Saturday, April 23, 2011

Compaq C500 - LED's Flash Rapidly, and never posts or starts

Unit arrives with following symptoms:
Flashing LED's on the power panel, and the power LED on the front edge flashes continuously. [See video]

When I connected the DC source with the battery inserted, I thought the charge light would come on as it should regularly. Since it didn't and flashed at the same rate as the other LED's this told me what area to check. When checking the MOSFETs PQ5, PQ49, and PQ4 all of them were of the same type [FDS4435] but PQ49 had a short in both directions with meter set to DIODE mode.  This indicates the FET is shorted. So that was the first thing to replace and test again. 

When I replaced the PQ49 FET and plugged in the charger (with battery inserted) the charge light came on, as it normally should.  Upon powering the unit, it posted normally on external monitor. 

Since it was a power in/charge circuit I assumed everything else was fine and reassembled the unit.  All items work. USB/Wifi and other peripherals.  So the unit was tested with OS loaded and everything fine there, and allowed the battery to fully charge. 

Friday, April 22, 2011

Gateway laptop M-Series - Shuts off immediately [M-6750]

This Gateway M- Series  6750 arrived with the following symptoms: 

Turns on, but shuts off after about 5 sec.  The lights would flash, then shut down. Counting seven (7) flashes before shutting off. Attempting a cold reset of the BIOS/CMOS it would not complete the reset, and continue to do the same thing. 

Because it would not write the CMOS, that shows me the first area to check.  Under close inspection of the board, I notice a coil is burnt, and the number is not visible/readable. [PL20]

The burnt coil had also burned the motherboard layer under itself and would require cleaning and insulating the second layer. This was done with a small piece of mylar tape that was placed under the coil before soldering it in place. This repair was a quick repair and not making a complete mask and copper pad for the exit contact of the coil. Works and sufficiently insulates the second layer of the groundplane.

MOSFET PQ54 was also replaced simply because the cost is negligible and would be easier to replace while the unit was open rather than to reassemble and possibly have to replace anyway. The FET was  a Fairchild FDS6900AS  which is a 30V Dual N-Channel PowerTrench® SyncFET™

After replacing the coil, I partially reassembled the laptop simply enough to get power and connect the screen.  (Image blurred while holding camera)  It Works!  Reassemble and test rest of peripherals, USB, Webcam, etc...

Compaq C500 - Bad Video Memory

Compaq C500 Arrives with bad video. 

The information is displayed just not properly. External LCD also displays same thing. This is related to the video memory of the unit and not the GPU in this case. 

This will require a low temp reflow. Because video memory is often not glued to the mother board, low heat has to be used so that the chips become separated from the motherboard. If that happens it will require special equipment and higher cost to the customer.

If you unit does not have dedicated memory for video, then it would be the chip that addresses (provides the row and column information for the GPU) the video memory.   This is usually a large chip directly to the side or near the GPU. Typically on Intel models rather than AMD. 

Middle picture shown after reflow.

 Link Video showing screen issue at Youtube 
if you have issues displaying from blog.

Tuesday, April 19, 2011

Acer 17.3" Shuts off when moved or touched

Newer Acer 17" Model [7736Z-4809] laptop arrives after being repaired.  

The initial repair was a DC jack. Seems that the jack didn't last long and was sold to a reseller. Physically broken and replaced.  

After replacement, the unit would charge and run.  Minor tests completed based upon user diagnosis.  However, if the laptop was moved when on/running, it would shut off immediately, and the lights would go off. (Charge and Run lights).

 When the unit was moved or tapped at the area on the right where logo stickers are, it would shut off -everytime. Thinking it may have been a ground issue with the case, I tapped the motherboard after it was removed, and still presented the same problem.  Repeated on/off testing located the area where the open circuit was occuring.  The PCI controller chip (indicated in square).  Since this model did not use high temp glue to hold the chip to the motherboard, the best reflow method for this model would be Low Temp reflow with liquid flux instead of the petroleum based flux.

Why? Because if the chip is having issues with the solderballs too much heat would separate the chip from the board, requiring reball or motherboard replacement. A High Temp reflow would cause that problem. And there would be no way to rescue it at this shop. So that is why low-temp reflow method was used. 

A low temp reflow involves liquid flux, and heating only half the chip at one time. First the lower half, then allowing it to cool a little, then the top half.  This method assures that the chip does not become unseated, as well as preventing solder balls pooling together and shorting out. The liquid flux aids in bonding the solderballs to the motherboard as well as removing oxidation (not much at low temp).

Saturday, April 16, 2011

Dell 1545 - No Power & Shorting Power Supply

Dell Inspiron 1545 arrives with No Power, No Lights and short Power Supply.

This unit arrived with because the customer thought it was a DC Jack repair because it was shorting out the power supply everytime it was plugged in. So it did not matter which power supply was used it would still short out. 

My way was to simply separate the power panel board from the laptop and see if it was the laptop having a short or the small power board. When the small board was removed, the power supply would still short out.  Looking into the jack itself, and testing from POS to NEG shows no short on meter. So tracing the path from the POS terminal on the back through the on board diode showed continuity.  

Reversing the meter's leads it shows 1 Ohm. So this tells me that the diode was shorting the unit. 

Extracting the power board only took 3 screws, and not having to remove the whole motherboard made things simpler and a quick repair. Then lifting the power board and replacing the diode.
Boots after replacing diode

After power board diode was replaced. To test, I simply plugged in the power supply before putting it back into the base. This would show me if the power supply was still shorting.  It did not short again, so I re-installed the power board and turned on the unit.  It works!  

Starting the OS everything appears to be OK, except for the message that Windows did not shut down correctly the last time.  This was to be expected, but everything works fine now.