Maker Pro
Maker Pro

LCD Monitor CCFL fault -- lamps or inverter?

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
Based on my recent successful monitor repair I have been passed an already disassembled 19 inch monitor that "went pink and died". This is the classic CCFL failure problem.

The owner of this monitor decided that replacing the CCFL lamps was going to be too hard and didn't go any further. As far as I can see, the replacement of the lamps (there's 4 of them) would be uneconomical in any case.

The big question though... Is the "going pink and dying" problem more likely to be an inverter fault or a tube fault. Indications are that the monitor went pink all over, and so I think that points to either an inverter problem or a tube pulling down the voltage on the inverter.

Is it low voltage ==> low current which causes the pink hue? My experience with xenon flash lamps is that very high voltages and low currents cause them to go the other way (violet).

*IF* the problem is the inverter, what is an acceptable method of testing it? I know that the current is in the range of 7 to 8 mA per lamp, but I have no idea of the voltage to expect. I also don't have any equipment for measuring particularly high voltages (nor the inclination to get too close to them).

Even considering the following, I'm still keen to test the inverter.

OK, some further research indicates that CCFL, like many discharge lamps, have a high striking voltage, and are then relatively low voltage devices driven from a current source. Too low a current may lead to premature death of the lamp :-(
 

Mitchekj

Jan 24, 2010
288
Joined
Jan 24, 2010
Messages
288
Bah, LED is the way to go. Heheh. None of that pesky mercury either! Any motivation to retrofit the CFLs to LED?
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
Any motivation to retrofit the CFLs to LED?

Lots. Have you been reading my mind? :D

I presume that since there are 4 CCFL's spread across the screen that changing the backlight to a number of point sources (instead of linear sources) won't cause any huge problems.

The next trick if I go this way is to somehow fit a power supply for these devices into the monitor housing. I was thinking about connecting them all in series and using a switchmode current regulator to power them.

I could probably design something to run from rectified mains, but obviously isolation is a major issue and I don't really want to go down this route. :eek:

I might mount a single LED behind the display, powering it externally to see what sort of illumination I get and to determine an approximate number of LEDs required.
 

Mitchekj

Jan 24, 2010
288
Joined
Jan 24, 2010
Messages
288
19" monitor? My best guess would be 30-50 LEDs, shot in the dark. Since, as you stated, the point source may lead to spotting if they're too far apart. Fewer LEDs spaced further apart, they'd need to be high power, then you start running into heat issues. There may even be LEDs with very wide viewing angles. (Most of the SMD ones I'm familiar with have ~120-130 degrees.) I'm sure there's a way to get around high part counts, tho.

To wire them all in series, you'd need something along the lines of 90-150V, which may make the SMPS more feasible, in theory. At least much more efficient. :)
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
Yeah, my rough guestimate was about 10 LEDs per CCFL, but more if I need more power (I think it's better to have more lower power diodes than a few really bright -- and hot -- ones).

SMPS is probably fairly simple, but I think I might have to wind my own transformer which is way outside my experience.

Interfacing the control over brightness to the existing signal may be interesting, but having a schematic of a similar (perhaps earlier) monitor means I have a rough chance.
 

Resqueline

Jul 31, 2009
2,848
Joined
Jul 31, 2009
Messages
2,848
The pink color comes when the neon gas is the primary conductor. The lack of mercury vapor can stem from it being depleted, or a too low tube current to vaporize it sufficiently.
All 4 tubes going pink at the same time would indicate the inverter being weakened. The HV transformer is a very stressed part.

Here's a place that makes & sells both CCFL's, -inverters, retrofit LED strips, & -drivers.
I have bought stuff from them but I don't know their prices on these products or their sales policy on non-business customers.
 
Last edited:

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
The monitor I have is not listed. Unfortunately, due to the price of a new one, it's hardly worth fixing this one up unless I can do it real cheap.

The monitor is not listed on their site either.

I'm doing this mostly to get familiar with what I might find when I take my laptop apart. The CCFL in that is worth replacing because of the value of the laptop.

Oh, and to correct an earlier misunderstanding; the 4 CCFLs are not spread across the monitor, 2 are at the top, and 2 are at the bottom.
 

jerryg50

Apr 18, 2010
57
Joined
Apr 18, 2010
Messages
57
Based on my recent successful monitor repair I have been passed an already disassembled 19 inch monitor that "went pink and died". This is the classic CCFL failure problem.

The owner of this monitor decided that replacing the CCFL lamps was going to be too hard and didn't go any further. As far as I can see, the replacement of the lamps (there's 4 of them) would be uneconomical in any case.

The big question though... Is the "going pink and dying" problem more likely to be an inverter fault or a tube fault. Indications are that the monitor went pink all over, and so I think that points to either an inverter problem or a tube pulling down the voltage on the inverter.

Is it low voltage ==> low current which causes the pink hue? My experience with xenon flash lamps is that very high voltages and low currents cause them to go the other way (violet).

*IF* the problem is the inverter, what is an acceptable method of testing it? I know that the current is in the range of 7 to 8 mA per lamp, but I have no idea of the voltage to expect. I also don't have any equipment for measuring particularly high voltages (nor the inclination to get too close to them).

Even considering the following, I'm still keen to test the inverter.

OK, some further research indicates that CCFL, like many discharge lamps, have a high striking voltage, and are then relatively low voltage devices driven from a current source. Too low a current may lead to premature death of the lamp :-(



With the LCD screens it is common practice to change both the lamps and their respective power supply at the same time. Both can be warn down, and there is no point in only changing one without the other.

Carefully evaluate the cost of servicing in relation to buying a new set.


Jerry G.
 
Top