Rack Power Becomes A Managed Network Node

On 6 October TASCAM introduced two rack power products aimed squarely at installed AV. The AV-P2803 is a 15 A distribution unit, UL derated to 12 A, with six switched outlets on the rear panel split across three timing stages, two unswitched outlets on front and rear, surge protection, EMI and RFI filtering and a front panel voltage display; dedicated control terminals let up to ten units be daisy chained so a whole multi-rack system sequences from one command. The AV-P5015N adds network management: eight NEMA 5-15R outlets under a web interface, per-outlet on and off delays, a scheduler for automatic power up and shutdown, event logs and Syslog, and an Auto Ping function that watches for network activity and reboots attached equipment when it stops responding. An optional EA1240 sensor adds rack temperature and humidity to the same dashboard. Both are 1U.

The sequencing half of that is not new in principle and is still routinely skipped in practice. A rack containing a 16 by 16 matrix, a PoE switch, two amplifiers and eight displays does not draw its steady state current at the instant the breaker closes. Switch mode supplies charge their input capacitors in the first few hundred milliseconds and the aggregate inrush can be several times the running load, which is how a circuit that is comfortable all day trips on restoration after an outage. Order matters as much as staggering: amplifiers should come up after the sources that feed them, or the room greets the morning with a thump through every ceiling speaker. A PoE switch is a special case, because its load arrives when the endpoints boot and request power, so sequencing the switch ahead of the decoders changes the shape of the current curve rather than just delaying it.

The Auto Ping feature is the more consequential change, because it converts a service visit into a policy. Remote reboot is valuable in an unmanned rack, and it is also easy to get wrong. If the device being power cycled is the only path back to the network, the administrator loses the very connection needed to undo the decision; the PDU itself therefore belongs on an unswitched outlet or a separate circuit with its own out of band access. A watchdog that reboots after two minutes of silence will spend a week hiding a failing power supply if nobody reads the log, so the useful configuration combines a long ping interval, a reboot count limit and alerting on the second event rather than the tenth. Syslog export matters here: logs kept only on the device disappear with the device.

Metering is the part that pays for itself during commissioning. Per-outlet current tells you what a display actually draws instead of what its label claims, which sizes a UPS on measured load rather than nameplate arithmetic, and a branch circuit running at 80 percent of derated capacity is a fact worth discovering before the rack is dressed. Temperature and humidity in the same dashboard catch a failed fan weeks before thermal shutdown does. None of this replaces the standby power limits that now apply to networked AV hardware in Europe, but it is the data those declarations are supposed to be based on. Amtune AV switches and splitters, AV over IP decoders and hybrid matrices are specified with documented inrush and running consumption so a rack can be sequenced and metered from the drawing rather than from guesswork. Rack power worksheets on request from sales@amtune.com.