Lpc2000

Hallo there,

I need some help with a rough powerestimation for an possible 
aplication an the LPC2105.
I know that an estimation without a dedicated Hardware and the 
Information about the used I/O loads is realy hard, thats also a 
reason why I try to get some information here.

In the Datasheet I found the "static" powerconsamtion of the core to 
be 30mA * 1.8V = 54mW only for the core. 
The apllication I would like to estimate would need one UART-Port 
running at upto 115200 baud perhapes only 38400 (not clear jet). In 
addition to that I would need 12 GPIOs to communicate with three 
different AD-converters. 4 outputs and 2 inputs should run at 2.2MHz. 
The other 4 pins should run at 1MHz (3 output, 1 input).
I think i´d need also the captuer compare timers and the interrupt 
controller to solve the required algorithm and to synchronize the 
converters. 

The Algorithm wich should run on the System would be an 1024-point 
FFT, wich leads me to the next question if this is possible to manage 
with the LPC2106 and wich clock I would need (think 25Mhz are not 
enough so I have to work with 60MHz or am I wron?).
In addition to that the 1-bit inputsignals from the GPIO should be 
serial-parallel converted to a 16-bit value. Some derivatives and 
calculation of Moments (out of the power density spectrum after the 
FFT) should also be implemented.

I am really not familiar with MCUs and programming in C or assembler. 
My main interst is to get a rough powerestimation in comparison to a 
much more consuming FPGA solution, wich will need several Watts.

I hope you could help me or give me some additional hint where to 
find some useable answers

Regards
Andreas Kirsch

Hi Andreas,

just an extimate but based on some measurements. All the peripherals
you mentioned should add less than 10 mA to the 1.8V power supply
running at the speeds you listed. The 3.3V supply depends on your
driver requirements in particular the outputs. Inputs are pretty much
negligible, outputs can add up if you drive external power-hungry
devices. 

The power consumption of all peripherals active at max. speed combined
 is a lot less than that of the CPU. We are currently conducting some
tests what peripherals add to the equation and I will post some results.

A qualified guess for your application for 1.8V is around 40 mA @ 60
MHz or roughly 20 mA when you run the device at 25 MHz no PLL but with
the peripherals you mentioned. 

Hope this helps, Robert

--- In lpc2000@yahoogroups.com, "didgiei" <Kirsch@i...> wrote:

> Hallo there,
> 
> I need some help with a rough powerestimation for an possible 
> aplication an the LPC2105.
> I know that an estimation without a dedicated Hardware and the 
> Information about the used I/O loads is realy hard, thats also a 
> reason why I try to get some information here.
> 
> In the Datasheet I found the "static" powerconsamtion of the core to 
> be 30mA * 1.8V = 54mW only for the core. 
> The apllication I would like to estimate would need one UART-Port 
> running at upto 115200 baud perhapes only 38400 (not clear jet). In 
> addition to that I would need 12 GPIOs to communicate with three 
> different AD-converters. 4 outputs and 2 inputs should run at 2.2MHz. 
> The other 4 pins should run at 1MHz (3 output, 1 input).
> I think i´d need also the captuer compare timers and the interrupt 
> controller to solve the required algorithm and to synchronize the 
> converters. 
> 
> The Algorithm wich should run on the System would be an 1024-point 
> FFT, wich leads me to the next question if this is possible to manage 
> with the LPC2106 and wich clock I would need (think 25Mhz are not 
> enough so I have to work with 60MHz or am I wron?).
> In addition to that the 1-bit inputsignals from the GPIO should be 
> serial-parallel converted to a 16-bit value. Some derivatives and 
> calculation of Moments (out of the power density spectrum after the 
> FFT) should also be implemented.
> 
> I am really not familiar with MCUs and programming in C or assembler. 
> My main interst is to get a rough powerestimation in comparison to a 
> much more consuming FPGA solution, wich will need several Watts.
> 
> I hope you could help me or give me some additional hint where to 
> find some useable answers
> 
> Regards
> Andreas Kirsch