Forums

several questions regarding lpc2xxx functions

Started by Thomas July 15, 2005
hello,
we are planning to use the lpc2xxx family in some of our products.
now i check if the lpc2xxx family fulfills all of our requirement.
i am new to the lpc2xxx family but have already studied several
manuals and faq's.

reading all of these documents some questions arised:

1) software breakpoints
obviously the ARM Embedded ICE supports up to 2 hardware breakpoints.
how does it look with software breakpoints in flash memory? is this
possible?

2) wake-up-timer
the wake-up-timer is mentioned in the 213x datasheet but there is no
hint how long this timer is running until the controller is starting
to run. what time can i expect?

3) removing of clock
the faq on the philips homepage says, that the pll is not longer
working in the case the clock is removed (e.g. crystal is defect).
because the watchdog-timer depends from the clock the watchdog will
not reset the lpc2xxx. Is this correct until now?
so if i would like to have a well defined situation (=reset) in this
case i have to use an external watchdog chip. Is this correct?
if the pll is not longer running and the exterenal watchdog makes a
reset will the port pins be switched to input as they are in normal
reset state?

many thanks in advance for answerring my question
thomas


An Engineer's Guide to the LPC2100 Series

Hi Thomas,

let me take a shot at your questions:
1. Software breakpoint in Flash can theoretically be done but is
tricky. In RAM you replace the target instruction with a branch to the
debugger. You can do the same thing in Flash but you might reprogram
one location very often, so the debug hardware could wear out and
reach the reprogramming cycles. Also you need to reprogram a larger
area due to the fact how the flash is origanized. Reprogramming just
one instruction is not possible.
The only tool that I know supporting it is J-Link from Segger (comes
with IAR evaluation kit)

2. Wake Up Timer
How long does it take to wake up? If the timer is the only thing
"alive" the clock is shut down, the time is basically the wake up time
of the oscillator. This is typically some (e.g. 10) milliseconds.
Everything else is negligible compared to the 10 ms

3. Watchdog
No clock, no internal watchdog, that is correct. The watchdog protects
from software problems only, not from a broken oscillator. So the WDT
will not reset the device.
The second part of your question I have to pass I do not know whether
the reset of the LPC2xxx is asyncronous. If yes, the pins would go
into reset within some nanoseconds, if it is a synchronous reset, a
clock is required to get the pins into reset state.

May be somebody else can help out here.

Hope this helps a little, Bob

--- In lpc2000@lpc2..., "Thomas" <cool_1969@y...> wrote:
> hello,
> we are planning to use the lpc2xxx family in some of our products.
> now i check if the lpc2xxx family fulfills all of our requirement.
> i am new to the lpc2xxx family but have already studied several
> manuals and faq's.
>
> reading all of these documents some questions arised:
>
> 1) software breakpoints
> obviously the ARM Embedded ICE supports up to 2 hardware breakpoints.
> how does it look with software breakpoints in flash memory? is this
> possible?
>
> 2) wake-up-timer
> the wake-up-timer is mentioned in the 213x datasheet but there is no
> hint how long this timer is running until the controller is starting
> to run. what time can i expect?
>
> 3) removing of clock
> the faq on the philips homepage says, that the pll is not longer
> working in the case the clock is removed (e.g. crystal is defect).
> because the watchdog-timer depends from the clock the watchdog will
> not reset the lpc2xxx. Is this correct until now?
> so if i would like to have a well defined situation (=reset) in this
> case i have to use an external watchdog chip. Is this correct?
> if the pll is not longer running and the exterenal watchdog makes a
> reset will the port pins be switched to input as they are in normal
> reset state?
>
> many thanks in advance for answerring my question
> thomas