Author Topic: Rant ... USB ....  (Read 7996 times)

0 Members and 1 Guest are viewing this topic.

Offline BravoVTopic starter

  • Super Contributor
  • ***
  • Posts: 7547
  • Country: 00
  • +++ ATH1
Rant ... USB ....
« on: December 18, 2017, 06:00:52 am »
Ok, maybe I'm too old ... in the beginning there was USB 1.0 .. life is starting to look good.

Then came terms for HighSpeed , FullSpeed , USB 2.0 ... Ok, abit confusing .. but most people can live with it, then everything went fine.

Then the scene starting to evolve into like these ...

usb 2.0  - Ok, thats nice, its settled now since 1.0
usb 3.0  .. Wow .. great ... better speed
usb 3.1 .. Duh .. geezz ?
usb 3.1 gen 1 ... huh ?
USB 3.1 Gen 2 ... wtf !!!  :wtf:
usb type c .... :palm:

Why don't they (USB consortium) makes an average Joe or non technical lay people out there life easier ?  :palm:

And this ?....  :wtf: :palm:
Yeah, it really happened to me actually, .... "Wifey : Honey, why its so slow ?" ... "Do I need to plug into the blue one or red one ?" etc ...  :palm:
Its not too hard to imagine a conversation like this ... "Damn it, is it the blue one or the teal blue one ? WTF, come here plug it your self  >:( ....



Not to mention the physical connectors.  :palm:


End rant.
« Last Edit: December 18, 2017, 06:02:24 am by BravoV »
 
The following users thanked this post: John C, jonovid, Electro Detective

Offline Rerouter

  • Super Contributor
  • ***
  • Posts: 4694
  • Country: au
  • Question Everything... Except This Statement
Re: Rant ... USB ....
« Reply #1 on: December 18, 2017, 06:18:40 am »
type C is just a physical connector for usb 3.0 and up, but i will admit laying out that connector almost guarantees a more expensive board,

honestly for most things i stop caring when stuff is copying faster than 120MB/s, thats an average movie in under a minute,
 
The following users thanked this post: sony mavica

Online Berni

  • Super Contributor
  • ***
  • Posts: 4924
  • Country: si
Re: Rant ... USB ....
« Reply #2 on: December 18, 2017, 06:21:16 am »
Yeah things have gotten pretty confusing with USB.

I never seen teal blue, green, yellow or orange before tho. And there i was thinking that making USB 3.0 blue was a genius move. Since some of the 3.0 connectors look identical its great to have a way to tell them apart, but this is starting to get out of hand now.

Tho my worse gripe is the Gen 1 and Gen 2 things. Just come up with a standard and stick with it like USB 2.0. I think some of the reason behind this was that other technologies started competing with USB such as Thunderbolt. They didn't want to seam slow compared to it, so they simply cranked up the clock speeds to make it go faster. Tho the two technologies are not the same thing and aimed at different price points.
 

Offline BravoVTopic starter

  • Super Contributor
  • ***
  • Posts: 7547
  • Country: 00
  • +++ ATH1
Re: Rant ... USB ....
« Reply #3 on: December 18, 2017, 06:48:45 am »
And finally the stupid pin assignment. You have up to 20V right next to a data line that goes to either an SBU pin or a CC pin, and a tin whisker or a grain of conductive dust can send 20V right into a 3.3V logic chip.

Meh ... conductive dust you say ?

I heard from friend, a true incident made by avg Joe, that a type C was pushed "snugly"  >:D into a micro usb at a phone, that wrecked the whole connector and broke the middle tab, there you go the 20V.  |O

Offline Halcyon

  • Global Moderator
  • *****
  • Posts: 5633
  • Country: au
Re: Rant ... USB ....
« Reply #4 on: December 18, 2017, 07:00:50 am »
OT: I've also inserted a type A plug into an Ethernet port (apparently without being able to see it under the desk), and I wonder why it's not enumerating...

Done that... many times!
 

Online Berni

  • Super Contributor
  • ***
  • Posts: 4924
  • Country: si
Re: Rant ... USB ....
« Reply #5 on: December 18, 2017, 07:57:21 am »
Oh yes the 100W over USB idea. Whatever engineer pitched that one in the meeting must have been high on something.

With how cost optimized USB cables are you get most cables barely capable of sustaining 500mA before the voltage starts to noticeably drop at the end. So now we are going to be putting 5A 20V trough similar Chinese bottom of the barrel cables. Okay so maybe we convince them to add some reasonable amount of copper on the power wires, then you have the user pulling out the tiny USB-C plug while 20V 5A DC is flowing trough it. That sort of power will certainly draw an arc for a short moment. In the good case it vaporizes off some of the contact material leading to a bad connection that gets hot after this happens enough times. In the bad case the arc jumps to one of the signal pins and blows the crap out of the delicate high speed USB transceivers.

We already had cases where this 20V USB power went wrong when a cheap usb-c cable between a laptop and phone caused the negotiation to go wrong and let the smoke out of them.
 

Offline HKJ

  • Super Contributor
  • ***
  • Posts: 2899
  • Country: dk
    • Tests
Re: Rant ... USB ....
« Reply #6 on: December 18, 2017, 08:32:24 am »
type C is just a physical connector for usb 3.0 and up

You also gets USB type-c in USB 2.0 flavor and that also includes cables.
 
The following users thanked this post: wraper

Offline BravoVTopic starter

  • Super Contributor
  • ***
  • Posts: 7547
  • Country: 00
  • +++ ATH1
Re: Rant ... USB ....
« Reply #7 on: December 18, 2017, 08:38:14 am »
type C is just a physical connector for usb 3.0 and up

You also gets USB type-c in USB 2.0 flavor and that also includes cables.

Its easy to imagine lay people will starting to shout out loud at the net ..... fake usb cable !

Its so slow to charge !  :-DD

Offline dmills

  • Super Contributor
  • ***
  • Posts: 2093
  • Country: gb
Re: Rant ... USB ....
« Reply #8 on: December 18, 2017, 04:25:45 pm »
USB is getting almost as annoying as SCSI did at the end.
There were :
  • SCSI
  • FAST SCSI
  • WIDE SCSI
  • FAST WIDE SCSI
  • ULTRA SCSI
  • ULTRA WIDE SCSI
  • ULTRA2 SCSI
  • ULTRA2 WIDE SCSI
  • ULTRA3 SCSI
  • ULTRA 320 SCSI


Several of these came on more then one connector variant.

Then there was the question of which type of bus terminator was required...

There was a common view that getting an arbitrary SCSI chain to work required black candles, a pentagram and the sacrifice of at least one goat to the great old ones.

Regards, Dan.
 

Offline bd139

  • Super Contributor
  • ***
  • Posts: 23018
  • Country: gb
Re: Rant ... USB ....
« Reply #9 on: December 18, 2017, 04:47:15 pm »
I don't mind it. I've had zero problems despite the total mess. This is the thing that annoys me:



Oh and the USB type A neatly fits in the ethernet hole on some laptops, if you're a fucking retard and you ram it in there hard enough. I've seen that at least 8 times this year already.
 

Offline Mr. Scram

  • Super Contributor
  • ***
  • Posts: 9810
  • Country: 00
  • Display aficionado
Re: Rant ... USB ....
« Reply #10 on: December 18, 2017, 04:56:32 pm »
I don't mind it. I've had zero problems despite the total mess. This is the thing that annoys me:



Oh and the USB type A neatly fits in the ethernet hole on some laptops, if you're a fucking retard and you ram it in there hard enough. I've seen that at least 8 times this year already.
Yeah, whoever came up with that should be banned to the deepest level of engineering hell. Many connectors before it had tabs or other visual cues which way it should go. Why on Earth would you omit that on a universal connector? And you would think you would have to try two times at most, but somehow it often ends up being 4, 5 or even 6 times, no matter how much you try to jiggle or guide the thing in. There's some quantum tomfoolery going on there for sure.

I noticed I still try to figure out which way's up when I use a USB-C connector  :palm:
 

Offline dmills

  • Super Contributor
  • ***
  • Posts: 2093
  • Country: gb
Re: Rant ... USB ....
« Reply #11 on: December 18, 2017, 04:57:11 pm »
Yep and I just love the refinement that is those USB drives with just a male contact tang that can be inserted either way up.

Annoying? Most.

Regards, Dan.
 

Offline wraper

  • Supporter
  • ****
  • Posts: 16805
  • Country: lv
Re: Rant ... USB ....
« Reply #12 on: December 18, 2017, 04:58:16 pm »
type C is just a physical connector for usb 3.0 and up

You also gets USB type-c in USB 2.0 flavor and that also includes cables.

Its easy to imagine lay people will starting to shout out loud at the net ..... fake usb cable !

Its so slow to charge !  :-DD
Actually I bought a fake Remax type C cable at supposedly reputable shop for 14 EUR WTF. Was visiting my parents and forgot the cable, bought one in the shop (quiet large chain) to be able to charge my phone. As my phone had USB type C but 2.0 only, found that it's fake only 6 months later. It had USB 3.0 on the package, "SS" on connector which was blue (but had only 4 terminals  :palm:). Turned out it had only 4 wires without any shield. Surprisingly wires were quiet thick so there was no issue with charging.

« Last Edit: December 18, 2017, 05:03:04 pm by wraper »
 

Offline tooki

  • Super Contributor
  • ***
  • Posts: 11343
  • Country: ch
Re: Rant ... USB ....
« Reply #13 on: December 20, 2017, 02:10:18 pm »
Oh yes the 100W over USB idea. Whatever engineer pitched that one in the meeting must have been high on something.
While I agree that the implementation may not be ideal, the idea is certainly reasonable. (Though frankly, the proliferation of crappy no-name cables isn’t the designers’ fault. USB has always required certification, after all.)

I mean, FireWire provided what, 45W of power? And it worked, and worked well. USB’s meager power availability was a huge downside for a long time, and the proliferation of proprietary high-power USB “standards” was the result. Or hacks like the dual-headed USB cables for external drives, that draw power from two USB ports.
 

Offline Mr. Scram

  • Super Contributor
  • ***
  • Posts: 9810
  • Country: 00
  • Display aficionado
Re: Rant ... USB ....
« Reply #14 on: December 20, 2017, 02:13:34 pm »
While I agree that the implementation may not be ideal, the idea is certainly reasonable. (Though frankly, the proliferation of crappy no-name cables isn’t the designers’ fault. USB has always required certification, after all.)

I mean, FireWire provided what, 45W of power? And it worked, and worked well. USB’s meager power availability was a huge downside for a long time, and the proliferation of proprietary high-power USB “standards” was the result. Or hacks like the dual-headed USB cables for external drives, that draw power from two USB ports.
Proper design takes human behaviour into account. It's not as if cheap Chinese knockoffs weren't a thing when they came up with this. That should have given them pause and motivated them to come up with a very robust design that can deal with most of that.
 

Offline wraper

  • Supporter
  • ****
  • Posts: 16805
  • Country: lv
Re: Rant ... USB ....
« Reply #15 on: December 20, 2017, 05:12:50 pm »
While I agree that the implementation may not be ideal, the idea is certainly reasonable. (Though frankly, the proliferation of crappy no-name cables isn’t the designers’ fault. USB has always required certification, after all.)

I mean, FireWire provided what, 45W of power? And it worked, and worked well. USB’s meager power availability was a huge downside for a long time, and the proliferation of proprietary high-power USB “standards” was the result. Or hacks like the dual-headed USB cables for external drives, that draw power from two USB ports.
Proper design takes human behaviour into account. It's not as if cheap Chinese knockoffs weren't a thing when they came up with this. That should have given them pause and motivated them to come up with a very robust design that can deal with most of that.
It's not like devices blindly pull high current. They first negotiate with charger or computer and roll back if there is high voltage drop. About crap cables, it's not problem existing solely for USB. For example, in local electronics shop I purchased fake C13 mains power cord with super thin wires, maybe good for 1-2A current.
 

Offline wraper

  • Supporter
  • ****
  • Posts: 16805
  • Country: lv
Re: Rant ... USB ....
« Reply #16 on: December 20, 2017, 05:16:11 pm »
There's some quantum tomfoolery going on there for sure.
 
The following users thanked this post: Berni, gamalot, JPortici, BrianHG, Mr. Scram

Offline bd139

  • Super Contributor
  • ***
  • Posts: 23018
  • Country: gb
Re: Rant ... USB ....
« Reply #17 on: December 20, 2017, 05:19:24 pm »
 :-DD that's hilarious :)
 

Offline Mr. Scram

  • Super Contributor
  • ***
  • Posts: 9810
  • Country: 00
  • Display aficionado
Re: Rant ... USB ....
« Reply #18 on: December 20, 2017, 05:27:30 pm »
It's not like devices blindly pull high current. They first negotiate with charger or computer and roll back if there is high voltage drop. About crap cables, it's not problem existing solely for USB. For example, in local electronics shop I purchased fake C13 mains power cord with super thin wires, maybe good for 1-2A current.
I know power is being negotiated, but I don't think it takes the actual physical wire into account. It would be good to have the system monitor a sudden rise in resistance or something, indicating a wire running hot.
« Last Edit: December 20, 2017, 05:33:11 pm by Mr. Scram »
 

Offline HKJ

  • Super Contributor
  • ***
  • Posts: 2899
  • Country: dk
    • Tests
Re: Rant ... USB ....
« Reply #19 on: December 20, 2017, 05:31:15 pm »
I know power is being negotiated, but I don't think it takes the actual physical wire into account. It would be good to have the system monitor a sudden rise in resistance or something, indicating a wire running hot.

Negotiating is a bit on the optional side in real life*, but usual devices will reduce current draw when the voltage drops.

*I believe phones always do it, but a lot of other equipment do not.
 

Online Berni

  • Super Contributor
  • ***
  • Posts: 4924
  • Country: si
Re: Rant ... USB ....
« Reply #20 on: December 21, 2017, 06:28:30 am »
Power is indeed supposed to be negotiated for 500mA in USB 2.0 but in practice nobody does it since pretty much all PC and USB hub manufacturers found it easier to just provide 500mA all the time. So as a result a lot of USB devices don't bother negotiating for it and just take the current.

As for >500mA devices like phones they simply try to take as much current as possible until they can see the voltage drop too much(usually around 4.5V) at that point they back off the current. Works pretty well in practice as it requires no fancy protocol that some manufacturers would ignore anyway (apple) and will reduce the current on a crappy copper saving USB cable as the resistance is visible as a voltage drops. The only annoying thing is that phones don't really show any indicator of how much power they managed to squeeze out.

A product i designed with USB did the same thing for USB charging, pulling up to 1A when possible but scaling down if the volts start to drop. Worked pretty well, the supplied wall charger gets the 1A while on most PCs it scales down to what its capable of. Then one of the software developers called me over to ask me why his prototype unit doesn't want to charge at all. I already thought "well crap the charger circuitry must have died somehow". Upon investigating further i found that he was using a USB cable so crappy that the charging circuit only got to about 150mA before the voltage would sag so much that it backed off the current. As a result the device was using slightly more power while running than it was getting from USB to charge. Gave him a new USB cable and everything was fine.
« Last Edit: December 21, 2017, 06:30:33 am by Berni »
 

Offline HKJ

  • Super Contributor
  • ***
  • Posts: 2899
  • Country: dk
    • Tests
Re: Rant ... USB ....
« Reply #21 on: December 21, 2017, 07:17:37 am »
As for >500mA devices like phones they simply try to take as much current as possible until they can see the voltage drop too much(usually around 4.5V) at that point they back off the current.

Phones usual looks for special coding on the data pins, most other devices just draws the current.
 

Online Berni

  • Super Contributor
  • ***
  • Posts: 4924
  • Country: si
Re: Rant ... USB ....
« Reply #22 on: December 21, 2017, 07:34:47 am »
Phones usual looks for special coding on the data pins, most other devices just draws the current.

True some phones check for certain things before attempting to pull >500mA. This is usualy the data pins shorted together or the apple approach of having a voltage divider on the data pins. But even when signaled to go in to this high power mode it will still reduce the current if the voltage sags so cheap high resistance cables still slow it down.

Then there is also Qualcomm QuickCharge standard. That instead uses digital communication over the data pins to allow the phone to tell the charger what it wants. Allows for up to 20V 60W to be output from the charger provided the charger is capable. This is separate from the USB 3.0 20V charging thing.
 

Offline HKJ

  • Super Contributor
  • ***
  • Posts: 2899
  • Country: dk
    • Tests
Re: Rant ... USB ....
« Reply #23 on: December 21, 2017, 07:47:22 am »
Phones usual looks for special coding on the data pins, most other devices just draws the current.

True some phones check for certain things before attempting to pull >500mA. This is usualy the data pins shorted together or the apple approach of having a voltage divider on the data pins. But even when signaled to go in to this high power mode it will still reduce the current if the voltage sags so cheap high resistance cables still slow it down.

The DCP and Apple is only two of the many different encoding schemes used today.


Then there is also Qualcomm QuickCharge standard. That instead uses digital communication over the data pins to allow the phone to tell the charger what it wants. Allows for up to 20V 60W to be output from the charger provided the charger is capable. This is separate from the USB 3.0 20V charging thing.

Digital communication? Maybe, sort off, but not anything like sending bits around. It is only some specific voltage levels on the data pins. QC usual tops at 18 Watt, that standard allows for more, but it is very uncommon.
 

Online Berni

  • Super Contributor
  • ***
  • Posts: 4924
  • Country: si
Re: Rant ... USB ....
« Reply #24 on: December 21, 2017, 08:29:35 am »
Well not sure if its actually exchanging bits, but there is some sort of handshake sequence at the start that switches the changer from being a normal 5V charger into quick charge mode, this ensures that it can't enter this mode by accident since that could fry the device.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf