Programmable System on Chip (PSoC®) gives you flexibility to create an awesome product, without huge design costs.

1609
 

Q. When morphing your PSoCs to be used in a touch scenario, how is cost being controlled? How is it better from a regular SoC? Isn’t a programmable SoC more expensive?
A. It definitely is more expensive and therefore, it will not be the same chip that is used in the touch family. We have a base line that comprises of all the IP blocks which go into the ‘mother chip’ called PSoC. From the ‘mother chip’, some features need to be eliminated and only essential items needed for touch are retained. These essential items include power system, serial communication block and the touch sub-system. Also, as we go from one generation to another, the touch system becomes more and more sophisticated in terms of channel engineering and it has developed into a beast of its own. However, at the end of the day the touch chips need to have ultra-low cost structure and Cypress has a very good DNA on that front.

Q. So the PSoC would be programmed first to work like touch, the prototype is created and tested, and finally the finished version for mass production of touch is created, right?
A. Not really. We cannot afford that time-line now. About four to five years back, we branched off the touch business unit as a separate business unit from the PSoC business unit. The touch market requirements and challenges are well understood. The market size warrants a separate wholly funded effort. In my previous response, I was just illustrating that PSoC is the mother of all ideas from where several ideas emerged and one of those ideas was touch. But once it was discovered, touch itself became a separate business unit for Cypress. They have their own operating philosophies, constraints and cost reduction timelines which are much more aggressive than what is required for broad-base PSoC customers.

Q. We recently read about nvSRAM and PSoC being used to create mixed-signal data loggers. What are the main applications of nvSRAM?
A. They are involved in a very broad range of applications, one example is metering. The name of the game is to use our serial nvSRAMs to record data over a 104 Mhz SPI interface, without any loss in content. The value of nvSRAM over a simple EEPROM is infinite endurance coupled with fast write times. Recently, we acquired FRAM, a non-volatile memory based on Ferro-electric technology. FRAM also has the same winning traits of nvSRAM, which is why we bought them. Let me give one more example. Most of the storage systems today use Redundant Array of Independent Disks (RAID) for higher performance and data reliability. RAID systems use a fast SDRAM as a write-back cache for high performance. Although it increases performance, a write-back cache increases the risk on the system data reliability. A write journal is a nonvolatile random access memory used in a RAID system for keeping record of the ongoing transactions and helps increase the system reliability. nvSRAM, with its infinite read and write endurance, high reliability, and high performance is the memory of choice for the write journaling application. nvSRAM and FRAMs are also used in a variety of automotive applications like blackbox and infotainment applications. If there is sudden loss of power, the non-volatile chip comes into the picture insuring data protection. Electrically Erasable Programmable Read-Only Memory (EEPROM) was used before this, but it suffers from endurance and long write- time issues.

Q. From the perspective of a system design engineer, what are the differences that he would face whilst using nvSRAM when compared to EEPROM?
A. In terms of the read operation, there is not much of a difference. But a data-logger requires efficient write, not read. One can write data extremely fast in nano-seconds into nvSRAM, while EEPROM takes milli-seconds. Another important difference is that many engineers don’t like EEPROM because of low endurance. If there are, for example, after 100,000 write operations over its life-time, the EEPROM functionality starts to deteriorate. FRAM (and nvSRAM, to an extent) does not suffer from an endurance problem i.e. practically infinite. Whenever an application requires writing to a memory more than 100K times in its life, FRAM or nvSRAM is the preferred choice.

[stextbox id=”info” caption=”About Cypress India”]

Cypress has been operating in India for more than 15 years. The company has 650 employees in its Bangalore office, including half of its global design engineers. Cypress is unique among multi-national semiconductor companies in that it conducts all aspects of product development in India, including design engineering, product engineering, testing and characterization, applications engineering, software development, IT, strategic marketing, marketing support, planning, operations and finance. Cypress offers Indian professionals unparalleled opportunities to grow careers in a variety of areas, unbounded by geographical barriers

About Cypress
Cypress delivers high-performance, mixed-signal, programmable solutions that provide customers with rapid time-to-market and exceptional system value. Cypress offerings include the flagship PSoC 1, PSoC 3, PSoC 4, and PSoC 5 programmable system-on-chip families. Cypress is the world leader in capacitive user interface solutions including CapSense® touch sensing, TrueTouch® touchscreens, and trackpad solutions for notebook PCs and peripherals. Cypress is the world leader in USB controllers, which enhance connectivity and performance in a wide range of consumer and industrial products. Cypress is also the world leader in SRAM and nonvolatile RAM memories. Cypress serves numerous major markets, including consumer, mobile handsets, computation, data communications, automotive, industrial, and military. Cypress trades on the NASDAQ Global Select Market under the ticker symbol CY. Visit Cypress online at www.cypress.co

[/stextbox]


SHARE YOUR THOUGHTS & COMMENTS

Please enter your comment!
Please enter your name here