Whenever a new communications technology comes along, engineers need test equipment before they start designing. USB4 is no exception. The USB Promoter Group announced on March 4 the forthcoming USB4 specification.

According to the USB Promoter Group, USB4 will support:

* Two-lane operation using existing USB Type-C cables and up to 40-Gbps operation over 40-Gbps–certified cables.

* Multiple data and display protocols to efficiently share the total available bandwidth over the bus.

* Backward compatibility with USB 3.2, USB 2.0, and Thunderbolt 3

* On March 5, Teledyne LeCroy became the first company to announce USB4 test equipment, the Voyager M4x protocol analyzer. It connects to a host PC through a USB3 port.

The merging of USB with Thunderbolt 3 into USB4 should eventually mean that users will have one I/O cable for everything. While consumers will have to wait until at least 2020 for the first USB4 products, engineers must first design and test those products.

In the days since the USB4 announcement, the consumer electronics press has been all abuzz, claiming that the 40-Gbps speed promised with a Type-C cable will become the standard. The current “standard” is USB 3.2 Gen 2×2, now officially known as SuperSpeed USB 20 Gbps, according to PC World.

Indeed, USB oscilloscope manufacturer Pico Technology this week released its latest product, a combination real-time and equivalent-time sampling oscilloscope that uses the classic USB 2.0 port with a Type-B connector. Just last week, I reviewed the Joulescope energy analyzer, currently a Kickstarter project, that uses the same connector. To find out why USB 2.0 remains popular, especially with test equipment, I asked both companies.

There are several reasons why USB 2.0 still thrives in applications where 12 Mbps is more than enough. Furthermore, USB 2.0 is mature and inexpensive to implement. “USB instruments can differ quite substantially,” said Mark Ashcroft, RF Business Development Manager at Pico Technology. “Some have a processor on board and, to a large extent, pass only control and results across to the controller device. Others may only have a PIC or FPGA on board and perhaps pass large amounts of data across to the PC for processing. The 9400 carries a fairly powerful ARM processor and Kintex 7 FPGA. It’s close to the former case. Other Pico products carry sizable FPGAs only and are closer to the latter case — in practice, only when passing large data captures (up to 2 Gsamples) for storage on the PC or when streaming trace data at up to 300 MSamples/s. We embrace the additional cost of USB3 (5.5-Gb/s transmit — receive, connectors, and cables) only when it benefits the customer.”

“USB 2.0 met the performance requirements for Joulescope, and I know how to implement it,” said Joulescope developer Matt Liberty. “USB3 is confusing to developers and end users alike, and it adds cost. Some of the Joulescope backers have requested USB-C connectors just because the full-sized USB Type-B connector is old and clunky. Other than wanting to be ‘cool,’ the right decision for the Joulescope was pretty clear.”

Ashcroft concurred with Liberty regarding the Type-B connector. “We also have to consider reliability, physical size, and the ongoing development of the USB3 interface. The standard USB3 connector is too large for some of our products, the smaller connectors are perhaps unreliable. Standardizing upon one USB3 interface is not as straightforward as we might want it to be.”

Personally, I’m still using mostly USB 2.0 links, with some devices (mostly external drives) running on USB 3.2 Gen 1, which was previously called USB 3.0, and then called USB 3.1 Gen 1. Confused about the designations? I can see why. Indeed, I have one USB-A to USB-C adapter cable, which was used to review a dock. The cable now sits in a box after I discovered that the dock continued to apply power to the external drives even when the PC was turned off. Other than that, I have no devices that use Type-C connectors.

By: DocMemory
Copyright © 2023 CST, Inc. All Rights Reserved