Last week in How to Electronically Modify Your Car, Part 5 we looked at how a cheap resistor or pot can be used to trick the engine management system into running more advanced ignition timing. This week we’re going to look in more detail at how pots can be used.
Resistors vs Pots
As we have already described in this series, a pot can be wired as a variable resistor. In this circuit diagram, when the pot’s wiper (indicated by the arrow-head) is at the bottom, no resistance is introduced into the circuit. When the wiper is at the top, maximum resistance is introduced into the circuit. If the pot had the right values of resistance and power dissipation, you could vary the brightness of the light by changing pot position.
A pot wired as a variable resistor is good when you want to add resistance to a sensor circuit. (The example we used last week was of an intake air temp sensor, where the resistance went up as intake air temp went down. Adding a resistance meant that the ECU thought the intake air temp was colder than it really was).
But that’s all a variable resistor can do – add resistance, or, if the wiring is done differently (ie in parallel), subtract resistance.
But most sensors in cars don’t use a variable resistance designs. Instead, they have inbuilt electronics that causes them to output a varying voltage. For example, most airflow meters, MAP (manifold absolute pressure) sensors, accelerometers, yaw sensors and throttle position sensors have an output voltage that varies across the range of about 1-5 volts. If you just insert a series variable resistor in their circuit, things are liable not to work in the way you might expect!*
But by using a pot in a different way, very good modification results can be gained with voltage outputting sensors. Before we look at how to do it, let’s examine the type of sensors that we’ll probably be applying this modification to.
(*Strictly speaking, the variable resistance sensors are actually variable voltage outputting sensors, but I am trying to keep things as simple as possible!)
Many voltage-outputting sensors have three connections. Here is a diagram of a MAP sensor. One wire (here it is red) is a regulated 5V supply from the ECU. (Regulated means it is held at a fixed voltage, irrespective of battery voltage.) There is also a ground wire (black) that somewhere is connected to the car’s body – probably back at the ECU. Finally, there is another wire (green here) that is the signal output.
If you connect a multimeter to the signal and ground wires, and then drive the car, you’ll be able to see on the multimeter how the signal varies in different driving conditions. For example, you might find that the lowest reading is 0.8V and the highest is 4.6V. You might also be able to see that the highest occurs at full throttle and the lowest occurs on the engine over-run.
The ECU uses this signal to detect what is occurring – in this case, what the intake manifold pressure is at any given moment.
OK, let’s go back to the MAP sensor. (Remember that the MAP sensor is connected to the ECU, although I haven’t bothered showing all the wiring.) The highest that the MAP sensor output can go is 5V, and the lowest it can go is 0V. But this sort of sensor is rarely holding a steady signal output – it’s up and down as the driver moves their foot.
So if we wanted the ECU to see a higher voltage coming from this sensor, we can’t just cut off the sensor wire and connect it to 5V, as shown here. Sure, that would mean that the signal voltage the ECU sees from the sensor is higher – but it’s also now fixed, not varying with engine conditions. For the same reason, we can’t just connect the signal wire to ground.
If we want to raise the sensor output voltage (so for example, the ECU thinks that manifold pressure is higher than it really is), we need to add a bit of voltage to the signal. So, how do we do that?
What we have done here is wire a 10 kilo-ohm pot between the signal wire and the 5V wire. Note how the connections have been made to each end of the pot’s resistance track. (This might look like we’re connecting the signal wire straight to 5V, but we’re not – the resistance of the pot is so high that almost no current flows through this connection.)
Let’s put a multimeter in the circuit, measuring the voltage on the wiper of the pot. (Note how the other probe of the multimeter still goes to ground - voltages are almost always measured with respect to ground.) The pot has been set so that its wiper is at the top – close to the 5V supply. This is the same as connecting the multimeter probe to the 5V supply, so the meter reads 5.0V.
No we’ve moved the pot wiper so that it’s down the bottom – effectively connecting it straight to the signal wire. Since the signal is at 1.6V, that’s also what the meter reads.
But now we’ve moved the wiper up just a bit. The multimeter is reading a combined signal – the signal voltage plus the little voltage we’ve added to it. So the meter reads 3.3V – we’ve added 1.7 volts to the signal.
There’s just one step left. If we connect the ECU signal wire to the pot wiper rather than to the MAP sensor, we can dial-in any addition to the sensor signal that we want. The signal will still rise and fall as it did before, but with an additional voltage on top. By changing the position of the pot, we can change how much voltage we add to the signal. (See later in this story for the way that this addition actually occurs.)
If we want to subtract a voltage, we just connect the pot between ground and the sensor signal, like this.
This modification is a very powerful one that can be applied in many different circumstances. We’ll show in detail a modification in a moment, but the first way I ever used this mod remains perhaps the most impressive.
I had a small 3-cyliner Daihatsu Mira turbo (660cc!) to which I’d fitted a bigger turbo and water/air intercooling. The injectors were running out of flow capability so I fitted larger injectors from a Charade GTti. But the ECU didn’t know that larger injectors were being used, so over-fuelled the car.
The ECU uses as its main load input a MAP sensor, much like the one described above. By using a pot on the output of the sensor, I was able to lower the voltage (and so the load) that the ECU thought it was experiencing. This resulted in the ECU pulsing the larger injectors at a reduced rate (technically speaking, at a reduced duty cycle) and so the fuelling was able to be adjusted so that it was again correct. And, at the top end, where the engine previously ran out of injector flow, the larger injectors could keep up!
The driveability of the car was perfect.
A 10 kilo-ohm pot can be used in nearly all cases without upsetting the original sensor behaviour (note: an exception is a narrow band oxygen sensor, where even a 10 k pot is too great a load.). By using a multi-turn pot, fine adjustments can be made.
As with any modifications that alter a sensor’s output, you need to be aware that the ECU will change all of its outputs that rely on that sensor’s input. In the case of the Honda Insight’s EGR valve, the lift sensor is used only for EGR feedback. However, other sensors (eg a MAP sensor) are used by the ECU for lots of things. For example, reducing the output of the MAP sensor will not only alter the pulsing of the injectors, but will also advance the ignition timing.
Another point to keep in mind is that the technique does not add a constant voltage to the signal. Instead, it does better than this by adding (or subtracting) a constant percentage. For example, if you set the pot to add 20 per cent to the sensor voltage, this percentage stays the same across the whole sensor output range. This is one reason that the pot approach works so well in practice.
The use of a pot to tweak the signal of a voltage outputting sensor is very powerful. It is also very simple to install, extremely cheap, and allows fine tuning of the modification.
Next week we’ll cover another overlooked but effective way of modifying cars – relays.