Hi Steve,
I found the following on the web, which seemed interesting to me:
"After the [USB] engineers made their decision about the maximum cable length, they used it directly in the specification to limit parameters which have physical properties. One of those properties is the cable delay. As you might know electricity travels almost by the speed of light through a cable, but it still takes some time to travel through a cable from one side to the other. This maximum delay time has been defined as 18 nsec for the low-speed USB 1.1 connections and 26 nsec for the high-speed USB 2.0 connections. The speed of electricity through a copper wire depends on the dielectric coefficient of the surrounding material. Although we don't know this parameter exactly for every USB cable out there, a speed of 65% of the speed of light is quite a realistic guess for a general cable. In that case we get 5.07 meter as maximum cable length for the high speed USB 2.0 connection and 3.51 meter for low speed USB 1.1.
This maximum time delay is fixed in the interface, so even with the highest quality cable made from the best materials which you buy in the most expensive computer shop you wouldn't get a significantly longer maximum USB cable length than with your cheap Chinese copy cable bought from an on-line store. This is because the speed of electricity through a copper cable is defined by simple physics laws which even the most expensive cables cannot change.
You might think that buying the cheapest cable would give the best price-to-quality ratio, but from personal experience I have to warn you. Although these cables may function with some devices, cheaper cables do not always function with all devices. This is because the USB specifications allow quite a large range in properties of driver chips, receiver chips and cables. If you have a cheap driver chip in your computer which is relatively slow and has low output power and want to connect it to peripheral which has also a poorly designed USB interface which operates just within the limits of the specification, an expensive cable might work, where over a cheap cable the communication will fail. Although the maximum distance between your devices will not vary dependent on the quality of the USB cable, the reliability of the connection certainly will."
I am not familiar with error recovery mechanisms within the USB protocol, but considering the above would it be possible that with longer cables and poorer usb ports:
a) data get corrupted
or
b) error recovery generates additional traffic to the point at which it affects the timely delivery of the information
Could this account for differences between USB cables?
On another note, are Locus Design cables still manufactured now that Lee Weiland is no longer with us?
Regards,
JPS
