18fX52 --> 18FX520, 18FX58 --> 18Fx580

valemike · Guest

As many are aware, Microchip is recommending the use of the extra "0" suffix part for new designs. However, they do have some subtle erratas.

Does anyone know if the latest versions of PCH/PCWH handle the "workarounds" that are recommended in the errata notes? For example, the newer chips have erratas in ISR handling. At one time, even the legacy 18FX52 chips had erratas where you would need to insert a NOP as the first instruction of an ISR, which CCS promptly handled to avoid user problems.

-Mike

newguy · Joined: 24 Jun 2004 Posts: 1907

Mike,

I played with a 18LF4580 (either that, or it was the regular F part) to get familiar with the CAN bus. I wrote some pretty basic code that used the RTCC (timer 0) interrupt and the CANRX interrupts. No problems at all - it worked just fine.

At the time, I believe that 3.230 was the latest compiler version.

Hope this helps.

MikeValencia · Joined: 04 Aug 2004 Posts: 238 Location: Chicago

Okay thanks. It does give me some assurance. Thing is, i'm about to set off a production run, and debating whether to use the more mature 458 than the 4580. After all, it does save money since i can utilize the internal oscillator on the 4580; thus a savings of a crystal, 2 18pF caps, and the $0.75 in savings of not populating them!

treitmey · Joined: 23 Jan 2004 Posts: 1094 Location: Appleton,WI USA

for my 2cents. Prototype that thing with the chip you want to use.
And test the hell out of it.

Don't waist a production run.

newguy · Joined: 24 Jun 2004 Posts: 1907

Ditto what treitmey said. Build one and test. Far less headaches that way.

JimBob · Joined: 12 Oct 2005 Posts: 1

Hi Guys,

Working on a product at the moment. Started developing on 18F452. Having problems when moving to 18F4620. Wandering baud rate on comms. Exactly same settings as for 18F452 project, so my recommendation is as already mentioned here. Test your code on BOTH if you are migrating.