I have a weird problem with an EIZO FlexScan SX3031W lcd monitor.
It shows vertical lines of garbled/distortion pixels around various parts of the displayed image.
Opened to check insides and nothing looks broken (no visible capacitors problem).
Please see attached images. Anyone know what can be the problem ?
+ Reply to Thread
Results 1 to 10 of 10
Any or a combination of the following can be the problem, possibly repairable:
- ungrounded connections somewhere, possibly because of a loose screw or cold solder joint
- cold solder joints elsewhere on the PCB in signal path
- the power supply is not shielded enough & is producing enough RF to distort video circuits (just because capacitors look good does not mean they are; they should be considered suspect if >5 years old)
- the CCFL inverter circuit is producing enough RF interference because some parts are defective
- another nearby PC or monitor with a badly shielded sympathetic power supply is producing enough RF interference
- the connector between video & LCD drive boards are not making good contact
I consider the Achilles' heel of LCD monitors to be the flat cable(s) between LCD driver board and LCD itself. Because it's not possible to solder or crimp connectors on the transparent leads on LCD itself, conductive glue is machine-applied to adhere and perfectly align the many thousands of connectors. Giving the LCD monitor a few sharp knocks or letting it fall even just a few inches onto a hard surface like concrete may dislodge some of these minute connections. This is, alas, not repairable.For the nth time, with the possible exception of certain Intel processors, I don't have/ever owned anything whose name starts with "i".
Thank you turk690 for the useful hints, you are very helpful.
I'll look further into it considering your advice and come back with details.
Looks a bit like poor quality/unshielded cable...
Opened it, checked the insides and because there are many big photos, I've uploaded them here.
All capacitors inside are Nichicon, around 5 years old, and look good. There is a set of 4 capacitors in the PSU that can be sensed warm to the hand, but they are rated 105'C.
Reconnected all visible cables and flex cables that can be seen in the photos.
I can try to disconnect the CCFL inverter circuit and check with a light if artifacts are still present, but as long as it does not make noise I would not suspect it.
May be a hint: The cooling fan starts after a few seconds from power on, don't know if it should come in so early.
I found someone with a similar problem that was solved by reconnecting all cables.
Video of the problem added: http://youtu.be/o2o9cVHDTSQ
Last edited by bassfreak; 28th Jun 2014 at 08:50. Reason: Added video
I can see this monitor has internal connectors galore. Unplug and plug them in, maybe 3x each, very carefully and see.For the nth time, with the possible exception of certain Intel processors, I don't have/ever owned anything whose name starts with "i".
Last edited by bassfreak; 28th Jun 2014 at 14:54.
Lol... sorry, "galore" means many, a lot, huge numbers of, etc.
OK, a typical LCD monitor has a few distinct parts:
- input circuit, which conditions input signals and to protect and isolate monitor from unexpected levels and waveforms
- A-to-D converter for the VGA analogue signals
- scaler circuit, to attempt to resize input signal resolutions to the native mechanical one of the LCD
- control processor for receiving user inputs and coordinating interaction between the many sections
- LCD driver board
- switch-mode power supply (SMPS)
- cold-cathode fluorescent lamps (CCFL) and their associated driver circuits
Integrating parts onto fewer boards make for a more compact monitor but also increases the chance for interference from each other. For example, the SMPS, if unshielded, badly shielded, or has defective parts, can literally be an antenna spewing out its switching frequency and its harmonics and interfere with all of the other circuitry around, and with other electronic appliances nearby. This interference is mitigated by shielding, grounding, ample DC filtration; the control processor can track the actual horizontal scanning frequency in use at the moment and try to make sure the SMPS switching frequency is a multitude of it.
I consider the flat cables between the LCD element and the LCD driver board as the weakest mechanical link in an LCD monitor. In your monitors case this would be the ones labeled YP 0750 03850E & YP 0747 03850E. On the LCD element side, this is machine-attached with conductive glue to its transparent leads. On the driver board side, it may also be glued, but in your case there is a rotate-snap connector. Subjecting the monitor to abrupt force, like sharply knocking it or letting it fall even just a few inches onto a hard surface can dislodge connections on both ends; the result are usually missing lines or pixels on the picture, or permanently of one color, intermittently or otherwise. If this happens to the LCD side, there is no way it can be fixed; the monitor has to be junked, same with if it's also glued to the driver board side. Since in your case there are connectors on the driver board side, I will probably detached them and put them back. This involves grasping the edge of the brown clasp, easing it out a bit, rotating it up, and letting the flat cable slide out. Then I will re-insert the flat cable as far as it will go in, hold it in place, and let the brown clasp snap back in. This has to be done very very carefully & slowly with a steady hand; it's easy to break the connector locks if done heavy-handedly.
You may try that, but I consider the problem there unlikely. As I mentioned if the problem was there, I'd expect missing or permanently black or colored pixels/lines. Your problem is more of interference, or some load or impedance not being met. The next suspect is the input to the LCD driver board; on the board this is the connector on the opposite side of the ones mentioned above, and it's labeled CN5 (as best as I can discern the photo). Connect and disconnect this about 3x; do the same to the end which goes to the input/scaler board. This connection is significant because it has the same form as HDMI (4 differential pairs plus control lines), such that in some current monitors it is the direct input connection, and input, scaling and driver parts are now on just one board. Impedances have to be met and appropriately terminated; if they are not, ringing and trailing will occur, like what's happening with yours. If one of the differential pairs is cut, the picture may pulsate, appear even worse, or even mute all together.
The lions share of power in a monitor like yours is taken by the CCFL and its driver circuitry. It is also the hottest, and and has the highest voltages (~1Kv). When a monitor ages or was subjected to shock, parts on the CCFL driver board are slowly cold-jointed, or get cracked or broken. This is true of the CCFL driver transformer which has a brittle ferrite core. When this transformer gets cracked it heats up and can actually become a fire hazard. So one of the first things I inspect when opening up an LCD monitor is this part; I resolder the transformer pins on PCB, as well as the actual CCFL connectors (assuming the CCFL is good). This is why LED backlighting has dramatically reduced the power consumption of current LCD monitors, as well as the complexity and chances of failure (for the first time there is no high-voltage present). LEDs do have their own problems (mainly involving color-rendering) but that's another story.
Interference problems can originate even from outside the monitor, even if no video is applied yet (but more so if video is applied). Some computers have defective or poorly-designed power supplies that spew out interference like there was no tomorrow. This is true of cheap, made-in-china units, where, to cut costs to the bone, they purposely omit the AC line filter components (filter chokes, capacitors, etc) and just put a short in its place. The power supply primary side will introduce switching frequency interference into the AC line, and any and all other electronic appliances plugged on the same may be affected. This is critical if the offending power supply switching frequency or its harmonics is just a few Hz up or down those in your monitor. Therefore have you tried to plug that monitor into another AC outlet of a different branch in your house? Have you tried another computer?
This is all for now; tell me how it goes should you do something further...
Last edited by turk690; 28th Jun 2014 at 19:59.For the nth time, with the possible exception of certain Intel processors, I don't have/ever owned anything whose name starts with "i".
Thank you turk690 for the comprehensive contribution.
Found out that the TCON board is the same as in the Samsung 305T which is more popular, and guess what, there are a lot of folks with this/very similar simptoms.
So the problem narrows down to the ALTERA chip which requires reballing. Many solved it by oven reflow, but I will try a professional reballing service for this expensive item.
I'll come back with results after this operation.
Very good news!
Professionaly reballed today and operation was successful. Briefly tested without input signal, the OSD looks clear, no artifacts, no left side vertical lines artifacts.
A knowledge stone added and a remarkable display saved from going to trash bin.
The chip is very hard to reball, because of the attached heat spreader. The balls just don't melt to the pads on the dark infrared BGA rework station and had to be reworked three times with the hot air gun.
All your advices were very helpful in diagnosing the problem.
Thanks for your time and for sharing the knowledge in helping this out!