Appunti Automotive electronic Systems completi

CONDITIONING

The purpose of conditioning is to:

• Adapt the signal to A/D converters requirements like voltage range;
• Preserve signal integrity which may be affect by noise and losses on the wires.

The connection can be:

• Single ended;
• Differential;
• Isolated.

Single-ended connection has a single common return line. The advantages of this solution is the fact that if many channels have to be acquired a single multiplexer can do the task, so the connection require less cable and it cheaper (multiple input cables share the same mass cable). The downside is the fact that different mass potential at the two ends will cause a voltage drop that adds up to the signal of the transducer and it also add noise. This voltage drop between the mass is due to the fact that a small amount of current can flow on the mass and the connecting wires have a parasite impedance.

Differential connection:

With now the two different mass potential will only be significantly attenuated and the signal.

will have less noise (now the current between the two mass is ero). The downside is the fact that now a single multiplexer isn’t enough because we have also to deliver the ground information of the transducer.

Isolated connection guarantee a galvanic isolation between the transducer and the acquisition circuit; they are typically realised with opto-isolators or fibers. Anyway they are expensive and bulky.

The noise in a signal arises from coupling with external sources and it can be:

• Capacitive: caused by high frequency components that run close to high impedance nodes;
• Inductive: due to a magnetic field.

In general it depends on many factors and it can be mitigate by:

• Using shielded cable (cable with a metallic shield which is connected to ground, to protect from magnetic field);
• Using short connection;
• Separate power lines from signals and also analog from digital line;
• Using coaxial cable for single-ended connection and twisted pairs for differential connection:

The first are cable that are surrounded by a dielectric material (like the one of the antenna) while the second is a technique where the cables are turned one around the other (this reduce the electromagnetic filed generated and also increase the rejection of external one).

A/D and D/A converters

Control system can be done with:

• No control: the process receive the input information and elaborate an output signal, but we don't know if the output will be the desire one;
• Open loop control: the desire output response if fed into an actuating device with send an output signals to the process. In this way we hope that the output signals is close to the desire one;
• Closed loop control: now the actual output signal is measured and compared to the desire one. A comparison circuit gave the information on how to modify the process to obtain the desire output signal.

Digital control system is based on microcontroller which process information that can have only 2 possible state.

• Extremely cheap;
• Flexible, because it can be reconfigured via software;
• Scalable and adaptable, because parameters can change over time and it's possible to use different sizes of memory;
• They are less prone to environmental conditions.

• Analog: continuously valued signal with infinite possible values in between. The world is analog;
• Digital: discretely valued signal encoded in binary. The world of microprocessors is digital and the value can be only 1 or 0; also these 2 values can be associated with the states ON and OFF.

The signal, and the precision is limited by the which represent infinite real values with a finite number of finite values. Sampling a signal is like taking a snapshot of the instant value with a defined discrete frequency; after a discrete amount of time a new snapshot is taken and so the continuous signal is now represented by a finite number of points which represent the amplitude of the signal at a given instant of time.

In reality we can have some problems:

• Jitter: this is caused by the fact that the sampling period may not be the same during the process;
• Aperture error: measuring the value take a finite amount of time so the sample taken has a small deviation.

The biggest problem of the sampling process is the this cause the appear of copies of the frequency spectrum on the sampling signal and this can bring the signal to be undistinguishable from others. This occur when sampling frequency is near the same or lower than the one of the signal; to avoid this issue the

The sampling frequency must be at least two times higher than the frequency of the signal itself in order to prevent aliasing. Anti-aliasing filters are used to restrict the bandwidth of the signal and satisfy the Nyquist theorem. In real circuits, aliasing is always present, but it can be mitigated by using a sampling frequency that is 5 to 10 times higher.

Quantization is necessary to reduce the precision of the signal from infinity to a finite value. This process replaces each real number with an approximation from a finite set of discrete values. Using an n-bit quantizer, we can have 2^N discrete values. Quantization noise is the difference between the input signal and the quantized signal. The use of a more finite set of values reduces the amount of noise in the signal. The quantizer sets two thresholds to determine which finite value to use for an infinite number of input values.

(ADC):Flash;

• Digital-ramp, dual slope, counter slope;
• Successive approximation;
• Sigma-delta ADC.

Flash ADC consists of a series of comparator, each one comparing the input to an unique reference voltage. The comparator outputs connect to the input of a priority encoder circuit which produce a binary code. As the analog input voltage exceed the reference voltage of the comparator, the output will saturate at +V SS and the priority encoder generate a binary code based on the highest-order active input. This ADC is:

• Simple and very fast because the comparation are done in parallel (only limited by the op-amp slew rate and the encoder speed);
• Low resolution (limited by the number of comparator);
• Expensive;
• For each additional output bit, the number of comparator required is doubled.

The encoder has a enable pulse which is used to refresh the read of the output of comparators.

Dual slope ADC is a more complex circuit that use an integrator, a comparator

and a digital counter together with a control logic. The input voltage is fed to the integrator for a fix amount of time; then the reference voltage is fed until the output voltage gets zero (this control is made using a comparator that compares the output of the integrator with the ground voltage). The time that occur for the output of the integrator to become zero is counted using a clock frequency by the counter device; for every period of the clock a bit is modified, starting by the LSB, and the numbers of period of clock that occur define the digital output of the system.

The main advantage of this system is that has an high resolution because to increase it we have to simply increase the clock frequency, anyway the count is very slow (because the count start always from the LSB) and the real resolution is limited by the precision of the integrator and the comparator. The slope of the integrator is defined by the values of its resistor and capacitor.

Successive approximation ADC works in

• Resolution: the number of possible output voltage that the device can reproduce;
• Maximum sampling frequency: the speed at which the device can produce the correct output;
• Monotonicity: the ability of the device's output to move only in the direction that the digital input moves. So for example when the digital input increase, the output will not decrease before reaching the correct output;
• Total harmonic distortion THD and noise: in general the output will be a sinusoidal signal with some noise and other harmonics that distort the output signal, so this noise needs to be low;
• Dynamic range: the difference between the largest and the smallest signal that the device can reproduce.