HSP50214 Datasheet, PDF (28/54 Pages) Intersil Corporation – Programmable Downconverter

HSP50214
Serial Direct Output Port Mode
The Serial Direct Output Port Mode offers the ability to construct
two serial output data streams, SEROUTA AND SEROUTB, from
16-bit I, Q, magnitude, phase, frequency, timing error, and AGC
level data words. The total number of data words (1 to 8) for serial
output, and the sequential order of these data word components
of the serial output are programmable. Each data word may be
used once in either the SEROUTA or SEROUTB data streams.
Figure 31 illustrates the conceptual implementation of the Serial
Direct Output Port Mode.
In the Serial Direct Mode, the output data is loaded into serial shift
registers and routed to two serial output pins, SEROUTA and
SEROUTB. The serial output shift clock, SERCLK, is PROCCLK
divided by 1, 2, 4, 8, or 16. The divide down ratio is programmed
using Control Word 20, bits 14-16. The data is shifted out on the
rising edge of the internal SERCLK. The external clock polarity of
SERCLK is programmable via Control Word 20, bit 18. A sync sig-
nal is provided for detection of the start or end of each word in the
serial sequence. Control Word 20, Bit 17, sets the SERSYNC sig-
nal location as either preceding the MSB (typical for interfacing
with microprocessors) or following the LSB (typical for interfacing
to D/A converters). Control Word 20, bit 19, sets the SERSYNC
polarity as active low or high. The LSB of each data word can be
configured as either the true LSB data, or set at a fixed logic “1” or
“0” for use as a tag bit. Control Word 20 bits 0-13 set the LSB of
each of the 7 types of data words that can be configured in the
serial output stream. Control Word 19, bits 21-24 set the number
of serial data words that will be linked to form the serial outputs.
Up to 7 data words can be linked to form the serial output.
SEROUTA and SEROUTB will have an identical number of words
in the serial output streams.
The 16-bit I, Q, magnitude, phase, frequency, timing error, AGC
level, and “zeros” data words for are loaded into their respective
shift registers. The Magnitude and AGC Level data word are
unsigned binary format with a leading zero, while the remaining
signals are 2’s complement format.
Any of the eight data sources can be selected as the first serial
word for SEROUTA or SEROUTB. Control Word 19, bits 25-30
set the data type for the first serial word for SEROUTA and
SEROUTB. The three bit data type identifier is shown both in
Table 13 and in Figure 30, to the right of the controls for the cross
matrix switch. Serial output data word sequences are formed by
linking data words by programming the data source for each shift
requester’s shift input signal. This programming links the shift reg-
isters together in one or two serial chains. Thus the Control Word
19 term “Link follows X data”, where X is one of the seven data
types. Once the data source data word is selected (by program-
ming a three bit word representing one of the data types into Con-
trol Word 19, bits 25-27 (SEROUTA), and 28-30 (SEROUTB)), the
process for identifying the next word is to select a three bit data
type identifier which represents the data type to follow the source
data type. Program these bits into the Control Word 19 field repre-
senting the “Link following X data”, where X = the source data
type, defines the second word in the sequence. Likewise, the third
data word is linked by selecting the Control Word 19 bits that
identify the “Link following X data”, where X = the data type of the
second word in the serial chain. The process continues until all
the desired data words have been linked.
NOTE: I and Q are sample aligned in time. |r| and f are sample
aligned in time, but one sample delayed from I or Q.
The frequency sample is delayed in time from I or Q by
1 sample time + 63 tap FIR impulse response. If the FIR
is set to decimate, the FIR output will be repeated every
sample time until a new value appears at the filter out-
put. (i.e., the frequency samples are clocked out at the
I, Q sample rate regardless of decimation.)
TABLE 13. LINKING CONTROL WORDS FOR SERIAL OUTPUT
DATA TYPE
IDENTIFIER
DATA TYPE
000
I Data
001
Q Data
010
Magnitude (MAG) Data
011
Phase (PHAS) Data
100
Frequency (FREQ) Data
101
Timing Error (TIMERR) Data
110
AGC Gain
111
Zeros
Two examples will illustrate the process of configuring a serial
output using the Serial Output mode.
Serial Output Configuration Example 1:
It is desired to output the I data word, followed by the Q data
word, followed by the Phase data word on the SEROUTA output.
Similarly, it is desired to output the Magnitude data word followed
by the Frequency data word, followed by the Timing Error data
word, followed by the AGC Level data word on the SEROUTB out-
put. Table 14 illustrates how Control Word 19 should be pro-
grammed.
TABLE 14. EXAMPLE 1 SERIAL OUTPUT CONTROL SETTINGS
CONTROL
WORD 19
BIT POSITION
FUNCTION
BIT
VALUE RESULT
30-28
SEROUTA Data Source 000
(I)
27-25
SEROUTB Data Source 010
(|r|)
24-21
Number of Serial Word
100
(4)
Links in a Chain
20-18
Link following I data
001
(Q)
17-15
Link following Q data
011
(φ)
14-12
Link following |r| data
100
(f)
11-9
Link following φ data
111
(Zeros)
8-6
Link following f data
101
(Timing)
5-3
Link following AGC data XXX
(N/A)
2-0
Link following Timing
Error data
110
(AGC)
NOTE: Because all but the first data word in the serial output
is identified by the data type that it follows, SEROUTB
can only be fully independent of the sequence in SE-
ROUTA if it does not use any of the same data word
types. This implies a partition as described in Example
1. Once a data word that is used in SEROUTA is called
out in SEROUTB, the remaining sequence in SEROUTB
will be identical to that portion of SEROUTA sequence
that follows the duplicate data type. This follows from
using the “Link follows ‘data type’ data” for word link-
age.
NOTE: Each type of data word should be used only once in
each data stream. If the “Link following I data” is pro-
grammed with the data type identifier for I, then the
part will repeat the I data word until all of the data word
locations are filled. In Example 1, if bits 20-18 were er-
roneously programmed to 000 (I data) then the SEROU-
TA would be four sequential repeats of the I data word.
28