Problem with Gateway 900G

[Mad Capper]

I have acquired a Gateway 900G model FPD 1960.

I have read and tried to do as much as I could to give the best account of what is occurring.


The story so far ...

It does not do, from my understanding what I read to be, the 2 seconds on then black screen. i.e. the LCD does not flash a normal screen then turn black.
The power button is lit and stays on.
Screen is black.
LCD is not cracked.
When viewed using flash light technique everything is there and looks like nothing is wrong with the LCD itself.

Looking at PSU ...
No capacitors show any signs of bloating on top or leakage on bottom.
Transformers both show 712 Ohm
Power output shows both 5 VDC and 12 VDC
Both Alpha&Omega 4606 IC's (U301 & U302) show signs of being burnt/melted on the corner of what it think is Pin 1.
Fuse has continuity

I tried to run the tests described here: https://www.badcaps.net/forum/showpo...2&postcount=33

When plugged in but power button turned OFF
blue circle 14.46
a to gnd 0.01
b to gnd 2.76
pin 6 0.01
pin 10 2.74


Plugged in and power button to ON
blue circle 14.62
a to gnd .01 holds until I hear a zits from IC then jumps to 3.02 when there is no signal then drops to 0 and simultaneously power button light turn from blue to amber. when there is an incoming monitor signal at this point it drops from the 3.02 to 2.98-9
b to gnd 2.74
pin 6.01
pin 10 2.73

I don't know if the tests described here should be done unplugged, plugged in and off, or plugged in and on
https://www.badcaps.net/forum/showpo...0&postcount=21

While using a flash light to see I tried to change the Brightness and Contrast and could not see any change. I don't know if this is a function of, and thus points to a problem with, the CCFL's or the two burnt IC's.

The logic board seems to be clean.

Additionally, as seen in the attached picture the red arrow points to what seems to be a failed capacitor (C325), with what looks like leakage on the board but I am uncertain. Is there a way to test the capacitor and if bad does anyone know its value so as I can replace it?

So the question I have is, Is everything functioning correctly and the only problem is the two burnt IC's(U301, U302)?

Are there any other tests I can perform to be certain that this is the only problem before spending the time and money to replace them?

What is the likelihood that, if they were replaced, they would burn up in a very short time? i.e. how can I tell if they are the problem and not something else which is burning up those IC's as a symptom?

Is there a way that I can test the CCFL's(without buying a case mod kit) to see if they are also a problem that I cannot see due to the inverter portion seemingly the culprit?


Anything else I am missing?


Thanks a bunch.
I appreciate any assistance you can provide.

[jetadm123]

Re: Problem with Gateway 900G

WoW! Great job on testing and describing the symptoms. Also, great job with the photos. You've verified the voltage outputs are good and the logic board is outputting video. And you've narrowed it down to the inverter section. Yes, U301 and U302 do look burnt. Try testing them for shorts:

1) look at http://www.alldatasheet.com/datashee...MD/AO4606.html for the datasheet.

2) with power off and your meter set to 200ohms resistance, place your probes across the following:

S1-G1
G1-D1
S1-D1
S2-G2
G2-D2
S2-D2

As for testing the ccfl's, if you have a working monitor lying around you can try plugging them in to that monitor. Otherwise, I can't think of any way to test them with your monitor in it's current state.

Also, caps C301 and C302 could be failing, thus causing U301 and U302 to work harder, overheat and burn out. With the exception of the large 450V cap, a complete recap (I counted 10 caps) might be in order.

[Mad Capper]

Re: Problem with Gateway 900G

Thanks jetadm123 for the quick response. Wish I could test that fast.

I did what you suggested, or at least my best attempt.

first 3 tests of each set I had to change the range to 2000 as it showed out of range when set at 200

U301
S1-G1 - 979
G1-D1 - 940 and slowly climbs
S1-D1 - 900 and climbs quickly past range
S2-G2 - fluctuates around 10
G2-D2 - fluctuates around 5
S2-D2 - fluctuates around 10

U302
S1-G1 - 984
G1-D1 - 1000 and slowly climbs
S1-D1 - 500 and climbs quickly past range
S2-G2 - fluctuates around 10
G2-D2 - fluctuates around 5
S2-D2 - fluctuates around 9-11

Are these appropriate values? I guess they mean that there are not shorts. Or should they be tested out of circuit?

If these tests show that the two IC's are good then what would the next step be? Replace the C301 and C302 first then see what happen and if nothing to a total recap as you suggested? Could the problem be the U303(BIT3193G)?

What are your thoughts as to where to go from here.

Thanks again.

[retiredcaps]

Re: Problem with Gateway 900G

first 3 tests of each set I had to change the range to 2000 as it showed out of range when set at 200

U301
S2-G2 - fluctuates around 10
G2-D2 - fluctuates around 5
S2-D2 - fluctuates around 10

U302
S2-G2 - fluctuates around 10
G2-D2 - fluctuates around 5
S2-D2 - fluctuates around 9-11

Are these appropriate values? I guess they mean that there are not shorts.

A reading of 5 to 10 ohms for both chips above indicate they are shorted (i.e. bad). Desolder both and repeat the same test for only the above pin combinations on the 200 ohm scale.

We only care about reading less than 200 ohms when testing for a short which is why the recommendation for setting your dial to 200 ohms. If the measurement is above 200 ohms, we don't care what it actually is. So switching to 2,000 ohms doesn't give us any relevant information for this test.

[Mad Capper]

Re: Problem with Gateway 900G

I wish I could say my desoldering skills aren't what they used to be, but the truth is closer to that they never were.

Thus Pin 1, i.e. S2, on both IC's broke off as did pin 2 on one of the IC's. I think the Pin 1's breaking off was caused somewhat by the IC being burnt to begin with as the corner pretty much crumbled. So I jammed one of the legs back in for the broken pin 2 and scraped away some of black part to get to a piece of metal for the pin 1's. I don't know how that would affect the measurements but here they are. I need two new IC's no matter what now, obviously, but I hope that the readings show they were bad anyway.

Additionally, I checked these out at different times and with different meters and again with the same meters and got different results. Due to the fact that I do not know how these work or what they should read I am rather confused as to why I get different readings, again perhaps because of the breakages. Or I am just doing it all wrong.

U301
S2-G2 - initial 4.5
G2-D2 - initial 5
subsequent tests for the above show readings varying widely and jumping around sometimes between 30-150 and slowly lowers at each jump., if there is any pattern readings seem to be around 30-60 many times

S2-D2 - 1-2 pretty much consistent


U302
S2-G2 - initial 10.3
G2-D2 - initial 11
subsequent tests for the above show readings varying widely and jumping around sometimes between 30-150 and slowly lowers at each jump., if there is any pattern readings seem to be around 30-60 many times


S2-D2 - 1.5-2 pretty much consistent

As for testing the CCFL's I do have another LCD monitor, not the same brand/model. Can one plug in the output of the good monitor to the CCFL's of the panel I am working on without adversely affecting either or is this unadvisable?

Again my thanks for helping me through this.

[jetadm123]

Re: Problem with Gateway 900G

Replacing U301 and U302 seems like a good idea. There's one us seller selling these chips on Ebay.

Yes, you can use your other monitor for testing the ccfl's. However, if your other monitor is a 15" or 17", it may not supply enough current to light your 19" ccfls. Also, if possible, try testing them in pairs.

[retiredcaps]

Re: Problem with Gateway 900G

Additionally, I checked these out at different times and with different meters and again with the same meters and got different results.

Post a picture of all your multimeters. I'll tell you which one is the most likely reliable and accurate one.

[Mad Capper]

Re: Problem with Gateway 900G

Plugged panel into another 19" LCD and CCFL's light right up.

So two IC's should get it functioning again right? I guess replacing the C301 and C302 for the sake of the new IC's would be advisable.

As for replacing the other capacitors is it necessary or just good preventative maintenance when it comes to these models?

As I look at the board I still wonder about C325, the one pointed out with the red arrow, I don't know whether to worry about it affecting anything.

The meters I have are a cheap Centech 90899 and an older Fluke 8020 B. I figure the Fluke is the better one but IIRC both gave varying readings.

Incidentally, what do find the best way to fasten a lifted pad back down?

[retiredcaps]

Re: Problem with Gateway 900G

As for replacing the other capacitors is it necessary or just good preventative maintenance when it comes to these models?

If you have an ESR meter, you can verify if the caps are good or not.

If those green caps are Samxon GF caps, they are a problematic series and will fail sooner than later.

As I look at the board I still wonder about C325, the one pointed out with the red arrow, I don't know whether to worry about it affecting anything.

What red arrow?

The meters I have are a cheap Centech 90899 and an older Fluke 8020 B. I figure the Fluke is the better one but IIRC both gave varying readings.

A 30 year Fluke is likely to more accurate.

Incidentally, what do find the best way to fasten a lifted pad back down?

https://www.badcaps.net/forum/showpo...8&postcount=12

If too much work, you can run a small jumper wire to the next component on the same circuit path.

[Mad Capper]

Re: Problem with Gateway 900G

The red arrow in my original attached picture, in the lower central portion of the back side of the board.

Unfortunately, yes they are the Samxon GF's and I do not have and ESR meter.

Thanks for the links for the videos. I'll have to see what I can do to fix my mistake.

Yeah, I figured the Fluke was better, maybe the bad IC was the reason for the varying readings.

I guess the next thing to do is order some parts and wait, I'm patient.

[retiredcaps]

Re: Problem with Gateway 900G

The red arrow in my original attached picture, in the lower central portion of the back side of the board.

I looked three times and still don't see it? If it is there, it must be very tiny red arrow.

[selldoor]

Re: Problem with Gateway 900G

Is this a poll - if so I see no arrow. I see the cap and the smut but no arrow.
Ceramic caps do not leak I was going to suggest trying to clean with isopropyl
alcohol (gently as they can become brittle).

[retiredcaps]

Re: Problem with Gateway 900G

Is this a poll - if so I see no arrow.

Thanks for confirming my eyesight is still fairly decent!

[Mad Capper]

Re: Problem with Gateway 900G

My apologies I had unknowingly uploaded the wrong image. Sorry for that.


Ok, if the ceramics don't leak then that's good I have one less thing to worry about.

When I find the time to order the parts and get back to this I will post the, hopefully, successful end to the story.


Again thanks all around for the help I appreciate it.

[Mad Capper]

Re: Problem with Gateway 900G

Here is the Update.

I order and installed two new 4606 MOSFET and there is no change. Still powers up and can see LCD but its as if CCFL's are bad. As I mentioned I plugged the LCD panel into another LCD power supply and they light up nice and bright.

I did not do a recap but after nothing worked I had two 220uf 35v caps lying around that I put in for the two C301 and C302 220uf 25v caps. The only change that I could see is that right after turning on the seems to flicker slightly for a second. It does not flicker from complete black to back on but rather like a brownout.

So I am at a loss as to what to do next or should I cut my losses.

Just as I did to check the CCFL's would plugging this power supply into my other monitor rule out a problem with the logic board if it lights up the other monitors CCFL's?

Could it be the transformers? If so, how could I test them to see if they are the problem?

Is it possible the inverter controller causing the problem? How would I confirm this?

Is it possible that all the capacitors need to be replaced for it to actually work?

Again thanks for all the help so far and anything info that will get this thing working is greatly appreciated.

[retiredcaps]

Re: Problem with Gateway 900G

I order and installed two new 4606 MOSFET and there is no change.

1) Did you measure the new mosfets to see if they are shorted before installing them?

2) Did you measure the new mosfets to see if they are shorted after soldering them in and before applying power?

3) Did you measure the new mosfets to see if they are shorted after powering up the lcd?

[Mad Capper]

Re: Problem with Gateway 900G

1) Did you measure the new mosfets to see if they are shorted before installing them?

No, I figured being brand new they were fine. I order more than I needed JIC and tested one that I did not install and it does not have any shorts. Therefore I would have a high degree of confidence the ones I did install were good also.

2) Did you measure the new mosfets to see if they are shorted after soldering them in and before applying power?

No, again did not see the need as they are new.

3) Did you measure the new mosfets to see if they are shorted after powering up the lcd?

Yes, they are not shorted.

I tested for shorts using the procedure

1) look at http://www.alldatasheet.com/datashee...MD/AO4606.html for the datasheet.

2) with power off and your meter set to 200ohms resistance, place your probes across the following:

S1-G1
G1-D1
S1-D1
S2-G2
G2-D2
S2-D2

No audible noise for continuity from meter or change in ohms from what meter reads when probes are not touching anything.

As they do not appear shorted I am not sure what to check next.

[retiredcaps]

Re: Problem with Gateway 900G

Is it possible the inverter controller causing the problem? How would I confirm this?

U303 is likely your inverter pwm controller. Find its datasheet. Measure the voltages on each pin and compare it to the expected values on the datasheet. The most important ones are probably VCC and EN (enable).

[selldoor]

Re: Problem with Gateway 900G

Hi - Continue as advised by retiredcaps but "I had two 220uf 35v caps lying around"
has me a little worried - what make are they and how old. This is a pressure point on
this board so you need to give it your best shot.
It would in any case be better to recap (except the BIG Cap)with new caps -panasonic FC or FMs.
I see that you have read some of the other threads.
In this one:
https://www.badcaps.net/forum/showth...highlight=700g
a few people make reference to the small 47uf cap which was often missed and was the cause or part of the cause of a few problems.
There was also a transistor burnt out on the video board so perhaps pics of that would be helpfull

[Mad Capper]

Re: Problem with Gateway 900G

After looking for a BIT3193G datasheet all I could find was this
http://html.alldatasheet.com/html-pd...IT3193-SO.html

I am hoping that is close enough otherwise I am not able to find the exact datasheet.

When checking voltages for the inverter controller should I check on the PSU alone, connected to the logic board, or everything connected together? I assume everything connected together and powered on so that is the readings I am posting

As per the pinout on the datasheet this is what I get:

1 1.307
2 4.51
3 0.7
4 1.41
5 3.96
6 3.85
7 0
8 0 (when turned on goes to 0.87 then to 0)
9 0 (when turned on goes to 0.49 then to 0)
10 7.77
11 1.77
12 1.81
13 1.28
14 0
15 0.39(when turned on goes to 1.89 then drops to 0.39)
16 0

I only watched for voltage change during the power on process for pins 8, 9, and 15 as they seemed from my interpretation of the datasheet(which is probably way off) and what retirecaps said as they should important. The others were take when it was already powered on.


Having read some I understand that good caps make a big difference, but after it did not work after the new IC's the ones lying around were all I had so I thought I would give it a shot. They might be just as good/bad as the ones on there but thought I would give it try to see it that made a difference.


I have done subsequent testing in various iterations and am perhaps more confused.

I replaced the 47Uf 25v with a 220Uf 25v and then a 100uf 50v, again that I happened to have, and this time when I turned it on the screen was brighter, flickered/brownout more noticably for a second or 2 then went dark. Does this point toward a complete recap as the solution?

Selldoor - It seems the link you referenced refers to a transistor on the PSU not video board

Post 21
https://www.badcaps.net/forum/showth...highlight=700g

First point: The shunt regulator is part of the regulator circuit. Check the output voltages of the power supply. The connector is labeled. If you have 5 volts at the 5V pin and 12 volts at the 12V pin (if there is one), the shunt regulator is not damaged.

As I have 5v and 12v would this point to the transistor working properly? But I don't know if when turned on the clicking and slight 5v fluctuation has to do with the 47uf cap or the transistor.

I did some more reading and saw that some one posted the way the CCFL's are supposed to be plugged in. I marked mine before removing and had a different order. So I tried what was posted and got the two seconds to black rather than the flickering. However, in changing the 47uf with others that I had and changing the CCFL order my results varied. I would get the flickering, two seconds to black or the clicking. At first there did not seem to be a particular setup that I tried that consistently produced one symptom vs the other, i.e. while keeping one thing constant and changing the other I would get different results even from one time I turned it on to another.

So I put back in all the original caps, and now it seems to consistently get the clicking. In time with the clicking the CCFL fire and shut off. 5V goes between 4.99 and 5.03.

So I tried other combinations of the components I had and it seems to still get the clicking.

So if the results I got for the inverter controller show it to be ok or make sense for the symptoms does a recap point the way to a working monitor? Should I retest the inverter controller again now that the main symptom is the clicking?


Thanks again