I plugged in a molex cable the wrong way and the power supply exploded. Did it kill my hard drive?

Question

I bought a cheap kit allowing me to use a 3.5 SATA hard drive with a laptop, by plugging it in with USB. It still requires external power supply so I used an adapter that came with this kit but I plugged it the wrong way - the molex cable has four wires - one yellow, two black and one red and they are labeled 1, 2, 3, 4.

The power supply exploded (the enclosure is fine but it sparked) when I plugged it in, probably because it was plugged in the wrong way. Could this kill my hard drive? There wasn't any smell coming from the hard drive but I'm not sure that it wasn't fried.

When I shake the power supply I hear a rattling sound (similar to what you hear when you shake a burned out light bulb) - is it a sign that the power supply exploded first and it saved my hard drive?

I found conflicting answers on the Internet: this link says that it shouldn't kill the hard drive but this said it should.

As I have no alternative power supply I can't check it right now (if there is another way to check if it's fine please post it).

Solution 1

Simply plugging the cable in backwards would not cause what you observed (although it isn't exactly good for the drive).

It should be noted that hard drive circuity has safety resistors on it to protect the logic board should there be an over-volt condition on the power connector. This is standard practice on most low-voltage electronic equipment. Even though the connectors are designed to not allow a backwards connection, the circuits themselves are still designed to handle the situation if someone somehow manages to do it. This is why the ground pins are in the center -- they're still ground no matter what way the connector goes in.

The hard drive is probably OK, but of course there are no guarantees here whatsover.

---

Here is what likely happened:

The SATA connector was either not crimped properly to the wires, or the components in the brick itself were not soldered correctly. Either way, that's a short circuit, and it's quite possible that not even a drop of power actually made it to the hard drive.

A standard PC AMP (Molex) power connector uses 18AWG wires, which can carry up to 10A of current each. SATA power connectors use 28AWG wires, which can carry only 1.5A of current each. Because hard drives can pull up to 6A of power on the 5V rail (which exceeds the current-carrying capacity of any single 28-gauge wire), each 18 gauge wire on the Molex connector is crimped to three of the 28 gauge wires on the SATA connector so that the current load is spread out over them. In electrical work, this is known as ganging.

Here is a pin diagram of a SATA conector. Note the matching Molex wire color in the right-hand column: (Also note that Molex connectors do not provide a 3V line).

Here is a picture of an un-crimped SATA power connector:

As you can see, the 28-gauge pins are tiny, and they MUST be lined up perfectly. If you've ever crimped an RJ-45 connector to an Ethernet cable, you probably understand how easy it is to accidentally crimp two wires down to the same pin if you don't line them up properly when you push the connector on.

This isn't a big deal with Ethernet, as even with PoE, you're only talking about 500mA of power there. But with SATA, it's literally 10 times that. Accidentally crimping a hot wire to ground creates a short circuit which is a very real fire hazard at those current loads.

I drew up this ugly diagram to show what would happen if the SATA crimp were crooked by just 1 pin.

All that power flows straight back to the brick over one ground wire with no resistance. Pop, sizzle, and fry. It would take a very well-designed, high quality power brick to handle that error safely. This is why you should never trust your electronic equipment with no-name Chinese garbage power supplies. This goes for PC's, phone chargers, and the like.

An improperly designed/built power supply can ruin your electronics at best, and start a fire at worst.

Solution 2

There's no real way to tell if the hard drive's fine but this shouldn't happen.

Molex cables are keyed - though I suppose if it was truly dodgy the end connecting to the external power supply might be odd. Even then, typically what goes into the device is 'pure' DC.

As for power supplies, you might have a fuse or a blown out component like a voltage regulator but this is before power goes into the circuit board that adjusts the voltages for the hard drive.

So, no promises. From your description, the PSU is probably toast, and hacking it open to see if its a fuse (if there's one) or another component is a diagnostic step you should try. If nothing else, its not particularly trustworthy.

As for the drive, I'd do a visual inspection, sudden death of its enclosure may cause data loss and you MIGHT have burnt something out by doing the impossible and plugging things in the wrong way. There's no way to check short of trying it in a known good enclosure or system.

Solution 3

In these situations it is more likely to kill the hard drive than it is to kill the power supply.

However there is no way for us to know and the only way for you to find out for sure is to test it with another supply.

Note that it is generally quite hard to plug a Molex in the wrong way round without significant force. It's potentially possible you had a faulty power supply and plugged it in the right way round.

Solution 4

Sometimes the damage is done just until where the short-circuit occurred or the magic smoke was release.

So it was power adapter -> enclosure -> SATA HDD.

If you inverted the cable, it means that where it was expected +5 V, you injected +12 V. Yes, it's more than enough to fry some chips, since they have strict limits, and +12V should have gone to the HDD motor.

If the enclosure didn't had any chips or protection, i.e., the 12V didn't make something blow earlier (inside the enclosure), it might have fried some chips from our drive.

But it doesn't mean that everything is lost. I mean, your drive won't work right know.

But if there was sensitive data you want to recover from it, you can always try to exchange the controller board from it from another one, of the same manufacturer, same model, etc. Because the chips control many things, but the plates inside the HDD are probably intact, you might have lucky recovering some (or all) data from the disk.

You know that Murphy plays some rule here: if you were trying to destroy your drive so that nobody would recover anything from it, just plugging an inverted cable isn't enough to prevent that. OTOH, in the exact moment you want to recover information, Murphy comes and makes all information irrecoverable even if you replace the controller board.

Solution 5

This happened to me too! I bought a cheap Chinese ATA to USB device. The device did work, but it was designed with the ATA connector upside-down, so that when the button and indicator lights on the device were facing upwards, the hard drive was label-side down with the delicate circuits exposed, so I preferred to instead keep the hard drive upright and the device upside-down. At one point while using it, I forgot to insert the molex power connector into the device likewise upside-down, and although the connector is notched to theoretically prevent misconnection, the cheap plastic had enough give that it quite easily went into the socket.

There was a loud bang. A wisp of smoke blew out of the hard drive, and there was a burning smell. Both the ATA to USB adapter and the hard drive were dead.

The only saving grace is that the USB cable wasn't plugged into my laptop at the time.

View more solutions

Comments

[Thalys]

Comments are not for extended discussion; this conversation has been moved to chat.

[DavidPostill]

Can you provide a photo of this cable? There have been suggestions that it is not actually a molex cable.

[Spehro Pefhany]

One thing I discovered, to my chagrin, is that the PSU end of modular power supplies is not standardized. Sad!

[JMac]

@Spehro That might be on purpose. I doubt it would be a problem; but you could run into issues if people started mixing and matching between PSU's. I doubt the rating of cable coming out of PSUs varies a ton; but it could be dangerous if one manufacturer chose to cut it close knowing they had low power draw and then someone uses those in a higher power PSU and starts a fire.

[Spehro Pefhany]

@JMac They plug in and destroy disk drives instantly. I hope that isn't deliberate.

[JMac]

@SpehroPefhany So wait, different manufactures have different PSU sides on their modular PSUs, and those can fit into hard drives? I just don't get what you mean. I assumed you meant that the cables for modular PSU's all have different PSU sided ends so you can't just buy the cables and use them on any modular PSU.

[Spehro Pefhany]

Both ends fit perfectly, they just happen to go to different voltages at the PSU end, depending on the supply manufacturer!

[DrZoo]

That's what my guess was. It was some extremely cheap Chinese made adapter that has no quality control.

[Yorik]

"Hacking it open [...] is a diagnostic step you should try." PSUs and transformers have capacitors which have it will kill you power even when unplugged and off for hours/days. The OP might become "poorly from too much electric"

[FreeMan]

@JMac - I believe he means the end that goes into the PSU. Those are often 6- or 8-pin connectors and many, many different cables will fit. If you don't get the correct one, wired for your PSU, you'll blow whatever you plug into the other end as it receives 12v where it was expecting a data signal.I've got a Corsair PSU, and having been warned of this issue, took a look at the cables I bought, then contacted Corsair to confirm the one I was buying was correct. It would only work with 3 of their PSU lines, others were wired differently.

[Thalys]

@yorik pretty sure most power supplies I've opened have been mostly safe.

[plugwash]

PSUs are supposed to have discharge resistors on the capacitors to render them safe after input power is disconnected. I have had a shock when one failed though.

[TOOGAM]

+1 for the weasel words, which help accuracy. In my experience talking with people who dropped phones in water, about 50% of them never recover, and over half of the remaining have shortened life spans. But why the difference? I figured it is probably because when electricity goes where it isn't designed to, results can be rather random and unpredictable. e.g., does an electrical arc get made, or not? If stuff gets damaged, what component gets damaged first? This is typically uncontrolled/undesigned, so results are not very predictable. Luck may be a significant factor of the results.

[Thalys]

This answer hasn't gotten the love it deserves. It clearly explains what might have happened. and even has beautifully ugly freehand arrows. Though, its worth remembering most gear comes with some flavour of chinese power supply, with various levels of no-name or fakenameness, and garbagivity.

[CausingUnderflowsEverywhere]

Have you ever successfully replaced a circuit board? I'm pretty sure the control board keeps track of where data is written. See article: research.ibm.com/research/gmr/basics.html "After consulting the table of stored data locations in the drive's electronics, the actuator moves the head over the track where the chosen data is located." By minimum the circuit board will keep track of where bad sectors are.

[Nick Matteo]

@JMac: Cables from one PSU will fit into a different PSU. Everything looks fine, but it will fry whatever components you plug in to it. It's happened to me.

[AStopher]

I agree that it's a decent answer, but IMHO it's too generalised and uses improper terms ('brick' is 'PSU', for example).

[AStopher]

This answer is based in assumption and personal experience, and not fact. From what the OP described I'd say that the PSU blew its fuse and probably fried the adapter, but the drive might just be fine.

[barlop]

why does this question talk of molex cable and yet never mention PATA drives, the question only mentions SATA drives. So why the talk of molex?

[Thierry]

@barlop : sata drive might have molex power supply. Seach images for sata drive with molex, you'll even see disks with both molex & sata power supply.

[Thalys]

Those would be early model/transitional ones.

[Chris H]

@JMac Dell used to (~15 years ago last time I did it) use the standard AT motherboard connector, with a nonstandard pinout. Plugging a standard PSU into a Dell motherboard shorted almost all of the lines together. Sometimes it let the magic smoke out, even quite impressively, but some PSUs had sufficient output protection to work again.

[woliveirajr]

Yes, I have... Last time was about 8 years ago. Same model, capacity, etc. Think I could recover all I wanted (no, at the time I didn't try to recover the whole disk, just 40% of its size was the data I was after).

[user1686]

As for Ethernet connectors, it's been a while since I had to, but the ones I've used all seem to have plastic separators between every pin – self-aligning.

[user]

@JourneymanGeek It's easy to find SAS breakout cables that use Molex connectors for power and SFF-8482 on the HDD end.

[Keith M]

Sometimes there is a small panel on the back of the PSU which gives access to the fuse for checking and replacing it. Should check if you have that on yours :)

[ggll]

There are exceptions to this rule: forums.hexus.net/chassis-mods/…

[Tomachi]

In my case, the cable was inserted "diagonally" to achieve the result.