Sending card versus receiving card: asynchronous or synchronous system?
Control of a synchronous LED wall is split between two pieces of hardware: the sending card takes the image at the signal-source end, and the receiving card distributes it among the modules behind the panels - while an asynchronous card replaces both, because it plays back from its own memory. The control side is the cheapest part of the wall to get wrong.
Sending card and receiving card: what is the difference?
Control of a synchronous LED wall is split between two pieces of hardware. The sending card sits at the signal-source end: it takes the image on an HDMI or DVI input from a computer, media player or camera, and passes it on towards the wall over network (Cat5e/Cat6) cable. The receiving card is on the back of the panels; it selects from the incoming data stream the part of the image that belongs to its modules, then distributes it among them through its HUB outputs. A wall typically contains one sending card but as many receiving cards as the number of modules demands: the 8, 12 and 16 HUB75 output receiving cards in our range differ in exactly this. On a larger wall, fewer receiving cards of higher capacity mean clearer cabling and faster fault-finding, because there are fewer connection points at which the signal can be interrupted.
Synchronous or asynchronous: the first and most important decision
A synchronous system mirrors in real time what the signal source shows, but for that it needs a continuously running computer or player. For a studio, a conference room, an event or a live broadcast this is the only workable route, because only this way does an image from a camera or a laptop appear without delay. An asynchronous card, by contrast, stores the content in its own memory: you upload once, after which the card plays back independently, with no computer and no permanent network connection, and resumes by itself after a power cut. For a shop window sign, a price list, a scrolling sign, a pharmacy cross sign or a facade advertising surface, asynchronous is the right choice, because you do not have to keep a PC running behind it twenty-four hours a day. The two are not mutually exclusive: on larger installations we often combine them, with the asynchronous card carrying the default, scheduled content and a synchronous branch with a video processor handling live events, so you can switch over for the duration of an event.
How to work out what size controller you need
Controller capacity is measured in pixels, not square metres, and this is the most misunderstood point in the whole subject. The calculation is simple: the width of the wall in millimetres divided by the pixel pitch gives the horizontal pixel count, and the same for the height gives the vertical. A 3,000 x 2,000 mm wall at P4 pitch is 750 x 500 = 375,000 pixels; the same surface at P10 is 300 x 200 = 60,000 pixels. So over the same area, the finer pitch puts 6.25 times the load on the controller. That is why a wall of the same size can be handled with a single card on one project and needs several receiving cards and a larger sending card on another. The exact figure always comes from the modules, because the wall is built from multiples of the module size: our outdoor P10 module is 320 x 160 mm, which is 32 x 16 = 512 pixels, and the outdoor P4 module is 160 x 320 mm, which is 40 x 80 = 3,200 pixels. In design, then, the module layout has to be fixed first, and the wall's true pixel resolution follows from it.
Load capacity: how many pixels can one port carry?
The limit in a synchronous system is the network port on the sending card. A figure that recurs on manufacturers' datasheets is that one gigabit port drives roughly 650,000 pixels at eight-bit colour depth, so a two-port sending card is usable up to broadly 1.3 million pixels. That number moves down if a higher refresh rate or greater colour depth is set, because the bandwidth stays the same while more data has to pass through it. A rule of thumb is that the planned pixel count should not exceed about 80 per cent of the port capacity: that leaves room for later expansion and for a higher refresh rate. On the asynchronous side, the limit is the card's own capacity. Our range starts at the smallest compact card of about 256x48 pixels, the entry series runs to about 256x64, 384x96 and 512x128 pixels, the middle range is covered by cards of about 384x128, 512x160, 512x256 and 640x160 pixels, and the largest version with five HUB75 outputs goes up to about 768x384 pixels. If the wall is larger than that, the answer is not a more powerful card but a synchronous architecture or a split across several cards.
Compared: asynchronous card or synchronous sending-receiving system
The table below summarises the practical differences between the two architectures, in the order in which they matter for the decision. The most common error is buying an asynchronous card for live video use, or the reverse: building a synchronous system for a simple price list, then realising you have to keep a computer running behind it permanently. Of the two mistakes, the second is the more expensive, because it costs more not only to buy but to run.
HUB75, HUB12 and connectors: what goes into what
The standard between the receiving card and the module is the HUB connector, which links the two with flat ribbon cable. On full-colour (RGB) modules, HUB75 is effectively the dominant standard: typically a 16-pin (in some builds 20-pin) ribbon-cable connector accepted by most manufacturers' modules. Single-colour and two-colour signs, typically LED scrolling signs, use a different HUB pinout, usually HUB12, and the two are not interchangeable: a single-colour scrolling sign panel will not work from a HUB75 output. So when buying, always settle the module type and the controller's output standard together. The good news is that HUB75 is manufacturer-independent, so an existing wall can be expanded or repaired with a receiving card from another source, provided the module's scan mode (1/8 or 1/16 scan, for example) and chipset match. That is described not by the connector but by the configuration file.
Cabling and distances: where things usually go wrong
Three cable runs have to be treated separately. From the signal source to the sending card, HDMI or DVI is used; here the length of passive copper cable is limited, and the higher the resolution the shorter the safe run - for longer distances you need active cable, a signal booster or an optical converter. From the sending card to the receiving cards, Cat5e or Cat6 network cable runs, with a maximum under the structured cabling standard of 100 metres of total channel length per run (of which the fixed run behind the wall may be at most 90 metres, the remainder being patch cables): above that you have to move to fibre-optic transmission, not add a joint. The HUB75 ribbon cable from the receiving card to the modules is the most sensitive; its factory length is typically 20-50 cm, and a cable much longer than needed, or extended, can cause flicker, a displaced row of modules or intermittent image faults. In fault-finding, therefore, always start with the cables: a bad ribbon cable, a loose connector and a kinked network cable are among the quickest to check and the most common causes.
Refresh rate and colour depth: what the controller decides
Many people attribute image quality to the module, when in fact the refresh rate and the grayscale depth are determined by the controller and its configuration. To the naked eye, no flicker is visible above 1,920 Hz, but camera use needs at least 3,840 Hz, or dark bands will run across the recorded image; for live broadcast, studio backdrops and television recording, 7,680 Hz is the safe level, because those work with short shutter speeds. Our own module range follows this: the coarser-pitch outdoor P8 and P10 modules run at 1,920 Hz, while the finer outdoor P2-P5 modules and the indoor modules run at 3,840 Hz. If the wall is going somewhere that is regularly photographed and filmed - an event space, a showroom or a sports hall - the required refresh rate has to be set down in writing in the quotation. It is worth asking about grayscale depth too: in practice 14-16 bit is recommended, because the higher it is, the finer the gradation in dark parts of the image. In night-time or studio operation at low brightness, this is precisely the parameter that decides whether the wall still shows detail or only blocks of colour.
Software: two layers, two users. Calibration software for the installer, content management for the daily user
Two kinds of software go with the controller, and it is worth separating them at the quotation stage. The low-level configuration and calibration software belongs to the installing specialist: it sets the wall's resolution, the modules' scan mode, and brightness and colour uniformity, and it saves the receiving cards' configuration file. The content management software or phone app belongs to the daily user: this is where you change text, upload images or video, and schedule what appears. Neither replaces the other, and typically they are not even the same program - we work through the specific software types and cloud content management in a separate article. What really matters from the client's point of view: ask for the configuration file at installation, and keep it. If the receiving card has to be replaced years later, that file restores the wall within minutes; without it, the settings have to be worked out from scratch.
Connectivity: USB, Wi-Fi, LAN or mobile network
Asynchronous cards also differ in how they receive content, and this decision depends on the site. USB or SD-card upload is the cheapest but requires physical access, and on a facade sign mounted high up that comes back to bite you at every update. Wi-Fi is the most convenient, but its range is limited, typically around 10-15 metres, and it weakens further behind a metal-framed support structure or through a thick wall, so the distance to the router has to be settled at the site survey. Wired LAN is the most stable, and we recommend it for industrial halls, shopping centres and anywhere the Wi-Fi is congested. A mobile network card with a SIM is justified where there is no fixed internet: a roadside surface, a temporary event site, a rural depot. Importantly, with an asynchronous card the network is only needed for uploads: if the Wi-Fi or mobile data drops, the wall keeps playing from memory rather than going dark. We have written up a detailed comparison of the four control routes - including cloud, multi-site management - in a separate article.
Controlling a scrolling sign: what to watch on single- and two-colour signs
The card in a LED scrolling sign belongs to the same family as a LED wall controller, only its job is narrower: it displays text, numbers and simple icons, typically in one or two colours, or in a seven-colour mode that cycles through shades. At entry level, a capacity of around 256x64 pixels is enough for small window signs; for medium-sized lettering, around 384x96; and for longer, multiple alternating messages, around 512x128 pixels is the realistic band. Our smallest compact card, which fits even a narrow sign housing, goes up to about 256x48 pixels. In choosing, the bottleneck is usually not capacity but memory: if you want to store a lot of long texts and several scheduled messages, the smaller card fills up. A colour build needs a full-colour card and HUB75. If you are refurbishing an existing sign, check the module's scan mode and HUB type before replacement, because that decides which card can be connected at all.
When do you need a separate video processor?
A video processor comes in when a live signal has to be displayed, because the resolution of the source does not match the resolution of the wall, or several sources have to be handled. The resolution of a LED wall is almost never a standard aspect ratio - the 750 x 500 pixel wall in the example above corresponds to no usual video format - so the processor scales the incoming HDMI or DVI signal to the wall's actual pixel dimensions. An entry-level processor works with HDMI and DVI inputs, is made for fewer sources, and is enough for small to medium walls. A multi-input version capable of cutting and picture-in-picture (PIP) is justified when a laptop presentation and a live camera have to be shown at once, with clean, uninterrupted switching between them: in a conference centre, a studio or event production that is a basic requirement. Its advantage is that it is a standalone device that goes into the control cabinet rather than behind the panels, so if it fails it can be replaced without opening up the wall.
Sensors: automatic brightness and temperature monitoring
The control system can be supplemented with sensors, and on an outdoor wall that is not a luxury. A light sensor continuously measures ambient light and adjusts the wall's brightness in real time: it turns it up during the day so the wall stays readable in blazing sun, and down at night. In our experience this can cut night-time consumption by 10-30 per cent, and it also spares the LEDs, because brightness degradation is fastest on walls run permanently at maximum. A temperature sensor monitors the internal or immediate ambient temperature of the wall, and on a larger outdoor wall this gives the first warning if the cooling cannot cope - in blazing sun the temperature behind the module can significantly exceed that of the outside air. There is also a separate temperature and humidity sensor, but that serves a different purpose: it is for displaying the measured data on scrolling signs and message signs, not for monitoring the state of the wall. Sensors can be connected retrospectively to most synchronous and asynchronous systems, and for a specialist the wiring is typically under half an hour's work, so if they fall out of the budget now they can be added later - just leave room for them in the cabling.
Spare cards and operational reliability
The greatest operational risk on the control side is not the likelihood of failure but the time it takes to get a spare part. Replacing a receiving card is 15-20 minutes' work in itself, provided the card is to hand and the configuration file exists; if it has to be imported, the same job can stop the wall for weeks. That is why on larger installations we recommend keeping at least one spare receiving card and one spare power supply on site. Held stock of parts and service backing follow the same logic: the question is never whether something will ever fail, but how long it takes to get back up. In event production and on commercially critical surfaces it is also worth discussing the redundant (loop) cabling common in synchronous systems, in which the signal can reach the receiving cards by two routes, so a single cable break does not black out the surface - but always ask whether the specific card type supports it.
What determines the price of a control system?
It would be irresponsible to write a specific figure, because it varies from project to project, but the pricing logic is transparent and checkable. Pushing the price up: greater pixel capacity, more network ports, a higher refresh rate and colour depth, a multi-input video processor, redundant build-out and fibre-optic transmission. An asynchronous, single-card system is orders of magnitude simpler and cheaper than a synchronous sending-receiving architecture, because in the latter a receiving card goes behind every module group. Pushing the cost down: keeping the wall's pixel count within the capacity of a single card, and not having to handle a live video signal. The quotation should always list the sending card, the receiving cards, the video processor, the sensors and the cabling as separate lines: that shows where you can meaningfully save and where you must not.
Checklist for requesting a quotation
Eight questions worth putting to the supplier in writing before you sign. One: what is the exact pixel resolution of the wall, and what percentage of the controller's capacity does it load? Two: is the system synchronous or asynchronous, and does it need a continuously running computer behind it? Three: what is the guaranteed refresh rate in hertz, and is it camera-friendly? Four: what HUB standard and scan mode do the modules use, and how many HUB75 outputs does the receiving card have? Five: which software will I manage daily content with, and will I get the configuration file? Six: how long is the longest network run, and where does it switch to fibre? Seven: are a spare receiving card and power supply held in stock? Eight: is on-site training included in the price? With us, the free site survey, on-site training in the control software, and telephone and remote-desktop support are part of handover, and we give a one-year warranty on the cards.
Expert tip
Ask the installer for the receiving cards' configuration file, and save it in two places: to the cloud and to a memory stick, alongside the wall's documentation. That file of a few kilobytes describes the modules' scan mode, the chipset, and the brightness and colour calibration. If a receiving card has to be replaced years later, with that file the swap takes 15-20 minutes; without it, the settings have to be rediscovered by trial and error, which can mean a whole day's site visit. In the same way, note down the wall's exact pixel resolution, the type of sending card and the name of the software: those three pieces of information are enough for anyone to give meaningful help over the phone.
Common mistakes
1. Thinking in square metres instead of pixels: the same 6 mÂČ surface is 60,000 pixels at P10 and 375,000 at P4, and the latter loads the controller 6.25 times as heavily. 2. Choosing capacity to match the current wall size exactly: with zero headroom, even the smallest future expansion means a complete replacement. 3. Buying an asynchronous card for live video use: camera images and real-time presentation need a synchronous branch with a video processor. 4. Making the HUB75 ribbon cable needlessly long or extending it: cable pulled to length instead of the factory 20-50 cm can cause flicker and a displaced row of modules. 5. Extending the network run beyond 100 metres: the standard channel limit for Cat5e/Cat6 is 100 metres, above which you need optical transmission, not a joint. 6. Accepting a refresh rate below 3,840 Hz on a filmed site, or below 7,680 Hz for live broadcast: dark bands run across the recording, and this cannot be freely corrected afterwards. 7. Not asking for the configuration file and the software name at handover: on a card swap, that is the difference between 20 minutes and a day of fault-finding. 8. Running a commercially critical wall with no spare receiving card and power supply: the risk is not the likelihood of a fault but the wait for the part.
| Criterion | Asynchronous card | Synchronous system (sending + receiving card) |
|---|---|---|
| How it works | Plays the uploaded content from the card's memory | Mirrors the source image in real time |
| Does it need a computer running behind it? | No; once uploaded it runs on its own | Yes; a continuously running PC or player is required |
| Live video, camera feed | No; that needs a synchronous branch with a video processor alongside | Yes, that is its basic function |
| If the network drops | Keeps playing the stored content | The image freezes or goes dark |
| Typical pixel capacity | approx. 256x48 - 768x384 pixels per card | approx. 650,000 pixels per gigabit port (8 bit) |
| Component requirement | One card, with HUB outputs | One sending card, plus a receiving card behind every module group |
| Number of HUB75 outputs | 1 on compact builds, 4-5 on modular versions | 8, 12 or 16 per receiving card |
| Connectivity | USB, SD, Wi-Fi, LAN or SIM-based mobile network | HDMI/DVI input, Cat5e/Cat6 or fibre towards the wall |
| Where it belongs | Shop window sign, price list, scrolling sign, facade advertising, pharmacy cross sign | Studio, event, conference, sports hall, live broadcast |
| Day-to-day operation | Phone app, a 2-3 minute text change | Playback and content management software on a computer |
| Main risk | Running out of the card's memory or pixel capacity | A failure of the source computer stops the whole wall |
A few questions and you get the recommended pixel pitch, the size and an indicative price.
Start the configurator â
