What Is LCD Power Sequencing?
Power sequencing is the order and timing in which an LCD panel receives the inputs it needs to display a correct image: its logic supply, video interface, and backlight control. Every panel has a required sequence for switching on and for switching off, with minimum, and sometimes maximum, intervals between the steps. Get it right, and nobody notices. Get it wrong, and the result is anything from a flash of white at power-on to a panel that will not start on a cold morning. The same display may work perfectly on a warm bench, which is what makes sequencing faults slow to find.
The three inputs a panel receives
Logic supply (VDD): Powers the panel's timing controller (T-CON) and its driver circuits; typically 3.3 V, 5 V, or 12 V depending on the panel. Converters inside the panel generate the row and column driver voltages from this one supply.
Video interface: The video data from the display controller over TTL/RGB, LVDS, V-by-One, eDP or MIPI DSI. With TTL/RGB and LVDS, there is no link-training stage: the controller drives the clock and data directly. V-by-One, eDP and MIPI DSI have interface-establishment or initialization stages before active video is available. Below, "valid video" means that the interface is in the state the panel specification requires and valid image data can be displayed.
Backlight control: The backlight has its own power (VLED) and two logic-level control lines, Backlight Enable (BL_EN) and PWM brightness, driven by the controller or by a backlight driver board under its control.
The order at power-on is logic supply, then interface ready and valid video, then backlight enable, subject to the panel specification. Switching off is the reverse. The backlight supply and the enable line may have separate requirements of their own; some specifications give the backlight its own power-sequence section, and where they do, it applies alongside the panel's.
What a power-sequence specification looks like
The full specification for a panel normally gives the sequence as a timing diagram and a table. Not every spec sheet has it, and the brief datasheet rarely does; if yours is missing it, ask the panel maker for the full specification. Where it exists, it contains three things.
An order, as above.
A set of windows, intervals between the transitions, each with a minimum and sometimes a maximum. Manufacturers usually label them on the diagram as T1, T2, T3 and so on, but the labels are not standardized: two makers, or two panel families from one maker, can assign the same label to different intervals. Always work from the definitions in the specific panel specification. The intervals that matter are:
| Interval | What it spans | What the sheet gives |
|---|---|---|
| Supply rise | VDD from 10% to 90% of its working level | A specified range or a maximum rise time, depending on the panel: a rise that is too slow can defeat the T-CON's power-on reset |
| Supply to interface | VDD stable → interface ready and valid video | A minimum, and sometimes a maximum |
| Video to backlight | Valid video → Backlight Enable on | A minimum, and sometimes a maximum |
| Backlight to video off | Backlight Enable off → video removed | A minimum |
| Video off to supply off | Video removed → VDD off | A minimum |
| Off interval | VDD off → VDD may be re-applied | A minimum, and occasionally a maximum |
The rule is the same for every row: stay inside the window. "Longer is safer" is not a general rule because some of these intervals have both a maximum and a minimum. A case from our own work illustrates this. A DT-1920-HDMI driving a Full-HD LVDS panel showed an abnormal image at start-up. The board was on its default panel-timing entry, which set the interval between the supply coming up and the LVDS signal starting to several tens of milliseconds. The panel's specification requires the supply to be established only 1 to 2 ms before the LVDS signal rises. Bringing the interval inside that window corrected the display, and nothing else was changed. The default was not too short; it was too long.
Two rules. The video interface must remain in the state the panel manufacturer specifies while VDD is off, which in most cases means that active clock and data must not be driven into an unpowered panel. And the panel must stay off for the specified minimum interval before power is applied again.
Why the order matters
Each rule guards against a specific physical effect.
Signal before supply. Input protection structures can conduct when an interface is driven while its supply domain is unpowered. Current then flows into the panel through its signal pins, so the panel is partly powered through the wrong path, and the result can be undefined operation or latch-up. Panel manufacturers warn about this explicitly, which is why the interface state while the panel is unpowered is a rule rather than a recommendation.
Backlight before a valid image. For the first frames after the signal arrives, the T-CON is still locking to it, and the driver outputs are settling. Enable the backlight during that time and the viewer sees noise, a white flash or a frozen frame. The video-to-backlight interval hides it. This is normally a visible start-up artifact rather than a panel-damage mechanism, but it is often the fault a customer notices first.
DC across the liquid crystal. Liquid crystal is driven with an alternating voltage around a common reference, VCOM; a net DC component causes ions in the material to migrate and the image to stick. Applying signal before VCOM is established, or removing VDD while signal is still present, can leave DC across the cells for a moment each time. Repeated violations can contribute to DC stress and image sticking, and may affect the panel's reliability.
Restarting too soon. The panel's internal driver voltages are held on capacitors, and its reset circuit needs VDD to fall below a threshold before it will fire again. Re-apply power before that has happened and the panel can start from a partially charged state, with the drivers in the wrong condition and a display that shows corruption until the unit is properly powered down. The specified minimum off interval is what protects against this.
What goes wrong when the sequence is violated
The symptoms are recognizable once you know them.
- No image at power-on, on some units or on some mornings, with everything else measuring fine. In the cold, supplies rise more slowly and the panel's internal converters and oscillator are slower to start, so a supply rise or a supply-to-interface interval that was inside its window at room temperature may no longer be.
- Starts on the second or third attempt. Behavior consistent with a marginal sequence; temperature-sensitive start-up margins are a common cause.
- A flash of white or noise at every power-on. The backlight is enabled before a valid image.
- Corrupted display after a brief power interruption, such as a brown-out, a watchdog reset, or an operator switching off and straight back on, that clears only after a full power-down: consistent with the minimum off interval not being respected.
- Image sticking developing over time on a unit that is power-cycled often: potentially associated with incorrect power-off sequencing.
The first two may hide because a warm laboratory can be considerably more forgiving than the environment in which the finished product will operate.
Power sequencing is not panel timing
People often confuse the two because a controller stores them together. Panel timing describes the video signal itself: resolution, refresh rate, horizontal and vertical totals, porches, pixel clock. Power sequencing describes the supply and control-line transitions around that signal. A controller can drive a perfectly timed signal in the wrong sequence, and a correct sequence with the wrong timing; the symptoms differ, and both must be right.
One consequence is worth drawing out. When a controller is set to a default or nearest-match panel entry, it inherits that entry's sequence values and timing. The picture can be stable and the sequence still outside the panel's specification, because most of the faults above are intermittent or slow to appear. A stable image is not evidence that the sequence is right. Only a comparison with the panel's own table is.
Where the controller comes in
The controller board provides all three inputs, generating VDD, driving the video interface, and controlling the backlight lines, so the controller's configuration is where the panel's sequence is implemented. That is why, when Digital View matches a controller to a panel, the power sequence for that panel is part of the match, alongside the panel timing, the supply voltage and the cable; it is held in the board's configuration for that panel. Digital View controllers represent the principal power-on and power-off delays as six parameters, T1 through T6, Digital View controller terminology configured to satisfy the sequencing requirements of the particular panel, and on the newer platforms they can be set over the serial interface, stored, and read back, so that a working configuration can be recorded and reproduced exactly. Whichever board you use, the question is the same: what sequence does it apply to this panel, and does that sit inside the panel's windows?
Questions and answers
Why does my display flash white when it switches on? The backlight enables before the T-CON locks to the video signal. The video-to-backlight interval needs correcting within the window the panel specifies.
Why does the panel start only on the second attempt? An interval is outside its window, often the supply rise or the supply-to-interface delay, and the margin changes with temperature. A longer delay does not fix a supply that rises too slowly, so the correction is to the parameter that is out, not to a delay after it. For a matched kit, ask us.
Does the order matter on power-off too? Yes. Removing VDD while the video signal is still present can unbalance VCOM and leave DC across the liquid crystal; repeated, that contributes to image sticking.
How long must the panel stay off before repowering? As long as the specification's minimum off interval says; it is often the longest interval in the table, and a few sheets give a maximum too. A system that can restart quickly, after a reset or a brown-out, has to enforce that interval itself.
Can the wrong sequence physically damage the panel? Driving an unpowered panel can cause back-powering, undefined operation or latch-up; in some circumstances physical damage is possible. More commonly, incorrect sequencing causes intermittent start-up faults or contributes to image sticking over time, and both arrive as field returns.
Do I have to work all this out myself? Not for a matched kit: the panel sequence is part of the configuration you receive. For a panel we have not matched, send us its full specification.
In short
Power sequencing is logic supply, then interface ready and valid video, then backlight enable, and the reverse on the way off, inside the windows the panel specifies, with a minimum off interval before a restart. The windows exist for physical reasons: input protection that conducts when driven unpowered, a timing controller that needs a moment to lock, DC balance across the liquid crystal, and internal supplies that need time to discharge. The labels on the panel's diagram are the panel maker's; get the table and make sure the sequence the controller applies fits within it.
If you already have a panel selected, enter its model number in the Digital View Controller Solution Generator to find compatible controller solutions. For a specific controller-and-panel pairing, or a panel we haven't previously matched, contact our support team.