Lpc2000

Hi Charles (and philips_apps ?) ,

> In some place in the datasheet, the external xtal is specified as 10
> to 25MHz. There might be problems trying to get the PLL running
> properly off a 4MHz clock.

The PLL divider can be set for eg. 16, and the prescaler still can be

set such that the CCO runs within its specified range ?

> If you need a 4MHz clock for some other part of your board then
> perhaps you can generate this out of the LPC2119.

I already have an accurate 4 MHz CLK output from an RF section on the

system (has its own crystal), and I was pondering over saving a crystal/caps

on the LPC2119/29 by injecting that 4 MHz into the CLKIN of LPC.

Board space needs to be minimal, an SMT crystal is quite big.

I might end up putting a low PPM crystal in after all, unless there is some clarification
to the actual freq range allowed with an Ext input Clock signal into LPC2000.

Robert at philips Apps team ?
Can you assist ?

Anywho,

Bit of a pain to bootstrap it, but it just comes down to a pull up resistor on

a control line - although the default clock on power up out of RF device is

500 kHz, until set to 4 MHz with an SPI instruction......

( I do this on MSP430, it clocks of its internal DCO, and the 4 MHz is used
as SMCLK for peripherals that need accuracy like timers and USARTs etc)
That could very well be the biggest problem : the LPC clock will only be

500 kHz at cold start for the first few mS.

I can't see why one couldn't inject a full rail-to-rail 50% duty signal like that

into the oscillator input of LPC2000, it *should* work.

Reliability is the issue of course.

I'm sure there's a reason for the restriction on EXT clock in, but I don't readily

see it. The 10-25 MHz as an oscillator with Xtal, that's understandable, it's quite

hard to provide an on-chip Pierce or similar that can cover a highly extended freq.

range.

The only other reason I could have seen for the equal freq. range restriction

on Ext Clock of 10-25 Mhz could have been some AC coupling time constant

that is too low for signals much lower than 10 MHz.

But that seems like it doesn't make sense either.

Thanks anyway !

;

B rgds

Kris