Author Topic: DANGEROUS chinese stuff  (Read 15586 times)
Medved
Member
*****
Offline

Gender: Male
View Posts
View Gallery

Re: DANGEROUS chinese stuff « Reply #60 on: July 14, 2020, 12:40:10 AM » Author: Medved
A bit out off-topic, but the resistance of the chinese xmas light cords is about 20 to 50 ohms, that's HUGE. If the cord shorts out internally all will melt in fire and set your house in fire before the breaker pops.

It depends how the resistance behaves on an overload. If it has a PTC characteristic (goes sharp up when reaching certain temperature), it maintains this temperature and remain safe. Only if ig would be solid metal, it would become a problem.
But dobt be fooled, the PTC characteristic is attainable very easily (Read: it is dirt cheap), once your system can tolerate some "cold" resistance. So I would expect this to be the main way, how the thing is designed to safely handle the eventual short circuit (when lamps start to pop and bypass shunts activate in a cascade).
Logged

No more selfballasted c***

Medved
Member
*****
Offline

Gender: Male
View Posts
View Gallery

Re: DANGEROUS chinese stuff « Reply #61 on: July 14, 2020, 12:48:31 AM » Author: Medved
I did a resistance test on the cable to know for sure and:
Positive wire: <0.2Ω
Negative wire: ~1.5Ω
Total Resistance: ~1.6Ω to 1.7Ω

Now the higher resistance could be inaccurate for a few reasons, as it could have been contact resistance from the cheap plugs. Or it could also be something like a fusible resistor in the plug.

I also did a test on a normal 3ft micro USB cable and it came out as <.2Ω per wire.


If you calculate how thick would have the copper be for the 1.5 Ohm and above, you will end up with ridiculously thin filaments. And if the resistance were on the contacts, it would be the heck unstable.
So the 0.2 Ohm per wire is a standard copper cable (non resistive), but the 1.7 Ohm one is for sure the resistive type (I guess something between "0.5A" or "1A" rated).
You may check, how tge resistance behaves at higher loadings, I wont be surprised it increases (using PTC materials is quite common way to prevent overheating in a cheap manner with these).
Logged

No more selfballasted c***

Binarix128
Member
*****
Offline

Gender: Male
View Posts
View Gallery
220V AC 50Hz, NTSC


GoL UCOUT2noI2R__jgPSJUjGRtA
Re: DANGEROUS chinese stuff « Reply #62 on: July 14, 2020, 01:10:10 AM » Author: Binarix128
@Medved I mean the cheap chinese LED xmas light, wich its mains contacts in the circuit board are too close that can short out in any moment. The cold resistance of the cord is 50 ohms, same as a 70w halogen lamp, so the wire will melt anyway consuming 300w for a second, and then all will melt in fire consuming 70w. PURE CHEAPNESS. A lower resistance wire should trip the breaker if shorts out.
Logged
High Intensity
Member
*****
Offline

Gender: Male
View Posts
View Gallery


Re: DANGEROUS chinese stuff « Reply #63 on: July 14, 2020, 01:22:18 AM » Author: High Intensity

If you calculate how thick would have the copper be for the 1.5 Ohm and above, you will end up with ridiculously thin filaments. And if the resistance were on the contacts, it would be the heck unstable.
So the 0.2 Ohm per wire is a standard copper cable (non resistive), but the 1.7 Ohm one is for sure the resistive type (I guess something between "0.5A" or "1A" rated).
You may check, how tge resistance behaves at higher loadings, I wont be surprised it increases (using PTC materials is quite common way to prevent overheating in a cheap manner with these).

The cable had a warning on it saying not to use it on USB power supplies over 1A, but i decided to to a heat test on it to see if it did have any PTC elements to it, so i heated it up with a heat gun and for a while, nothing was happening, until it shot up 1Ω when the micro USB end was being testing, however, this was after it got hot enough to start softening the plastic, so i stopped the test before anything else happened.

Now another thing to note, my sister's pair of earbuds are pretty much identical to mine in design (just a different brand), but (according to her) that pair did not have any warning about swapping the cable or anything like that, so what's different about mine?
Logged
Binarix128
Member
*****
Offline

Gender: Male
View Posts
View Gallery
220V AC 50Hz, NTSC


GoL UCOUT2noI2R__jgPSJUjGRtA
Re: DANGEROUS chinese stuff « Reply #64 on: July 14, 2020, 01:32:44 AM » Author: Binarix128
A good way to test the PTC feature of a wire is to take one wire and shorting a car battery with it...
Logged
Medved
Member
*****
Offline

Gender: Male
View Posts
View Gallery

Re: DANGEROUS chinese stuff « Reply #65 on: July 14, 2020, 02:06:47 AM » Author: Medved
A good way to test the PTC feature of a wire is to take one wire and shorting a car battery with it...

No, that is plain silly...
The cables are designed for 5V supply and any PTC functionality is supply voltage limited (the PTC action has limits in its range, plus often it is a form of a semiconductor, which increases its conductivity at higher voltage). By exceeding the voltage you may severely overheat it because the PTC action wont be able to counteract it anymore and just blow a perfectly good PTC protected cable up.

But trying it on a 5V may work, but I would better increase the voltage/current gradually and observe the response, both resistance, as well as temperature.
Logged

No more selfballasted c***

Rommie
Administrator
Member
*****
Online

Gender: Female
View Posts
View Gallery

Andromeda Ascendant


Re: DANGEROUS chinese stuff « Reply #66 on: July 14, 2020, 08:19:46 AM » Author: Rommie
Generally, I use Sennheiser or Audio Technica headphones.
However, I bought the Samsung Buds+ recently, for very good reasons:
-I'm old and now need glasses, headphones and glasses are uncomfortable
-When I'm at work I need to wear a head sock to protect my white boy skin from the sun, but I want tunes at the same time (I work on traffic signals)
-The Samsungs have an ambient aware function that allows outside sounds to be heard when the buds are in, it is adjustable on my smartphone, I can have total noise cancelling or 3 levels of ambient sound. Helps me to hear someone who may want to attack me or a car sneaking up on me.
-I considered Apple, but they are only in white and I hate Apple products.

Best of all, NOT MADE IN CHINA!

Well that's a good reason..! But I can't get the things to stay in, must be the shape of my ears or something.

I'm not getting any younger myself, and I wear glasses, don't have any problems with the Beyers, they're always very comfortable.
Logged

Ria (aka Rommie) in Aberdeen
Administrator, UK & European time zones. Any questions or problems, please feel free to get in touch :love:

"What greater gift than the love of a cat..?" - Charles Dickens
*** No smiley-only replies, please ***

Ash
Member
*****
Offline

View Posts
View Gallery


Re: DANGEROUS chinese stuff « Reply #67 on: July 14, 2020, 06:42:11 PM » Author: Ash
Interesting that the resistance is in the negative wire. That would easily mean going out of spec for devices that use the data lines... (Vd+ or Vd- will be <0V at the device end)
Logged
Binarix128
Member
*****
Offline

Gender: Male
View Posts
View Gallery
220V AC 50Hz, NTSC


GoL UCOUT2noI2R__jgPSJUjGRtA
Re: DANGEROUS chinese stuff « Reply #68 on: July 15, 2020, 12:36:35 AM » Author: Binarix128
Cheap chinese usb charging cables are just long thermistors...
Logged
Medved
Member
*****
Offline

Gender: Male
View Posts
View Gallery

Re: DANGEROUS chinese stuff « Reply #69 on: July 15, 2020, 01:08:19 AM » Author: Medved
Interesting that the resistance is in the negative wire. That would easily mean going out of spec for devices that use the data lines... (Vd+ or Vd- will be <0V at the device end)


First it wont, because the driver holds tye data lines just a bit around 1.5V and the voltage shift can not be more than 0.6V (with the allowed maximum 0.5A for USB).
Seconds the receiver is supposed to handle voltages 0.5V above and below supply and respond only on difference. Of course, during the first identification the lines are used for single ended signalling, but at that time no significant current is allowed in the supply lines.

And when operating the charger interface (that is a dedicated "protocol" for battery chargers, the only common thing with USB are the connectors and cables and the relative ease to implement interface for both at the same connector), the signal lines are either just a pull up/downs evaluated only without current or open drain from the charged device to the adapter to control the voltage, so with too high current the device just wont be able to boost the adapters output. The interface has current limit of few 100s uA (10s kOhm resistors) for the adapter to device direction, the mA range drivers are only for the device to adapter direction. So no danger for any side either, even when the charger allows larger drops.

Logged

No more selfballasted c***

Ash
Member
*****
Offline

View Posts
View Gallery


Re: DANGEROUS chinese stuff « Reply #70 on: July 15, 2020, 05:41:43 PM » Author: Ash
Most chips have ESD protection diodes inside from any data input/output pin to +Vcc and Earth. Those would clamp the data lines to -0.5V on the low side. But this also means, that once current draw of the device in question exceeds ((Vdata_pullup + 0.5V) / Rcable), the current limiting provided by the cable will no longer work (more current draw will not lead to more voltage drop on the cable), instead, the current will return to the host through the data lines

So i could see this even damaging a host (like a PC) if the device tries to draw excessive current
Logged
Medved
Member
*****
Offline

Gender: Male
View Posts
View Gallery

Re: DANGEROUS chinese stuff « Reply #71 on: July 16, 2020, 01:16:05 AM » Author: Medved
Most chips have ESD protection diodes inside from any data input/output pin to +Vcc and Earth. Those would clamp the data lines to -0.5V on the low side. But this also means, that once current draw of the device in question exceeds ((Vdata_pullup + 0.5V) / Rcable), the current limiting provided by the cable will no longer work (more current draw will not lead to more voltage drop on the cable), instead, the current will return to the host through the data lines

So i could see this even damaging a host (like a PC) if the device tries to draw excessive current

Generally dont worry for the USB safety vs wiring drop, the standard explicitely counts on drop down to 3V on supplied side from more than 5V on the computer side.
The diodes are supposed to have certain drop, but most importand if they restrict the voltage beliw the -2..7V range, they are required to handle some current indefinitely and without affecting any other circuits (the second is important when completely integrated within an ic; withnormal ESDs the current can flow only in very short pulses and once it flows, the device may not function).
Even when many DYI designs use standard ucontroller port for driving the USB lines and the USB pins on many microcontrollers are shared with "standard" port functions, the genuine USB interface uses completely separated electronic, designed according the standard. Even the other electronic on that pin, not related to USB, needs to be designed to not interfere under all conditions listed in the standard.
The data current is limited, the -2..7V includes that...
Logged

No more selfballasted c***

Ash
Member
*****
Offline

View Posts
View Gallery


Re: DANGEROUS chinese stuff « Reply #72 on: July 16, 2020, 03:46:38 PM » Author: Ash
The USB spec used to require the host to withstand a short between any wires in the cable. (due to worn out cable)

At some point the spec explicitly changed to allow the host to burn out if +5V is shorted to any of the data lines. The excuse was to allow miniaturization of host circuits (so the IC in which the root hub is located). I think the underlying reason is to allow manufacturers to make less reliable hardware in general (read : thinner gadgets that are not intended to last more than the life of one battery anyway), and there i would not count on any voltage level tolerances to be designed to survive such situations anymore
Logged
Medved
Member
*****
Offline

Gender: Male
View Posts
View Gallery

Re: DANGEROUS chinese stuff « Reply #73 on: July 17, 2020, 12:47:10 AM » Author: Medved
Data lines are easy to protect: Either tge transciever chip, include the ESD protections can tolerate the 2V or more above and beliw supply (when the ESDs are integrated), or tyere are series resistors, or the external ESD diodes are designed to handle the short circuit current before the power is cut out (as required by the USB spec). All three approaches are used, each has its pros and cons, mostly depend on the USB interface chip.
Plus that short circuit can not draw more than the device and computer data output current (which of these depends on which lines are shorted).
E.g. in mobile phones the USB connector is shared for a charger connection, USB interface itself and a passive headset connection. For this you need switches reconnecting the data lines batween the USB interface itself, the audio amplifiers or the charger controller. Usually this is in the form of a dedicated IC for this switch, which usually integrates the required protection as well. As mainly the headphone output implies negative voltages, it means the connector side has to be designed to handle them. So it then directly leads to a -7..+12V (or similar) tolerant interface, so no short in the cable could harm it. This IC is rather simple, so a lower integration process with a cheaper brutt mm^2 is used, so integrating the ESDs (which require brute force silicon area to handle the energy) is economically viable, so not that big problem to make tyem according to the voltage tolerance needs.

On the other hand a mouse or keyboard is never expected to draw more than 20mA, so the standard protection diodes could handle that easily.
Logged

No more selfballasted c***

Binarix128
Member
*****
Offline

Gender: Male
View Posts
View Gallery
220V AC 50Hz, NTSC


GoL UCOUT2noI2R__jgPSJUjGRtA
Re: DANGEROUS chinese stuff « Reply #74 on: July 17, 2020, 12:54:12 AM » Author: Binarix128
I did myself a wire for connect my external DVD burner. It was an adapter to connect the 5V from the phone charcher without the data lines and the 5v and the data lines from my computer to the external unit, for it can take the enough current for work. Then I decided to cut the 5v wires of the adapter that goes to the computer only leaving the data lines, and it didn't worked anymore. Is bad for the computer to mix 5v power supplies? Why the computed didn't detect my device when I cutted the 5v cables?
Logged
Print 
© 2005-2024 Lighting-Gallery.net | SMF 2.0.19 | SMF © 2021, Simple Machines | Terms and Policies