Thread Subject: Simulink: Modelling a DAC conversion and continuous time transfer function

Hi all,

I am in the process of learning simulink in order to model a signal
processing system for wireless RF communications, however I have hit a
bit of a dead-end on what must be a pretty simple problem.

To cut a long story short, I want to produce 4 level symbols, apply
pulse shaping (in a DSP), then convert to a continuous signal and apply
a filter that models an analog receiver circuit.

The 4 level symbols and pulse shaping are all done in the discrete time
world, with a symbol rate of 9600 and 6 samples per symbol. This is
working perfectly.

What I can't work out is how to:
A) convert the discrete signal to a continuous signal. Is it just a
ZOH+LPF? Or should I use some kind of generic DAC unit?

B) produce a continuous time filter to model the receiver circuit. I
have its frequency response, and I am quite capable of producing a
digital filter from this frequency response, but when it comes to the
continuous time world I am a little clueless as to what block elements
I should be using.

Any advice would be appreciated

regards,
Nick

I am also trying to model a DAC conversion in simulink. Basically I
have an OFDM signal around 8MHz BW. So this discrete signal is
converted into analog by first going into ZOH sampling 1/20M and then
a LPF with cutoff at 10MHz. So my question is whether this model a
DAC conversion or not. Any suggestion will be appreciated.

Thanks

npelly wrote:
>
>
> Hi all,
>
> I am in the process of learning simulink in order to model a signal
> processing system for wireless RF communications, however I have
> hit a
> bit of a dead-end on what must be a pretty simple problem.
>
> To cut a long story short, I want to produce 4 level symbols, apply
> pulse shaping (in a DSP), then convert to a continuous signal and
> apply
> a filter that models an analog receiver circuit.
>
> The 4 level symbols and pulse shaping are all done in the discrete
> time
> world, with a symbol rate of 9600 and 6 samples per symbol. This is
> working perfectly.
>
> What I can't work out is how to:
> A) convert the discrete signal to a continuous signal. Is it just a
> ZOH+LPF? Or should I use some kind of generic DAC unit?
>
> B) produce a continuous time filter to model the receiver circuit.
> I
> have its frequency response, and I am quite capable of producing a
> digital filter from this frequency response, but when it comes to
> the
> continuous time world I am a little clueless as to what block
> elements
> I should be using.
>
> Any advice would be appreciated
>
> regards,
> Nick
>
>

hi all
I need to convert a continuous sine signal to a 4 digit binary signal

thank u
npelly@gmail.com wrote in message <1103097289.775096.83700@c13g2000cwb.googlegroups.com>...
> Hi all,
>
> I am in the process of learning simulink in order to model a signal
> processing system for wireless RF communications, however I have hit a
> bit of a dead-end on what must be a pretty simple problem.
>
> To cut a long story short, I want to produce 4 level symbols, apply
> pulse shaping (in a DSP), then convert to a continuous signal and apply
> a filter that models an analog receiver circuit.
>
> The 4 level symbols and pulse shaping are all done in the discrete time
> world, with a symbol rate of 9600 and 6 samples per symbol. This is
> working perfectly.
>
> What I can't work out is how to:
> A) convert the discrete signal to a continuous signal. Is it just a
> ZOH+LPF? Or should I use some kind of generic DAC unit?
>
> B) produce a continuous time filter to model the receiver circuit. I
> have its frequency response, and I am quite capable of producing a
> digital filter from this frequency response, but when it comes to the
> continuous time world I am a little clueless as to what block elements
> I should be using.
>
> Any advice would be appreciated
>
> regards,
> Nick
>