Since the RHIC Engineering Run in 1998, several issues have surfaced that make the current setup
less suitable for MVD's purpose.

A. Current setup

There are four pin-diode modules on each of the PHENIX nosecones, for a total of eight.
The powers and output signals to each module are carried by a ribbon cable that connects to
the module via 10-pin connector.  The ribbon goes to a passive junction box located on top
of a cable tray on top of the Beam-Beam Counter rack on the west side of PEH.  The junction
box is connected to the counting house by a multi-conductor copper cable that runs 150-plus
feet in length.  This cable carries both powers and signals, and is 150-plus feet in length.
In the counting house, the cable is connected to a power supply/signal junction box.
This box was located in the RHIC Controls Rack, and was powered by the 208VAC.
The signal from a pin-diode module was simply fed into a discriminator so that it would
give a NIM logic signal whenever the pin-diodes were hit.  The NIM signal was fed into
either a visual scaler or the RHIC's scaler readout system.  Four of the eight modules were
instrumented in this fashion.

STAR Experiment has a similar setup, except their power supply/signal junction box sits
in the experimental hall behind the shielding.  Also, their signal gets converted to optical signal
at the junction box, and goes to their counting house, where it gets converted back to logic
signal and fed into the RHIC scaler readout system.

B. Problems with the current setup

The power supply sitting in the counting house is not configured to satisfy the RHIC safety requirements.
During the Engineering Run, the radiation monitoring system was given a go-ahead  since it was
virtually identical to the STAR's system, and that was RHIC-approved.  After the ER, however,
closer scrutiny of the power supply by RHIC determined that it is not OK.  The fact that STAR'r
power suppy is OK may have to do with the fact that STAR's power supply sits inside the shielding
while ours sits in the counting house, where there is lot more people traffic.

The power supply itself is not reliable.  While we did not have the power supply turned on long
enought to notice any failures, the STAR's was, and they had problems.  And the power supply
being behind the shielding makes any fix impossible while the interaction region (IR) is secure.

Connecting the IR to the counting house with copper is a big no-no due to ground loop problems.
Thus, an optical link must be installed.  This means that analog-to-optical junction must be made
in the IR, and reverse in the counting house for eight channels.  Optical fibers (8) must also be laid
between the IR and the counting house.

Readout and archiving of the signals might be a problem.  RHIC Controls does have a readout
system in place  to read the scaler readings from the Zero-Degree Calorimeter (ZDC).  However,
during the ER, the readouts were not archived for later analysis.  It is not clear at this moment whether
RHIC plans to do this or not.  For MVD, recorded data with time stamps are critical in order to gauge
the radiation danger for the detector.

C. Proposal

Given the problems listed above, I propose the following idea.  It's actually the original idea that we
considered -- and that is, make the radiation monitor be part of the MVD ancillary system.  By this I
mean utilizing the powers available to the ancillary system, and also use the readout system.  The pin-diode
modules are very simple to operate.  Each module requires +5VDC (~45mA), -5VDC (~72mA),
and +25VDC (~4mA) for power.  The output signal is a positive TTL (50 ohm).  Both powers and
signal are carried to and from the module via 10-pin ribbon cable.

I propose that the +25VDC power be tapped out from the MVD High-Voltage system.  A full MVD requires
4 LeCroy 1461P-M410 modules, and we have 6 total.  Each module has 12 independent channels.  The
1461P variant M410 has a current limit of 0.256 mA, far less than the required 4 mA for the +25VDC.
The 1461P module, however, can be reconfigured to provide higher currents to satisfy the 4 mA
requirements.  The power can be derived directly from the LeCroy crate, or from the HV distribution
crate.

The +5VDC and -5VDC can be tapped out of the MVD Low-Voltage system.    These powers can
be derived from the LV distributiion crate.

The readout will be done along with all the other MVD ancillary readouts.  Appropriate VME module
will have to be found, that will allow scaler values to be converted to ADC or read out directly.

The big advantage of this proposal is that everything resides in the MVD ancillary crate, close to the
nosecones where the pin-diodes are located.  The cabling of these modules can also be bundled
together with the other MVD cables.  The other advantage is that the system will be totally under the
MVD control without having to rely on other systems for readout or power.