer screen there.
- Making fuel hoses without knowing what you're doing from wrong materials.
- Using standard metric adapters in various parts of the 928 fuel system. Just because it's metric doesn't mean it'll seal...
- Too small fuel lines, especially the return line which is often overlooked. I have a big pressure line and the stock return line, which only makes sense if one is planning for an electronic control of the fuel pump to match the fuel demand. Otherwise, this is a big "duh!"
- Tight turns and many bends in the fuel lines. Every turn and every constriction heats up the fuel hugely. This is an easy way to turn your fuel pump into a fuel heater.
- Removing dampers. Guess what, it turns out that the fuel dampers on a batch fire system serve a purposes other than just decorating the engine bay. With big injectors and high fuel pressures, the batch fire can create significant resonances in the fuel system. I now have more, not fewer, dampers on my fuel system (4 + mechanical regulator) than the stock system (3 + mechanical regulator).

andy-gts

Tuomo,

Is your twin monster down for the winter or is the fuel issue only with wot for prolonged periods and really a non issue with tootling around town?

ptuomov

The car was drivable with what it had before this latest round of enhancements.

(A digression: I wouldn't drive it out in the NE winter, though. The 2011 E63 is plenty powerful and safer for my winter driving. The highways here are covered with a bottom layer of ice and a top layer of massholes -- a loud horn, traction control, well-exercised middle finger, and a full insurance coverage with an umbrella are pretty much required.)

The 928 TT was basically ok to the point of problems not being detectable with the pump driven in the low mode and the AC on, as long as the tank wasn't less than 20% full. Even then, it wasn't bad, just detectable. Still, it wasn't as good as it could be, and it _needs_ to be as good as it can be.

My hope is that we can make the pump less loud in cruise and fuel temperature more stable in general with the EFPR. Perhaps counterintuitively, the fuel temp is _not_ a problem at WOT for prolonged time, it's a problem when cruising at a very low load with a close-to-empty tank. The goal is that to make the fuel system work well until the last drop of fuel is gone and the tank has been sucked dry. There's no uncertainty about whether this goal will be met, it will be met. I just don't know whether it will be met over the next week and with the Fuelab EFPR.

By the way, with a "small" pump like Bosch 044 the fuel system would not have any of these problems, we're trying to design something that works as _a system_ with a really big pump. The Bosch 044 setup that comes with youknowwho's standard kit is completely issue free for lower peak power levels.

I tried blowing the stock engine up the one last time two months ago, and failed. The attempt got us some very useful feedback on youknowwho's twin turbo kit, though. The kit was designed up to about 14 psi of boost, and it works great in meeting the design requirements. However, at 20 psi, which is outside the original design requirements, problems appear. Youknowwho is now designing a higher pressure version of the kit for my use, the main two enhancements being grooving the MAF (a proven solution by the supercharger crowd, Dave Lomas posted photos of his similar solution for example) and a different anchoring system for the air box. The connection between the MAF and the throttle body can now easily support the weight of a grown man without any help from the air box and the 600+ pound engine can be hanged from the air box with no problem. The bracket design is such that differing thermal expansion of the block, intake, and the air box will also not cause a problem, that was a design requirement for the new bracket system.

Here's a photo of the grooved MAF housing:

andy-gts

is your fuel system of a larger capacity than the one that Hacker talks about and posted about earlier this year? I guess you have probable spoken to him what he is using for his turbo beast....wow just awesome "going where no man has gone before" oops showing my age...

great work and thanks for posting this increadible journey !!!

17prospective buyer

Wow... the money invested in this :O i cannot comprehend.

ptuomov

Todd Tremel's twin turbo I believe runs two Bosch 044's and burns alcohol (E85). By my understanding the second pump is switched on based on boost, although I might be mistaken. The system is a traditional return line system, with the one open loop control aspect being the second pump on-switch point.

My car runs pump gas and has a single Fuelab Prodigy 42401 pump. The goal is to make it pump run in a closed loop control based on the fuel demand using the Fuelab EFPR 52091 as the control unit.

These two systems are just two different ways to skin the same cat. A single Bosch 044 flows 53 gph (200 l/h) against 72.5 psi (5 bar) pressure. Two of them flow double that, or 106 gph (400 l/h). A single Prodigy 42401 flows about 160 gph against 72.5 psi pressure at high speed and about 100 gph at normal speed. It's not rated for continuous duty at the max speed, however, it only has stree-car-like indefinite-run-time reliability when run at normal speed.

In words, the two alternative systems are "close enough for goverment work" equivalent in terms of flow capacity. It's really the control logic where the differences will be, if there will be any meaningful differences.

ptuomov

The parts that are actually currently installed in the car are surprisingly affordable. One could recreate the setup with a phone call and $20k. The dead ends and detours have been more expensive, though. I guess there's enough of an R&D aspect in this project since we're trying to do things a bit differently every time.

As far as money goes, it's all relative. For example, I today learned that our family has adopted a rhino. Yes, today's news is that I am sponsoring a rhinoceros. You know, the actual live mammal and not some sports team with the name "The Rhinos". What this cost me and why it was done was never communicated to me, and I am figuring that I will be happier not asking any questions. Nobody's asking me any questions about what the things in the garage have costs either.

[Edit: Apparently, the rhino was a gift. But the general point still applies.]

Rob Edwards

LOL, good stuff, Tuomo. My bride and I have developed a DA,DT approach to the checkbook. I don't ask about her vacation budget, and I don't tell about my German carparts problem....

AO

I see our issues are universal when it come to looking the other way re: $$$$.

Yes. Todd's car runs dual 044 pumps and the 2nd one is triggered somehow to come on at the right time. Todd didnt even try t use the stock LH and when to an aftermarket controller. Autronic I believe, which has many cool gizmos and controls and customizable parameters.

Lizard928

Why not just use the flappy output to trigger a relay to switch your pump to high gear.

ptuomov

Because it's not the best possible solution. With a pump this large, I need continuous adjustability to create the most stable possible fuel temperatures. Recall that it pumps as much as two Bosch 044's at the lower setting.

dprantl

This begs the question, do you really need a fuel pump that has more flow than two Bosch 044 pumps? It seems Todd is not having fuel delivery issues and his setup has a bunch more power than yours.

Dan
'928GT S/C 475hp/460lb.ft