TK,

there is a fairly new Application Note on the Philips website here:
http://www.standardics.philips.com/support/appnotes/microcontrollers/pdf/an=
10404.pdf

Although the title of the AN talks about initialization, have a look at
chapter 4, "reducing power consumption". Ports 1,2,3 have internal
pull-ups, always active, these pull-ups can add up to a noticeable
current if they are externally pulled down during power down mode or if
the software writes a "0" to the pins before it enters power down.

This would be my strongest suggestion, check pin level when entering
power down and get as many as possible to high.
------- Cut and paste from the AN  ------------
7=2E To reduce current consumption on the 3.3 V rail in power-down mode
(let's consider
the 144-pin devices):
a=2E There are 80 pins with internal pull-ups on Ports 1, 2 and 3. These
pull-ups are
always active.
b=2E Set those pins high or keep them floating before entering power down
mode. If set
low, they will consume around 50 =B5A (typical) on each pin.
c=2E Port 0 does not have any internal pull-ups. These pins can be set
low.
-------------------------------------------------------------


In addition, if not already done so, you could post LPC2000 related
questions in the Yahoo LPC2000 forum here:
http://groups.yahoo.com/group/lpc2000/

Hope this helps, An Schwob



tkeski@vaz.embedtronics.fi wrote:
> Hi,
>
> I have design using Philips LPC2114 processor. Its battery driven so mini=
mal
> power consumption is essential.
>
> Problem is that even with cpu power down the circuit still uses about
> 2-3mA. :I I have measured that almost 2mA goes to CPU so something there
> is not right. I've read erratas and application notes and lots of other
> stuff and done the things they suggest to get power consumption down, but
> I can't get power consumption lower than that about 2mA.
>
> Has anyone else have same kind of problems and maybe even solved them? An=
d no,
> changing cpu is _not_ a option. ;)
>=20
> Cheers.
>=20
> - TK

tkeski@vaz.embedtronics.fi wrote:
> Hi,
>
> I have design using Philips LPC2114 processor. Its battery driven so minimal
> power consumption is essential.
>
> Problem is that even with cpu power down the circuit still uses about
> 2-3mA. :I I have measured that almost 2mA goes to CPU so something there
> is not right. I've read erratas and application notes and lots of other
> stuff and done the things they suggest to get power consumption down, but
> I can't get power consumption lower than that about 2mA.
>
> Has anyone else have same kind of problems and maybe even solved them? And no,
> changing cpu is _not_ a option. ;)
>
> Cheers.
>
> - TK

My first thing to look for is the external pins to see if something
attached to the device is drawing current.

Cheers

PeteS

<tkeski@vaz.embedtronics.fi> wrote in message
news:878xuxkkz1.fsf@vaz.embedtronics.fi...
> Hi,
>
> I have design using Philips LPC2114 processor. Its battery driven so
minimal
> power consumption is essential.
>
> Problem is that even with cpu power down the circuit still uses about
> 2-3mA. :I I have measured that almost 2mA goes to CPU so something there
> is not right. I've read erratas and application notes and lots of other
> stuff and done the things they suggest to get power consumption down, but
> I can't get power consumption lower than that about 2mA.

What does the datasheet say about the power consumption in power down?
Are you really in power down mode or perhaps in sleep mode. The difference
could be whether the clock oscillator keeps on running or not. Also, if you
have any I/O configured with pull-ups on (does the LPC support that?) while
you have those inputs tied to ground, these will contribute to the power
consumption too.

Meindert

If you have stopped the clock and made sure there
is no load on some of the outputs of the MCU,
there is one relatively new possibility. If the MCU is
a new design or a new mask (I just don't know Philips
processors), at 130nm and below geometries the
leakage is pretty great, e.g. Motorolas MPC5200 specifies
52 mW with all clocks stopped - it is a pretty big
chip, in other cases this would be a lot less but
still far from the practically zero consumption at stopped
clock we have been used to for decades.

Dimiter

------------------------------------------------------
Dimiter Popoff               Transgalactic Instruments

http://www.tgi-sci.com
------------------------------------------------------

Hi,

I have design using Philips LPC2114 processor. Its battery driven so minimal
power consumption is essential. 

Problem is that even with cpu power down the circuit still uses about 
2-3mA. :I I have measured that almost 2mA goes to CPU so something there
is not right. I've read erratas and application notes and lots of other
stuff and done the things they suggest to get power consumption down, but
I can't get power consumption lower than that about 2mA. 

Has anyone else have same kind of problems and maybe even solved them? And no,
changing cpu is _not_ a option. ;)

Cheers.

- TK