The acceleration of the wave of intelligent network connection makes today's automobile CAN bus face more and more pressure.As the highest transmission rate is only 1Mbit/s CAN bus technology (usually the actual use rate of automotive CAN system is up to 500kbit/s), it is difficult to withstand the high load rate of the bus and network congestion caused by a large number of ECU and external equipment.However, to realize the intelligent network connection of automobiles, the increase of ECU and external units is inevitable, and the efficiency of the whole system should be improved. From the perspective of CAN bus, the reliability and real-time performance of data and information transmission should be improved to become a 'curve to save the country' path.

In fact, the main source of demand for CAN FD today is a large number of external ECU units. At the SBC tcan4550-q1 new product launch held in TI's shenzhen office recently, Charles(Chuck)Sanna, product line manager of TI interface product department, said:'ECU is to cooperate with the car now more intelligent, more like ADAS, quite apart from the audio and video data, control the amount of data that is in a substantial increase, there is need to CAN the application of FD in ADAS, or HEV (hybrid electric vehicle) and motor vehicle related areas, in addition to the demand rate high, have more CAN interface.The OEM requirements of our advanced vehicles are from the perspectives of intelligence and electricization, which requires the future BUS architecture to switch from the traditional CAN to the requirements of CAN FD.'In general, it is domestic demand that is causing the chipmakers to change our direction.'

From CAN to CAN FD: traditional architecture may be 'out of date'

From the perspective of three dimensions, the difference between traditional CAN and CAN FD lies in the transmission rate. Charles(Chuck)Sanna summarized in the field: 'the traditional CAN rate is about 1Mbps, while CAN FD is about 5Mbps, which CAN reach 64 bytes.In addition, CAN FD CAN be compatible with CAN downward without a series of adjustments, but CAN cannot be compatible with CAN FD upward.Finally, our existing solutions and CAN FD solutions CAN help customers reduce the size and cost of PCB.

And this a series of new features, but also makes the existing structure and design method of 'backward' step by step, because if manufacturers still use the existing architecture, although most of the MCU CAN not loss of FD high rate, but because of its high transmission rate CAN only reach 1 MBPS, it will also greatly limits CAN FD potential, become nowadays some ADAS or other Prochain data the main bottleneck of large quantity scenario;On the other hand, in order to upgrade or extend the function of CAN FD, designers have to integrate multiple discrete components or completely modify microcontrollers in their designs, which is usually a time-consuming and expensive project.

In response to this, Charles(Chuck)Sanna said, 'traditional MCU or traditional programs do not actually have the ability to support CAN FD. Most of them support SPI. If you want to update the traditional architecture to support CAN FD at a high rate, the most direct way is to change MCU, but the cost may be relatively high.Many applications need multiple CAN FDS and multiple data transmission. For example, most new energy vehicles at present may have controllers of CAN FD or one or two POTS (interfaces). If extended to 3-5 cases, adding MCU infinitely is obviously not the optimal solution due to cost.That only exacerbates the problem of high costs.

A new way to 'reduce cost and increase efficiency'

This case, the best way is to increase a CAN separate the function of FD, such as CAN increase a TI through SPI latest integrates the CAN controller and transceiver FD system based on chip TCAN4550 - Q1, such traditional MCU or process through the SPI and TCAN4550 - Q1 docking, rate CAN be improved, at the same time also more economic cost, for solutions integrator, is undoubtedly a good means of 'authors efficiency'.

Richard Hubbard, systems engineer, interface products division, Texas instruments (TI), said: 'this approach ensures that there is no need to change the existing architecture too much, and no need to replace the MCU.To optimize the implementation of the customer's real needs, at the same time, the cost will be much cheaper.At the same time, the customer CAN still use the CAN FD controller contained in MCU and make an external CAN, so that the whole architecture CAN meet the requirements without increasing the cost of MCU.

How to achieve the effect of 'reducing cost and increasing efficiency'?Charles (Chuck) Sanna, said: 'the chip CAN make PCB size decreased by 50%, it put the Watchdog,', CAN FD controller, and CAN FD transceiver are integrated in the four samples, in fact these used in the external discrete device, may also need to resistor capacitor, such as to do some decoupling (decouple).Compared with our product, youshang's product only contains a controller, but still needs some transceiver emulators outside. If we want to add a Watchdog, we must add a third device, or an LDO, which will add a fourth device, and its overall cost will be further increased.All in all, our TCAN4550 - Q1 Watchdog, 'and the controller, transceiver inside, so the interface is simple, the cost is more economic, our SBC also supports the cars to 8 Mbps based on maximum data transfer rate during assembly software program to write faster cars, beyond CAN FD agreement 5 Mbps maximum data transfer rate, is a very cost-effective solution.'

summary

From the previous Antenna-on-Package technology of MMW radar chips to today's SBC chip that integrates Watchdog, LDO, CAN FD controller and transceiver, it CAN be noticed that TI has its own unique innovation ability in highly integrated chips.This advantage will be the source for TI to continue to lead the automotive, industrial and other chip markets, create the 'killer mace' of ultra-cost-effective chips for oems and solution integrators, and 'differentiate' TI products in the market.