Connecting devices using the high-speed versions of USB requires some care with respect to cable lengths as well as the actual connected devices. In this paper we will point out the considerations one should be aware of when setting up a test configuration and the potential pitfalls that may be encountered.
A bit of history regarding the USB data connection will shed some light on the many different ‘flavors’ of USB and how they came about. This isn’t for the sake of history, but to help one understand how we got to the point of sending vast amounts of data over what any capable RF engineer would consider to be a lousy, and lossy, cable interconnect!
As computers became more ubiquitous, there were many devices such as printers, scanners, external hard drives and the like that needed a common interface to the computer. Manufacturers got together and the USB-IF (USB Implementers Forum) was formed to codify interface specifications to allow easy, universal connections between devices. The USB-IF developed specifications that the various hardware manufacturers could follow which would allow standard and repeatable interconnection between any and all USB compliant devices. While there were some early growing pains, in general, users found that it became easy to connect a printer, camera or storage device, without having to set hardware DIP switches, deal with computer Interrupt settings or other confusing and complex configurations. Initially, USB1 was capable of the blazing speed of 1.5 to 12 Mbps! Cable lengths of 10 meters or more were common and there was no concern over lengths until one was talking 30 meters or more. At these low data speeds, cable attenuation was very low, generally just a few dB per hundred feet.
Each successive iteration of USB increased data speeds, but there was little attempt to improve the actual cable medium that the data passed through. As an aside, consider cable television, you have seen the large trunk and feeder cables on the power poles, usually ½” in diameter using foam filled, low loss dielectric and robust aluminum conductors. These cables carry frequencies far lower than what is on USB-C cables! Sure, they carry signals a few hundred feet between amplifiers, but the fact remains, that if you carry higher frequencies, you need better cable with lower loss – or limit the length of the cable in order to reduce the cable loss.
Within the USB-C cable there are actually two independent USB connections. One using a single pair of data conductors which provide a standard USB2 data connection and another set of conductors to provide the USB-C data connection. While the USB2 data connection uses a single pair of conductors which can pass signals in just one direction at a time – think push-to-talk on a walkie talkie, or a single lane road that can carry just one car at a time and in just one direction, the USB-C data connection uses multiple conductors, like a multi-lane freeway. There are two data connection lines in each directions, exactly like a 4 lane freeway with two lanes in each direction, clearly capable of moving more cars along than a single one lane road. USB3 has only one data connection line in each direction, similar to a two lane highway, faster than USB2 with a single lane, but slower than USB-C with double the lanes. The discussion still stands, where cable loss is the determining factor.
The higher speed versions of USB are extremely dependent upon having a low loss connection between the connected devices. If the cabling has high loss, the data waveform will be attenuated and distorted, making the connection fail. The beauty of the USB protocol is that if the devices find the high-speed connection unusable it will throttle back to USB2 speeds using the USB2 connection within the USB-C cable. I mention this since one may see the devices connect, but only at the slower USB2 speed.
With all the above under our belt, let’s get to the heart of the issue and see if we can answer the question, “Well, just how long can my cables be while still passing the higher USB-C speeds?” Seeing that the latest buzz is AI, let’s ask that same question to Google Search Labs!
AI generated response from Google Search Labs:
The maximum length of a USB-C cable that can maintain high speeds depends on the USB version and the specific requirements of the application, but generally, 3 meters (9.8 feet) is a common limit for USB 3.2 Gen 1/Gen 2 and USB 3.0. For USB4, which supports 40 Gbps speeds, the recommended maximum length is 0.8 meters (2.6 feet). Exceeding these limits can result in decreased data transfer speeds and potential connectivity issues. [1, 2, 3, 4]
Here's a more detailed breakdown: [2, 3]
- USB 3.2 Gen 1 (5 Gbps): 3 meters (9.8 feet) is the recommended maximum length.
- USB 3.2 Gen 2 (10 Gbps): 3 meters (9.8 feet) is the recommended maximum length.
- USB 3.2 Gen 2x2 (20 Gbps): 3 meters (9.8 feet) is the recommended maximum length.
- USB4 (40 Gbps): 0.8 meters (2.6 feet) is the recommended maximum length.
Factors Affecting Cable Length Limits: [1, 1, 4, 4]
- Data Rate: Higher data rates require more stringent signal requirements, leading to shorter maximum lengths. [1, 1, 4, 4]
- Signal Loss: Longer cables can experience signal degradation, especially at higher frequencies, reducing data transfer speeds. [1, 1, 4, 4]
- Voltage Drop: Power delivery over longer distances can lead to voltage drop, which can impact charging speeds and stability. [5, 5, 6, 6]
- Cable Quality: Thicker conductors and better shielding can help mitigate signal loss and voltage drop, allowing for slightly longer cables while maintaining performance. [5, 5, 6, 6, 7]
Alternatives for Longer Distances: [8, 8]
- Active Extension Cables: These cables use a small amplifier to boost the signal, allowing for longer lengths while maintaining performance. [8, 8, 9, 10, 11]
- USB Hubs: Using a USB hub can effectively extend the cable length to the hub and then from the hub to the device. [12, 12, 13]
- Fiber Optic Cables: For very long distances and high data rates, fiber optic cables can be used. [14, 14]
Good job, Google Search Labs! Now, let’s see how this applies to the JRE Test USB-C and USB3 filtered interfaces.
The JRE USB-C and USB3 interface filters are completely passive with no ‘active’ internal devices within. One must be extremely cognizant of the data speeds that USB-C is capable of, with ‘revisions’ upping the speed from 5 Gbps (SS) to 10 Gbps (SS+) to 20 Gbps and now 40Gbps - Thunderbolt/USB4. In all cases, the cable attenuation is the primary determinate for speed. Note that all connectors, no matter the speed are the same whether USB-C, USB-4 or Thunderbolt. All of this is discussed in great detail on the USB Implementers web site, https://www.usb.org/ and USB-C in particular here: https://www.usb.org/usbc
The JRE USB-C and USB3 filters ‘appear’ or ‘look’ like a meter long cable and as such will be the determining factor on data speeds. For example, if you are using a 1 meter cable with your JRE USB-C filter, the total cable ‘length’ will work as though you are using 2 meters worth of cable.
The cable length is the limiting factor on Thunderbolt and USB-C and if you wish to pass the fastest data rates, you must limit cable length or use an active cable to mitigate cable losses.
The JRE USB-C and USB3 filters are provided with a short jumper cable and using it along with a similar length cable to device outside the test chamber will pass 5 Gbps speeds, for faster speeds, the use of an active cable is needed. The active cable is selected depending upon your particular devices and cable lengths. Bear in mind that the filters simply look like a meter long cable – this is all you need to know regarding speed limitations. There is no upper frequency limit since the filters use a patented filter topology that does not use low pass filtering, it cleverly discerns between the data signals and the interfering signals. JRE Test has two patents on this technology that has no speed limit or active circuitry that could be targeted by cyber attack.
In closing, we can safely say that cables matter, and to connect at the higher USB data speeds, short, good quality cables are of paramount importance. If a short cable connection cannot be made due to your test set-up, the use of an active cable or USB hub can be used.
[1] https://www.anker.com/blogs/cables/usb-cable-max-length
[3] https://www.cablecreation.com/blogs/news/how-long-can-a-usb-cable-be
[4] https://cabletimetech.com/blogs/knowledge/usb-cable-length-restrictions-how-to-bypass-and-extend
[6] https://www.anker.com/blogs/cables/are-all-usb-c-cables-the-same-for-charging
[7] https://www.securitycameraking.com/securitynews/how-far-can-you-run-a-security-camera-cable/
[8] https://www.quora.com/How-long-is-too-long-for-a-USB-cable
[9] https://www.avaccess.com/blogs/guides/usb-extender-a-useful-tech/
[10] https://www.cablewholesale.com/blog/index.php/2024/01/29/what-is-the-length-limit-for-usb-cables/
[11] https://www.av-connection.com/?PGr=17257
[13] https://www.anker.com/blogs/cables/usb-cable-max-length
