INSIGNIA LCD HDTV help

[Blueteeth]

Not trying to turn this thread into one about tin-whiskers But yeah, seems NASA has blamed the recent toyota failures on tin-whiskers too. I can't see it being a major problem since almost all consumer, military, and medical equipment these days uses lead-free, and we haven't had 'catastrophic' failures everywhere - just a few more than usual. My guess is, that it's a rare combination of lead-free (basically pure tin) along with manufacturing faults which would otherwise pass QC, where joints aren't well formed/rounded, leading to areas where tin crystals can grow out and cause shorts. I seem to remember a paper on this written over a decade ago, where they pionted out that better quality control was required if the switch was made to lead-free. Seems like this hasn't be adhered to in the industry. I dont' blame lead-free, I blame the lack of companies ability to treat it as different from working with food 'ol 60/40.

 

[electronicsfreak]

Believe it or not, I'm posting here again! (unfortunately....)

Anyways, while I do believe tin whiskers could still be valid, I really think just board warping caused the solder under this BGA chip to crack. Here's why.

The same sort of failure described below had occurred again, good proof that something was still fishy (I believe this was caused by merely shuffling the set around as it awaited for a proper mount/stand to be constructed) So myself not really wanting to have to do this same re-flow process every few weeks or so, and having seen a few guides/videos of a successful re-ball of a BGA without any special equipment I thought I'd try to do so myself to replace the brittle solder underneath (heheh, seemed like a good idea at the time, but I guess that's what they all say)

In the end, I was not able to get it to successfully re-bond every connection, but to be honest I can't actually be sure of this. I had re-"heat" soldered this BGA multiple times, and out of every try, I had two main results follow.
1) Either nothing happened, same unresponsive board as before
2) Immediately after the power switch was depressed, the back light fired, and the entire LCD panel would cycle through Red, Blue, Green, White, Black at roughly 2seconds per "color".

My thoughts on 2) an error response of some form, but Insignia wasn't willing to tell me what it was. So, with a failed attempt and board traces beginning to separate from my "tries" of heat applied, I opted in for a replacement. Problem solved right? I wish.


I ordered a cross referenced board 6HA0136910-12V-LH32, and triple checked with a technician over the phone to make sure this is an exact replacement. This "Drop in" replacement gave an interesting response. Different problem, but still.

Upon power-up, the set did turn on, but something seemed different. All of the User interface Menu displays were similar, but slightly off color. When an input was given (regardless of source, but primarily composite and VGA were tested), the color was off drastically, and in many areas, particularly shading, pixels would dance between one color and another as though the board was trying to guess what color it should be.

One board warranty replacement and many, many phone-calls later I believe this fault could be a few items

1) Wrong board was sent/ by some lucky coincidence, the two sent to me had the exact same fault

2) The board is correct and faultless, but needs to be calibrated through a service menu, access information Insignia was not willing to divulge

3) The board is correct and faultless, but the timing board has developed a problem throughout my experimentation flaws (interface between the LCD panel and Logic controller)

I've primarily been searching for information on 2), but not much of use has come up.

So, for those who've read this far, any thoughts or suggestions? I have already checked the LVDS cable with my multimeter, and it is making a solid connection.

Also, A photo of SMPTE color bars as an input will be added tomorrow to visually show what I'm talking about.

 

[unclejed613]

iirc, most Insignias are made with Samsung parts. could it be a wrong version main pcb?

 

[electronicsfreak]

It could be possible, though on three separate occasions I tried to confirm that this board was the correct replacement, as a technician over the phone asked me about specific IC locations, wether or not a component was on a labeled pad, and any version stickers.

As promised, attached is that photo I was talking about, hopefully describing the new problem I am trying to diagnose.

Through some of my research, I have not found a leaked service manual yet, however I have found a code that will access a hidden menu. Using the remote, entering the 4 digit sequence:

->
2 5 8 0

in any setting but TV will access a menu with a list of options including AutoColor, White Balance, and Service Menu. Tried a few of these with no success, and when selecting "Service Menu" I believe either another code/some form of equipment may be required as the selection simply makes the hidden menu dissapear

 

[electronicsfreak]

Well, this past weekend I came across a tad bit of success in my search. Googling that code mentioned above with NS-LCD gave me a promising link:
https://www.electro-tech-online.com/custompdfs/2012/02/LCD20TV20NS-LCD37_SM.pdf
This link is the service manual for the NS-LCD37, but aside from some programmable output settings, it's practically the same main board as from this NS-LCD32 set.
Now, to be honest, it really didn't help me out a whole lot (atleast at this point in this project), but rather gave me an idea to try. In the manual, it describes calibrating the set using a few particular signals from a video signal generator/calibrator .

So, I thought maybe if I try to "create" my own signal generator using the Ladder 16 grayscale described, and the VGA out of my computer, I could try to get something "close". This proved to be just that though: close. The attached image is the set running with the described image sent to it. Running through the auto color configuration did clean up the output a tad, but I can't get it to be exact, hence the result of my poor man's video signal generator.

I'll call a few local repair shops, describe to them exactly the routine I need completed, and pray something I can afford can be negotiated. We'll see though. Ah, if only proper equipment was accessible......

 

[Blueteeth]

Might just seem like words of encouragement, but when I was younger (maybe 13/14) I was surprised at how helpful the folks at repair shops were. They would often let me do the work using their equipment, and giving my pointers here and there - as long as I did the work because they were too busy. Probably realised they couldn't make any money form me, but I learned a hell of a lot from a few hours in a TV repair shop. I'm sure they'll give out some free advice to narrow it down, and might even let you borrow some test equipment.

 

[electronicsfreak]

Well, a bit more of an update with.......success? Well, kinda. I guess more of a step forward than anything else.

So, I had spoken with a few local shops. None I had spoken were willing to, nor had the tools necessary to complete the calibration I was asking for. However after mentioning Tri-state Module (company I had purchased a replacement from), a couple of shops had mentioned that they had ceased doing business with Tri-State due to faulty components being sent when replacements were purchased......This sounded a bit fishy to me. So, after building a cleaner image and one last try at my "calibration" attempt, I called them, told them my situation again, and set up a third warranty replacement. This time, I was to send in my old board as well so one of their tech's could do a comparison.

Well, 3 weeks later, the "new" board came in the mail. This time a board with a differently labeled refurbished part sticker # was sent (6HA0136911, rather that the previous two with 6HA0136910) I'm not sure if this really was much of a difference, but after hooking everything up and testing one of the composite inputs was a success, I re-assembled the unit.

I was initially pretty happy with the results, the composite video channels and VGA input worked, without a problem, and navigating the menu's didn't yield any distorted picture as the other two boards did.

However, I can't get any of the HDMI channel's to display an image. When feeding a signal, the set will recognize it and respond with the resolution and frequency of the signal sent, but the screen will remain dark. The component inputs are a bit finicky as well, though the set will only recognize there is anything present when a signal is applied or removed.

I'm not really sure what to think right now. This project was a bit more enjoyable when it involved actual circuit analysis (and staring at the schematics is pointing to a few key IC's to start searching around), but right now I'm pretty frustrated with this company. Personally I would expect a purchased component to work properly from the start, but after three boards and still no dice, I'm not really sure what to do.

 

[Blueteeth]

Sounds like you've hit a major snag. Debugging a 'faulty' board, only to be sent one with little confidence in its viability (= more debugging from scratch, something you shouldn't have to do).

If you have the original board back form them, you might be able to probe the outputs/pins/test points on the original, and compare to the 'new' one, see what gives. Could well be they upgraded the firmware that required a minor hardware change on one of the sub boards (like a logic line high/low). Soundsl ike its starting to venture into '16-channel logic analyser' terroritory


I'm afraid I was out of my depth fairly early on in your debugging/repair, now I don't think I can do anything. I would end up just tracing the entire tihng out again, which is far too much work :/