Chapter I: The First Encounter §

Back in May of 2015, I was going to do a data operation of some sort (I don't remember exactly what, probably secure erasure) on my SK Hynix SH910 128GB SSD I purchased for $85 in late March of 2013. I connected it to my Gateway GT5422E, and, for some odd reason, it'd shut off immediately after I turned it on. From experience, I knew this meant a short circuit was happening, and, eventually, I traced it down to the SSD. The SSD would spark if it touched the case, so my workaround to this was simply not having it touch the case.

About two months later, I used this SSD in my ThinkPad, as a hard drive proved less responsive than I'd like when running Windows 8.

Later, the previous issue I had encountered came to bite me. Whenever I pressed down on the keyboard in a certain area (about 3/4 from the left, around K), the ThinkPad would immediately shut off. Here we go again. My workaround to this was insulating the SSD from the rest of the computer with electrical tape. Nothing bad ever happened again...

...until one Saturday night, the tenth of October, 2015. I was using my ThinkPad, and it was doing fine.

When I got home, though, I couldn't get my ThinkPad to boot anymore; it would shut off immediately upon turning it on. Since I was already familiar with this problem before, I knew straight where to look. Sure enough, the ThinkPad would power up just fine without the SSD, but, when I would plug the SSD in, it would not boot anymore. I kept attempting to improve the insulation, but no dice. Finally, I took the SSD to my desktop computer. I have a handy little SATA extension cable that I use for plugging in drives on a whim, so, I plugged the SSD into it.

I turned the desktop computer on, and it booted swiftly as usual, and everything seemed fine so far.

Something strange happened though; at the Windows login screen, it hung at "Welcome". I knew that, sometimes, this process would be a
bit slow, so I waited a bit, but then I looked at my computer.

My SATA extension cable was letting out billows of smoke!!

I swiftly flipped the power switch on the power supply (good thing it has one), and I unplugged the SSD. The SSD was warm, and I
inspected the SATA extension cable. It looked fine, but I looked at it closer, and this is what I saw:

The SATA extension cable, with melted insulation on the yellow +12V wire.

I also found something else when I was looking closer at this cable to take a picture of it:

One of the two connectors on the SATA extension cable, with melted plastic covering a pin.

Well, I took apart the SSD, and I saw one half of the blue PCB, which had the memory chips on it. Everything looked absolutely fine, and unharmed. I flipped it over, and I saw the controller chip. It looked like it may have been somewhat damaged, but I went ahead and I plugged it into the Gatway GT5422E, but since I knew that computer was capable of stopping the same drive from shorting out. I did not plug in any SATA cable.

Armed at the power plug, I turned the computer on, and I saw the green light of the SSD turn on, which told me that the SSD's controller chip was probably okay. I had turned the computer off quick. though, as I didn't want anything bad to happen.

I retrieved a SATA cable, and I tried the drive out; it was recognized in the BIOS! This tells me that the drive is very likely to be functioning perfectly fine, and that my data was probably okay.

Since I knew the SSD wasn't a fire hazard with its casing off, I knew my next course of action; back up the data. I plugged the SSD into my Acer Aspire M5630, my main desktop from 2008, and fired it up. Sure enough, the Acer also detected the drive, just as the Gateway did. I booted Windows Vista from the Acer's old, original 500GB Western Digital drive, and I initiated a backup with Acronis True Image WD
Edition.

Now, the casing of the SSD was acting as a heatsink, and, since I don't yet dare put the PCB back in its case, I put a fan near the SSD to keep it cool. Maybe it wasn't going to be quite as effective, but we need to keep the danger down, here.

The SSD's blue circuit board, with a small red fan blowing at it.

The backup's done, and the Acer computer is off. The data is safe! My next task is to make the SSD usable again, or pitch it in the
garbage. Stay tuned for what I'll do next!

Well, I went ahead and transferred my backed-up data to a Kingston SSD I have, then secure-erased the Hynix. I discovered that the
short-circuiting issue seems to happen only sometimes. It seems to depend on the case's screw tightness. I managed to assemble the SSD so
that the short-circuiting would happen again, and I will be looking into RMAing the drive, next. After all, the drive acted up without me
tampering with it, so, if I have to tamper with it to get all this done, that's fine.


Later, I arranged contact with NewEgg, and they naturally said to contact the manufacturer, SK Hynix. NewEgg also said to get back with them if they won't help. SK Hynix's website in itself was already posing a problem, and even after somehow getting their broken file selection for uploading a receipt to work, sure enough, they weren't helpful.

Time passed, and the drive sat there.

Chapter II: The Unexpected Gift §

On December 22nd of 2015, finally contact was made back with NewEgg about the situation, and they said that they'd give credit for any "comparable" 128GB SSD from them! I picked the first choice in the list of the best-reputation 128GB SSDs, and that was the Samsung 850 PRO 128GB, for what was otherwise $89. Thank you, NewEgg!

Now, here's the next mission: since I still get to keep the smoking SSD, it's time to see what I can do to get it working again. Clearly, the case is shorting out the SSD, but where, and how? It's time to find the conductivity in the case!

The SSD's casing laid out, and a digital multimeter showing no conductivity between two points on the casing.
Nothing here.

The multimeter, showing conductivity at the screw holes.
Nothing's going on here.

As before, with two other points.
Two other points once more.
Aha! The screw holes are doing it. The screw holes also do touch the circuit board, which basically tells us everything we need to know.

One of the circuit board's corner holes, showing a slightly discolored area around it.
Here's something that looks a bit interesting. I'm unsure as to what this is or why it is there, but this is one of the eight points that the casing touches the circuit board.
As above, but another hole, showing less intense discoloration.
This other corner isn't as bad.

The circuit board, now with electrical tape covering all of the holes.
Now I've insulated the PCB from the case entirely. This ought to fix the issue for good!

I've got it all closed up. Up next, the test!

The SSD, showing as detected in a Dell OptiPlex GX620's BIOS.
Yay, it works! This doesn't guarantee against failure, but it's a good sign. Hopfeully, it will never smoke cables again!

Chapter III: Revenge of The Smoking SSD §

On October 15th of 2016, I was upgrading my Dell Latitude D630, and something scary happened not long after. I turned on the laptop, and it started smoking!

Upon discovering that the laptop's pointing devices were not operational, I started to disassemble the laptop, and I found that the palmrest connector (for the touchpad and Bluetooth module) was melted! I had to push melted plastic bits out of the way and unbend pins on the motherboard's connector, and I had to steal "sorta-borrow" the palmrest cable of my other D630, but I got it all working again.

The fried palmrest connector.

It took me a while to figure it out, but I realized that the reason why the palmrest connector fried, of all things, was most probably because, as I know, The Smoking SSD has +5V applied to the metal casing, and so any conductor touching the SSD's unpainted metal will be given the same charge. The palmrest's metal likely functions as a ground, and The Smoking SSD must have been touching the grounded metal, resulting in a short-circuit that fried the weakest link, that being the connector.

After this second scary incident, I moved the SSD into my ThinkPad, as my ThinkPad handles SATA short-circuits better.

Render Time: 1.83 PPS