Dell Inspiron 5150 Random Shutdowns
After researching known problems, I bought a two-year-old out-of-warranty Dell Inspiron 5150 on eBay with the random shutdown problem. Based on eBay bids for units with the random shutdown problem, I believe many people think they can cheaply fix a defective unit. Their bids reflect no price reduction for a possible repair failure and little price reduction for the time and effort involved trying to fix the system board.
There are two approaches. One involves disassembling and removing the system board prior to attempting the repair. The other approach prepares the unit solely by removing the plastic protective sheet in the cavity where the modem card and/or wireless card fits.
It is well documented that the LVC14A IC under one of the side "C" cover tabs is repeatedly physically stressed by the tab resulting in fracture of the solder connection. The cited cure is either replace the system board or "refloat" the solder connections of the LVC14A pins to the system board.
My effort involved three full disassemblies and system board removals. I became more efficient after the first experience. The first attempt resulted in causing the unit to no longer attempt a boot. The flashing LED indicated a defective system board. The second attempt got me back to random shutdowns. The third attempt fixed the random shutdowns. While the experience was rewarding, it involved several hours of effort eating up any savings. Purchasing a replacement system board on eBay would have made the total cost of the laptop higher than the going auction price of a working Inspiron 5150.
Soldering surface mount IC's requires skill and proper tools or a large dose of luck. If the soldering iron is too hot, the IC can be destroyed or the solder pad on the system board separated from the PCB. If the iron is too cool, you might as well be poking things with a pencil. If the point it too big or misplaced, you may create a solder bridge between the pins. Refloating all 14 pins on the IC may create problems with the good solder connections in existence. You will be working on an IC that is about 1/4" long with two rows of seven pins on two sides. These pins are separated by a little less than the thickness of a razor blade.
Here is what I learned from my experience and details of my "overkill" approach that worked. First I tried to melt a couple of mounds of solder nearby, on the opposite side of the system board from where capacitors had been placed. There was a large amount of solder meaning it would take longer to melt than the solder affixing an IC pin to the PCB. The temperature needed was 500 degrees Fahrenheit. My solder iron permits setting a precise temperature and continually maintains and confirms the set temperature.
The first effort bridged two pins with solder, even though I did not add any solder. I had "refloated" each pin's solder connection for 3 seconds at 500 degrees Fahrenheit. I cut the bridge during the second effort. During the third effort, I spend more analytical time to determine how many pins had a fractured solder connection and specifically identified which pin(s) needed repair.
I purchased a AAA battery holder and soldered two leads to the holder. I soldered a sewing needle to the positive lead and the ground was attached to my instrument. I used an oscilloscope but a needle based voltmeter or needle based multimeter would have worked. A digital multimeter with an oscilloscope function would also have worked. A digital multimeter that takes some time to display a reading would not work for finding an intermittent fault.
I theorized the pin, actually a leg similar to a human leg in the seated position from the thigh to the foot , had been pressed against the PCB until the solder fractured. I expected the foot would have moved outbound from the IC as the only "give" available. The effect would be a lifting of the foot from the PCB. This theory would explain why my initial refloating of the solder failed. Sure I may have melted the solder but when I removed the solder iron, the foot just lifted out of the still molten solder.
I used the sewing needle and a magnifying glass for precise sequential placement on the PCB just beyond each foot while holding the oscilloscope probe where the thigh entered the IC. Only one pin exhibited an intermittent failure. The pin was located directly under the inboard side tab of cover "C". This time I changed my solder strategy to heating 4 seconds and holding the foot to PCB with the sewing needle. I then retested the electrical connect that confirmed success. When the computer was reassembled, it worked without random shutdowns.
Now that I know which pin was at fault and specifically how to effect the repair, I would not need to disassemble the computer to effect a good repair. There are risks to the plastic using a solder iron in a tight situation. There are also risks disassembling a system. I also filed 1/8" of the tab on the "C" cover to avoid additional physical stress on the IC.
I paid $386.50 plus $35 shipping or $421.50 for a low end Inspiron 5150. It had 2.66 MHz P4, 256 MB RAM, 30 MB hard drive, DVD only, and no WiFi card. I knew of the intermittent shutdown problem, but I didn't know of the bent-in keyboard that also pushed in the RF shield below. [Perhaps an effect of a closed fist blow, in frustration.] I bought two RAM strips of 512 MB each and a WiFi card. The CPU speed difference is not so important to me. I would like to have the a burner version of the DVD but I have a networked unit with that feature. The computer came with the original CD's for reinstalling the operating system, drivers, and applications. It also came with an original COA for XP Home Edition, a power supply, and a main battery. I prefer XP Professional which I use on another machine.
You can tell my purchase was a compromise to reduce price because I was uncertain I could fix the machine. My fall back was to just resell it as defective. With inbound shipping, auction cost, and a small amount for a poor auction, I expected the experience to cost no more than $100 and a few hours time should I fail. Some eBayers do not realistically price out failure and that causes bids that are too high for defective goods.
As more people attempt a fix, my speculation is that the quality of defective 5150's offered at auction may include botched repair attempts. That means the market should lower the price of defective goods. A false sense of how easy the repair can be will tend to drive the price of defective goods higher. My situation worked out. I feel extremely fortunate.
Remember this article documents my experience and frustrations. I take no responsibility for the ill advised activities of others.