Samsung Service Tech 2nd visit 12/21, and my results.
I haven't had time to update this post until now, so I thought I would catch everyone up. While the Samsung Technician was at my home 2011-11-30 for the capacitor procedure, he replaced ONLY the four 1000uF with new capacitors and reinstalled all the other caps I had removed and tested. That's right, instead of installing new capacitors, he put back the old ones. Before he replaced the 1000uFs, he had me sign an electronic form stating I would be responsible for all further repairs (parts & labor) related to this problem if the capacitors do not fix it. Hmmm, that was strange, but I had to sign it. Not a problem for me since I could get the main board for ~$130 online with a 6 month warranty from 2 or 3 online stores when it was in stock (...popular item). What about the poor sap that had to pay $500.00 for the Samsung repair!
The tech left and I sat with my broken TV and my schematic determined to troubleshoot the cause. Over the next week I solder tacked 19 small wires on key signals and voltages. Armed with a multimeter, oscilloscope, and a 16 channel logic analyzer I zeroed in on one IC that controlled the power sequencing and reset. It was apparent that the heart of the main board is a small micro controller (MICOM) WT61P6S combined with a 2K AT24C02 Atmel serial EEPROM which controls the following:
+ power-on/off
+ monitors voltage levels
+ IR remote and perimeter touch buttons
+ communicates with two main processors (A1F & A1B) over a simple UART
+ Issues reset signals to the main processors and other system.
+ Audio control
+ HDMI Identifier
What I discovered was that all the voltages were all within normal limits including those on the main processors. The DDR memory on the two processors was active and appeared to be functioning fine and all clock signals were stable and active. What I found strange was that the watchdog timer in the MICOM was timing out every 29.4 seconds, turn off for 5 seconds, and then turn back on for another 29.4 seconds. On the oscilloscope, you could see the SW_POWER (turns on & off TV) signal pin on the MICOM (PD1) go Hi-Z briefly (mid-rail voltage) and then get configured as an output to drive the signal at full voltage for another 29.4 seconds. All I had to do was plug the TV in and it started cycling on and off automatically.
I measured and /or disabled all other signals from the MICOM one by one to isolate the cause. The last suspect was the MICOM code. What I determined was that the firmware in the AT24C02 seriel EEPROM was somehow corrupted or bad data written to the EEPROM caused the MICOM to enter an undefined state which eventually caused it to time out (no watchdog kick) and reset. When I ground the clock signal to the EEPROM, the TV will not start at all, but during the 29 seconds, I see plenty of data between the EEPROM and MICOM. Again, all other circuits appeared and measured normal. No overheating ICs or bad or missing clock signals.
In any case, I was now convinced that this was a software problem caused by my son powering off the TV off while the Blu-ray was playing which may have caused the AnyNet+ feature to write (or not write) incorrect data to non-volatile memory. When my wife turned the TV back on an hour later, it would play the Samsung startup melody, would not power up fully (black screen w/ backlight), and began cycling.....over and over.
While I was waiting for one of the three on-line vendors (ie. shoppjimmy.com, rivervalleyelectronics.net) to get the main board back in stock, I surfed around on Samsung's own web site and stumbled upon a very important tip in the "Troubleshooting Guide" for my specific TV titled:
"LN52A750R1F TV Turns Off and On Continuously"
In it they described my issue exactly and recommended a firmware update that was released October 2008 to resolve the issue. I purchased my TV August 2008 (built in June '08). The last paragraph said "If your TV is turning on and off continuously and you are unable to install the firmware then your TV WILL REQUIRE SERVICE to resolve the issue". Basically Samsung is acknowledging that there is a software problem causing the On/Off cycling which needs the new firmware to correct. If you can't power up the TV, you can't load the firmware through the USB port on the TV. This to me amounts to a Samsung-admitted RECALL to change the main board free of charge!
On December 7th 2011, I picked up the phone a called Samsung customer service and explained that the service tech came out 11/30 and replaced the capacitors, and also mentioned the Troubleshoot Guide firmware issue I found on the Samsung web site. He responded by insisting that the web site I was looking at was a third party site, not the Samsung official site. He then reminded me that I had signed the form accepting all financial responsibility for repairs other than the capacitors for this power on/off cycling problem. I offered to send him the web link and even offered to guide him to it so he could verify it himself. He flat out refused to let me show him and stood firm. I'm now PO'ed so I got forceful and said I didn't care what I signed and that Samsung basically admitted the firmware problem so they should fix my TV for free as a recall for this obvious design flaw. It was dead quiet for a few seconds, then I said "you obviously aren't going to help so I need to talk to someone who can - your manager" He put me on hold for 5 minutes and then came back and said Samsung would activate a warranty window (parts and labor) to fix my TV free of charge.
December 21st, 2011, the Samsung repair tech replaced the main board like I expected and the TV worked fine. I removed all the wires on the old board and he took it with him. No doubt, Samsung will probably slap it in a test fixture and directly reprogram the 2k EEPROM using the factory test points on the PCB and it will work fine again....for someone else's TV...with the same issue! I'm just glad I got it fixed for free, but I had to fall back to my 22" TV for almost a month. Surprisingly, the firmware in the new main board is newer (higher # = 1014.0, 2009-05-22) than the downloadable upgrade from Samsung's web site (October 2008 release 1012.2) so you know they still needed to tweak for other issues...
Now we have both hardware and software contributing to defective stuff. Consider if all electronics have software running them, what percentage of failures are just bad software code instead of an actual hardware failures. If it's any conciliation, Samsung settled its class action law suit on February 23rd 2012. It seems they agreed to fix all TVs purchased before December 31st 2008 and give up to $300 to each TV owner experiencing power cycling problem. Hey, where is my $300 bucks...for pain (dealing with Samsung customer service), suffering (no TV), and compensation for my troubleshooting abilities!
Conclusion:
If you have an LN52A750 (or 40, 46) with a manufacturers date in 2008, turn off the Anynet+ feature IMMEDIATELY or risk ending up in my situation. You will just have to change the input source with the remote instead of automatic source switching. My profession and abilities armed me with facts to build a case for a free repair. Most of you would be stuck with a $130 to $500 repair bill for a TV you bought 3 years ago for $2200!
Good Luck!
Was this reply helpful? (0) (0)
Staff pick