LED Control System: The Complete Technical Guide to Sending Cards, Receiving Cards, and System Architecture (2026)

Why the Control System Is the Brain of Every LED Display — and Why It Deserves More Attention Than the LEDs Themselves

An LED display without a control system is a collection of diodes that will never emit a single photon of coordinated light. The LED control system — comprising sending cards (or independent controllers), receiving cards, and the software that orchestrates them — is what transforms a grid of individual LEDs into a synchronized display capable of rendering 4K video at 3,840 Hz refresh rates with per-pixel color accuracy. Yet in B2B procurement, the control system is routinely treated as an afterthought — a commodity checkbox item rather than the performance-critical subsystem it actually is.

This guide explains the architecture, components, and selection criteria for LED display control systems. Drawing on the LED Display Application technical reference series — particularly the intermediate and advanced volumes which devote extensive coverage to control system design, configuration, and troubleshooting — it provides B2B buyers, AV integrators, and system designers with the technical foundation needed to specify, evaluate, and verify LED control system performance. For background on the LED modules these systems drive, see our indoor LED pixel pitch selection guide.

System Architecture: The Three-Layer Control Hierarchy

Every LED control system follows a three-layer architecture that maps cleanly to the physical structure of the display itself:

Layer	Component	Function	Key Specification
Layer 1: Source & Processing	Sending Card / Independent Controller / Video Controller	Receives video input (HDMI, DVI, SDI, DisplayPort), performs color space conversion and Gamma correction, packages pixel data into network protocol packets	Maximum pixel loading capacity (e.g., 650,000 pixels per Gigabit Ethernet port; 2.3M pixels for 4-port controller)
Layer 2: Distribution	Gigabit Ethernet Network (Cat5e/Cat6)	Transports packetized video data from sending card to multiple receiving cards; each port drives up to 650,000 pixels	Maximum cable length: 100m (copper); 10km+ (fiber with media converter)
Layer 3: Display Execution	Receiving Card + Driver ICs	Decodes network packets via FPGA, generates PWM timing signals for each row/column, drives individual LED channels through constant-current driver ICs	Maximum pixel loading per receiving card (varies by model; typically 256×256 to 512×512 pixels)

This architecture is fundamentally a serial distribution model: a single high-bandwidth video source is decomposed into parallel data streams at the sending card, distributed over standard Ethernet infrastructure, and reconstructed into synchronized PWM signals at each receiving card. The elegance of this approach is that Gigabit Ethernet — a mature, low-cost, well-understood physical layer — provides the backbone, while the proprietary intelligence resides in the FPGA firmware on the sending and receiving cards. For a detailed look at how controller selection affects system design, see NovaStar VX vs MCTRL controller comparison.

Understanding Sending Card Capacity: The Math Behind Pixel Loading

The sending card’s pixel capacity is the primary constraint in LED control system design. Each Gigabit Ethernet output port on a sending card can drive a maximum of 650,000 pixels at standard frame rates. This limit is not arbitrary — it is determined by the net bandwidth available after network protocol overhead:

Available bandwidth per port ≈ 1 Gbps / (bit depth × pixels per frame × frames per second)

For a typical configuration at 8-bit color depth, 60 fps: each pixel requires 24 bits (3 colors × 8 bits) per frame. At 60 frames per second: 24 × 60 = 1,440 bits per second per pixel. A 1 Gbps port (approximately 950 Mbps after overhead) can therefore support approximately 950,000,000 / 1,440 ≈ 660,000 pixels — consistent with the industry-standard 650,000-pixel rating, which includes a safety margin for control overhead and blanking intervals.

The practical implications for system design are straightforward but must be calculated precisely:

• A 2-port sending card (1.3 million pixels total) can drive a 1920×1080 (2.07 MP) display — but only just. The 2.07M pixels exceed 1.3M, requiring either a 4-port card or a lower-resolution input.
• A 4-port sending card (2.6 million pixels) comfortably supports 1920×1200 (2.3 MP).
• A 16-port sending card or 4K controller (10.4 million pixels) supports 4K UHD (3840×2160 = 8.3 MP) with capacity to spare for redundancy.
• A P0.9 display measuring 5m × 2.8m produces 5,556 × 3,111 = 17.3 million pixels — exceeding even a 16-port card. This highlights why ultra-fine-pitch displays require multiple synchronized controllers or the latest generation of high-capacity 8K controllers.

For guidance on calculating display resolution from physical dimensions and pixel pitch, see our indoor LED display resolution and size calculator.

Receiving Card Configuration: Where the Physical Display Meets the Digital Signal

The receiving card is the interface between the standardized Ethernet video stream and the physically unique LED module it drives. Each receiving card must be configured with a configuration file (.rcfg) that tells the card’s FPGA exactly how the connected LED module is organized: the driver IC type and channel count, the scan mode, the pixel arrangement (row/column mapping), and the HUB connector pinout. A misconfigured receiving card produces anything from swapped colors to completely scrambled images — and this configuration is what makes LED control system integration a skilled engineering task rather than a plug-and-play exercise.

The receiving card configuration workflow, as documented in the LED Display Application intermediate volume, follows these steps:

1. Identify the driver IC on the LED module (e.g., ICN2053, MBI5153, MBI5124). The driver IC type determines the data protocol, channel count, and grayscale bit depth the receiving card must output.
2. Determine the scan mode of the module (e.g., 1/16 scan, 1/32 scan). The scan mode tells the receiving card how many rows are illuminated simultaneously versus sequentially. For a detailed explanation of scan mode identification, see our LED module scan mode guide.
3. Map the HUB connector pinout to the receiving card’s output interface. Standard HUB75 connectors have defined pin assignments for RGB data lines, clock, latch, and output enable — but the mapping of specific data lines to specific LED positions on the module is unique to each module design.
4. Load the configuration file onto the receiving card via the control software (e.g., NovaStar NovaLCT, Linsn LEDStudio). The configuration is stored in non-volatile memory on the receiving card and persists across power cycles.
5. Verify the display output with a test pattern (color bars, gradient, grid) to confirm correct pixel mapping, color order, and scan direction before deploying content.

The market for receiving cards is concentrated among a few dominant brands: NovaStar (MRV series), Colorlight (5A-75 series), and Linsn (RV series). These three brands collectively represent the vast majority of receiving cards in commercial LED installations worldwide. For B2B buyers, specifying a receiving card from one of these established brands ensures long-term availability of configuration software, firmware updates, and replacement units — considerations that matter over a display’s 10+ year operational lifespan. Many of these controllers are available through specialist distributors; leading LED display manufacturers typically pair their modules with specific receiving card models optimized for their products.

Synchronous vs. Asynchronous Control: Two Fundamentally Different Operating Modes

The LED control system market divides into two architectural approaches that serve fundamentally different use cases:

Synchronous Control Systems

A synchronous system requires a live video source — typically a computer, media server, or video processor — continuously connected to the sending card. The display shows exactly what the source outputs, in real time, with latency measured in milliseconds. This is the architecture used for: live events and broadcast studios (camera feeds must be real-time), corporate presentations (presenter’s laptop mirrored to display), command and control centers (operator workstations driving video walls), and any application where the content changes dynamically based on a live source.

The synchronous control chain — from video source through video processor through sending card through receiving card to LED — introduces approximately 30-60ms of total system latency in a well-designed system. For most applications this is imperceptible; for esports and high-frequency financial trading displays, latency must be specified and verified below 16ms (one frame at 60Hz).

Asynchronous Control Systems

An asynchronous system stores content locally on the display — typically on an embedded multimedia player with internal storage — and plays it back without requiring a connected computer. Content is loaded onto the player via USB, Wi-Fi, or cloud platform, and the player outputs directly to the receiving cards. This architecture is used for: digital signage networks (content updates scheduled remotely), information displays (fixed content with periodic updates), and installations where running a permanent video cable to the display location is impractical or cost-prohibitive.

Modern asynchronous systems connect to cloud platforms (e.g., NovaStar ViCloud, Colorlight Cloud) that enable remote content scheduling, display health monitoring, brightness adjustment, and firmware updates across hundreds of geographically distributed displays from a single web interface. This cloud management capability is what distinguishes a modern asynchronous LED control system from the standalone USB-update players of the previous generation. For B2B buyers managing multi-site digital signage networks, cloud-based asynchronous control can reduce ongoing operational costs by 40-60% compared to maintaining on-site media players or sending technicians for manual content updates. For more on the infrastructure implications, see our outdoor LED display installation planning.

Video Processors and All-in-One Controllers: The Convergence Trend

The traditional separation between “video processor” and “sending card” — two distinct hardware boxes in the signal chain — is being steadily eliminated by all-in-one video controllers that integrate both functions. This convergence is driven by practical engineering advantages: fewer devices means fewer points of failure, lower total system latency, simpler cabling, and reduced rack space. The integrated architecture is now the dominant form factor for new installations. For B2B buyers evaluating NovaStar VX1000 and similar all-in-one controllers, the key evaluation criteria are:

• Input versatility: HDMI 2.0/2.1, DisplayPort 1.4, 12G-SDI, and USB-C inputs provide compatibility with diverse video sources
• Output pixel capacity: 2.3M (entry-level), 6.5M (mid-range), 10.4M+ (4K/8K class)
• Layer management: Multi-layer compositing (picture-in-picture, side-by-side) enables flexible content presentation without external switchers
• Genlock/frame sync: Essential for multi-camera broadcast environments to prevent video tearing
• Low-latency mode: Bypasses frame-buffer processing for applications requiring sub-16ms glass-to-glass latency

 

Control System Redundancy: Designing for Mission-Critical Uptime

For applications where display downtime is unacceptable — broadcast studios, command and control centers, financial trading floors, 24/7 network operations centers — the LED control system must incorporate redundancy at every layer. The industry-standard redundancy architecture includes:

• Dual sending cards with automatic failover: if the primary sending card fails, the backup takes over within 0.1 seconds, typically with no visible disruption to the displayed content
• Dual power supplies per cabinet: each power supply operates at 50% load under normal conditions; if one fails, the other instantly takes the full load
• Dual-redundant receiving cards per cabinet: both cards receive and process the same video stream; if one fails, the other continues output without interruption
• Redundant network paths: separate Ethernet cables from separate sending card ports provide physical path diversity within each cabinet

The cost premium for full redundancy — typically 30-50% above a non-redundant equivalent — must be weighed against the cost of a single hour of downtime. For a broadcast studio losing $50,000-$500,000 per hour of off-air time, or a command center where display failure could have safety implications, the redundancy investment is not a technical preference — it is an operational requirement. For more on system-level reliability engineering, see our LED display troubleshooting guide and LED display manufacturing quality standards.

---

Key Takeaways

1. The control system is not a commodity — it determines the display’s performance ceiling. The sending card’s pixel capacity, the receiving card’s grayscale bit depth, and the network topology’s latency characteristics collectively set the maximum achievable resolution, refresh rate, and color accuracy of the entire display system.
2. Calculate pixel capacity before purchasing. A sending card rated for 650,000 pixels per port cannot drive 700,000 pixels. The math is unforgiving and must be verified during system design, not discovered during commissioning.
3. Synchronous for live, asynchronous for scheduled. The choice between synchronous and asynchronous control architecture determines how content reaches the display — and this choice has operational cost implications that compound over the display’s service life.
4. All-in-one controllers reduce complexity and failure points. The convergence of video processing and sending card functions into single devices eliminates cabling, reduces latency, and simplifies troubleshooting. For new installations, integrated controllers are the default recommendation unless specific requirements dictate separate components.
5. Redundancy is an operational insurance policy. The 30-50% cost premium for full system redundancy buys protection against the single-point failures that cause display downtime. For mission-critical installations, redundancy is not optional.

For more technical guidance on LED display system design, component selection, and performance optimization, visit the LED screen manufacturer knowledge base and explore our guides on outdoor LED brightness, LED color calibration, and LED display supply chain trends.