Hacker News new | past | comments | ask | show | jobs | submit login
Ever Plugged a USB in Wrong? (2019) (npr.org)
42 points by Tomte 5 days ago | hide | past | web | favorite | 178 comments

Back in the early '90s, when 100 MB was a big hard drive, we were doing a project for a company that had access to some early, very hard to get, units of a 1 GB drive that IBM was going to be introducing later that year.

I somehow managed to fry one by plugging in the power connector upside down. This should be possible, as it used the 4-pin connector from the middle of the top row here [1]. That and its mate are not symmetrical and are made of a pretty stiff plastic.

When I told the company we were doing the project from that I had fried one of the pretty much irreplaceable test drives I expected them to be very upset.

Instead, their owner laughed and admitted that he had also fried two of them that way.

Even better, his contacts inside IBM told him that something like 30% of the drives were failing testing during manufacturing, and that had been traced to IBM testers plugging in the power connector upside down and frying the drive!

It turned out that for some unfathomable reason they were using a power socket on the drive that had a very flexible plastic hood. It turned out that the force needed to plug in the keyed connector backwards, bending the hood out of the way, was not much more than the normal force to plug the thing in the right way because the plastic was so flexible.

[1] http://www.playtool.com/pages/psuconnectors/connectors.html

It doesn't really address why one plugs it in wrong more than 50% of the time. John Siracusa had a great analysis on this on the ATP podcast a few months ago: Even if you have it in the correct orientation the first try, it inevitably catches a bit and you think you've done it wrong. You don't want to break anything, so you flip it to try the other way. That is now the wrong way and won't fit at all. So you flip it again and it goes in. So, 3 tries if you had it right the first time, 2 tries if you had it wrong the first time.

Naw, they're just 4-dimensional. So 540 degrees of rotation are needed to flip them 180 degrees in 3 spatial dimensions.

I thought it was because you had to rotate the usb plug through 4d space to get it into the correct orientation.

My (non-technical) friend solved this ten years ago: "holes up"

Go through your drawer of USB cables. You'll see that a lot of Type-A connectors now have the holes on both sides!

There are several ways to tell visually which way it goes. Holes up and the fat plastic part (look into the connector on the cable) down. On top of that there is often a USB logo on the 'top'. Finally you can remember. heheh. With a bit of effort I've managed to get my USB plugging in to be about 90% accurate but I'm still caught out by that factor they mention in the OP - when it catches, you doubt and reverse.

I go by the seam on the metal shroud. Seam down. Port is vertical? Well shit.

The article addresses the reason it was made non-reversible (cost) but does not address why the plug was made rectangular, which gives it no visual or tactile feedback for orientation. HDMI for example is non-reversible but it's obvious (more or less) which way to insert it.

The only reference in the article is,

> The Intel team led by Bhatt anticipated the user frustration and opted for a rectangular design and a 50-50 chance to plug it in correctly, versus a round connector with less room for error.

Which sounds like a post-hoc rationalisation for a bad design to me. I find it hard to believe they did this intentionally.

I really don't have any better luck with HDMI than USB. The connectors are usually at an awkward angle due to being on the GPU/back of a TV. Gotta say I enjoy the Lightning and USB-C connectors that have no "right-side" from a dumb users point of view the most.

In theory an asymmetric connector can have a "correct" design orientation, like "for horizontal insertions the wider end is always down toward the floor" and "for vertical insertions the wider end is toward the body of the appliance."

Not everyone will notice/remember, but some will, especially if it is logical - like 'the "heavier" end is down for more stability.'

Unfortunately it looks like the accepted orientation of HDMI connectors is a precarious "heavy side up."

Not quite the same but usb has the logo, the usb-logo on the cable goes up. A thumb drive often don't have the usb-logo but determining up is often easy enough.

Now "up" isn't always up, connectors on the side of a monitor for instance should often have the logo towards you since they are usually vertically mounted.

With a second of thought it mostly works as expected in my eyes. But there are exceptions.

HDMI is a lot easier: It's easier to see which way it goes, it's easier to feel which way it goes, and it's easier to know when it's being inserted backwards since it offers immediate mechanical feedback.

HDMI isn't so bad.

S-Video connectors were terrible. Worse than USB in every way possible, if you can imagine that.

For an example of closer vintage to USB-A, FireWire 400 had a great connector shape. You could tell what side you were looking at from any angle.

Svideo wasn't bad, better than HDMI for me. Though I wouldn't have believed that myself if it weren't for all the nightmare situations I've had with HDMI. At least it have given me some use for the selfie camera on my phone...

HDMI and Displayport are just hard to come in from an angle. DVI and VGA are much simpler in that regard.

Svideo was often quite simple, just find the whole and you are halfway done. Then carefully twist it and you can feel when it is correct. Even when you have no idea where the slot even is to start.

With HDMI and DP being so similar it is easy to mix them up.

The problem is the seeing part. If you have to see in order to plug in a cable then it is inherently harder to work with.

You don't, for example, hear the same complaints about DVI, VGA, Serial, or Parallel ports. Why? Because you can feel the right orientation when you plug things in.

If you can't feel the right orientation, then the next best thing is to make the orientation not matter.

This is why S-Video sucks. Because you MUST see to work with it.

Of course, the main reason people have a problem with USB is because they are constantly interacting with it, plugging and unplugging things.

Not all vga connectors are made equal, I had some that would quite easily plug in the wrong way, luckily no damage was caused by that and it was back in the day.

That sounds more like "Well, it's not as bad as PS2 connector at least."

The worst part is that they already had a good example of a connector you couldn't plug in backwards, was sturdy, hot-pluggable, and could easily accommodate the number of pins they needed: The Gameboy Link Cable.

PS/2 seems entirely rotationally ambiguous without looking at the pins, or at least seemed so to the younger version of me. It still stands out as the most painful part of computing during my childhood.

PS/2 male connectors have a set of grooves and rotationally asymmetrical rim. I had always been able to plug them in blind on the first try.

And the "top" is usually flat so you can put your thumb on it and push in.

I broke countless keyboards and mice (and not a few motherboards) due to PS2 rotation

PS/2 ports were the most painful? I guess you never used dial-up.

The Gameboy Link Cable is a fine design, but it just reminds me of all if the other fairly decent connectors that Nintendo has developed, implemented, and sold millions of.

I guess the real LPT is to always ask the oldest remaining provider of digital home electronics for advice on these matters.

Heh, you do have to hand it to them for making stuff where your use cases include "Robust and Intuitive enough for a six year old at the playground"

25-pin din parallel port connectors were D-shaped and invented in the 1970s. And there were directional connectors that long predated that.

USB's rectangular is simply a horrible oversight. Comparing it to PS2 is pretty lame. What about comparing it to nearly every other connector out there?

I'm guessing they didn't want to enlarge the serial port to one almost large enough to stick a deck of cards into.

They wouldn't have to. The point is that the D shape was a proven design at the time, and it (or something similar like the HDMI shape) could have been scaled down to USB-A size.

The PS2 plug on my keyboard has a flat surface at the 12 o'clock position which makes it very easy so plug in without looking at the pins.

I needed that info three decades ago...

iirc, the Gameboy Link Cable was physically (though not electrically) compatible with Firewire, USB's main rival, so I can imagine wanting to distance themselves from that.

USB type-B never gave me the same problems as type a, maybe they could have used that everywhere.

I used to know a guy at Intel who was the leader or one of the leaders of the industry standardization part of the USB initiative. He told me once that reversability (or at least a better scheme for easily orienting) is the one thing he wished they had done differently with the original USB.

Also micro USB and mini USB were asymmetric, right? I think the "wrong all the time" angle just didn't register when they developed the rectangular one. And I don't blame them, I wouldn't have thought of it, and probably user testing was inconclusive.

I actually hate HDMI as a connector. It has to do with the way the spade is directly in the middle. I'm more technically minded than the average consumer I guess, but with USB A the general rule is to insert the cable with the open side "up".

HDMI only keys based on the metal shielding. If you're trying to insert this without looking, it's almost impossible. I've seen that HDMI ports don't really conform to a specific orientation and very frequently get soldered on the underside of boards instead of on top.

you go to plug in your thumb drive. You try once. Failure ensues. Metal clashes with metal. Humiliated and discouraged, you flip it and try again. Failure, again! How could this be possible?

I hoped the article would explain also why the above happens, not just why the design decision which leads to the above was made.

Although I think it's not just the shape, but a combination with other things. Thingking in the direction of: most ports are all black and/or hard to reach so from a distance you just don't see where exactly it will fit so you just try and don't get a proper match. I.e. I never ever have this problem when pluggin into a socket which is right before my eyes and/or e.g. has the central plastic piece in white since then you can clearly see how to match them. Further ideas?

"I hoped the article would explain also why the above happens"

If I had to guess, I'd suspect that it's because there are a couple of ways to get it wrong when plugging in a USB drive or cable. You can get the orientation wrong but you can also accidentally put it in at a slight angle and have it catch. In the later case, just slightly adjusting your approach would get it in, but because you know that there is a chance of you having got the orientation wrong, you automatically assume that that was what happened and flip the orientation instead and end up flipping it back when you realize you were right the first time.

I suspect that the reason we make that assumption, that we got the orientation wrong and not the angle, is that that's actually what happens the majority of the time when we fail to insert a USB connector. We discount the angle problem because it happens so infrequently, and when it does happen we just get the impression that we got the orientation wrong, twice, which is something that sticks with us. Hence, the sense that there is a paradox about the orientation of USB connectors.

RJ-45 and USB ports are exactly the same width. If you plug in a USB port blindly, you can feel it make contact with the sides of an RJ-45 port.

I found this out trying to fix someone's printer. I even un- and re-plugged the cable and it still wasn't detected. It wasn't until I actually looked at the back panel that I noticed the printer's USB cable was plugged into an RJ-45 network jack.

I also discovered this while trying to plug my telescope's camera USB cable into my PC in the total darkness, and wondering why the camera didn't work.

I did this exact thing on my laptop once, and was very confused for about 5 minutes until I went to plug my Ethernet in.

I’ve also blindly plugged a USB cable onto pins of a DB9 serial port.

(1) I remember the time I mashed a USB plug into an e-SATA socket and thought I did something horribly wrong, but then I heard the beep saying that the computer recognized the device and I looked it up and found out you really can mash a USB plug into an e-SATA socket and have it work.

(2) USB-C eliminates the problem of putting the cable in upside down, but it adds the problem that there are different kinds of USB-C cable, you might pick the wrong one, get a fake, etc. It's a tough problem to make one cable that can "do everything" but not make the cable to plug your mouse in crazy expensive.

(3) I've seen quite a few devices (PCs and tablets) where the Type-C connectors seem to be floppy and not well connect to the circuit board. I haven't had one outright fail, but I think it's just a matter of time.

#1 only worked because the machine you are talking about had one of those nifty dual USB / eSATA ports. If I recall right, a LOT of the Dell laptops my company had -- for years -- had this.

I had a Nexus 5X for a few years and the USB C port started progressively giving out at around 8 months. I had to lay it down flat find a position where it would charge and hope no gust of wind would hit it. Forget charging it while using it for navigation.

I like the idea of USB C, but actual my experience with it makes me happy to currently be on a micro. You just learn which way the plug goes. USB A on the other hand, that's trickier...

If I am only charging and not using the cable for data, I have started to rely on imitation Mag safe type plugs for phones and other portables. Since the lifespan on both USB-A and USB-C is limited, hope to make stuff useable until the battery needs to be replaced (if possible) instead of a mechanical failure of the port itself or wearing down the contacts too much for use - my 2012 Macbook pro USB ports are super loose and only work if I plug the cable in just right for this reason I assume.

They make 20-pin magnetic adapters now that can carry data as well as power. Last I checked they were like $30 each.

I had a similar problem with my Nexus 5X. I worked around it as you've described for quite some time before I realized that a small amount of lint had been compressed into the port, which was preventing the connector from going all the way in.

While it's possible that the socket is damaged, USB C is designed so that the parts that are likely to break or wear out are in the cable.

I believe micro USB is the same way.

One of the things I don’t like about USB-C when compared to Lightning is that the USB-C socket has some kind of wafer inside it which strikes me as very breakable. I’ve not managed to break one yet, but why was it not put on the cable end as opposed to having the strange male-but-also-female cables we have now.

Do you know what the reasoning behind this decision was?

The wafer has fixed metal contacts on it. The part that mates with the wafer has springy metal contacts, and those wear out. This is one of the major problems with the Lightning port — the springy parts are in the port, not on the cable.

Lightning also appears to have severe problems with arcing.

> Lightning also appears to have severe problems with arcing.

Well, it IS "Lightning"...

I have never seen or even heard of the spring contacts in the lightning port failing. The biggest issue I've ever seen is lint ingress preventing good contact/locking, which can be cleaned out.

I can't speak to arcing problems.

I’ve seen attempts to clean the contacts damage them because they’re springy. I’ve also just seen them get deformed enough that they stop making good contact.

I see! That makes perfect sense, thanks for replying. I had to search about the arcing problems you mentioned, and I now have an explanation for the blackened pins on my older cable!

My Nexus 5X had decided to only accept about 20mA from USB-C (via common A to C cables) if the cable was oriented wrong. Fun times figuring that one out; this is after it had the board swapped or reworked for bootlooping around the original warranty expiration.

Not likely to spend real money on something Google put their name on again.

After like 2-3 months, every micro I've ever had has required me to stretch the cable in some way (i.e., place the phone on top of the cable) in order for it to make contact. Never had this problem with USB C.

It's usually pocket funk, in my experience.

Make a small tool (I use plastic food packaging from something like fresh strawberries) to grab and hook that stuff out of there.

And then, it works fine.

This is rather embarrassing: one of the design goals of USB-C is that it’s rated for a very large number of insertion cycles.

Wikipedia: "Micro-USB and USB-C receptacles are both designed for a minimum rated lifetime of 10,000 cycles of insertion and removal".

AFAIK an insertion cycles test is linear and extremely repetitive e.g. no bending forces and no foreign objects. It doesn't cover the wide range of real world stresses.

My Nexus 6 regularly ate micro-usb cables, even the good ones from Anker. Eventually even the port failed and it could only be charged wirelessly. So far failure rate on my Pixel 2 and Pixel 4 has been nearly zero and the only cables to fail are ones I believe I tripped over. Perhaps it has gotten better? If you're still using a 5x you might want to grab a 3a or 3b cheap for an upgrade.

Charging ports are often replaceable. I don't know about the Nexus but I paid €25 to have the charging port replaced on my Huawei phone, within an hour.

Excellent advice and I considered that, but the Nexus 6 had other problems that caused me to upgrade. Also the camera updates on the new phones were compelling.

Re: (1). That's only true for an eSATAp port. Most eSATA ports are not dual purpose like that.

On my work Macbook Air, which has 2 usb-c ports, both have failed over a course of 6 months and now I can't use the laptop at all.

There are plenty of wrong and fake regular USB cables. It has nothing to do with USB-C.

I am not sure about others, but I can not identify, by sight, a USB-C cable capable of safe high rate charging. Nor Thunderbolt. I can’t even tell which charger will work with a device with a USB-C port.

For example, I have a Fiio bluetooth adapter which will not charge with a high rate charger, only a USB-A 5v charger via a USB-A -> USB-C adapter. Is it my MacBook and iPad charger’s fault? The cable’s fault? Fiio’s USB-C chipset fault? Beats me - I have no way to know or find out, since I don’t have the tools to sniff out who is not handling the charging protocol properly (and yes, this could be the cable, since there is silicon attached to the USB-C plugs on both sides of the cable which may be involved in the negotiation as well).

This isn’t a huge problem for me now, since I have very few USB-C cables, and none which are not effectively always plugged into a charger. However, in a decade I’ll have to figure this out with a box of cables with wildly differing capabilities.

Whatever bug you're hitting is unfortunate but it's not really related to telling cable capabilities by sight.

Knowing the data rate of a cable is an utter mess. But for charging, every cable can do 60 watts. For most purposes the difference between 60 and 100 watts is pretty minor.

>Nor Thunderbolt

It's possible they exist, but I have yet to have personal experience of an actual Thunderbolt cable that didn't have a thunderbolt mark on the connector.

I have a USB power meter and learned that most of my microUSB cables can't handle 2A charging either.

Can you tell, just by looking, that this specific micro-usb cable will do data and not only power?

> adds the problem that there are different kinds of USB-C cable

Yes. We went from something remediated by looking at the plug, to something that can’t be remediated without trying it and seeing if it fails.

I've sometimes stuffed a USB cable into an Ethernet jack. It has just enough resistance to feel right.

I sincerely don't understand why ports are soldered to main boards. mechanical decoupling would help

Same reason batteries are glued to cases, or RAM/SSD is soldered in. Cost, weight, and thickness.

BoM cost.

I really thought not having to embed them in the pcb design and manufacturing would save enough to compensate for the screws on the frame.

When the port fails you buy a new one: ka-ching!

My playstation vita has limped along for years with a finicky edge connector that takes some care to get it to charge. The vita is astonishingly easy to do field repairs on, but you can't replace that connector. It's a good thing that the kind of games I like to play on the Vita come out on Steam these days.

Many electronic devices EOL because of problems with connectors. When I was an undergrad, I think every engineering student in the world had an HP28S calculator, in which the springs would eventually push the batteries right through the battery door. People tried taping it, but short from running it on a bench power supply, none of those engineering students found a great answer to fix it.

"Having to embed" them means it's more easily automated, and screws are cheaper than a flex cable and two sockets.

Sometimes when plugging in USB blind I manage to force it into 8P8C Ethernet port..

One cool trick: the USB standard [0] dictates that USB A type ports should be oriented "toward the host". On a computer, this means that the thick part of the plug should be facing toward the top surface motherboard. As long as you start in this position, you've got a good change of getting it in correctly on the first try, assuming you aren't blindly feeling and end up putting it in an RJ45 or HDMI port.

[0] https://www.phoenixcontact.com/assets/downloads_ed/global/we...

Seconded. Once I figured this out my days of plugging USB cables in the wrong way were over.

Look for the two little open holes on the face of the plug and you're golden

I don't know which way all my motherboards are oriented. Opening the case isn't easy.

motherboards are almost always going to be mounted on the right side of a tower (facing left) or facing up in a sff/nuc/laptop. if all else fails, look at the io shield. the side where all the ports line up is where the io ports are attached to the motherboard.

> A USB that could plug in correctly both ways would have required double the wires and circuits, which would have then doubled the cost.

I don't think this is quite true. You wouldn't need double the wires.

That's correct. In fact, if everything were AC coupled (signals and power) it would be easy to reverse 180 degrees. The original USB was to replace kbd/mouse 'PS2' connectors, and some audio devices - the original maxed out at 12 Mb/s raw. Given the mishmash of connectors at the time, USB even with 50% insertion rate success was pretty darn good. It seems much less good in hindsight. One does have to wonder why they didn't try to fix this when they went USB-C -- they did only half the job.

yeah, loads better than the connectors at the time that had easily-bent pins.

which part didn't USB-C fix?

Still gotta line it up.

And push it all the way in until it clicks home

On the flip side, I think the connectors would have been more than double the cost, and likely event more expensive than the cables themselves.

You could have a cable that is unidirectional, with a bidirectional connector. On a $100 motherboard, who cares if the connector raises the price by another 50 cents?

You can even have a round cable with pins radiating out from the center, and the connector would be concentric circles. Then the plug will always go in no matter what.

> On a $100 motherboard, who cares if the connector raises the price by another 50 cents?

Manufacturers do. I can't count the number of times that I've seen hardware manufacturers spend four-to-five figures in order to find a way to eliminate a 2 cent part. At scale, even a 2 cent part adds up to serious money.

So why don't they increase the cost of the motherboard by 2 cents and make serious profit?

Pricing stuff is a very complicated topic, so this is an oversimplification, but...

Because people are always looking for the lowest price for things, so manufacturers have a very strong incentive to reduce those prices. If they don't, then people will buy the competitor's gear that costs 2 cents less. In that sort of market, manufacturers won't raise prices for gear unless they have no other option. If they need to get more money out of it (because of cost increases, or because they want a higher profit margin), then the first thing they'll do is find a way to reduce manufacturing costs. Increasing the price is the last thing they'll do.

There are a lot of 2 cent parts already on the board.

Because the other guy didn't do that. Capitalism

In some industries that makes sense, and that is what happens. But where prices on parts could change between units, the sale price is often set despite those small changes in build costs.

We're talking about an industry that in 2014 included about 600 million units [0] - times maybe 4 USB ports per unit, and we're nearing 2.4 billion USB ports. That money adds up fast.

[0] - https://www.statista.com/statistics/272595/global-shipments-...

Not just motherboards though.. every USB device, not just the MB... even then, many people have 1 cable per device, so the cost is roughly the same overall. And on a MB, there isn't just one USB port, even most laptops have a couple type A's and one or two type Cs.

Both port and wire don't need to be bidirectional, it's fine if only one of them is. Making the wire side bidirectional would probably be easier.

True. The Nonda Zus car charger has reversible USB outlets for charging.


I still prefer USB-A fitting on the third try over USB-C, which may fit on the first try, but then mysteriously fails to work properly half of the time due to factors like badly designed cables or worn solder joints on some of the "reversible" pins.

> The Intel team led by Bhatt anticipated the user frustration and opted for a rectangular design and a 50-50 chance to plug it in correctly, versus a round connector with less room for error.

Why not a shape that can't be rotated, like Displayport or HDMI?

Oddly enough, I find HDMI cables while looking squarely at them to be more of a pain in the ass than trying to plug in USB-A blindly.

Both of those came after USB 1, yes? So they would have that history behind them. A precursor can't be criticised for laying the foundation for better descendants.

Later usb dont have that defense. Those micro usb abominations have no excuse. They are just waiting to break and become landfill.

Let me tell you as someone that was there, man, when USB came out. Everything else that plugged into my computer had a non-symmetrical shape. Serial ports, VGA, DVI, SCSI, even PS/2 which had a flat spot that told which way was up.

Now imagine someone showed me a new connector that looks like it could go either way, but only one way would work. They had a clean slate, and that's what they came up with? From my perspective at the time, that was the dumbest fucking thing I'd seen in a long while.

You forgot to mention IEEE 488 which was both a great thing and the cause of many failed experiments. But I digress. Yeah I saw all those and a lot of others you didn't mention. Remember DIN connectors? They were nicely round, dimpled and standard. Didn't stop students plugging them in upside down. Students also plugged vga into 25 pin serial! Or 9pin serial for that matter. I've seen students try sd cards into usb then complain it didn't work.

You didn't address microusb or even its cousin the lightning cable. Talk to anyone who fixes phones and, after the screen or battery, those damned connectors are broken by design. Flimsy. They are D shaped but only barely. I can easily plug microusb in either way. Oops. Lightning breaks because you unplugged it too many times. Its not D shaped and works both ways. Still made-for-landfill.

Yeah, Usb was a game changer went it came out. Intel did that design for size, cost and just plain "ability to get it out there". Firewire didn't really pan out. Too expensive.

So many attempts at the perfect cable. They all work in their own way. USB A did ok in general.

> Everything else that plugged into my computer had a non-symmetrical shape. Serial ports, VGA, DVI, SCSI, even PS/2

Heck, even USB B-type connectors.

I've only ever seen female B connectors on peripherals. Do they make motherboards that have them?

No, B port should be only on peripheral.

Raspberry Pi?

The type B port is for connecting the Pi as a peripheral. It has separate type A ports for connecting peripherals to it.

That’s because you’re forgetting older technologies. DB9, DB25, and VGA were all trapezoidal.

HDMI learnt nothing from the old db connectors like VGA. Try it with HDMI when the cable has to go in but you can't see the port. Get gud as they say.

At least they didn't learn lessons from PS/2, which is circular and mostly rotationally ambiguous, or at least seemed so to the younger version of me. That was the bane of computing during my childhood.

Never had trouble with PS/2. The connectors were flat on top, so it was pretty easy to "aim" them.

Triangle comes to mind

I feel like I am the only person who can plug one in correctly the first try >90% of the time. The trick is simple, look at the damn connector first.

Hard to do when you're plugging into the back of a workstation that's located under your desk. Or in a lot of other situations really.

Actually still pretty simple - In almost all scenarios, the USB connections on the PC will be oriented to point downwards or to the right. You tell what's 'down' on your USB cord by looking for the seam on the outside metal band that runs around the USB connector itself.

Orient that seam downwards or (if the ports are vertical) to the right, and most of the time it'll fit first go.

I have come across a few setups that had a USB port 'upside down', but mostly this is on laptops where they've got the connector board fitted upside-down inside the chassis.

Likewise here; I rarely experience the "catch" that would make me think I have the connector upside-down. The trick is to approach at an angle, giving a lot more room for the connector to fit.

>The trick is simple, look at the damn connector first.

Wouldn't do you any good, as the port can be on different orientation (up/down) in different devices.

If you mean look at both ends, then duh!. That doesn't count as "first try" though as it's not an actual "try" (and see), but a in-advance planned action...

Nearly all devices that have a horizontal USB port expect the "holes with shadows" side up. Look, then feel a little to jiggle it in. It takes just 2 mins to learn this and it will (well would have) save you years of frustration.

Not every cable is labeled with the USB logo. And sometimes even that is on the wrong side.

True that's not foolproof but you also have the fat plastic part down if you just glance into your cable before you go to plug. That's foolproof. But you still have to know what is 'down' on the device you are plugging into...

That is also not foolproof.

There is no method or approach that is foolproof for all cables and ports. I have cables with no logo, cables with rounded and heavy cable parts on both sides, cables with holes in the metal jacketing on both sides, and so on.

The only truly foolproof method is to look inside both the plug and the port before attempting to plug it in, to see where the metal is. Of course, that only works when you can clearly see into the port, so it doesn't work in the dark or for ports behind other things.

So yeah, nothing is foolproof when it comes to USB-A. Not 100% of the time.

You're not the only one!

Okay, and what’s your trick for knowing which way around the plug is?

Edit: Whoops, meant port.

No idea why no one else has mentioned: if the usb is horizontal, the side with the two open holes is 99% of the time the "up" side.

No need for USB logos or anything.

The cable has the “thick” part on the bottom, when you hold it with the USB symbol up.

The port is orientated “up” on the device (that is, the thick portion is on top), or orientated up with regards to the motherboard if it’s sideways. Most devices that accept USB provide these orientations.

It has a symbol printed on the side that's "up".

People keep saying this, but I have at least a half-dozen (admittedly cheap) USB cables with no symbol or logo printed on them at all.

Oh that's no good. Time to grab a pocket knife and make some marks. :)

The logo on the connector usually corresponds to the top/right of the plug

Every device I've come across has it set up so that the plastic part should be inserted closest to the floor (horizontal plugs), or closest to the centre of gravity of the device (vertical plugs).

I don't know if this is specced—otherwise it's very interesting that all the manufacturers agreed to this kind of arrangement.

Actually I just found this:


USB-A plugs have spin 2/3.

Proof That USB Cables Exist In a 4-Dimensional Space: http://www.smbc-comics.com/?id=2388

> A USB that could plug in correctly both ways would have required double the wires and circuits, which would have then doubled the cost.

why couldn't they have used a notch / tab like other connectors do? not durable enough?

I'm pretty sure some Apple USB had a notch. But I think it wasn't to help to plug it in, I think it was to ensure it was directly plugged into the computer without using a cable extension. But I don't remember well, it's only speculations.

You got it backwards. The notch was for compatibility with the extension cable, which had a lug because USB extension cables are not allowed by the spec (too much distance for the signal), so the lug prevented you from using the extension cable with a device that might fail.

Also this way you can't plug two extension cables into each other, because the cable didn't have the notch, only the lug.

You are right, it was on the extension cable!

My guess is that it's harder to manufacture. Probably not a good reason though as those harder to manufacture things already existed with Serial and Parallel ports.

Text-only link, for those who haven't got a redirect rule: https://text.npr.org/s.php?sId=734451600

I've found that if you try to plug a USB male into a USB 3.0 jack backwards, you can wedge the male connector substrate under the female connector substrate, causing it's connectors to short circuit with the housing, blowing a polyfuse that shuts the PC down abruptly. The problem is the substrate on the female connectors isn't adhered to the wall of the connector, and can bend far enough to touch the opposite wall and blow.

I'm surprised there's not more stories like this lately.

>I'm surprised there's not more stories like this lately.

Probably because what you describe requires an unusually high amount of force to be applied to the connector, causing most people to stop before damage is done.

can != its easy. Your story reminds me of my girlfriend plugging old PATA Hard drive MOLEX power cable backwards (12V to 5V and vice versa), I noticed, at the last moment, because it 'only' plugged 3/4 the way in. Mind you MOLEX is designed in a way making such connection impossible, at least in theory :).

dear cosmo...

i thought it was because of usb superposition:


This "Decline and visit plain text site" button is by far the best functionnality that they could've added

> In fact, Bhatt has not made a single penny from his USB design, because Intel owns the patent.

Does Intel not award patent bonuses? I think they're finally being phased out but historically big tech companies have given bonuses to employees who are awarded a patent.

Any insights as to why we don't have more connectors in a round form factor, such as headphone jack?

For what it's worth, I think this is the thing that USB-C has really gotten right: it works both ways.


for a USB-C to USB-C cable, you can use either end of the cable!

Grab one end of the cable, either end, and it plugs in (flipped either way).

You can plug a phone into a laptop. You can make either one of them be the "host" and you can control which one will charge the other one, or that no charging will occur at all. (At least with Android and a Pixelbook.)

this is an anti-feature in my eyes. there's no way to predict what will happen when you connect two host-capable devices.

With a Pixelbook and Android phone, it is completely predictable.

You choose which does what using UI controls on each device.

You can specify whether this device will provide charge to the other device. Whether this device will share files with the other device. And the other device can specify that it will be charged, and will receive files.

There is no guesswork about it. No mystery.

I believe the control is based on which device you plug the cable into first, except it's impossible to remember and confusing.

Not as good as the 3.5mm jack, which works in all ways.

And all that was ever needed was a colored bar on the outside of the plug and the jack.

My experience with Apple laptop, iPhone, etc. plugs that are reversible is that you inevitably end up twisting the cord to a point where its rubbery shell breaks. That never happens with USB, where the minimal viable twist is "twice over" which the outer material doesn't accept, making it untwist on the other end.

The fact that a plug is reversible doesn't mean you keep endlessly twisting it in the same direction, nor that it's immune to the forces that make the 'outer material' untwist it.

That really isn't my experience. Plugs break usually because they're pulled, for example from using the device while it's plugged in to charge. I'm not talking about violent pulls from a phone falling of the desk, just the stresses, bends and twists of normal usage.

This happens eventually on all types of cables, USB, Apple branded cables, earphone cables, everything. Higher quality cables last for longer, even much longer, but not forever.

But earphone cables rotate in place. So do the thicker kind of ordinary audio cable.

USB cables don't rotate in place, but you have to exert a lot more effort to twist them twice (so they plug in). Reversible cables have a lower barrier to twisting.

(I'm repeating myself a little, but that's because what I had just written sounds confusing in re-reading.)

(1) I plug them in correctly one time.

(2) I mark the top side outline of the USB symbol with a white paint marker.

(3) ...

(4) Profit!

Why is 2 after 1? It would help 1 succeed.

Apple solved this for all time with their magnetic orientation-agnostic connectors, and then didn't let the rest of the world have it. Vantablack bastards!

The last thing I would want is a peripheral whose data connection was on a MagSafe connector. It’s specifically designed to be able to disconnect easily by accident.

Does the article actually explain WHY?

Yes. It would have required double the wires and circuits, increasing the expense of the cable. That would have made it a difficult sell to computer manufacturers.

Not quite sure I follow that argument, but I'm not a hardware person. Nor am I looking to be one.

Doesn't explain why it wasn't made a trapezoid shape like many common connectors (serial, parallel, VGA, SCSI) of the time were.

The part of the plug with the usb logo (the wire thing splitting in three) goes on top.

Got a desktop in a tower case, USB connectors on the back plate( | | ) have no "top"

The "top" is the side furthest from the motherboard.

Interesting, is this a universal thing?

It was the plan, and the most common result. But no, it isn't universal. For one, not all usb cables even include a logo or marking for the top of the cable.

You can, with a cable, rotate it another 90 degrees and visually identify the top.

where i paint over with bright nail polish to make it easier to see (and also have seen the usb symbol on the wrong side)

This depends on the machine! But for a given machine, the logo should be oriented the same way for all cables, at least.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact