[Gb-ccb] Caltech Backend Telecon Monday 07 July 4pm EDT

[Martin Shepherd]

On Thu, 3 Jul 2003, John Ford wrote:
> Martin Shepherd writes:
> I don't know if it's necessary or desirable to add optos into the
> analog signals.

Agreed. I think that for the detected signals, differential-pairs make
the best sense, and in this case I would have no qualms of following
Richard's recommendation of grounding the cable shields at both ends.

> Seems quite attractive for noise abatement on the
> digital signals.  If I recall the earlier meeting, we decided that the
> optos were an unnecessary feature.  Maybe we should do it, since it is
> pretty cheap and a definite win.

That would be fine with me. However, assuming that by opto-isolation,
we are talking about optic fibres, rather than opto-isolator chips and
wires, the feasibility of this would presumably hinge on the
availability of fibre-connectors with a metal screen going right up to
the borders of the fibre. I'll see what I can find.

>...
> In another mail, off the mailing list, I suggested that we could build
> the detector preamp electronics with a frequency response matching
> your digitizing system, i.e. build in the anti-aliasing upstream of
> your digitizing system.  Dunno if this is practical in this case The
> circuits we've used in the past would lend themselves to this.  I
> think we could do it.

It would be great if you could, especially if we do adopt the balanced
differential approach.  If the filtering were done in the receiver box
before the conversion to differential, this would eliminate not just
one, but three analog stages from each signal path in the backend,
because if the anti-aliasing filters were placed in the backend box,
the backend first would have to convert from differential to
single-ended to drive the filters, then back to differential to drive
the ADCs. With the filters moved into the receiver box, the backend
analog electronics would be trivial, involving no more than a simple
differential buffer amplifier between the signal connector and the
ADC. Giving this buffer a simple single-pole frequency roll-off ought
to be sufficient to suppress any small-signal high-frequency
interference and noise picked up in transit following the
anti-aliasing filter.

Martin