compare them to the conventional return-style

POINT OF NO RETURN There s a lot more to returnless fuel systems than just moving the pressure regulator from the rail to the tank. Here s what you ll need to know to service these systems as they become more and more prominent. BY MARK WARREN Tightening federal evaporative emissions requirements are driving dramatic changes in fuel delivery systems. One of the main problems is the soaring temperature of the fuel in the rails due to high underhood temperatures. Most of this fuel makes its way back to the tank. In-tank fuel temperatures in summer can be greater than 160 F today, and are a major contributor to excess evaporative emissions. As a result, Chrysler has gone to returnless fuel systems on all its 1998 vehicles, with most manufacturers introducing similar setups on at least some of their models this year. Estimates are that by the year 2000, all cars and lightduty trucks will be using these so-called returnless, or in-tank return, systems. To help you understand the changes that these new fuel systems bring, let s compare them to the conventional return-style fuel delivery system. Standard passenger-car fuel systems pump about 30 gallons per hour (gph), or a half-gallon per minute. A car traveling 60 mph that gets 30 miles to the gallon, therefore, consumes about 2 gph. That means 28 gph are filtered and returned to the tank. On conventional systems, the fuel is pulled through a screen sock, pressurized in the pump and then routed through a filter. From there it goes into the fuel rail, past the injectors and through the pressure regulator, then is returned to the tank. The screen sock is designed to filter out large contaminants that would quickly damage the pump assembly. The screen holes are large enough to keep the screen from getting plugged easily. The downside is that some smaller particles have to travel through the pump. Photo: Bob Savasta 36 July 1998

July 1998 37

In designing returnless systems, engineers have three filtration options: They can filter the fuel before the pump (in-tank), filter it after the pump at the pressure regulator return (intank) or filter the fuel outside the tank. Let s look at the last option first. Fig. 1 4 Screen Sock Fig. 2 3 Fuel Reservoir 2 5 1 6 Fuel Level Sensor Fuel Pressure Regulator 1 Floated Jet Pump Assembly 2 Secondary Umbrella Valve 3 Primary Umbrella Valve 4 Rollervane Fuel Pump 5 Fuel Directly From Pump 6 Venturi Area Fuel Flow To & From Engine Aspirated/Recirculated Reservoir Fuel Illustrations: Harold A. Perry With the filter outside the tank, service obviously is much easier. However, using this method means that the fuel returning to the tank is never filtered, except by the screen sock. Fig. 1 shows this option. Note the in-tank pressure regulator. With the filter outside the tank, the pump can suck in rust from bad tanks at filling stations. Although some of this rust goes to the filter, most is returned to the tank (the ratio of returned vs. consumed fuel is about 15:1 under normal operating conditions). This rust is then ground up by the pump and returned to the tank again, along with ground up parts from the pump itself. This cycle repeats itself, getting worse each time. I call it the fuel pump grinder effect. In areas with poor-quality fuel (fuel that contains particulate matter), some of these pumps aren t lasting 8000 miles. Factory repair procedures call for removing and disposing of the fuel (in an appropriate manner) and steamcleaning the tank. Many dealer techs aren t following the procedures, causing rapid new-pump failure and increasing warranty costs. As a result, more than one manufacturer is considering offering pumps only as a complete tank assembly. Every fuel pump replacement, returnless system or not, should include a tank inspection, cleaning and new filter sock. The first option mentioned filtering the fuel before the pump looks like the best once you re past the hassle of in-tank filter service. However, when this filter gets plugged, the pump has to work hard to suck through it. This creates low pressure (suction) that causes the fuel to boil and the pump to cavitate. The pump is also starved for its coolant and lubricant (fuel). Many manufacturers that picked this option are using a large lifetime filter to combat the problem. The second option filtering at the pressure regulator return is probably the best. The fuel is picked up through a standard screen sock, with a replaceable filter built into the regulator assembly on top of the pump. Using this method, all the fuel in the tank is continually filtered. Many manufacturers are installing fuel pump access panels to make service faster and easier. Prior to returnless systems, OEMs went to great lengths to keep the fuel in the tank cool. Fig. 2 is a typical fuel pump module off a conventional return-style system. Note its complexity and the extent to which the manufac- 38 July 1998

turers went to keep hot returning fuel from coming into contact with the bulk fuel in the tank. The fuel pump is inside a sealed housing. In addition to isolating the hot fuel, this works as a noise barrier and keeps the car from misfiring under low-fuel/sloshing conditions. Look at points 5 and 6 in Fig. 2. This is the real key to the operation of these systems. High-pressure/volume fuel is taken from the pump at point 5 and directed right at point 6, a small venturi. Remember, when fuel rushes over the venturi, a local low-pressure area is created. This draws fuel from under the float (point 1) when there s adequate fuel in the tank or from the secondary umbrella valve (point 2) when low-fuel conditions close the float valve. This is where the term floated jet valve assembly comes from. Point 3 is called the primary umbrella valve here. This valve is stronger and is used as backup in the event the secondary umbrella valve fails closed. Also notice the extra sock filter at the base of the pump, which serves to filter and exclude, or break up, bubbles in the fuel. Photo: Larry Gouge Vaporization & Cavitation Fuel vaporization and cavitation are major problems in fuel delivery systems. Never actually seen them? Me neither, until I got the Fuel System Analyzer from Emi-Tech, a 1997 MOTOR Top 20 Tools award winner. This tool measures fuel pressure, vacuum and, most importantly, fuel flow and quality. I define quality as the absence of water or vapor (bubbles) in the fuel. This can be easily observed in the tubed sight glass. Good fuel volume is critical to maintaining fuel pressure. Remember, we calculated that most fuel systems deliver 15 times more fuel than is consumed by the engine under normal operating conditions. Now let s assume we re operating under a killer load at, say, 6 miles per gallon and we re doing 60 mph. Here we re consuming 10 gallons of fuel per hour. Using 10 gph with a pump that can deliver 30 gives us a 3:1 ratio. As long as it can deliver enough volume, the pressure will be okay. In testing for flow with the Fuel System Analyzer, I ve found that normal fuel flow should be.3 to.7 gph. Anything less than.3 indicates a flow problem that needs to be addressed. Fuel flow restriction often can be caused by a plugged filter, although pinpoint testing may be necessary to confirm the problem area. All of us have strapped a fuel pressure gauge onto a windshield and gone for a killer loaded test drive to look for a pressure drop. But this method works only when the pump has dropped to less than a 1:1 ratio of volume vs. demand. Usually, the fuel pump will suffer a major performance loss long before the pressure drops. I ve seen pumps with bad bearings take a full two hours of run time to degrade to the volume/pressure failure threshold. Now that s a long test drive! Look closely at the gauge readings on the Emi-Tech analyzer in the photo (left). This system has low vacuum, normal fuel pressure (for this vacuum) and almost no fuel flow in the flow meter. This car is under load (low vacuum) and the flow is at the edge of the failure threshold. Using just a pressure 40 July 1998

gauge, this diagnosis would have been missed; using the flow gauge, the imminent failure is obvious. Just as in electrical testing, where voltage is only half the power equation, the same is true for pressure testing in hydraulic systems. The high failure rate of fuel pumps makes this an important issue. Fig. 3 Fig. 4 C A D B D Pressure Regulation Most fuel pressure regulators on return-style fuel systems have a vacuum diaphragm-controlled regulator. When engine vacuum drops (high load), the fuel pressure gets bumped up to deliver more fuel. These are called constant flow systems. Here s C B A Waveforms: Mark Warren what I mean by constant flow: A total vacuum is about 29.9 in./hg at sea level, while atmospheric pressure is about 14.7 psi at sea level. So 2 in./hg is about equal to 1 psi when we re talking atmospheric pressure. Now let s not get trapped into that does-the-engine-suck-in-air-or-is-itpushed-in-by-atmospheric-pressure argument. It all depends on your point of reference. If you set the baseline at sea level, then the engine sucks; if you reference relative to zero atmosphere, then we re under 14.7 psi all the time. Note that your pressure gauge shows 0 psi even though atmospheric pressure is actually 14.7 psi. This is why most gauges show psig (psi gauge) on the face. Stay with me here. If an engine is sucking at 20 in./hg, this is the equivalent of 10 psi acting on the intake side of the injector, right? So, if the fuel pressure is 30 psi and the intake is sucking at 10 psi, then the effective pressure driving the fuel into the intake is 40 psi. Still with me? Now suppose I jam the throttle wide open. Vacuum drops to zero and the intake goes to atmospheric pressure (a 10-psi increase). Without any change in regulated pressure, the effective fuel pressure would be 30 psi. Notice how all vacuum-controlled pressure regulators change about 10-psi full range. Look at Fig. 3. These waveforms show the relationship between intake vacuum and fuel pressure. When vacuum is high (point A), fuel pressure is running at about 36 psi (point B). At point C we put the engine under heavy load. Vacuum drops to zero, while fuel pressure rises to about 45 psi (point D). This is why these systems are called constant volume systems. With changes to intake vacuum, the fuel pressure is compensated for constant flow. I don t recommend disconnecting the vacuum source to the regulator to jack up fuel pressure to test for leanrunning conditions. Usually the engine management system will compensate for the excess fuel. Then when you accelerate, the engine will stumble because of the lack of fuel pressure enrichment. The fuel pressure regulators in returnless systems have no provision for 42 July 1998

constant volume delivery relative to load/intake pressure changes. These systems require rapid injector pulse width response and longer injector open times. Faster processors and higher average fuel pressure make these compensations possible. Higher fuel pressures are also necessary in returnless systems to prevent the fuel from boiling in the hot injector rail. The downside of higher fuel pressure is smaller injector openings (which are more sensitive to deposits) and harder-to-control idle fuel delivery. Fuel pressure and injector flow are hard to balance with the conflicting demands of tight idle control vs. fullload/high-flow demand. Other Diagnostic Methods Many diagnostic tool suppliers sell fuel pressure transducers. The transducer is connected to the fuel pressure test port, then the wiring is run from underhood to your DVOM, graphing meter or scope. Long record times offer the advantage of not having to watch a gauge while driving. You can record the pressure, then analyze the data on your return to the shop. An intermittent fuel pressure problem can be detected easier this way. Also, testing for fuel pressure decay (leaking injectors or a bad check valve in the pump assembly) can be done without your having to babysit your meter. Fig. 4 on page 42 is a high-resolution waveform of fuel pressure taken with Edge Diagnostic s Engine-Tech and a pressure transducer. Right after point A is the pressure drop in the fuel rail relative to the injector opening; point B is the injector closing. You might guess that the valley between A and B would be the injector closing and the fuel pressure recovering, but the injector waveform correlated perfectly to A and B. This test is used to approximate injector flow and compare cylinders. You can see that the injectors for the three cylinders shown here all flow about the same. This was confirmed by a good cylinder balance test. To be valid, this test must be run on an engine with sequential fuel injection. Testing on an engine that uses gangfired injectors would show an imbalance only from bank to bank. Fuel rail pulsations caused by the opening and closing of injectors combined with pulses from the fuel pump itself are often dampened by pressure dampeners on the pump or fuel rail. GM puts them on the pump; Toyota prefers the rail. Be aware that GM had a high failure rate on its early dampeners, so an improved unit was designed. I m wondering how many older GM cars are suffering driveability problems from these bad dampeners, and aren t being diagnosed properly. I sure haven t fixed many. Any major pulsation in a gauge should be cause for further investigation. Refer back to Fig. 4. Note that the pulsations in the waveform are almost 44 July 1998

1 psi peak (point C) to valley (point D). I thought this was the regulator s response to the vacuum pulses in the intake manifold, but disconnecting the manifold vacuum source from the regulator and substituting a vacuum pump didn t change anything. This may be normal harmonic pulsation or a reflection of the pump characteristics. I don t know; I m going to have to compare more systems to see. Current Testing Revisited I covered fuel pump testing with a current probe briefly in the March issue, but I have just a few more thoughts here. If you have a dropout on one winding, then the pump is bad. If the amperage is low, the problem could be low fuel, low fuel pressure (bad regulator) or a few other possibilities. If the amperage is high, you may have high fuel pressure (bad regulator), a plugged fuel filter or some problem other than the pump. Current testing can be tricky if you don t consider all the possibilities! One brave tech on iatn (International Automotive Technicians Network on the web) had the guts to admit he replaced a fuel pump drawing too little current, only to have the same problem with the new pump. Further testing revealed no fuel in the tank. Ouch! Remember how I said to check the tank for rust and crud before dropping in a new pump? Well, doing that here would have revealed the problem a little sooner. An eyeball on the gas gauge would have saved the job even sooner than that. I applaud this tech for having the guts to share his story, to save us all from the same fate. We all suffer from brain fade from time to time. Considerations for Returnless Failures We have yet to experience the in-use age deterioration of returnless systems; they simply haven t been around that long. However, I believe we re going to see new diagnostic problems. The injector that s now at the end of the line rather than in the loop will probably experience greater filter screen contamination as a result (use a pressuredrop test). To date I have seen no vapor lock or fuel rail boiling/hard-restart problems. As long as there are no major changes to fuel Reid Vapor Pressures, we probably will be in pretty good shape there. Returnless systems really are a simplification over fuel systems we ve dealt with in the past. So be sure to explain how wonderful this new system is when you hand your customer the bill for his first in-tank filter replacement. For a free copy of this article, write to: Fulfillment Dept., MOTOR Magazine, 5600 Crooks Rd., Troy, MI 48098. Additional copies are $2 each. Send check or money order. July 1998 45