Best CPLD choice for a Stellaris platform

to

a

us

I'd use a bigger ARM chip. However, I like the Altera MAX II CPLDs. The software is easier to use than Xilinx's, and they seem to do more development with their devices than the latter.

to

a

us

I have been using the Lattice XP family which has a 3 kLUT part in a

100 pin QFP package (62 IOs). It has internal configuration Flash so you don't need to use flash in your MCU. I can get this for under $10 in qty 100 which is very competitive. I think when I have posted about this being the most cost effective FPGA I can find, others have posted parts that are similar, but not better that I recall. Lattice is stocked by Mouser and is easy to use. I think you have to pay for a version of their tools that includes a simulator, but a base version is free.

Si Blue has some pretty nice parts when it comes to package and price, but I can't say how real they are. If you look into this company, please report back here what you find.

Rick

to

a

us

I almost forgot. There are SPI and I2C I/O expander chips out there that are pretty inexpensive, up to 16 bits of I/O for $1 or $2.

Rick

And don't forget the good old 74HC595 for output expansion at about $0.30 for 8 lines.

An external low-end PIC/AVR/whatever is another option for IO expansion

I've never done a price tradeoff to see if adding multiple MCUs is cheaper than using one with a larger pin count. As the pin count goes up, they typically provide more peripherals, more Flash, more RAM, etc, so I would think multiple MCUs could be less expensive. But it is more work getting the extra software written and debugged. But once you write an I/O expander function, it should be reusable.

Rick

On Oct 3, 8:39=A0pm, "abrous" wrote: I was thinking about to use as a companion chip my

What issues did you hit with ATF1504, as it shows as being widely available ? There is also ATF1508ASx/ATF1508BE/ATF1508RE in 100pins, and in 44 pins, the ATF1502 too .

Alternatives would be XC2C64 and LC2064, but the Atmel parts have better logic density, well suited to IO expanders.

If it is a fixed IO mix, then simple shift registers are still hard to beat. i2c expanders are more costly.

-jg

to

a

us

Without a speed requirement your question is nonsence. Without a volume requirement, your price/availability concern is nonsence as well. You could allways google "shift register" and save everyone alot of time and effort.