LCD Backlights

Either the lamps are failing and causing the ballast to overheat, or the ballast itself is failing. Suggest swapping the ballast to the other lamp and see if you maintain the same heat rise, if so then you know the lamps probably are ok. Overheating ballast could be caused by failing electrolytic capacitors or other devices.

I thought electrolytics, but there is only one on this board and its common to both PSU. I'll try swapping the lamps around and re-check.

Would a failing lamp be any dimmer ? The light across the screen seems equal. No dark areas or bands. Turning down the brightness doesn't alter the length of time they stay on. Another thing I noticed was that the output capacitors, there is two in series for each lamp, seem to be very small values. They are marked 18J 3Kv. I assume that is

On Wed, 19 Mar 2008 20:57:29 +0000, Baron put finger to keyboard and composed:

An 18pF capacitor would have an impedance at 100kHz of ...

= 1/(2 x 3 x 10^5 x 18 x 10^-12)

= 1/(108 x 10^-7)

= 100Kohm

If 100Kohm were in series with the lamp, then the lamp would be drawing a miniscule amount of current.

- Franc Zabkar

At 1000 volts and 100kohms there is 10ma current which is ten watts. This is about what these lamps run at.

Thanks Franc, I followed your maths perfectly.

I found a lot of references including the circuit diagram for my particular monitor backlight inverter.

It seems that 18pf is indeed correct and that there are in fact four back light tubes each with 18pf in series with it. The tubes need about 1100 volts + to strike and then run at around 700 volts.

In response to another post, because there are two inverters I swapped ends and the overheating stays at the same end. There is a 0.22uf 100 volt capacitor across each of the transformer primary which in the case of one of them gets hot and changes value resulting in the current limiter shutting them both down.

The bad capacitor rapidly looses value when hot air is blown on it. I hope that replacing this will effect a cure. I intend replacing both caps to make sure that I don't have to come back and replace the other later.

Thanks for the support it's appreciated.

On Thu, 20 Mar 2008 15:52:52 +1100, Franc Zabkar put finger to keyboard and composed:

After some research, it appears that 18pF makes sense.

The following datasheet uses an example with a 50kHz resonant frequency, 5mA operating current, and a ballast capacitance of 22pF:

The lamp specs (for a notebook) are:

Lamp Length 250mm (10?) Lamp Diameter 6mm Striking Voltage (20°C) 1000V (PEAK) Operating Voltage (5mA) 375V (RMS) Full Rated Current 5mA Full Rated Power 1.9W

Sorry for my error.

- Franc Zabkar