Rack Planning for Video Equipment

Budget rack units, including the ones you will need later

Count units for the equipment you are buying, then add for what the installation will accumulate: patch panels, a shelf, spare capacity for the channels you will add next year. A rack that is full at commissioning is a rack that gets extended badly, because the next device goes wherever it fits rather than where it should.

Form factor drives the count. Box-style encoders are compact but consume units fast if you have many of them, and each brings its own power adapter. Chassis-based systems concentrate them: a 3U frame accepts multiple encoder cards and handles power and cooling centrally, so sixteen channels occupy three units instead of sixteen. The ZY-EH1416-3U is the chassis approach; the ZY-EH1411-1U is a single-channel 1U unit for installations that want one channel per unit with a front-panel display.

A useful middle path is the modular frame: a 1U chassis that accepts one to four cards, which lets you start at one channel and grow inside the same unit. That is what the ZY-EHP1414 is for. Decide which of the three patterns you are building before ordering, because they are not interchangeable later.

Depth, rails and the clearance people forget

Rack depth is the constraint that gets discovered on installation day. Check the depth of the deepest device, add the bend radius of the connectors at the rear, and confirm the rack is deeper than that sum. Cables that have to be folded to shut the rear door will fail, and they will fail intermittently.

At the front, leave room to read the device. Front-panel status is genuinely useful during commissioning and fault-finding: the ZY-EH1411-1U displays IP address and input resolution on the unit, and the ZY-DH901 shows device name, address, resolution and working status. A display you cannot see is a display you will not use, and you will end up connecting a laptop instead.

Power: one supply or many adapters

The choice is between distributed adapters and centralised supply, and for anything beyond a handful of devices the answer is centralised.

• Individual adapters are fine for one or two boxes on a shelf. At ten devices they become ten failure points, ten cable runs into a power strip, and a tangle nobody wants to trace.
• A rack power distribution unit with a documented per-outlet budget gives you one place to measure total draw and one place to look when something is dead.
• Chassis-based power goes further: the frame supplies the cards, so there is one mains connection and one place to add redundancy.

Budget power by adding nameplate ratings, then confirm the measured draw once running. Leave margin — supplies run hotter and age faster at the top of their range, and the measured figure is what you should record at handover, not the sum of the labels.

Where the site has it, Power over Ethernet removes a cable per device. Several encoders here offer PoE as an option, which is worth specifying when devices sit away from the rack, in ceiling voids or at remote positions where mains is inconvenient to run.

Redundancy: what is worth duplicating

Duplicating everything is expensive and rarely necessary. The useful split:

• Worth duplicating: the supply to a chassis that carries many channels, and the uplink feeding a whole zone. A card-based 3U frame with dual redundant power protects sixteen channels for the cost of one extra supply.
• Usually not worth duplicating: individual field devices. Hold a configured spare instead. Swapping a decoder is faster and cheaper than maintaining a redundant path to every screen.

Where a chassis is the design, confirm whether the frame supports redundant supply before ordering it, and label the feeds A and B physically so a single circuit failure cannot take both.

Cooling: airflow is a direction, not a fan

Racks fail thermally because of airflow path, not fan count. Cool air enters low at the front and warm air leaves high at the rear; anything that lets exhaust recirculate to the intake raises the temperature of everything above it.

• Fill blanking panels. Gaps between devices let hot exhaust wrap around to the front. It is the cheapest thermal improvement available and the most often skipped.
• Load heavy devices low. Heat rises; putting the hottest equipment at the bottom keeps it out of the intake of everything above.
• Do not block rear exhaust. A rear door with no perforation, or cables bundled across the exhaust path, does more harm than an undersized fan.
• Check intake temperature, not room temperature. The thermostat on the wall is not the air the equipment breathes. Measure at the rack front, low.

Chassis systems help here too: one frame with centralised cooling is easier to keep cool than sixteen boxes each with their own convection path and their own dust intake.

Cable separation and bend radius

Keep copper data and mains physically separated, and give fibre its own route. Fibre is immune to interference, but it is not immune to being crushed: respect the minimum bend radius and do not zip-tie it tight against a rail.

Practical rules that hold up:

• Run mains down one side, signal down the other, crossing only at right angles where crossing is unavoidable.
• Secure bundles with hook-and-loop rather than tight plastic ties, which deform cable and are hostile to later changes.
• Leave a service loop at each device so it can be pulled forward for service without unplugging.
• Specify cable category explicitly and check what arrives. Copper-clad aluminium sold as Category 6 does not deliver the same reach; see Twisted Pair Cable Basics.
• Clean and cap optical connectors. Contamination is the leading cause of intermittent fibre faults and it is entirely avoidable.

Labelling and serviceability

Label both ends of every cable, at the rack and at the device, with the same identifier. Record device name, address, position and function in a schedule kept with the rack. This is the difference between a twenty-minute fault and a day of tracing.

Choose hardware that supports remote service: a web interface, remote management and upgrade, and configuration backup and restore all reduce site visits. Where a device has a front-panel display, make sure the rack layout leaves it readable. And confirm the encoder exposes what you need for integration — protocol documentation and API availability matter more once the system is in service than at purchase.

Commissioning checks

Item	Pass criterion
Unit budget	Installed equipment plus documented spare capacity; rack not full at handover
Depth clearance	Rear door closes without compressing connectors or cable
Power draw	Measured total recorded, with margin against supply rating
Redundant feeds	Where fitted, A and B traced to separate circuits and labelled
Intake temperature	Measured at rack front, low, not read from the room thermostat
Blanking panels	Fitted to every unused unit
Labelling	Both ends of every cable identified, schedule matches the rack
Spare	A configured spare device held, with its configuration backed up

Where to start

FAQ

Related reading

• Twisted Pair Cable Basics for AV Installers
• How Much Bandwidth Does an IP Video Link Need?
• Designing a Rail Transit PIS Video Network
• Video Encoders