New Tech Tuesdays: Scaling 48V Data Center Power for Open Rack V3
Key Highlights
- Increasing rack power density requires careful design of conversion efficiency, cooling, and distribution systems to maintain performance and reliability.
- Modular power shelves and hot-pluggable modules enable incremental expansion, reducing downtime and supporting future load growth.
- Effective telemetry and monitoring are essential for early fault detection, troubleshooting, and predictive maintenance in high-density environments.
New Tech Tuesdays
Join Mouser's Technical Content team for a weekly look at all things interesting, new, and noteworthy for design engineers.
Expanding a server rack once involved little more than locating an open slot and connecting the necessary cables. However, at today’s rack-level power densities, expansion also affects conversion efficiency, cooling, distribution, monitoring, and maintenance.
Designers planning an Open Rack V3 deployment must design around a nominal 48V rack-level power system. In such designs, the applicable specifications and selected equipment define the exact operating range.
This week’s New Tech Tuesdays explores scalable 48V power architectures for Open Rack V3 and the design considerations required to balance efficiency, redundancy, serviceability, and monitoring in high-density data centers.
Why 48V Data Center Power Matters
Densely packed high-performance compute equipment in server racks requires more power per rack than traditional setups. Increasing the power level by increasing voltage helps mitigate resistive conduction losses that scale with the square of current, expressed as Ploss = I2R. For the same delivered DC power, and neglecting conversion losses, increasing the distribution voltage from 12V to 48V reduces current to one-quarter. Actual system performance still depends on factors such as busbar resistance, connector design, loading, and conversion efficiency.
Designers must still account for voltage-drop limits, thermal rise, insulation, protection, connector ratings, and downstream conversion. Voltage selection is only one part of the design. The power shelf, busbar, protection, downstream converters, cooling, and monitoring interfaces must operate as a coordinated system.
Building Scalable Rack Power
A scalable design must account for future load as well as the initial deployment, including expected growth, transient demand, reserve capacity, redundancy, upstream power availability, and thermal limits.
Modular power shelves enable capacity to be installed incrementally rather than requiring the complete system to be replaced when demand increases. Incremental expansion still depends on sufficient capacity in the surrounding architecture. Busbars, connectors, branch protection, airflow, and downstream conversion stages must all have sufficient headroom.
Conversion efficiency also becomes increasingly important at scale. Nearly all electrical power lost during conversion ultimately becomes heat, and the location and thermal design of the conversion equipment determine where additional cooling may be needed. Efficiency across the expected operating range provides a more useful design measure than the peak rating alone.
For instance, a conversion stage delivering 30kW at 97 percent efficiency draws approximately 30.93kW, leaving about 0.93kW for the cooling system to remove. This example calculation shows why even a few percentage points of conversion loss matter in a high-power rack.
Design for Serviceability and Visibility
Building resilience into a power architecture requires more than redundant hardware; it also demands service-friendly design and the visibility needed to detect issues early. Hot-pluggable power modules can be inserted or removed while the shelf remains energized. With sufficient redundant capacity and a supported replacement procedure, technicians can replace a failed module without shutting down the powered equipment. In an N+1 configuration, the remaining modules must carry the full load after input voltage, temperature, cooling, and derating limits are applied. Validation must cover normal operation, maintenance, and fault conditions.
Further adding visibility, telemetry provides power-system data for fault detection, troubleshooting, and maintenance. When combined with defined response procedures, this visibility helps operators detect and address emerging problems. Available telemetry may include electrical measurements, module status, temperature, and fault indications. Before selecting the telemetry interface, designers need to identify the required measurements and alarms and determine how that data will integrate with the existing management environment.
Telemetry can also support troubleshooting, trend analysis, and maintenance planning. Predictive maintenance, however, requires additional historical data, analytical models, and operational validation.
The Newest Products for Your Newest Designs®
The Bel Power Solutions Data Solutions portfolio addresses multiple stages of data center power delivery. The portfolio includes AC-DC and DC-DC front-end power supplies ready to fit into hot-pluggable power shelves. These high-performance power supplies provide up to 28kW of power in a single power shelf within a single-busbar Open Rack V3 configuration, enabling designs to use more next-generation compute hardware per rack.
Each selected shelf, rack, controller, and distribution product must meet the applicable mechanical, electrical, communications, and power-interface requirements. General Open Compute Project (OCP) compatibility does not necessarily establish Open Rack V3 conformity or formal recognition as an OCP product.
Product selection, therefore, requires a review of documented Open Rack V3 compatibility, including:
- output-voltage range,
- power rating,
- efficiency,
- derating curves,
- cooling requirements,
- redundancy behavior,
- current sharing,
- busbar interface,
- mechanical fit, addressing and management interfaces,
- protection requirements, and
- certifications.
Compatibility or compliance demonstrated within one product family should not be extended to the complete portfolio without product-specific documentation. Still, these front-end power supplies comply with UL/CSA/IEC/EN 62368-1 international safety standards and are available in various form factors to accommodate specific environments, operating within the -5°C to 65°C temperature range.
Tuesday’s Takeaway
Supply capacity is only the starting point for a scalable 48V data center power architecture. Designers must balance efficiency, density, thermal performance, serviceability, redundancy, telemetry, and downstream distribution across the entire rack.
Modular shelves, hot-pluggable power modules, redundant configurations, and networked monitoring provide key building blocks for scalable designs. A well-planned architecture supports the initial deployment and reserves verified electrical, thermal, mechanical, protection, monitoring, and facility capacity for planned expansion.
Mouser Electronics, founded in 1964, is a globally authorized distributor of semiconductors and electronic components for over 1,200 industry-leading manufacturer brands. We specialize in the rapid introduction of the newest products and technologies targeting the design engineer and buyer communities. Mouser has 28 offices located around the globe. We conduct business in 23 different languages and 34 currencies. Our global distribution center is equipped with state-of-the-art wireless warehouse management systems that enable us to process orders 24/7, and deliver nearly perfect pick-and-ship operations.
