Cypress PSOC programmers please comment.

Hello. Myapp is for a PSC motor controller (speed and direction) there are many app notes for various chips eg Microchip AN967 gives decent details, code and so on. I am used to 8051 and with the given data from Microchip appnote AN967 would be able to impliment a similar control system.

I went to my chip vendor who suggested instead of staying with the 8051 family to try Cypress PSOC because he said it was so easy to use as to almost make the programmer redundant. However, in reality it seems that when one has unusual requirements there may be a similar workload as would be the case if I stuck with my tried and trusty 8051 variants (inlude Atmel Mega).

I am wondering if the learning curve for a newbie to create such a driver for PSOC would take me longer to implement than for the likes of familiar 8051 or AVR in C.

If anyone here has been using the PSOC system, and has previous experience with other micros I'd be particularly interested in your comments as to the development curve for new addons, and comments in general. Kind regards, Alistair.

Reply to
Alistair George
Loading thread data ...

My PSOC Experience is over a year old, so It may have improved. The analog components are poor. they work, but are not hi precision. The temperature range is lower that the PIC. The power consumption is quite high with the analog components on. The pin configure-ability is limited. You can not put any pin anywhere. The tools are good. You need them the number on configuration registers is huge. The learning curve is not too bad for the tools. The digital blocks worked very well. I liked them. The core does not have bit instructions, you have to mask. The compiler is $145 It is a great $145 compiler. It is not a good $1000 compiler. The code is large. The interrupt handling is poor, you may have to do them in asm. The Tool box has a lot of pre-done code modules you can just click to include. There was no bootloader, or room for one (in my app). It programs SPI like the PIC. The limitations are glossed over in the docs. I had to trip over them. Like the temperature sensor is useless. The chip draws enough current to rise 10-20C So you temperature is within that range.

I was disappointed with the chip. There is a good web site for it

formatting link

The chip is a good choice for many things. Their marketing pushes it to hard. It is not the 8-bitter to replace all others.

Reply to
Neil

Thanks Neil. A lot of hype from the company as you say. I have been playing around with the IDE today. So far it makes little sense. Al.

Reply to
Alistair George

I helped a student with a simple project using a PSOC a couple of years ago when I was visiting a friend's lab. It went quite well but I feel that a conventional MCU like an AVR or a PIC would be just about as easy to use, and a lot better for most applications.

Leon

Reply to
Leon

Alistair,

I have worked with many micros from way back starting in 1976. It seems t me that people's opinions are highly coloured by which micros they firs used and the subsequent paths and experience. I came to the PIC 16XX family relatively recently and I regard them as a throwback to the 8748 only worse in some ways. Now I don't mean to start a flame thread sinc there is a sizable cadre of PIC users who swear by its capabilities. It i rare to find a micro that is so perfect for a task that no other micro wil do. And if it is the case, what happens on the next project?

Each micro has its advantages and disadvantages. If you are concerne about the capabilty of any 8 bit micro in terms of execution speed etc then you should be looking at 16 bits or better anyway. The PSoC micro i a middle of the road micro- no great shakes, but my applications are no cutting edge in terms of speed. But the PSoC is more than a micro. Asid from the CPU it has an array of analog and digital blocks which yo configure to realize the I/O configuration that you need. Unlike an 805 where you get a fixed number of timers and UARTs, here you get to make th choice within the limitations of the device. If you need 3 UARTS well yo can do it. You can also allocate I/O pins: I once used a single UART t communicate with 4 external devices in turn by changing the I/O pin dynamically. The blocks are somewhat like an FPGA and can be modifie dynamically so that you can change the configuration on the fly. Let's sa you are working with a half duplex serial port, and you have maxed out th resources. A UART takes two logic blocks, (Tx and Rx) but on the PSoC yo can use a single logic block as a TX to send the message and the reconfigure to an Rx for the response.

Yes there are limitations on the analog circuitry. Keep in mind it is $3-4 device with both digital, analog and the reconfigurabilty running o the same piece of silicon. Just because you can configure a 14 bit A/D doesn't mean you get even close to an external part made exclusively t convert perfectly.

But the whole package comes together well. As an 8051 programmer, I don' think you will have any trouble making the transition, and both Cypress and the users on PSoCDeveloper.com (of whom I am one) are extremel helpful. Some of the time I may sound like an apologist, but what I woul like to convey is this. I remember the good old days where there was a ne peripheral to consider almost every week- I remember the 8255, the 8251 th Zilog PIO, memory mapped I/O versus I/O mapped I/O; it was an exciting tim where you could use your ingenuity. The PSoC re-captures that.

-Aubrey

Reply to
antedeluvian

The PSoC is a 8051 core, more or less. They've added a bunch of configuration and paging registers to fit in all the things the chips can do. For the price, they are good chips; I like them a lot to replace low-precision op-amps, comparators, etc. I've uses them as a slave display driver by putting a UART and a timer in, and used them as active filters by plugging in a 4-pole LPF with some gain on other occasions. They are very low cost in quantity; most projects I've used them in end up with the chip in die form, no external crystal. My stuff is normally room temperature, so that's not an issue. The wide supply voltage is a plus (2.7-5.5V), making it easy and cheap to do battery-powered devices. The architecture is limited, being based on the 8051. Their tools are reliable, but quirky. Expect to spend some time figuring out how to really use them. Cypress does have some PSoC-family parts that do USB, and 2.4GHz radio. They have announced plans for improved parts with ARM-based cores and higher-precision analog later this year.

-Travis

Reply to
Travis Hayes

Aubrey and Travis thanks for your comments. Aubrey I already joined the PSOCDeveloper forum and see you are very active there. Downloaded the flash video and thats sold me on the flexibility of the beast(s). I'd love a low-cost emulator (the cypress one comes out at $US600) but for me thats a big investment.

But its a no-brainer to build a specific task one as per Cypress appnote an2323.pdf

Otherwise, Aubrey et all, if any know of a redundant ICE somewhere please lets know! Alistair. PS I found one already sold Cypress ICe on Ebay; sold for $US36!

Reply to
Alistair George

Be careful of the ones on e-bay. If they are the older variety and may no support all the new devices. An ICE 4000 is OK for a normal 8 bi application, provided you don't need expanded memory and some of the mor specialised devices. I guess $40 would be fine in this regard. Check th Cypress web site for compatibilty. It also operates on the parallel por of the PC, so you need a PC that will support it. Even then some peopl had problems with some of the parallel ports and versions of Windows if remember correctly.

The current emulators are called ICE-Cubes.

-Aubrey

Reply to
antedeluvian

Alistair

I forgot to mention that if you buy an ICE cube, the C compiler i included in the price. (in case this fact helps in your financia calculations)

-Aubrey

Reply to
antedeluvian

Select and place the blocks. Connect the cross bars. Configure the clocks. Then configure the pins. The environment should make a skeleton program.

Reply to
Neil

The PSoC is most Certainly NOT a 8051 core. It is the core from the low end Cyress USB chips. It would be great if it was. Then it could be used with 8051 tools. The core has its own pluses and minuses. It has some features like the 8052 core. But you would expect the designers to pick features they liked. I feel the lack of bit instructions makes the code bigger.

I would put it more in a PIC16 "Class" then 8052. Due to the Smaller memory. TI and Silab have better analog sections in their chips. But there cost more and do not have the blocks.

The chip certainly has many uses. Cypress should accept what it can not do. The "Data Sheets" should have the information. You should not have to look the the block data to guess if it is good. The power consumption is high they give that info now, but back then they did not.

As for me I would have to be dead sure it would work. My boss would kill me if I used it again and failed.

It is certainly a useful chip for the right applications.

Reply to
Neil

I'm convinced its perfect for my app. The design specification has a huge wishlist that is only partially going to be incorporated at the start. I was blown away by the tutorial for the PSOC which of course starts off by expounding forth on how good it is! Future upgrading of specifications means the Programmable System fits well with me. Did the first simple course today to switch LED on and off. Came up with a warning which made me hunt high and low until I found in the PSOC forum that 'you ignore this warning' hmmmmm.

Aubrey I know the C compiler is included in the price, but unfortunately its just window dressing to me for now due to financial constraints. Appreciate the warning you offer on older IDE. Kind regards, Alistair.

Reply to
Alistair George

can

Ok, I will admit it's not 100% compatible to an 8051. My original statement of "the PSoC is a 8051 core, more or less" perhaps was misleading. The original poster said a vendor "suggested instead of staying with the 8051 family to try Cypress PSOC." Prformance-wise, I think it is on the same level as many of the 8051-like chips out there. I agree that the lack of bit manipulations is a hinderance. My biggest gripe is the accumulator-based instruction set. It seems that just about everything has to go through the accumulator, which is a serious bottleneck, compared to register-to-register based architectures out there. It is definitely targeted to low-end stuff; the largest part you can get is 32K, with 2K of RAM. Recently, it seems that Cypress is pushing them hard for capacative-touch applications, and all but forgot that many of us aren't building iPods.

We both agree that it is a fine chip in its place, and not suitable for everything. The price is right, if the part fits the application.

-Travis

Reply to
Travis Hayes

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.