The Dangers Of Cheap Power Cables

[iamdarkyoshi]

Just a slight PSA here. I've got lots of power cables around my house, some of them are good quality, and some of them came from cheap stuff from ebay, like those "replacement chargers" we all love...

 

Anyway, the power leads that are supplied with them also seem to be quite unsafe. 

I've splayed the prongs out on this one and made this more readable using thermal paste. This cable should have no issues passing 10A through live and neutral.

 

 

To test this, I'll cut both ends off, and connect one end to my bench PSU, and twist the other ends together (I ended up soldering them after this image was taken)

Those wires look awfully thin, but they've done their research, right?

 

Anyway, my bench power supply can only put out up to 5.1A, which is just a little more than HALF this cable's rated 10A. Lets switch the supply on.

That cable should NOT be dropping over 14V, dissipating over 75W! Lets let it run for a little while longer...

 

 

Oh dear, the increased temps has increased the resistance, causing the voltage to climb, we are now dissipating over 100W of power in this cable.

 

 

I'm going to turn that off before my garage fills with smoke, but look a the state of the cable!

 

 

Had this been on an average gaming desktop, this would have EASILY caused a fire. The sad part is how many of these exist. I've got more than I can count of these crappy power leads.

 

How can you tell if you've got a poor quality cable?

 

Look for a UL number on the cable (at least for those in the US) These can easily be faked, but not having one is an easy sign. Other countries may have different safety markings, look for the one your country uses.

 

The feel of these cables feels different to a proper cable. These can be bent very easily around tight corners compared to a proper certified cable.

 

The prongs are weak. I can break a good quality cable's prongs, but this one was so much easier than a good one. That photo at the start of the topic with the bent prongs took almost no effort.

 

You could measure the resistance with a multimeter from one end of the cable to the other, it should be very low. These were several ohms per conductor, totally not acceptable.

 

If you've got one of these, destroy the ends so they cannot be plugged in, and dispose of them.

 

 

As usual guys, stay safe!

[vanished]

30 minutes ago, iamdarkyoshi said:

[...] You could measure the resistance with a multimeter from one end of the cable to the other, it should be very low. These were several ohms per conductor, totally not acceptable. [...]

For anyone wondering how this works, here's a little crash course in electricity.

 

Voltage is the driving force behind it, just as pressure moves things mechanically and temperature moves heat.

A voltage applied across any particular resistance will cause a certain amount of current to flow, as per ohms law: V = I * R, where V is voltage in volts, I is current in amps (don't ask), and R is resistance in ohms.

 

 

In this example, the "power source" represents the wall, a 120 V source.  Both the wire, and whatever you are driving (say, a PC) are resistances, or loads, measured in ohms. (in real life with AC things are much more complicated but for the sake of explaining the wire thing this will do)

 

Say the wire has a resistance of 5 ohms, and the load has a resistance of 20 ohms.  The total resistance in the circuit is thus 25 ohms, and thus the 120 volt source will force 120 / 25, or in other words 4.8 amps through the circuit (both the load and the wire).

 

Now, we know that the voltage drop across this circuit must be 120 V since that's what the source has defined, but how much of that occurs in each item?  It turns out it is proportional to the resistance.  So the wire gets a 120 * (5/25) V drop across it, and the load gets a 120 * (20/25) volt drop across it.  Those work out to 24 and 96 V respectively.

 

Now hold on, if there is a 24 volt drop across the wire, and current is flowing through at 4.8 amps, that means the power it is "consuming", and thus the heat it is releasing to the room is 24 V * 4.8 A, which is 115.2 W!  That's huge!  Well, I said it is releasing that to the room... that might not be true.  That's what it must release to the room if it intends to remain at room temperature.  In reality, it will not release heat at that rate, and thus it will heat up... and then melt

 

You can see if the wire had a resistance of only 0.1 ohms, things would be very different.  Assuming the PC load stays the same, the current through the system would now be 120 / 20.1 = 5.97 A, and the voltage drop of the wire would be 120 * (0.1/20.1) = 0.6 V, and the power dissipated by the wire must then be 0.6 V * 5.97 A = 3.56 W.  Probably not even enough to make it noticeably warmer to the touch.  

[iamdarkyoshi]

2 minutes ago, Ryan_Vickers said:

 

In terms of measuring resistance, just put the meter on ohms and measure from one side to the other.

[themctipers]

i just use whatever came with whatever

 

monitor? use the power cable provided in the box.

psu? power cable from box..

laptop? power cable from box....

 

 

other shit? unopened power cables from.. psus. 

[vanished]

1 minute ago, themctipers said:

i just use whatever came with whatever

 

monitor? use the power cable provided in the box.

psu? power cable from box..

laptop? power cable from box....

 

 

other shit? unopened power cables from.. psus. 

If it's a reputable item certified in Canada that should be safe, but it never hurts to check them!  In fact, it could hurt quite a bit not to so it's probably a good idea regardless

[Skanky Sylveon]

So, if the cable feels like it is of high quality, i'm probably safe?

[themctipers]

Just now, Ryan_Vickers said:

If it's a reputable item certified in Canada that should be safe, but it never hurts to check them!  In fact, it could hurt quite a bit not to so it's probably a good idea regardless

$25 evga 600b

lol

 

i can't imagine why anyone would need more power cables than the amount of electronics you own..

how did you get 5 devices when you've only bought 4? even at that, if you needed to borrow one, use one from your other 4! 

[ARikozuM]

35 minutes ago, iamdarkyoshi said:

Interesting thread... I initially wondered if the above was why the cheap ones are bad... I thought "how did he plug it on like that?!"...

[iamdarkyoshi]

Just now, Cinnabar Sonar said:

So, if the cable feels like it is of high quality, i'm probably safe?

Most likely. If you suspect otherwise though, just put a decent load on it, and feel if it gets warm. It should stay cold.

[Guest]

A cable costs like $10 (good ones) 

Your entire system probably costs way more.

 

But not sure if buying a multimeter to test if a garbage cable is working would be worth it

 

2 minutes ago, ARikozuM said:

Interesting thread... I initially wondered if the above was why the cheap ones are bad... I thought "how did he plug it on like that?!"...

You also spread the holes 

[Skanky Sylveon]

6 minutes ago, iamdarkyoshi said:

Most likely. If you suspect otherwise though, just put a decent load on it, and feel if it gets warm. It should stay cold.

Okay, Thanks.

3 minutes ago, deXxterlab97 said:

But not sure if buying a multimeter to test if a garbage cable is working would be worth it

 

I believe that multimeters/PSU testers are well worth the investment.

[unknownmiscreant]

28 minutes ago, iamdarkyoshi said:

In terms of measuring resistance, just put the meter on ohms and measure from one side to the other.

 

8 minutes ago, deXxterlab97 said:

A cable costs like $10 (good ones) 

Your entire system probably costs way more.

 

But not sure if buying a multimeter to test if a garbage cable is working would be worth it

 

You also spread the holes 

Measuring less than 5ohms (cables etc) can't be done accurately with a multimeter. Even my Fluke 28-II (same specs as the industry standard fluke 87v) can't measure cable resistance accurately. Using the 4-wire kelvin method is much better. Granted it is more complicated, but if done properly yields considerably better results. This video is quite a good explanation:

3 minutes ago, Cinnabar Sonar said:

Okay, Thanks.

I believe that multimeters/PSU testers are well worth the investment.

Very true. Also add oscilloscope to that list.

[Guest]

7 minutes ago, Cinnabar Sonar said:

Okay, Thanks.

I believe that multimeters/PSU testers are well worth the investment.

 

1 minute ago, unknownmiscreant said:

 

Measuring less than 5ohms (cables etc) can't be done accurately with a multimeter. Even my Fluke 28-II (same specs as the industry standard fluke 87v) can't measure cable resistance accurately. Using the 4-wire kelvin method is much better. Granted it is more complicated, but if done properly yields considerably better results. This video is quite a good explanation:

Very true. Also add oscilloscope to that list.

Right. I never had to use other cables than ones provided though apart from data/video cables (SATA and HDMI) so for me it's not something needed. But for people who buy used stuff for cheap a lot I can see it's useful, especially when some cables can be picked up for free lying around your house or in some junkyard

[iamdarkyoshi]

2 minutes ago, unknownmiscreant said:

 

Measuring less than 5ohms (cables etc) can't be done accurately with a multimeter. Even my Fluke 28-II (same specs as the industry standard fluke 87v) can't measure cable resistance accurately. Using the 4-wire kelvin method is much better. Granted it is more complicated, but if done properly yields considerably better results. This video is quite a good explanation:

Very true. Also add oscilloscope to that list.

Even my 10$ meter can distinguish between this cable and a good one. You just short the probes to get a "0" measurement, and a good cable should be as close as possible to that reading, whereas this one was several ohms, easily detectable.

[Skanky Sylveon]

2 minutes ago, deXxterlab97 said:

 

Right. I never had to use other cables than ones provided though apart from data/video cables (SATA and HDMI) so for me it's not something needed. But for people who buy used stuff for cheap a lot I can see it's useful, especially when some cables can be picked up for free lying around your house or in some junkyard

Sometimes, you just need longer cables.

[Guest]

Just now, Cinnabar Sonar said:

Sometimes, you just need longer cables.

Ah yeah that make sense. But I have a surge protector for that and it comes with quite long cables. Easier to unplug replug things in case of shit happens with PC

[unknownmiscreant]

16 minutes ago, iamdarkyoshi said:

Even my 10$ meter can distinguish between this cable and a good one. You just short the probes to get a "0" measurement, and a good cable should be as close as possible to that reading, whereas this one was several ohms, easily detectable.

Oh okay, thats a different level of crap to what I'm used to dealing with. Generally when I start trying to measure a cable, its under 1 ohm. I can short the test leads, and stick the meter to relative, so its zeroed, but the reading fluctuates too much to be useful. IO think the only time I've encountered this was with a USB cable I was using to power an AV recorder that had such high resistance, the AV recorder (a very low powered 240p affair) could not pull enough current to operate. Pulling the cable apart revealed that the power lines were the same thickness as the data lines.

 

$10 multimeters are surprisingly good, I've purchased $10 multi meters in the past that out of the box match the readings of the fluke (albeit with less decimal places). Having said that, I also have a $40 multimeter thats pretty crap, its as accurate as a $10 one, all you get for the extra $$ is a rather terrible (almost useless) continuity check, and auto ranging.

 

Edit: looking at your first picture, another tell-tale sign of crap quality is that the earth prong is hollow, rather than solid. Feel free to correct me, I have not seen alot of US plugs in my time, buts the ones I have seen were all bundled with decent quality equipment from reputable manufacturers, and they all have solid earth prongs.

[Skanky Sylveon]

25 minutes ago, deXxterlab97 said:

Ah yeah that make sense. But I have a surge protector for that and it comes with quite long cables. Easier to unplug replug things in case of shit happens with PC

I have surge protectors too, just some of the equipment is positioned in a way where a long cable is better (no dangling surge protectors please ).

But yeah, you generally get what you pay for.  You buy cheap shit, expect it to be shit.

[iamdarkyoshi]

7 minutes ago, CUDA_Cores said:

 You need to attach a load to the end of the cable. This could be a giant dummy resistor, a giant lightbulb of a spool of steel wire wrapped around a piece of wood it doesn't matter. No wonder it's dissipating 75w you just dead shorted power cables, duh! The electricity has nothing to power on the other end, so you are asking a dinky little wire to dissipate 75 watts of power. Take your multimeter and measure the total voltage drop across the entire cable then multiply this value by the current. If you want to calculate the resistance of the wire, take the voltage drop and divide it by the current to calculate the total resistance of the wire. 

I'm going to have to disagree here. My bench power supply was set to 40v, but it is current limited to 5.1A, so it dropped the voltage down to 20V. If I put a 12v 5.1A lamp in series, I'd still be at 5.1A, but at 32v, 5.1A. The cable is still dropping the same amount of voltage (20V) at the same current, resulting in the same power loss.

[mariushm]

The PLUG is rated for 10A , the cable isn't

 

Yeah, Chinese crap... they use steel coated with copper or aluminum coated with copper (or they don't even bother coating with copper) in order to produce cheaper wire, which means the resistance of the wire will be much  higher

Just don't buy the cheapest shit, or just make your own cables.

[unknownmiscreant]

@CUDA_Cores I'm going to agree with this:

43 minutes ago, iamdarkyoshi said:

I'm going to have to disagree here. My bench power supply was set to 40v, but it is current limited to 5.1A, so it dropped the voltage down to 20V. If I put a 12v 5.1A lamp in series, I'd still be at 5.1A, but at 32v, 5.1A. The cable is still dropping the same amount of voltage (20V) at the same current, resulting in the same power loss.

 

4 minutes ago, CUDA_Cores said:

100w/120v means out device will use 0.83 amps from the wall. Because our load is much smaller now, our wire will no longer be dropping 20v. This is why we calculated the wire resistance in ohms before. 4 ohms x 0.83 amps = 3.32 volts drop. 3.32 volts x 0.83 amps only leaves us with ~2.75w to dissipate over our wire.

I think you're missing the point that the resistance of the cable changes with temperature. Sure the cable will be fine with small loads, but at larger loads, the amount of heat dissipated within the cable (assuming an ideal case where the cable resistance remained constant) is O(n^2). P=I^2*R. Adding in the fact that resistance increases with temperature, it gets quite ugly quite fast.

 

9 minutes ago, CUDA_Cores said:

Even after all this, I still doubt your numbers. Whenever I measure wire resistances I measure the voltage with a seperate decive from my power supply to provide a more accurate reading since sometimes these bench power supplies can display an incorrect output voltage after a load has been applied. This is why I always measure voltage under load with a separate device.

Depends on the power supply. Decent quality offerings have very accurate voltage and current readings. But for low quality affairs, you definitely should measure using a separate device.

 

The numbers may be a bit off, but the bottom line is the cable still burned up while carrying half its rated current. Thats really not good.

 

20 minutes ago, mariushm said:

The PLUG is rated for 10A , the cable isn't

 

Yeah, Chinese crap... they use steel coated with copper or aluminum coated with copper (or they don't even bother coating with copper) in order to produce cheaper wire, which means the resistance of the wire will be much  higher

Just don't buy the cheapest shit, or just make your own cables.

Aluminium cables are actually used in the power grid. However they are a fair bit larger than their copper counter parts to counter the higher lineal resistance of aluminium.

 

Making your own cables is very rarely worth it. Unless I need something with a very specific pin out, or special cabling I always buy a cable and cut it up. Usually the cost of one connector, let alone the wire is the same as the cost of a middle of the road pre-made cable that is plenty good enough for my purposes.

[vanished]

10 hours ago, Cinnabar Sonar said:

So, if the cable feels like it is of high quality, i'm probably safe?

Remember some places have learned to cheat though.  Linus talked about how you used to be able to get a rough idea of PSU quality just by the weight, until some cheap manufacturers started adding dead weight to them  

 

Do the feel test for temperature, and/or check resistance with a meter.  If it's several ohms, it's probably going to cause problems.  I described how this works above, but that was also a hybrid sort of explanation of how electricity works for those who don't know, and for the purposes of actually figuring out the wattage "in" the cable, there's a simpler way.  Power = Current² * Resistance.  The resistance is what you measured in ohms, and the current is the current in the wire.  If you have a rough idea of how much power the item you have plugged in consumes, you can divide that by your wall voltage* (120 probably) to get this number.

 

*OK technically this needs to be divided by the voltage drop across that item but you may or may not be able to actually test that, so just know that this will underestimate the problem by a bit... moreso the worse your cable is.  Bottom line, if it's 50, 100, or more watts, that's WAY too much!

10 hours ago, deXxterlab97 said:

You also spread the holes 

( ͡° ͜ʖ ͡°)

[Skanky Sylveon]

2 minutes ago, Ryan_Vickers said:

Remember some places have learned to cheat though.  Linus talked about how you used to be able to get a rough idea of PSU quality just by the weight, until some cheap manufacturers started adding dead weight to them  

 

Do the feel test for temperature, and/or check resistance with a meter.  If it's several ohms, it's probably going to cause problems.  I described how this works above, but that was also a hybrid sort of explanation of how electricity works for those who don't know, and for the purposes of actually figuring out the wattage "in" the cable, there's a simpler way.  Power = Current² * Resistance.  The resistance is what you measured in ohms, and the current is the current in the wire.  If you have a rough idea of how much power the item you have plugged in consumes, you can divide that by your wall voltage (120 probably) to get this number.

I have a general idea on how that all works, Ohms = resistance, resistance = heat, heat =

 

It's nice to have a refresher course though.  Thanks.

Do you know of any equipment that can test the cables without splicing them?

[vanished]

Just now, Cinnabar Sonar said:

I have a general idea on how that all works, Ohms = resistance, resistance = heat, heat =

It's nice to have a refresher course though.  Thanks.

Do you know of any equipment that can test the cables without splicing them?

of course, you just need to probe both ends

 

 

If you get a reading of infinite ohms, you poked the wrong hole, so try again

[Skanky Sylveon]

2 minutes ago, Ryan_Vickers said:

of course, you just need to probe both ends

 

 

If you get a reading of infinite ohms, you poked the wrong hole, so try again

Now I feel like an idiot...