MC9S08DZ60 32.789 kHz crystal external clock problem

cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 

MC9S08DZ60 32.789 kHz crystal external clock problem

1,971 Views
shadoooo
Contributor I
We are using Freescale MC9S08DZ60 microprocessors in our products. We had problems using these ICs with an external 32.789kHz XTAL using the first masks, as I discussed in my precedent post here:

http://forums.freescale.com/freescale/board/message?board.id=8BITCOMM&message.id=7237

Now we are in full production since some months, but we are reporting a 10% of the pieces reporting the same problems. The devices suddenly reboot itself, after a time variyng from a few of seconds to some hours. This problem is very bad to be discovered, as it isn't deterministic.
Normally the problem, when it appears frequently, is fully solved changing the uP on the PCBs, but of course this is awful, because the devices are very small and require the human operation of desoldering-resoldering. After that, we cannot undestand why the devices have this kind of behaviour, and if this problem is a known-case or not.  Note that if we modify the firmware to use the internal oscillator trimmed to almost the same frequency, the problem doesn't appear anymore, but of course the system isn't working well, because of the imprecise clock frequency.
All of this is very disappointing, especially from a brand like Freescale, comprehending that we need to lay the devices for a week turned-on and wait, to undestand if they "seem" to be working or not. And consider also that we are spending a lot of money to replace the "defective" processors, not to throw in the garbage the entire PCB.
Please give me the way on how to solve this!!!

PS-Where could I find mask errata for this IC? I have printed copies of 2M05C and 3M05C, but where can I find original PDFs in freescale website?
Thanks
    Andrea
Labels (1)
0 Kudos
5 Replies

481 Views
bigmac
Specialist III
Hello Andrea,
 
Your problem seems suggestive of the crystal not operating correctly with some the boards.  Do you have the resistor Rs in series with the crystal (normally fitted to prevent over-driving the low frequency crystals)?  If so, what value resistor?  If relatively high, perhaps 100k or so, you might try reducing the value, to see if this makes any difference.
 
Do you examine the SRS register after each reset, to see whether this indicates the cause?  The LOC bit should be set if the reset was caused by loss of the external clock.
 
Are you using the PLL or the FLL for the clock?  The PLL may be more susceptable to loss of lock.
 
Is the MCU operating from a regulated supply?  If not, this is another potential cause of reset, perhaps due to unexpected "dips" in the supply voltage.
 
Regards,
Mac
 
0 Kudos

481 Views
shadoooo
Contributor I
Hello.
Unfortunately I'm sure that the problem is not related to external components or power supply. I don't have any loss of lock or bad power supply detected. Furthermore, if I put the BDIV to 2 instead of 1, I don't have problems anymore. The problem definitely seems to be related to what described on mask errata 3M05C, "MCGV1 can cause a very short clock pulse". As far as I know, there's no solution for this problem even in the latest masks, so I this seems to go in a very bad direction... :smileysad:
0 Kudos

481 Views
JimDon
Senior Contributor III
Did you submit a service request ?
0 Kudos

481 Views
shadoooo
Contributor I
Yes.
0 Kudos

481 Views
JimDon
Senior Contributor III
The errata is supposed to be here, but it is not.
You need to submit a service request to get the errata.
Also submit a service request about the problem.
You will need to dialog with service to clarify it, but they are EE's, so they will understand what you are saying.

I am sure you tried this, but if you change out the xtal, does that help?

For now, I would save the bad boards and if the service request does not resolve the problem, you  should tell them you want to send a board to engineering.







0 Kudos