The LPC932 was (probably due to marketing pressure) out with some serious problems. I am happy to report that rev K is, as far as anyone knows, errata free.

When Atmel brought out the first flash '51s (their first '51s) they got to rev S before the chip worked as specified.

Philips broke the mold and started from scratch with the LPC89 series (2 clocks, 3.3V, a load of peripherals and extremely low cost). It is unfortunate that the chip manufactureres need to rely on us to test their chip, but how often have we relied in our customers to test our products.

I know of no reasonably complex product that did not have problems found once the users got to play with it.

This is by no means a defense of the past performance of the 932, just a recommendation that the fact that the early versions of the chip were awful should not scare you away from trying it now the bugs have been worked out.

One word of 'praise': Philips, at least has not been bashful publishing the errata once the problems were reported, that was not the case with Atmel.

Erik