[Gb-ccb] Single supply operation of the AD8138

[Martin Shepherd]

While in the process of updating my CCB external interfaces document
to reflect changes that have been proposed in subsequent discussions,
I encountered a problem.  Randy's suggestion of using a single power
supply for the AD8138 in the CCB, instead of the dual +5V and -5V
supplies that I originally proposed, seemed like a good one,
especially wrt being able to get rid of the ADC clamping diodes and
some associated liabilities.  Unfortunately, it turns out that this is
potentially problematic, because in single supply mode, the AD8138
data-sheet says that its DC input voltages can only swing between 1V
and 3.8V. While in theory this is enough to accomodate the 2.5V swing
arriving on each line of each of the differential pairs from the
receiver, to do so would require that the common mode voltage arriving
at the inputs to the CCB be 2.4V +/- 0.15V, whereas the common-mode
voltage following the matching resistors at the outputs of the
receiver, in the currently proposed video driver circuit, is just
1.25V.  Given that before these matching resistors, the common-mode
output voltage is double that seen after the matching resistors, to
present the correct 2.4V common-mode voltage at the input to the CCB,
would require increasing the common-mode output of the AD8138 driver
in the receiver to 4.8V, while maintaining the 2.5V swing of each
output above and below this voltage. In the current video driver
circuit this isn't possible, because the +5V supply limits the output
of the AD8138 to less than 5V.

Having just updated the power-management circuit, the analog input
circuit, power-connector pinout diagram, and the text of my document
to accomodate the removal of the -5V supply, among other changes, it
would be great if we can find a way to resolve this problem without
having to revert to using dual +/-5V analog power-supplies in the CCB.

One possibility would be to offset the signal inputs of the AD8138 in
the receiver, to make them always positive. This would allow the
receiver AD8138s to be powered from a single positive power-supply.
According to the data-sheet, in this mode the power-supply voltage can
be up to 11V, which would be enough to accomodate the required 4.8V
common-mode output voltage.

This also brings up a question about clamping. In Randy's circuit, the
outputs of the receiver AD8138s are clamped at the input to the
cables. Since the parts that will break if the input voltages go out
of range, are the AD8138s in the CCB, doesn't it make more sense to
place clamping diodes at the CCB inputs than the receiver outputs?

Also, since I have brought up the subject of power-supplies in the
CCB, is NRAO or Caltech now responsible for the power-management
circuit, now that the responsibility for the main PCB has changed
hands from Caltech to NRAO?

Martin