Auto Service World
Feature   December 1, 2004   by Jim Anderton

Without a Trace

You can diagnose fuel supply problems without 'scoping the system, but why?

If you’re beyond a certain age, you probably remember when getting the fuel to the engine was a simple matter of sucking it out of the tank, through a mechanical diaphragm pump and up to the carb. Diagnosis was possible with nothing more sophisticated than a Mason jar and a foot of Neoprene tubing. Not so today. Fuel injection has demanded higher system pressures, which combined with the automaker’s drive to reduce costs, has moved the now electric pumps into one of the few places where techs can’t get at it easily: the fuel tank. Pressure and volume are still important, but a few checks at the front of the vehicle can save lots of time and frustration later.

No-start/hard start strategies

The no-start/hard start issue is common enough, but what if OBD II can’t give you clear direction? Getting adequate current to the fuel pump is essential and directing that current is the ECU, acting through the fuel pump relay. Why use a relay? Pump currents are too much for the ECU, so the relay is a cheap way to turn tiny control currents into serious pump power. They all work in the same way. The control current passes through a solenoid type coil, which magnetically pulls in a switch contact, passing higher current to the pump. There are lots of relay controlled devices in modern vehicles, some responding to currents delivered by modules, while others are always “hot”, relying on the module to provide a ground to complete the relay’s coil circuit. The key to thinking about fuel pump relays as a potential trouble source is to avoid thinking about them as “on-off” devices. Time is the key. In ‘real’ time, they turn on and off like a switch. If you measure time in milliseconds, however, the inrush of current into the relay coil can tell an interesting story. Diagram 1 shows a typical ‘scope trace of the inrush current into a relay coil as the ignition key is turned to the “on” position. If you chose a ‘scope time base in the millisecond range (experiment for the best resolution) then you should be able to see the relay contacts closing as a small blip in the coil’s inrush current. Since the relay has to hold during pump operation, it should stabilize somewhere near battery voltage.

That’s fine, unless the ECU, as it does in most vehicles, has some safety or anti-flooding code written into its software. The idea is to prevent the pump from operating if, after a certain time, the engine isn’t operating, preventing it from feeding a post accident fire, or flooding an engine that just won’t start. To see this function, watch the relay coil trace during cranking (no-start condition). You’ll need a time base that’s longer than the millisecond-range needed to see the coil inrush current. Most vehicles will give the engine several seconds of cranking before they shut the pump down, so make sure your time base accommodates it. Diagram Two shows a typical case. The trace starts with full relay current, which dips as the starter loads the electrical system. After the predetermined time, the relay current is commanded “off” by the ECU and there’s no point cranking any further. The MIL should be lit by now, and if so, it’s probably a good idea to think about systems and sensors that can tell the ECU whether or not the engine has started. For example, the crank angle sensor is under suspicion, while it’s unlikely that the throttle position sensor is at issue. Trouble codes help, but a no-start can confuse the system, especially if there are multiple sensor issues. It’s unlikely that two or more sensors will fail at once, but it can happen if the customer drives for weeks or months with the MIL glowing and the ECU in limp mode, until the engine fails completely. Makes sure to ask if the light was not just on, but for how long.

Diagram Three extends the trace into engine start. Although the traces shown here are cleaned-up representations of a fictional generic vehicle, during cranking, engine RPM is low enough for the electrical system loading from individual injector firing to be visible in the fuel pump relay coil current. Once the engine fires, of course, the alternator will feed the system, and coil voltage will rise to above battery cranking voltage, but during that critical period during cranking, the trace dips during injector cycling give another clue about what’s going on under the hood.

‘Scoping the pump

The fuel pump feed is solid, fine, but the engine is still showing drivability issues. You can hear the pump whirring in the tank, so the wiring between relay and tank is probably fine. A pressure check is a logical approach, as well as a visual inspection of the lines, but if you have a current clamp for your ‘scope, you can connect to the pump feed wire and learn a lot about the pump’s condition. How? The DC motor that drives the pump impeller uses a commutator that has multiple segments, giving a saw tooth wave pattern on the ‘scope as the brushes sweep past each conductor/insulator segment. If you look carefully at the trace, choosing a time base that gives a long pattern, there should be a repeating cycle visible, probably as a subtle difference in one of the peaks. Count the number of ‘scope time base divisions between the two slightly different peaks, and you can calculate the time it takes for one revolution of the pump commutator. A little math, and you can determine the pump’s operating RPM. And the current draw itself is an important parameter. A pump that’s drawing only a couple of amps might function well enough for idle and low speed operation, then create drivability issues like lean-induced misfire at higher speeds. Note that on many newer models, there’s a pump driver module that may control pump RPM. To isolate this module, it will be necessary to either command it to go to full speed, or jumper the module to bring full battery voltage directly to the pump feed. If you don’t bypass the module, keep in mind that some toggle the pump’s ground path on and off at high frequency as a control strategy, making it difficult or impossible to check pump speed. Of course, that ground path cycling is itself worth looking at if you’re checking module performance. For module-controlled pumps, a quick look at a wiring or block diagram before testing is a must.

There’s a whole textbook full of concepts and ideas on ‘scope testing of fuel pump electrics, and as the vehicle fleet moves toward sophisticated control of pump speed, they will become more and more model-specific. For many cars and light trucks, however, these simple tests can narrow the troubleshooting task down significantly, saving you time and your customer money.

Print this page


Have your say:

Your email address will not be published. Required fields are marked *