Документ взят из кэша поисковой машины. Адрес
оригинального документа
: http://www.naic.edu/~phil/hardware/dds/dds.html
Дата изменения: Tue Feb 15 22:24:41 2011
Дата индексирования: Sun Feb 3 19:59:17 2013
Кодировка:
Поисковые слова: пппп
The dds synthesizers are used for the doppler
correction during radar experiments. There are two synthsizers that are
drivenv by the dap1 computer in the da crate. The dap1 computer (mv162)
has 4 industry pack interfaces (ip interfaces). One of these has bill
sisk's custom ip board that drives the dds synthesizers (where are in
an exteral chassis).
Names/locations of some
software:
/home/phil/vw/diag/ipSynDg0[1-3].c
diagnostics for dds's. Access via rundiag (with cpu dap1) and
selelction 11.
/home/phil/vw/devs/ipSyn/ipSynDrv.c
vxWorks driver that interfaces to ip card.
/home/phil/vw/datatk/dds/ddsProg.c
datatking program that runs on dap1. It does the doppler
updates during a datataking run.
Documentation:
Bill Sisk's documentation:
Notes: this was copied from maestro server , bill
sisk's disc on 15feb11.
14feb11 -
ddsProg stops doppler updates during moon430 run
The 430 Mhz moon bistatic observing had trouble with
the data on 12jul10. Looking at the datataking log file
(/share/obs4/msgs/dtmsgs.2010) the ddsProg had some messages saying
that it was late for interrupt. These errors occurred on:
10jul10, 11jul10,12jul10,29oct10, and 31oct10.
When the error occurs, the ddsProg stops the
doppler updates and waits for a command to start another look (or
run). For asteroids this will occur every roundtrip time. For the moon
430 this is a bistatic run and it lasts for up to 2 hours.
The dds update rate (dopipp) for most radar runs is
2000 usecs. For the moon430 runs it is set at 1000 usecs. On 15feb11 i
ran the diagnostic program ipDgSyn02.c to check on how short an
ipp would work:
ipp=1000 usecs. Usually missed an interrupt within a few minutes
ipp=1500 usecs. missed interrupts
ipp=1800 usecs. ran for an hour or more without missing an
interrupt.
The test program has less to do between interrupts than the
actual datataking routine.
the datataking routine must evalutate an Nth order polynomial
at each ipp.
The test program uses a linear polynomial to compute the
frequency step.
The solution to the problem is to change the ipp
update rate of the moon datatking to 2000 usecs.