Hi,

Anyone else having problems getting the AT91SAM7S256C revC silicone main xtal oscillator to run?

We have a product which works fine with revB silicone. Swap the chip for a revC silicone part and the main oscillator will not run. Tried changing the caps around the crystal to various values 22pf, 10pf, 3.3pf no cap at all. No luck.

This is happening on 30+ boards so its not just a one off.

The crystal is 18.432MHz and meets the data sheet spec. Digikey type XC1286CT-ND

Anyone with experiance of this?

Regards mpc.

Hi.

I have had to download the latest version 6 IAR kickstart. This worked.

Atmel suggested I use SAMBA and sent a new programming file: isp-flash-at91sam7s256.bin

You have to load this into the following directory: \ATMEL Corporation\SAM-BA v2.10\tcl_lib\at91sam7s256-ek\

Not tried this yet.

Anybody else having issues with the revision C silicon?

MPC.

Hi

There was a new PLL restriction introduced in newer SAM7X parts (I don't know about the SAM7S parts but presumably similar). Original PLL configuration had difficulty when used with newer chips and had to be modified to respect the details in the newest data sheet (compared to the original data sheet there are changes which are not listed in the documentation history).

After carefully following the recommendation, the PLL could again be set reliably:

while (!(PMC_SR & PLL_LOCK)) {} // wait for PLL lock before continuing
PMC_MCKR = CLK_PRES; // {74} first program the prescale and wait for ready
while (!(PMC_SR & MCKRDY)) {} // {74} wait for the master clock status to indicate ready
PMC_MCKR = (SELECT_PLL | CLK_PRES); // finally set the master clock to half the PLL value
while (!(PMC_SR & MCKRDY)) {} // {74} wait for the master clock status to indicate ready

The lines marked with {74} had to be added. Originally the prescaler and master clock settings could be programmed in one go but that changed at some time and they need to be programmed separately.

Regards

Mark

www.uTasker.com

Do you have any resolution on this problem?

Hi

The resolution was in the last post (at least this worked for us...)

Regards

Mark

Does it mean you actually had the main oscillator running, but you had problems to start the PLL?

rgds
Balint

Hi

The PLL was starting but it important to follow the steps before connecting it the processor:
- originally these steps were not required (they were not specified in initial user's manual and also SAM7 parts didn't have any problems).
- when the SAM7X512 was introduced it was also noted that the user's guide had been generally modified to require following a careful ordering of setting the PLL (see user's guide since it is specified there). Also, not following this sequence with the SAM7512 parts caused the PLL configuration to not be able to complete (although sometimes it did after a long delay).
- generally following the new recommendations solves the problem with old and new parts.

Regards

Mark