Communication Protocols in Embedded Systems Questions and Answers

Answer: C

Correct answer: another chip select line

SPI has no addressing: a device is selected by pulling its own select line low, so every additional device costs another pin on the master. That is the bus’s principal limitation and the main argument for I2C, which addresses devices in the protocol and so adds nothing per device. The clock and data lines are shared and need no change.

Answer: D

Push-pull outputs and a dedicated clock let SPI run an order of magnitude faster than I2C's open-drain bus. It uses more wires for one device, has no addressing and no error detection at all, which are the prices paid.

Answer: A

Any device may pull the line low and none drives it high, which is what allows arbitration and clock stretching to work. The resistor value sets the rise time and therefore the maximum bus speed for a given capacitance.

Answer: B

The rise is an RC exponential against the bus capacitance, so a smaller resistor charges it faster and draws more current when a device pulls low. Choosing it is the standard trade between speed and power on an I2C bus.

Answer: C

A slow device that needs time between bytes simply holds the clock down and the master waits, which is much simpler than a flow control protocol. Not every master implements it, which is a compatibility question worth checking.

Answer: D

Each master watches the line while driving it, and the one that intended a one but sees a zero knows it has lost and withdraws without any data being corrupted. The open-drain bus is what makes this non-destructive arbitration possible.

Answer: A

Seven bits give 128 combinations, of which sixteen are reserved for special purposes, leaving 112. Two devices with the same fixed address on one bus is a common and awkward problem, solved by a bus multiplexer or a second bus.

Answer: B

Correct answer: a differential pair for noise immunity

I2C is single-ended: two open-drain lines, clock and data, referred to a common ground and pulled up by resistors. Differential signalling belongs to CAN and RS-485. The other three are all genuine I2C features - multi-master with non-destructive arbitration, a 7-bit address in the first byte, and an acknowledgement bit after every byte.

Answer: C

Differential signalling with 120 ohm terminations at both ends is what gives it immunity in a vehicle's electrical environment. It needs a transceiver between the controller and the bus, which is an extra device on the board.

Answer: D

Nodes transmit their identifiers simultaneously and a dominant zero overrides a recessive one, so the numerically lowest identifier wins and continues without any loss of data. Message priority is therefore fixed at design time by the identifier allocation.