Richard, For the record, we never did a full mechanical study proving that an enclosure with the the two rectangles, plus faceplate, could be rack mounted without additional rail or shelf support. Anyone planning to do that needs to do a little homework.
Our racking setup is often stacks of (1 rack mount shelf with 2 Roach2 chassis on top of each other).
Thus, we could take some shortcuts on the impact of the extra holes for the ADC16x250 coax inputs. Matt On Fri, 7 Mar 2014, Richard Black wrote:
Date: Fri, 7 Mar 2014 11:53:13 -0700 From: Richard Black <[email protected]> To: Matt Dexter <[email protected]> Cc: casper <[email protected]> Subject: Re: [casper] Mounting ADCs in ROACH-2 Enclosures Matt, Thanks for the info. That answers my question. I was considering combining the two "small" rectangles into one longer one to accommodate the x64 ADC, but if it is structurally necessary for rack mounting and displaces the custom PCB, then I'll hold off on that. Thanks a ton! Richard Black On Fri, Mar 7, 2014 at 11:43 AM, Matt Dexter <[email protected]> wrote: Hi Richard, Assuming your chassis are affixed to rails or rest on shelves and ignoring all of the EMI/EMC issues that arise when making all sorts of holes in a metal enclosure it should be OK to have holes even larger than the two called out for the ADC16x250-8 coax rev2 input connectors. These holes, also shown in the attachment, reach down almost to the bottom of chassis - there isn't much farther to go that way. The holes could go a bit higher. There's a chassis fan on the left so there's not much room to go that way. The space between the 2 rectangles could be removed and the holes could be extend towards each other. The holes may increase a bit to the right but there's not all that much room due to the other set of chassis fans. Even the 2 "small" rectangles mentioned here forced the design of a custom small power and reset buttons + 5 LED board so that users would have most of the standard chassis front panel switches and LEDs. Perhaps there's a way to mount those on the extreme right hand side of the chassis. Or just remove them entirely. Matt On Fri, 7 Mar 2014, Richard Black wrote: Date: Fri, 7 Mar 2014 11:17:20 -0700 From: Richard Black <[email protected]> To: Matt Dexter <[email protected]> Cc: casper <[email protected]> Subject: Re: [casper] Mounting ADCs in ROACH-2 Enclosures Matt, That's good to know. Thanks for getting me that information. I failed to mention that our ADC boards are the coax versions, so the cutting measurements provided on the wiki page do not apply. I know that the page suggest cutting two small rectangles for the ADC input connectors, but is there any reason why we couldn't just cut a larger rectangle so the x64 ADC could fit as well? Thanks, Richard Black On Thu, Mar 6, 2014 at 2:30 PM, Matt Dexter <[email protected]> wrote: Hi Richard, The Roach2 enclosure changes I know about may be found at https://casper.berkeley.edu/wiki/ROACH_2_Enclosure I work with at least 4 different versions of the Roach2 enclosure. Each customized by us (the end-user) or by special arrangement with Digicom. Maybe others know of tricky mounting schemes that require no drilling/filing/... ? Matt On Thu, 6 Mar 2014, Richard Black wrote: Date: Thu, 6 Mar 2014 14:21:40 -0700 From: Richard Black <[email protected]> To: casper <[email protected]> Subject: [casper] Mounting ADCs in ROACH-2 Enclosures Hi all, I was just attempting to connect the 64ADCx64-12 into a ROACH-2 enclosure. I took off the front-panel that came with the ROACH-2 board to find a completely forged enclosure with the same holes drilled out as in the out-of-the-box front panel. At first glance, this makes the idea of mounting an ADC impossible without further drilling. We also have some ADC16x250-8 RJ45 cards with compatible front panels. With the current ROACH-2 enclosure, I do not see any way of fitting these ADCs into the enclosure either without drilling more holes. Was the expectation for the user to drill new holes into the ROACH-2 enclosure, or is there some way of mounting these ADCs that I'm missing. Thanks, Richard Black
