Dual Fuel Pump Rales

[lmcchesney]

The fuel pressure of the D-jet using one pump to supply four injectors is not as stable as it should be. Ray Green suggested the development of a dual fuel pump system to supply stable fuel pressure to the injectors. I have designed a plan for the use of two pump in the front trunk (avoidance of vapor lock) feeding the engine compartment via two 3/8" fuel lines which feed the separate rales combining to enter into one fuel pressure regulator, (to assure equal pressure in each rale), with a return line via a single 5/16" fuel line to the front trunk. The question is weather the return line to both pumps can be done in series or required to be parallel. What is the thoughts?

Thanks,
L. McChesney
Lamcchesney@netzero.net

[Kenny Powell]

1 isn't enough for you??

[lmcchesney]

Read the script Kenny. Ray believes that a single fuel pump feeding both fuel rales results in wide flucuations in fuel pressure throught the throttle range. Since the fuel injectors and subsequent fuel mixture is a timexpressure formula, variations in real pressure results in lean spots in the throttle range. Thus, using a parallel fuel pumps and larger volume fuel rales hopefully will result in a stable fuel pressure and thus absence of lean spots.
L. McChesney

[norustscott]

I would not agrue the variation in pressure...but what is the head room built into the system. if the regulator bleeds off @ 30 whatever psi...the pump must put out ... big guess here 45 psi. If I have a constant regulated pressure of 30 psi I would guess that the injectors, designed to operate and deliver fuel at a constant PSI, would only require 20 or so psi. So the system could vary as much as 33% and still function without adverse conditions... tat is a variation in PSI @ the rail...

My 2 Cents

[lmcchesney]

Scott,
It has been my understanding that the volume injected is effected by an increased pressure. Thus, increasing the resting or working pressure of the fuel rales will result in an enriched mixture. With increased volume in the rales, removal of the injector volume will resulting in a decreased change in pressure in the rales and therefore, a more linear enrichment.
Thanks,
L. McChesney

[norustscott]

With the regulator you cannot increase the system pressure without adjusting the regulator. Pressure failures are related to leaks and bad pumps. The only way to reduce volume is to down size the fuel lines (restrict flow) or reduce pressure at the regulator or with the wrong pump.

For a stock engine I think the dual rail feed is folly. For a radical 2.2 with fuel injection ...? naw, still folly. make sure all of the stock components are working correctly, I would be willing to bet the system components and design will handle any application. Der Engineers put the system to the TEST or sure.

[lmcchesney]

Scott,
Ray has experience and data on the fuel pressure variations on the stock setup. We plan to have both rales feed into a single regulator to assure matching fuel rale pressure seen by all four injectors.
I can't follow your reduced volume statement. I want to increase fuel rale volume. Yes, the system is for a 96 x 78 build.
L. McChesney

[Bleyseng]

I a higher than 35psi the injectors squirt of stream of fuel instead of a cone shaped fog. This is the problem. Try is it out to see for yourself!
I will have to get a better fuel pressure gauge that is way more sensitive to see what Ray is taking about.

Geoff

[red-beard]

I have access to some high speed pressure transmitters if someone wants to really see what is going on.

[lmcchesney]

Interesting finding. So as I understand it, we would need to keep the fuel rales at 35psi to maintain cone atomization. Thus, we will need to increase duration to adjust for increased volume to keep the same A/F ratio. This can be approximated by adjustments in the MPS and or CHT.
I think we are discussing connected issures on this forum an on my email from Shoptalkforum. Either way the flow of information is fantistic. Thus, we could run into a problem by the need to increase duration of injection while decreasing cam lift duration to accomidate for the larger intake valves.
I really can't wait to get this experiment into the dyno.

Thanks,
L. McChesney

[lmcchesney]

Read-Beard,
Ray Green says he has seen this in his combination. I don't know that your need a fast reacting pressure transducer vs a serial recording to see the trends. Maybe Ray can add some light.
L. McChesney

[ray greenwood]

OK...hello, a few notes. (1) the pump puts out "0" pressure. Think about it. It is a volume instrument. Wire it up with no hose...and it puts out "0" pressure. Intriqued? Thats because pressure is only built by volume against a restriction. That would be the fuel pressure regulator. The stock D-jet pump is capable of about 70 psi. But...you will also find that the pressure at that level is very unstable. The issue here is not pressure capability per-se....it is the ability to refill the volume taken from the fuel line...without showing pressure variation.

Drain the fuel system and measure the volume...including the fuel filter and all the lines and the pump cavities. Thats a lot of fuel. More than you think. All of that volume has to be moved forward and brought back up to pressure from the average 6000 injections per minute. Thats a constant pressure bleed. The problem is, and this depends upon the condition of your pump and your rate of tune, and the condition of the fuel regulator.....is that even a brand new stock pump sometimes cannot replenish volume fast enough to keep the needle from moving more than an adequate 1 psi. Sound stupid? Think about it.

28 psi....if you are losing 3 psi.. thats 10.71% of your fuel pressure....in a system that is supposed to be running a FIXED pressure because the ECU has no sensor to tell it otherwise.
Thats 3.57% of loss for every 1 psi of pressure dip. The average D-system has about 3 psi of flux. Bad news.

Think about this scenario...when adjusting the MPS, do you often find that in order to get the best possible mid-range running....you have to leave the idle a little richer than you wanted? And STILL occasionally get a very top end clatter for a split second around 4300 rpm? Thats fuel pressure.
The largest dips in pressure are immediatel when opening the throttle...and when at WOT and high rpm. This is not only at max injection duration...its happening more times per second at the high rpm.

Also...its more noticable when you are using advanced timing. The vacume signature and the injection timing...which advances with the timing for obvious reasons...can produce a very high speed pressure needle swing that can nearly mimic rpm. It fools you into thinkingits justthe enginevibration...but its not. Its higher injection rate...and poor replenishment from the pump.

More fuel PRESSURE is not what we are looking for. Its more fuel volume. The fuel pump is not different than any other pump. The higher the pressure output it is restricted to....the lower the volume output...because the presure column limits fuel ingress into the pump.

Also...unless you check fuel pressure on both rails...you have no idea if you have balanced pressure or even have this problem at all. In the 24 years of working on almost nothing but D-jet...about 3 of 4 cars have this problem to some degree.

Don't want to run 2 pumps? Ok..then shorten the pressure distance and get 1 higher volume pump. Like a CIS pump...but put it in the back. Its better to suck father than try to push it farther. But guess what?....70% of all CIS pumps need a booster pump. The high head pressure needed, gives very poor fuel ingress and poor suction....so putting it in the back...without a boster pump is a moot point. So its back to 2 pumps.

Silly?...Porsche GT-2 and GT-3 many times run 3 pumps. 2 main and 1 for reliability. The pumps are very high volume...and yet the fuel usage is not very much higher than any other six cylinder Turbo.

Once I got my fuel pressure to a rock solid 1 psi....I had to re-do all 3 of the main fuel mixture adjustments in the MPS. It made that much of a diference. Through all of the strange nit-picky tuning in D-jet.....I was eventually able to get a hair over 100 hp out of my 1.7L. Check into it...or not. Ray

[Scrappy]

I have always been told that an electric pump is a "pusher" not a "sucker" and that mechanical pumps are "suckers" not "pushers".

To overcome this tencency, a booster pump is used when the electric pump is mounted far from the tank AND if the fuel has to rise in height to get to it's destination. A flooded suction pump would have the same effect as a booster pump and the 914 from the way the tank is mounted and fuel lines run would qualify as flooded suction, so mounting it in the rear would not have as dramatic effect unless it were mounted higher than the tank fuel level. I will give you that the amount of flooded suction pressure will vary over the fill of the tank.

Since you want volume only, it should only be necessary to increase the fuel line and fuel rail diameter to increase volume, assuming the stock pump is capable of flowing the required volume of fuel.

And, the stock pump is not a zero pressure pump. Your instruments are incapable of measuring the pressure in a free discharge situation. The pressure side of the pump must overcome the frictional loss created the attraction of the fuel molecule and the interior wall of the the hose, pipe or whatever. No pump creates pressure until there is a frictional force resisting flow.

In any system, you will not be able to stop a pressure loss as the injector is nothing but a poppet valve. If you have static pressure in the system, opening the poppet valve will immediately release some of the pressure. Your only hope is to minimize the pressure loss.

Also, the feed of the injector (volume) is based on pressure - so many mls per minute at a given pressure. While volume will minimize the effective pressure drop, the amount of fuel injected will vary with any drop.

Perhaps the solution may be to activite a second pump on WOT by tying into the TPS.

Hey, this was my $1.00 worth.

[ray greenwood]

Scrappy, the part you are missing, is that the volume IS pressure...when pushed against a uniform restriction. The restriction is uniform...because a spring in the reulator is maintaining it. When you increase the volume, it will take a slight tweek to the adjustment to make sure enough tension is REDUCED against the increased volume....which is CAPABLE of generating more pressure...but not required to.

Water pressure is the same way...why? A standard tap at 35 psi of street pressure puts out aboyt 2.5 gpm. If too many people are using it (too many taps open...thinking like injectors open)...the pressure drops in the pipe. Why? Certainly not because the pump is putting out less pressure....but simply because it could not increase volume fast enough to keep the pipe filled at the pressure it was at. Like a fuel line....a larger pipe is NOT required to flow more volume. Simply a larger volume pump to put that line under more initial pressure ...created by that volume is all that is required.

Try it yourself. Put two pumps on a standard fuel line and let them pump into a 1 liter jar. 1 pump will do about 1 liter in 60 seconds (depending on the pump). Two...through the same line, will do 2 liters in 60 seconds. Yes! it squirts out darther...because that extra volume created more pressure squeezing itself through that line! But...no line diameter increase is needed! What will have to be done to keep the pressure from going too high...is to back the spring pressure off the regulator. But with twice the refill speed and capacity....its much easier to keep the line filled after each injection

The issue about sucker pumps is not correct. It depends on teh cavity, the number and size of the rollers...and if the pump has a pressure relief valve internally. D-jet pumps can literally suck a hole in filter media. This is because the internal pressure is bled off by the relief valve to keep flow high. L-jet pumps are totally different design. Less pressure and suction capability. Most are actually vane type pumps.

CIS pumps are a bit different still. They are roller cell high pressure with a foot or check valve to maintain the pressure stack...and keep fuel from "back bleeding". That also limits the length of the column in the pump and the roller cells being acted on.

The level/height of CIS pumps has very little to do with the need for a feeder pump. The SAAB 900 pumps are mid-tank in depth. Even at a full tank, they cannot suck. They have to have a pressurized cell because they are have no way of drawing in fuel. It has to start with some pressure to be compressed further. The compression point for the CIS pumpis actually on the other side of the check valve. The combination of the 5-10 psi of the feeder and the pressure of the roller cell is just enough to repetedly unseat and flow past the check valve to add pressure to the main line. This "unseating action" is the reason CIS pumps are so much noisier. The high pressure restriction that lets the volume build...is actually in the fuel distributor. Speaking of volume...CIS pumps have much larger cells. The VOLUME is what allows building pressure. The pump itself, without the check valves...is capable of no pressure at all, because there is no "thumb" over the end of the hose to create it.

As far as injector pressure goes, 35 PSI is just about max on D-jet injectors. After that, they revery to a stream spray. The ground shape of the pintle is not designed for that pressure. At 37-38 psi, they become undependable. The pressure tends to overcome the springs ability to re-seat the injector. They bounce...just like floating a valve.

Much over 34 psi in a D- system..and its time to have a hard look at the fuel curve and think about larger injectors.

D-jet injectors are some of the largest in the first place. The system simply has less cycle time available to adjust with compared to other systems. Increasing fuel pressure too high, places the system in a position of not being able to adjust properly in the lower rpm ranges and at idle, because the cycle rate does not go low enough to overcome the added fixed pressure. You get a rich idle and flat spots if you are not careful,. Ray

[Dave at Pelican Parts]

A lot of this discussion is over my head, but I do have one comment to make.

Originally posted by ray greenwood
....L-jet pumps are totally different design [than D-jet pumps].

Not on the 914--they have the same part number. In 75, both D-jet and L-jet switched from the three-port pump (S, D, R) to the two port pump (S, D). I don't know about later L-jet Volksies for certain, but I suspect that they used that same type of pump at least for a while.

--DD

[Scrappy]

Scrappy, the part you are missing, is that the volume IS pressure...when pushed against a uniform restriction. The restriction is uniform...because a spring in the reulator is maintaining it. When you increase the volume, it will take a slight tweek to the adjustment to make sure enough tension is REDUCED against the increased volume....which is CAPABLE of generating more pressure...but not required to.

Ok, what you're trying to tell me here is that if I have a pump capable of 35 psig and regulate it at 30 psig with a spring (relief valve) and say a flow rate of 60 gph is:

"if I fill a 5 gallon container with this system and fill a 30 gallon container with the same system, the pressure is going to be different due to the increased volume of liquid available"

This I find hard to swallow, other than where you measure the pressure and accounting for the weight of the liquid if pressure is measured at the bottom of the container.

Now, if you are telling me that running two pumps into the same circuit or container, with pressure regulated at 30 psig by a single valve, and the increased volume is now 120 gph and will now elevate the pressure by volume alone - I find that hard to swallow also as the pressure in the container is still regulated at 30 psig.

I will agree that if a pressure drop is seen due to the opening of a valve, recovery would be much faster with two pumps rated at 60 gph rather than one rated 60 gph. Of course then, we could always use one pump rated at 120 gph.

And when I was speaking of "pushers and suckers", the suckers were mechanical diaphram type (positive displacement). Roller electric types are great because of the seal the roller achieves against the body. Vane types are ok, but cavitate to easily and the "poor" quality of the seal leads to fluid bypass.

Gawd, I love this. Ray makes me think and I can't even spell engineer.

[M. Hendrix]

HAH! My pump doesn't weigh 30lbs! You are all stupid!


M

[mike95125]

special olimpics

[Bleyseng]

Much easier to get a larger volume pump, thats what we do for turbo motors. You also have to increase the feed tube to satify the new pump. Also new size the rest of the injector feeds to a larger size.

Hmm, maybe some math wizard can figure out the sizing the volume capablity of the stock system.
Maybe Porsche did that 35 years ago, dunno.

I think a slightly larger pump is the simple solution here....

Geoff

[ray greenwood]

Bleyseng, yes...you are correct that the very best way is with a larger VOLUME pump. But, in stock engines a larger injector FEED line is just not necsessary. Yes...if the pump gets large enough...you will need a larger SUPPLY line from the tank to keep from starving it, but if you increase the feed lines in diameter, the larger mass/volume of fuel that must be moved...defeats the purpose of the larger pump. Also, since the stock fuel supply lines already supply more fuel volume than any four injectors on the planet could possibly use...hence the need for a return line to carry away the excess, don't bother the feed lines.

Scrappy...you missed it again. Pumps are not rated by PSI.! They are rated by VOLUME. Logic is not going to solve this for you. Get a pump , a gauge and a graduated cylinder and discover this for yourself.

If your line volume is 1 liter...and the pump is rated at one liter per minute, it will take one minute to fill the line. The restriction that creates pressure...is NOT part of the pump. It is the regulator. If another pump is rated at 2 liters per minute...it fills the line twice as fast...but will be running the same pressure...because it has the same restriction. It uses the same regulator. It has 30 psi of spring against the end of the hose. So the higher volume...will push harder against the plate in the regulator...moving it further out of the way...relieving more of that excess pressure potential from the higher volume.....UNTIL....the space around the metering plate in the regulator is large enough that the regulator plate quits moving away...because there is equilibrium. Guess what...you get the same pressure. But...since there is twice the MAKE-UP volume available....its able to replenish the flow twice as fast...so that metering plate and plunger do not have to move in and out as much or as far, to keep the pressure on the fuel stream constant (the movement of the diaphram is what causes the needle fluctuation...and the fuel pressure fluctuation).

Think of this...when you go to adjust the fuel pressure, when is the last time you had to adjust the pump? Never. Its about volume. Pressure is created by volume against the restriction. Pressure is maintained by volume from the pump. Not enough volume.....pressure is hard to maintain. again, the point you missed...is that we are NOT trying to increase pressure. Thats easy to do on any pump. We are trying to maintain the pressure we have without flucuation. If the pump pushing against the diaphram in the regulator....is what causes pressure on the line (and it is)....then if the needle on the gausge moves....you don't have enough fuel pushing against the plate. (uh....that would be volume).

You really should gets some pumps and test them. I already know this works. I use it every day. Engineering and the laws of physics also support it. Its not a theory...and if it were...its not mine anyway. Its a function that has been corrected in most modern fuel injection systems....by having pumps properly sized in volume,to the systems. I know how well it makes my cars run....and Im happy I have rock solid fuel pressure. It opened up an entirely new level of tuning with D-jet a while back. What you do with your system is your own business. Please... Don't take it from me...get out there and do the testing to prove it like I have. You will be happy when you have learned what I have...and applied it.

One quick note, the D-jet pumps are fully capable of over 70 PSI of pressure. Some of the early ones can go almost 100 psi. Of course they cannot supply the volume to keep a system at that pressure. How do we know this? Try installing one on a CIS system. Once the car starts...if it starts.....the guage drops to about 50 psi on the main pressure. It cannot keep up the volume to maintain pressure. Pressure is a FUNCTION of volume. If you have variable pressure...you have variable volume.

Dave, you are correct...I forgot that they changed to the L-jet style pump in the later years. But they are different type pumps internally, (the L and original D). They have very different tolerances. Ray