INSIGNIA LCD HDTV help

[sheldonstv]

well i was very scared today as i come across a samsung set which had the same power supply fitted as the phillips set which gives trouble and the same faulty electrolytics with the wrong voltage rating had been fitted from new...also these were not hi temp types ...i think someone must have been asleep when they designed this particular unit....

 

[unclejed613]

i've seen some "crossover" manufacture in the last few years. it's hard to keep score. philips makes samsung soundbars, samsung makes philips dvd players, insignia makes pioneer receivers denon makes marantz receivers and harmon kardon amp boards. samsung makes insignia tvs and LG makes insignia dvd players and DTV boxes.

a list of brands of caps to be avoided is available at Badcaps.net - Badcaps Home and they also have a list of good cap manufacturers

 

[mittens]

if the backlight isn't working the unit won't be allowed to power on.

i apologize in advance for what is surely a stupid question--but i really want to know the answer and these things aren't so easy to google:

when you say the unit won't be allowed to power on, do you mean:

a) power button led won't respond at all,

or

b) power button led will come on and then quickly go out

or

c) power button led will be the only sign of life?

 

[johnbarko@msn.com]

Insignia LCD

10uf @ 50v cap in the power supply (with heatshrink on it).

 

[johnbarko@msn.com]

I wonder if he's getting the PWON 5vdc from the power supply?

 

[unclejed613]

one of the most common problems with the insignia power supplies (and probably samsung too....) is the diodes on the secondary side of the switching transformer. apparently there was a very large batch of fast recovery rectifiers that are either too slow for the switching frequency used, or just plain substandard. the diodes tend to overheat themselves to death, quite often in less than the warranty period. check the diodes........ replace any shorted ones with silicon carbide rectifiers (which have ZERO recovery time, and so run much cooler). they're a bit more expensive, but they won't fail again.

 

[johnbarko@msn.com]

UncleJed can you get schematics on the Insignia LCD tv's.
Let me refraze that, do you know where I can get them?

 

[electronicsfreak]

i apologize in advance for what is surely a stupid question--but i really want to know the answer and these things aren't so easy to google:

when you say the unit won't be allowed to power on, do you mean:

a) power button led won't respond at all,

or

b) power button led will come on and then quickly go out

or

c) power button led will be the only sign of life?

The LED is the only sign of life. On top of that, the unit will not respond to using the remote control to shut it off, almost as if the logic has been frozen or it is stuck trying to power on the florescent tubes.

I tested the power supply and it is outputting what it needs to on the 5v, 12v, and 24v lines.

 

[unclejed613]

i'll see what i can do....... at the very least, i can find out the name of the OEM company, and you might find a schematic listed under one of their model numbers. a lot of this stuff comes out with just a troubleshooting manual and bill of materials with the part numbers of the boards.... unfortunately that's the way things are in the industry today...... the company i work for sells some of these brands, and the equipment cost is so low, that we aren't authorized by the manufacturer to do much more than swap it for a new one. if it's out of warranty there aren't even boards, DVD mechs, etc... available for them. most items that sell for less than $200.00 are in this category. i'll look it up and see what i find.....

i've seen this happen with brand names like Boston Acoustic and Polk as well. there's a sound bar that Boston makes with a wireless subwoofer, and i've "repaired" a few of them (if you can call board swapping a repair..., but boards are all that's available for them), and i just saw a Polk Audio sound bar last week that internally is 100% identical to the Boston model, with the same wall wart power supply and metal rear panel. the only thing that's different is the styling of the plastic frame of the speaker bar and plastic subwoofer box. the wireless cards even operate on the same frequencies.

 

[electronicsfreak]

Yes I am completely aware that I am reviving an old post. However, seeing as I did start the thread, and have brought this project back to my table, I thought it would be best to return here.

Anyways, upon finding this old thing in my parents garage, I have been able to get it to turn on again, but just that one time.
I have since replaced the inverter board, but unfortunately this was not the culprit.
I have also tested the power supply again, and the 3.3v, 5v, 9v, 12v, and 24v supplies are all where they should be, though both the 12v and 24v lines are sitting about 0.5v lower than what they are listed as. Also, when testing the 24v supply with a known resistive load, the actual current was quite a bit lower than what was expected.

Any thoughts? I am running a little low on ideas.

Thank you guys in advance!

 

[electronicsfreak]

Kinda an update from what I've found today:
After scanning and following the inverter board for any possible sign of current, I had tested to find any AC current present on what would normally be feeding the two coils present using a voltmeter. Nothing was present, though the board was supplied its 24v rail. Also, testing a pin labeled "on/off" on the inverter board, when the TV set is first plugged in, sits around 0.6v. After hitting the Power button, it's brought down to 0v.

Any thoughts?

 

[Blueteeth]

Maybe detach the 'on/off' pin connection, but leave the other connections (power, feedback) intact. Pull this pin to ground with say a 2.2k resistor. When its powered up, when the inverter board gets its 24V. Try putting 3.3V - 5V , via a 470 ohm resistor on the 'on/off' pin, and see if the tubes light.

The only reason I brought resistors into this is to limit current on the on/off signal. Its most likely a digital input to the inverter controller chip, so it'll draw uA. But if the input pin has blown, you'll be connecting your 5V to 0.6V and make matters worse.

If its a new inverter board, then if it lights up, the controller board is 'duff', if it doesn't then its most likely the 24V rail - did you measure this whilst it was being powered?

 

[electronicsfreak]

Maybe detach the 'on/off' pin connection, but leave the other connections (power, feedback) intact. Pull this pin to ground with say a 2.2k resistor. When its powered up, when the inverter board gets its 24V. Try putting 3.3V - 5V , via a 470 ohm resistor on the 'on/off' pin, and see if the tubes light.

The only reason I brought resistors into this is to limit current on the on/off signal. Its most likely a digital input to the inverter controller chip, so it'll draw uA. But if the input pin has blown, you'll be connecting your 5V to 0.6V and make matters worse.

If its a new inverter board, then if it lights up, the controller board is 'duff', if it doesn't then its most likely the 24V rail - did you measure this whilst it was being powered?

Hey, thanks for the input man. I was beginning to think this thread wouldn't start up again.

--The current limiting resistors make sense, especially as the "on/off" pin appears to be a logic level input.

--I had tested the 24v rail of the PSU with the inverter disconnected using 2 car headlamps (of a known resistance) connected in series while the PSU was fired up and comunicating with the controller board to make sure the PSU was actually on when the set was "turned on". From what I remember, there was a small voltage drop from the static level that was present, and the current was a little bit lower that what was expected (yeah, I know actual values are helpful, and during this test I did write them down, but being the wonderfully organized student I am I don't actually know where that sheet of paper is..)


I plan to test what you've suggested this weekend, as well as re-measure the values described above of the PSU. Expect an update about this time next week.

 

[Blueteeth]

No problem man, I have a back log of 'broken crap' I've been meaning to repair, much of it is long term and eventually gets done bit by bit over a few months.

Although us engineers always like specific numbers, for the most part, a power supply will either put out its rated voltage, give or take, or nothing (dancing around between 0 - 1V) so if it kicks out 22-26V when powering the inverterboard or your dummy load (good call on that btw) I'm sure its fine.

Projects like this, are pretty rewarding, sometimes frustrating, but awesome when you actually repair something. If not, still got to fiddle around with electronics!

 

[electronicsfreak]

No problem man, I have a back log of 'broken crap' I've been meaning to repair, much of it is long term and eventually gets done bit by bit over a few months.

........

Projects like this, are pretty rewarding, sometimes frustrating, but awesome when you actually repair something. If not, still got to fiddle around with electronics!

heheh, this kinda actually sounds like myself as I too have a few other odds and ends waiting for their time to be repaired/scrapped. And yes, in the end, even if I don't fix this thing, starting this old project up again has been fun so far.


Welp, I've finally found the time to return to this project. I've since cut the lead running from the controller to the BL_ON/OFF pin on the inverter, and ran the inverter's input end to it's own little Logic level High/low voltage divider with a 470Ω and a 10kΩ resistor. When pulled high the backlight fired up . However, I can't get any image to be produced in the screen, even from trying to display the menu and such.

I've already begun searching for an affordable replacement controller board as a back up plan, but I don't want to resort to this quite yet.

Any thoughts or suggestions on where to start hunting on this board for the culprit, or shall I resort to a replacement board?

 

[Blueteeth]

Hi again,

Good to hear it powered up! So at least the 'power side' of hte backlight, as well as the tubes are kosher. Guess its all about the process of elimination.
In situations like this, I generally first scour ebay for an entire 'replacement board', as you probably did, just as back up. Its money, but it's a fall back plan in case you release the magic smoke out of your crrent board whilst testing. So again, good call on that!

Now, if the power supply *did* fail (perhaps not the power to the backlight inverter, but the logic board) there is a chance the logic board is fried. In my experience, its never 'everything' thats freid, but usually the chips/semiconductors, most sensitive to power/voltage surges. And thats where the trouble lies - if its the 'big chips' that have gone, replacing them will cost mroe than a new board.

Probably should point out here I have never repaired an LCD/plasma tv, but I've done my fair share of portable dvd players, laptops, netbooks, ipads etc.. (as the components get smaller, my eye sight has started to fail lol).

I usually start with a good 'ol multimeter, probing the board, and looking at it to find out where the power lines are. And checking those for shorts. Sometimes you get lucky - as with a recent 2.5" SATA harddisk I debugged - and the power lines are shorted because a TVS diode blew, and shorted across the power lines. Replacing that sorted it. But as its going to be a complicated board, it may be a case of getting as many part numbers of the devices as possible, googling for dataheets/pinouts and checking the power pins of each device. When semiconductors fail, be it catastrophically, or just a single IO pin, they close circuit (sorry if you know all this). So with one probe attached to the system ground, and using a continuity tester on your meter, which should bleep for <200 ohms, see if you can find areas/pins/devices which really shouldn't be shunted to ground by appear to be so. I usually start with all the power lines ( shouldn't be <100ohms to ground or power) and all the input/output connectors on the board. Knock on wood, the decoder chip on the board is fine, and its something minor, like an on board switching power supply.

One godsend in repairing, is, google. I realise since you've posted you've probably trawled the web for any information/similar problems, but its always great when you 'ping' a forum post where someone has the same problem, and a random guy replies saying 'check D405 for short as its a common problem'. Gives you more stuff to check

Now, what if the board is actually fine? Then it seems to be a case that, its not displaying anything because it's not getting the signals it expects fm other parts of the cicuit (like 'power good' signals, or IO flags from other bits of hardware to tell the board everything is fine) then you could look at what is attached to this board, what comes out of it (control signals, display lines, LED status outputs etc..) and what comes into it.

I've never had training in repairs, just far too many years 'taking crap apart' and fiddling, so take my psots with a pinch of salt I have destroyed a lot of stuff in my time.

Good Luck! And if you're stuck, I'll continue to throw idea's out, and let you decide if I'm jsut ranting or if it seems worthy of a check. Also a photo of PCB's, and the part number of the tv should get others on board. Aside from 'a picture telling a thousand words', PCB's are basically digital erotica for guys like us

 

[electronicsfreak]

Hey Blueteeth, thanks again man for the help! Seeing even a portion of this thing repeatibly live is very encouraging.

I'm pretty sure the TV's part number can be found in previous posts of years past, but for sanity's sake:

NS-LCD32

Manufacturer: Insignia

the board I'll be staring at this evening is: 782.26HA37-690E

And, attached are a few front/back photos of the board itself
I made a short video showing it's "life", but YouTube's up loader is being a bit finicky this evening. I'll try uploading it another time.

**broken link removed**
**broken link removed**

**broken link removed**
**broken link removed**

 

[Blueteeth]

Nice pics! Thats a beast of a board eh? Since I don't see a SMPS transformer and some BIG fet switches, I assume thats just the logic board and the power board is elsewhere. Apologies for missing the part number, I'll go through the thread before I post

You've already nicely ruled out the inverter board and tubes, and the power supply lines (12V, 24V, 5V...any others? mentioned in post #30) seems its narrowed down to that board. Although, with your 'resistance load' trick, might be prudent to check the other voltages from the PSU whilst current is being drawn. I'm unsure a good value to start with, because the 5v/12V lines may not be required to source over a couple of amps. Sometimes when a power supply is on its last legs, it'll kick out a good regulated voltage with little or no current draw, but soon buckle under a load as the output caps can't hold the output voltage. Drawing too much, sohuldn't blow anything (it'll be current limited) but will put it under stress. I'd say a good amp for each of the rails apart from 24V which you've tested

Aside from that, the power supply connector on the logic board (with no PSU attached of course), resistance checks from each power rail to ground. I'm sure there'll be a few ground connections, but you can see those from the PCB traces connected to it. Dodgy chips/power components on the board will bring the resistance down dramatically - example, an old acer mobo blew on me, the 5V line on a healthy board was ~4.6K to ground. On this one ~21 ohms. 'Resistance' per se isn't always the best way to check things, but does highlight shorts and open circuits rather well - exact values aren't critical.

The logic board will probably step down the 12v or 5V lines to 3.3V or lower for the logic using a synchronous buck converter (note the SMT inductors and big caps on top centre of that pic). Which may be a good place to check for blown MOSFETs/chips. When not powered, diodes and MOSFETs can be tested with continuity/diode check. when a MOSFET goes, it'll show very low resistance, either polarity, when its 'ok', and not powered up, should be fairly high, with a diode drop of 0.3-0.6 from drain to source. - as its in circuit, the value will vary, but '0.0v' is a definate sign that 'this has met its maker'. As with a very low ohm value.

Apologies if I seem patronising, telling you tihngs you already know, or if I'm goiong over your head. I'm never sure how to gauge someones experience, so I just risk 'sounding bossy' rather than providing vague advice that someone doesn't understand (thus the long posts). Good call on the youtube vid! If the PSU checks out, and its defeinately that logic board.

If you get stuck with things to try, post (or PM me if its a massive photo) a high resolution of the top of that logic board I'll highlight some devices for checking. Notice how I've avoided saying 'BAD CAPS!' even though its a common fault, they're just a bit of a ***** to check without an ESR meter so I always leave that check once all else has been elimiated (pure laziness on my part, not good practice).

Soo, things to try over a cup of coffee, armed with a meter - sounds like much of my evenings, how depressing...

Blueteeth.

 

[electronicsfreak]

Blueteeth, that post is gold. Thanks man! Also, I should apologize, as I also looked over previous posts, and I guess I never did post the TV's model #

Anyways,
here is that video I had taken earlier of the operational backlight: https://www.youtube.com/watch?v=bTysz8mRJ4Y&feature=relmfu

Unfortunately, being a full time student plus everyday life, It's beginning to be a bit difficult to find time to dive into this project, but I have made a little bit of progress.

I do not have an ESR meter, but I do have access to a scope/function generator. So, I have been looking over the caps in the main board today. I haven't found anything useful yet, aside from a few fishy surface mount electrolytics surrounding a couple of IC's meant to process audio signals. For now I'll order the replacements, but I don't actually think these are the main reason it won't produce an image on the LCD screen. Video of today: https://www.youtube.com/watch?v=-CTXZ5lsnjA&feature=channel_video_title


For now though I'm still slowly checking off a few things you've suggested, as I find the time.

Also, kinda a random update for anyone else following this thread, from some of the research I've done, I've found that this board:
Insignia DSP 782.26HA37-690E

Is almost identical to:

Insignia DSP 6HA0136910-12V-LH32

Vizio DSP 569HA0969E

 

[Blueteeth]

Yeah, since everything on the main baord is linked up, any one of them could be raising an error flag to cause the main CPU to 'shut down' before it properly powers up. I would be surprised the the audio part has blown, and that being the cause of the issue. That bit can short and draw too much current, but its doubtful it would effect the rest of the system. I'm still putting my money on the power going to the CPU/controller chip.

When you checked the voltage on the 'power on' connection from the logic to the inverter board, you mentioned it was either 0v, or ~0.6V. now to me that says the either the line is pulled low by the chip, with a pull up resistor - and the voltage that pull up is connected to is only 0.6V (one of the power supply lines is duff), or, the chip itself is only sourcing 0.6V - which coiuld mean the power to the chip is only 0.6V, OR, the chip has in fact blown, taking some of its pin-protection diodes with it. Now I could be wrong of course could just be the fact the chip flags an error, and doesn't fully 'enable' theinverter line, leaving it at ~0.6V.

I know schematic woudl be a godsend, and I'm sure you've searched, I drew a blank as well. One thing that oculd be prudent in terms of 'thigns to do' would be to google the part numbers of chips you find. MOSFET's, and diodes - being SMT - wil have part markings perhaps unreliated to the part number, making them very difficult to find the datasheet for, but usually IC's had the full part number. From the datasheet, you can see what should be connected to what, as they almost always have a 'test schematic' of how it sohuld be connected. Good way to check power pins, and see if any logic IO's have been shunted to V+/GND.

Great vid btw! Seems you're far more organised than myself when repairing things - but the croc clips, wires breadboard and multimeter is true 'techs' setup

I know how it is with 'finding time'. Sure, those who do this for a living can spend 8-hours a day, over several days on it, but as a 'hobby' thing in ones spare time, the odd hour here or there, projects can last months. I have a backlog of over 40 things 'to do' some of which started back in 2005, PCB's everywhere, fixing screws in plastic bags labelled 'ACER screws!' :/ But its cool to have stuff like this to go back to when you have time, or when you're bored, its also a hell of a lot less stressful than working on it constantly

Good luck!