Prev: Caveats for Multi Layer Boards?
Next: As oscilloscope I'd like to see Dave Jones review
From: MooseFET on 16 Jun 2010 10:41
On Jun 16, 3:28 pm, Fred Bartoli <myname_with_a_dot_inbetw...(a)free.fr>
wrote:
> MooseFET a écrit :
>
>
>
> > On Jun 15, 2:42 am, n...(a)puntnl.niks (Nico Coesel) wrote:
> >> MooseFET <kensm...(a)rahul.net> wrote:
>
> >>> I have been looking at the Analog Designs and NXPs and a few other
> >>> ARM chips. Does anyone have much experience with using an ARM
> >>> near analog workings?
> >>> My main requirements (in no order) are:
> >>> (1) Easy to develop and design with.
> >> NXP. A serial port is enough to program it.
>
> >>> (2) Low power for a given clock speed.
> >> LPC1100 series. AFAIK these also have an internal RC oscillator (1%
> >> accuracy).
>
> >>> (3) Low amounts of electrical noise.
> >>> The processor will be an inch away from signals down in the 1mV
> >>> range.
> >> That shouldn't be a problem with proper decoupling.
>
> > It often is. a problem with 1mV signals. Note that I didn't
> > say that the noise was 1mV. Thats the signal.
>
> Oh, 1mV is plentiful.
>
> I'm currently working on a 1uV level signal board lost in volt level CM,
> with onboard really sub ppm distortion level mid-power amplifiers @
> 50kHZ/10V, 1 FPGA, a bucket load of PWMs, 1.5MSPS 16b ADC and DACs, just
> one 4"x6" PCB, has to work up to 85°C, and so and so...

My board is 2"x4" so:

1uV * (4*6)/(2*4) = 3uV

Assuming it scales by area. This should be good enough :)


BTW: My signal is fairly broad band and needs to not have a lot
of phase shift added.

First  |  Prev  | 
Pages: 1 2 3
Prev: Caveats for Multi Layer Boards?
Next: As oscilloscope I'd like to see Dave Jones review