16f876 and 16f876A

[PicFan]

We have a commercial product that has been refined through extensive field testing that uses the PIC16F876-20I/SO. The non_A version of this chip is becoming relatively more expensive and it might just be time for us to consider replacing it with the PIC16F876A-I/SO.

I've read the data sheets and errata on both silicon revs and for our application it seems like a simple changeover. However, I realize that sometimes unpredictable problems can lie just below the surface only to be discovered once the application is actually running on the new silicon.

Recognizing and appreciating the talent in this forum, I would like to ask if you could give me some advise on what I might need to watch out for when we attempt to implement the 'A' version of the 16f876?

With Best Regards,
PicFan

[valemike]

We've migrated to the PIC16F873A from the 16F873 and from the 16F876 to the 873A with no problems.

You'll encounter hardware problems if your circuit's hardware was marginally designed to begin with.

Example: We used the Motorola MC78HC705C8 (PIC16C877 equivalent) with no problems. Migrating to the MC78HC705C8A caused things to malfunction (i.e. motor control going haywire, motor drives drawing too much current and getting damaged, etc.).

The reason for this was because the "A" Mot. part got cheaper due to a die shrink. Thus more yield, and the price savings gets passed to the customer. With the die shrink unfortunately bought with it less robustness in the micro and more susceptibility to problems in marginally designed circuits.

[ckielstra]

Make sure to check your programming equipment. Our old PicStart Plus can't program the PIC16F876A and the the old design we use it in has no ICD connection so we reverted to the non-'A' version.