[Gb-ccb] suggested changes to ccb library (fwd)

[Martin Shepherd]

Sorry. This was meant to go to the mailing list, not just to John
(John - note that there is a correction at the end).

On Tue, 12 Aug 2003, John Ford wrote:
>...
> It is not a problem.  It is swamped by other delays, and as you point
> out, the start of the scan is limited to 1 second boundaries by the 1
> PPS tick.  I'm pretty sure the intent of the question was that we have
> no multi-second or minute long delays waiting for the darned thing to
> stop before we could start another scan.

The CCB integrates within 32-bit integers, so since we will be
sampling at 10MHz with a resolution of 12 bits, the maximum practical
CCB integration time, ignoring cal-diode delays, will be just a little
over 100ms before overflow. Thus the only reason that one might have
an integration period much longer than this would be if the cal-diodes
took a lot longer to settle than we are hoping.

[The aforementioned correction to this, is that in the 1cm receiver
 there are up to 4 32-bit accumulation bins, and thus the maximum
 integration period is actually 400ms. For the 3mm receiver with its
 single phase switch, the corresponding maximum is 200ms.]

Martin