Anyone have experience with ccfl/inveters in TVs?

[Bobblank77]

Hey all, I bought a "Broken" TV for really cheap and am trying to fix it up. So the problem is that the ccfl backlight shuts off after ~30 mins of being on. Here's what ive found out

• All the caps other components on the main power supply board look fine. No obvious bulging/burn marks
• It looks the the main board provides high voltage AC to the inverter boards which also have a bunch of transformers
• There is nothing obviously wrong with the inverter board that I have access to (I havent taken a look at the other one as it is below the LCD - kinda scared to move it)
• The signals coming from the main board to the power supply are normal (Inverter on pin is high, PWM for brightness is working fine)
• High voltage caps on main PS board remain charged at ~380v even when ccfls are shut off

Anyone have any idea what this problem could be? I'm thinking its not the ccfl tubes themselves as the backlight definitely turns on.
Also there is a ribbon cable between the inverter and power supply with markings vcc, FB1, FB2, gnd, and LB. Any idea what these mean? FB are always 0v when i measured and LB is 12v

I'm not sure how to test the inverter board when its on because I believe that it is too much for my multimeter. Any ideas?

Thanks everyone!

[paulhm81]

FB should be feedback. If you have a spare ccfl you could try to test it by placing it instead of the original one by one. Usually it shuts when it senses something wrong in the feedback. It could also be one of the inverters. If it works with the spare ccfl you should look at the end of the ccfl for burned solder joint or burned isolation. If the isolation is burned it could ark to the chassis and the feedback would kick in.

[Bobblank77]

Thanks! I will take a look at it!

[tautech]

Got a monitor that does similar. 
Improved but not yet completely fixed but it's a good bet to work through replacing the PSU secondary side caps and those on the supply rail for the backlight inverter.
Basic cap value checks will give some idea as to those that need replacing and I set a limit of ~5% below labeled rating should be replaced.

[Bobblank77]

I don't think it would be the caps. They all look fine. I did a bit more poking around and I found that if I have a fan blow across the power supply the TV stays on for much longer (14+ hours rather than 3). So it looks like the heat is causing something to fail. After the TV backlight goes out I checked the high voltage input between the PSU and the inverter and got nothing. My plan is to go backwards from there to try and figure out whats going on. Do you know if MOSFETs can temporarily fail open due to heat? Maybe a bad solder joint between the die and pins that opens the circuit due to thermal expansion?

[tautech]

I don't think it would be the caps. They all look fine.

Don't be deceived, they can look perfect and be completely f***ed.
A basic uF capacitance test with a DMM will find the culprits, any 5% from nominal value....biff.

I did a bit more poking around and I found that if I have a fan blow across the power supply the TV stays on for much longer (14+ hours rather than 3). So it looks like the heat is causing something to fail. After the TV backlight goes out I checked the high voltage input between the PSU and the inverter and got nothing. My plan is to go backwards from there to try and figure out whats going on.

A pic of your PSU might help with some more ideas.

Do you know if MOSFETs can temporarily fail open due to heat? Maybe a bad solder joint between the die and pins that opens the circuit due to thermal expansion?

Quite possibly solder joints.....seen these types of intermittent faults before.
A PCB in a small Mig welder had a heat related fault, I was positive it was a cracked joint but could never find it, even with magnification but a hand reflow of ALL the componentry fixed it. 

[Bobblank77]

Okay Ill check the caps when i open it again. Ill post a pic too.

[Bobblank77]

Quick question for you Tautech. Do I have to desolder the caps before taking measurements?

[tautech]

Quick question for you Tautech. Do I have to desolder the caps before taking measurements?

Best to.
Depends sometimes as to how many caps are on the same rail/track.
Since Ecaps are a common point of failure there's often no time or effort wasted by removing for tests as you mostly have to replace then anyway.
Sometimes when the solder is sucked away, the lead hole is large enough to give disconnection of 1 pin and that's all you need. 

In the old days when PCB's weren't so densely populated and mostly one sided an accepted technique was to isolate a component by cutting a track and resoldering after testing. Occasionally I still do so. 
Known as the razor blade technique.

[Bobblank77]

Here are some pictures of the board. I took measurements of the caps.

Red one measured 625nF when its supposed to be 684nF
Green boxed ones are wired in parallel. Measures 183 uF when its supposed to 200 uF.
Blue ones are 80pf each spot on.

Should I be replacing any of these?
Ill upload what ive figured out of the schematic soon

[Bobblank77]

Here's what i could figure out of the schematic. Sorry its not much

[tautech]

The inverter (top right) is likely sourcing it's supply from the LV (top left) and that's where I'd be focussing my efforts on those LV caps.
If you DMM has sufficient frequency range and you can get to the LV rails with it powered on you might be able to measure the ripple on the AC range. You'd want to see much less than 100mV ripple on the DC rails.
It's likely each rail's voltage is marked on either the PSU PCB or the main MB so try to see each rail is up to voltage.

[Bobblank77]

What are LV caps? Low Voltage? should I be looking at those electrolytics on the top left?

[tautech]

What are LV caps? Low Voltage? should I be looking at those electrolytics on the top left?

Yep x3

[Bobblank77]

Just measured all the LV caps. Everything checks out. Also no jitter on the 12V rail

[Bobblank77]

The gates of the mosfets that drive the HV transformer are connected to a smaller transformer (Just north of the red boxed capacitor in the picture). I measured the AC voltage across this transformer and it went from 2.8v to 0.7v to 0.1v when the backlight shut off. I guessing that whatever is controlling this is at fault for the backlight failure.