From: "Peter Onaka" <onaka@ifa.hawaii.edu>
Date: Apr 2, 2013 8:47 AM
Subject: iSHELL review feedback
To: "Donald N.B. Hall" <hall@ifa.hawaii.edu>
Cc: "Jay Elias" <jelias@noao.edu>, "Alan Tokunaga"
<tokunaga@ifa.hawaii.edu>, "Randy Chung" <rchung@ifa.hawaii.edu>,
"Doug Toomey" <toomey@mkir.com>, "Mike Mumma"
<michael.j.mumma@nasa.gov>, "David Warren"
<telecentric@sbcglobal.net>, "Klaus Hodapp" <hodapp@ifa.hawaii.edu>,
"Dan Jaffe" <dtj@astro.as.utexas.edu>, "John Rayner"
<rayner@ifa.hawaii.edu>, "Eric Warmbier" <warmbier@ifa.hawaii.edu>,
"Darryl Watanabe" <watanabe@ifa.hawaii.edu>, "Ken Cheung"
<kcheung@oceanit.com>

Hi All,
 Sending out a quick write up:


Issues:
=========

1. Suggest an operational lifetime be defined in the "Facility
Requirements" section of the SRD. This will help define a flow down
requirement for equipment spares.

2. A spares policy should be defined and cost budgeted to make sure
technology that will leave the market (PCIe, parallel ports) will be
supported throughout the instrument lifetime.

3. Disabling the Kaman displacement position sensor emitter during
detector readouts will require an RF (preferably terminated with an
appropriate load) cutoff. This cutoff should be placed as close to the
control electronics and as far away from the cryostat as the sensor
wiring will allow.
  Kaman transmission frequencies appear to be in the 500Khz to 1Mhz
range and will couple to the detector wiring....

4. If it does not require too much effort, the option to have the
detector controller readouts interleaved (or the slit guider held off)
should be implemented.


 The typical IRTF mechanical development *does not* include the
detector (and all sensor) wiring paths in the 3D model. Also, as
illustrated in the Instrument Controller block diagram, the system
power supplies are co-located in the cooled racks at the MIM (we have
seen sub-read noise, intermittent pattern noise on other multi
detector instruments with decent physical separation between the
detetctors). The IRTF has been able to achieve good noise performance
with multiple detector instruments before, but component and cabling
shielding, routing and design has always been critical. The cryogenic
ribbon cabling is not shielded. Historically, the final actual wiring
was done by the instrument technical (with guidance from the
electronic engineer).

 If this option is not supported then the 3D mechanical model should
include all the detector and sensor wiring.

Discussion:
===========

1. What are the H2RG cable length differences between NSFCAM2 and ISHELL?

2. How is the GPS timestamp requirement of 1usec going to be implemented?


Comments:
==========

1. The IRTF approach of strict minimization powered devices in the
cryostat when doing detector readouts should be maintained.

2. The thermal shielding on the Phytron motors should be developed/reviewed.


Regards,
--
Peter Onaka
Institute for Astronomy
tel 808-956-7747
onaka@ifa.hawaii.edu