Strange action of TI chip

[[email protected]]

Bought some old LCD monitors with non-functioning backlights. In
examining the CCFL driver circuit, it appears to be at Texas
Instruments TL1451 PWM regulator chip driving essentially a solid state
tesla coils for the HV/HF output.

Did some checking with my scope and found the 1451 didn't appear to be
switching the output transistors, so I replaced it with a new 1451 chip
and everything works great. I shut the power off and come back 10
minutes later to button up the case and the lights won't turn on again!

To make a long story short, through trial and error, I found that if I
touch any lead of the chip for a few seconds with a soldering iron
(heat it up) the chip works and back lights come on. If it cools off
to room temp, the chip won't work. So now I have two monitors with the
original chips and two monitors with new chips that do the same thing.
I sat down with one chip and could literally touch a lead for 3 seconds
with the soldering iron and get it to function normally, then put a
couple drops of liquid freeze on it and the chip would quit.

Anybody know why my chips work at say 120F, but not at room temp?

Thanks
Corey

 

[G]

Howdy Corey.....No specific help but it sure sounds more like a
mechanical connection that's being influenced by the heat rather than
a circuit problem (inside or outside of the chip). Is the 1451 an SMD?
Was wondering if maybe you could socket it otherwise. Might just be a
case of a little too much heat when soldering if it's seperating inside
the chip perhaps. Looking forward to see what you might find

Gord

 

[Damir]

Bought some old LCD monitors with non-functioning backlights. In
examining the CCFL driver circuit, it appears to be at Texas
Instruments TL1451 PWM regulator chip driving essentially a solid state
tesla coils for the HV/HF output.

Did some checking with my scope and found the 1451 didn't appear to be
switching the output transistors, so I replaced it with a new 1451 chip
and everything works great. I shut the power off and come back 10
minutes later to button up the case and the lights won't turn on again!

To make a long story short, through trial and error, I found that if I
touch any lead of the chip for a few seconds with a soldering iron
(heat it up) the chip works and back lights come on. If it cools off
to room temp, the chip won't work. So now I have two monitors with the
original chips and two monitors with new chips that do the same thing.
I sat down with one chip and could literally touch a lead for 3 seconds
with the soldering iron and get it to function normally, then put a
couple drops of liquid freeze on it and the chip would quit.

Anybody know why my chips work at say 120F, but not at room temp?

Thanks
Corey

Do you have some electrolytic caps near the chip?

 

[Dave D]

On 5 Mar 2006 07:37:29 -0800, [email protected] wrote:

Do you have some electrolytic caps near the chip?

That's exactly what I was going to ask. I'll bet the heat from the IC is
simply being conducted to nearby faulty electrolytic caps.

Dave

 

[[email protected]]

Gord - Thanks for the input. I guess at this point I would be willing
to consider anything. The 1451 is a 16-SOP package, so it is surface
mount and pretty tiny!. I took a look at my soldering under a 40x
microscope...it looked acceptable, but who knows!

Damir - Yes there are a couple of electrolytics near the chip. I put a
scan of the board up at:

http://coreyonline.tripod.com/CCFLdriver.jpg

C1 is 100uf, 25V and C2 is 10uF, 25V. These are really the only two
electrolytics on the board. There is a filter cap (blue) for each HV
output and then some surface mounts scattered on the board.

 

[[email protected]]

Gord - Thanks for the input. I guess at this point I would be willing
to consider anything. The 1451 is a 16-SOP package, so it is surface
mount and pretty tiny!. I took a look at my soldering under a 40x
microscope...it looked acceptable, but who knows!

Damir - Yes there are a couple of electrolytics near the chip. I put a
scan of the board up at:

http://coreyonline.tripod.com/CCFLdriver.jpg

C1 is 100uf, 25V and C2 is 10uF, 25V. These are really the only two
electrolytics on the board. There is a filter cap (blue) for each HV
output and then some surface mounts scattered on the board.

 

[[email protected]]

Gord - Thanks for the input. I guess at this point I would be willing
to consider anything. The 1451 is a 16-SOP package, so it is surface
mount and pretty tiny!. I took a look at my soldering under a 40x
microscope...it looked acceptable, but who knows!

Damir - Yes there are a couple of electrolytics near the chip. I put a
scan of the board up at:

http://coreyonline.tripod.com/CCFLdriver.jpg

C1 is 100uf, 25V and C2 is 10uF, 25V. These are really the only two
electrolytics on the board. There is a filter cap (blue) for each HV
output and then some surface mounts scattered on the board.

 

[Damir]

Gord - Thanks for the input. I guess at this point I would be willing
to consider anything. The 1451 is a 16-SOP package, so it is surface
mount and pretty tiny!. I took a look at my soldering under a 40x
microscope...it looked acceptable, but who knows!

Damir - Yes there are a couple of electrolytics near the chip. I put a
scan of the board up at:

http://coreyonline.tripod.com/CCFLdriver.jpg

C1 is 100uf, 25V and C2 is 10uF, 25V. These are really the only two
electrolytics on the board. There is a filter cap (blue) for each HV
output and then some surface mounts scattered on the board.

Check or replace C2.

 

[Jerry G.]

Look for high ESR caps in the supply that you are working on, and
replace as necessary.

--

Jerry G.


Bought some old LCD monitors with non-functioning backlights. In
examining the CCFL driver circuit, it appears to be at Texas
Instruments TL1451 PWM regulator chip driving essentially a solid state
tesla coils for the HV/HF output.

Did some checking with my scope and found the 1451 didn't appear to be
switching the output transistors, so I replaced it with a new 1451 chip
and everything works great. I shut the power off and come back 10
minutes later to button up the case and the lights won't turn on again!

To make a long story short, through trial and error, I found that if I
touch any lead of the chip for a few seconds with a soldering iron
(heat it up) the chip works and back lights come on. If it cools off
to room temp, the chip won't work. So now I have two monitors with the
original chips and two monitors with new chips that do the same thing.
I sat down with one chip and could literally touch a lead for 3 seconds
with the soldering iron and get it to function normally, then put a
couple drops of liquid freeze on it and the chip would quit.

Anybody know why my chips work at say 120F, but not at room temp?

Thanks
Corey

 

[Jamie]

Bought some old LCD monitors with non-functioning backlights. In
examining the CCFL driver circuit, it appears to be at Texas
Instruments TL1451 PWM regulator chip driving essentially a solid state
tesla coils for the HV/HF output.

Did some checking with my scope and found the 1451 didn't appear to be
switching the output transistors, so I replaced it with a new 1451 chip
and everything works great. I shut the power off and come back 10
minutes later to button up the case and the lights won't turn on again!

To make a long story short, through trial and error, I found that if I
touch any lead of the chip for a few seconds with a soldering iron
(heat it up) the chip works and back lights come on. If it cools off
to room temp, the chip won't work. So now I have two monitors with the
original chips and two monitors with new chips that do the same thing.
I sat down with one chip and could literally touch a lead for 3 seconds
with the soldering iron and get it to function normally, then put a
couple drops of liquid freeze on it and the chip would quit.

Anybody know why my chips work at say 120F, but not at room temp?

Thanks
Corey

seems strange, always thought the 1451 was a dual generic op-amp?
anyways, i would test the by pass cap at the Vcc, Vee area.
and other components in that area for solder cracks.
could even be the driver transistor that is just around the
area.

 

[Jamie]

seems strange, always thought the 1451 was a dual generic op-amp?
anyways, i would test the by pass cap at the Vcc, Vee area.
and other components in that area for solder cracks.
could even be the driver transistor that is just around the
area.

excuse me. it was the 1458 i was thinking of )

 

[[email protected]]

All,

Thanks for all the posts. As usual, the wealth of information on this
forum is stunning. As suggested, I dug deeper into the caps and found
some issues. I replaced them as suggested and now have all the
inverters operational hot or cold.

Thanks again for the input.

Corey

 

[budgie]

All,

Thanks for all the posts. As usual, the wealth of information on this
forum is stunning. As suggested, I dug deeper into the caps and found
some issues. I replaced them as suggested and now have all the
inverters operational hot or cold.

I'm sure all readers would benefit from knowing what those issues were.

 

[[email protected]]

The issues were that the C2 caps Damir suggested appeared to be shorted
- at least to the extent that my meter can test. I replaced that cap
on each of the 4 boards and all were immediately brought back to life.

Going back to my OP, I still don't know why the chip would work when I
warmed it up and fail at room temp when the board had the bad cap.
Maybe a warm chip was slightly less sensitive to input voltage
fluctuation with the bad cap?

Corey

 

[Dave D]

The issues were that the C2 caps Damir suggested appeared to be shorted
- at least to the extent that my meter can test. I replaced that cap
on each of the 4 boards and all were immediately brought back to life.

Going back to my OP, I still don't know why the chip would work when I
warmed it up and fail at room temp when the board had the bad cap.
Maybe a warm chip was slightly less sensitive to input voltage
fluctuation with the bad cap?

Corey

If there's one component most influenced by temperature changes it's a
failing electrolytic cap. You've fixed the problem by changing the cap,
forget about the IC, it's a red herring!

Dave