PIC C software question...and AUSART vs EUSART

[weg22]

Sorry for the excessive posts...I'm trying to spec out a PIC for my application and was wondering if there was a difference between AUSART and EUSART. Basically, I want to use the INT_RDA interrupt and have done so previously with AUSART. I wanted to confirm that this code would work just the same with a PIC which had EUSART?

Oh yeah, and I also have to purchase the new version of PIC C to program the 18F8722...this is what I should buy, right?
http://www.ccsinfo.com/product_info.php?cPath=Store_Software_Compilers&products_id=PCH_full

Can I still program the 16F series with this?

Thanks again!
-weg22

[Mark]

You have to use PCM to program the 16F's. Make sure that you keep both compilers up to date. Sometimes there are problems if the versions are too far apart. PCM Programmer has posted on this before. Be careful of the EUARTS. They are quite buggy. Read the errata on the part first.

[Ttelmah]

Unfortunately, Microchip don't make this easy...
Most of the older chips (18F452 for example), have USARTs (they call it 'addressable', but tend not to add the 'A' to the abbreviation, except on a couple of chips). Then chip like the 18F1320, have EUSARTs, and most of the latter chips apparently have these. However there appear to be multiple versions of the silicon for both components. Historically for example, some chips need the TRIS registers set as inputs, while others with the same 'module', need them set as outputs for the section to work. Microchip even have this changing sometimes on the cores, with (for example), their own ICE version of the 18Fxx2, requiring different settings in it's initial revision, to the chip itself, and the 'datasheet' settings refusing to work with the first processor revision...
Now the comment about looking at the errata sheets, is spot on. On the EUSART modules (together with the latter MSSP modules in particular), there are a lot of individual problems with some chips (especially the 25xx chips), and some 'generic' ones across parts of the whole line. Now the 18F8722, has less problems than most of the chips, and I have used this in a product without problems, with the EUSART functioning OK.
The 18 chips as a whole, do tend to be more sensitive than the older chips to supply noise, and careful oscillator design (things that people 'got away with' on the older chips, often give problems on these). So provided you are careful with your design, the chip should be OK.

Best Wishes

[weg22]

[Ttelmah]

It has the 'drive' component of an oscillator, with the possibility to use this with an internal R/C network to form an oscillator, or with external components. Remember though that if you use the 'all internal' route, the accuracy is limited, and may be too poor for many applications. If instead you use an external resonator or crystal, driven by this component, you run into the potential problem, that the PLL component internally is very sensitive to being overdriven, and the gain of the actual drive gate is hiugher than on older chips. Hence the actual crystal selected, is more fussy than on the 16 chips, and may need a series resistor, to keep the loop gain lower, and avoid overdriving.
Historically, this was the cause of a lot of problems on the earliest 18F chips, with people using crystals that had functioned fine on 16 chips, and finding that they were getting 'odd' frequencies. Latter chips have improved on this, but if you intend to use the PLL module in particular, it is well worth being careful in the crystal choice,and layout round this part of the chip.
The internal R/C solution, is not adequate if you want to operate over the full temperature range of the chip, and carry out serial I/O, or similar timing functions, without at least the addition of temperature compensation. The cost of doing this, usually means that the crystal solution is cheaper, but with the caveat mentioned...

Best Wishes