Inside USB-C Cables

[MarkF]

Adam Savage posted a very interesting look at USB-C cables.
Adam is joined by electrical engineer Zach Radding to examine Lumafield's CT scans of
various USB-C cables to understand the engineering and components that make up these cables.

 

[Someone]

Right in the clickbaity land image for that video, a generic (unmarked, presumably USB) cable compared to a thunderbolt cable.

Not USB-C vs USB-C

[SiliconWizard]

Yes.

And even without Thunderbolt, there are a myriad of different configurations for USB-C cables just from the specs, so even for manufacturers that fully comply, a lot of variation out there.
USB-C is a very complex spec and there's a general misconception that anything marked "USB-C" would universally cover the full spec. Not at all, it's nearly impossible anyway.

[MarkF]

I just don't care if they meet USB-C specifications...

I don't even care if they are USB-C cables or not...

I just found it interesting how the cables were constructed and the different wiring being used.
I didn't know they could do a CT-Scan of cables and get that kind of detail when metal was involved.
I have had to get many CT-Scans and always have to remove anything metal.

[tooki]

I just don't care if they meet USB-C specifications...

I don't even care if they are USB-C cables or not...

I just found it interesting how the cables were constructed and the different wiring being used.
I didn't know they could do a CT-Scan of cables and get that kind of detail when metal was involved.
I have had to get many CT-Scans and always have to remove anything metal.

You have to remove metal for an MRI. You don’t need to remove metal for a CT scan for safety, only to prevent it from affecting image quality, and that problem can be corrected to a significant degree by special methods the doctor can order (e.g. when a patient has metal implants).

CT is just computerized 3D x-ray scanning. Basically, a medical CT scanner is optimized for visualizing biological tissues, not metal. Industrial CT scanners are optimized for inspecting objects, not biological tissues.

[tom66]

Yup.  I think people get confused because both MRI and CT scanners are torodial devices and to the inexperienced eye they look similar - a toroid in which you place the patient.  A major difference is inside the CT scanner is a rotating X-ray imaging head which must be precisely balanced, whereas an MRI scanner does not rotate internally (3D imaging is achieved by modulating the fields and a lot of digital processing.)  Both have specific uses in medicine, though MRIs tend to see more use these days.

Another really cool field is 3D X-ray which is used for mechanical and electronic inspection.  I've used one of those machines at our CEM when inspecting some boards, and you can rotate the image in real time to inspect individual BGA balls.   Since the product isn't going to get cancer(!), you can do continuous imaging, allowing rotation of the part throughout the session and inspection of individual parts on a board in detail.  Medical CT scanners need to get the job done quickly.

[tooki]

Yep. In fact, that video itself is sponsored by the manufacturer of the industrial CT scanner used.

I’ll also proudly state that I gave the video a thumbs-down due to the blatant misrepresentation of a Thunderbolt cable as a USB-C cable in the title, and failure to properly explain the difference subsequently. The video gives the impression that the Apple cable is an impressive feat of gratuitous overengineering, without going into _why_ it’s so complex. They don’t compare any other Thunderbolt cables, nor even any USB 3 cables, just USB 2 cables of various quality. And the technical explanations are… I’ll generously call them “oversimplifications for a lay audience”.

So very cool imaging, very poor video.

[ejeffrey]

I mean... Thunderbolt 3 cables *are* USB C, and thunderbolt is now officially part of USB 4.
.I mean yeah you can also talk about how USb C / thunderbolt have many levels of specifications in the same form factor, but "these two things look the same but are totally different on the inside" is pretty cool especially when you have CT images.

[newbrain]

I’ll also proudly state that I gave the video a thumbs-down due to the blatant misrepresentation of a Thunderbolt cable as a USB-C cable in the title, and failure to properly explain the difference subsequently.
[...]
So very cool imaging, very poor video.

I thought I remembered that video, checked on YT, and yes, there's my thumb down - probably given when it came out.
In addition to the misrepresentation, I find that guy grating and unfunny.

[tooki]

I mean... Thunderbolt 3 cables *are* USB C, and thunderbolt is now officially part of USB 4.
.I mean yeah you can also talk about how USb C / thunderbolt have many levels of specifications in the same form factor, but "these two things look the same but are totally different on the inside" is pretty cool especially when you have CT images.

Thunderbolt cables use the USB-C connector, but they are not USB per se. Apparently, many, but not all, USB-C (without Thunderbolt) computers will support using a Thunderbolt 3 cable. USB 4 is where the line really begins to blur.

[EPAIII]

Well, I always thought that I could tackle anything electronic, but I think they now have a cable that I wouldn't even dare to try to repair. I mean, dissolve away the plastic, work under a microscope with micro-manipulators, and heaven knows what else. The easiest step would be new plastic when the repair was complete.

This makes me wonder if there are any USB-C cables with just wires from pin to pin and with all the pins connected.

And I am wondering about the Apple cable that came with the iPad I just got.

[coppercone2]

you use solder paste for that job all you need to do is position them

[tooki]

And I am wondering about the Apple cable that came with the iPad I just got.

What’s there to wonder about? Apple states in the specs that a “USB-C Charge Cable (1 meter)” is included in the box. The product page for the 1m cable says “It supports charging of up to 60 watts and transfers data at USB 2 rates.”

There’s certainly no need for a far more expensive Thunderbolt cable in the box.

[tooki]

Well, I always thought that I could tackle anything electronic, but I think they now have a cable that I wouldn't even dare to try to repair. I mean, dissolve away the plastic, work under a microscope with micro-manipulators, and heaven knows what else. The easiest step would be new plastic when the repair was complete.

Eh, it’s nothing new. Various Apple adapters (like Lightning to HDMI, for example) have been full of tiny electronics for years.

This makes me wonder if there are any USB-C cables with just wires from pin to pin and with all the pins connected.

There shouldn’t be. The USB-C standard requires a “channel configuration” chip to be present.

And it makes sense, because the pins can be used very differently depending on what the application is. USB 2 doesn’t need the high-speed lines, USB 3.0 uses the high-speed lines with twisted pair wiring, USB 3.1 uses them with micro-coaxial cables, and I don’t know what DisplayPort mode actually requires…

[coppercone2]

does anyone else want a phone with banana jacks on it instead of this crazy cable??? The 2mm banana jacks would work here OK, or even a recessed binding post, so you can stick a wire in there

[SiliconWizard]

does anyone else want a phone with banana jacks on it instead of this crazy cable??? The 2mm banana jacks would work here OK, or even a recessed binding post, so you can stick a wire in there

Only if it can be used as a small lab power supply.

[tom66]

There shouldn’t be. The USB-C standard requires a “channel configuration” chip to be present.

And it makes sense, because the pins can be used very differently depending on what the application is. USB 2 doesn’t need the high-speed lines, USB 3.0 uses the high-speed lines with twisted pair wiring, USB 3.1 uses them with micro-coaxial cables, and I don’t know what DisplayPort mode actually requires…

Are you sure a chip is required?  To make a charging or USB data cable, we just use a resistor (5.1k) from CC to GND.

[Someone]

There shouldn’t be. The USB-C standard requires a “channel configuration” chip to be present.

And it makes sense, because the pins can be used very differently depending on what the application is. USB 2 doesn’t need the high-speed lines, USB 3.0 uses the high-speed lines with twisted pair wiring, USB 3.1 uses them with micro-coaxial cables, and I don’t know what DisplayPort mode actually requires…

Are you sure a chip is required?  To make a charging or USB data cable, we just use a resistor (5.1k) from CC to GND.

Yes, passive cables are A-OK for some modes. But the usual suspects are muddying the waters and pretending like everyone only wants a bleeding edge "full featured" cable:
https://en.wikipedia.org/wiki/USB-C