One of the amazing things about USB-C is its high-speed capabilities. The pinout gives you four high-speed differential pairs and several low-speed differential pairs, allowing you to transfer large amounts of data through connectors for less than a dime. Not all devices use this feature, nor should they – USB-C was designed to be accessible to all portable devices. However, when your device needs high speed over USB-C, you will find that USB-C can give you that high speed and how well it performs.
The ability to get a high-speed interface from USB-C is called Alternate Mode, or Alternate Mode for short. The three alternatives you may encounter today are USB3, DisplayPort, and Thunderbolt, with some already fading, like HDMI and VirtualLink, and some on the rise, like USB4. Most alternate modes require USB-C digital communication using some type of PD link messaging. However, not all USB3s are the simplest. Let’s see what the alternative template does.
If you’ve seen the pinout, you’ve seen the high speed pins. Today I want to show you what interfaces are available from these pins today. This is not a complete or extensive list – I won’t talk about things like USB4, for example, partly because I don’t know enough about it or have experience with it; it’s safe to assume that we’ll get more USB-equipped devices in the future -C for high-speed devices. Also, USB-C is flexible enough that hackers can expose Ethernet or SATA in a USB-C compatible way – if that’s what you’re looking for, maybe this review can help you figure it out.
USB3 is very, very simple – just a couple of TX and a couple of RX, although the transfer rate is much higher than USB2, it is controllable for hackers. If you are using a multilayer PCB with USB3 signal impedance control and respect for differential pairs, your USB3 connection will usually work fine.
Not much has changed for USB3 over USB-C – you’ll have a multiplexer to handle rotation, but that’s about it. USB3 multiplexers abound, so if you add a USB3-enabled USB-C port to your motherboard, you’re unlikely to run into problems. There is also Dual Channel USB3, which uses two parallel USB3 channels to increase bandwidth, but hackers don’t usually run into or need this, and Thunderbolt tends to cover this area better. Want to convert a USB3 device to a USB-C device? All you really need is a multiplexer. If you are thinking about installing a MicroUSB 3.0 connector on your motherboard for your high-speed devices, then I politely but strongly ask you to change your mind and install a USB-C connector and VL160 on it.
If you’re designing a USB3 device with a plug, you don’t even need a multiplexer to handle rotation – in fact, you don’t need any rotation detection. A single uncontrolled 5.1kΩ resistor is enough to create a USB3 flash drive that plugs directly into a USB-C port, or to create a USB-C male to female USB-A 3.0 adapter. As far as sockets go, you can avoid using a multiplexer if you have free USB3 connections to sacrifice, which of course isn’t all that much. I don’t know enough about dual channel USB3 to be sure if dual channel USB3 supports such a connection, but I think the answer “no” would be more likely than “yes”!
DisplayPort (DP) is a great interface for connecting high-resolution displays – it has overtaken HDMI on desktops, dominating the built-in display space in the form of eDP, and delivering high resolution over a single cable, often better than HDMI. It can be converted to DVI or HDMI using an inexpensive adapter that uses the DP++ standard and is royalty-free like HDMI. It makes sense for the VESA alliance to work with the USB group to implement DisplayPort support, especially as DisplayPort transmitters in SoCs become more and more popular.
If you’re using a dock with an HDMI or VGA output, it uses DisplayPort Alternate Mode behind the scenes. Monitors increasingly come with a DisplayPort input over USB-C, and thanks to a feature called MST, you can link monitors, giving you a multi-monitor configuration with a single cable – unless you’re using a Macbook, as Apple has abandoned with macOS. MST is supported in .
Also, interesting fact – DP Alternate Mode is one of the few Alternate Modes that uses SBU pins that are remapped to DisplayPort AUX pair. The general lack of USB-C pins also means that DP configuration pins must be excluded, except for DP++ HDMI/DVI compatibility mode, so all USB-C DP-HDMI adapters are effectively active DP-HDMI converters. Masking – Unlike DP++, DP++ allows you to use level switches for HDMI support.
If you want to change the DisplayPort, you will probably need a DP-enabled multiplexer, but most importantly, you must be able to send custom PD messages. First, the whole “grant/request alternate DP mode” part is done through the PD – there aren’t enough resistors. There are also no free pins for the HPD, which is a critical signal in DisplayPort, so hotplug and abort events are sent as messages over the PD link. That said, it’s not very hard to implement, and I’m thinking of a hacker-friendly implementation – until then, if you need to use DP Alternate Mode to output DP or HDMI over a USB-C port, there are chips like the CYPD3120 that allows you to write firmware for this.
One of the things that makes DP Alternate Mode stand out is that it has four high-speed lanes on USB-C, allowing you to combine a USB3 connection on one side of the USB-C port and a dual-link DisplayPort connection on the other. This is how all “USB3 Ports, Peripherals, and HDMI Out” docks work. If two-lane resolution is a limitation for you, you can also buy a quad-lane adapter – due to the lack of USB3, there will be no data transfer, but you can get higher resolution or frame rates with two additional DisplayPort lanes.
I find DisplayPort Alternate Mode to be one of the best things about USB-C, and while the cheapest (or most unfortunate) laptops and phones don’t support it, it’s nice to have a device that does. Of course, sometimes a big company gets that joy directly, as Google did.
In particular, via USB-C you can get Thunderbolt 3, and soon Thunderbolt 4, but so far it’s just fantastic. Thunderbolt 3 was originally a proprietary specification that was eventually open sourced by Intel. Apparently they aren’t open enough or have another caveat, and since Thunderbolt 3 devices in the wild are still being built exclusively with Intel chips, I’m guessing the lack of competition is the reason why prices remain treble stable. digital territory. Why are you looking for Thunderbolt devices in the first place? In addition to higher speed, there is another killer feature.
You get PCIe bandwidth over Thunderbolt as well as up to 4x the bandwidth! This has been a hot topic for those who need eGPU support or fast external storage in the form of NVMe drives that some hackers use for PCIe-attached FPGAs. If you have two Thunderbolt-enabled computers (for example, two laptops), you can also connect them using a Thunderbolt-enabled cable – this creates a high-speed network interface between them without additional components. Yes, of course, Thunderbolt can easily tunnel DisplayPort and USB3 internally. Thunderbolt technology is very powerful and delicious for advanced users.
However, all this coolness is achieved through a proprietary and complex technology stack. Thunderbolt is not something that a lone hacker can easily create, although someone should try it someday. And despite the Thunderbolt dock’s many features, the software side often causes problems, especially when it comes to things like trying to get sleep to work on a laptop without crashing the eGPU core. If it’s not obvious yet, I’m looking forward to Intel putting it together.
I keep saying “multiplexer”. What is this? In short, this part helps to handle the high-speed handshake according to USB-C rotation.
High-Speed Lane is the part of USB-C that is most affected by port rotation. If your USB-C port uses High Speed Lane, you will need a multiplexer (multiplexer) chip to manage the two possible USB-C turns – aligning the orientation of the ports and cables at both ends with the actual internal high speed receivers. and transmitters are matched to the connected device. Sometimes, if the high speed chip is designed for USB-C, these multiplexers are inside the high speed chip, but often they are separate chips. Want to add Hi-Speed USB-C support to a device that doesn’t already support Hi-Speed USB-C? Multiplexers will underpin high-speed communications operations.
If your device has a USB-C connector with High Speed Lane, you will need a multiplexer – fixed cables and devices with connectors do not need it. Generally, if you’re using a cable to connect two high-speed devices with USB-C slots, they’ll both need a multiplexer—controlling cable rotation is the responsibility of each device. On both sides, the multiplexer (or the PD controller connected to the multiplexer) will control the direction of the CC pin and act accordingly. Also, many of these multiplexers are used for different purposes, depending on what you want from the port.
You will see multiplexers for USB3 in cheap laptops that only implement USB 3.0 on a Type-C port, and if it supports DisplayPort, you will have a multiplexer with an additional input to mix these device signals. In Thunderbolt, the multiplexer will be built into the Thunderbolt chip. For hackers who work with USB-C but don’t have access to Thunderbolt or don’t need Thunderbolt, TI and VLI offer a number of good multiplexers for a variety of purposes. For example, I’ve been using DisplayPort over USB-C lately, and the VL170 (seems to be a 1:1 clone of TI’s HD3SS460) looks like a great chip for DisplayPort + USB3 combo use.
USB-C multiplexers that support DisplayPort (like the HD3SS460) don’t natively do CC pin control and turn detection, but that’s a reasonable limitation – DisplayPort requires a fairly application-specific PD link, which is very important. multiplexer capabilities. Are you happy with USB3 that doesn’t require a PD connection? The VL161 is a simple USB3 multiplexer IC with a polarity input, so you can define the polarity yourself.
If you also don’t need polarity detection – is a 5v only analog PD sufficient for your USB3 needs? Use something like the VL160 – it combines analog PD receivers and sources, processing power and high-speed track interleaving all in one. It’s a real chip “I want USB3 over USB-C, I want everything to be managed for me”; for example, recent open source HDMI capture cards use the VL160 for their USB-C ports. To be fair, I don’t need to single out the VL160 – there are dozens of such microcircuits; “USB3 mux for USB-C, do it all” is probably the most popular type of USB-C related chip.
There are several legacy USB-C alternate modes. The first one, which I won’t shed a tear for, is HDMI Alternate Mode; it simply places the pins of the HDMI connector over the pins of the USB-C connector. It can give you HDMI over USB-C, and it appears to have been available on smartphones for a short time. However, it has to compete with the ease of converting to HDMI DisplayPort Alternate Mode, while HDMI-DP conversion is often costly and cannot be used in conjunction with USB 3.0 because HDMI requires four differential pairs and HDMI licensing baggage, according to appears to be spurring the development of HDMI Alt Mode into the ground. I truly believe it should stay there because I don’t believe our world can be improved by adding more HDMI.
However, another one is actually quite interesting – it is called VirtualLink. Some big tech companies are working on USB-C capabilities in VR – after all, it’s pretty cool when your VR headset only needs one cable for everything. However, VR goggles require high-resolution dual-display, high-frame-rate video interfaces, as well as high-speed data connections for additional cameras and sensors, and the usual “dual-link DisplayPort + USB3″ combination cannot provide such features at the time. And what do you do then
The VirtualLink team says it’s easy: you can connect two USB2 redundant pairs to a USB-C connector and use four pins to connect USB3. Remember the USB2 to USB3 conversion chip I mentioned in a short article half a year ago? Yes, its original target was VirtualLink. Of course, this setup requires a more expensive custom cable and two additional shielded pairs, and requires up to 27W of power from the PC, i.e. a 9V output, which is rarely seen on USB-C wall chargers or mobile devices. power. The difference between USB2 and USB3 is frustrating for some, but for VR VirtualLink looks very useful.
Some GPUs come with VirtualLink support, but that’s not enough in the long run, and laptops notorious for often lacking USB-C ports don’t either. This caused Valve, a key player in the agreement, to back out of adding VirtualLink integration to the Valve Index, and everything went downhill from there. Unfortunately, VirtualLink never became popular. It would be an interesting alternative – a single cable would be a great choice for VR users, and requiring a higher voltage over USB-C would also give us more than 5V with PD functionality. Ports – Neither laptops nor PCs offer these features these days. Yes, just a reminder – if you have a USB-C port on your desktop or laptop, it will certainly give you 5V, but you won’t get anything higher.
However, let’s look at the bright side. If you have one of these GPUs with a USB-C port, it will support both USB3 and DisplayPort!
The great thing about USB-C is that vendors or hackers can definitely define their own alternate mode if they want to, and while the adapter will be semi-proprietary, it’s essentially still a USB-C port for charging and data transfer. Want Ethernet Alternate Mode or Dual Port SATA? do it. Gone are the days of having to hunt down extremely obscure connectors for your devices as each dock and charge connector is different and can cost upwards of $10 each if rare enough to find.
Not every USB-C port needs to implement all of these features, and many do not. However, many people do, and as time goes by, we get more and more functionality from regular USB-C ports. This unification and standardization will pay off in the long run, and although there will be deviations from time to time, manufacturers will learn to deal with them smarter.
But one thing I’ve always wondered is why the rotation of the plug is not handled by placing the + and – wires on opposite sides. Thus, if the plug is connected in the “wrong” way, + will be connected to – and – will be connected to +. After decoding the signal at the receiver, all you have to do is reverse the bits to get the correct data.
Essentially, the problem is signal integrity and crosstalk. Imagine, say, an 8-pin connector, two rows of four, 1/2/3/4 on one side and 5/6/7/8 on the other, where 1 is opposite 5. Let’s say you want a pair of +/- receive /broadcast. You could try putting Tx+ on pin 1, Tx- on pin 8, Rx+ on pin 4, and Rx- on pin 5. Obviously, inserting back only swaps +/-.
But the electrical signal doesn’t actually travel across the signal pin, it travels between the signal and its return in the electric field. Tx-/Rx- should be the “return” of Tx+/Rx+ (and obviously vice versa). This means that the Tx and Rx signals actually intersect.
You “could” try to fix this by making the signals complementary unbalanced – essentially putting a very tight ground plane next to each signal. But in this case, you lose the differential pair’s common-mode noise immunity, which means that simple crosstalk from Tx+/Rx- opposite each other does not cancel out.
If you compare this to placing Tx+/Tx- on pins 1/2 and 7/8 and Rx+/Rx- on pins 3/4 and 5/6 via a multiplexer, now the Tx/Rx signals do not cross and all crosstalk caused on contacts Tx or Rx, will be somewhat common for both pairs and partially compensated.
(Obviously, a real connector will also have many ground pins, I just didn’t mention it for the sake of brevity.)
> Unification brings compatibility that’s hard to tell, IMO what USB-C brings is just a world of hidden incompatibilities that are hard to understand for the tech-savvy since the specs don’t even state what it can/can’t do. and it will only get worse as more alternate modes are added, and those same cables have issues too…
Most pre-USB-C power connectors were barrel connectors, which are much cheaper than USB-C. While most brands of docking stations can have weird connectors that are a nuisance, they also often have direct access to PCI-E and other buses, and usually have a significant amount of lanes – faster than USB-C, at least relatively your time. … USB-C wasn’t a nightmare for hackers who only wanted USB-2, just an expensive connector, and the dock connector wasn’t ideal, but when you really need complex. When it comes to high-speed capabilities, USB-C takes it to another level of performance.
Indeed, that was my impression as well. The standard allows everything, but no one will implement anything that would make it difficult for any two USB-C devices to work together. I’ve been through it; I’ve powered my tablet via a USB-A power adapter and a USB-A to USB-C cable for years. This allows me to carry an adapter for my tablet and phone. Bought a new laptop and the old adapter won’t charge it – after reading the previous post, I realized that it probably needs one of the higher voltages that the USB-A adapter can’t provide. But if you do not know the specifics of this very complex interface, then it is not at all clear why the old cable does not work.
Even one provider cannot do this. We got everything from Dell in the office. Dell laptop, Dell docking station (USB3), and Dell monitor.
No matter which dock I use, I get a “Display connection limit” error, “Charging limit” error, only one of the two screens works, or won’t connect to the dock at all. It’s a mess.
Firmware updates must be performed on the motherboard, docking station, and drivers must also be updated. It finally made the damn thing work. USB-C has always been a headache.
I use non-Dell docking stations and everything went smoothly! =D Making a decent USB-C dock doesn’t seem that hard – they usually work pretty well until you run into Thunderbolt oddities, and even then there are problems in the “plug, unplug, work” realm. I won’t lie, at this point I wanted to see a schematic of a motherboard for a Dell laptop with these docking stations.
Arya is right. All problems disappeared when I bought a cheap USB-C powered splitter from Amazon. Keyboards, webcams, USB dongles can be plugged in, the monitor plugs into the USB-C, HDMI, or DP port on the laptop, and it’s ready to go. I was told what to do by an IT guy who said the Dell dock was not worth the money.
No, these are just Dell idiots – apparently they decided to make the product incompatible with USB-C when using the same connector.
Yes, if you ask me, a device like a tablet needs to be more specific about “why is it not fully charged”. The pop-up message “At least 9V @ 3A USB-C charger required” will solve people’s problems like this and do exactly what the tablet manufacturer expects. However, we can’t even believe that any of them will release even one firmware update after the device goes on sale.
Not only cheaper, but also stronger. How many broken USB connectors have you seen on various devices? I often do this – and usually such a device is thrown away, because it is not economically feasible to repair it …
USB connectors, starting with micro USB, have been quite flimsy, and having to constantly plug and unplug them, usually by people who don’t align them properly, use too much force, wiggling them from side to side, makes the connectors terrible. For data, this might be tolerable, but given that USB-C is now also being used to power everything from smartwatches to entire laptops and all sorts of electronic gadgets that don’t use data at all, damaged connectors will become more and more common. The more it worries us – and for no good reason.
That’s right, I’ve only seen one broken barrel connector and it’s fairly easy to fix (aside from the Dell BS version, it only works on a proprietary charger that can communicate with it, which is pretty flimsy, you could damage it even if you never ride a bike..) Even for an experienced repairman, the USB-C connector will be PITA, with more PCB area, smaller solder pins…
Barrel connectors are typically rated for half a cycle (or less) of regular USB-C connectors. This is because the center pin flexes each time it is inserted, and with USB, the lever arm is shorter. I have seen many barrel jacks that have been damaged by use.
One of the reasons USB-C seems less reliable is cheap connectors or cables. If you find a product that looks “stylish” or “cooler” with injection molding or whatever, it’s probably crap. Only available from major cable manufacturers with specifications and drawings.
Another reason is that you’re using USB-C more than barrel-shaped connectors. Phones connect and disconnect every day, sometimes several times.
Post time: Jun-24-2023