LG 42PC1RR-ZL y sus problem

[luiznunes]

Re: LG 42PC1RR-ZL y sus problem

only after 9 minutes the Vy is a stable -200V

[luiznunes]

Re: LG 42PC1RR-ZL y sus problem

the values of the output of the transformer are RMS AC values

[luiznunes]

Re: LG 42PC1RR-ZL y sus problem

I didn't fully inderstand how this dc-dc converter works..
when everything is ok, after 10 minutes, the V on the primary side of the transformer is higher that on the secundary. but at the begginnig, lets say first 6 minutes, it is the oposite..

[PlainBill]

Re: LG 42PC1RR-ZL y sus problem

I didn't fully inderstand how this dc-dc converter works..
when everything is ok, after 10 minutes, the V on the primary side of the transformer is higher that on the secundary. but at the begginnig, lets say first 6 minutes, it is the oposite..

A dc-dc converter is a stripped down version of the power supply used in a monitor. I'll describe it by saying that this one consists of a controller that generates pulses at around 30 - 70 Khz, a driver transistor that drives a transformer that steps the voltage up and isolates the input from the output. The transformer secondary feeds a diode which charges a cap. A voltage regulator senses the voltage across the cap and passes a signal back to the controller through an opto-isolator.

The key to the problem is the regulator. It works by using a voltage divider to drop the output voltage to 2.5 volts if the output is at the nominal output voltage. If the output voltage gets too high, the regulator conducts, sending a signal to the controller through the opto-isolator. The controller respdons by reducing the width of the pulse sent to the transformer.

The regulator is a TL431A. The key to troubleshooting is to measure the voltage at pin 1. If it slowly rises while the output voltage drops, there is probably a bad cap. If it stays steady, the problem is the regulator or the opto-isolator.

PlainBill

[luiznunes]

Re: LG 42PC1RR-ZL y sus problem

A dc-dc converter is a stripped down version of the power supply used in a monitor. I'll describe it by saying that this one consists of a controller that generates pulses at around 30 - 70 Khz, a driver transistor that drives a transformer that steps the voltage up and isolates the input from the output. The transformer secondary feeds a diode which charges a cap. A voltage regulator senses the voltage across the cap and passes a signal back to the controller through an opto-isolator.

The key to the problem is the regulator. It works by using a voltage divider to drop the output voltage to 2.5 volts if the output is at the nominal output voltage. If the output voltage gets too high, the regulator conducts, sending a signal to the controller through the opto-isolator. The controller respdons by reducing the width of the pulse sent to the transformer.

The regulator is a TL431A. The key to troubleshooting is to measure the voltage at pin 1. If it slowly rises while the output voltage drops, there is probably a bad cap. If it stays steady, the problem is the regulator or the opto-isolator.

PlainBill

So you think probably it is a bad cap or the TL431, right? I will check it. bu the way, did you saw the pdf with the graphic I posted 3 posts ago?

makes me confused how it can after a while get to the correct voltage.. in the begginnig, at the first seconds the Vy ir right, but after the 5th or 6th second it gets crazy for 8 or 9 minutes..

[luiznunes]

Re: LG 42PC1RR-ZL y sus problem

pin1, you mean the cathode, right? and measure between the pin 1 and the Vy's GND?

[PlainBill]

Re: LG 42PC1RR-ZL y sus problem

pin1, you mean the cathode, right? and measure between the pin 1 and the Vy's GND?

The unit I have uses a TL431A as the regulator. Pin 1 is the ref input, and is tied to the voltage divider. Pin 2 is the anode and pin 3 is the cathode. Measure the voltage between pin 1 and pin 3. The voltage should stay at 2.5 volts. If it does, the problem is in the components making up the voltage divider. If it goes higher, the problem is either the regulator or the opto-isolator.

PlainBill

[luiznunes]

Re: LG 42PC1RR-ZL y sus problem

The unit I have uses a TL431A as the regulator. Pin 1 is the ref input, and is tied to the voltage divider. Pin 2 is the anode and pin 3 is the cathode. Measure the voltage between pin 1 and pin 3. The voltage should stay at 2.5 volts. If it does, the problem is in the components making up the voltage divider. If it goes higher, the problem is either the regulator or the opto-isolator.

PlainBill

ok, this is what I measure. between pin 1 and 3, I found 0.6V. when the Vy reaches the 199V ( correct value ) between pin 1 and 3 I get between 2.6 V and 3V. In other words, the TL stays below 2.5 until the Vy reaches the correct value. but why it takes soooooo long to get there?

[luiznunes]

Re: LG 42PC1RR-ZL y sus problem

just to remember, the Vy begins too high ( 250V ) and then drops after 4 or 5 minutes

[luiznunes]

Re: LG 42PC1RR-ZL y sus problem

well.. too high, but negative.. so too low maybe ( lol )

[PlainBill]

Re: LG 42PC1RR-ZL y sus problem

ok, this is what I measure. between pin 1 and 3, I found 0.6V. when the Vy reaches the 199V ( correct value ) between pin 1 and 3 I get between 2.6 V and 3V. In other words, the TL stays below 2.5 until the Vy reaches the correct value. but why it takes soooooo long to get there?

You have it backwards. The voltage on pin 1 of the TL431 rises to 2.5 volts, at which point it starts conducting, turning on the emitter in the opto-isolator, causing the SMPS controller to reduce the drive to the power FET.

I've spent some time searching for anything to illustrate what is happening to no avail, so I was forced to draw the attached diagram. (Anyone laughing will be ignored).

Under normal operation as the voltage out rises the voltage into the R input of the TL431 rises until it reaches 2.5 volts. At that point the TL431 begins to conduct cathode - anode, causing current flow through the LED (5), reducing the drive to the power FET.

In your case, the voltage at R is increasing very slowly (several minutes rather than several milliseconds). There are several possible explanations for that - resistors 1, 2, or 3 changing value; the TL431 (6), or the capacitor (7) having a high leakage current. I'm inclined to think the capacitor is more likely to be the culprit, but don't rule anything out.

A simple test would be to remove the capacitor. On the supply I have it's the small ceramic cap next to R25, just above the TL431 (on the back of the board). If the voltage never rises above 200 volts, that confirms it.

Alternatively, you could remove the TL431, and observe the voltage at hole for pin 1. If it rises slowly, that tends to confirm the cap is the problem; if it rises quickly, the TL431 was the problem.

If you haven't already tried tweaking the pot, turn it through it's full range several times, then return it to the present position. (That should have been the first thing I suggested).

PlainBill

[luiznunes]

Re: LG 42PC1RR-ZL y sus problem

You have it backwards. The voltage on pin 1 of the TL431 rises to 2.5 volts, at which point it starts conducting, turning on the emitter in the opto-isolator, causing the SMPS controller to reduce the drive to the power FET.

I've spent some time searching for anything to illustrate what is happening to no avail, so I was forced to draw the attached diagram. (Anyone laughing will be ignored).

Under normal operation as the voltage out rises the voltage into the R input of the TL431 rises until it reaches 2.5 volts. At that point the TL431 begins to conduct cathode - anode, causing current flow through the LED (5), reducing the drive to the power FET.

In your case, the voltage at R is increasing very slowly (several minutes rather than several milliseconds). There are several possible explanations for that - resistors 1, 2, or 3 changing value; the TL431 (6), or the capacitor (7) having a high leakage current. I'm inclined to think the capacitor is more likely to be the culprit, but don't rule anything out.

A simple test would be to remove the capacitor. On the supply I have it's the small ceramic cap next to R25, just above the TL431 (on the back of the board). If the voltage never rises above 200 volts, that confirms it.

Alternatively, you could remove the TL431, and observe the voltage at hole for pin 1. If it rises slowly, that tends to confirm the cap is the problem; if it rises quickly, the TL431 was the problem.

If you haven't already tried tweaking the pot, turn it through it's full range several times, then return it to the present position. (That should have been the first thing I suggested).

PlainBill

already turn the pot through it's full range several times and nothing changes.

also removed the cap and everything stays the same. between pin 1 and 3 still 0.6V and Vy -253V..

[luiznunes]

Re: LG 42PC1RR-ZL y sus problem

I have a question : here, the job of the TL431 is to stop the rising of the voltage when it reaches the correct value, right?
this job, I think its done. when the Vy is -199, the TL431 reaches th 2.5V and the Vy stops rising

[Th3_uN1Qu3]

Re: LG 42PC1RR-ZL y sus problem

It's job isn't just to stop the rising of the voltage - it's to KEEP it at that value. Anyway, it sounds like the TL431 and associated circuitry of that power supply are working fine. Look for the problem elsewhere... or just buy a new board and be done with it.

[PlainBill]

Re: LG 42PC1RR-ZL y sus problem

I have a question : here, the job of the TL431 is to stop the rising of the voltage when it reaches the correct value, right?
this job, I think its done. when the Vy is -199, the TL431 reaches th 2.5V and the Vy stops rising

Poorly stated, but essentially correct - Th3_uN1Qu3 said it best, it regulates the output voltage.

The problem is simple - either something in parallel with resistor 3 in my sketch is holding the voltage at the R input of the TL341 low, or either resistor 1 or 2 is initially high in value.

Resolder the SMT resistors; a bad solder joint might do this.

Check for any other parallel resistances.

Remove the TL431. and check the voltage at the hole for pin 1. If it immediately jumps to over 2.5 volts when the set is turned on, replace the TL431. Don't operate the set with the TL431 removed for more than a minute.

PlainBill

[Th3_uN1Qu3]

Re: LG 42PC1RR-ZL y sus problem

I wouldn't remove the 431 if i were him... you never know how high that voltage can go.

[PlainBill]

Re: LG 42PC1RR-ZL y sus problem

I wouldn't remove the 431 if i were him... you never know how high that voltage can go.

That's right, you wouldn't. I tend to be cautious, and reviewed the tests performed. I am certain the voltage will go to 250 volts, and so would you if you had read the thread.

Luiznunes has already found that the output voltage starts at 250 volts. While this is happening, the voltage at the reference input starts at .6 volts. With the reference voltage at .6 volts, the Tl431 isn't conducting, the power supply is putting out the maximum voltage. What is more, since the TV works perfectly after the voltage drops, it is highly unlikely that an additional exposure to 250 volts is going to do any damage.

Now, if you have a TL431 you can deliver to luiznunes, please do so.

PlainBill

[Th3_uN1Qu3]

Re: LG 42PC1RR-ZL y sus problem

Luiznunes has already found that the output voltage starts at 250 volts. While this is happening, the voltage at the reference input starts at .6 volts. With the reference voltage at .6 volts, the Tl431 isn't conducting, the power supply is putting out the maximum voltage.

What you basically said here is that the TL431 operates in reverse in this power supply. Which i believe is not true. The previous reports of luiznunes indicate that there either isn't enough drive to the 431 - check resistor in TL431 cathode - or high leakage in the capacitor (which you have mentioned already). I would also replace the optoisolator.

Instead of simply removing the 431 i would replace the bottom feedback resistor (between ground and reference of 431) with a multiturn variable resistor of a larger value. Set to the value the old resistor had, power up the TV, and tweak the VR while watching the output voltage. But since the reference voltage starts low while output is high there is little point in doing this here - refer to the above paragraph first.

[PlainBill]

Re: LG 42PC1RR-ZL y sus problem

What you basically said here is that the TL431 operates in reverse in this power supply. Which i believe is not true. The previous reports of luiznunes indicate that there either isn't enough drive to the 431 - check resistor in TL431 cathode - or high leakage in the capacitor (which you have mentioned already). I would also replace the optoisolator.

Instead of simply removing the 431 i would replace the bottom feedback resistor (between ground and reference of 431) with a multiturn variable resistor of a larger value. Set to the value the old resistor had, power up the TV, and tweak the VR while watching the output voltage. But since the reference voltage starts low while output is high there is little point in doing this here - refer to the above paragraph first.

You are wrong. The TL431 conducts anode to cathode when the voltage at the reference input rises above 2.5 volts. That cause the LED in the optoisolator to begin emitting; the phototransistor begins conducting, and the drive pulses to the power fet are reduced. That is a very standard design, and can be verified by studying a schematic of a SMPS.

Your suggestion to replace the optoisolator is nonsense. Get it through your head - with the reference input below 2.5 volts, no current would be flowing through it. About the only thing you are right about is replacing the 'feedback' (it isn't) resistor is pointless.

Do yourself a favor - look at the schematic referenced in post 18 (note this is for a -75 volt supply, not 200 volt). STUDY IT!!! Compare it to the datasheet for the TL431.

Oh - your suggestion about the cathode resistor? Also nonsense. If it was conducting, it would REDUCE the output of the supply.

The fact is that unless I pointed out the wrong capacitor - something that is certainly possible - the problem is the TL431 OR some sort of contamination on the board, or a bad solder joint.

I WILL concede one point - removing the entire TL431 is not necessary - just isolating pin 1 would be sufficient.

PlainBill

[Th3_uN1Qu3]

Re: LG 42PC1RR-ZL y sus problem

Well, everybody makes mistakes, me included, so you help him fix his TV while i step aside and study the circuit in more detail.

What i do know about it is that in this circuit the 431 is not only used as a voltage reference but as a transconductance error amplifier, and my knowledge on this subject is slim, hence my initial confusion.