LCD got dark, any idea why?

I've had many HD44780-based/clone displays and I've not seen this sort of thing happen just to the middle. I would suspect heat or maybe pressure on the glass sandwich, but perhaps maybe it's a problem with that particular batch or model of display ... or maybe the way they are mounted causes some "warp" or pressure to accumulate in the centre.

Usually if it were a problem with the analog contrast voltage, then it would affect the display uniformly, rather than in the centre.

- Gough

Could it be related to an improper firmware/driver, that messes with the display drive waveforms? If they aren't properly balanced, LCDs can become charged. Wondering because all 4 have a similar shadow pattern.

https://www.sparkfun.com/datasheets/LCD/HD44780.pdf page 37.

Thank you for the replies,. I also haven't seen this before. I suspected heat, but I guess it wouldn't be that similar on all 4 of them. I never noticed the unit hot, but it definitely can generate some heat, since the thermal design includes a fan which is blowing air from the side.

I will check the mechanical pressure, it is not obvious without opening. The displays are under some front panel that is fully covered by a foil (transparent over the displays) and foil keys under the displays.

As mentioned it is likely related to stress and strain on the glass. You can gently push on different areas or gently lift different areas with some tape to see the effect of pressure.

Display contrast and angle control using bias voltage adjustment can optimize display characteristics, possibly enough to improve this situation. Some displays have a pot to adjust the bias.

dougw said:

related to stress and strain on the glass

If it was one, or some displays, yes. But here we see the exact same flaw on all four displays (lower row central characters impacted). I don't know how that can happen on all of them, gradually, in 4 years.

All causes should be low probability, Here is a more in depth list of some possibilities...

• The LCD sandwiches have this bias when built due to manufacturing tolerances, but it can be overcome by the correct electrical bias, however over time the electrical bias has drifted, exposing the manufacturing tolerance issue.
• The LCD sandwiches are clamped in frames that apply pressure around the edges. Over time the center sections bow out a bit causing the middle areas to need a different drive bias.
• The LCDs are mounted in the panel using screws that apply an uneven stress on the LCD and over time the LCD deforms slightly.
• If the LCDs use an elastomeric connector on a PCB, the PCB may bow over time causing a different impedance to some signals (this is common with elastomeric connectors)
• There is some hot component under that area in each display that is causing thermal expansion and something has fatigued enough for it to cause an issue.

Like I said, so far only low probability ideas...

dougw Good list. 

Fluke displays exhibited loss of characters due to the elastomeric connector oxidizing.  A little alcohol and allowing them to flex a bit usually brought them back in line.

My guess is it's related to heat.  The output levels listed on the front of the charger are pretty beefy.  AND... the pushbuttons are on the bottom.  Maybe there's some correlation there.

One of the older data sheets contains operation cautions: https://datasheet.datasheetarchive.com/originals/scans/Scans-049/DSADA001252.pdf 

It’s likely due to prolonged use, which is quite normal for performance degradation.

Not when properly driven though. Degradation because of improper balanced signals usually shows after intensive use. Properly driven displays stay clear for a lifetime.