Closed loop return fuel system

Can I run closed loop fuel pressure system with a return and mechanical FPR in a boosted setup.?

Can I tighten down the mechanical regulator and pwm against that?

Example- set base pressure at 45psi with pump at approx 40% duty.

Let the ecu pwm against it until demand is high and the pump is at 100% duty then the mechanical regulator will be maintaining rail pressure. This would require the boost ref to be connected to the regulator..

Could the regulator be used with no boost reference and basically just act as a bleed off to stop the pump from stalling.?

Thanks 

You seem to be making things a lot more complicated than they need to be. What is wrong with a conventional regulator?  Or at most a low/high relay/resistors like Toyota does for the fuel pump at idle so the regulator isn't overrun?

My fuel system suffers from over heating.. Base pressure drops by 5psi on a real hot day.

Ecu is adding a bunch of fuel, then if I auto tune on a hot day.. it over fuels on a cold day. I hit shutdown on afr protection and the ecu is hitting max correction.. 

My tank is close to exhaust.. Trying to keep pump from over working and making things worse..

Put a cooler in the return line? < the pump is not heating the fuel much/

Put a heat sheild between the tank and the exhaust?

Theses are all things I'm going to do this season. Hottest part of the entire system is the line exiting the tank after pump.  I'm currently an open loop pwm setup atm. Lower duty at below 100kpa and 100% over 100kpa. This has helped but just need the pressure to remain consistent. That's my main aim. 

Ceramic coat the exhaust pipes near the tank.

This solved my issue with melting control arm bushings.

It would also cut down on heat in the cabin and keep it in the exhaust to help it flow faster, etc.

You don't have to do ALL of it, just the parts you're concerned about.

You could also look at adding a fuel cooler loop/surge tank somewhere closer to the engine and move the primary pump there, with a lower pressure pump keeping the surge tank full and the return line dumping into the surge tank.

Ya you need to fix the heat problem and not band aid the problem with a work around in the fuel system. 

Its pretty much unavoidable, but I'll be doing everything possible this season.

But I still want to get the system in a closed loop mode. Really want to keep the pressure consistent and datalog it. Anyone care to say how it'll work with the regulator and boost?

My megasquirt ecu has 3 different options for closed loop, Gauge, absolute or differential. Which should I use.?

The first question is what have you got for an ECU?  That will control most of what is possible......

The simplest answer to fuel heating is a solid state relay to the pump, then PWM the pump just high enough to maintain pressure..aka a small amount of flow in your return hose.  I use a table with V bat vs MAF,,,the MAF tell you how much fuel you need and Vbat is how fast the pump runs at any given PWM, this isn't tuned yey but this is what it looks like:

The next step up would be to use the table above as a feedforward term to a PID controller so the fuel pressure will snap high enough quick enough when you rump the throttle, then the PID will get it perfect...at least that's how I would do it.  This would get rid of the need for the mechanical FPR but...but a FPR is by far the easiest way to control fuel pressure. 

As you get into closed loop fuel control you probably also want all the fancy stuff the OEMs use these days with the injectors.  

You need to deal with th viscosity and density of the fuel, here are the inputs I use:

Then control fuel flow as a mass flow which means well characterized injectors.  I use injector dynamics injectors and the cal data in GM format:

Its not so easy but certainly possible and how all modern OEM and better after market ECUs do it.

If you don't have an ECU that will permit this type of control a much lower tech solution is 2 pumps.  One runs all the time and is sized for like heavy cruise.....this supplies may 1/3 your total HP needs so most of the time you are only pump what you need.  Then a second pump in parallel (which means it needs a check valve) that you can turn on at a set rpm or throttle position of manifold pressure either with the ECU if you can or a separate switch.

For boost you need the fuel pres reference to the manifold pressure....the simple way is a FPR that accepts vacuum hose and you're done.  In the most fancy world you include it in your PID system.

Hope this helps, if you still have questions ask.

Craigorypeck said:

My fuel system suffers from over heating.. Base pressure drops by 5psi on a real hot day.

Ecu is adding a bunch of fuel, then if I auto tune on a hot day.. it over fuels on a cold day. I hit shutdown on afr protection and the ecu is hitting max correction.. 

There are 2 things here.

1) turn off auto tune.  The only think that is good for is bugging up your tune.  If you want to do a closed loop will at highway cruise thats fine as long as its clamped to no more than say 10% correction and clamped to a a tight MAP /TPS/MAF window where it operates....but it should be off above about 1/3 power.  I know the ECU makers love to preach how wonderful auto tune is and how its the answer to everything....but really its only good for 1st pass type tuning and even then seems to be more bother than its worth, best to turn it off.

2) If you're using a mechanical FPR and pressure is dropping something is wrong.  Either the  regulator has a problem or much more likely the pump is too small of failing.  This needs to be corrected before anything else will help.

There are a few factors working against me. Black car that's parked outside for 9+ hours a day in 30c days with no Sun shelter. Commute to work in the mornings and all is fine. Below luke warm tank.

Tank is in trunk so pretty much subjected to cabin temps all day. Commute home... 45mins traffic. It gets nasty, by the time I'm home its dangerously hot. Below half tank and it's off the scale...

The fuel may be vaporizing and I'm getting pressure loss there, the clearances within the pump tighten and I lose pressure there... Maybe even a hot spring in the regulator doesn't hold the exact same pressure as when cold.

I've got a new pump for free from the manufacturer and same problem with that brand new pump.

This year I'm doing everything i can. 

Deleting coolant line on inlet manifold with a pheniloc spacer to attempt keeping heat off the rail.

Insulating trunk floor to keep the heat from the muffler below away. 

Installing a cooler in the trunk with ecu operated fan.

I've removed 75% of the trunk seal to try getting some circulating air in the trunk. 

I'm already running a pwm open loop setup.

Fuel pump manufacturer went through my entire system to check fitting ID's filter type. Hoses ID's. They say its a rail issue. What else can i do..!!?

Ive tuned the car on hot fuel as id rather run rich when cold abd try and be in the ballpark when hot.

In reply to mke :

Ecu is megasquirt ms3 pro.

Full pid pwm control option for fuel pump.

No MAF in my setup. Just MAP and TPS.

Maybe when i do all the things I intend on doing the pressure drop issue may go away. But I'm installing another pressure sensor and wiring it to the ecu. If nothing else I'll be able to log it.

There may be a way to open loop the pump based on the fuel pressure and not based on the </> 100kpa MAP I have it set up with atm. 

Pressure drops, PWM just ramps up to a higher duty. 

Craigorypeck said:

Pressure drops, PWM just ramps up to a higher duty. 

If you are already running the fuel pump 100% and pressure drops then you can't fix it with the ECU. 

If you are not running at 100%, and you've now said you're running open loop so probably not 100%????,  I suggest you disable that feature and run at 100% and see if it's your open loop settings causing your issue.

This is a weird problem.....but now that you've said the ECU is messing with your fuel pump 

Normally the only issue with fuel heating is excess fuel vapor emissions escaping the tank....you basically can't have vapor in the pressurized portion of the system so that is pretty unlikely to be the cause.

I don't run an MAF sensor either, but I calculate MAF as part of the fuel control...its the only way to know how much fuel mass to deliver and why when I had a fuel pump failing my pressure was down to about 10psi before I even noticed....the ecu just kept calculating the correct injector time for the current Vbat, fuel press, and temp and it wasn't until the injectors were at basically 100% the problem showed up as a poor running engine.   More basic systems like MS that use ....well what is the polite word for obsolete these days...aged? processors can't handle that kind of math or the storage needed to do it and still use a simple error table approach to fuel tuning where the "VE" table is really just an injector pulse width at its core not actual VE which would let you calculate MAF. ...but it works well enough for most stuff though and isn't the issue you're having.  but for  god's sake turn off the auto-tune and turn off the fuel pump pwm until whatever is causing your pressure loss to be found and corrected.

In reply to mke :

oh, and how were you setting the PWM for the fuel pump without a fuel pressure sensor?  Reading the pressure gage on the regulator or just dropping PWM until you saw a change in O2 readings?  

My issue was worse when the pump was running flat out all the time. 

Reducing the duty when off boost helped. 

Autotune is something I run now and again when i change a parameter like cam timing for example. Otherwise its off. 

I have a pressure gauge in the dash. 

I set base pressure then reduced duty until it started to drop off. Then just increased the duty back up a bit.

Thats my in vacuum duty. 

Once I get on boost the ecu just ramps pump up to 100%

Craigorypeck said:

Autotune is something I run now and again when i change a parameter like cam timing for example. Otherwise its off. 

Good, off is what you want. when you are troubleshooting.  If I'm not mistaken the MS autotune is an O2 based system...changing cam timing often changes the optimal O2 reading a little, many thinkings do so as I said earlier its a fine 1st pass kind of thing but nothing you want to count on to deliver optimal results.

My issue was worse when the pump was running flat out all the time. 

Reducing the duty when off boost helped. 

Ok but here's the thing....if reducing the pump speed in the past helped lessen your problem then how can increasing the pump speed now help you problem?  they are opposites are they not?

Did you/do you have the issue at all rpm and power levels?  Normally this is a high hp power thing that moves down toward lower hp levels as the problem worsens.  I say hp because its normally fuel flow rate dependent not MAP dependent...more MAP*RPM which is basically like MAF or hp, with the wrong units but probably close enough.

and you're seeing the dropping pressure on your dash gauge and in you mixture? 

Do you also have a fuel temp gauge to confirm its fuel temp related and not simple say fuel pump temperature related?

If you had the same issue with the pump at 100% then it still sounds like a bad pump honestly...what fuel pump are you running?

hmmm....I think you said 5 psi drop.....always around 5 psi at all MAP and rpm?  If so that's that's kind of weird....pumps have flow curves so you'd expect more pressure drop the more flow you ask for as you bang up against the pump curve, only the regulator should be able to produce an across the board pressure drop.   What regulator are you using BTW?

mke said:

Craigorypeck said:

Autotune is something I run now and again when i change a parameter like cam timing for example. Otherwise its off. 

Good, off is what you want. when you are troubleshooting.  If I'm not mistaken the MS autotune is an O2 based system...changing cam timing often changes the optimal O2 reading a little, many thinkings do so as I said earlier its a fine 1st pass kind of thing but nothing you want to count on to deliver optimal results.

My issue was worse when the pump was running flat out all the time. 

Reducing the duty when off boost helped. 

Ok but here's the thing....if reducing the pump speed in the past helped lessen your problem then how can increasing the pump speed now help you problem?  they are opposites are they not?

Currently running a simple open loop tactic, below 100kpa 70% duty above 100kpa 100% duty. Staying out of 100% duty helped my system temps. Running closed loop could essentially reduce those duties further helping my case. Simple...?

Did you/do you have the issue at all rpm and power levels?  Normally this is a high hp power thing that moves down toward lower hp levels as the problem worsens.  I say hp because its normally fuel flow rate dependent not MAP dependent...more MAP*RPM which is basically like MAF or hp, with the wrong units but probably close enough.

and you're seeing the dropping pressure on your dash gauge and in you mixture? 

Yes. First hot day of summer I seen the gauge reading low.. Boosting hard and it hit lean afr shut down. So i autotuned on hot fuel to try and avoid that. Id rather run rich when cool than lean when hot. Not exactly sure on the fuel pressures when on boost as its hard to read that gauge when driving. I have way too many gauges! 

Do you also have a fuel temp gauge to confirm its fuel temp related and not simple say fuel pump temperature related?

I'm installing a flex fuel sensor atm that also has a temp sensor within. 

If you had the same issue with the pump at 100% then it still sounds like a bad pump honestly...what fuel pump are you running?

Aeromotive stealth 340 intank. They even sent me a brand new pump for free .When i answered all their queries without any advisories.. Everythings as it should be.. Fittings, hose, filters. I even raised the pump up on the hanger to prove it wasn't too close to the tank floor.

hmmm....I think you said 5 psi drop.....always around 5 psi at all MAP and rpm?  If so that's that's kind of weird....pumps have flow curves so you'd expect more pressure drop the more flow you ask for as you bang up against the pump curve, only the regulator should be able to produce an across the board pressure drop.   What regulator are you using BTW?

Aeromotive 13109 regulator.

After the cooler is installed and I do all the other preventional tasks ive lined up Maybe the issue will go away.

Still wanna know who has run a closed loop setup with a mechanical regulator and a return line. The PID could be adjusted so the mechanical regulator does its thing first then the ecu could fine tune it...?

At this point I can't have it like it was last summer. If I have to over complicate it to make it work like I wish, well so be it... 

I think I understand. 

1) At low (no boost) power you never have a fuel pressure issue

2) On hot days on boost you do have an issue

3) the hotter the fuel gets the worse your problem gets

Correct?

This says your pump is the issue, its just to small.  Certainly all the things you're doing to keep the fuel cooler will help and may get you there but not likely since you aren't going to the heart of the issue.

They make an intake version of the A1000

In tank A1000

I ran used the remote version of this pump on my last engine, a 3.0L at 24psi boost making about 650hp and it was just barely big enough...it would pop a 15A fuse at high boost, 20A fixed it.  I had base pressure set at  45 as that was about as high as I could go and maintain pressure.......

OOOH!!!!  Where is your base pressure set?  Pump ALL flow more at lower pressure.  If you are at 58-60 you could drop to 40-45 and re-tune and probably solve your problem.  It will heat less at lower pressure too.

I'm using the same A1000 pump on the now about 900hp 5.4liter, but its naturally aspirated meaning lower pressure so I think it will work but I've not gotten to the high power tasting to know for sure....even though they say 1000hp they mean under ideal (for the pump) conditions.

Another option is use small lift tank and a remote pump.  The intank pump fills the lift tank and the extra free returns to the main tank, 0 psi.  Then you can use a large remote pressure pump. Road racers and sometimes drag racers use this setup to prevent fuel starvation at high G but it would give you more options on component choices.

To your question about closed loop.....sure you can crank up the regulator pressure and run closed loop at a lower pressure.    The regulator just acts like a blow-off safely valve. just be sure you give yourself a decent buffer...say 10psi because if you get close the regulator will start bleeding off fuel land confuse you PID settings. ....but this won't likely fix your undersized fuel pump issue.  It will probably delay the issue by reducing fuel heating rate but its still heating so it seems like a long shot compared to the sure think of a bigger pump.

mke said:

I think I understand.

1) At low (no boost) power you never have a fuel pressure issue

No. Once hot base pressure drops from 42.5 to 38psi

2) On hot days on boost you do have an issue

On hot days all over the map not just on boost.

3) the hotter the fuel gets the worse your problem gets

Correct?

Yes

This says your pump is the issue, its just to small.  Certainly all the things you're doing to keep the fuel cooler will help and may get you there but not likely since you aren't going to the heart of the issue.

They make an intake version of the A1000

In tank A1000

I ran used the remote version of this pump on my last engine, a 3.0L at 24psi boost making about 650hp and it was just barely big enough...it would pop a 15A fuse at high boost, 20A fixed it.  I had base pressure set at  45 as that was about as high as I could go and maintain pressure.......

OOOH!!!!  Where is your base pressure set?  Pump ALL flow more at lower pressure.  If you are at 58-60 you could drop to 40-45 and re-tune and probably solve your problem.  It will heat less at lower pressure too.

I'm using the same A1000 pump on the now about 900hp 5.4liter, but its naturally aspirated meaning lower pressure so I think it will work but I've not gotten to the high power tasting to know for sure....even though they say 1000hp they mean under ideal (for the pump) conditions.

Another option is use small lift tank and a remote pump.  The intank pump fills the lift tank and the extra free returns to the main tank, 0 psi.  Then you can use a large remote pressure pump. Road racers and sometimes drag racers use this setup to prevent fuel starvation at high G but it would give you more options on component choices.

To your question about closed loop.....sure you can crank up the regulator pressure and run closed loop at a lower pressure.    The regulator just acts like a blow-off safely valve. just be sure you give yourself a decent buffer...say 10psi because if you get close the regulator will start bleeding off fuel

What if the boost reference is still connected to the regulator. ?

 

land confuse you PID settings. ....but this won't likely fix your undersized fuel pump issue.  It will probably delay the issue by reducing fuel heating rate but its still heating so it seems like a long shot compared to the sure think of a bigger pump.

 

Pump is 750hp capable. I'll be lucky if I can get near 450-500 this season. 

I'd run the fuel pump speed control in open loop - more of an "X duty cycle for Y RPM" - and let the regulator handle the pressure control. Having two active control mechanisms going on at once is likely to end up chasing its tail.

Craigorypeck said:

mke said:

I think I understand.

1) At low (no boost) power you never have a fuel pressure issue

No. Once hot base pressure drops from 42.5 to 38psi

That is a little confusing.  Does running the pump at 100% fix the base pressure?

If you drop your base pressure to say 35psi or even maybe 30 you might be fine if your injectors will still flow enough at the lower pressure.  Otherwise I really think you need a bigger pump

To your question about closed loop.....sure you can crank up the regulator pressure and run closed loop at a lower pressure.    The regulator just acts like a blow-off safely valve. just be sure you give yourself a decent buffer...say 10psi because if you get close the regulator will start bleeding off fuel

What if the boost reference is still connected to the regulator. ?

Something like +10psi  would be good with the reference connected 

+10psi+ boost is what you'd want with the reference disconnected I think.

.....but if the pump can't even maintain base pressure with no flow none of this is likely to help.  I am honestly surprised the pump is acting like that but it is what it is I guess.  

What are you planning to run for a fuel cooler?  Having the fuel temp will help in knowing how large the cooler needs to be since the hotter the fuel the more effective any cooler is so knowing temps helps in picking a size....or just install the largest that will fit in the space you have and know you did the bast you could.

Craigorypeck said:

Pump is 750hp capable. I'll be lucky if I can get near 450-500 this season. 

Thats the NA hp number....the A1000 1000hp pump I run was just barely big enough on my 650hp 24 psi engine, the higher the pressure the less fuel it will flow.  I never had the problem you're having though....my pump isn't in the tank but its pretty near the headers so I know its hot and I had it running 100%, I just added the control table mostly to cut down pump noise and my fuel tanks are in the engine bay so the fuel has to be 200F or so...I just added the temp sensor too so no historical data to look at.

In reply to MadScientistMatt :

Yes, you don't want systems fighting.  Either set the PWM high enough to KNOW the regulator is in charge or set the regulator high enough the KNOW the ECU is in charge.....anywhere in between is a failure.

mke said:

Craigorypeck said:

Pump is 750hp capable. I'll be lucky if I can get near 450-500 this season. 

Thats the NA hp number....the A1000 1000hp pump I run was just barely big enough on my 650hp 24 psi engine, the higher the pressure the less fuel it will flow.  I never had the problem you're having though....my pump isn't in the tank but its pretty near the headers so I know its hot and I had it running 100%, I just added the control table mostly to cut down pump noise and my fuel tanks are in the engine bay so the fuel has to be 200F or so...I just added the temp sensor too so no historical data to look at.

Google found this article

Pump flow test

Here's the info that makes it easy...hp vs pressure.  The higher the pressure the lower the hp.  With boost you need to add boot+base pressure to find the right column but the pump you have is probably not going to get you much over 500hp even under ideal conditions.  When I was originally building my boosted engine I remember planning for about 12 psi 400 hp and thinking the A1000 was the pump....but that was some time ago and options were more limited but I wanted a bit of freedom wih base pressure increase I needed more than the injectors had at 43psi.  On the 2nd build the same pump gave me 650hp  ut with big injectors so 43 psi bse pressure was enough.