Would this work? Connect output from existing Bosch pump to fuel supply and eliminate the other pump. Essentially the Bosch 044 becomes an in tank pressure pump. Surge tank will just stay empty since pump pulls from main tank and return line dumps into main tank. Will the pump overheat without being submersed in fuel?
Doesn't the return go into the top of the IST? That would keep it full and overflow like it does right now.
In reply to jgrewe :
I'd be changing that so the return dumps into the main tank. The surge tank would basically no longer have fluid in it other than what seeps or splashes in.
This would effectively turn the Bosch lift pump into an in tank pressure pump (with no surge tank).
I'm thinking of this as a way to test the theory that the pressure pump can't pull enough fuel to maintain pressure under boost. I could do this for free as a test, then decide what the final solution looks like. Ideally, having a surge tank again.
In reply to sevenracer :
If you left the return into the surge tank it will still go back to the tank eventually but it will keep the pump covered/cooled for the test.
In reply to jgrewe :
Good point. Really no need to re-route the return line.
Agreed leave the return to the surge to cool the pump. The 044 still wont like to pull fuel but its not as far. Do the test w a fullish tank.
Ok, will give it a try. May need to order a fitting or two to get it set up.
I did hear back from the guy that sold me the surge tank setup. He has not seen this issue in his personal car (boosting to 20 psi), and he was not aware of any other customers running into this. He asked for some more details on my install which I sent, and I'm waiting to hear back. He is using a Bosch 044 for the pressure pump, so that is one notable difference.
What do these make at 20psi? I have seen an 044 pull through a stock lift pump and make low 300s
Paul_VR6 (Forum Supporter) said:What do these make at 20psi? I have seen an 044 pull through a stock lift pump and make low 300s
Not sure. I think I read something he posted online about his car making around 500 whp. I want to say it was a ported motor with a larger than stock turbo.
20psi might be 350 crank horsepower or it might be 800 crank horsepower, depending on the turbo and the porting.
Rotaries are weird under boost, they can make more power than just the pressure increase suggests, for various reasons.
There is a use-case thing here too. What works fine for a couple pulls on the dyno or a pass down the dragstrip might not work well when driven on the street or track where everything gets to heat-soak.
Pulled the fuel tank and surge tank back out tonight. I discovered I neglected to buy one of the fittings that I needed, but I'm not sure how much that matters. There doesn't seem to be enough space to install fittings and a loop of hose to connect the pump outlet to the supply fitting in the lid. I don't think I have enough space between the pump and the lid to bend the rubber hose over without it kinking.
I am debating whether I want to remove the AN8 bulkhead fitting from the lid, and run a rubber hose from the pump through that hole to the pressure pump. Seal it up with a grommet and silicone adhesive. That would let me test the fuel pressure rise with the lift pump directly feeding the pressure pump.
Thinking of the final solution, in the surge tank, one side has the lift pump, and the other has the supply tube and the level switch. There is probably space for two pumps in there - one to fill the surge tank, and a pressure pump like a Walbro in tank 450. But it doesn't look likely that I could keep the level switch.
What is the purpose of the level switch?
It notifies when the surge tank level falls below 20%. Allows advance warning of fuel system losing fuel supply.
Intended to signal ecu to enter engine protection mode.
The more I think about it the more goofy this setup seems. It seems like it has way more failure modes than a stock, unmodified tank.
Paul_VR6 (Forum Supporter) said:The more I think about it the more goofy this setup seems. It seems like it has way more failure modes than a stock, unmodified tank.
Sure, if you're talking about a stock unmodified tank that was designed and tested to meet the needs of a fuel injected, forced induction motor. In this case, the stock tank was designed and tested to support an external pump delivering about 2.5psi to a carbed engine. And I can attest to the fact that the stock fuel pickup can starve for fuel with anything less than a 1/3 of a tank (on track).
I opted for this because it appeared to be a proven solution for my particular application, which hopefully would reduce my time investment to develop and fab my own solution. Obviously, it hasn't worked out that way, but so it goes sometimes.
I hear ya and maybe there is no good answer! After looking at what you've been through a decent intank with a hydramat would be my go-to.
I have fuel starved at half tank at rallycross before I made a surge tank.
The pickup is at the front left of the tank, so left turns or accelerating can uncover it...
Spent some time on this over the weekend.
I connected a 3/8 hose to the output of the surge tank pump and removed the AN fitting in the lid to provide a hole to get the hose through since there wasn't space to connect the pump to the AN fitting. Added some silcone sealant around the hose as a temporary seal. I connected that to the external pressure pump.
Wasn't sure there would be enough space between the top of the tank and the body to loop the hose without kinking it, but it worked out ok.
Tested the flow of just the in tank pump: 6.35L/min or 381 lph unrestricted.
Tested the whole system, including the external pump, by measuring output of the return line (car not running). 6.05L/min or 363 lph with regulator set at 44.8 psi.
Also retested the fuel pressure rise using shop air providing signal to ECU, FPR, and BOV all on same circuit. Looks pretty much dead on - the base fuel pressure was set about 1 psi too high when I logged this. Not sure if the pressure decay was leakage from the shop air or something on the car (BOV maybe?).
Then, I logged the car idling. I think the fuel pressure is above expected because I set base pressure with batt voltage, not running voltage:
Then some pulls to 6300 rpm. Still seeing a gap from actual vs expected fuel pressure, though maybe less than the 5psi gap I had before. I did notice that the measured AFR is richer than target from about 2800 rpm on up. An example:
Not sure what I learned with all this, other than connecting the pumps in series didn't really address the fuel pressure rise gap.
Interesting data for sure. 1psi on your bench test looks good. The 5psi difference in the road test seems like it's catching the problem still. It would be interesting to see if setting the FPR 5psi higher/lower makes that delta smaller/larger. If you are already on the rich side it should be safe enough. My hypothesis is if the pump is near the limit the higher pressure will show more dP than if it's lower, as it strains it more.
I can try bumping up the fpr and see how that changes dP.
I don't know how reliable it is, but the ecu 'fuel flow estimated' is 1.6L/min. Way less then the 6L/min flow I measured from the return line. If that's accurate, seems hard to believe the pump can't deliver enough fuel.
You can do the math yourself if you look at the duty cycle * injector rate * number of injectors
Just closing the loop on this. I was hoping to find a true root cause for why the fuel pressure doesn't rise equally with boost, but never really got there. I contacted AEM support for the FPR, and they thought the 5/16" return line back to the tank might be a restriction and causing the delta. Seems a bit odd since the supply line is also 5/16".
The tuner I'm working with, without knowing AEM's response, also thought there was a slight restriction in the return line, and he recommended just bumping up the fpr by ~5psi and he could tune it that way. So that's what we did. Went back to the dyno and had no issues with fuel supply this time. Limiting factor at this point is the stock intercooler, and there's not much headroom left in the injectors either.
Anyway, I'm going to drive it as is for a while, but eventually would like to revisit the surge tank setup. I'm happy to have the full rev range to play with(I had the redline set to 5900 because of lack of fuel at high rpm), and the motor is plenty stout.
You'll need to log in to post.
