I am replacing SHF LO boards, which put out frequencies between 540 and
580
MHz and start from a crystal around 100 MHz. It is easier to remove the
SHF LO and feed the transverter at
higher frequencies, as below.
For 2304, 720
for 3456, 828
for 5760, 702
|
|
16F688
|
Band
|
IF
freq |
LO
source |
LO
HEX |
144
|
28
|
116
|
116
|
222
|
28
|
97
|
97
|
| 432 |
28 |
288 |
288 |
432
|
144
|
578
|
578
|
903
|
50
|
|
|
| 903 |
144 |
758 |
758 |
| 1296 |
144 |
96 |
96 |
| 1296 |
144 |
192 |
192 |
| 1296 |
144 |
384 |
384 |
| 1296 |
144 |
768 |
768 |
2304
|
144
|
90
|
90
|
2304
|
144
|
720
|
720
|
3456
|
144
|
92
|
92
|
3456
|
144
|
184 |
184 |
3456
|
144
|
828
|
828
|
5760
|
144
|
93.6
|
93.6
|
5760
|
144
|
702
|
702
|
| 10368 |
144 |
106.5 |
106.5 |
| 10368 |
432 |
852 |
852 |
| 24192 |
144 |
100.2 |
100.2 |
| 24192 |
432 |
99 |
99 |
The PLL boards are much smaller than the original SHF LO, and there is
lots of room.
In this instance, there is no need to filter the output. The next thing
we try to do is
generate harmonics in the transverter, so it makes no sense, to me, to
throw away the power in the harmonics that might already be present.
If you are replacing a "Micro LO" this approach might work. However, if
you have a "MicroLO" then the "Apollo32" boards are a drop in
replacement.
On my 220
transverter, there was no separate board, so I removed the power to the
oscillator circuit and injected the LO into the filter right after the
point originally fed by the LO . Again no filtering was used in the PLL
output since the helical filters were already in place on the 220 board.
The SI41xx family of synthesizers have an interesting operation that
compensates for changes
over a narrow range. As the limits of the lock range get close to the
original settings a pulse is generated. By reprogramming the
frequency, the limits are recentered. This happens very fast, and is
evident as a glitch in the frequency. This should happen very
infrequently, and has never been observed in my 222 or 432 transverter,
which are in a stable environment .
PLL lock is monitored by the PIC even though it is sleeping. If an
event is generated, the PIC wakes up and lights the LOCK light,
reprograms the frequency, and if lock is achieved, turns off
the
light and goes back to sleep. If lock is not achieved, the
light
remains on and the program continues to try to achieve lock.
The PIC uses its internal oscillator which saves a few parts and some
board space. Having the PIC enter sleep mode keeps digital noise from
getting into the PLL.
The math that describes the register settings includes a number for the
"R" register. For the SI4133 family, with a 10 MHz reference input, 10
is the lowest number we can have which equates to an update rate of 1
MHz. Anything less than 10 will have more noise in the synthesizer
output. 5760 is a goofy frequency in the scheme of things, but can be
handled as 702 MHz for the "one board 5760 transverter". I have not
encountered any other cases that are this odd.
PROGRAMMING NEW FREQUENCIES
* marker freq with SI4133-GT
or SI4112-GT, NOT SI4113-GT
** marker freq is 1152.0 harmonic with SI4133-GT
or SI4113-GT, NOT SI4112-GT
This has value for DSP-10
owners and anyone else
with a need to convert 432 to 144. For the DSP-10 the apparent reversal
of
tuning can be handled in the "CONFIG FILE". Others will have to get
used to tuning backwards in LSB. Useful if you have higher bands (like
10, 24 or 48 GHZ) using a 432 IF frequency.
BACK