Lots of modern monitors feature an indication of incoming video signals, even if the monitors are switched off. In most cases, an LED provides this indication by fading (slowly increasing and decreasing) its light intensity. This type of indication is very specific, and thus convenient, for use in other devices for the same purpose.
Monitors typically stay connected to the power line, even when they're switched off. However, this doesn't apply for all video devices. Some may need a similar indication even when they're not connected to power. Therefore, a "powerless" fading indication seems to be a useful design.
The circuits presented here indicate the presence of a video signal on a VGA input. The circuits don't need an additional supply voltage, because they're supplied by the energy of the horizontal synchronization pulses (HSYNC) of the video signal itself. The circuits are designed to work with all Video Electronic Standards Association (VESA) video signals. Other video signals may work as well but weren't tested.
Because the HSYNC output of a graphics card can deliver a limited current only during its High state, we can consider it as a dc source with a duty cycle whose value depends on the VESA mode used. Depending on the type of graphics card and the logic family used, the High-level HSYNC pulses may vary from 3.3 to 5 V, and the source resistance from 30 to 62 .
As a worst case, it was assumed that the graphics card has a 3.3-V HSYNC level, and the video mode is 800 by 600 at 85 Hz. In this mode, the duty cycle of the HSYNC signal is only 6%. The output resistance of the graphics card was assumed to be about 60 .
Circuit 1: The first circuit uses a PIC10F200 (Fig. 1). This microcontroller holds several advantages. It's very small (available in a six-pin SOT-23 package), and it's a low-cost, low-power device. The fact that it uses only a few instructions makes it easy to program. An internal 4-MHz precision oscillator saves an additional component. The circuit works with a supply voltage from 2 to 5.5 V. Also employed is an L-934SEC-H, a Kingbright LED that features high illuminating power with a relatively low forward voltage.
In the absence of VCC, the HSYNC pulses charge the capacitor, which supplies the PIC10F200. In series with the LED is a resistor that limits the current, so that the voltage across the capacitor doesn't fall below the minimum of 2 V (in the worst-case scenario).
Fading of the LED is controlled by the PIC10F200, using pulse-width modulation (PWM). Because this PIC doesn't have an integrated PWM module, this function had to be implemented in software. The technique is based on a program from the Ingenieurbüro Lehmann (www.iL-online.de). English versions of the code are available with the online version of this article at www. electronicdesign.com. When VCC appears, the program changes to VCC mode, and the LED stops fading and is permanently on.
Circuit 2: For fans of analog techniques, an alternative solution uses 555 timers (Fig. 2). To minimize power consumption and reduce the capacitor values, it's necessary to use the CMOS version of the timer. U1 is astable and generates a sawtooth voltage that modulates the pulse width of the U2 astable. To minimize power consumption, the oscillating frequency of U2 should be as low as possible. Because it's not possible to reduce the pulse width down to zero and to get the LED completely dark with this IC, a 300-nF capacitor (C1) is added after the current-limiting resistor (R3). R3 and C1 build a low-pass filter that integrates U2's PWM output signal to a fading dc voltage. That voltage's lowest value falls below the LED's forward voltage during the minimum pulse width.
Of course, the two 555 ICs could be replaced by one 556. But, unfortunately, there was no CMOS 556 timer for less than $1 available on the market.