EmbeddedRelated.com
Forums

WDT feed

Started by extra300_it February 3, 2005

Hello,
One question about WDT feeding.

The user manual is (once more) not exatcly clear about WDT refresh.
It says that after writing 0xAA to WDTFEED, 'the next operation in
watchdog register space should be a write 0x55 to WDTFEED', otherwise
the watchdog will reset the cpu even if the timer is not yet elapsed.

What happens if one interrupt triggers between 0xAA and 0x55?

I see the wdt resets cpu, even if the ISR does not affect WDT
register space.

Do I need to disable interrupts before feeding WDT?

Thank you.


An Engineer's Guide to the LPC2100 Series

As far as I know, yes. Feed sequence must not be interrupted.

Kerem

----- Original Message -----
From: "extra300_it" <>
To: <>
Sent: Thursday, February 03, 2005 4:49 PM
Subject: [lpc2000] WDT feed >
>
> Hello,
> One question about WDT feeding.
>
> The user manual is (once more) not exatcly clear about WDT refresh.
> It says that after writing 0xAA to WDTFEED, 'the next operation in
> watchdog register space should be a write 0x55 to WDTFEED', otherwise
> the watchdog will reset the cpu even if the timer is not yet elapsed.
>
> What happens if one interrupt triggers between 0xAA and 0x55?
>
> I see the wdt resets cpu, even if the ISR does not affect WDT
> register space.
>
> Do I need to disable interrupts before feeding WDT?
>
> Thank you. > Yahoo! Groups Links




At 02:49 PM 2/3/05 +0000, extra300_it wrote:
>The user manual is (once more) not exatcly clear about WDT refresh.
>It says that after writing 0xAA to WDTFEED, 'the next operation in
>watchdog register space should be a write 0x55 to WDTFEED', otherwise
>the watchdog will reset the cpu even if the timer is not yet elapsed.
>
>What happens if one interrupt triggers between 0xAA and 0x55?

The WD feed fails.

>I see the wdt resets cpu, even if the ISR does not affect WDT
>register space.

That's to be expected.

>Do I need to disable interrupts before feeding WDT?


Yes.

Robert

" 'Freedom' has no meaning of itself. There are always restrictions,
be they legal, genetic, or physical. If you don't believe me, try to
chew a radio signal. "

Kelvin Throop, III


Yes, disable ALL interrupts when writing the AA and 55 then re-enable them.
-----Original Message-----
From: extra300_it [mailto:]
Sent: Thursday, February 03, 2005 9:50 AM
To:
Subject: [lpc2000] WDT feed
Hello,
One question about WDT feeding.

The user manual is (once more) not exatcly clear about WDT refresh.
It says that after writing 0xAA to WDTFEED, 'the next operation in
watchdog register space should be a write 0x55 to WDTFEED', otherwise
the watchdog will reset the cpu even if the timer is not yet elapsed.

What happens if one interrupt triggers between 0xAA and 0x55?

I see the wdt resets cpu, even if the ISR does not affect WDT
register space.

Do I need to disable interrupts before feeding WDT?

Thank you.

----
--
Yahoo! Groups Links

a.. To



When I first started with the LPC210x and Watchdog my board would
sporadically crash - mainly then when I was dumping data to the
serial port (lots of SCI interrupts).

I decided to protect the feed sequence from interrupts and this
cured the problem - since then the system is as solid as a rock.

Therefore the feed sequence should absolutely be protected against
interrupts!!!

regards

Mark Butcher

www.mjbc.ch
--- In , "extra300_it" <roberto.labaa@l...>
wrote:
>
> Hello,
> One question about WDT feeding.
>
> The user manual is (once more) not exatcly clear about WDT refresh.
> It says that after writing 0xAA to WDTFEED, 'the next operation in
> watchdog register space should be a write 0x55 to WDTFEED',
otherwise
> the watchdog will reset the cpu even if the timer is not yet
elapsed.
>
> What happens if one interrupt triggers between 0xAA and 0x55?
>
> I see the wdt resets cpu, even if the ISR does not affect WDT
> register space.
>
> Do I need to disable interrupts before feeding WDT?
>
> Thank you.