252 or 2520

Hi
trying to pick a PIC
i need 28 pins and low power as in 'L'
so far i have narrowed to
PIC18LF252
or
PIC18LF2520
what are the advantages of the 2520 over the 252 ?
pkg to be SO28
need 4 A/D
any other candidates out there ?

Have you checked Microchip's site? There is a parametric search that you can use that will list all of the different features that each micro has to offer.

From Microchip website

Quote:

---

18F252
Not Recommended for new design
Please consider using device PIC18F2520

---

If you are not planning to make your device in the future, then either option will work. If you are going to make them in the future, the 18F252 might not be around and you will have to migrate to the 18F2520.

Paul Borgmeier
Salt Lake City, Utah
USA

I've used the 252 for a long time, and it's nice. The 2520 has a boat-load of problems. You might want to take a look at the latest errata sheet for the 2520 before committing to that one.

If you can live with the problems/work arounds shown in the errata sheet, then go with the 2520. If not, you might want to shop around for a similar 18F series without a 16 page errata sheet..;o}

Yikes !
bruce
this is pretty scary (errata)
have they respun the chip yet ?
any major issues to note?

the MEL parametic PIC picker is pretty nice

I'm not sure what silicon revision they're up to yet on this part, but it's worth looking into if you plan on using one. That is indeed a hefty bug list for the 2520.

I use the smaller RAM 2420 extensively without any issues. Admittedly, I'm not using anything over the ADC's and PWM modules (and all the other pins as Digital I/O) but they all work flawlessly. And I've almost maxed out the Program Codespace and PBP handles it all just perfect.

The big advantage of the 2420 over the 242 (2520 over 252) is on-board handling of MCLR and Internal Oscillator. This means you have 25 usable I/O pins on the 28-pin package...

thanks melanie for the encouragement
some of the errata applies to the I2C
have you used that ?

a quick scan of the errata sheet shows that the bugs
seem to apply equally to the 2420 and 2520

considering that the price is pretty close i always wondered why
they bothered with both sizes of memory

have you had or heard of any reports of strange field falures?
i have heard one, not sure how trustworthy or relevant it is
ds

pretty close price or not is not revealant on a single of few device. But when you do a project and create 100 000 or million of them... few cents of difference made enough difference to be considered.

Uneless you use the PBP I2CREAD/I2CWRITE, you shouldn't experiment any problem because it doesn't use the PIC MSSP... it's a simple software solution who handle everything for you on the i/o you want to use for your I2C bus.

Now wich one trust? I can't say, but before, be sure of they do all you need, be sure that a lower model wouldn't fit, and so on. If it was me to decide, i would check to some other wich have less errata. But it's just me, i prefer to leave possible problem to other. I have those in stock now in case of, but i've never messed with them.

I admit that i'm kinda addictive to those pinned black box :)

Yes, I use I2C, but PBP's I2CREAD and I2CWRITE and not the chips hardware module. Everything is fine as mister_e suggested. I also use internal EEPROM READ/WRITE also from PBP on this chip, again no problems with that either. The Timers work too and so do the ADC's and the pwm modules as I mentioned earlier. But on the negative side, I can't vouch for any hardware components I've not listed. The Labour and Parts saving in not having to add a Resonator or Xtal is worth the the few cents extra. I have about 100/day of the 18F2420 (and the LF version when earlier we couldn't get stock of the straight one) going into the field, with the first units being shipped at the start of the year - nothing's bounced back. This is a major product badged for a blue-chip company that earns me big-bucks in royalties, I wouldn't risk that on a suspect PIC.

Im trying to communicate with i2c two Pic´s a master and a slave of course. I attemp to do a communication with 3 slaves later but at first i have to do it with a single one. Im having problems initiating the i2c protocol.
Im using MPLAB to create the master and slave code. Using MPSIM and checking the ssp registers i see that the code is right in theory but once i try it on the chip it doesnt work. I cant initiate the Start condition i activate the SSPEN flag in SSPCON2 in the simulation it works pefectly but in the chip polling the register i can see that the flag have never been set i dont know which problem i have but please i need some help