The failureWell, I carried this card in my pocket without a protection cover. It was probably a bad idea. Bad the card sometimes reported an error on write from the early beginning. Later, this card worked in one of ten attempts on my Zaurus, but worked on the desktop reader. Finally, it stopped to work on a desktop reader as well. that s why I decided to look inside.
I supposed that the error is caused by the mechanical failure somewhere nearly the SD controller. that is why I decided to attempt to reflow the chip.
The card contains chip marked SK6626ACSC S0810 H28171, the SDHC card driver, two memory chips Samsung K9HCG08UIM, and few passive parts.

Having
no reflow oven, I tried to use gas cooker hotplate. Using a cheap
Chinese infrared thermometer from Ebay, I fired up a manually regulated
heat and tuned the hotplate to 150°C. I did not want to reflow the card
completely, that is why I used a gas soldering iron (Portasol Pro Piezo
with a hot air tip). It was pretty easy to remove the chip, much easier
than using only the soldering iron.
But things went bad. After removal of the chip, four soldering pads on the PCB were detached and lost. I probably used too much heat, but it is possible that some of the pads broke even before desoldering. Maybe the chip was not correctly soldered. Who knows.
Never mind, I decided to continue in reflow. I put a bit of Aoyue Soldering Paste W002, and soldered again with a hotplate at 150°C plus soldering iron hot air. reflow went well.
I tried to fix broken pads using liquid silver conductor ELCHEMCo EL-1.
As you may expect, my attempt to fix it failed and the card did not work.
So I decided to reach terminal stage of the teardown and learn the reflow. Now I used higher temperature. When the hotplate reached about 240°C and the PCB about 220°C, all chips were released and easily removed with tweezers. Now all soldering pads were saved.
I did not fix the card. But I got a lesson: Reflow is possible. Heat dosage is far the most critical operation of the reflow repair works. Keep the temperature as low as possible, even at cost of longer rework time and longer exposition of parts to that heat.
|
|