TFT backlight shuts off after 2 seconds..

[[email protected]]

I have a more than one tft screen where the backlight shuts off after
~2 seconds. By turning the monitor off and on again I get the
backlight to work for another 2s.
The screen functionality for the TFT matrix seems ok. And because the
tubes actually gives light even if only for 2s I assume they work. And
conclude it's the backlight inverter pcb.

My general understanding of this pcb is that it will take DC (usually
12V?) and chop it up and then transform it up by some means of smps. I
have access to multimeter and oscilloscope. Considering this there
should be a at least one power semiconductor acting as a switch.
Controller circuit and an output transforrmer.

The switch should have pulsed dc input and a pulsed voltage should
occour over the emitter-collector (or source-drain etc..). The
transformer primary should have pulsed dc aswell. Any break in this
chain should limit where the fault might be.
The output of the transformer (1300 kVrms?) is proberbly not a good
idea to measure as it's likely to blow the measurement equipment.

What strategy should one use..?

(I have been thinking about using a CCFL driver used for computer
decoration, but those lack brightness control..)

Two of the screens with these symptoms are:
NEC LCD1810 (Backlight HIU-391C)
ViewSonic VE710b

 

[Arfa Daily]

I have a more than one tft screen where the backlight shuts off after
~2 seconds. By turning the monitor off and on again I get the
backlight to work for another 2s.
The screen functionality for the TFT matrix seems ok. And because the
tubes actually gives light even if only for 2s I assume they work. And
conclude it's the backlight inverter pcb.

My general understanding of this pcb is that it will take DC (usually
12V?) and chop it up and then transform it up by some means of smps. I
have access to multimeter and oscilloscope. Considering this there
should be a at least one power semiconductor acting as a switch.
Controller circuit and an output transforrmer.

The switch should have pulsed dc input and a pulsed voltage should
occour over the emitter-collector (or source-drain etc..). The
transformer primary should have pulsed dc aswell. Any break in this
chain should limit where the fault might be.
The output of the transformer (1300 kVrms?) is proberbly not a good
idea to measure as it's likely to blow the measurement equipment.

What strategy should one use..?

(I have been thinking about using a CCFL driver used for computer
decoration, but those lack brightness control..)

Two of the screens with these symptoms are:
NEC LCD1810 (Backlight HIU-391C)
ViewSonic VE710b

The inverters may be shutting down as a result of detecting an abnormal
load - such as a defective lamp - or because the system is telling them to,
as it would if going into power save. It could also be that the inverter
itself is faulty. You could start by looking to see if there is any input
line to the inverter board, which changes state when the lamps go out.

Many parts available from http://www.lcdparts.net/

Arfa

 

[[email protected]]

The inverters may be shutting down as a result of detecting an abnormal
load - such as a defective lamp - or because the system is telling them to,
as it would if going into power save. It could also be that the inverter
itself is faulty. You could start by looking to see if there is any input
line to the inverter board, which changes state when the lamps go out.

The lamp seems to work (from what I remember) if just for 2s. And I
have not spotted any change in voltage on the input/connections to the
ccfl-inverter with the multimeter on board before and after shutdown.
The monitor doesn't indicate any powersave mode. I think I even could
see very dimly the stuff that's supposed to be on the screen.
So I'm looking for a fault search on the inverter board.

lcdparts.com are nice, but they charge 70 USD, and with freight, vat
etc.. it doesn't make sense.

 

[JW]

I have a more than one tft screen where the backlight shuts off after
~2 seconds. By turning the monitor off and on again I get the
backlight to work for another 2s.
The screen functionality for the TFT matrix seems ok. And because the
tubes actually gives light even if only for 2s I assume they work. And
conclude it's the backlight inverter pcb.

My general understanding of this pcb is that it will take DC (usually
12V?) and chop it up and then transform it up by some means of smps. I
have access to multimeter and oscilloscope. Considering this there
should be a at least one power semiconductor acting as a switch.
Controller circuit and an output transforrmer.

The switch should have pulsed dc input and a pulsed voltage should
occour over the emitter-collector (or source-drain etc..). The
transformer primary should have pulsed dc aswell. Any break in this
chain should limit where the fault might be.
The output of the transformer (1300 kVrms?) is proberbly not a good
idea to measure as it's likely to blow the measurement equipment.

What strategy should one use..?

(I have been thinking about using a CCFL driver used for computer
decoration, but those lack brightness control..)

Two of the screens with these symptoms are:
NEC LCD1810 (Backlight HIU-391C)
ViewSonic VE710b

That particular model has a habit of blowing one or more picofuses for no
apparent reason. Check all the fuses on the inverter - likely one is
blown. I fixed several of those with the same problem.

 

[Ol' Duffer]

I have a more than one tft screen where the backlight shuts off after
~2 seconds. By turning the monitor off and on again I get the
backlight to work for another 2s.

CCFL backlight driver IC's are usually configured to allow an
initial surge of current to get things started, then current
limiting kicks in via a capacitor charging up. The symptom
you describe is usually caused by worn-out tubes that will no
longer stay lit at the current limit levels designed into the
circuit.

 

[Andy Cuffe]

I have a more than one tft screen where the backlight shuts off after
~2 seconds. By turning the monitor off and on again I get the
backlight to work for another 2s.
The screen functionality for the TFT matrix seems ok. And because the
tubes actually gives light even if only for 2s I assume they work. And
conclude it's the backlight inverter pcb.

My general understanding of this pcb is that it will take DC (usually
12V?) and chop it up and then transform it up by some means of smps. I
have access to multimeter and oscilloscope. Considering this there
should be a at least one power semiconductor acting as a switch.
Controller circuit and an output transforrmer.

The switch should have pulsed dc input and a pulsed voltage should
occour over the emitter-collector (or source-drain etc..). The
transformer primary should have pulsed dc aswell. Any break in this
chain should limit where the fault might be.
The output of the transformer (1300 kVrms?) is proberbly not a good
idea to measure as it's likely to blow the measurement equipment.

What strategy should one use..?

(I have been thinking about using a CCFL driver used for computer
decoration, but those lack brightness control..)

Two of the screens with these symptoms are:
NEC LCD1810 (Backlight HIU-391C)
ViewSonic VE710b

If it has more than one bulb, then one of them is probably not
lighting at all. Most inverters will shut down unless all the bulbs
are working. I've never seen a bulb fail completely and cause this,
but one of the inverter sections often fails. To figure out which one
it is, power it up with only one bulb at a time connected until you
find the one that's not working. Then you at least know which part of
the inverter to focus on. To check the bulb, simply plug it into one
of the inverter's other outputs. Typical failures include failed
transformers, shorted transistors and diodes, and open fuses.
Andy Cuffe

[email protected]

 

[Michael Kennedy]

The lamp seems to work (from what I remember) if just for 2s. And I
have not spotted any change in voltage on the input/connections to the
ccfl-inverter with the multimeter on board before and after shutdown.
The monitor doesn't indicate any powersave mode. I think I even could
see very dimly the stuff that's supposed to be on the screen.
So I'm looking for a fault search on the inverter board.

lcdparts.com are nice, but they charge 70 USD, and with freight, vat
etc.. it doesn't make sense.

Check to make sure both ccfl lamps are working.. If only one is working it
will usually do exactly what you are describing.

Mike

 

[jango2]

I just fixed a similar problem in a Sony KLV-L32M1. The service manual
did'nt have a schematic of the invertor section, so i pulled up
datasheets of the driver ICs and the main controller. The main
controller had a selectable option of reporting CCFL and peripheral
circuitry failure by either turning off all CCFLs together, or (by
removing a ) ohm to ground jumper on 1 of its pin thereby lifting the
pin high) running it in IND (individual) mode where only the faulty
drive circuit and CCFL would be turned off. Hence the tv would now
switch on with all lamps glowing while the defective one would fire
and then quit. Replacing the faulty lamp (which mind you seemed to be
fine, it glowed if forced, slight blackening at one end observed)
solved the problem. Returned the ic to group mode by re-installing the
grounding jumper.

 

[[email protected]]

I have a more than one tft screen where the backlight shuts off after
~2 seconds. By turning the monitor off and on again I get the
backlight to work for another 2s.
The screen functionality for the TFT matrix seems ok. And because the
tubes actually gives light even if only for 2s I assume they work. And
conclude it's the backlight inverter pcb.

This page
http://www.hackaday.com/2005/11/01/lcd-backlight-replacement/
was my inspriation when I fixed my Sony LCD.
I didn't replace the bulbs, but I did use an external invertor PCB to
drive the bulbs. I used one from a cold cathode case mod kit.

 

[[email protected]]

This pagehttp://www.hackaday.com/2005/11/01/lcd-backlight-replacement/
was my inspriation when I fixed my Sony LCD.
I didn't replace the bulbs, but I did use an external invertor PCB to
drive the bulbs. I used one from a cold cathode case mod kit.

I have considered this path aswell, but the catch is that then you can
no longer adjust brightness..
Any idea on how to solve that ..?