You dont have javascript enabled! Please enable it!

Duty cycle and PWM control

Subjects:

  • General
  • Measuring on a duty cycle
  • Duty cycle with a plus circuit
  • Duty cycle with a ground connection
  • Duty cycle measured from power supply
  • Troubleshooting the PWM Controlled Fuel Pressure Regulator

General:
With a duty-cycle circuit, the current intensity can be regulated by a consumer. The current intensity can be regulated without loss of power, as is the case with a series resistor. In automotive technology, the duty cycle can be used, among other things, to regulate the speed of the heater fan, the position of the throttle valve position motor, for example, or to keep lights on.

When applying a duty cycle to a lamp, it can be ensured that the lamp burns less brightly. This is applied, among other things, to taillights, where one lamp can burn at two different strengths, namely for normal lighting and the brake light. In normal lighting, the lamp burns weakly (here a duty cycle is applied to limit the current through the lamp). At the brake light, the lamp will change the duty cycle so that the lamp burns brighter.

The image shows a rear light of a BMW 5-series, where the left lamp of the rear light also functions as a brake light by making it burn more brightly.

Measuring on a duty cycle:
The duty cycle can be measured with an oscilloscope. The oscilloscope will graphically display the voltage trend over time.

When a duty cycle is measured with a multimeter, the correct voltage value will never be displayed. Since the voltage varies constantly with a duty cycle, the multimeter will indicate the average voltage because it is too slow.

Duty cycle with a plus circuit:
In the image below is a waterfall diagram with the plus of the battery (12 volts) at the top, followed by the fuse, the ECU (the electronic switch), the consumer (in this case a lamp) and finally the ground. The ECU constantly switches the power supply on and off.
The oscilloscope measures the voltage between the plus of the lamp and the ground of the vehicle. The oscilloscope settings are as follows: 2 volts per division and 5 milliseconds per division. That means that each square from bottom to top is 2 volts, so when the squares of the ascending line are added up (6 pieces in total), the highest voltage measured is 12 volts.
The duration is from left to right. Each box (division) is set to 5 milliseconds. Looking from left to right, the line is 10 milliseconds high and 10 milliseconds low.

Just like with the multimeter, the oscilloscope measures the voltage difference between the positive and negative cables connected to the meter. If the lamp is switched on in the diagram below, the positive cable has a voltage of 12 volts and the negative cable (always) 0 volts because it is connected to ground. The difference between them is indicated by the meter; the difference between 12 volts and 0 volts is 12 volts. This 12 volt is displayed in the meter's display. When the duty cycle is high, the lamp is on. This is not the case with a ground connection. This is explained in the next section.

To determine the duty cycle, it is important to know what 1 period means. In a period, the voltage is once high and once low. After this period, the next period starts. In the scope image below, 1 period is marked in blue. This shows that the period lasts a total of 20 milliseconds, namely 10 ms high and 10 ms low. So it can be read that half of the time the voltage is high, and the other half is low. The duty cycle in this scope image is therefore 50%. In this case, the lamp burns weakly.

In the image below, the period has remained the same (20 ms), but in this case the voltage is only high for a quarter of the time (5 ms) and low for three quarters of the time (15 ms). In this measurement the duty cycle is 25%. This means that the lamp now burns even weaker than with the 50% duty cycle, because the lamp is only supplied with power for a quarter of the total period.

Duty cycle with a ground connection:
In automotive technology, ground circuits are usually used. With a mass-switched consumer, the duty cycle will be reversed compared to a positive circuit. An example of this can be seen in the image below.
When the lamp is switched off, the ECU has interrupted the connection to ground. This means that the circuit is interrupted. In that case, the voltage of 12 volts is applied to the input of the ECU. This means that this voltage is also on the negative terminal of the lamp. In this case, the voltage difference when the lamp is switched off is 12 volts.

As soon as the ECU switches the lamp to ground, the lamp will light up. A current flows from plus to minus. The lamp uses the 12 volts to burn, so there is 0 volts on the negative terminal of the lamp. In that case there is 0 volts on the positive cable and 0 volts on the negative cable. The voltage difference is then 0 volts. This means that at 0 volts the lamp is switched on and at 12 volts the lamp is switched off.

To make the lamp burn more weakly, the time in which the lamp is supplied with power must be shortened. This can be seen in the image below. In one period, the voltage is 15 ms high (lamp is off) and 5 ms low (lamp is on). In this case, the lamp is only switched on for a quarter of the period, so that it will burn more weakly.

Duty cycle measured from power supply:
The foregoing measurements have all been made with respect to the mass of the vehicle. Another possibility is to measure from the plus of the battery to the mass of the consumer, as shown in the figure below.

The lamp will light up as soon as the ECU has switched the ground. In that case, the supply voltage of 12 volts is consumed by the lamp to light up. So there will be a voltage of 0 volts on the negative lead of the oscilloscope. The positive cable has a voltage of 12 volts. In that case, there is a voltage difference of 12 volts between the test leads, so the 12 volt line in the display will indicate that the lamp is switched on. So this is 25% of the period.

As soon as the ECU breaks the connection to ground, the voltage of 12 volts will also be on the negative side of the lamp. The voltage difference between the measuring leads of the oscilloscope will then be 0 volts. 0 volts will then be displayed on the screen when the lamp is switched off.

Troubleshooting the PWM Controlled Fuel Pressure Regulator:
On the page ECU circuit of a PWM valve explains what the circuit in the ECU of a PWM controlled rail pressure regulator looks like. That is why it is recommended to read the information on that page first.

The rail pressure regulator on the high pressure rail of the common rail diesel engine is by the engine control unit controlled by a PWM (Pulse Width Modulation).
At rest, the valve in the pressure regulator is open, allowing the fuel pressure to exit the high pressure rail through the return. The valve closes the moment it is actuated. The pressure in the rail is rising. When the rail pressure sensor registers a (too) high pressure, the ECU adjusts the PWM signal.

The figure below shows the schematic of the engine control unit (J623) and the rail pressure regulator (N276). The rail pressure regulator is supplied on pin 2 with a voltage between 13 and 14,6 volts (depending on the charging voltage with the engine running). The ECU puts pin 45 to ground when the valve needs to be actuated. A current will flow through the coil of N276 as soon as pin 45 is connected to ground. The pressure in the common rail is rising. The moment the ECU interrupts the connection between pin 45 and ground, the pressure build-up in the fuel rail stops. The spring in the pressure regulator opens the valve slightly again, allowing the fuel to rush back to the tank through the return lines.

The scope image shows a supply voltage (blue) and the PWM control (red). The supply voltage is around 13,5 volts and is constant.
The voltage of the PWM control signal (red) is between 0 and 13,5 volts. This scope image shows that the valve is constantly being turned on and off. 
The current (green) increases when the valve is energized and decreases after switching off.

At rest, the voltage is 13,5 volts. The PWM valve is not actuated. 
The spring located in the valve ensures that the valve is open at rest. 
The moment the ECU switches on the ground (this can be seen in the scope image when the red signal is 0 volts), a current will flow through the coil (the green image), causing the valve to close.

The scope image shows that the valve is switched on for a short time and switched off for a longer time. This means that the fuel pressure must be relatively low.

We read out the car and view the live data. The fuel pressure is almost 300 bar at idling speed. This is OK.

Fault: engine no longer starts when starting.
The engine does not start during starting. We are sure that there is sufficient fuel in the tank. We will of course start with reading out faults. In this case, no faults are stored. That's why we look at the live data (in VCDS these are called the measurement value blocks). During starting, the starting speed is 231 rpm. The ECU receives the crankshaft signal. That is good.
The fuel pressure during starting is 7.1 bar. That is too low to start the engine.

Low fuel pressure can have the following causes:

  • too little fuel in the tank
  • fuel pump (lift pump or high-pressure pump) defective
  • clogged fuel filter
  • faulty fuel pressure control valve

To determine why the fuel pressure remains too low, we check the voltages of the electrical components with the oscilloscope.
Earlier in this section, the scope image of the properly functioning PWM fuel pressure regulator is shown. The next scope image is again a measurement of this pressure regulator, but this time with a malfunction. 

As the current increases, the supply voltage decreases. The supply voltage therefore drops when current flows. In addition, the following points stand out:

  • When switched on, the supply voltage drops to a lower value, normally a contact resistance causes an abrupt drop (a vertical line in the scope image to a lower voltage);
  • The current build-up follows the characteristic charging curve according to the e-power after switching on the coil. The current course during discharge is mirrored by the gradual build-up of the supply voltage. The current does not drop to 0 A. So current continues to flow after the control has ended.
  • As soon as the coil is switched off, there is no induction peak in the red image (where the voltage rises from 0 to 14 volts). Think of switching off the injector coil, which can cause a peak of up to 60 volts.

So there is a contact resistance in the supply wire to the fuel pressure regulator. Only when current flows does a voltage drop occur as a result of the contact resistance. When the ground is switched off, no current flows and the supply voltage remains the same as the battery voltage.

Now back to the schematic: the power wire is circled in red. The next step is actually locating the damaged wire. Damage can occur as a result of chafing against engine parts, or because the wire has been pinched during previous assembly work. Once the damage has been found, it can be repaired.

It is now clear what resulted in the transition resistance. You may have noticed that there has been talk of a missing inductance peak in the scope signal. When the coil is switched off, the current image slowly decreases to a lower value. So there is no interruption of the control; it is terminated, but current continues to flow through the coil.

The moment the FET has been made conductive by the microprocessor, a current can flow from the drain to source and thus also through the coil. The coil is thus energized and the control valve can close against the spring force due to the generated magnetic field.

As soon as the driving of the FET ends, no more current flows through the coil to ground. The freewheeling diode ensures that the induction current is fed to the plus as a result of the residual energy in the coil. This ensures a gradual flow reduction and prevents an induction from taking place. This process is indicated in the figure by the red arrows.

This explains why a current course is still visible in the scope image after the actuation has already ended.