[time-nuts] 3.0GHz Channel 3 installation in Agilent 53132A counter
dave.martindale at gmail.com
Wed Mar 19 11:56:00 EDT 2014
No, the suggestion makes perfect sense. A 12 GHz input board likely
has an internal prescaler with a division ratio that is 4 times that
of a 3 GHz input board, in order to have the main counter running at
the same internal frequency. For example, a 3 GHz input might divide
by 32 while a 12 GHz input might divide by 128, so that both feed a
signal in the 0 to 100 MHz range to the main counter circuits. (And
the 5 GHz input might divide by 64...)
When the firmware goes to display what the main counter measured, it
has to multiply by the prescaler ratio to obtain the original input
frequency. If this multiply happens on the main board (e.g. in a FPGA
or CPU), it needs to be told which prescaler division ratio was used.
If the firmware thinks the 12 GHz input is installed instead of the 3
GHz input, it will multiply by 128 instead of 32, and the displayed
frequency will be 4 times the actual frequency (since the prescaler
actually divided by 32).
(The prescaler division ratios may not be 32 and 128; the actual
values don't matter. As long as the 12 GHz option uses 4 times the
prescaler divisor of the 3 GHz option, the display error will be a
factor of 4).
On Wed, Mar 19, 2014 at 11:26 AM, James Robbins <jsrobbins at earthlink.net> wrote:
> Someone suggested that maybe the new main board had been set up for the 12.5 or 6GHz channel 3 but was sold without that channel. The idea was that such a main board would cause a 4x reading. To my mind this is opposite to what I would think in that the division ratio for 12.5 or 6GHz would be higher than the ratio for the 3.0GHz board and would result in a fraction of the frequency rather than 4x frequency.
> and follow the instructions there.
More information about the time-nuts