HP w20 LCD Display Inverter, running out of options.

[CapBlown]

Re: HP w20 LCD Display Inverter, running out of options.

A few questions:
- When you did the heating and freezing you say "without any success",
does that mean it still takes 1 hour before the monitor shuts down ?

- Yes and no. I heated the board and caps with hairdryer, yet it took about an hour or so before backlights shut down.

- I've applied cold spray to components, one by one and when I got to HV caps it once booted and stayed on longer, so I switched those caps but the problem remained. And later I've noticed that sometimes (very rarely) it may boot up almost immediately afted lights go off.

- Monitor is gettin much worse again. It now keeps lights on only about 15 to 20 minutes which is close to what it was when I got it.

- Does it take longer to shut down if the brightness is set at ~10% vs 90% ?

- There seems to be no correlation anymore. For instance last evening:

50% brightness:
-> 28 minutes on -> Lights off-> Turned off/on -> 2 secs to black
-> 15 minutes off
-> 14 minutues on -> Lights off -> Turned off/on ->2 secs to black
-> 16 minutes off
-> 2 hours 14 minutes on->Lights off -> Turned off/on -> 2 secs to black

Today:
50% brighness:
-> 34 minutes on -> Lights off -> Turned off/on ->2 secs to black
-> 4 minutes off
-> turned brighness to 90%
-> 1 hour and 4 minutes on -> Lights off -> Turned off/on ->2 secs to black
-> 2 hours off while I watched a movie
-> 12 minutes on -> Lights off -> Turned off/on ->2 secs to black
-> 5 minutes off
-> 15 minutes on -> Lights off -> Turned off/on ->2 secs to black
-> 1-2 minutes off
-> already 20 minutes on -> Lights off -> Left it (on) for 15 minutes
-> Turned off/on and on immediately.

- Once it is shut down, can you turn the AC power off then back on immediately and does it take yet another hour before it will shut down again ?

I can turn it on but 95% of the cases it's 2 secs to black. Usually it has to cool down first. Sometimes it takes only a few minutes to get lights back on, then it usually does not last on for very long before lights go off again. It may stay on for an hour even so (this really drives me crazy).

When backlights shut down, green light always stays on and LCD and sound works fine. If I let it stay on "on" for about 15 minutes and then turn it off/ on again, it starts operating normally. It heals even with powers on as long as backlights are off for a while.

- Once it is shut down, does freeze spraying make it recover ?

No, and I've tried this several times.

- Are the main voltages like 5V, 12V from the P/S still present when it is shut down ?

Power is there and when it's on operating normally voltages are 5.1 V and 12.5 V. When lights go off they change to 4.95 V and 14.11 V and they stay there until power is switched off/on again.

CCFL's are allright, with rev 3 board this monitor does not shut down.

On the rev2 to rev3 question, it looks like there was a functional change needed on
the rev2 boards and was reworked into them. The rev3 boards has this "rework" all
layed out neatly with etch on the pcb. To know the reason why they did this is going
to be hard to find out unless you can get hold of the service bulletin or ECO
(engineering change order). This change is not your problem at this moment. This
monitor was working x years before it failed. What you need to find out what has
failed and replace it/them and first have a working monitor. Then you can go and
investigate (if you want to) what this change is between rev2 to rev3 and decide if
this change affects your specific situation. Changes are sometimes specific like you
only need this change if you are operating the monitor in 35-40 degree C ambient.

Thanks. I just got curious as that line going to logic board had 1000 ohms instead of zero in rev 3. It looked like a huge difference to me.

[CapBlown]

Re: HP w20 LCD Display Inverter, running out of options.

I measured all wires between logic and inverter board. Voltages change when lights go off (or because they go off)

Brown 1.67 V -> 1.60 V
Yellow 3.05 V -> 0.58 V
Blue 5.11 V -> 4.91 V
Red 12.76 -> 14.10 V
Extra Black 0.01 V -> 0.01 V

So I guess yellow line is the one that keeps monitor going? Something causes it's voltage to drop down.

[CapBlown]

Re: HP w20 LCD Display Inverter, running out of options.

I measured all wires between logic and inverter board. Voltages change when lights go off (or because they go off)

Yellow 3.05 V -> 0.58 V

So I guess yellow line is the one that keeps monitor going? Something causes it's voltage to drop down.

After letting it rest for 15 minutes, Yellow gets now to 0.61 V and lights lid up.

[Aruba]

Re: HP w20 LCD Display Inverter, running out of options.

After letting it rest for 15 minutes, Yellow gets now to 0.61 V and lights lid up.

I am confused. On your previous post you mention that the "yellow" 3.05V
turns to 0.58V when the monitor shuts off. On this post you say the 3.05V
goes to 0.61V and yet the monitor turns on. Am I reading this right ?

[CapBlown]

Re: HP w20 LCD Display Inverter, running out of options.

That's what I thought I measured... Yellow is now 0.62 V at the beginning. Tested it also with power off now and it's 4.51 V... Maybe I made a mistake last time. I better measure these several times in case there really is some strange variation.

[lucky13]

Re: HP w20 LCD Display Inverter, running out of options.

Did you manage to get another set of CCFL lamps to swap in to rule out the lamps?

Your description sure sounds like a dying lamp to me....

You don't need 20" lamps. Even 17" will do as you are only testing them.

[retiredcaps]

Re: HP w20 LCD Display Inverter, running out of options.

Did you manage to get another set of CCFL lamps to swap in to rule out the lamps?

Your description sure sounds like a dying lamp to me....

I think in post #36, he moved the inverter into another LCD and it fails too. So it looks like he narrowed down the problem to the inverter.

On a similar note, I finally was able to get 2 identical LCDs and figured out by moving power/inverter board into the other LCD that the original LCD panel is likely dying. One LCD had similar problems with it working for 15 to 20 minutes and then shutting off. If you let it cool, it will work again.

[lucky13]

Re: HP w20 LCD Display Inverter, running out of options.

I think in post #36, he moved the inverter into another LCD and it fails too. So it looks like he narrowed down the problem to the inverter.

My bad! Yes he did all that already to narrow down to the inverter board.

Well then, the other possibility is that the output transformer are going south. Do you have a ring meter to test?

[retiredcaps]

Re: HP w20 LCD Display Inverter, running out of options.

Well then, the other possibility is that the output transformer are going south. Do you have a ring meter to test?

That was what I was thinking to try next too. He has already resoldered everything on the board (more than once I believe).

[CapBlown]

Re: HP w20 LCD Display Inverter, running out of options.

I have to admit that I have no idea what is ring meter or what "output transformer are going south" means.

Please explain.

P.S I have only cheap multimeter.

[PlainBill]

Re: HP w20 LCD Display Inverter, running out of options.

I have to admit that I have no idea what is ring meter or what "output transformer are going south" means.

Please explain.

P.S I have only cheap multimeter.

'Going south' is a slang term for failing (see below for the history of this).

One possible cause of an inverter failure is an open or a short in the inverter transformer. A 'ring tester' tests a transformer by hitting it with a pulse and determining the number of oscillations (rings) that occur after each pulse. A transformer with shorted turns will have fewer oscillations; an open winding will have none.

The origin of 'Going south' is fairly recent. When making a line chart of sales of a product, the line rises toward the top of the chart with increasing sales, and of course drops with decreasing sales. By convention, maps are drawn with North on the top and South on the bottom. Thus if product sales are dropping, they (and the viability of the product) are 'Going South'.

PlainBill

[PlainBill]

Re: HP w20 LCD Display Inverter, running out of options.

What is the part number of IC1 - the 16 pin IC in the inverter area? Perhaps it will be possible to determine what is triggering the shutdown.

PlainBill

[retiredcaps]

Re: HP w20 LCD Display Inverter, running out of options.

P.S I have only cheap multimeter.

You can see a ring tester in action with this youtube video. 8 min 8 secs long

http://www.youtube.com/user/preherte.../1/HMUazhb0bhc

You can try this test with a multimeter, but it won't find what PlainBill is talking about.

I can't make out the part number of the transformers, but measure the resistance between the secondary pins. Since I don't know which one is the secondary pin, you can do this.

Number the pins on the transformers anyway you like.

black probe on pin 1, red on pin 2 - record ohm
black probe on pin 1, red on pin 3 - record ohm
black probe on pin 1, red on pin 4 - record ohm
etc
black probe on pin 1, red on pin last - record ohm

black probe on pin 2, red on pin 3 - record ohm
black probe on pin 2, red on pin 4 - record ohm
etc
black probe on pin 2, red on pin last - record ohm

black probe on pin 3, red on pin 4 - record ohm
etc

finally,

black probe on pin last-1, red on pin last - record ohm

I'm guessing the secondaries should read 950 ohm. The primaries should read 1.3 ohm. Some will be 0L. If one secondary is 950ohm and the other is 750ohm, that would indicate a problem.

[CapBlown]

Re: HP w20 LCD Display Inverter, running out of options.

Thank you very much for your comprehensive explanation to my questions PlainBill. It's always great to learn new terms and things.

What is the part number of IC1 - the 16 pin IC in the inverter area? Perhaps it will be possible to determine what is triggering the shutdown.
PlainBill

That would be great, to me this is beginning to look like rocket science.

IC is (attached):

OZ9939GN
A542E2.1G
0551A

You can see a ring tester in action with this youtube video. 8 min 8 secs long

http://www.youtube.com/user/preherte.../1/HMUazhb0bhc

You can try this test with a multimeter... ...that would indicate a problem.

Nice video, thanks retiredcaps.

I did measure transformers some time ago and they were pretty much identical, but would it be possible that they are ok when they are "cold" and after board is heating up for an hour, one of them gets heated enough to be shorted? That might explain why everything is working fine until... I would need to solder them out of the board when they are "hot", but that takes
at least 2-5 minutes while the problem may have disappeared already.

[retiredcaps]

Re: HP w20 LCD Display Inverter, running out of options.

I would need to solder them out of the board when they are "hot", but that takes at least 2-5 minutes while the problem may have disappeared already.

I'm *guessing* you should be able to measure them in circuit after the shutdown occurs, but this test is not as comprehensive as the ring tester.

Let's see what PlainBill says before you go down this "rabbit hole". It could be a "red herring".

[PlainBill]

Re: HP w20 LCD Display Inverter, running out of options.

Thank you very much for your comprehensive explanation to my questions PlainBill. It's always great to learn new terms and things.

That would be great, to me this is beginning to look like rocket science.

IC is (attached):

OZ9939GN
A542E2.1G
0551A

Cowabunga!!! We have a winner!!!

We're going to try working on this a little bit differently. Hopefully you can access IC1 while the monitor is in operation. The datasheet for the OZ9939GN is located here. This design appears to have both voltage and current protection, so let's monitor pins 5 and 6 to see which one changes and causes the shutdown. Use pin 14 as the ground point.

If you can't reach those points while the monitor is working, solder fine wires (wire wrap wires would be excellent) to the pins and run them to an accessible point. (Avoid high voltage points like the inverter output and the AC input).

PlainBill

[CapBlown]

Re: HP w20 LCD Display Inverter, running out of options.

Cowabunga!!! We have a winner!!!
We're going to try working on this a little bit differently. Hopefully you can access IC1 while the monitor is in operation. The datasheet for the OZ9939GN is located here. This design appears to have both voltage and current protection, so let's monitor pins 5 and 6 to see which one changes and causes the shutdown. Use pin 14 as the ground point.
PlainBill

Thanks PlainBill. I did as you suggested and measured also voltage from pin 2, which seems to stay at 4.98V when monitor is on.

Here are the results on pins 5 & 6. Seems that I get overvoltage at some point if I understood this correctly? What could cause this sort of peak?

Powers on 20% brightness (Or=Orange/Ow=orange/White):
Ow 5 ISEN: 0.52 V
Or 6 VSEN: 0.52 V

Set brightness to 90% (at 15 mins on)
Ow 5 iSEN: 1.23 V
Or 6 VSEN: 0.84 V

(30 mins on)
Ow 5 iSEN: 1.23 V
Or 6 VSEN: 0.84 V

(Lights go off at 40 mins)
Ow 5 ISEN: 0 V
Or 6 VSEN: 0 V

Turned off and on again, lights turned on immediately:
Ow 5 iSEN: 1.23 V
Or 6 VSEN: 0.84 V

(Lights go off 13 mins)
Ow 5 ISEN: 0 V
Or 6 VSEN: 0 V

Turning on gets these peak values and lights go off immediately
Ow 5 ISEN: 0.83 Vpeaks
Or 6 VSEN: 2.11 V peaks

Suddenly turns back on:

Ow 5 ISEN: 1.23 V
Or 6 VSEN: 0.84 V

(still 15 mins on)
Ow 5 ISEN: 1.22 V
Or 6 VSEN: 0.85 V

Lights go off... turning on gets these peak values and lights go off

Ow 5 ISEN: 0.47 V
Or 6 VSEN: 1.85 V

Couple of minutes off, and here we go again (15 mins on)
Ow 5 ISEN: 1.23 V
Or 6 VSEN: 0.83 V

[PlainBill]

Re: HP w20 LCD Display Inverter, running out of options.

Thanks PlainBill. I did as you suggested and measured also voltage from pin 2, which seems to stay at 4.98V when monitor is on.

Here are the results on pins 5 & 6. Seems that I get overvoltage at some point if I understood this correctly? What could cause this sort of peak?

Powers on 20% brightness (Or=Orange/Ow=orange/White):
Ow 5 ISEN: 0.52 V
Or 6 VSEN: 0.52 V

Set brightness to 90% (at 15 mins on)
Ow 5 iSEN: 1.23 V
Or 6 VSEN: 0.84 V

(30 mins on)
Ow 5 iSEN: 1.23 V
Or 6 VSEN: 0.84 V

(Lights go off at 40 mins)
Ow 5 ISEN: 0 V
Or 6 VSEN: 0 V

Turned off and on again, lights turned on immediately:
Ow 5 iSEN: 1.23 V
Or 6 VSEN: 0.84 V

(Lights go off 13 mins)
Ow 5 ISEN: 0 V
Or 6 VSEN: 0 V

Turning on gets these peak values and lights go off immediately
Ow 5 ISEN: 0.83 Vpeaks
Or 6 VSEN: 2.11 V peaks

Suddenly turns back on:

Ow 5 ISEN: 1.23 V
Or 6 VSEN: 0.84 V

(still 15 mins on)
Ow 5 ISEN: 1.22 V
Or 6 VSEN: 0.85 V

Lights go off... turning on gets these peak values and lights go off

Ow 5 ISEN: 0.47 V
Or 6 VSEN: 1.85 V

Couple of minutes off, and here we go again (15 mins on)
Ow 5 ISEN: 1.23 V
Or 6 VSEN: 0.83 V

Excellent job of testing, thanks for the data.

Forgive me, I haven't read the complete thread. Please verify this - the CCFLs suddenly turned off, then back on without any action on your part?

That is very interesting, and would really limit the number of possible causes. The OZ9939 uses a standard technique for regulating the outputs. The current through the CCFLs' return lines are measured, and the total current fed back to pin 5. It adjusts the output drive to maintain a constant CCFL current, based on the brightness setting. If a return line is suddenly shorted to ground, it appears this would not cause a fault condition.

The output voltage is also monitored. These are 'ored' and pin 6 senses the highest voltage across any CCFL. If this rises above the threshold, the inverter shuts down and stays down until the control input is cycled to off, then on.

It appears the ONLY way the CCFLS could go dark, then come back on again is if the voltage at pin 10 drops below 1 volt, then rises above 2 volts.

PlainBill

[CapBlown]

Re: HP w20 LCD Display Inverter, running out of options.

Excellent job of testing, thanks for the data.

Forgive me, I haven't read the complete thread. Please verify this - the CCFLs suddenly turned off, then back on without any action on your part?

Thanks PlainBill.

For the CCFL's to get back on I need to push Power button twice to get lights back on. First push shuts down the monitor, second lights it up. Image and sound are there even if I don't shut it down, just lights are missing.

Usually when I turn monitor off and on again, lights are immediately turned off again. But sometimes they lid back up and may stay on for some 10 to 20 minutes. There is no clear pattern and behaviour seems pretty random to me.

That is very interesting, and would really limit the number of possible causes. The OZ9939 uses a standard technique for regulating the outputs. The current through the CCFLs' return lines are measured, and the total current fed back to pin 5. It adjusts the output drive to maintain a constant CCFL current, based on the brightness setting. If a return line is suddenly shorted to ground, it appears this would not cause a fault condition.

The output voltage is also monitored. These are 'ored' and pin 6 senses the highest voltage across any CCFL. If this rises above the threshold, the inverter shuts down and stays down until the control input is cycled to off, then on.

It appears the ONLY way the CCFLS could go dark, then come back on again is if the voltage at pin 10 drops below 1 volt, then rises above 2 volts.

PlainBill

I was reading electrical characteristics of this IC chip, I noted that ISEN is on it's Max level (1.23 V) all the time (except when shit happens) while VSEN is not even close to it's minimums (2.78)? see page 4. I might be wrong though, these are the things I don't really understand.

I quess my next step is to measure pin 10. Should I check preset voltage at pin 7 too?

[PlainBill]

Re: HP w20 LCD Display Inverter, running out of options.

Thanks PlainBill.

For the CCFL's to get back on I need to push Power button twice to get lights back on. First push shuts down the monitor, second lights it up. Image and sound are there even if I don't shut it down, just lights are missing.

Usually when I turn monitor off and on again, lights are immediately turned off again. But sometimes they lid back up and may stay on for some 10 to 20 minutes. There is no clear pattern and behaviour seems pretty random to me.



I was reading electrical characteristics of this IC chip, I noted that ISEN is on it's Max level (1.23 V) all the time (except when shit happens) while VSEN is not even close to it's minimums (2.78)? see page 4. I might be wrong though, these are the things I don't really understand.

I quess my next step is to measure pin 10. Should I check preset voltage at pin 7 too?

The voltage at ISEN is normal. The inverter is a constant current source, so the current will stay at the regulation point. If you reduce brightness on the monitor, the voltage at ISEN should drop.

Go ahead an check the voltage at pin 10. I suspect it will be either well above 2 volts (on) or below 1 volt (when turned off from the front panel). I am also interested in the voltage at pin 7. I had expected to see Vsen much higher.

PlainBill