thanks a lot for the photos, they are very helpful. This method looks very
I would have to talk with the people in our machine shop, but I’m confident that
they can make the modifications.
Thanks also to Dan and Francois.
From: Jonathan Weintroub [mailto:jweintr...@cfa.harvard.edu]
Sent: Montag, 19. September 2016 18:45
To: Guenter Knittel
Cc: casper list; Dan Werthimer
Subject: Re: [casper] Roach2 enclosure and power supply
Dan’s absolutely correct, the enclosure was designed to be “RFI-aware” for
We (SAO), have recently decided to mount the power supplies in just the way you
propose. See photo attached, of an early prototype.
We have worked with Digicom to produce recent ROACH2 assemblies according to
this scheme using modifications to the enclosure. Digicom subcontracts these
modifications to a local machine shop, and we pay extra for them. (We had
prior modifications to the case, hence the residual circular cutout and
additional screw holes which would not be there if a virgin enclosure was the
We are motivated by improved cooling rather than space because our telescope is
at 4000m. We operate at ~1mm wavelength so emissions from the ROACH2 are out
of band for us so we can compromise the RFI characteristics of the case without
impact in our specific application.
The power supply in question is FSP250-60FAG which was recommended for ROACH2
by Alex Rust in a note to this mailing list on 6 July this year. It is a tag
higher than the iSTAR it replaced. A benefit to the modification is we can
mount it with screws to the back panel.
There are some drawings which we can supply, or Digicom may be able to help
On Sep 19, 2016, at 5:42 AM, Dan Werthimer <d...@ssl.berkeley.edu> wrote:
i think this was designed for RFI reasons,
to enable installation of a 220/110VAC input module that has good LC filtering,
to keep RFI from getting out through the line cord.
On Mon, Sep 19, 2016 at 8:22 AM, Guenter Knittel <gknit...@mpifr-bonn.mpg.de>
we have a Roach2-unit and consider adding a few components into the case.
I wonder: why wasn’t the power supply moved flush to the rear panel, with a
large cutout for power connector and fan? Was it because it is just a little too
much in height?
Because if one did, there would be a lot of free space in the enclosure.
Any info appreciated