Title: 3. Digital Implementation of Mo/Demodulators
13. Digital Implementation of Mo/Demodulators
2General Structure of a Mo/Demodulator
3Single Side Band (SSB) Modulator
MOD
SSB
4Implementation using Real Components
SSB
where
5Single Side Band (SSB) Demodulator
DEM
LPF
6Single Side Band (SSB) Modulator in Discrete Time
Modulator Implemented in two stages
Digital Up Converter
DUC
Analog MOD
ZOH
DISCRETE TIME
CONTINUOUS TIME
7Single Side Band (SSB) Demodulator in Discrete
Time
Demodulator Implemented in two stages
Digital Down Converter
DDC
Analog DEM
ZOH
DISCRETE TIME
CONTINUOUS TIME
8Digital Down (DDC) and UP (DUC) Converters
DUC
DDC
RF
Baseband
- MHz for voice
- GHz for data
- kHz for voice
- MHz for data
Order of magnitude of resampling
9Problem with Large Upsampling Factor
if M is large, very small transition region
high complexity filter
10Problem with Large Downsampling Factor
LPF
LPF
if M is large, very small transition region
high complexity filter
11Solution Upsample in Stages
In order to make it more efficient we upsample in
L stages
12i-th Stage of Upsampling
13Example Upsample in One Stage
This is not only a filter with high complexity,
but also it is computed at a high sampling rate.
14Same Example in Three Stages
Total Number of operations/sec
a 95 savings!!!!
15Downsample in Stages
16i-th Stage of Downsampling
noise
keep aliased noise away from signal
17Example Downsample in One Stage
18Same Example in Three Stages
Total Number of operations/sec
a savings of almost 99 !!!
19Stages at the Highest Rates
highest rates
- the highest sampling rates are close to carrier
frequencies, thus very high - properly choose intermediate frequencies to have
simple filters at highest rates
20Last Stage in UpSampling
wide region
21First Stage in DownSampling
wide region
22Very simple Low Pass Filter the Comb Integrator
Cascade (CIC)
same!!!
Comb
Integrator
these two are the same!
Notice no multiplications!
23Frequency Response of the Comb Filter
like a comb!
24Impulse Response of the CIC
interpolating sequence
25The CIC in the Time Domain
like a discrete time ZOH!
26Two Important Identities The Noble Identities
Same !!!
As a consequence we have one of two Noble
Identities
Same!!!
27Other Noble Identity
Same !!!
As a consequence we have the other of the two
Noble Identities
28Efficient Implementation of Upsampling CIC
Use Noble Identity
Very simple implementation (no multiplications)
29Efficient Implementation of Downsampling CIC
Use Noble Identity
Very simple implementation (no multiplications)
30Frequency Response of the CIC
Not a very good Low Pass Filter. We want a better
attenuation in the stopband!
31Put M Stages together
Frequency Response
32Improved Frequency Response of CIC Filter
Resampling Factor N10
With M4 or 5 we already get a very good
attenuation.
33Example M4 Stages
34Implementation of M Stage CIC Filter Upsampling
Use Noble Identity
35Implementation of M Stage CIC Filter Downsampling
Use Noble Identity
36Problem DownSampling CIC is Unstable
Now we have to be careful the output of the
integrator will easily go to infinity
37CIC Implementation.
At the p stage
This implies
and
38If we use Q bits for the integrators then we
need to guarantee
Let the input data use L bits
Then
decimation factor
input bits
number of stages
39Application Software Defined Radio
- Definitions
- Software Defined Radio modulation, bandwidth
allocation all in software - Field Programmable Gate Array (FPGA)
reprogrammable logic device which is able to
perform a number of operations in parallel. They
can process data at a rate of several 100s of MHz - DSP Chip optimized for DSP operations by some
hardwired ops (such as multiplies).
40An HF SSB Software Defined Radio by Dick Benson,
The Mathworks,
64MHz
15.6kHz
7.8kHz
RF
IQ
AUDIO
Rec.
Rec.
Rec/Tr
Trans.
Trans.
DAC
IQ
AUDIO
RF
FPGA
DSP Chip
41Transmitter
AUDIO
I
FIR
SSB
Q
FIR
DSP Chip
Xilinx Library Modules
I
FIR
FIR
CIC
RF
Q
FIR
FIR
CIC
FPGA
42Receiver
Xilinx Library Modules
I
RF
CIC
FIR
FIR
Q
CIC
FIR
FIR
FPGA
I
FIR
AUDIO
Q
FIR
DSP Chip