Possibly OT: 1U 'rack'

[Part Timer]

I know many here have experience of sound and computer installs.

I'm looking for a way to 'retain' a 1U high pair of UHF Ch70 radio mic receivers horizontally on a solid surface without actually placing them in a 'real' rack. This is to enable access by the users (not me) to turn them on/off (when mains power is turned on at the wall warts they require their front panel buttons pressing to turn them on). Also they have fixed aerials, which would otherwise be shielded inside a rack. Otherwise I would hide them out of the way on long mains and XLR cables.

I have found the following items so far:

receivers measure 16.5cm deep. Does a variation of the above items exist where the unit is 'boxed in' for a few cm of its depth, thus preventing the back from being lifted?

Not after high security, just to prevent them walking from a room which is locked out of hours but may not be watched over _all_ the time. I plan to use 'security' bolts on the rack ears.

[The Natural Philosopher]

lots people make cases out of wood: a pair of wood battens form the rail and mirror screws can hold the kit in..

[Theo Markettos]

off the legs to the appropriate length and bolt them to the surface. (Note that Lack tables are mostly made of paper honeycomb, so I wouldn't expect them to be as strong in the vertical direction afterwards)

Theo

[Jeremy Nicoll - news posts]

Neat, but a bit alarming from a cooling point of view, with some units shown mounted right up against the underneath of the table.

[dennis]

"Jeremy Nicoll - news posts" wrote in message news: [e-mail]...

Rack mounted stuff has to be able to cope with other heat producing stuff stacked above and below it. Sitting on or below a surface should be safe, if it isn't its time to send the stuff back with a very stroppy letter.

[Steve Firth]

Heck, add datacentre design to the long list of subjects that Dennis knows nothing about.

[Man at B&Q]

e

In what way is the underneath of the table "heat producing"?

No, it's time to admit you don't have clue what you are talking about.

MBQ

[Tim Watts]

Except (unusually) I agree with him.

The table is not heat producing, but does impede airflow over the equipment surface which I believe is what Dennis was alluding to. This should not be a problem as rack equipment, at least in the computer realm, assumes it will be fitted beween other kit (and, worse than having no cool air flow over the flat faces, it will actually be subjected to some heat from its neighbours). Thus it can be assumed to have forced air cooling, in front, out rear or utilise so little power that it does not need to.

The only issue is wheather the thing the kit is bolted to is a liability from a flammability POV - not a problem I would worry about with computer kit, but the OPs kit should be checked (ie how likely is it to catch fire in the event of a fault?).

[NT]

All you need for rackless rack mounting is a pair of steel brackets. Position them like book ends, bolting the mounting points on the equipment to them. You will need to drill extra holes in them. I'd get some heavy duty ones, not flimsy london type. For security, put epoxy on the threads when you do the bolts up.

NT

[dennis]

Heck Steve obviously hasn't seen a rack. Its bloody obvious that you put stuff above and below them or they wouldn't be rack mount.

BTW I did do some data centre design but as it was a few years ago I wouldn't want to do it again. I don't even know if the companies still exist, the products wont.

I often wonder what people like Steve do for a living, they obviously don't do much in the way of variety or they wouldn't think the number of things I have done was odd. Where I worked very few did the same thing over and over again, just the programmers really. Software is pretty much the same now as it was a decade ago, hardware and systems design was always changing. I guess that's the problem with being semi-skilled like a plumber/electrician no variety so no gain in skills.

[Part Timer]

Each receiver runs off a 12V 500mA wallwart (these will be hidden away). So not much heat generation. The issues with positioning the receivers is access to turn them on easily and their ability to pick up RF. Wasn't thinking Ikea or Dennis' Datacentre!

[dennis]

Who said it was? Is it a hot plate for keeping food hot?

So where do you think rack mount stuff vents its heat too then?

Go on admit that you are venting yours from your arse.

[Tim Streater]

This is all quite correct. Something like a large Juniper router, fully loaded, might be drawing 90 amps at 48VDC, though more than one power supply. Forced air cooling is how that's dealt with, in the front and out the back. That's why it's so effing noisy at a CoLo.

In the case of 1U high stuff, a fully loaded rack could well have these in every U position, so a 47U high rack (2200mm) might have 47 1U high devices in it. Each is cooling itself - no convection cooling.

[Tim Watts]

And in the worst case I have personally seen, 42 Dell servers doing flat out computational work (= max power draw) in a single tall rack had a total power consumption of 13.8kW (I measured it).

That gets warm at the rear!

The best cooling arrangement I have ever seen with my own eyes was pumped liquid CO2, supplied to an evaporator coil with fins covering the whole of a secondary rear door to a similar rack. Effectively air conditioning with the evaporator at the immediate point of use. And lots of extra fans on the door...

The air coming out the back of that lot was not far above ambient. Very impressive way to cool dense clusters, and eye wateringly expensive I was told (they only had 8 IIRC racks with this arrangement for special high power density loads). Not surprising given the arrays of high pressure distribution pipes, drop hoses (very thick and reinforced) and control/monitoring equipment it needed.

Water cooling is becoming popular too as a cheaper alternative, either to similar "special rear doors" or to coils in covered "ceilings" that span to rows of racks forming a covered hot aisle. Both rows are arranged to point the rear of the equipment into this space which is made fully enclosed with doors or plastic curtains at each end and again, large fans assist with drawing the hot air up through teh coils and fins. I'd hate to have to work in that space for long given it could easily get to the high 30Cs

In all cases, some regular aircon remains installed to provide cool ambient air to the fronts of the racks.

[The Natural Philosopher]

I think he probably is an IT support manager for a major bank: What this means is he has a mobile phone with a list of numbers to call when something goes wrong. Then he takes the credit for getting it fixed.

[Steve Firth]

[snip]

It'll never happen.

[The Natural Philosopher]

And no doubt all running windows..

I just accepted that a 64 bit Intel Atom was in fact about 10 times faster than the 486's and low end SPARCS I used to use that were perfectly adequate - and now have a 30W server that does everythng I want..

CPU cycles per watt is a very highly used quality when picking servers, these days.

[Owain]

You could use metal U-bars over the items, with the ends going through the surface and bolted on the underside. Or bolt the items down with builders band.

Go and see how they bolt display items down to the perforated metal shelves in Argos for the method.

Owain

[Man at B&Q]

[e-mail]...

That depends on the design of the "stuff" and the "rack".

What it will not do is vent through an impermeable mounting face.

I have come across a fair few rack that vent the heat through their arses, so to speak.

MBQ

[Tim Streater]

Unlikely I'd have thought. I would have expected linux in general.