Looking for help with a Viewsonic VP230mb where the backlight comes on only momentarily

[lostone]

Hello, I came here after finding some similar questions asked on other monitors and hope someone here can help me get an ancient Viewsonic VP230mp (23" 4:3 LCD) back to working condition.

First a little background, I have a background in computer and electrical engineering, and I'm smart enough to know I know nothing about LCD displays and how they work. For this topic I'm as green as they come.
When the display is turned on the viewsonic logo displays in the uppper left for 2 seconds then the screen goes black, the power LED is green (on) and when a computer is connected the display can track the systems monitor state. Meaning if the computer sleeps/wakes the monitor goes in/out of power saving mode. I've tried to shine a flashlight at the panel to see if htere is a a display or not, but given it's face down on the bench in a disassembled state this is not an easy task. I can't say there's an image but expect there is and this is an inverter/backlight problem. But without experience with backlights and LCD panels I have concerns about working on those. It appears there may be 4 backlight tubes and I can see light on all four edges when the display wakes up/powers on,

That said here's what I know....
The display is a Viewsonic VP230mb which has a HOAU04F power supply board and uses a Fujitsu TFT-LCD panel FLC59UXC8V-03 which has s HIU520A HPC-1411A inverter board with 6 outputs (CH2-CH7).
The power supply is a basic 12V 100W (simple) design with a feedback channel from the main board.
I have the service manual for the monitor and a spec sheet for the older sister panel: FLC59UXC8V-02 so some tech stuff is known, but more is not.


Suspect number one, the supply... so recap time and while the supply was disassembled and waiting for replacements I swapped in a DC supply power the video board (main board) and found the same results. Even though I found two bad caps on the supply, I was pretty sure the supply was not the problem.

I also recapped all 15 or so radial can electrolytic caps on the main board and also replaced 2 of the smd caps whose ESR values seemed bad.

With the supply good I started probing the board and here's what I know... well I guess I should say on the PS board the board was really clean and not even dusty but on the bottom side there's an LM385 which has something on the pins. Because of how clean the rest of the board is this was suspicious... but it could just flux.

the serial eprom was read and has data, but if that data is correct is a different story as I couldn't find anything to compare with on the web. I'm also expecting the video board flash chip and its contents are good. I could read the flash, but without anything to compare to that seems like a waste of time.

I have followed many of the service manual trouble shooting steps, but when the question is "is the back light on" and the answer no says: "check J7 and J5; is the inverter ok?" I'm not clear what to check or how to determine the answer to the question.

I have probed signals which I felt were more interesting... vsync, hsync and other video signals to the panel look like what I'd expect.
The video board jumper to the inverter has 4 signals... GND(4 lines), vdd (12V)(4 lines), GND (for VEE?) , VEE, and BRICON (4 lines all tied together).
BRICON is connected to a i2c controlled ADC that, when powered on receives an message to set the output to 0xF8 (basically just shy of 3.3V) which measured at 3.2V. and BRICON is 3.2V, then 200ms later i2c message to turn output "off" actually 0x0A matches the BRICON signal dropping to basically 0V. This 200ms pulse is noticed on the inverter board where all outputs start pulsing while BRICON is high, then when BRICON drops the inverter output on all 6 channels looks fairly stable over those 2 seconds before shutting off:



Knowing I know nothing about servicing the LCD panel, what should I do next? If there's something bad on the inverter board I could rework if parts are available. I don't have a second inverter board to try.

I've got lots of pictures and am happy to share but without known what would be of interest, I'll just start with pictures of some of the boards.
The four images:
5438 is the video board
5452 and 5453 are two shots of the inverter board topside
5453 - shows J7 which is the connection from video board to inverter showing the signal note BRICON source is Pin2 of U3 which is located just south of the electrolytic caps along that left edge in that picture

[lostone]

Little update because I'm looking to move this to the side for a while, I was able to confirm the screen does have an image!
I'm also sharing a video of the display being switched on to demo the dim backlight behavior. I hope I can share .mov file!

[lostone]

With a little more digging I found the data sheet for the -04 panel which is still not the -03 version but does provide additional information about the signals to the inverter board. Seems the -03 version has the most wires, but this looks promising!

I was correct with regard to two different grounds, VEE and BRICON would use the single line ground while the 12V "HV" lines are paired with the group of 4 and would explain why I had to connect the board directly to AC ground rather than just use DC ground. I had the BRICON signal backwards lower voltage is brighter, so the switch of the ADC from 0xF8 to 0x0A should put the monitor at near max brightness. I had observed VEE never went above 0V which indicates the BL should always be on. According to panelook.com there are 6 backlight tubes, which jives with the 6 HV outputs but adds confusion since all 6 outputs are different.

I had previously been avoiding an LED BL + inverter option because I didn't know if I'd find something compatible but now believe this may be a standard control protocol for the BL system. If anyone can confirm that could help.
I also discovered a site ccflwarehouse that may have replacement bulbs so unless someone suggest to steer clear of them I could reach out to them this week.

[CapLeaker]

I read all what you did and din't do to the LCD. I have to admit you did pretty good, but there are a couple of crucial things that were either missed or not written down. Symptoms on LCD like this aka “2 seconds to black” the most common reasons for it to do so are capacitors bad on the inverter board, the small little transformers on the inverter board, cold solder joints / connections or the blue 1kv high voltage caps at the ccfl connectors.
If the ccfls are displaying a pinkish hue, they are at the end of their life.

Check the little transformers on the inverter board and measure the resistance on the primary and secondary windings of each transformer and note them down. The transformers got to be within 3%. My bet is that at least one is out of spec. There is no real fix for this, unless you either find a replacement board / transformer or you disable the protection circuit on the inverter.

[lostone]

Check the little transformers on the inverter board and measure the resistance on the primary and secondary windings of each transformer and note them down. The transformers got to be within 3%. My bet is that at least one is out of spec. There is no real fix for this, unless you either find a replacement board / transformer or you disable the protection circuit on the inverter.

Thanks for the reply, yes without a schematic for the inverter board I was not comfortable given I also know nothing about those boards. 🙂 I would have to get some help finding replacements for those transformers who knows if there's a compatible part available today!
I didn't think of checking the transformer resistances, will do once I get it back on the bench.

So the caps like 25-33 would be the HV caps you called out? I could ESR check those, but given their voltage I'll just be able to check similar look similar.
Are the J parts that look like large resistors, just a bridge (0 Ohm resistor) between traces? I really wish I had some prior experience with those inverters, but other than knowing it's converting DC to AC and stepping it waaaay up I have no idea what any of those outputs should be at.

[CapLeaker]

Thanks for the reply, yes without a schematic for the inverter board I was not comfortable given I also know nothing about those boards. 🙂 I would have to get some help finding replacements for those transformers who knows if there's a compatible part available today!
I didn't think of checking the transformer resistances, will do once I get it back on the bench.

So the caps like 25-33 would be the HV caps you called out? I could ESR check those, but given their voltage I'll just be able to check similar look similar.
Are the J parts that look like large resistors, just a bridge (0 Ohm resistor) between traces? I really wish I had some prior experience with those inverters, but other than knowing it's converting DC to AC and stepping it waaaay up I have no idea what any of those outputs should be at.

You did all right, you just started wrong. First things I check on the CCFL inverter boards are: comparing transformer winding resistance, check the high voltage (1KV usually blue) capacitors, followed by ESR checking the electrolytic capacitors and any cold solder joints. Don't bother checking the ESR on any non electrolytic capacitors.
Don't worry... knowledge comes with experience. Been there done that and I also started at some point and learned when I didn't know chit. 10 years ago this forum was littered with backlights problems, now not so much.
Usually if you diag a bad transformer, yes 3% of difference is all you need on the transformers and the inverter board turns the backlights off after 2 seconds, it is game over most of the time because you can't source a replacement transformer nor a inverter board.

Fawk, now what?! If you really need to get this going, because it is a specialty monitor etc. convert it over to LED backlights. You could also hack a different inverter board in that has the same amount of CCFLs and wattage than the old one.

[lostone]

Thanks again, right now I'm toying with the idea of i2c injection to see at what point the display cuts off, I know it's on at min brightness and off at max, but where does it go south may help me. I have questions about just wiring in an embedded dev board to the monitor main board and while I think that's safe if running directly off the display's power I still got questions I hope someone can confirm my thinking.... we'll see what the collective says.

There are no electrolytic caps on the inverter, at least none I've seen. I shared a picture before, and I don't see anything other than the previously mentioned C25-32 caps as being tied directly to output. I wonder if I should start with the outputs that are dramatically lower than others.

[CapLeaker]

Ah! Back on track, eh!

[lostone]

Well this is promising... it's great to have a lot of components to probe on the board. I started with the HV transformers measuring the resistance between the input and output pins, what I noticed is the top side (CN2-CN4) had fairly high resistances (450kOhm) pin 1 and the others. Moving to the other side Cn5-CN7 were all almost 0. Odd so I checked the ESR on the big caps neraby and found, pictures are easier than words, so these are measured ESR values, no surprise the two on the right are equal because they are in parallel...




So I'm feeling confident things aren't quite right since both sides are show some very different readings.


Next moved on the the BJT transistors (Q6, Q7, Q8, and Q9) and found Q9 (pictured in the second ESR picture seemed to show shorts between B/E and B/C while the others tested fine.

Removing the C5707 and testing confirmed:




Looking at the back of the board I do see a slight brown discoloration on the green soldermask right where this part is mounted on the other side. If that's the result of my removal or some other overheating I'm not sure but I'm going with it wasn't my work. =)

I'll get a few of these with my next parts order and see what magic follows!

[lostone]

Update...
I got a new C5707 (a few actually in case I missed something and destroyed another one). What's interesting now, is I can see the discoloration on the board around Q9. After replacing I powered it up and same thing!! WTH?!?!

I started tracing the circuit back to the input and soon found two 63V 4A fuses, I had originally found a single 63V 8A fuse and found that was fine. But the 4A fuse in line with Q9 was blown. After waiting for mouser to week to fill a tiny order I got the new fuse installed and the backlight stayed lit!!

YEAAAAAAHHH!!!
Lots of learning along they way, and I never wanna do that again!

[CapLeaker]

Sweet! Another one kept from the heap!