PIC16LF877A vs PIC16F946

[Christophe]

Hi ,

for a new project I'm planning on using a new PIC that has more I/O then the current PIC16LF877A. I have some questions about the differences I found from the microchip website:

* What is the difference between enhanced flash (PIC16LF877A) and standard flash (PIC16F946) ?

* What is the difference between I²C and mI²C ?

* What is the goal of an internal oscillator (8Mhz for PIC16F946) ?

* What is LVD ? ( PLVD for PIC16F946)

* What extra is nanoWAtt technology ( 946 )

* Why isn't there an 'L' version of the PIC16F946 ?

* Why is the PIC16F946 cheaper than the PIC16LF877A ?

tks

[Ttelmah]

1) Look at the quoted life in the data sheet. Also look at the voltage range over which all functions are supported.

2) Depends on the context. mI2C, is used as a 'trademark', by Mentor graphics, for their implementation of the I2C bus. MI2C, is used by Microchip, for versions of the I2C peripheral, that support _separate_ interupts for the transmit and receive operations, with 'MI2C', standing for the 'master I2C' interrupt.

3) Saves having an external oscillator. Potentially two more pins available for I/O.

4) Generates a 'low voltage' warning interrupt, when the supply voltage falls below a defined point. Allows you to perform a 'clean' shutdown (provided _you_ ensuure that the supply rail falls slowly enough for everything you want 'saved', to be stored before the rail falls below the operational voltage. 'P' means it is programmable 'on the fly', while the standard versions are only changeable when the chip is originally programmed.

5) A 'generic' name for a number of features. Typically chips with this draw less power, and also offer options to turn more individual coponents 'off', to reduce power consumption further. Some modules in particular, draw a lot of power on the standard chips (watchdog on a 16F877, draws about 9uA. The same module on the 'NanoWatt' chip draws about 1/10th of this.

6) It is an 'L' chip. They just never produced a 'non L' version. 'L', is generically used, where this is a lower voltage 'variant' of an existing high voltage only chip. Several chips launched in low voltage versions, do not use this designator.

7) Commercial. All down to quantities being sold. Basically, new, chip variants, are often sold in quantities that make the sales of some of the older chips look 'silly'. A an example, I used one of the smaller RFID products in a module recently. Initial order, was 10million units over 6 months.

Best Wishes

[Christophe]

[Ttelmah]

_Read_ the data sheets. Look at the quoted cycle life, and then the voltages over which the memory will work. On the 'standard' flash, the life is 1/10th that of the 'enhanced', and the erase cycle, won't work at the lower supply voltages.
The accuracy of the oscillator (a trimmed RC oscillator, not a 'crystal'), is again, _in the data sheet_.

Best Wishes

[Christophe]

I've read.. that was 1%. But how is that accuracy compared to an external ceramic resonator?

[ckielstra]

The design of the internal oscillator is something like an R/C oscillator, never as accurate as an external crystal. The internal oscilator is specified with a 1% accuracy because it is factory calibrated, but only at a certain temperature. With temperature changes the internal oscillator will drift a bit, this can be adjusted with the OSCTUNE register.

[Ttelmah]

The same comment applies for the accuracy of the ceramic resonator. Look at _it's_ data sheet, and the ones for any crystals you are thinking about. Generally, the better resonators will be quite good (much better than the 1% quoted for the internal RC oscillator), but some cheaper units, and especially if the temperature gets extreme, have quite poor accuracies (cause of many failures in USB peripherals, using these, when they are moved outside comfortable 'room' temperatures).
Accuracies in general would run:

RC
Resonator
Crystal

Going from 'worst' to 'best', but their is a lot of overlap from different units.

Best Wishes